Veterinary surgical oncology 9780470963210, 0470963212

Table of contents :
Veterinary Surgical Oncology
Contents
List of Contributors
Preface
1: Principles of surgical oncology
2: Multimodal therapy
3: Interventional oncology
4: Skin and subcutaneous tumors
5: Head and neck tumors
6: Oral tumors
7: Alimentary tract
8: Respiratory tract and thorax
9: Cardiovascular system
10: Reproductive system
11: Urinary tract
12: Eyelids, eye, and orbit
13: Endocrine system
14: Hemolymphatic system
15: Nervous system
16: Musculoskeletal system
Index

Citation preview

Veterinary Surgical Oncology

Veterinary Surgical Oncology Simon T. Kudnig, BVSc, MVS, MS, FACVSc, Diplomate ACVS Melbourne Veterinary Specialist Centre Glen Waverley, Victoria, Australia

Bernard Séguin, DVM, MS, Diplomate ACVS Associate Professor College of Veterinary Medicine Oregon State University Corvallis, Oregon USA

Illustrations by Dave Carlson

A John Wiley & Sons, Inc., Publication

This edition first published 2012 © 2012 by John Wiley & Sons, Ltd. Wiley-Blackwell is an imprint of John Wiley & Sons, formed by the merger of Wiley’s global Scientific, Technical and Medical business with Blackwell Publishing. Registered office: John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial offices: 2121 State Avenue, Ames, Iowa 50014-8300, USA The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK 9600 Garsington Road, Oxford, OX4 2DQ, UK For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our Website at www.wiley.com/ wiley-blackwell. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Blackwell Publishing, provided that the base fee is paid directly to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license by CCC, a separate system of payments has been arranged. The fee codes for users of the Transactional Reporting Service are ISBN-13: 978-0-8138-0542-9/2012. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data Veterinary surgical oncology / [edited by] Simon T. Kudnig, Bernard Séguin ; illustrations by Dave Carlson.    p. ; cm.   Includes bibliographical references and index.   ISBN-13: 978-0-8138-0542-9 (hardcover : alk. paper)   ISBN-10: 0-8138-0542-2 (hardcover : alk. paper)   I.  Kudnig, Simon T.  II.  Séguin, Bernard, 1968–   [DNLM:  1.  Neoplasms–surgery.  2.  Neoplasms–veterinary.  3.  Surgery, Veterinary–methods.  SF 910.T8]   LC classification not assigned   636.089'7–dc23 2011032156 A catalogue record for this book is available from the British Library. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Set in 10.5 on 12.5 pt Minion by Toppan Best-set Premedia Limited

Disclaimer The publisher and the author make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation warranties of fitness for a particular purpose. No warranty may be created or extended by sales or promotional materials. The advice and strategies contained herein may not be suitable for every situation. This work is sold with the understanding that the publisher is not engaged in rendering legal, accounting, or other professional services. If professional assistance is required, the services of a competent professional person should be sought. Neither the publisher nor the author shall be liable for damages arising herefrom. The fact that an organization or Website is referred to in this work as a citation and/or a potential source of further information does not mean that the author or the publisher endorses the information the organization or Website may provide or recommendations it may make. Further, readers should be aware that Internet Websites listed in this work may have changed or disappeared between when this work was written and when it is read.

1  2012

To my wife Narelle, my parents Philip and Judy, my brother Martin, and my sister Mandy for their unending support and understanding, and to my daughter Samantha for the joy and perspective on life you have brought. S.K. To my wife Lisa, my parents Gisèle and René, and my brother Jean-François for their encouragement, support, and love and to my children Alexandre and Gabrielle, for being my inspiration and teaching me so much more. B.S. To Dr. Stephen J. Withrow for teaching us, among many other things: “Success is the ability to move forward in the face of failure.” S.K. and B.S.

Contents

List of Contributors Preface

ix xiii

  1

Principles of Surgical Oncology Nicole Ehrhart, William T.N. Culp

  2

Multimodal Therapy Tania A. Banks

15

  3

Interventional Oncology William T.N. Culp

35

  4

Skin and Subcutaneous Tumors Stewart Ryan, Erik G.H. Wouters, Sebastiaan van Nimwegen, Jolle Kirpensteijn

55

  5

Head and Neck Tumors Sara A. Ayres, Julius M. Liptak

87

  6

Oral Tumors 119 Julius M. Liptak, B. Duncan X. Lascelles

  7

Alimentary Tract William T.N. Culp, Ryan P. Cavanaugh, Earl F. Calfee III, Paolo Buracco, Tania A. Banks

  8

3

179

  9

Cardiovascular System Simon T. Kudnig, Eric Monnet

329

10

Reproductive System Maurine J. Thomson, Tara A. Britt

341

11

Urinary Tract Nicholas J. Bacon, James P. Farese

365

12

Eyelids, Eye, and Orbit B. Duncan X. Lascelles, Michael Davidson

383

13

Endocrine System Bernard Séguin, Lisa Brownlee, Peter J. Walsh

405

14

Hemolymphatic System Christine C. Warzee

443

15

Nervous System Elaine R. Caplan

465

16

Musculoskeletal System Julius M. Liptak, William S. Dernell, James P. Farese, Deanna R. Worley

491

Index

569

Respiratory Tract and Thorax 273 Marina Martano, Sarah Boston, Emanuela Morello, Stephen J. Withrow

vii

List of Contributors

Sara A. Ayres, DVM, DVSc, Diplomate ACVS Surgeon, Simcoe Veterinary Surgical Referral Ontario, Canada Head and Neck Tumors Nicholas J. Bacon, MA, VetMB, CertVR, CertSAS, Diplomate ECVS, Diplomate ACVS Clinical Assistant Professor, Surgical Oncology University of Florida College of Veterinary Medicine Gainesville, Florida, USA Urinary Tract Tania A. Banks, BVSc, FACVSc Lecturer, Small Animal Surgery The School of Veterinary Science, The University of Queensland Gatton Campus, QLD, Australia Multimodal Therapy Alimentary Tract: Pancreas Sarah Boston, DVM, DVSc, Diplomate ACVS Assistant Professor, Small Animal Surgery (Surgical Oncology) Department of Clinical Studies Ontario Veterinary College, University of Guelph, Guelph, ON, Canada Respiratory Tract and Thorax: Chest Wall Tumors, Laryngeal Tumors, Tracheal Tumors Tara A. Britt, VMD, Diplomate ACVS Veterinary Surgical Services Veterinary Referral Center of Colorado Englewood, CO, USA Reproductive System: The Male Lisa Brownlee, DVM, MS, Diplomate ACVIM (Internal Medicine) Assistant Professor, Department of Clinical Sciences College of Veterinary Medicine, Oregon State University Corvallis, OR, USA Endocrine System

Paolo Buracco, DVM, Diplomate ECVS Professor of Veterinary Surgery School of Veterinary Medicine Grugliasco, Turin, Italy Alimentary Tract: Colorectal Tumors and Perianal Tumors Earl F. Calfee, III, DVM, MS, Diplomate ACVS Nashville Veterinary Specialists Nashville, TN, USA Alimentary Tract: Stomach, Liver and Gall Bladder, Pancreas, Small Intestine Elaine R. Caplan, DVM, Diplomate ACVS, Diplomate ABVP Texas Veterinary Oncology, Capital Area Veterinary Specialists, Inc. Austin, TX, USA Nervous System Ryan P. Cavanaugh, DVM, Diplomate ACVS Staff Surgeon, VCA Alameda East Veterinary Hospital Denver, CO, USA Alimentary Tract: Stomach, Liver and Gall Bladder, Pancreas, Small Intestine William T.N. Culp, VMD, Diplomate ACVS Assistant Professor, Small Animal Surgery (Surgical Oncology/Interventional Radiology), Department of Surgical and Radiological Sciences School of Veterinary Medicine, University of California, Davis Davis, CA, USA Principles of Surgical Oncology Interventional Oncology Alimentary Tract: Esophagus Michael Davidson, DVM, Diplomate ACVO Professor, Ophthalmology; Associate Dean and Director of Veterinary Medical Services College of Veterinary Medicine, North Carolina State University Raleigh, NC, USA Eyelids, Eye, and Orbit ix

x  List of Contributors

William S. Dernell, DVM, MS, Diplomate ACVS Professor and Chair, Department of Veterinary Clinical Sciences College of Veterinary Medicine, Washington State University Pullman, WA, USA Musculoskeletal System Nicole Ehrhart, VMD, MS, Diplomate ACVS Professor, Surgical Oncology, Department of Clinical Sciences College of Veterinary Medicine and Biomedical Sciences, Colorado State University Fort Collins, CO, USA Principles of Surgical Oncology James P. Farese, DVM, Diplomate ACVS Associate Professor, Surgical Oncology, Department of Small Animal Clinical Sciences College of Veterinary Medicine, University of Florida Gainesville, FL, USA Urinary Tract Musculoskeletal System Jolle Kirpensteijn, DVM, PhD, Diplomate ACVS, Diplomate ECVS Professor, Surgery, Department of Clinical Sciences of Companion Animals Faculty of Veterinary Medicine, Utrecht University Utrecht, The Netherlands Skin and Subcutaneous Tumors: Skin Tumors General Principles, Soft Tissue Sarcomas Simon T. Kudnig BVSc, MVS, MS, FACVSc, Diplomate ACVS Staff Surgeon, Melbourne Veterinary Specialist Centre Melbourne, Victoria, Australia Cardiovascular System B. Duncan X. Lascelles, BSc, BVSc, PhD, CertVA, DSAS(ST), Diplomate ECVS, Diplomate ACVS Professor of Surgery and Pain Management, Surgery Section and Comparative Pain Research Laboratory North Carolina State University College of Veterinary Medicine Raleigh, NC, USA Oral Tumors Eyelids, Eye, and Orbit

Julius M. Liptak, BVSc, MVetClinStud, FACVSc, Diplomate ACVS, Diplomate ECVS Specialist Small Animal Surgeon, Alta Vista Animal Hospital Ottawa, Ontario, Canada Head and Neck Tumors Oral Tumors Musculoskeletal System Marina Martano, DMV, PhD Assistant Professor of Veterinary Surgery School of Veterinary Medicine, University of Turin Grugliasco (TO), Italy Respiratory Tract and Thorax: Thoracotomy, Rhinotomy Eric Monnet, DVM, PhD, Diplomate ACVS, Diplomate ECVS Professor, Small Animal Surgery, Department of Clinical Sciences Colorado State University College of Veterinary Medicine and Biomedical Sciences Fort Collins, CO, USA Cardiovascular System Emanuela Morello, DMV, PhD Assistant Professor of Veterinary Surgery School of Veterinary Medicine, University of Turin Grugliasco (TO), Italy Respiratory Tract and Thorax: Lung. Stewart Ryan, BVSc, MS, Diplomate ACVS Assistant Professor, Musculoskeletal and Surgical Oncology, Department of Clinical Sciences College of Veterinary Medicine and Biomedical Sciences Colorado State University Fort Collins, CO, USA Skin and Subcutaneous Tumors: Skin Tumors General Principles, Mast Cell Tumors Bernard Séguin, DVM, MS, Diplomate ACVS Associate Professor, Department of Clinical Sciences College of Veterinary Medicine, Oregon State University Corvallis, OR, USA Endocrine System Maurine J. Thomson, BVSc, FACVSc Specialist, Surgical Oncologist Veterinary Specialist Services, Springwood Centre Underwood, Qld, Australia Reproductive System: The Female

List of Contributors  xi

Sebastiaan van Nimwegen, DVM, PhD Department of Clinical Sciences of Companion Animals Faculty of Veterinary Medicine, Utrecht University Utrecht, The Netherlands. Skin and Subcutaneous Tumors: Skin Tumors General Principles, Soft Tissue Sarcomas

Deanna R. Worley, DVM, Diplomate ACVS Assistant Professor, Surgical Oncology, Department of Clinical Sciences College of Veterinary Medicine and Biomedical Sciences, Colorado State University Fort Collins, CO, USA Musculoskeletal System

Peter J. Walsh, DVM, MVetSc, Diplomate ACVS Veterinary Specialty Group West Sacramento, CA, USA Endocrine System

Erik G.H. Wouters, DVM Department of Clinical Sciences of Companion Animals Faculty of Veterinary Medicine, Utrecht University Utrecht, The Netherlands Skin and Subcutaneous Tumors: Skin Tumors General Principles, Soft Tissue Sarcomas

Christine C. Warzee, DVM, Diplomate ACVS Assistant Professor, Surgical Oncology Center for Comparative Oncology, College of Veterinary Medicine Michigan State University East Lansing, MI, USA Hemolymphatic System Stephen J. Withrow, DVM, Diplomate ACVS, Dipomate ACVIM (Oncology) University Distinguished Professor, Surgical Oncology, Department of Clinical Sciences College of Veterinary Medicine and Biomedical Sciences, Colorado State University Fort Collins, CO, USA Respiratory Tract and Thorax: Metastasectomy for Sarcomas

Preface

This book is the result of the collaboration between many contributors who belong to the Veterinary Society of Surgical Oncology (VSSO). At its inception, the impetus to write this book was to help fulfill the goals of the VSSO, which include “to disseminate knowledge to help provide the highest possible standard of surgical treatment for cancer and to encourage and promote education in surgical oncology for professional veterinary students, graduate students and house officers, and graduated veterinarians and veterinary surgeons” (www. vsso.org/aims.html). The field of surgical oncology has greatly expanded in recent years. The creation of the VSSO reflects this growth. The idea of the VSSO was the brainchild of Dr. Steve Withrow. Dr. Withrow is, for many of us, the pioneer of surgical oncology in veterinary medicine, and he instituted the first fellowship in veterinary surgical oncology in 1988. Many of the original members of the VSSO are graduates of the fellowship. Under the leadership of Dr. Julius Liptak, the VSSO was officially created in 2006. In the first year of the VSSO, there were less than 30 members, whereas at the time of publication there are more than 235. Members are from North America, Europe and the United Kingdom, Australia and New Zealand, and Asia. The American College of Veterinary Surgeons (ACVS) has announced that it will recognize further training and expertise in certain fields of surgery, one of which is oncologic surgery. This is affirmation of the expanding body of knowledge in surgery in general as well as that focusing on a certain field is necessary to remain the most proficient. The recognition of advanced training in a field will best promote continued development of novel ideas that will increase our understanding of the diseases and their treatment. We hope this textbook will serve as a repository of knowledge for anyone with an interest in surgical oncology to use and to build upon in the future.

The emphasis of this book is on the surgical aspect of treating small animals afflicted by cancer. This book is not meant to be a full review of small animal oncology as there are several excellent existing textbooks doing so. For instance, this book was not meant to be a comprehensive review of how to diagnose the diseases. Rather, we wanted to concentrate on the surgical procedures, such as those that are not well covered in the literature. Our goal is to assist decision making and to cover controversies in the field. The reader is expected to have a basic knowledge of general surgical principles and surgical techniques. We are indebted to all the contributors for their remarkable contributions. The excellence of the chapters is to their credit and not ours, but any errors are our responsibility. We want to thank Erin Gardner, Erica Judisch, Nancy Turner, Erin Magnani and Susan Engelken from Wiley-Blackwell for their assistance and patience during this whole process, which was for the most part new to both of us. We also want to thank Jane Loftus for copy-editing the chapters; and Dave Johnson and Lorie Kennerly from the Information Technology Services at the College of Veterinary Medicine, Oregon State University, for their assistance when needed. We also thank Jill Bartlett from Oregon State University and Jean-François Séguin for their technical assistance with some of the figures. We need to thank our colleagues, house officers, students, and staff for their support and the motivation they supplied. And most importantly, we thank our families who by extension and default have lived through the creation of this book. Without their support and understanding, this would not have been possible. We hope you find this book helpful in your practice and education and welcome any comments you may have. Simon T. Kudnig and Bernard Séguin

xiii

Veterinary Surgical Oncology

1 Principles of surgical oncology Nicole Ehrhart, William T.N. Culp

Cancer treatment is a rapidly changing and evolving area involving the use of multiple diagnostic and therapeutic modalities to achieve the most optimal outcome. Surgical intervention remains a pivotal aspect of the treatment of cancer. Surgery cures more cancer than any other single modality. Nonetheless, the optimal treatment pathway for any given animal patient with cancer most often involves several adjuvant treatment modalities. Adjuvant treatments significantly affect the success of surgery, and likewise, surgery affects the outcome of adjuvant treatments. It is widely recognized in human cancer centers that patient outcome is greatly improved when surgery is performed by a surgeon with specialized training in oncologic procedures. These surgeons have expertise in selecting surgical treatment options in combination with other forms of cancer treatment, as well as knowledge of the benefits and risks associated with a multidisciplinary approach beyond that which can be mastered within a 3-year surgery residency training program. This level of expertise requires an understanding of the fundamental biology of cancer, clinical pharmacology, tumor immunology and endocrinology, as well as a thorough understanding of potential complications of multimodality therapy. Veterinary training programs in surgical oncology have been in existence for the last 14 years. With the development of new treatments such as small molecule inhibitors, gene therapy, and new forms of radiation, the role of the surgical oncologist is constantly evolving and changing (O’Reilly et al. 1997; Drixler et al. 2000). Therapeutic goals (e.g., curative intent, cytoreduction, or palliation) for each case should be established with the pet owners before surgery is initiated. The efficacy of surgical therapy in any patient with cancer is heavily dependent upon the surgeon’s global under-

standing of the patient’s general health status, lifestyle, and activity level; type and stage of cancer; adjuvant therapies available; alternatives to surgery; and expected prognosis. To maximize effectiveness, the optimal treatment pathway for each case should be strategically assessed before initiating treatment. This planning should always include a frank and thorough discussion with the owner regarding preoperative diagnostic tests, stage of cancer, palliative options, surgical options, adjuvant treatments likely to be needed, costs, postoperative care, and expected function, cosmesis and prognosis including risks of complications. The goal of this discussion is to provide owners with enough information to help them make an informed choice regarding the best treatment plan for their companion. Highly individualized initial planning will allow for the best overall outcome for each patient.

Preoperative Considerations Signalment The patient’s age, gender, breed, and weight are important factors in the determination of appropriate recommendations. Advanced age is not necessarily a negative prognostic factor. Comorbidities common to geriatric veterinary patients such as renal insufficiency, hepatic disease, or osteoarthritis may limit or change specific treatment recommendations; however, the age of the patient alone should not. Certain neoplastic diseases are common in a particular gender or breed. The surgical oncologist should always bear in mind the role that gender and breed play in the diagnosis of neoplasia. As an example, the differential list for a flat-coated retriever with a femoral bony lesion noted on radiographs that has been referred

Veterinary Surgical Oncology, First Edition. Edited by Simon T. Kudnig, Bernard Séguin. © 2012 by John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.

3

4  Veterinary Surgical Oncology

for a suspected diagnosis of osteosarcoma should be expanded to include histiocytic sarcoma; other diagnostics such as an abdominal ultrasound would be recommended to look for other foci of histiocytic disease. Other portions of the signalment are also important to note, including the patient’s weight and body condition. Patients that are morbidly obese or those in poor body condition may not be able to function effectively or may be more severely debilitated by a major surgery. For example, a patient with cancer cachexia can have such profound alterations of their carbohydrate, protein, and fat metabolism that recovery may be compromised (Ogilvie 1998). Staging and concomitant disease Staging diagnostics such as a complete blood count, chemistry profile, urinalysis, thoracic radiographs, and abdominal ultrasound are essential components for the preoperative assessment of veterinary oncology patients. While there is debate about the timing of some of these diagnostics (i.e., before or after biopsy), for many patients thorough preoperative staging diagnostics can unmask an underlying condition that may alter the plan or better assist the surgeon to provide a more accurate prognosis. Alternative surgical dose may also be recommended based on the results of staging. Neoadjuvant therapy The surgical oncologist is often presented with extremely large tumors or tumors located in difficult anatomical locations. It is important to consider neoadjuvant treatments, if available and warranted, such as chemotherapy and radiotherapy before proceeding with surgery. In some cases, these treatments may decrease the overall surgical dose needed to achieve local control. Most commonly, recommendations about chemotherapy and radiation therapy are made after the grade of the tumor and the surgical margins have been determined. In tumors that are suspected to be sensitive to chemotherapy based on published literature or previous experience, a postoperative protocol can be discussed prior to surgery. Neoadjuvant chemotherapy is rarely pursued in veterinary medicine. However, for certain tumor types, this may prove to be a beneficial adjunct to surgery. In human cases of osteosarcoma, neoadjuvant chemotherapy is commonly used prior to surgery and local tumor response (as measured by percentage of tumor necrosis) has been shown to be associated with increased survival. A recent veterinary study showed that neoadjuvant chemotherapy with prednisone administered to a group of dogs with intermediate grade mast cell tumors resulted in tumor size reduction; surgical excision of very large

mast cell tumors or tumors that were in an anatomical site that precluded wide (3 cm lateral and one facial plane deep) excision was more successful (Stanclift and Gilson 2008). Microscopically complete margins were achieved in many of the pretreated cases. These patients would not likely have had complete surgical margins otherwise (Stanclift and Gilson 2008). Long-term follow-up was not the focus of this study, however, and controversy exists as to the risk of local recurrence in patients where neoadjuvant chemotherapy is used to shrink gross tumor volume to allow a less aggressive surgical margin. Further study is needed to assess the benefit of neoadjuvant chemotherapy in veterinary cancer patients. Neoadjuvant radiation therapy has also been advocated as a method of treating neoplastic disease to reduce the need for radical surgery (McEntee 2006). Advantages to neoadjuvant radiation therapy include a smaller radiation field, intact tissue planes, better tissue oxygenation, and a reduction in the number of viable neoplastic cells that may be left within a postoperative seroma or hematoma following microscopically incomplete margins. Complications such as poor wound healing may occur more commonly in irradiated surgical sites than in nonirradiated tissue due to the effects of radiation on fibroblasts and blood vessels (Séguin et al. 2005). Even so, surgery in previously irradiated fields can be quite successful provided care is taken to ensure minimum tension, careful surgical technique, and appropriate timing (either before or after acute effects have occurred). Consultation with a radiation oncologist prior to surgery can help the surgeon identify those patients who may be good candidates. Considerations such as whether or not preoperative radiation will diminish the surgical dose and what type of reconstruction will be needed to ensure a tension-free closure in an irradiated surgical field should be discussed at length prior to deciding if neoadjuvant radiation is warranted.

Surgical Planning Fine-needle aspirate Fine-needle aspiration is often the most minimally invasive technique for obtaining critical information about a newly identified mass prior to surgery. The accuracy of a fine-needle aspirate depends on many factors, including the tumor type, location, and amount of inflammation. Overall sensitivity and specificity of cytology has been reported to be 89% and 100%, respectively (Eich et al. 2000; Cohen et al. 2003). Imaging tools such as ultrasound and fluoroscopy can increase the chance of obtaining a diagnostic sample.

Principles of Surgical Oncology  5

In most patients, a fine-needle aspirate of cutaneous or subcutaneous lesions can be obtained with no sedation and a minimal amount of discomfort. Fine-needle aspiration has been compared to histopathological samples in several studies. In a recent study of the correlation between cytology generated from fine-needle aspiration and histopathology in cutaneous and subcutaneous masses, the diagnosis was in agreement in close to 91% of cases (Ghisleni et al. 2006). Cytology was 89% sensitive and 98% specific for diagnosing neoplasia (Ghisleni et al. 2006). The goal of fine-needle aspiration is to differentiate between an inflammatory or neoplastic process, and if neoplastic, whether the tumor is benign or malignant. In some cases, the specific tumor type can be determined (e.g., mast cell tumor). In other cases, the class of tumor may be identified (e.g., sarcoma), but the specific diagnosis requires histopathology (e.g., chondrosarcoma versus osteosarcoma). The overall purpose of performing the fine-needle aspiration is to guide the staging diagnostics (where to look for metastasis or paraneoplastic diseases) and surgical dose. For example, a fine-needle aspirate of a mass showing normal adipocytes would indicate that the mass is not inflammatory; rather, it is a neoplastic process and is benign (lipoma). Based on the knowledge of the biological behavior of this tumor we would perform no other staging tests and prescribe a minimal surgical dose (marginal resection). Alternatively, if the fine-needle aspirate of a mass indicated carcinoma cells, we would be prompted to perform more advanced staging (three-view thoracic radiographs, abdominal ultrasound, lymph node aspirates) and would prescribe a larger surgical dose. Fine-needle aspiration of internal organs can also be performed and may be helpful in guiding diagnostic and treatment choices. Image guidance should be used when obtaining tissue from fine-needle aspirations of masses within a body cavity. Aspirates of lung and other thoracic organs can be performed safely in most cases. In one study, fine-needle aspiration of lung masses had a sensitivity of 77% and a specificity of 100% (DeBerry et al. 2002). The aspiration of cranial mediastinal masses is beneficial, as thymomas can be diagnosed by cytology (Rae et al. 1989; Atwater et al. 1994; Lana et al. 2006). Cytological diagnosis of thymoma requires the presence of a population of unequivocal malignant epithelial cells. The presence of mast cells is also common in thymoma and often supports the diagnosis (Atwater et al. 1994). Flow cytometry is another diagnostic tool that will differentiate thymoma from lymphoma using a fine-needle aspirate sample. Thymomas will contain both CD4+ and CD8+ lymphoctyes, whereas lymphoma would typically contain a clonal expansion of one lymphocyte type (Lana et al. 2006).

Fine-needle aspiration of hepatic and splenic neoplasia has been described in several studies (Osborne et al. 1974; Hanson et al. 2001; Roth 2001; Wang et al. 2004). Successful diagnosis of hepatic neoplasia with fineneedle aspiration is variable. A study has reported diagnostic rates for liver cytology of multiple pathologies (including neoplasia) as high as 80% (Roth 2001); however, another study demonstrated less success with diagnostic rates of 14% in dogs and 33% in cats for fineneedle aspiration of hepatic neoplasia (Wang et al. 2004). In cases of suspected splenic hemangiosarcoma, fine-needle aspiration is generally not recommended, as an accurate diagnosis is unlikely due to the abundance of blood-filled cavities. Additionally, complications may include severe bleeding from the aspiration site. Fineneedle aspiration of splenic neoplasia such as lymphoma and mast cell tumors is often diagnostic (Hanson et al. 2001). Other tumors in which fine-needle aspiration has been used to obtain diagnostic information include gastrointestinal tumors and bony tumors. The accuracy of fine-needle aspiration in the diagnosis of gastrointestinal neoplasia is often dependent on the type of neoplasia present. For instance, fine-needle aspiration of gastrointestinal lymphoma tends to have a higher sensitivity than aspiration of gastrointestinal carcinoma/ adenocarcinoma or leiomyoma/leiomyosarcoma (Bonfanti et al. 2006). The specificity of the diagnosis is similar among these neoplastic diseases with fine-needle aspiration (Bonfanti et al. 2006). In a recent report, ultrasound-guided fine-needle aspiration of osteosarcoma lesions was found to have a sensitivity of 97% and specificity of 100% for the diagnosis of a sarcoma (Britt et al. 2007). Another study found that cytology after fine-needle aspiration agreed with incisional and excisional biopsies of bony lesions in 71% of cases (Berzina et al. 2008). As with any procedure, fine-needle aspirates are not without risk. In certain cases, bleeding or fluid leakage can be problematic, especially within a closed body cavity where it cannot be easily controlled. Tumor seeding and implantation along the needle tract is a rare occurrence but in certain tumors has been reported more frequently. Localized tumor implantation following ultrasound-guided fine-needle aspiration of transitional cell carcinoma of the bladder has been reported (Nyland et al. 2002) and should be a consideration when deciding on methods for diagnosing bladder masses. Fine-needle aspiration of mast cell tumors can cause massive degranulation, and clinicians should be prepared to treat untoward systemic effects following aspiration of a suspicious or known mast cell tumor. Despite the risks associated with fine-needle aspiration,

6  Veterinary Surgical Oncology

it remains an effective, inexpensive, and valuable tool in the preoperative planning process. Biopsy Clinicians often use the term biopsy as a nonspecific description of obtaining a tissue sample for histopathological interpretation. For this reason, we will designate biopsy procedures into two major categories: pretreatment biopsy (tissue obtained before treatment initiation) or posttreatment biopsy (tissue obtained at the time of definitive tumor resection). We will also give examples of specific biopsy techniques. All biopsy procedures, whether pretreatment or posttreatment, should be carefully planned with several factors in mind. These factors include known patient comorbidities, anatomical location of the mass, differential diagnoses, biopsy technique, eventual definitive treatment, and any neoadjuvant or adjuvant therapies that may need to be incorporated. Pretreatment biopsy Needle core biopsy This technique is commonly used for soft tissue, visceral, and thoracic masses (Osborne et al. 1974; Atwater et al. 1994; deRycke et al. 1999). Image guidance is recommended when using this technique in closed body cavities. Most patients require sedation and local anesthesia but do not need general anesthesia. Instrumentation includes a needle core biopsy instrument (automated or manual), no. 11 scalpel blade, local anesthetic, and a 22-gauge hypodermic needle. To perform the procedure, the area surrounding the mass is clipped free of fur and prepared with aseptic technique. If intact skin is to be penetrated and the animal is not anesthetized, the skin overlying the area to be penetrated is anesthetized with lidocaine or bupivicaine. A 1–2 mm stab incision is made over the mass to allow for placement of the needle core biopsy instrument. The instrument is oriented properly and fired, and the instrument is withdrawn. The 22-gauge needle can be used to gently remove the biopsy from the trough of the needle core instrument. This identical procedure is performed for masses within a body cavity; however, it is necessary to use image guidance for proper placement of the instrument within the desired tissue. Imaging can be used to determine the depth of penetration and to safely avoid nearby vital structures. Punch biopsy The punch biopsy technique is most effective for cutaneous lesions as well as intraoperatively for biopsies of masses within organs such as the liver, spleen, and

kidney. Subcutaneous lesions can be biopsied using this method, but it is best to incise the skin overlying the mass and then obtain the sample using the biopsy instrument. Instrumentation includes a punch biopsy instrument, no. 11 scalpel blade, local anesthetic, Metzenbaum scissors, forceps, and suture. The area containing the mass is clipped free of fur and prepared with aseptic technique. If intact skin will be penetrated and the animal is not anesthetized, the skin overlying the lesion is anesthetized with lidocaine or bupivicaine. For cutaneous masses, an incision is not necessary. For subcutaneous masses, make an incision in the skin over the mass to allow a better sample to be procured. The skin incision should be large enough for the punch biopsy instrument to be placed and allow it to be twisted without engaging skin. Twist the punch biopsy instrument until the device is embedded into the mass to the hub. The punch biopsy instrument is then withdrawn from the mass to expose the tissue sample. Gently grasp the sample with forceps, use Metzenbaum scissors to sever the deep aspect of the sample from the rest of the tissue, and remove the sample. A single suture is generally sufficient to close the incision. The same procedure can be performed on visceral organs. Incisional (wedge) biopsy The incisional biopsy technique is effective for masses in all locations and generates a larger sample for histopathological evaluation as compared to the needle core biopsy. The location of the incision should be carefully planned, as the biopsy incision will need to be removed during the definitive treatment. Care should be taken to avoid dissection and prevent hematoma or seroma formation as these may potentially seed tumor cells into the adjacent subcutaneous space. Although the junction of normal and abnormal tissue is often mentioned as the ideal place to obtain a biopsy sample, one should take care to avoid entering uninvolved tissues. Obtaining a representative sample of the mass is the most important principle to consider. It is also important to obtain a sample that is deep enough and that contains the actual tumor, rather than just the fibrous capsule surrounding the mass. Incisional biopsy has a higher potential for complications such as bleeding, swelling, and infection due to the increase in incision size and dissection. Instrumentation includes a no. 11 or no. 15 scalpel blade, local anesthetic, Metzenbaum scissors, forceps, suture, and hemostats. A gelpi retractor or similar selfretaining retractor aids in visualization if the mass is covered by skin. If the skin is intact and moveable over the mass, a single incision is made in the skin. Once the tissue layer containing the tumor is exposed, two parallel

Principles of Surgical Oncology  7

incisions are started superficially and then meet at a deep location to form a wedge. The wedge is then grasped with forceps and removed. If the deep margin of the wedge is still attached, the Metzenbaum scissors can be used to sever the biopsy sample free of the parent tumor. The wedge site is then closed with suture. Posttreatment biopsy (excisional biopsy) The approach to an excisional biopsy varies based on location, goal of surgery, and predetermined adjuvant therapy. An excisional biopsy has the advantage of being both a diagnostic technique as well as a treatment modality. A great deal of caution should be exercised in cases where the diagnosis is unclear. At a minimum, a fine-needle aspirate should be obtained to discern if a given mass is inflammatory or neoplastic, and if neoplastic, whether benign or malignant. This information is imperative in order to determine surgical dose. There are cases where an excisional biopsy may be a reasonable option if doubt remains after fine-needle aspiration, depending on the size and location of the tumor. In these instances, the surgeon must contemplate if an excisional biopsy will compromise the ability to enact a cure by using a wide excision. If it is deemed that an excisional biopsy can be performed while leaving this option, an excisional biopsy may be considered. Once an excision is performed, the local anatomy is forever altered; both deep and wide tissue planes to the tumor are invaded, providing the opportunity for tumor cells to extend and seed deeper and wider into tissues. For this reason, the best chance for complete excision is at the time of the first surgical excision. In order to perform a curative surgery, the surgeon must take the appropriate margin of tissue for the tumor type. In some cases (e.g., lipoma), this margin is minimal or even intralesional. In other cases (e.g., soft tissue sarcoma), the margin should be much more extensive. Unless the tumor type is known at the time of excision, the surgeon may compromise the patient by doing too little or too much surgery. Specific biopsy techniques Bone biopsy The clinician performing the bone biopsy procedure should consider the eventual definitive treatment that is likely to be pursued for each case. The biopsy tract or incision needs to be in a location that can be removed during the definitive treatment. A reactive zone of bone exists in the periphery of most bone tumors, and samples taken from this region are more likely to result in an incorrect diagnosis (Wykes et al. 1985; Liptak et al. 2004). The surgeon should target the anatomical center

of the bony lesion. Two radiographic views of the involved bone should be available during the procedure as this will aid in optimal sampling. The majority of bone biopsies are performed using either a Michele trephine or a Jamshidi needle (Wykes et al. 1985; Powers et al. 1988; Liptak et al. 2004). A trephine instrument provides a large sample and has been associated with 93.8% diagnostic accuracy (Wykes et al. 1985). The disadvantages of the trephine technique include increased likelihood of fracture as compared to other techniques, requirement of a surgical approach, and a more lengthy decalcification time prior to sectioning (Wykes et al. 1985; Ehrhart 1998). Michele trephines are available in variable diameters. As a small surgical approach is required, a simple surgical pack is needed for the procedure. The biopsy site is clipped free of fur, and the patient is prepared with aseptic technique and draped. A 1–3 cm incision is made over the bony lesion, and the soft tissues are dissected from the surface of the tumor. The trephine is then seated into the tumor using a twisting motion. The trephine is advanced through the cis cortex. An effort should be made to not penetrate both the cis and trans cortex as fracture of the bone is more likely (Liptak et al. 2004). Once the trephine is within the medullary cavity, the trephine is rocked backed and forth to loosen the sample and then removed. A stylet is introduced into the trephine to push the sample out of the trephine onto a gauze square. The Jamshidi needle technique is considered a less invasive means of obtaining a bone biopsy as compared to a Michele trephine. A small stab incision is necessary to introduce this device and fractures are unlikely. In approximately 92% of cases, a correct diagnosis of tumor versus nontumor is achieved when using a Jamshidi needle (Powers et al. 1988). Instrumentation includes a no. 11 scalpel blade and a Jamshidi needle. The surgical site is clipped free of fur, and the patient is prepared with aseptic technique and draped. A 1–2 mm stab incision is made over the bony lesion. The Jamshidi needle is introduced into the stab incision and pressed onto the bony lesion. The stylet is then removed from the needle, and the needle is twisted until the cis cortex is penetrated. The Jamshidi needle is rocked back and forth to loosen the sample and then removed. The stylet is reintroduced into the needle in the opposite direction of the initial location. As the stylet is moved through the Jamshidi needle, the biopsy will be ejected from the base of the Jamshidi needle. Lymph node biopsy Treatment and biopsy of lymph nodes in neoplastic disease remains controversial (Gilson 1995). Removing

8  Veterinary Surgical Oncology

a lymph node or performing an incisional biopsy of a lymph node can aid in staging the patient and assist in determining prognosis or treatment options. The surgical oncologist should have a thorough knowledge of the anatomical location of the probable draining lymph node for a mass in a particular location. The excisional biopsy of superficial lymph nodes such as the mandibular, prescapular, axillary, inguinal, or popliteal lymph nodes is described below. For removal of lymph nodes within the thorax or abdomen, an exploration of that body cavity is performed, and the lymph nodes are removed by careful dissection and maintenance of hemostasis. Instrumentation includes a no. 10 or no. 15 scalpel blade, Metzenbaum scissors, forceps, suture, and suture scissors. The surgical site is clipped free of fur, and the patient is prepared with aseptic technique and draped. An incision slightly larger than the palpable lymph node is made parallel to the axis of the lymph node. The superficial tissue overlying the lymph node is bluntly and sharply dissected. The lymph node capsule is then grasped with the forceps and blunt or sharp dissection is performed around the lymph node to free it from the surrounding tissue. Vessels that are encountered may need to be ligated. The lymph node is then removed, and the subcutaneous tissue and skin are closed. Endoscopic biopsy Esophagoscopy, gastroscopy, duodenoscopy, and colonoscopy are routinely performed in veterinary medicine as minimally invasive techniques to attain biopsy tissue from the gastrointestinal tract. Biopsies attained during these procedures are generally smaller than that which can be achieved with an open procedure; however, the biopsies are often diagnostic, and the morbidity associated with these procedures is reduced over open procedures (Magne 1995; Moore 2003). Laparoscopy and thoracoscopy are still relatively underused modalities, but successful procurement of kidney, bladder, liver, spleen, adrenal gland, pancreas, stomach, intestine, and lung biopsies have been described by use of these procedures (Rawlings et al. 2002; Lansdowne et al. 2005; Vaden 2005; Barnes et al. 2006). Case selection is essential when considering these minimally invasive alternatives, as cases that have excessively large tumors or other potential contraindications should undergo an open procedure. Laparoscopy and thoracoscopy may have a role in the staging of veterinary patients as the use of these techniques increases. In cases where lymph node evaluation and biopsy would assist in predicting outcome or determining treatment, these procedures could be per-

formed by minimally invasive techniques (Fagotti et al. 2007). Surgical considerations for curative-intent surgery Certain surgical technical principles will improve the chance of success and minimize the risk of local or distant seeding of tumor cells. The tumor should be draped off from the rest of the surgical field. Surgeons should avoid contact with ulcerated or open areas of tumor with gloves or instruments. Sharp dissection is preferred over blunt dissection when possible, as this will decrease the likelihood of leaving neoplastic cells within the patient and decrease the risk of straying from the preestablished margin. Tension on skin closures should be avoided whenever possible, especially in cases that have undergone radiotherapy. Proper knowledge of tension-relieving techniques such as tension-relieving sutures and flaps can assist in closure (Soderstrom and Gilson 1995; Aiken 2003). If an indwelling drain is deemed necessary in a tumor resection site, the drain should be located in an area that can be resected during a subsequent surgery or in an area that will not compromise radiation therapy and can easily be included in the radiation field. Lastly, control of hemostasis and prevention of seroma or abscess development due to dead space is encouraged. Seromas or hematomas following an incomplete resection allow tumor cells to gain access to areas beyond the surgical field as these fluids may be widely dispersed throughout the subcutaneous space during movement. To decrease the risk of recurrence after tumor resection, there are several techniques the surgeon should practice. For tumors that have been previously biopsied or for which a drain has been placed, the biopsy tract and/or drain hole need to be removed en bloc with the tumor. Similarly, adhesions should be removed with the tumor, when possible. Leaving any of these can result in an increased risk of tumor recurrence. Additionally, when establishing a margin during surgical dissection, this margin must be maintained around the periphery of the tumor down to the deep margin. Straying from this may result in an incomplete resection. Similarly, the pseudocapsule present around a tumor should not be penetrated, as this pseudocapsule is constructed of a compressed layer of neoplastic cells (Soderstrom and Gilson 1995). Seeding of these cells will likely result in recurrence, and healing may be inhibited. Lastly, it is essential that a new set of instruments, gloves, and possibly drapes be used for closure of a wound created by tumor removal or reconstruction of a wound. This principle applies to the removal of subsequent tumors on the

Principles of Surgical Oncology  9

same patient as these items should not be transferred from one surgical site to another. Defining and evaluating surgical margins The evaluation of surgical margins of an excised specimen is an essential component to appropriate care in a cancer patient. A surgical margin denotes a tissue plane established at the time of surgical excision, the tissue beyond which remains in the patient. Excised masses should be submitted in their entirety for evaluation of the completeness of excision. The surgeon should indicate the margins with ink or some other method prior to placing the specimen in formalin to aid the pathologist in identifying the actual surgical margin. Because the larger tumor specimen is trimmed by a technician to fit on a microscope slide, the pathologist may not be oriented as to what represents a surgical margin versus a sectioning “margin”. Tissue ink on the surgical margin allows orientation throughout sectioning. The ink is present throughout the processing of the tumor specimen and is visible on the slide. If tumor cells are seen at the inked margin under the microscope, the surgical margin is by definition “dirty” or incomplete. The surgical techniques used to remove tumors define the type and magnitude of intended surgical margin. When tumors are removed using an intracapsular technique, dissection occurs within the dimensions of the tumor and residual microscopic disease always remains (Soderstrom and Gilson 1995). Marginal excision refers to tumors excised with a 1 cm or less cuff of normal tissue surrounding the mass. Marginal excision may be quite appropriate for certain tumors such as lipomas but is often not sufficient for malignant tumors (Ehrhart and Powers 2007). Wide excision refers to tumors removed with 1–3 cm of normal tissue in all directions, including a deep margin. To achieve wide excision, the mass needs to be removed en bloc and the pseudocapsule and reactive zone should be completely contained within a cuff of normal tissue. Because dissection for a wide excision is intracompartmental, it is distinguished from a radical excision. A radical excision is considered an excision of normal tissue surrounding the mass of greater than 3 cm or the entire anatomical compartment (e.g., amputation). Extracompartmental excision is defined by a plane of excision beyond the anatomical compartment considered to have a cancer-resistant tissue barrier (Soderstrom and Gilson 1995). Special focus is usually placed on mast cell tumors and soft tissue sarcomas when considering surgical margins. These tumor types generally have a bulky mass that is easily palpable; however, microscopic projections

of tumor cells extend out from the main tumor bed (Séguin et al. 2001; Murphy et al. 2004; Ehrhart 2005). These tendrils of tumor cells need to be considered preoperatively so that a proper surgical dose can be determined. Historically, 3 cm margins were recommended for excision of mast cell tumors and soft tissue sarcomas. Recently, though, studies have shown that 2 cm margins are sufficient for complete excision of 91%–100% of grade 2 mast cell tumors (Simpson et al. 2004; Fulcher et al. 2006). Recommendations for surgical margins around soft tissue sarcomas, however, continue to be at least 3 cm (Aiken 2003; Ehrhart 2005; Liptak and Forrest 2007). In many cases, the deep margin of a tumor excision can be less than 2–3 cm from the tumor if removal of one tissue plane deep to the last tissue plane the tumor touches is achieved. For example, if the tumor is freely moveable in the subcutaneous tissue of the thigh, removal of the fascia lata as the deep margin will often be sufficient to achieve a clean margin. On the other hand, if the tumor is attached to the fascia lata, a muscle plane deep to this layer must be removed to achieve a clean margin. Unfortunately, the true definition of a “fascial plane” is lacking in medicine, and specific guidelines remain elusive (Fasel et al. 2007). While to some authors the definition of fascia has included adipose tissue, this concept is not universally supported (Fasel et al. 2007). A current definition of fascia is considered “sheaths, sheets, or other dissectible connective tissue aggregations visible to the unaided eye” (Wendell-Smith 1997; Fasel et al. 2007). Furthermore, fascia can be “considered as gross structures enveloping and/or supporting other formations” (Fasel et al. 2007). These definitions support the removal of a deep layer of connective tissue (not including adipose tissue) when considering a deep margin. When an incomplete margin is noted on histopathological evaluation, the surgeon must decide on the next appropriate course of action. Options include intensive monitoring for recurrence, reexcision, and chemotherapy and radiation therapy. Both human and veterinary studies support early reexcision of a surgical wound bed when an incomplete margin is achieved during the primary surgery (Raney et al. 1982; Gibbs et al. 1997; Bacon et al. 2007). The goal during a reexcision surgery is to achieve tumor-free margins. Therefore, the entire wound bed must be treated as a dirty site and must be completely removed with a margin of normal tissue around it so that all tumor cells and microscopic extensions previously left in the patient will be removed. This always requires a more extensive surgery than the original surgical attempt.

10  Veterinary Surgical Oncology

Palliative and cytoreductive surgery The decision to perform a palliative or cytoreductive surgery is often a difficult one, and the surgeon needs to educate the client and referring veterinarian about the risks and benefits of such surgery. Piecemeal removal (debulking) of a mass should generally only be performed when the mass is physically causing obstruction or altering function. There is little advantage to debulking otherwise, unless the removal results in only microscopic amounts of disease left behind. Palliation of symptoms caused by obstructive masses by removing most of or portions of large masses can temporarily improve quality of life in some cases. This should be performed only when necessary as excessive bleeding can often occur and dehiscence is very common.

Postoperative Considerations Tissue marking As discussed above, following an excisional biopsy, the surgical margins of the mass should be clearly indicated in some way so that the histopathologist can accurately evaluate the mass for complete excision. Several methods have been proposed to do this, including specialized sectioning techniques, suture markers, inking, and the submission of adjacent tissue as a separate sample (Rochat et al. 1992; Mann and Pace 1993; Seitz et al. 1995). Inappropriate sectioning can result in neoplastic cells being noted at the cut margin, and a false-positive result can occur. Sutures can be used to mark a particular area of interest or for tumor orientation, but sutures need to be removed before sectioning to prevent microscopic artifact (Mann and Pace 1993). A sample of tissue surrounding the surgical wound can also be submitted for evaluation. However, this increases the size of the wound bed, and added expense may be seen due to the submission of extra biopsy samples. In general, the marking of tumor margins with inks or dyes is recommended. Several types of inks and dyes have been evaluated, including merbromin, laundry bluing, India ink, alcian blue, typists’ correction fluid, commercial acrylic pigments, and artists’ pigment in acetone (Rochat et al. 1992; Mann and Pace 1993; Seitz et al. 1995; Chiam et al. 2003). Alcian blue has been shown to be the best marking material; however, india ink and commercial kits (Davidson Marking System, IMEB Inc., San Diego, CA) are reasonable alternatives (Seitz et al. 1995). One of the benefits of the commercial kits is that multiple colors are provided. When using these kits, all the margins can be marked in different colors, but at a minimum, the lateral margin can be marked in one color and the deep margin in a different color. Yellow, black, and blue are considered the best

colors to use, whereas red and green are less ideal (Seitz et al. 1995). Guidelines for fixation of surgical tissue specimens Small biopsy samples should be placed in fixative immediately to prevent drying of the sample. Early fixation will initiate changes in the sample that will prevent autolysis and bacterial alteration of the sample (Stevens et al. 1974). In large biopsy submissions, the sample should be sliced evenly to allow more complete fixation (Dernell and Withrow 1998; Ehrhart and Withrow 2007). Many fixatives, including formalin, Bouin’s fluid, chilled isopentane, Zenker’s fluid, and glutaraldehyde have been described in veterinary medicine (Osborne 1974; Stevens et al. 1974), but in general, 10% buffered formalin is sufficient for almost all biopsies. A biopsy sample should be fixed in formalin in a 1:10 solution of tissue to formalin (Ehrhart and Withrow 2007). Frozen sections The use of frozen sections is common in human medicine. (Lessells and Simpson 1976; Kaufman et al. 1986). Frozen sections generate an accurate diagnosis in greater than 97% of human biopsy samples (Lessells and Simpson 1976; Kaufman et al. 1986). The process requires highly trained personnel and equipment specific to the procedure, and thus veterinary facilities that have the capability are limited (Ehrhart 1998). In one veterinary study, the accuracy of frozen sections in determining a specific diagnosis was 83% (Whitehair et al. 1993). In that same study, frozen sections were able to make a determination between neoplastic and nonneoplastic diseases in 93% of cases (Whitehair et al. 1993). Wound healing The veterinary oncology patient has several risk factors that may increase the frequency of complications associated with wound healing (Cornell and Waters 1995). Nutritional compromise and concomitant disease can be treated to improve the outcome of wound healing, but other factors such as tumor type and completeness of surgical excision have to be considered as well. Neoadjuvant and adjuvant therapies such as chemotherapy, radiotherapy, and antiangiogenic medications have also been documented to impair wound healing (Devereux et al. 1979; Cornell and Waters 1995; te Velde et al. 2002; Séguin et al. 2005). Proper surgical technique as described above can be employed to decrease the chance of wound complications. Regular communication with the patient’s owner

Principles of Surgical Oncology  11

both before and after surgery will help to preemptively prepare for complications or aid in rapid identification and intervention when complications arise. Prevention of self-trauma should be routinely discussed with the owner, and methods of prevention such as bandaging or having the patient wear an Elizabethan collar should be included in the postoperative care. Adjuvant therapy The time to discuss the potential need for adjuvant therapy in a tumor patient is prior to any surgical intervention. This allows owners to make informed choices and to better prepare for the financial burden, time required, and potential complications associated with this type of therapy. Failing to properly prepare the client for these additional treatments and the benefits and challenges unique to each one may leave the patient’s owner feeling overwhelmed, underinformed, and may expose the patient to unnecessary morbidity or delay in treatment. Chemotherapy in the adjuvant setting is generally administered after wound healing has been completed. Experimentally, it has been shown that administering certain types of chemotherapy before or at the same time as surgery may retard wound healing (Shamberger et al. 1981; de Roy van Zuidewijn et al. 1986; Lawrence, Talbot et al. 1986; Lawrence, Norton, et al. 1986). By the time a patient is ready for suture or staple removal, a wound is generally healed sufficiently, and chemotherapy may be administered. The results of the biopsy will also be accessible at a similar time, and these can help to guide chemotherapeutic recommendations. Radiation therapy may be administered preoperatively or postoperatively. In general, radiation therapy will slow wound healing. In cases where radiation is administered either before or after surgery, it is important to ensure that there is minimal tension on the wound closure. This requires careful planning prior to and during the initial surgery. In some cases, if local flaps require extensive dissection in areas away from the tumor bed and outside the proposed radiation field, it may be better to delay primary closure until it is known if tumor margins are clean. This will help prevent seeding of tumor cells along the dissection planes where the flap will be raised. In postoperative patients who require radiation therapy but have wound complications such as infection or dehiscence, it is often better to try to manage the wound complication before beginning radiation. This may not always be possible, as tumor remaining in the wound may prevent wound healing. In these cases, it may be necessary to go forward with radiation in an open wound setting. In many cases, once acute effects have resolved, the wound can be closed. In these

cases, strict adherence to the “no skin tension” rule is imperative. While certain basic concepts of surgery will remain static for the treatment of neoplasia, pursuit of better options for our patients will require that the surgical oncologist be able to adapt. It is hoped that the desire for improved outcomes will continue to improve the lives of our patients as well as their owners. Prolonging a quality of life for veterinary patients and advising their owners appropriately about the options that we have to offer should remain our goal as advances in therapy occur.

References Aiken, S.W. 2003. Principles of surgery for the cancer patient. Clin Tech Small An P 18:75–81. Atwater, S.W., B.E. Powers, R.D. Park, et al. 1994. Thymoma in dogs: 23 cases (1980–1991). J Am Vet Med Assoc 205:1007–1013. Bacon, N.J., W.S. Dernell, N. Ehrhart, et al. 2007. Evaluation of primary re-excision after recent inadequate resection of soft tissue sarcomas in dogs: 41 cases (1999–2004). J Am Vet Med Assoc 230:548–554. Barnes, R.F., C.L. Greenfield, D.J. Schaeffer, et al. 2006. Comparison of biopsy samples obtained using standard endoscopic instruments and the harmonic scalpel during laparoscopic and laparoscopicassisted surgery in normal dogs. Vet Surg 35:243–251. Berzina, I., L.C. Sharkey, I. Matise, et al. 2008. Correlation between cytologic and histopathologic diagnoses of bone lesions in dogs: A study of the diagnostic accuracy of bone cytology. Vet Clin Path 37:332–338. Bonfanti, U., W. Bertazzolo, E. Bottero, et al. 2006. Diagnostic value of cytologic examination of gastrointestinal tract tumors in dogs and cats: 83 cases (2001–2004). J Am Vet Med Assoc 229: 1130–1133. Britt, T., C. Clifford, A. Barger, et al. 2007. Diagnosing appendicular osteosarcoma with ultrasound guided fine-needle aspiration: 36 cases. J Small Anim Pract 48:145–150. Chiam, H.W., P.G. Maslen, and G.J. Hoffman. 2003. Marking the surgical margins of specimens: Commercial acrylic pigments are reliable, rapid and safe. Pathology 35:204–206. Cohen, M., M.W. Bohling, J.C. Wright, et al. 2003. Evaluation of sensitivity and specificity of cytologic examination: 269 cases (1999–2000). J Am Vet Med Assoc 222:964–967. Cornell, K. and D.J. Waters. 1995. Impaired wound healing in the cancer patient: Effects of cytotoxic therapy and pharmacologic modulation by growth factors. Vet Clin N Am-Small 25:111–131. DeBerry, J.D., C.R. Norris, V.F. Samii, et al. 2002. Correlation between fine-needle aspiration cytopathology and histopathology of the lung in dogs and cats. J Am Anim Hosp Assoc 38:327–336. Dernell, W.S. and S.J. Withrow. 1998. Preoperative patient planning and margin evaluation. Clin Tech Small An P 13:17–21. De Roy van Zuidewijn, D.B.W., T. Wobbes, T. Hendriks, et al. 1986. The effect of antineoplastic agents on the healing of small intestinal anastomoses in the rat. Cancer 58:62–66. DeRycke, L.M.J.H., J.J. van Bree, and P.J.M. Simoens. 1999. Ultrasoundguided tissue-core biopsy of liver, spleen and kidney in normal dogs. Vet Radiol Ultrasoun 40:294–299. Devereux, D.F., L. Thibault, J. Boretos, et al. 1979. The quantitative and qualitative impairment of wound healing by adriamycin. Cancer 43:932–938.

12  Veterinary Surgical Oncology Drixler, T.A., E.E. Voest, T.J.M.V. van Vroonhoven, et al. 2000. Angiogenesis and surgery: From mice to man. Eur J Surg 166:435–446. Eich, C.S., J.G. Whitehair, S.D. Moroff, et al. 2000. The accuracy of intraoperative cytopathological diagnosis compared with conventional histopathological diagnosis. J Am Anim Hosp Assoc 36:16–18. Ehrhart, E.J. and B.E. Powers. 2007. The pathology of neoplasia. In Small Animal Clinical Oncology, 4th edition, pp. 54–67. Stephen Withrow and David Vail, editors. Philadelphia: Saunders. Ehrhart, N. 1998. Principles of tumor biopsy. Clin Tech Small An P 13:10–16. Ehrhart, N. 2005. Soft-tissue sarcomas in dogs. J Am Anim Hosp Assoc 41:241–246. Ehrhart, N.P. and S.J. Withrow. 2007. Biopsy principles. In Small Animal Clinical Oncology, 4th edition, pp. 147–153. Stephen Withrow and David Vail, editors. Philadelphia: Saunders. Fagotti, A., F. Fanfani, R. Longo, et al. 2007. Which role for pre-treatment laparoscopic staging? Gynecol Oncol 107:S101– S105. Fasel, J.H., J.C. Dembé, and P.E. Majno. 2007. Fascia: A pragmatic overview for surgeons. Am Surgeon 73:451–453. Fulcher, R.P., L.L. Ludwig, P.J. Bergman, et al. 2006. Evaluation of a two-centimeter lateral surgical margin for excision of grade I and grade II cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 228:210–215. Ghisleni, G., P. Roccabianca, R. Ceruti, et al. 2006. Correlation between fine-needle aspiration cytology and histopathology in the evaluation of cutaneous and subcutaneous masses from dogs and cats. Vet Clin Path 35:24–30. Gibbs, C.P., T.D. Peabody, A.J. Mundt, et al. 1997. Oncological outcomes of operative treatment of subcutaneous soft-tissue sarcomas of the extremities. J Bone Joint Surg Am 79:888–897. Gilson, S.D. 1995. Clinical management of the regional lymph node. Vet Clin N Am-Small 24:149–167. Hanson, J.A., M. Papageorges, E. Girard, et al. 2001. Ultrasongoraphic appearance of splenic disease in 101 cats. Vet Radiol Ultrasoun 42:441–445. Kaufman, Z., S. Lew, B. Griffel, et al. 1986. Frozen-section diagnosis in surgical pathology. A prospective analysis of 526 frozen sections. Cancer 57:377–379. Lana, S., S. Plaza, K. Hampe, et al. 2006. Diagnosis of mediastinal masses in dogs by flow cytometry. J Vet Intern Med 20:1161– 1165. Lansdowne, J.L., E. Monnet, D.C. Twedt, et al. 2005. Thoracoscopic lung lobectomy for treatment of lung tumors in dogs. Vet Surg 34:530–535. Lawrence, W.T., T.L. Talbot, and J.A. Norton. 1986. Preoperative or postoperative doxorubicin hydrochloride (adriamycin): Which is better for wound healing? Surgery 100:9–13. Lawrence, W.T., J.A. Norton, A.K. Harvey, et al. 1986. Doxorubicininduced impairment of wound healing in rats. J Natl Cancer Inst 76:119–126. Lessells, A.M. and J.G. Simpson. 1976. A retrospective analysis of the accuracy of immediate frozen section diagnosis in surgical pathology. Br J Surg 63:327–329. Liptak, J.M., W.S. Dernell, N. Ehrhart, et al. 2004. Canine appendicular osteosarcoma: Diagnosis and palliative treatment. Comp Cont Educ Pract 26:172–183. Liptak, J.M. and L.J. Forrest. 2007. Soft tissue sarcomas. In Small Animal Clinical Oncology, 4th edition, pp. 425–454. Stephen Withrow and David Vail, editors. Philadelphia: Saunders. Magne, M.L. 1995. Oncologic applications of endoscopy. Vet Clin N Am-Small 25:169–183.

Mann, F.A. and L.W. Pace. 1993. Marking margins of tumorectomies and excisional biopsies to facilitate histological assessment of excision completeness. Semin Vet Med Surg 8:279–283. McEntee, M.C. 2006. Veterinary radiation therapy: review and current state of the art. J Amer Anim Hosp Assoc 42:94–109. Moore, L.E. 2003. The advantages and disadvantages of endoscopy. Clin Tech Small An P 18:250–253. Murphy, S., A.H. Sparkes, K.C. Smith, et al. 2004. Relationships between the histological grade of cutaneous mast cell tumours in dogs, their survival and the efficacy of surgical resection. Vet Rec 154:743–746. Nyland, T.G., S.T. Wallack, and E.R. Wisner. 2002. Needle-tract implantation following US-guided fine-needle aspiration biopsy of transitional cell carcinoma of the bladder, urethra, and prostate. Vet Radiol Ultrasoun 43:50–53. Ogilvie, G.K. 1998. Interventional nutrition for the cancer patient. Clin Tech Small An P 13:224–231. O’Reilly, M.S., T. Boehm, Y. Shing, et al. 1997. Endostatin: An endogenous inhibitor of angiogenesis and tumor growth. Cell 88:277–285. Osborne, C.A., V. Perman, and J.B. Stevens. 1974. Needle biopsy of the spleen. Vet Clin N Am-Small 4:311–316. Osborne, C.A. 1974. General principles of biopsy. Vet Clin N AmSmall 4:213–232. Powers, B.E., S.M. LaRue, S.J. Withrow, et al. 1988. Jamshidi needle biopsy for diagnosis of bone lesions in small animals. J Am Vet Med Assoc 193:205–210. Rae, C.A., R.M. Jacobs, and C.G. Couto. 1989. A comparison between the cytological and histological characteristics in thirteen canine and feline thymomas. Can Vet J 30:497–500. Raney, R.B., A.H. Ragab, F.B. Ruymann, et al. 1982. Softtissue sarcoma of the trunk in childhood. Cancer 49:2612– 2616. Rawlings, C.A., E.W. Howerth, S. Bement, et al. 2002. Laparoscopicassisted enterostomy tube placement and full-thickness biopsy of the jejunum with serosal patching in dogs. J Am Vet Med Assoc 63:1313–1319. Rochat, M.C., F.A. Mann, L.W. Pace, et al. 1992. Identification of surgical biopsy borders by use of India ink. J Am Vet Med Assoc 201:873–878. Roth, L. 2001. Comparison of liver cytology and biopsy diagnoses in dogs and cats: 56 cats. Vet Clin Path 30:35–38. Séguin, B., N.F. Leibman, V.S. Bregazzi, et al. 2001. Clinical outcome of dogs with grade-II mast cell tumors treated with surgery alone: 55 cases (1996–1999). J Am Vet Med Assoc 218:1120–1123. Séguin, B., D.E. McDonald, M.S. Kent, et al. 2005. Tolerance of cutaneous or mucosal flaps placed into a radiation therapy field in dogs. Vet Surg 34:214–222. Seitz, S.E., G.L. Foley, and S.M. Maretta. 1995. Evaluation of marking materials for cutaneous surgical margins. Am J Vet Res 56: 826–833. Shamberger, R.C., D.F. Devereux, and M.F. Brennan. 1981. The effect of chemotherapeutic agents on wound healing. Int Adv Surg Oncol 4:15–58. Simpson, A.M., L.L. Ludwig, S.J. Newman, et al. 2004. Evaluation of surgical margins required for complete excision of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 224:236–240. Soderstrom, M.J. and S.D. Gilson. 1995. Principles of surgical oncology. Vet Clin N Am-Small 25:97–110. Stanclift, R.M. and S.D. Gilson. 2008. Evaluation of neoadjuvant prednisone administration and surgical excision in treatment of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 232:53–62.

Principles of Surgical Oncology  13 Stevens, J.B., V. Perman, and C.A. Osborne. 1974. Biopsy sample management, staining, and examination. Vet Clin N Am-Small 4:233–253. Te Velde, E.A., E.E. Voest, J.M. van Gorp, et al. 2002. Adverse effects of the antiangiogenic agent angiostatin on the healing of experimental colonic anastomoses. Ann Surg Oncol 9:303–309. Vaden, S.L. 2005. Renal biopsy of dogs and cats. Clin Tech Small An P 20:11–22. Wang, K.Y., D.L. Panciera, R.K. Al-Rukibat, et al. 2004. Accuracy of ultrasound-guided fine-needle aspiration of the liver and cytologic

findings in dogs and cats: 97 cases (1990–2000). J Am Vet Med Assoc 224:75–78. Wendell-Smith, C.P. 1997. Fascia: An illustrative problem in international terminology. Surg Radiol Anat 19:273–277. Whitehair, J.G., S.M. Griffey, H.J. Olander, et al. 1993. The accuracy of intraoperative diagnoses based on examination of frozen sections. Vet Surg 22:255–259. Wykes, P.M., S.J. Withrow, and B.E. Powers. 1985. Closed biopsy for diagnosis of bone lesions in small animals. J Amer Anim Hosp Assoc 21:489–494.

2 Multimodal therapy Tania A. Banks

The key to success for the effective management of many cancers in animals today and in the future lies in employing a multipronged attack. Multimodal therapy requires not only a sound understanding of the strengths and weaknesses of each modality used, the tumor’s response to each treatment modality, and an accurate understanding of the various specific toxicities and interactions but also—and most importantly—a cooperative, communicative, interactive, and integrated team. The traditional trilogy in the oncology arsenal, surgery, radiation, and chemotherapy, remains the essence of the current mainstream multimodal protocol using a combination of some or all. The best outcomes will be realized when the multimodal therapy is planned and coordinated. The surgeon talks to and collaborates with the radiation oncologist while involving the medical oncologist from the outset as well. This also includes empowerment and involvement of the pet’s owners and family. These protocols are costly and time-consuming, and they require tweaking and adjustment throughout the course of treatment. A healthy client–doctor relationship is born from careful consultation with all specialists and the owner. All specialists must know the animal’s status and special situation as well knowing the owner. This way nothing is “lost in translation” and confidence is maintained, resulting in a strong sense of trust. Over the years, a volume of experiences has amassed and provided the wisdom to make these firm recommendations: plan, cooperate, and communicate. The surgeon, who operates on an animal with a solid tumor, then refers to a radiation oncologist to “mop up” residual disease has likely done the animal a disservice. Upfront surgical and radiation therapy planning minimizes surgical morbidity and minimizes normal tissue radia-

tion injury while maximizing the efficacy of the union of modalities. In the following sections the technology of multimodal therapy is presented in some detail using specific tumor settings to illustrate them. Underscoring this technology is the philosophy of this introduction and it cannot be overstated: the effective management of cancer in patients is a team event. Surgery remains the mainstay for treating many types of cancer in pets, and the value of a competent surgical oncologist cannot be understated. The type of surgical mind-set, specific knowledge, and technical prowess required is a specialized skill. Such a surgeon appreciates and can deliver what is required to expertly attempt a surgical cure or completely change tack for a palliative or diagnostic approach and employ other modalities synergistically. Radiation therapy as an addition to the available treatment options allows a greater scope and choice of therapy in many instances. Examples of tumors commonly treated with radiation include certain oral tumors, nasal tumors, brain tumors, mast cell tumors, and soft tissue sarcomas. Radiation can be used with palliative (e.g., to palliate bone pain with primary appendicular osteosarcoma) or curative intent and be used on its own or with surgery. For example, radiation can be used with surgery to treat a solid tumor in a location where wide clean margins cannot be achieved without limb amputation to save the limb. In this scenario, the radiation oncologist should be involved prior to surgery so that he or she can see if this approach is feasible; appreciate the size, fixation, and exact location of mass; plan the radiation field size and shape; determine how to spare normal tissue and include a large enough field; meet the owners; discuss complications, costs, expected outcome;

Veterinary Surgical Oncology, First Edition. Edited by Simon T. Kudnig, Bernard Séguin. © 2012 by John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.

15

16  Veterinary Surgical Oncology

etc. The surgeon’s role in this setting is a delicate, minimal surgery with intent to preserve blood and oxygen supply to the tissue to increase the effectiveness of radiation. A marginal resection to remove all the macroscopic tumor and allow primary closure is performed, and radiation therapy is used with surgery to provide long-term tumor control or cure. This approach differs greatly from a failed curative-intent surgery and a poorly healed or open, hypoxic, radiation-resistant wound and a delayed start to radiation therapy: a situation that is avoided by a team approach and good planning. When adjuvant radiation is planned, the surgeon can help the radiation oncologist by decreasing wound complications such as infection, dehiscence, and seroma formation. Preservation of blood supply, gentle tissue handling, aseptic technique, attention to hemostasis, use of fine, nonirritating (inert) suture material in minimal amounts, obliteration of dead space in the wound, avoidance of tension, postoperative rest, and use of bandages are all important. Drains should be avoided if possible, and if they are used, drainage entry and exit holes are included in the radiation field. Hemoclips can be placed in the wound intraoperatively to delineate the boundaries of the excised gross tumor burden to assist the radiation oncologist in planning the radiation field. Radiation can be used postoperatively (as in the above scenario) or preoperatively or intraoperatively, depending on tumor type and location. Sources of radiation therapy include megavoltage (>1 million electron volts of photon energy = high energy; maximum dose to tumor rather than skin) and orthovoltage (150–500 kVp = low energy; maximum dose to skin surface) external beam radiation, brachytherapy (interstitial placement of radioactive isotopes), or systemic or cavitary injection of radioisotopes (e.g., iodine131). Megavoltage irradiation has advanced with 3D imaging and planning, multileaf collimators, custommade blocks, etc. Chemotherapy can sometimes be used neoadjuvantly to “down-stage” (shrink) a primary tumor prior to surgery, and thus make it more amenable to surgical resection with clean margins. This may be appropriate for cutaneous and subcutaneous masses such as mast cell tumors and hemangiosarcomas. In this setting the surgeon needs to involve the medical oncologist prior to surgery. Chemotherapy also can prolong life postoperatively by addressing systemic metastasis; the classic example is appendicular osteosarcoma in dogs. Chemotherapy can be used immediately postoperatively or once the wound has healed, at the discretion of the medical oncologist and the surgeon. Surgery may have only a small role, such as for diagnostic biopsy, with the

sole treatment being chemotherapy, as is the case with lymphoma. Metronomic chemotherapy uses standard chemotherapy agents in a continuous administration, which requires lower doses to be used. The target of the drug is the tumor’s continually proliferating microvasculature, which is susceptible to chemotherapeutic effects with minimal systemic toxicity (Gately and Kerbel 2001). Bisphosphonates concentrate within areas of active bone remodeling and induce osteoclast apoptosis, which is of therapeutic benefit in managing pathological bone resorptive conditions such as osteosarcoma, multiple myeloma, and metastatic bone cancer. Bone pain is decreased, quality of life is improved, and progression of bone lesions is delayed (Fan et al. 2007). Other therapies that can be combined with more traditional therapies such as surgery, radiation, and chemotherapy include gene therapy, immunotherapy, and photodynamic therapy (PDT). Molecular and targeted therapies show great potential. These therapies include gene therapy (e.g., viral and nonviral vectors); targeting signal transduction that regulates cell growth, differentiation, survival, and death (e.g., via inhibition of protein kinase); RNA (ribonucleic acid) interference (the use of double-stranded RNA to cause posttranscriptional gene silencing); antiangiogenic factors (including metronomic chemotherapy and cyclo-oxygenase-2 inhibitors); and telomerase (enzyme that maintains telomeres or the protective structures at ends of chromosomes). Ninety-five percent of all canine cancers are associated with telomerase activity, whereas almost all normal cells have no telomerase activity (Argyle et al. 2007). Embolization treatments include “bland arterial embolization” (without chemotherapy) and chemoembolization (embolization with chemotherapeutic agents) that can be used as sole therapy or preoperatively to decrease tumor mass and size. Chemoembolization delivers chemotherapy to the tumor, allowing prolonged contact of the tumor to the chemotherapy without high systemic toxicity (Granov et al. 2005) and augmenting tumor ischemia (Weisse et al. 2002). There are several experimental studies of embolization treatments in healthy dogs, including chemoembolization with gemcitabine (Granov et al. 2005), carboplatin (Chen et al. 2004), and cisplatin (Nishioko et al. 1992). Bland arterial embolization resulted in decreased tumor growth, pain palliation, and control of hemorrhage in two dogs and one goat (Weisse et al. 2002) and decreased primary tumor size in a dog with a soft tissue sarcoma (Sun et al. 2002). A thinking surgical oncologist is always aware of the animal as a whole and how the behavior of the

Multimodal Therapy  17

specific cancer in the specific patient influences the surgeon’s role. The surgeon is cognizant of paraneoplastic syndromes, appropriate imaging and staging prior to and during surgery, appropriate supportive and follow-up care, and how various modalities can be used

synergistically to achieve maximal outcome with minimal morbidity. Tables 2.1, 2.2, and 2.3 outline various treatment modalities and published outcomes of these treatments for various types of cancers in dogs and cats.

Table 2.1.  Epithelial. Neoplasia

Researched Treatment Options and Outcomes

Cutaneous Squamous Cell Carcinoma

Modalities include surgery, radiation therapy, surgery combined with radiation therapy, photodynamic therapy, imiquimod, intralesional chemotherapy (alone or combined with hyperthermia or radiation), vitamin A–related synthetic retinoids. Cryosurgery is used for small lesions, and there are partial responses with piroxicam in dogs. For curative-intent radiation therapy (RT), median survival time (MST) is approximately 8–20 months. Curative-intent RT followed by surgical debulk (13 dogs) has MST of 47 months (Adams et al. 1987; Adams et al. 1998; Adams et al. 2005; Lana et al. 2004; McEntee et al. 1991; Nadeau et al. 2004; Theon et al. 1993). Curative–intent RT with CT planning has MST of approximately 11–20 months. Coarse-fraction RT (56 dogs) has MST of 7 months (Mellanby et al. 2002). Palliative 3D conformal RT (38 dogs) has overall median progression-free interval of 10 months (Buchholz et al. 2009). Chemotherapy alone (small numbers of dogs) is investigational (Hahn, Knapp, et al. 1992; Langova et al. 2004). Radiation sensitizers (some investigational) is generally not better than RT alone. Other treatments include brachytherapy, immunotherapy, cryotherapy, and PDT (all investigational) (Lucroy, Long, et al. 2003; MacEwen et al. 1977; Thompson et al. 1992; White et al. 1990; Withrow 1982). Transitional cell carcinoma (TCC) (dogs): Bladder/urethra: Surgery includes debulk, stent, bypass (e.g., prepubic cystostomy catheter for palliation if obstructed). Rarely is there complete excision. Nephrectomy is performed if one ureter is obstructed; if both ureters are obstructed, there is no benefit to surgery. Debulking surgery alone for bladder TCC has MST of 109 days (Lengerich et al 1992). One dog treated with resection of proximal urethra and bladder neck and bilateral ureteroneocystostomy and adjuvant chemotherapy for bladder survived 580 days (Saulnier-Troff et al. 2008). Radiation: Complications of urinary incontinence and cystitis occurs with whole-bladder intraoperative radiation (Walker and Breider 1987; Withrow et al. 1989). Coarse fractionation external beam radiation (with mitoxantrone-piroxicam) showed no benefit over mitoxantrone-piroxicam chemotherapy alone (Poirier, Forrest et al. 2004). Laparoscopically implanted tissue-expander radiotherapy shows promise in reducing radiation damage to surrounding tissues (one dog still alive at 21 months) (Murphy et al. 2008). Medical: Mitoxantrone-piroxicam combination with minimal toxicity has a MST of 291 days (Henry et al. 2003). For piroxicam alone, palliative MST is 195 days (20% alive after 1 year) (Mutsaers et al. 2002). Concurrent antibiotics are commonly needed. PDT: PDT is currently under investigation (Lucroy, Ridgway et al. 2003; Ridgway and Lucroy 2003). Prostatic carcinoma: Prostatectomy is performed for early stage disease confined to prostate capsule (but there is a high rate of urinary incontinence) (Basinger et al. 1989; Hardie et al. 1984). Other treatment modalities include transurethral-resection (TUR) (electrosurgical and investigational; relieves urethral obstruction, but 2 of 3 dogs had perforated urethra) (Liptak, Brutscher et al. 2004); Nd:YAG laser (investigational: 8 dogs had MST of 103 days; 3 died from complications within 16 days) (L’Eplattenier et al. 2006); stenting (investigational but very promising, with good to excellent outcome in 8 dogs) (Weisse et al. 2006); bypass obstruction (prepubic catheter); chemotherapy (investigational); NSAIDs (MST 6.9 months in 16 dogs, compared to 0.7 months with no cancer therapy in 15 dogs) (Sorenmo, Goldschmidt, et al. 2004); intraoperative prostatic radiation (MST for 10 dogs was 114 days) (Turrel 1987a); PDT (investigational) (Lucroy, Bowles et al. 2003; L’Eplattenier et al. 2008); and palliative radiation for skeletal metastasis. (Continued )

Intranasal Carcinoma

Transitional Cell Carcinoma (urogenital)

Table 2.1.  (Continued ) Neoplasia

Researched Treatment Options and Outcomes

Solitary Primary Lung Mass

Surgery (lung lobectomy) is performed. All adjuvant chemotherapy is investigational at this stage: systemic chemotherapy (vinorelbine) (Poirier, Burgess et al. 2004); inhalational chemotherapy (Hershey et al. 1999; Vail et al. 2000); intrapleural chemotherapy for malignant pleural effusion (Moore, Kirk et al. 1991). Resection is done if possible (about 70%), along with chemotherapy (Aronsohn 1985; Atwater et al. 1994; Willard et al. 1980; Martin et al. 1986, especially if concurrent with megaesophagus relating to a poor surgical candidates); radiation therapy has complete to partial responses, and many also received concurrent surgery or chemotherapy (Hitt et al. 1987; Kaser-Hotz et al. 2001; Smith et al. 2001). Treatment of concurrent myasthenia gravis is accomplished with immunosuppressive and or anticholinesterase therapy, H2 blockers, other supportive care. Intestinal: Modalities include surgery with wide margins (at least 5 cm); adjuvant doxorubicin chemotherapy in cats with colonic adenocarcinoma (Slaweinski et al. 1997); intracavitary chemotherapy for carcinomatosis (Moore, Kirk et al. 1991); adjuvant doxorubicin chemotherapy in dogs (Paoloni et al. 2002); piroxicam palliative for rectal tubulopapillary polyps if unresectable or as alternative to surgery (Knottenbelt et al. 2000). Surgery for primary tumor and regional (sublumbar) lymph nodes (repeated surgical removal of metastatic lesions may afford prolonged survival) (Hobson et al. 2006); debulking and omentalization of sublumbar nodes when nonresectable (Hoelzler et al. 2001); adjuvant radiation for local tumor and nodal metastasis; systemic chemotherapy (various protocols) (Goldschmidt and Zoltowski 1981; Williams et al. 2003; Turek et al. 2003; Ross et al. 1991; Bennett et al. 2002). Surgery is the treatment of choice, except for inflammatory carcinoma or if distant metastasis is present. Surgery includes nodulectomy, mammectomy, regional mastectomy, unilateral or bilateral mastectomy, and also lymph node removal for staging. The surgical choice depends on benign versus malignant, size, number, and species (cat versus dog). In cats, adequate surgical treatment combined with adjuvant chemotherapy may be of benefit to prolong survival time over surgery alone. In one paper, the MST of cats that received surgery and doxorubicin was 448 days, and the median disease-free interval (DFI) was 255 days (Novosad et al. 2006). There is no known proven effective adjuvant chemotherapy protocol for malignant or metastatic mammary tumors in dogs. Some preliminary studies show promise; in 14 dogs with stage III disease (T3 N0 M0) and 2 dogs with stage IV disease (any T N1 M0), half had cyclophosphamide and 5-FU, and half had regional mastectomy alone. The dogs receiving adjuvant chemotherapy had improved survival and disease-free interval (Karayannopoulou et al. 2001). Another study showed adjuvant gemcitabine chemotherapy postsurgery in dogs had no benefit (Marconato et al 2008). Doxorubicin and cyclophosphamide or cisplatin has some antitumor effect against mammary adenocarcinoma. Doxorubicin is associated with partial response with duration of 12 and 16 months in 2 dogs with metastatic mammary adenocarcinoma (Hahn, Richardson et al. 1992). Doxorubicin increases membrane viscosity and lipid hydroxyperoxide, and this effect is increased with concurrent medroxyprogesterone acetate (Pagnini et al. 2000). Doxorubicin has better efficacy than platinum drugs, and carboplatin and cisplatin have the same efficacy in mammary tumor cell culture; efficacy is not affected by cell type (adenocarcinoma, solid, mixed cell) (Simon et al. 2001). Piroxicam plus radiation therapy has produced best results with inflammatory carcinoma. In 44 human patients with inflammatory breast carcinoma, an 81% response rate was achieved with combination therapy (fluorouracil, doxorubicin, cyclophosphamide, mastectomy, and adjuvant paclitaxel) (Cristofanilli et al. 2001). Adjuvant chemotherapy is investigational, considered if poor prognostic factors are present (e.g., large, lymph node–positive, invasive, high grade), and administered after complete surgical removal (Lana et al. 2007).

Thymoma

Intestinal

Anal Sac Adenocarcinoma (see Figure 2.1)

Mammary

18

Table 2.1.  (Continued ) Neoplasia

Salivary Gland Carcinoma

Ear Carcinoma

Ovarian Carcinoma

Uterine Carcinoma Insulinoma

Thyroid Carcinoma

Hyperthyroid Cats

Researched Treatment Options and Outcomes Immunomodulation appears ineffective to date (Lana et al. 2007). Ovariohysterectomy early in life is preventative. Ovariohysterectomy as part of treatment is still not proven clearly to be of benefit (Fowler et al. 1974; Brodey et al. 1983; Morris et al. 1998; Yamagami et al. 1996; Sorenmo, Shofer et al. 2000). Tamoxifen is not recommended (Morris et al. 1993). Surgery is used for aggressive removal where possible, along with adjuvant radiation if there is incomplete resection (Carberry et al. 1987; Hammer et al. 2001; Evans and Thrall 1983; Carberry et al. 1988), and chemotherapy (investigational). Surgery is used (conservative if benign, radical if malignant, resectable, and no metastases) (Marino et al. 1994; Marino et al. 1993; London et al. 1996). Radiation is used as an alternative to surgery if unresectable or as adjuvant to incomplete resection (Theon et al. 1994); PDT used for local disease. Ovariohysterectomy: Intracavitary cisplatin for malignant effusion (Moore, Kirk et al. 1991; Olsen et al. 1994). Platinum drugs with tamoxifen used in metastatic human ovarian tumors. Hectate-b significantly reduces tumor burden (Gawronska et al. 2002). Chemotherapy has potential to prolong life in animals with metastatic ovarian cancer. Uterine adenocarcinoma in cats: Treatment is ovariohysterectomy. Role and effectiveness of radiation and chemotherapy is unknown. Modalities include surgery (resection primary, staging, debulking metastases), frequent feeds, prednisolone, streptozotocin, diazoxide, and octreotide (Feldman and Nelson 2004; Leifer et al. 1986; Tobin et al. 1999; Robben et al. 1997; Moore et al. 2002). Dogs: Mobile thyroidectomy (Carver et al. 1995; Klein et al. 1995; Panciera et al. 2004), fixed/ nonresectable radiation therapy treatment of choice 80% for 1-year survival, 72% for 3-year survival (Theon et al. 2000). 131I thyroid ablation can give prolonged survival in dogs with nonresectable thyroid carcinoma, with local/regional tumor MST at 839 days and MST 366 days for metastasis (Adams et al. 1995; Panciera et al. 2004; Peterson et al. 1989; Turrel et al. 2006; Worth et al. 2005). Chemotherapy is considered as adjuvant treatment for nonresectable primary or large primary carcinoma (>27 cm3), bilateral disease, or for gross metastatic disease) (Theon et al. 2000; Jeglum and Whereat 1983; Fineman et al. 1998; Post and Mauldin 1992; Ogilvie et al. 1991; Hammer et al. 1994; Gallick et al. 1993; Leav et al. 1976). Boron neutron capture therapy is investigational (Pisarev et al. 2006). Multinodular adenomatous hyperplasia (70%–75%), solitary benign adenomas (20%–25%), malignant carcinomas (1%–3%) (Bailey and Page 2007). 131I thyroid ablation is treatment of choice: oral antithyroid medication, topical methimasole to pinna, thyroidectomy (Padgett 2002; Flanders 1999), ultrasound-guided percutaneous ethanol injections (Wells et al. 2001), ultrasound-guided percutaneous radiofrequency ablation (Mellary et al. 2003). Preoperative scintigraphy is ideal (Bailey and Page 2007).

19

(a)

(b)

(c)

Figure 2.1.  (A) Anal sac adenocarcinomas treated with adjuvant megavoltage radiation. (B) Lead block used to spare normal tissue from RT. (C) Final setup including a tissue–equivalent “bolus” to allow the maximum dose of radiation to reach the tumor. (Courtesy of Mary-Kay Klein)

20

Table 2.2.  Round cell. Neoplasia

Researched Treatment Options and Outcomes

Mast Cell Tumor

Marginal surgery with adjuvant radiation results in 85%–95% 2-year control for stage 0, grade I or II (Turrel et al. 1988; Al-Sarraf et al. 1996; Frimberger et al. 1997; LaDue et al. 1998). Other modalities include marginal surgery with adjuvant chemotherapy (vinblastine and prednisolone; see Figure 2.2) (Davies et al. 2004), surgery with curative intent (2–3 cm margins depending on grade) (Simpson et al. 2004), vinblastine-prednisolone chemotherapy as adjuvant to surgery for high risk (mucous membrane origin, node positive, high-grade) (Thamm et al. 2006). Chemotherapy may also be used for dogs with multiple cutaneous mast cell tumors. The need for adjuvant chemotherapy for completely excised grade II tumors (when not in a poor prognostic location) is unpredictable; close monitoring is advisable (Seguin et al. 2001). A recent paper suggested dogs with a mitotic index (MI; number of mitoses per 10 high-power fields) is prognostic, with animals with MI = 0 not reaching median survival, animals with MI between 1 and 7 with MST of 18 months, and dogs with MI > 7 with a MST of 3 months (Elston et al. 2009). A higher MI may help identify which subset of grade II mast cell tumors would benefit from adjuvant chemotherapy. Other chemotherapy agents include lomustine, vincristine, prednisolone/ cyclophosphamide/vinblastine, cyclophosphamide/vincristine/prednisolone/hydroxyurea (Elmslie 1997; McCaw et al. 1997; Rassnick et al. 1999; Davies et al. 2004; Thamm et al. 1999; Gerritsen et al. 1988), vinorelbine (Grant et al. 2008), inhibitors of tyrosine kinase (SU11654), both direct antitumor and antiangiogenic activity (London et al. 2003; Liao et al. 2002). Other adjunctive medical therapies include H1 blocker, H2 blocker, omeprazole, sucralfate, and misoprostol. Pretreatment with prednisone prior to surgery (neoadjuvant) can reduce the size of mast cell tumors, facilitating resection with adequate margins in situations where margins cannot be confidently attained because of mass location or size or both (Stanclift and Gilson 2008). Multiple myeloma treatment modalities include chemotherapy using melphalan and prednisolone standard, as well as cyclophosphamide, CCNU, chlorambucil, doxorubicin, vincristine (Matus et al. 1986; MacEwen and Hurvitz 1977; Hanna 2005; Drazner 1982; Brunnert et al. 1992; Osborne et al. 1968; Gentilini et al. 2005; Fan et al. 2002; Vail 2007); surgery (stabilization of pathological fractures; see Figure 2.3) (Vail 2007; Banks et al. 2003) with or without adjuvant radiation therapy, bisphosphonates (Vail 2007); tyrosine kinase– inhibitor therapy SU11654 (London et al. 2003). Extramedullary treatment modalities (cutaneous) include conservative surgical resection (can add chemotherapy if local recurrence or incomplete margins) (Rusbridge et al. 1999; Kryiazidou et al. 1989). Radiation alone for stable solitary osseous plasmacytoma (MacEwen et al. 1984; Rusbridge et al. 1999; Meis et al. 1987). Surgery plus radiation for solitary osseous plasmacytoma resulting in an unstable long-bone fracture or surgery with or without radiation for solitary osseous plasmacytoma resulting in neurological compromise (Vail 2007). Various chemotherapy protocols (Boyce and Kitchell 2000; Carter et al. 1987; Cotter and Goldstein 1983; Garrett et al. 2002; Greenlee et al. 1990; Khanna et al. 1998; Keller et al. 1993; MacEwen et al. 1981; MacEwen et al. 1987; Morrison-Collister et al. 2003; Mutsaers et al. 2002; Myers et al. 1997; Page et al. 1992; Postorino et al. 1989; Stone et al. 1991; Valerius et al. 1997; Zenman et al. 1998) immunotherapy (investigational) (MacEwen et al. 1985; Crow et al. 1977; Jeglum et al. 1988; Steplewski et al. 1990; Rosales et al. 1988; Jeglum 1996); radiation therapy for whole body (localize stage I or stage II disease for nasal or CNS lymphoma, palliation of local disease) (Vail and Young 2007); bone marrow transplantation and staged half-body radiation after remission with induction of chemotherapy—both investigational (Williams et al. 2004; Gustavson et al. 2004)—or surgery for solitary lymphoma (early stage I) or solitary extranodal, or splenectomy for massive splenomegaly due to lymphosarcoma (Moldovanu et al. 1966; Brooks et al. 1987) or surgery for obstructive or ruptured gastrointestinal lymphoma (Marks 2001).

Plasma Cell Tumor

Lymphoma

21

(a)

(b)

Figure 2.2.  (A) Terrier breed dog with concurrent multiple mast cell tumors and previous history of having had several other mast cell tumors removed. (B) Same dog as (A), with tumor in a poor prognostic location (prepuce). This dog was treated with multiple marginal resections and adjuvant chemotherapy.

(a)

(b)

Figure 2.3.  (A) Multiple myeloma causing a pathological fracture of T12, treated with surgical stabilization and adjuvant chemotherapy. (B) Same dog as in (A), with surgical stabilization of a subsequent pathological fracture of the humerus. This dog survived approximately 8 months due to a combination of surgery and chemotherapy.

22

Table 2.3.  Mesenchymal. Neoplasia

Researched Treatment Options and Outcomes

Soft Tissue Sarcoma (Schwannoma, neurofibroma, peripheral nerve sheath tumor, etc.)

Surgery, wide margins, with curative intent (Baez et al. 2004; Banks, Straw et al. 2003; Banks, Straw et al. 2004; Dernell, Withrow et al. 1998; Kuntz et al. 1997; Posterino et al. 1988), surgery-marginal resection with adjuvant radiation (Evans 1987; Forrest et al. 2000; Graves et al. 1988; McKnight et al. 2000); systemic chemotherapy of possible benefit for highly anaplastic tumors but as yet unproven for grade III soft tissue carcinomas (Selting et al. 2000). Marginal resection and localized cisplatin chemotherapy into wound bed (OPLA-Pt/ Atrigel) as yet still investigational (Banks and Straw 2003). Metronomic chemotherapy (continuous low-dose chemotherapy) with cyclophosphamide and piroxicam significantly increased disease-free interval for incompletely resected soft tissue sarcomas compared to control dogs (Elmslie et al. 2008). Surgery (Davidson et al. 1997; Hershey et al. 2000; Kuntz CA unpublished data; Lidbetter et al. 2002; McEntee and Page 2001); surgery and radiation therapy (Cohen et al. 2001; Cronin et al. 1998; Bregazzi et al. 2001; Kobayashi et al. 2002), chemotherapy (Barber et al. 2000; Poirier et al. 2004; Bregazzi et al. 2001); immunotherapy (Jourdier et al. 2003; Kent 1993; King et al. 1995; Quintin-Colonna et al. 1996). Careful surgical dissection (peeling out), excellent prognosis (Thomson et al. 1999). Aggressive surgical resection, adjuvant radiation if margins incomplete (McEntee and Thrall 2001). Wide surgical resection with clean margins yields good prognosis (Baez et al. 2004). Adjuvant radiation if incomplete resection. Surgery, usually debulking, pericardiectomy for palliation (surgical or thoracoscopic), intracavitary and/or intravenous chemotherapy (Closa et al. 1999; Dunning et al. 1998; Jackson et al. 1999; Kerstetter et al. 1997; Moore, Kirk et al. 1991; Stepien et al. 2000; Seo et al. 2007; Sparkes et al. 2005; Spugnini et al. 2008). Early metastasis a concern even if complete resection achieved (Liptak and Brebner 2006). Surgery, chemotherapy, radiation therapy (all investigational, as very little reported) (Itoh et al. 2004). Surgery (high amputation is treatment of choice because local recurrence is higher with marginal or wide resection) (Vail et al. 1994); chemotherapy may be of benefit if sarcoma is high grade and there is no metastasis, or if the node is positive (Vail et al. 1994; Tilmant et al. 1986). Adjuvant radiation for incomplete excision investigational. Surgical resection with wide margins is treatment of choice (Lascelles et al. 2003; Schwarz et al. 1991a; Schwarz et al. 1991b; White 1991); if not resectable with clean margins, surgery and radiation therapy (unreported) or radiation therapy alone (palliative) (Thrall 1981). Systemic chemotherapy has no known benefit. For histologically low-grade, biologically high-grade oral fibrosarcoma, prognosis depends upon early diagnosis and aggressive treatment. Prolonged survival can be achieved in some dogs with surgery, radiotherapy alone, surgery and radiotherapy, and radiotherapy and local hyperthermia (Ciekot et al. 1994). For local disease, surgery with wide clean margins (Kudnig et al. 2003; Ramos-Vara et al. 2000; Kosovsky et al. 1991; Wallace et al. 1992; Schwarz et al. 1991a; Schwarz et al 1991b; Overly et al. 2001; MacEwen et al. 1986; Harvey et al. 1981; Hahn et al. 1994); repeat surgery with wide margins or adjuvant radiation therapy if margins incomplete; radiation therapy alone (Freeman et al. 2003; Bateman et al. 1994; Blackwood and Dobson 1996; Theon et al. 1997; Turrel 1987b; Proulx et al. 2003; Farrelly et al. 2004). Regional lymph node metastasis treatment includes surgery and or radiation therapy; chemotherapy (partial responses) (Kudnig et al. 2003; Overly et al. 2001; Rassnick et al. 2001; Page et al. 1991); immunotherapy (investigational) (Alexander et al. 2006; Bergman, Camps-Palau et al. 2003; Bergman, MacEwen et al. 2003; Bergman et al. 2004; Dow et al. 1998; Elmslie et al. 1994; Elmslie et al. 1995; MacEwen et al. 1999; Moore et al. 1991; Quintin-Colonna et al. 1996).

Vaccine-Associated Sarcomas in Cats

Intermuscular Lipoma Infiltrative Lipoma Liposarcoma Mesothelioma

Lymphangiosarcoma Synovial Cell Sarcoma

Oral Fibrosarcoma

Oral Melanoma

(Continued )

23

Table 2.3.  (Continued ) Neoplasia

Researched Treatment Options and Outcomes

Cutaneous Melanoma

Surgical excision is treatment of choice (Bolon et al. 1990; Aronsohn and Carpenter 1990); chemotherapy shows little response (Ogilvie et al. 1991; Moore 1993; Gillick and Spiegle 1987; Rassnick et al. 2001); hyperthermia and intralesional cisplatin/carboplatin (Theon et al. 1991) and PDT (Cheli et al. 1987; Dougherty et al. 1981) have short-lived responses. Radiation therapy likely to be of use if melanoma not surgically excisable (Vail and Withrow 2007). Immunomodulation is investigational (Hogge et al. 1999; Dow et al. 1998; Hajduch et al. 1997; Quintin-Colonna et al. 1996; Alexander et al. 2006; MacEwen et al. 1999; Bergman et al. 2006; Bianco et al. 2003; Gyorffy et al. 2005). Surgery (amputation/limb-spare) (Vasseur 1987; Berg et al. 1992; LaRue et al. 1989; Thrall et al. 1990; Withrow et al. 1993; Morello et al. 2003; Buracco et al. 2002; Ehrhart 2005; Tomamassini et al. 2000; Ehrhart et al. 2002; Rovesti et al. 2002; Seguin et al. 2003; Pooya et al. 2004; Liptak, Dernell, et al. 2004; Huber et al. 2000); hemipelvectomy (Straw et al. 1992); partial scapulectomy (Trout et al. 1995; Kirpensteijn et al. 1994); ulnectomy (Straw et al. 1991). Local chemotherapy as adjuvant to limb-sparing (OPLA-Pt) reduced local recurrence rate (Straw et al. 1994; Withrow et al. 2004). Local chemotherapy (isolated limb perfusion) (Van Ginkel et al. 1995) as adjuvant to limb-sparing (investigational); radiation therapy to primary site (palliative as alternative to amputation/limb-spare) (McEntee et al. 1993; Ramirez et al. 1999; Mueller et al. 2005; Green et al. 2002; Heidner et al. 1991); adjunctive to limb-spare (Thrall et al. 1990; Withrow et al. 1993), radioisotopes (Milner et al. 1998; Aas et al. 1999); chemotherapy (various protocols) adjuvant to limb-spare or amputation (clear benefit) (Thompson and Fugent 1992; Bergman et al. 1996; Berg et al. 1995; Berg et al. 1997; Kent et al. 2004); chemotherapy neoadjuvant to limb-sparing to downstage disease presurgery (Withrow et al. 1993; O’Brien et al. 1996); chemotherapy as an adjuvant to palliative radiation (role unclear) (Walter et al. 2005). Surgery as cure of local disease, prolonged survival if local disease cured (MST 14 months even if metastasis present at diagnosis) (Dernell, Straw et al. 1998). If surgical removal not possible, consider surgical debulk plus adjuvant radiation therapy (Straw et al. 1989). Wide surgical excision significantly improves survival. Median survival time is 540 days treated with amputation alone (Popovitch et al. 1994); chest wall resection MST 1,080 days (Pirkey-Ehrhart et al. 1995); wide surgical excision for nonnasal sites MST of 3,097 days and did not reach MST (Waltman et al. 2007); and MST 979 days for 25 dogs with appendicular chondrosarcoma treated with amputation alone, although grade was found to be prognostic (Farese et al. 2009). Debulking and adjuvant radiation therapy if location is not amenable to curative resection, to radiation alone (Popovitch et al. 1994; Lana et al. 1997), or have objective responses to coarse fraction radiation alone (Dernell 2007). Metastasis still occurs in about 25%, even after surgical resection. Grade may be prognostic for survival (Waltman et al. 2007; Farese et al. 2009). Dogs: Stage I surgery (MST 780 days), stage II and III surgery (and adjuvant doxorubicin chemotherapy should be considered) (Ward et al. 1994). Twenty-one dogs with subcutaneous (17) and intramuscular (4) hemangiosarcomas, with adequate local tumor control and no metastasis at presentation, were treated with adjuvant doxorubicin. Five dogs also received adjuvant radiation therapy. The MST for subcutaneous HSA was 1,189 days and for intramuscular was 272.5 days (Bulakowski et al 2008). Cats: Wide surgical excision (metastasis occurs less frequently than dogs, but adjuvant chemotherapy may have a role, depending on the case) (Miller et al. 1992; Kraje et al. 1999; McAbee et al. 2005). Radiation therapy is considered adjuvantly if incompletely resected local disease.

Appendicular Osteosarcoma

Multilobular Osteochondrosarcoma Chondrosarcoma

Cutaneous Hemangiosarcomas (HSA)

24

Multimodal Therapy  25 Table 2.3.  (Continued ) Neoplasia

Researched Treatment Options and Outcomes

Visceral HSA

Surgery (e.g., splenectomy) (Spangler and Culbertson 1992; Spangler and Kass 1997; Brown et al. 1985; Sorenmo, Baez, et al. 2004; Prymak et al. 1988); adjuvant chemotherapy of various types can be considered for splenic hemangiosarcomas, with median survival times of 141–179 days reported (Ogilvie et al. 1996; Hammer et al. 1991; Sorenmo, Duda, et al. 2000; Sorenmo, Baez et al. 2004; Sorenmo, Samluk, et al. 2004; Sorenmo et al. 1993; Vail et al. 1995). Immunotherapy (Vail et al. 1995) and angiogenic therapy are investigational (Sorenmo, Duda, et al. 2000). Localized (skin/subcutis): Aggressive surgery with clean margins (Affolter and Moore 2002); adjuvant radiation therapy if incomplete resection; adjuvant chemotherapy (unknown role but likely to be warranted due to high metastatic potential) (Liptak and Forrest 2007). Disseminated/malignant histiocytosis: Chemotherapy can give durable partial responses but is generally unrewarding (Skorupski et al. 2003). Prognosis good with complete surgical resection.

Histiocytic Sarcomas

Uterine Leiomyoma and Leiomyosarcoma in Dogs Vaginal and Vulval Tumors

Most are benign (leiomyoma and fibroma in cat, leiomyoma and lipoma in dog).

References Aas, M., L. Moe, and H. Gamlenm. 1999. Internal radionuclide therapy of primary osteosarcoma in dogs, using 153Sm-ethylenediamino-tetramethylene-phosphonate (EDTMP). Clin Cancer Res 5:3148s–3152s. Adams, W.M., D.E. Bjorling, J.E. McAnulty, et al. 2005. Outcome of accelerated radiotherapy alone or accelerated radiotherapy followed by exenteration of the nasal cavity in dogs with intranasal neoplasia: 53 cases (1990–2002). J Am Vet Med Assoc 227(6):936–941. Adams, W.M., P.E. Miller, D.M. Vail, et al. 1998. An accelerated technique for irradiation of malignant canine nasal and paranasal sinus tumors. Vet Radiol Ultrasound 39(5):475–481. Adams, W.H., M.A. Walker, G.B. Daniel, et al. 1995. Treatment of differentiated thyroid carcinoma in 7 dogs utilizing 131I. Vet Radiol Ultrasound 36:417. Adams, W.M., S.J. Withrow, R. Walshaw, et al. 1987. Radiotherapy of malignant nasal tumors in 67 dogs. J Am Vet Med Assoc 191:311–315. Affolter, V.K. and P.F. Moore. 2002. Localised and disseminated histiocytic sarcoma of dendritic cell origin in dogs. Vet Pathol 39:74–83. Alexander, A.N., M.K. Huelsmeyer, M. Mitzey, et al. 2006. Development of an allogeneic whole-cell tumor vaccine expressing xenogeneic gp100 and its implementation in a Phase II clinical trial canine patients with malignant melanoma. Cancer Immunol Immunother 18:1–10. al-Sarraf, R., G.N. Mauldin, A.K. Patnaik, et al. 1996. A prospective study of radiation therapy for the treatment of grade 2 mast cell tumors in 32 dogs. J Vet Intern Med 10:376–378. Argyle, D.J., C. London, R. Chun, et al. 2007. Molecular/targeted therapy of cancer. In Withrow & MacEwen’s Small Animal Clinical Onoclogy, 4th edition, pp. 236–274. S.J. Withrow and D.M. Vail, editors. St. Louis: Saunders.

Aronsohn, M. 1985. Canine thymoma. Vet Clin North Am Small Anim Pract 15(4):755–767. Aronsohn, M.G. and J.L. Carpenter. 1990. Distal extremity melanocytic nevi and malignant melanomas in dogs. J Am Anim Hosp Assoc 26:605–612. Atwater, S.W., B.E. Powers, R.D. Park, et al. 1994. Thymoma in dogs: 23 cases (1980–1991). J Am Vet Med Assoc 205(7):1007– 1013. Baez, J.L., M.J. Hendrick, F.S. Shofer, et al. 2004. Liposarcomas in dogs: 56 cases (1989–2000). J Am Vet Med Assoc 224:887–891. Bailey, D.B. and R.L. Page. 2007. Tumors of the endocrine system. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 583–609. S.J. Withrow and D.M. Vail, editors. St. Louis: Saunders. Banks, T.A. and R.C. Straw. 2003. Canine soft tissue sarcomas: Incomplete resection treated with opla-pt sponges. AVA Annual Conference, Science Week, Australian College of Veterinary Scientists, Surgery Chapter. Gold Coast, Queensland. Banks, T.A., R.C. Straw, B.E. Powers, et al. 2004. Soft tissue sarcoma in dogs: A study assessing surgical margin, tumor grade and clinical outcome. Aust Vet Pract 34:158–163. Banks, T.A., R.C. Straw, S.J. Withrow, et al. 2003. Prospective study of canine soft tissue sarcoma treated by wide surgical excision: Quantitative evaluation of surgical margins. Proceedings of the 23rd Annual Conference of the Veterinary Cancer Society, Madison, WI, p. 21. Banks, T.A., V. Langova, and R.C. Straw. 2003. Repair of three pathological fractures in a dog with multiple myeloma. Aust Vet Pract 33:98–102. Barber, L.G., K.U. Sorenmo, K.L. Cronin, et al. 2000. Combined doxorubicin and cyclophosphamide chemotherapy for nonresectable feline fibrosarcoma. J Am Anim Hosp Assoc 36:416–421. Basinger, R.R., C.A. Rawlings, J.A. Barsanti, et al. 1989. Urodynamic alterations associated with clinical prostatic diseases and prostatic surgery in 23 dogs. J Am Anim Hosp Assoc 25:385–392.

26  Veterinary Surgical Oncology Bateman, K.E., P.A. Catton, P.W. Pennock, et al. 1994. 0-7-21 radiation therapy for the treatment of canine oral melanoma. J Vet Intern Med 8:267–272. Bennett, P.F., D.B. DeNicola, P. Bonney, et al. 2002. Canine anal sac adenocarcinomas: Clinical presentation and response to therapy. J Vet Intern Med 16(1):100–104. Berg, J., M.C. Geghard, and W.M. Rand. 1997. Effect of timing on postoperative chemotherapy on survival of dogs with osteosarcoma. Cancer 79:1343–1350. Berg, J., M.J. Weinstein, D.S. Springfield, et al. 1995. Response of osteosarcoma in the dog to surgery and chemotherapy with doxorubicin. J Am Vet Med Assoc 206:1555–1560. Berg, J., M.L. Weinstein, S.H, Schelling, W.M. Rand. 1992. Treatment of dogs with osteosarcoma by administration of cisplatin after amputation or limb-sparing surgery: 22 cases (1987–1990). J Am Vet Med Assoc 200:2005–2008 Bergman, P.J., M.A. Camps-Palau, J.A. McKnight, et al. 2003. Development of a xenogeneic DNA vaccine program for canine malignant melanoma with prolongation of survival in dogs with locoregionally controlled stage II–II disease. Vet Cancer Soc Proc 23:40. Bergman, P.J., M.A. Camps-Palau, J.A. McKnight, et al. 2004. Phase I & IB trials of murine tyrosinase ± human GM-CSF DNA vaccination in dogs with advanced malignant melanoma. Vet Cancer Soc Proc 24:55. Bergman, P.J., M.A. Camps-Palau, J.A. McKnight, et al. 2006. Development of a xenogeneic DNA vaccine program for canine malignant melanoma at the Animal Medical Centre. Vaccine 24:4582–4585. Bergman, P.J., E.G. MacEwen, I.D. Kurzman, et al. 1996. Amputation and carboplatin for treatment of dogs with osteosarcoma: 48 cases (1991–1993). J Vet Int Med 10:76–81. Bergman, P.J., J. McKnight, A. Novosad, et al. 2003. Long-term survival of dogs with advanced malignant melanoma after DNA vaccination with xenogeneic human tyrosinase: A phase I trial. Clin Cancer Res 9:1284. Bianco, S.R., J. Sun, S.P. Fosmire, et al. 2003. Enhancing antimelanoma immune responses through apoptosis. Cancer Gene Ther 10:726–736. Blackwood, L. and J.M. Dobson. 1996. Radiotherapy of oral malignant melanomas in dogs. J Am Vet Med Assoc 209:98. Bolon, B., M.B. Calderwood Mays, and B.J. Hall. 1990. Characteristics of canine melanoma and comparison of histology and DNA ploidy to their biologic effect. Vet Pathol 27:96–102. Boyce, K.L. and B.E. Kitchell. 2000. Treatment of canine lymphoma with COPLA/LVP. J Am Anim Hosp Assoc 36:395–403. Bregazzi, V.S., S.M. LaRue, E. McNiel, et al. 2001. Treatment with a combination of doxorubicin, surgery, and radiation versus surgery and radiation alone for cats with vaccine-associated sarcomas: 25 cases (1995–2000). J Am Vet Med Assoc, 218:547–550. Brodey, R.S., M.A. Goldschmidt, and J.R. Rozel. 1983. Canine mammary gland neoplasms. J Am Anim Hosp Assoc 19:61–90. Brooks, M.B., R.E. Matus, C.E. Leifer, et al. 1987. Use of splenectomy in the management of lymphoma in dogs: 16 cases (1976–1985). J Am Vet Med Assoc 191:1008–1010. Brown, N.O., A.K. Patnaik, and E.G. MacEwen. 1985. Canine haemangiosarcoma: Retrospective analysis of 104 cases. J Am Vet Med Assoc 186:56–58. Brunnert, S.R., L.A. Dee, A.J. Herron, et al. 1992. Gastric extramedullary plasmacytoma in a dog. J Am Vet Med Assoc 200:1501–1502. Buchholz, J., R. Hagen, C. Leo, et al. 2009. 3D conformal radiation therapy for palliative treatment of canine nasal tumors. Vet Radiol Ultrasound 50(6):679–683.

Bulakowski, E.J., J.C. Philibert, S. Siegel, et al. 2008. Evaluation of outcome associated with subcutaneous and intramuscular haemangiosarcoma treated with adjuvant doxorubicin in dogs: 21 cases (2001–2006). J Am Vet Med Assoc 233:122–128. Buracco, P., E. Morello, M. Martano, et al. 2002. Pastuerised tumoral autograft as a novel procedure for limb sparing in the dog: A clinical report. Vet Surg 31:525–532. Carberry, C.A., J.A. Flanders, W.I. Anderson, et al. 1987. Mast cell tumor in the mandibular salivary gland in a dog. Cornell Vet 77:362–366. Carberry, C.A., J.A. Glanders, H.J. Harveyet al. 1988. Salivary gland tumors in dogs and cats: A literature and case review. J Am Anim Hosp Assoc 24:561–567. Carter, R.F., C.K. Harris, S.J. Withrow, et al. 1987. Chemotherapy of canine lymphoma with histopathological correlation: Doxorubicin alone compared to COP as first treatment regimen. J Am Anim Hosp Assoc 23:221–225. Carver, J.R., A. Kapatkin, and A.K. Patnaik. 1995. A comparison of medullary thyroid carcinoma and thyroid adenocarcinoma in dogs: A retrospective study of 38 cases. Vet Surg 24:315. Cheli, R., F. Addis, C.M. Mortellaro, et al. 1987. Photodynamic therapy of spontaneous animal tumors using the active component of hematoporphyrin derivative (DHE) as photosensitizing drug: clinical results. Cancer Lett 38:101–105. Chen, C.L., M.H. Sun, D.C. Tan, et al. 2004. Comparison of pelvic transarterial chemoembolization with lipiodol ultra-fluid carboplatin and transarterial carboplatin through experiment in dogs. Ai Zheng 23 (11 Suppl):1405–1408. Ciekot, P.A., B.E. Powers, S.J. Withrow, et al. 1994. Histologically lowgrade, yet biologically high-grade, fibrosarcomas of the mandible and maxilla in dogs: 25 cases (1982–1991). J Am Vet Med Assoc 14(204):610–615. Closa, J.M., A. Font, and J. Mascort. 1999. Pericardial mesothelioma in a dog: Long-term survival after pericardiectomy in combination with chemotherapy. J Small Anim Pract 40:383–386. Cohen, M., J.C. Wright, W.R. Brawner, et al. 2001. Use of surgery and electron beam irradiation, with or without chemotherapy, for treatment of vaccine-associated sarcomas in cats: 78 cases (1996– 2000). J Am Vet Med Assoc 219:1582–1589. Cotter, S.M. and M.A. Goldstein. 1983. Treatment of lymphoma and leukaemia with cyclophosphamide, vincristine and prednisolone: I. Treatment of the dog. J Am Anim hosp Assoc 19:159–165. Cristofanilli, M., A.U. Buzdar, N. Sneige, et al. 2001. Paclitaxel in the multimodality treatment for inflammatory breast carcinoma. Cancer 92:1775–1782. Cronin, K., R.L. Page, G. Spodnick, et al. 1998. Radiation therapy and surgery for fibrosarcoma in 33 cats. Vet Radiol Ultrasound 39:51–56. Crow, S.E., G.H. Theilen, E. Benjamini, et al. 1977. Chemoimmunotherapy for canine lymphosarcoma. Cancer 40:2102–2108. Davidson, E.B., C.R. Gregory, and P.H. Kass. 1997. Surgical excision of soft tissue fibrosarcomas in cats. Vet Surg 26:265–269. Davies, D.R., K.M. Wyatt, J.E. Jardine, et al. 2004. Vinblastine and prednisolone as adjunctive therapy for canine cutaneous mast cell tumors. J Am Anim Hosp Assoc 40:124–130. Dernell, W.S. 2007. Tumors of the skeletal system. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 540– 582. S.J. Withrow and D.M. Vail, editors. St. Louis: Saunders. Dernell, W.S., R.C. Straw, M.F. Cooper, et al. 1998. Multilobular osteochondrosarcoma in 39 dogs: 1979–1993. J Am Anim Hosp Assoc 34:11–18. Dernell, W.S., S.J. Withrow, C.A. Kuntz, et al. 1998. Principles of treatment for soft tissue sarcoma. Clin Tech Small Anim Pract 13:59.

Multimodal Therapy  27 Dougherty, T.J. 1981. Photoradiation therapy for cutaneous and subcutaneous malignancies. J Invest Dermatol 77:122–124. Dow, S.W., R.E. Elmslie, A.P. Wilson, et al. 1998. In vivo tumor transfection with superantigen plus ctyokine genes induces tumor regression and prolongs survival in dogs with malignant melanoma. J Clin Invest 101:2406–2414. Drazner, F.H. 1982. Multiple myeloma in the cat. Comp Cont Ed Pract Vet 4:206–216. Dunning, D., E. Monnet, E.C. Orton, et al. 1998. Analysis of prognostic indicators for dogs with pericardial effusion: 46 cases (1985– 1996). J Am Vet Med Assoc 212:1276–1280. Ehrhart, N. 2005. Longitundinal bone transport for treatment of primary bone tumors in dogs: Technique description and outcome in 9 dogs. Vet Surg 34:24–34. Ehrhart, N., J.E. Eurell, M. Tomassini, et al. 2002. The effect of cisplatin chemotherapy on regenerate bone formation. Am J Vet Res 63:703–711. Elmslie, R. 1997. Combination chemotherapy with and without surgery for dogs with high grade mast cell tumors with regional lymph node metastases. Vet Cancer Soc Newsl 20:6–7. Elmslie, R.E., S.W. Dow, and T.A. Potter. 1994. Genetic immunotherapy of canine oral melanoma. Vet Cancer Soc Proc 14:111. Elmslie, R.E., P. Glawe, and S.W. Dow. 2008. Metronomic therapy with cyclophosphamide and piroxicam effectively delays tumor recurrence in dogs with incompletely resected soft tissue sarcomas. J Vet Int Med 22(6)1373–1379. Elmslie, R.E., T.A. Potter, and S.W. Dow. 1995. Direct DNA injection for the treatment of malignant melanoma. Vet Cancer Soc Proc 15:52. Elston, L., F.A. Sueiro, J. Cavalcanti, et al. 2009. The importance of the mitotic index as a prognostic factor for canine cutaneous mast cell tumors—a validation study. Vet Pathol, Jan 13. (Epub ahead of print). Evans, S.M. 1987. Canine haemangiopericytoma: A retrospective analysis of response to surgery and orthovoltage radiation. Vet Radiol 28:13. Evans, S.M., and D.E. Thrall. 1983. Postoperative orthovoltage radiation therapy of parotid salivary gland adenocarcinoma in three dogs. J Am Vet Med Assoc 182:993–994. Fan, T.M., L.P. de Lorimier, K. O’Dell-Anderson, et al. 2007. Singleagent pamidronate for palliative therapy of canine appendicular osteosarcoma bone pain. J Vet Intern Med 21(3):431–439. Fan, T.M., B.E. Kitchell, R.S. Dhaliwal, et al. 2002. Haematological toxicity and therapeutic efficacy of lomustine in 20 tumor-bearing cats: Critical assessment of a practical dosing regimen. J Am Anim Hosp Assoc 38:357–363. Farese, J.P., J. Kirpensteijn, M. Kik, et al. 2009. Biologic behavior and clinical outcome of 25 dogs with canine appendicular chondrosarcoma treated by amputation: A Veterinary Society of Surgical Oncology retrospective study. Vet Surg 38:914–919. Farrelly, J., D.L. Denman, A.E. Hohenhaus, et al. 2004. Hypofractionated radiation therapy of oral melanoma in five cats. Vet Radiol Ultrasound 45:91–93. Feldman, E.C. and R.W. Nelson. 2004. Beta-cell neoplasia: Insulinomas. In Canine and Feline Endocrinology and Reproduction, 3rd edition. St. Louis: Saunders. Fineman, L.S., T.A. Hamilton, A. de Gotarib et al. 1998. Cisplatin chemotherapy for treatment of thyroid carcinoma in dogs: 13 cases. J Am Anim Hosp Assoc 34:109. Flanders, J.A. 1999. Surgical options for the treatment of hyperthyroidism in the cat. J Feline Med Surg 1:127. Forrest, L.J., R. Chun, W.M. Adams, et al. 2000. Postoperative radiotherapy for canine soft tissue sarcoma. J Vet Intern Med 14:578–82.

Fowler, E.H., G.P. Wilson, and A.A. Koestner. 1974. Biologic behaviour of canine mammary neoplasms based on a histogenetic classification. Vet Pathol 11:212–229. Freeman, K.P., K.A. Hahn, F.D. Harris, et al. 2003. Treatment of dogs with oral melanoma by hypofractionated radiation therapy and platinum-based chemotherapy (1987–1997). J Vet Int Med 17:96–101. Frimberger, A.E., A.S. Moore, S.M. LaRue, et al. 1997. Radiotherapy of incompletely resected, moderately differentiated mast cell tumors in the dog: 37 cases (1989–1993). J Am Anim Hosp Assoc 33:320–324. Gallick, G.E., M.S. Talamonti, and G.L. Nicholson. 1993. Protooncogenes and tumor-suppressing genes in endocrine. In Endocrine Tumors, E.L. Mazzaferri and N.A. Sammaan, editors. Cambridge: Blackwell Scientific. Garrett, L.D., D.H. Thamm, R. Chun et al. 2002. Evaluation of a 6-month chemotherapy protocol with no maintenance therapy for dogs with lymphoma. J Vet Int Med 16:704–709. Gately, S. and R. Kerbel. 2001. Antiangiogenic scheduling of lower dose cancer chemotherapy. Cancer J 7(5):427–436. Gawronska, B., C. Leuschner, F.M. Enright, et al. 2002. Effects of a lytic peptide conjugated to beta HCG on ovarian cancer: Studies in vitro and in vivo. Gynecol Oncol 85:45–52. Gentilini, F., C. Calzolari, A. Buonacucina, et al. 2005. Different biological behaviour of Waldenstrom macroglobulinemia in two dogs. Vet Comp Oncol 3:87–97. Gerritsen, R.J., E. Teske, J.S. Kraus, et al. 1988. Multiagent chemotherapy for mast cell tumors in the dog. Vet Q 20:28–31. Gillick, A. and M. Spiegle. 1987. Decarbazine treatment of malignant melanoma in a dog. Can Vet J 28:204. Goldschmidt, M.H. and C. Zoltowski. 1981. Anal sac gland carcinoma in the dog: 14 cases. J Small Anim Pract 22:119–128. Granov, A.M., A.V. Pavlovskii, D.A. Granov, et al. 2005. Radiopaque oil chemical arterial embolisation in treatment of pancreatic cancer. Vestn Ross Akad Med Nauk 8:16–21. Grant, I.A., C.O. Rodriguez, M.S. Kent, et al. 2008. A phase II clinical trial of vinorelbine in dogs with cutaneous mast cell tumours. J Vet Intern Med 22:388–393. Graves, G.M., D.E. Bjorling, and E. Mahaffey. 1988. Canine haemangiopericytoma: 23 cases (1067–1984). J Am Vet Med Assoc 192:99–102. Green, E.M., W.M. Adams, and L.J. Forrest. 2002. Four fraction palliative radiotherapy for osteosarcoma in 24 dogs. J Am Anim Hosp Assoc 38:4450451. Greenlee, P.G., D.A. Filippa, F.W. Quimby, et al. 1990. Lymphoma in dogs: A morphologic, immunologic, and clinical study. Cancer 66:480–490. Gustavson, N.R., S.E. Lana, M.N. Mayer, et al. 2004. A preliminary assessment of whole-body radiotherapy interposed within a chemotherapy protocol for canine lymphoma. Vet Comp Oncol 2:125–131. Gyorffy, S., J.C. Rodriguez-Locompte, J.P. Woods, et al. 2005. Bone marrow-derived dendritic cell vaccination of dogs with naturally occurring melanoma by using human gp100 antigen. J Vet Intern Med 19:56–63. Hahn, K.A., D.B. DeNicola, R.C. Richardson, et al. 1994. Canine oral malignant melanoma: Prognostic utility of an alternative staging system. J Small Anim Pract 35:251–256. Hahn, K.A., D.W. Knapp, R.C. Richardson, et al. 1992. Clinical response of nasal adenocarcinoma to cisplatin chemotherapy in 11 dogs. J Am Vet Med Assoc 200:255–357. Hahn, K.A., R.C. Richardson, and D.W. Knapp. 1992. Canine malignant mammary neoplasia: Biological behaviour, diagnosis, and treatment alternatives. J Am Anim Hosp Assoc 28:251–256.

28  Veterinary Surgical Oncology Hajduch, M., Z. Kolar, R. Novotny, et al. 1997. Induction of apoptosis and regression of spontaneous dog melanoma following in vivo application of synthetic cyclin-dependent kinase inhibitor olomoucine. Anticancer Drugs 8:1007–1013. Hammer, A.S., C.C. Couto, and R.D. Ayl. 1994. Treatment of tumorbearing dogs with actinomycin D. J Vet Intern Med 8:236. Hammer, A.S., C.G. Couto, J. Filppi, et al. 1991. Efficacy and toxicity of VAC chemotherapy (vincristine, doxorubicin, and cyclophosphamide) in dogs with haemangiosarcoma. J Vet Intern Med 5:160–166. Hammer, A., D. Getzy, G. Olgilvie, et al. 2001. Salivary gland neoplasia in the dog and cat: Survival times and prognostic factors. J Am Vet Med Assoc 37:478–482. Hanna, F. 2005. Multiple myeloma in cats. J Feline Med Surg 7:275–287. Hardie, E.M., J.A. Barsanti, and C.A. Rawlings. 1984. Complications of prostatic surgery. J Am Anim Hosp Asooc 20:50–56. Harvey, H.J., G.E. MacEwen, D. Braun, et al. 1981. Prognostic criteria for dogs with oral melanoma. J Am Vet Med Assoc 178:580–582. Heidner, G.L., R.L. Page, M.C. McEntee, et al. 1991. Treatment of canine appendicular osteosarcoma using cobalt 60 radiation and intra-arterial cisplatin. J Vet Intern Med 5:313–316. Henry, C.J., D.L. McCaw , S.E. Turnquist, et al. 2003. Clinical evaluation of mitoxantrone and piroxicam in a canine model of human invasive urinary bladder carcinoma. Clin Cancer Res 9:906–911. Hershey, A.E., I.D. Kurzman, L.J. Forrest, et al. 1999. Inhalation chemotherapy for macroscopic primary or metastatic lung tumors: Proof of principle using dogs with spontaneously occurring tumors as a model. Clin Cancer Res 5:2653–2659. Hershey, A.E., K.U. Sorenmo, M.J. Hendrick, et al. 2000. Prognosis for presumed feline vaccine-associated sarcoma after excision: 61 cases (1986–1996). J Am Vet Med Assoc 216:58–61. Hitt, M.E., D.P. Shaw, P.M. Hogan, et al. 1987. Radiation treatment for thymoma in a dog. J Am Vet Med Assoc 190:1187–1190. Hobson, H.P., M.R. Brown, and K.S. Rogers. 2006. Surgery of metastatic anal sac adenocarcinoma in five dogs. Vet Surg 35:267– 70. Hoelzler, M.G., J.R. Bellah, and M.C. Donofro. 2001. Omentalisation of cystic sublumbar lymph nodes metastasis for long-term palliation of tenesmus and dysuria in a dog with anal sac adenocarcinoma. J Am Vet Med Assoc 219:1729–1731. Hogge, G.S., J.K. Burkholder, J. Culp, et al. 1999. Preclinical development of human granulocyte-macrophage colony-stimulating factor-transfected melanoma cell vaccine using established canine cell lines and normal dogs. Cancer Gene Ther 6:26–36. Huber, D.J., S.J. Withrow, S.M. LaRue, et al. 2000. Limb sparing with intraoperative radiation for bone sarcomas. Vet Surg 29:464– 456. Itoh T, K. Mikawa, M. Mikawa, et al. 2004. Lymphangiosarcoma in a dog treated with surgery and chemotherapy. J Vet Med Sci 66:197–199. Jackson, J., K. Richter, and D. Launer. 1999. Thoracoscopic partial pericardiectomy in 13 dogs. J Vet Int Med 13:529–533. Jeglum, K.A. 1996. Chemoimmunotherapy of canine lymphoma with adjuvant canine monoclonal antibody. Vet Clin North Am Small Anim Pract 26:73–85. Jeglum, K.A., and A. Whereat. 1983. Chemotherapy of canine thyroid carcinoma. Compend Contin Educ Pract Vet 5:96–98. Jeglum, K.A., K.M. Young, K. Barnsley, et al. 1988. Chemotherapy versus chemotherapy with intralymphatic tumor cell vaccine in canine lymphoma. Cancer 61:2042–2050. Jourdier, T.M., C. Moste, M.C. Bonnet, et al. 2003. Local immunotherapy of spontaneous feline fibrosarcoma using recombinant

poxviruses expressing interleukin 2 (IL2). Gene Ther 10: 2126–2132. Karayannopoulou, M., E. Kaldrymidou, T.C. Constantinidis, et al. 2001. Adjuvant post-operative chemotherapy in bitches with mammary cancer. J Vet Med A Physiol Pathol Clin Med 48:85–96. Kaser-Hotz, B., C.R. Rohrer, J.L. Fidel, et al. 2001. Radiotherapy in three suspect cases of feline thymoma. J Am Anim Hosp Assoc 37:483–488. Keller, E.T., E.G. MacEwen, and R.C. Rosenthal. 1993. Evaluation of prognostic factors and sequential combination chemotherapy with doxorubicin for canine lymphoma. J Vet Int Med 7:289– 295. Kent, E.M. 1993. Use of an immunostimulant as an aid in treatment and management of fibrosarcoma in three cats. Fel Pract 21:13. Kent, M.S., A. Strom, C.A. London, et al. 2004. Alternating carboplatin and doxorubicin as an adjunctive chemotherapy to amputation or limb-sparing surgery in the treatment of appendicular osteosarcoma in dogs. J Vet Intern Med 18:540–544. Kerstetter, K.K., D.J. Krahwinkel, D.L. Millis, et al. 1997. Pericardiectomy in dogs, 22 cases (1978–1994). J Am Vet Med Assoc 211:736–740. Khanna, C., E.M. Lund, K.A. Redic, et al. 1998. Randomised controlled trial of doxorubicn versus dactinomycin in a multiagent protocol for treatment of dogs with lymphoma. J Am Vet Med Assoc 213:985–990. King, G.K., K.M. Yates, P.G. Greenlace, et al. 1995. The effect of acemannan immunostimulant in combination with surgery and radiation therapy on spontaneous canine and feline fibrosarcomas. J Am Anim Hosp Assoc 31:439–447. Kirpensteijn, J., R.C. Straw, A.D. Pardo, et al. 1994. Partial and total scapulectomy in the dog. J Am Anim Hosp Assoc 30:313–319. Klein, M.L., B.E Powers, S.J. Withrow, et al. 1995. Treatment of thyroid carcinoma in dogs by surgical resection alone: 20 cases (1981– 1989). J Am Vet Med Assoc 206:1007. Knottenbelt, C.M., J.W. Simpson, S. Tasker, et al. 2000. Preliminary clinical observations on the use of piroxicam in the management of rectal tubulopapillary polyps. J Small Anim Pract 41:393– 397. Kobayashi, T., M.L. Hauck, R. Dodge, et al. 2002. Preoperative radiotherapy for vaccine-associated sarcoma in 92 cats. Vet Radiol Ultrasound 43:473–479. Kosovsky, J.K., D.T. Matthiesen, S.M. Maretta, et al. 1991. Results of partial mandibulectomy for the treatment of oral tumous in 142 dogs. Vet Surg 20:397–401. Kraje, A.C., E.A. Mears, K.A. Hahn, et al. 1999. Unusual metastatic behaviour and clinicopathologic findings in eight cats with cutaneous or visceral haemangiosarcoma. J Am Vet Med Assoc 214:670–672. Kryiazidou, A., P.J. Brown, and V.M. Lucke. 1989. Immunohistochemical staining of neoplastic and inflammatory plasma cell lesions in feline tissues. J Comp Pathol 100:337–341. Kudnig, S.T., N. Ehrhart, S.J. Withrow, et al. 2003. Survival analysis of oral melanoma in dogs. Vet Cancer Soc Proc 23:39. Kuntz, C.A., W.S. Dernell, B.E. Powers, et al. 1997. Prognostic factors for surgical treatment of soft-tissue sarcomas in dogs: 75 cases (1986–1996). J Am Vet Med Assoc 211:1147–1151. Lengerich, E.J., R.F. Teclaw, M. Mendlein, et al. 1992. Pet populations in the catchment area of the Purdue Comparative Oncology Program. J Am Vet Med Assoc 200:51–56. L’Eplattenier, H.F., B. Klem, E. Teske, et al. 2008. Preliminary results of intra-operative photodynamic therapy with 5-aminolevulinic acid in dogs with prostate carcinoma. Vet J 178:202–207.

Multimodal Therapy  29 L’Eplattenier, H.F., S.A. van Nimwegen, F.J. van Sluijs, et al. 2006. Partial prostatectomy using Nd: YAG laser for management of canine prostatic carcinoma. Vet Surg 35:406–411. LaDue, T., G.S. Price, R. Dodge, et al. 1998. Radiation therapy for incompletely resected canine mast cell tumors. Vet Radiol Ultrasound 39:57–62. Lana, S.E., W.S. Dernell, M.S. Lafferty, et al. 2004. Use of radiation and a slow-release cisplatin formulation for treatment of canine nasal tumors. Vet Radiol Ultrasound 45:1–5. Lana, S.A., W.S. Dernell, S.M. LaRue, et al. 1997. Slow release cisplatin combined with radiation for the treatment of canine nasal tumors. Vet Radiol 38:474–478. Lana, S.E., G.R. Rutteman, and S.J. Withrow. 2007. Tumors of the mammary gland. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 619–636. S.J. Withrow and D.M. Vail, editors. St. Louis. Langova, V., A.J. Mutsaers, B. Philips, et al. 2004. Treatment of eight dogs with nasal tumors with alternating doses of doxorubicin and carboplatin in conjunction with oral piroxicam. Aust Vet J 82:676–680. LaRue, S.M., S.J. Withrow, B.E. Powers, et al. 1989. Limb-sparing treatment for osteosarcoma in dogs. J Am Vet Med Assoc 195:1734–1744. Lascelles, B.D., M.J. Thomson, W.S. Dernell, et al. 2003. Combined dorsolateral and intraoral approach for the resection of tumors of the maxilla in the dog. J Am Anim Hosp Assoc 39:294–305. Leav, I., A.L. Schiller, A. Rijnberk, et al. 1976. Adenomas and carcinomas of the canine and feline thyroid. Am J Pathol 83:61. Leifer, C.A., M.E. Peterson, and R.E. Matus. 1986. Insulin-secreting tumor diagnosis and medical and surgical management in 55 dogs. J Am Vet Med Assoc 188:60. Liao, A.T., M.B. Chien, N. Shenoy, et al. 2002. Inhibition of constituitively active forms of mutant kit by multitargeted indolinone tyrosine kinase inhibitors. Blood 100:585–593. Lidbetter, D.A., F.A. Williams, D.J. Krahwinkel, et al. 2002. Radical lateral body-wall resection for fibrosarcoma with reconstruction using polypropylene mesh and a caudal superficial epigastric axial pattern flap: A retrospective clinical study of the technique and results in 6 cats. Vet Surg 31:57–64. Liptak, J.M. and N.S. Brebner. 2006. Hemidiaphragmatic reconstruction with a transversus abdominis muscle flap after resection of a solitary diaphragmatic mesothelioma in a dog. J Am Vet Med Assoc 15(228):1204–1208. Liptak, J.M., S.P. Brutscher, E. Monnet, et al. 2004. Transurethral resection in the management of urethral and prostatic neoplasia in 6 dogs. Vet Surg 33:505–516. Liptak, J.M., W.S. Dernell, B.D. Lascelles, et al. 2004. Intra-operative extracorporeal irradiation for limb sparing in 13 dogs. Vet Surg 33:446–456. Liptak, J.M. and L.J. Forrest. 2007. Soft tissue sarcomas. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 435–454. S.J. Withrow and D.M. Vail, editors. St. Louis: Saunders. London, C.A., R.R. Dubilzeig, D.M. Vail, et al. 1996. Evaluation of dogs and cats with tumors of the ear canal: 145 cases (1978–1992). J Am Vet Med Assoc 208:1431–1418. London, C.A., A.L. Hannah, R. Zodovoskaya, et al. 2003. Phase I dose escalating study of SU11654, a small molecule receptor tyrosine kinase inhibitor, in dogs with spontaneous malignancies. Clin Cancer Res 9:2755–2768. Lucroy, M.D., M.H. Bowles, R.G. Higbee, et al. 2003. Photodynamic therapy for prostatic carcinoma in a dog. J Vet Int Med 17:235–237.

Lucroy, M.D., K.R. Long, M.A. Blaik, et al. 2003. Photodynamic therapy for the treatment of intranasal tumors in 3 dogs and 1 cat. J Vet Intern Med 17:727–729. Lucroy, M.D., T.D. Ridgway, G.M. Peavy, et al. 2003. Preclinical evaluation of 5-aminolevulinic acid-based photodynamic therapy for canine transitional cell carcinoma. Vet Comp Oncol 1:76–85. MacEwen, E.G., N.O. Brown, A.K. Patnaik, et al. 1981. Cyclic combination chemotherapy of canine lymphosarcoma. J Am Vet Med Assoc 178:1178–1181. MacEwen, E.G., A.A. Hayes, R.E. Matus, et al. 1987. Evaluation of some prognostic factors for advanced multicentric lymphosarcoma in the dog: 147 cases (1987– 1981). J Am Vet Med Assoc 190:564–568. MacEwen, E.G., A.A. Hayes, S. Mooney, et al. 1985. Levamisole as adjunctive to chemotherapy for canine lymphosarcoma. J Biol Response Mod 4:427–433. MacEwen, E.G. and A.I. Hurvitz. 1977. Diagnosis and management of monoclonal gammopathies. Vet Clin North Am Small Anim Pract 7:119–132. MacEwen, E.G., I.D. Kurzman, D.M. Vail, et al. 1999. Adjuvant therapy for melanoma in dogs: Results of randomised clinical trials using surgery, liposome-encapsulated muramyl tripeptide and granulocyte-macrophage colony-stimulating factor. Clin Cancer Res 5:42–49. MacEwen, E.G., A.K. Patnaik, H.J. Harvey, et al. 1986. Canine oral melanoma: comparison of surgery versus surgery plus Corynebacterium parvum. Cancer Invest 4:397–402. MacEwen, E.G., A.K. Patnaik, A.I. Hurvitz, et al. 1979. Nonsecretory multiple myeloma in two dogs. Am Vet Med Assoc 174:1321–1325. MacEwen, E.G., A.K. Patnaik, G.F. Johnson, et al. 1984. Extramedullary plasmacytoma of the gastrointestinal tract in two dogs. J Am Vet Med Assoc 184:1396–1398. MacEwen, E.G., S.J. Withrow, and A.K. Patnaik. 1977. Nasal tumors in the dog: Retrospective evaluation of diagnosis, prognosis, and treatment. J Am Vet Med Assoc 170:45–48. Marconato, L., R.M. Lorenzo, F. Abramo, et al. 2008. Adjuvant gemcitabine after surgical removal of aggressive malignant mammary tumors in dogs. Vet Comp Oncol 6:90–101. Marino, D.J., J.M. MacDonald, D.T. Matthiesen, et al. 1993. Results of surgery and long-term follow-up in dogs with ceruminous gland adenocarcinoma. J Am Anim Hosp Assoc 29:560–563. Marino, D.J., J.M. MacDonald, D.T. Matthiesen, et al. 1994. Results of surgery in cats with ceruminous gland adenocarcinoma. J Am Anim Hosp Assoc 30:54–58. Marks, S. 2001. Update on canine and feline gastrointestinal neoplasia. Proceedings of the 26th World Small Animal Veterinary Association World Congress, Vancouver, BC. Martin, R.A., E.W. Evans, J.R. August, et al. 1986. Surgical treatment of a thymoma in a cat. J Am Anim Hosp Assoc 22:347–354. Matus, R.E., C.E. Leifer, E.G. MacEwen, et al. 1986. Prognostic factors for multiple myeloma in the dog. J Am Vet Med Assoc 188:1288–1291. McAbee, K.P., L.L. Ludwig, P.J. Bergman, et al. 2005. Feline cutaneous haemangiosarcoma: A retrospective study of 18 cases (1988–2003). J Am Anim Hosp Assoc 41:110–116. McCaw, D.L., M.A. Miller, P.J. Bergman, et al. 1997. Vincristine therapy for mast cell tumors in dogs. J Vet Intern Med. 11:375–378. McEntee, M.C. and R.L. Page. 2001. Feline vaccine-associated sarcomas. J Vet Int Med 15:176–182. McEntee, M.C., R.L. Page, G.L. Heidner, et al. 1991. A retrospective study of 27 dogs with intranasal neoplasms treated with cobalt radiation. Vet Radiol Ultrasound 32:135–139.

30  Veterinary Surgical Oncology McEntee, M.C., R.L. Page, and C.A. Novotney. 1993. Palliative radiotherapy for canine appendicular osteosarcoma. Vet Radiol Ultrasound 34:367–370. McEntee, M.C. and D.E. Thrall. 2001. Computed tomographic imaging of infiltrative lipoma in 22 dogs. Vet Radiol Ultrasound 42:221–225. McKnight, J.A., N. Mauldin, M.C. McEntee, et al. 2000. Radiation treatment for incompletely resected STS in dogs. J Am Vet Med Assoc 217:205–210. Meis, J.M., J.J. Butler, B.M. Osborne, et al. 1987. Solitary plasmacytomas of bone and extramedullary plasmacytomas. Cancer 59:1475–1487. Mellanby, R.J., R.K. Stevenson, M.E. Herrtage, et al. 2002. Long-term outcome of 56 dogs with nasal tumors treated with four doses of radiation at intervals of seven days. Vet Rec 151:253–257. Mellary, K.F., R.E. Pollard, R.W. Nelson, et al. 2003. Percutaneous ultrasound-guided radiofrequency heat ablation for treatment of hyperthyroidism in cats. J Am Vet Med Assoc 223:1602. Miller, M.A., J.A. Ramos, and J.M. Kreeger. 1992. Cutaneous vascular neoplasia in 15 cats: Clinical, morphologic, and immunohistochemical studies. Vet Pathol 29:329–336. Milner, R.J., I. Dormehl, W.K. Louw, et al. 1998. Targeted radiotherapy with Sm-153-EDTMP in nine cases of canine primary bone tumors. J S Afr Vet Assoc 69:12–17. Moldovanu, G., M. Friedman, and D.G. Miller. 1966. Treatment of canine malignant lymphoma with surgery and chemotherapy. J Am Vet Med Assoc 148:153–156. Moore, A.S. 1993. Recent advances in chemotherapy for nonlymphoid malignant neoplasms. Comp Cont Ed Pract Vet 15:1039–1050. Moore, A.S., C. Kirk, and A. Cardona. 1991. Intracavitary cisplatin chemotherapy experience with six dogs. J Vet Intern Med 5:277–231. Moore, A.S., R.W. Nelson, C.J. Henry, et al. 2002. Streptozotocin for treatment of pancreatic islet cell tumors in dogs: 17 cases (1989– 1999). J Am Vet Med Assoc 221:811. Moore, A.S., G.H. Theilen, A.D. Newell, et al. 1991. Preclinical study of sequential tumor necrosis factor and interleukin 2 in the treatment of spontaneous canine neoplasms. Cancer Res 51:233–238. Morello, E., E. Vasconi, M. Martano, et al. 2003. Pasteurised tumoral autograft and adjuvant chemotherapy for the treatment of canine distal radial osteosarcoma: 13 cases. Vet Surg 32:539–544. Morris, J.S., J.M. Dobson, and D.E. Bostock. 1993. Use of tamoxifen in the control of mammary neoplasia. Vet Rec 133:539–542. Morris, J.S., J.M. Dobson, D.E. Bostock, et al. 1998. Effect of ovariohysterectomy in bitches within mammary neoplasms. Vet Rec 142:656–658. Morrison-Collister, K.M. Rassnick, N.C. Northrup, et al. 2003. A combination chemotherapy protocol with MOPP and CCNU consolidation (Tufts VELCAP-SC) for the treatment of canine lymphoma. Vet Comp Oncol 1:180–190. Mueller, F., V. Priorier, K. Melzer, et al. 2005. Palliative radiotherapy with electrons of appendicular osteosarcoma in 54 dogs. In Vivo, 19:713–716. Murphy, S., A. Gutierrez, and J. Lawrence. 2008. Laparoscopically implanted tissue expander radiotherapy in canine transitional cell carcinoma. Vet Radiol Ultrasound 49:400–405. Mutsaers, A.J., N.W. Glickman, D.B. DeNicola, et al. 2002. Evaluation of treatment with doxorubicin and piroxicam or doxorubicin alone for multicentric lymphoma in dogs. J Am Vet Med Assoc 220:1813–1817. Myers, N.C., A.S. Moore, W.M. Rand, et al. 1997. Evaulation of a multidrug chemotherapy protocol (ACOPA II) in dogs with lymphoma. J Vet Int Med 11:333–339.

Nadeau, M.E., B.E. Kitchell, R.L. Rooks, et al. 2004. Cobalt radiation with or without low-dose cisplatin for treatment of canine nasosinus carcinomas. Vet Radiol Ultrasound 45:362–367. Nishioko, Y., S. Kyotani, M. Okamura, et al. 1992. A study of embolizing materials for chemo-embolisation therapy of hepatocellular carcinoma: Embolic effect of cisplatin albumin microspheres using chitin and chitosan in dogs, and changes of cisplatin content in blood and tissue. Chem Pharm Bull (Tokyo) 40:267–268. Novosad, C.A., P.J. Bergman, M.G. O’Brien, et al. 2006. Retrospective evaluation of adjunctive doxorubicin for the treatment of feline mammary gland adenocarcinoma: 67 cases. J Am Anim Hosp Assoc 42:110–120. O’Brien, M.G., S.J. Withrow, and R.C. Straw. 1996. Limb-sparing surgery versus amputation for dogs with bone tumors. Vet Clin North Am 26:135–143. Ogilvie, G.K., J.E. Obradovich, R.E. Elmslie, et al. 1991. Efficacy of mitoxantrone against various neoplasms in dogs. J Am Vet Med Assoc 198:1618–1621. Ogilvie, G.K., B.E. Powers, C.H. Mallinckrodt, et al. 1996. Surgery and doxorubicin in dogs with haemangiosarcoma. J Vet Intern Med 10:379–384. Olsen, J., J. Komtebedde, A. Lackner, et al. 1994. Cytoreductive treatment of ovarian carcinoma in a dog. J Vet Int Med 8:133–134. Osborne, C.A., V. Perman, J.H. Sauter, et al. 1968. Multiple myeloma in the dog. J Am Vet Med Assoc 153:1300–1319. Overly, B., M. Goldschmidt, F. Schofer, et al. 2001. Canine oral melanoma: A retrospective study. Vet Cancer Soc Proc 21:43. Padgett, S. 2002. Feline thyroid surgery. Vet Clin North Am Small Pract 32:851. Page, R.L., D.W. Macy, G.K. Ogilvie, et al. 1992. Phase III evaluation of doxorubicin and whole body hyperthermia in dogs with lymphoma. Int J Hyperthermia 8:187–197. Page, R.L., D.E. Thrall, M.W. Dewhirst, et al. 1991. Phase I study of melphalan alone and melphalan plus whole-body hyperthermia in dogs with malignant melanoma. Int J Hyperthermia 7:559. Pagnini, U., S. Florio, P. Lombardi, et al. 2000. Modulation of athracycline activity in canine mammary tumour cells in vitro by medroxyprogesterone acetate. Res Vet Sci 69:255–262. Panciera, D.L., O.I. Lanz, and D.M. Vail. 2004. Treating thyroid and parathyroid neoplasia in dogs and cats. Vet Med 99:154. Paoloni, M.C., D.G. Penninck, and A.S. Moore. 2002. Ultrasonographic and clinicopathological findings in 21 dogs with intestinal adenocarcinoma. Vet Radiol Ultrasound 43:562–567. Peterson, M.E., P.P. Kintzer, J.R. Hurley, et al. 1989. Radioactive iodine treatment of a functional thyroid carcinoma producing hyperthyroidism in a dog. J Vet Intern Med 3:20. Pirkey-Ehrhart, N., S.J. Withrow, R.C. Straw, et al. 1995. Primary rib tumors in 54 dogs. J Am Anim Hsop Assoc 31:65–69. Pisarev, M.A., M.A. Dagrosa, L. Thomasz, et al. 2006. Boron neutron capture therapy applied to undifferentiated thyroid carcinoma. Medicina (B Aires) 66:569–573. Poirier, V.J., K.E. Burgess, W.M. Adams, et al. 2004. Toxicity, dosage, and efficacy of vinorelbine (Navelbine) in dogs with spontaneous neoplasia. J Vet Intern Med 18:536–539. Poirier, V.J., L.J. Forrest, W.M. Adams, et al. 2004. Piroxicam, mitoxantrone, and coarse fractionation radiotherapy for the treatment of transitional cell carcinoma of the bladder in 10 dogs: A pilot study. J Am An Hosp Assoc 40:131–136. Pooya, H.A., B. Seguin, D.R. Mason, et al. 2004. Biomechanical comparison of cortical radial graft versus ulnar transposition graft limb-sparing techniques for the distal radial site in dogs. Vet Surg 33:301–308.

Multimodal Therapy  31 Popovitch, C.A., M.J. Weinstein, M.H. Goldschmidt, et al. 1994. Chondrosarcoma: A retrospective study of 97 dogs (1987–1990). J Am Anim Hosp Assoc 30:81–85. Post, G.S. and G.N. Mauldin. 1992. Radiation and adjuvant chemotherapy for the treatment of thyroid adenocarcinoma in dogs. Proceedings Veterinary Cancer Society, 12th Annual Conference, Pacific Grove, CA, Oct. 18–21 (abstract). Columbia, MO: Veterinary Cancer Society. Posterino, N.C., R.J. Berg, B.E. Powers, et al. 1988. Prognostic variables for canine haemangiopericytoma: 50 cases (1979–1984). J Am Anim Hosp Assoc 24:501–509. Postorino, N.C., S.J. Susaneck, S.J. Withrow, et al. 1989. Single agent therapy with Adriamycin for canine lymphosarcoma. J Am Anim Hosp Assoc 25:221–225. Proulx, D.R., D.M. Ruslander, R.K. Dodge, et al. 2003. A retrospective analysis of 140 dogs with oral melanoma treated with external beam radiation. Vet Radiol Ultrasound 44:352. Prymak, C., I.J. McKee, M.H. Goldschmidt, et al. 1988. Epidemiologic, clinical, pathologic, and prognostic characteristics of splenic haemangiosarcoma and splenic haematoma in dogs: 217 cases (1985). J Am Vet Med Assoc 193:706–712. Quintin-Colonna, F., P. Devauchelle, D. Fradelizi, et al. 1996. Gene therapy of spontaneous canine melanoma and feline fibrosarcoma by intratumoral administration of histoincompatible cells expressing human interleukin-2. Gene Ther 3:1104–1112. Ramirez, O., R.K. Dodge, R.L. Page, et al. 1999. Palliative radiotherapy of appendicaulr osteosarcoma in 95 dogs. Vet Radiol Ultrasound 40:517–522. Ramos-Vara, J.A., M.E. Beissenherz, M.A. Miller, et al. 2000. Retrospective study of 338 canine oral melanomas with clinical, histologic, and immunohistochemical review of 129 cases. Vet Pathol 37:597. Rassnick, K.M., A.S. Moore, L.E. Williams, et al. 1999. Treatment of canine mast cell tumors with CCNU (lomustine). J Vet Int Med 13:601–605. Rassnick, K.M., D.M. Ruslander, S.M. Cotter, et al. 2001. Use of carboplatin for treatment of dogs with malignant melanoma: 27 cases (1989–2000). J Am Vet Med Assoc 218:1444–1448. Ridgway, T.D. and M.D. Lucroy. 2003. Phototoxic effects of 635-nm light on canine transitional cell carcinoma cells incubated with 5-aminolevulinic acid. Am J Vet Res 64:131–136. Robben, J.H., H.A. Vissner-Wisselaar, G.R. Rutteman, et al. 1997. In vitro and in vivo detection of functional somatostatin receptors in canine insulinomas. J Nucl Med 38:1036. Rosales, C., K.A. Jeglum, M. Obracka, et al. 1988. Cytolytic activity of murine anti-dog lymphoma monoclonal antibodies with canine effector cells and complement. Cell Immunol 115:420– 428. Ross, J.T., T.C. Scavelli, and D.T. Matthiesen. 1991. Adenocarcinoma of the apocrine glands of the anal sac in dogs: A review of 32 cases. J Am Anim Hosp Assoc 27:349–355. Rovesti, G.L., M. Bascucci, K. Schmidt, et al. 2002. Limb-sparing using a double bone-transport technique for treatment of a distal tibial osteosarcoma in a dog. Vet Surg 31:70–77. Rusbridge, C., S.J. Wheeler, C.R. Lamb, et al. 1999. Vertebral plasma cell tumors in 8 dogs. J Vet Intern Med 13:126–133. Saulnier-Troff, F.G., V. Busoni, and A. Hamaide. 2008. A technique for resection of invasive tumous involving the trigone area of the bladder in dogs: Preliminary results in two dogs. Vet Surg 37:427–437. Schwarz, P.D., S.J. Withrow, C.R. Curtis, et al. 1991a. Partial maxillary resection as a treatment for oral cancer in 61 dogs. J Am Anim Hosp Assoc 27:617–624.

Schwarz, P.D., S.J. Withrow, C.R. Curtis, et al. 1991b. Mandibular resection as a treatment for oral cancer in 81 dogs. J Am Anim Hosp Assoc 27:601–610. Seguin, B., N.F. Leibman, V.S. Bregazzi, et al. 2001. Clinical outcome of dogs with grade II mast cell tumors treated with surgery alone: 55 cases (1996–1999). J Am Vet Med Assoc 218:1120–1123. Seguin, B., P.J. Walsh, D.R. Mason, et al. 2003. Use of an ipsilateral vascularised ulna transposition autograft for limb-sparing surgery of the distal radius in dogs: An anatomic and clinical study. Vet Surg 32:69–79. Selting, K.A., B.E. Powers, L.J. Thomson, et al. 2000. Outcome of dogs with high-grade soft-tissue sarcomas treated with and without adjuvant doxorubicin chemotherapy: 39 cases (1996–2004). J Am Vet Med Assoc 227:1442–1448. Seo, K.W., U.S. Choi, Y.C. Jung, et al. 2007. Palliative intravenous cisplatin treatment for concurrent peritoneal and pleural mesothelioma in a dog. J Vet Med Sci 69:201–204. Simon, D., J.W. Knebel, W. Baumgartner, et al. 2001. In vitro efficacy of chemotherapeutics as determined by 50% inhibitory concentrations in cell cultures of mammary tumors obtained from dogs. Am J Vet Res 62:1825–1830. Simpson, A.M., L.L. Ludwig, S.J. Newman, et al. 2004. Evaluation of surgical margins required for complete excision of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 224:236–240. Skorupski, K.A., C.A. Clifford, M.C. Paoloni, et al. 2003. CCNU for the treatment of dogs with metastatic or disseminated histiocytic sarcoma. Vet Cancer Soc Proc 23:36. Slaweinski, M.J., G.E. Mauldin, G.N. Mauldin, et al. 1997. Malignant colonic neoplasia in cats: 46 cases (1990–1996). J Am Vet Med Assoc 211:878–881. Smith, A.N., J.C. Wright, W.R. Brawner Jr, et al. 2001. Radiation therapy in the treatment of canine and feline thymomas: A retrospective study (1986–1999). J Am Anim Hosp Assoc 37:489–496. Sorenmo, K., L. Duda, L. Barber, et al. 2000. Canine haemangiosarcoma treated with standard chemotherapy and monocycline J Vet Intern Med 14:395–398. Sorenmo, K.U., J.L. Baez, C.A. Clifford et al. 2004. Efficacy and toxicity of a dose-intensified doxorubicin protocol in canine haemangiosarcoma. J Vet Int Med 18:209–213. Sorenmo, K.U., M.H. Goldschmidt, F.S. Shofer, et al. 2004. Evaluation of cyclooxygenase-1 and cyclooxygenase-2 expression and the effect of cyclooxygenase inhibitors in canine prostatic carcinoma. Vet Comp Oncol 2:13–23. Sorenmo, K.U., K.A. Jeglum, and S.C. Hefland. 1993. Chemotherapy of canine splenic haemangiosarcoma with doxorubicin and cyclophosphamide J Vet Intern Med 7:370–376. Sorenmo, K.U., M. Samluk, C.A. Clifford, et al. 2004. Efficacy and toxicity of intracavitary administration of pegylated liposomal encapsulated doxorubicin (Doxil) in dogs with haemangiosarcoma. Proceedings of the Veterinary Cancer Society, Kansas City, MO. Sorenmo, K.U., F.S. Shofer, and M.H. Goldschmidt. 2000. Effects of spaying and timing of spaying on survival of dogs with mammary carcinoma. J Vet Int Med 14:226–270. Spangler, W.L. and M.R. Culbertson. 1992. Prevalence, type, and importance of splenic diseases in dogs: 1,480 cases (1985–1989). J Am Vet Med Assoc 200:829–834. Spangler, W.L. and P.H. Kass. 1997. Pathologic factors affecting postsplenectomy survival in dogs. J Vet Intern Med 11:166–171. Sparkes, A., S. Murphy, F. McConnell, et al. 2005. Palliative intracavitary carcboplatin therapy in a cat with suspected pleural mesothelioma. J Feline Med Surg 7:313–316.

32  Veterinary Surgical Oncology Spugnini, E.P., S. Crispi, A. Scarabello, et al. 2008. Piroxicam and intracavitary platinum-based chemotherapy for the treatment of advanced mesothelioma in pets: Preliminary observations. J Exp Clin Cancer Res 19(27):6. Stanclift, R.M. and S.D. Gilson. 2008. Evaluation of neoadjuvant prednisone administration and surgical excision in treatment of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 232:53–56. Stepien, R.L., N.T. Whitley, and R.R. Dubielzig. 2000. Idiopathic or mesothelioma-related pericardial effusion: Clinical findings and survival in 17 dogs studied retrospectively. J Small Anim Pract 41:342–347. Steplewski, Z., C. Rosales, K.A. Jeglum, et al. 1990. In vivo destruction of canine lymphoma with adjuvant canine monoclonal antibodies. In Vivo 4:231–234. Stone, M.S., M.A. Goldstein, and S.M. Cotter. 1991. Comparison of two protocols for induction of remission in dogs with lymphoma. J Am Anim Hosp Assoc 27:315–321. Straw, R.C., R.A. LeCouteur, B.E. Powers et al. 1989. Multilobular osteochondrosarcoma of the canine skull: 16 cases (1978–1988). J Am Vet Med Assoc 195:1764–1749. Straw, R.C., S.J. Withrow, E.B. Douple, et al. 1994. The effects of cisdiamminedichloroplatinum II released from D,L,-polylactic acid implanted adjacent to cortical allografts in dogs. J Orthop Res 12:871–877. Straw, R.C., S.J. Withrow, and B.E. Powers. 1991. Primary Osteosarcoma of the ulna in 12 dogs. J Am Anim Hosp Assoc 27:323–326. Straw, R.C., S.J. Withrow, and B.E. Powers. 1992. Partial or total hemipelvectomy in the management of sarcomas in seven dogs and two cats. Vet Surg 21:183–188. Sun, F., J. Hernandez, J. Ezquerra, et al. 2002. Angiographic study and therapeutic embolization of soft-tissue fibrosarcoma in a dog: Case report and literature. J Am Anim Hosp Assoc 38(5):452–457. Thamm, D.H., E.A. Mauldin, and D.M. Vail. 1999. Prednisolone and vinblastine chemotherapy for canine mast cell tumor: 41 cases (1992–1998). J Vet Int Med 13:491–497. Thamm, D.H., M.M. Turek, and D.M. Vail. 2006. Outcome and prognostic factors following adjuvant prednisolone/vinblastine chemotherapy for high-risk canine mast cell tumor: 61 cases. J Vet Med Sci 68:581–587. Theon, A.P., P.Y. Barthez, B.R. Madewell, et al. 1994. Radiation therapy of ceruminous gland carcinoma in dogs and cats. J Am Vet Med Assoc 205:566–569. Theon, A.P, B.R. Madewell, M.F. Harb, et al. 1993. Megavoltage irradiation of neoplasms of the nasal and paranasal cavities in 77 dogs. J Am Vet Med Assoc 202:1469–1475. Theon, A.P., B.R. Madewell, A.S. Moore, et al. 1991. Localised thermocisplatin therapy: A pilot study in spontaneous canine and feline tumors. Int J Hyperthermia 7:881–892. Theon, A.P., S.L. Marks, E.S. Feldman, et al. 2000. Prognostic factors and patterns of treatment failure in dogs with unresectable differentiated thyroid carcinomas treated with megavoltage irradiation. J Am Vet Med Assoc 216:1775. Theon, A.P., C. Rodriguez, and B.R. Madewell. 1997. Analysis of prognostic factors and patters of failure in dogs with malignant oral tumors treated with megavoltage irradiation. J Am Vet Med Assoc 210:778. Thompson, J.P., N. Ackerman, J.R. Bellah , et al. 1992. 192Iridium brachytherapy, using an intracavitary afterload device, for treatment of intranasal neoplasms in dogs. Am J Vet Res 53:617–622. Thompson, J.P. and M.J. Fugent. 1992. Evaluation of survival times after limba amputation, with and without subsequent administration of cisplatin, for treatment of osteosarcoma in dogs: 30 cases (1979–1990). J Am Vet Med Assoc 200:531–533.

Thomson, M.J., S.J. Withrow, W.S. Dernell, et al. 1999. Intermuscular lipomas of the thigh region in dogs: 11 cases. J Am Anim Hosp Assoc 35:165–167. Thrall, D.E. 1981. Orthovoltage radiotherapy of oral fibrosarcomas in dogs. J Am Vet Med Assoc 172:159–162. Thrall, D.E., S.J. Withrow, B.E. Powers, et al. 1990. Radiotherapy prior to cortical allograft limb-sparing in dogs with osteosarcoma: A dose response assay. Int J Rad Onc Biol Phys 18:1354–1357. Tilmant, L.L., N.T. Gorman, N. Ackerman, et al. 1986. Chemotherapy of synovial cell sarcoma in a dog. J Am Vet Med Assoc 188:530–532. Tobin, R.L., R.W. Nelson, M.D. Lucroy, et al. 1999. Outcome of surgical versus medical treatment of dogs with beta cell neoplasia: 39 cases (1990–1997). J Am Vet Med Assoc 215:226. Tomamassini, M., N. Ehrhart, A. Ferretti, et al. 2000. Bone transport osteogenesis for limb salvage following resection of primary bone tumors: Experience with 6 cases (1991–1996). Vet Comp Orthop Trauma 23:43–51. Trout, N.J., M.M. Pavletic, and K.H. Kraus. 1995. Partial scapulectomy for the management of sarcoma in three dogs and two cats. J Am Vet Med Assoc 207:585–587. Turek, M.M., L.J. Forrest, W.M. Adams, et al. 2003. Postoperative radiotherapy and mitoxantrone for anal sac adenocarcinoma in the dog: 15 cases (1991–2001). Vet Comp Oncol 1:94–104. Turrel, J.M. 1987a. Intraoperative radiotherapy of carcinoma of the prostate gland in ten dogs. J Am Vet Med Assoc 190:48–52. Turrel, J.M. 1987b. Principles of radiation therapy. In Veterinary Cancer Medicine. GH Thielen and BR Madewell editors. Philadelphia: Lea and Febiger. Turrel, J.M., B.E. Kitchell, L.M. Miller, et al. 1988. Prognostic factors for radiation treatment of mast cell tumors in 85 dogs. J Am Vet Med Assoc 226:1368–1374. Turrel, J.M., M.C. McEntee, B.P. Burke, et al. 2006. Sodium iodide I 131 treatment of dogs with nonresectable thyroid tumors: 39 cases (1990–2003). J Am Vet Med Assoc 15:229:542–548. Vail, D.M. 2007. Plasma cell tumors. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 769–784. S.J. Withrow and D.M. Vail, editors. St Louis: Saunders. Vail, D.M., A.E. Hershey, and A.S. Moore. 2000. Efficacy of adjuvant inhalational chemotherapy following surgical cytoreduction in dogs with spontaneously arising node positive primary lung cancer. Proceedings Annual Meeting of American Association for Cancer Research, San Francisco, (abstract 1272). Philadephia: American Association for Cancer Research. Vail, D.M., E.G. MacEwen, I.D. Kurzman, et al. 1995. Liposomeencapsulated muramyl tripeptide phosphatidylethanolamine adjuvant immunotherapy for splenic haemangiosarcoma in the dog: A randomised multi-institutional clinical trial. Clin Cancer Res 1:1165–1170. Vail, D.M., B.E. Powers, D.M. Getzy, et al. 1994. Evaluation of prognostic factors for dogs with synovial cell sarcoma: 36 cases (1986– 1991). J Am Vet Med Assoc 205:1300–1307. Vail, D.M. and K.M. Young. 2007. Canine lymphoma and lymphoid leukaemia. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 699–733. S.J. Withrow and D.M. Vail, editors. St Louis: Saunders. Vail, D.M. and S.J. Withrow. 2007. Tumors of the skin and subcutaneous tissues. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 375–401. S.J. Withrow and D.M. Vail, editors. St Louis: Saunders. Valerius, K.D., G.K. Ogilvie, C.H. Mallinckrodt, et al. 1997. Doxorubicin alone or in combination with asparaginase, followed by cyclophosphamide, vincristine, and prednisolone for treatment of

Multimodal Therapy  33 multicentric lymphoma in dogs: 121 cases (1987–1995). J Am Vet Med Assoc 210:512–516. Van Ginkel, R.J., H.J. Hoekstra, F.J. Meutstege, et al. 1995. Hyperthermic isolated regional perfusion with cisplatin in the local treatment of spontaneous canine osteosarcoma: Assessment of short-term effects. J Surg Oncol 59:169–176. Vasseur, P. 1987. Limb preservation in dogs with primary bone tumors. Vet Clin North Am 17:889–993. Walker, M. and M. Breider. 1987. Intraoperative radiotherapy of canine bladder cancer. Vet Radiol 28:200–204. Wallace, J., D.T. Matthiesen, and A.K. Patnaik. 1992. Hemimaxillectomy for the treatment of oral tumors in 69 dogs. Vet Surg 21: 337. Walter, W.S., S.M. Dernell, S.M. LaRue, et al. 2005. Curative-intent radiation therapy as a treatment modality for appendicular and axial osteosarcoma: A preliminary retrospective evaluation of 14 dogs with the disease. Vet Comp Oncol 3:1–7. Waltman, S.S., B. Seguin, B.J. Cooper, et al. 2007. Clinical outcome of nonnasal chondrosarcoma in dogs: Thirty-one cases (1986–2003). Vet Surg 36:266–71. Ward, H., L.E. Fox, M.B. Calderwood-Mays, et al. 1994. Cutaneous haemangiosarcoma in 25 dogs: A retrospective study. J Vet Intern Med 8:345–348. Weisse, C., A. Berent, K. Todd, et al. 2006. Evaluation of palliative stenting for management of malignant urethral obstructions in dogs. J Am Vet Med Assoc 29:226–234. Weisse, C., C.A. Clifford, D. Holt, et al. 2002. Percutaneous arterial embolization and chemoembolization for treatment of benign and malignant tumors in three dogs and a goat. J Am Vet Med Assoc 221(10):1430–1436. Wells, A.L., C.D. Long, W.J. Hornof, et al. 2001. Use of percutaneous ethanol injection for treatment of bilateral hyperplastic thyroid nodules in cats. J Am Vet Med Assoc 218:1293.

White, R. 1991. Mandibulectomy and maxillectomy in the dog: Longterm survival in 100 cases. J Small Anim Pract 32:69–74. White, R., M. Walker, A.M. Legendre, et al. 1990. Development of brachytherapy technique for nasal tumors in dogs. Am J Vet Res 51:1250–1256. Willard, M.D., H. Tvedten, R. Walshaw, et al. 1980. Thymoma in a cat. J Am Vet Med Assoc 176:451–453. Williams, L.E., J.M. Gliatto, R.K. Dodge, et al. 2003. Carcinoma of the apocrine glands of the anal sac in dogs: 113 cases (1985–1995). J Am Vet Med Assoc 223:825–831. Williams, L.E., J.L. Johnson, M.L. Hauck, et al. 2004. Chemotherapy followed by half-body radiation therapy for canine lymphoma. J Vet Int Med 18:703–709. Withrow, S.J. 1982. Cryosurgical therapy for nasal tumors in the dog. J Am An Hosp Assoc 18:585–589. Withrow, S.J., E.L. Gillette, P.J. Hoopes, et al. 1989. Intraoperative irradiation of 16 spontaneously occurring canine neoplasms. Vet Surg 18:7–11. Withrow, S.J., J. Liptak, and R.C. Straw. 2004. Biodegradable cisplatin polymer in limb-sparing surgery for canine osteosarcoma. Ann Surg Oncol 11:705–713. Withrow, S.J., D.E. Thrall, R.C. Straw, et al. 1993. Intra-arterial cisplatin with or without radiation in limb sparing for canine osteosarcoma. Cancer 71:2484–2490. Worth, A.J., R.M. Zuber, and M. Hocking. 2005. Radioiodide (131I) therapy for the treatment of canine thyroid carcinoma. Aust Vet J 83:208–214. Yamagami, T., T. Kobayashi, K. Takahashi, et al. 1996. Prognosis for canine malignant mammary tumors based on TNM and histologic classification. J Vet Med Sci 58:1079–1083. Zenman, B.I., A.S. Moore, W.M. Rand, et al. 1998. A combination chemotherapy protocol (VELCAP-L) for dogs with lymphoma. J Vet Int Med 12:465–470.

3 Interventional oncology William T.N. Culp

Interventional radiology (IR) is a specialty that uses different imaging modalities to direct minimally invasive diagnostic and therapeutic procedures. IR has become a well-established and integral speciality in human medicine and is rapidly growing in veterinary medicine. The influx of IR techniques in veterinary medicine allows veterinary clinicians the ability to offer patients advanced treatment options that were previously unavailable. Interventional oncology (IO) is a subspecialty of IR that is focused on the treatment of oncologic disease. When performing IO procedures, it is essential for the veterinary clinician to have a firm grasp of different imaging modalities and basic surgical procedures, as surgically approaching blood vessels is often neces­sary. IO procedures such as vascular stenting, intra­arterial chemotherapy, and transarterial embolization/ chemoembolization are performed intravascularly, and specialized sheaths, guidewires, and catheters are needed for these interventions. Nonvascular diseases such as malignant obstructions and effusions can also be treated with IO techniques and involve the placement of stents and long-term catheters. Many of the current applications of IO in veterinary patients are palliative; in these cases, the primary goal is to improve quality of life while causing minimal morbidity. IO can also provide treatment options in cases that were previously considered untreatable. Reports on the use of IO in veterinary patients are limited, but investigation of IO applications in human medicine offers insight into the vast benefits that this expanding specialty can offer for our veterinary patients. A systematic discussion of the imaging, instrumentation, and techniques involved in IO will be discussed below.

Imaging A complete knowledge of the vascular anatomy is mandatory for performing vascular interventions. Additionally, the interventional radiologist should have a thorough understanding of the imaging modalities and contrast agents that are used to perform IO procedures. While imaging modalities such as fluoroscopy, computed tomography, and magnetic resonance imaging are commonly employed by veterinary clinicians, the use of these modalities for IO treatments is largely unreported, aside from isolated case reports and small case series. Modalities Stenting procedures can be performed solely with digital radiography, although fluoroscopy is superior as it allows for real-time evaluation of the anatomy. Fluoroscopy is mandatory when performing IO procedures that require vascular interventions. A fluoroscopy unit (Carm) with specifications including digital subtraction, road-mapping ability, collimation, and low patient radiation dosing are ideal. Ceiling mounting should be pursued when possible, and the C-arm should have the ability to acquire complex oblique views. Newer units allow the interventional radiologist to perform image acquisition and most other C-arm operations at the bedside, eliminating the need for an assistant to perform these tasks in a control room. While angiography performed with fluoroscopic guidance allows for excellent evaluation of the direction and velocity of blood flow, the images obtained are in two-dimensional planes and only display the lumen of the vessel (Green and Parker 2003). Computed tomographic angiography (CTA) and magnetic resonance

Veterinary Surgical Oncology, First Edition. Edited by Simon T. Kudnig, Bernard Séguin. © 2012 by John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.

35

36  Veterinary Surgical Oncology

angiography (MRA) are rapidly developing imaging modalities that have certain advantages over fluoroscopy, including noninvasive angiographic image acquisition, less patient postprocedure discomfort, and volumetric and cross-sectional image analysis (Green and Parker 2003; Hellinger and Rubin 2006; Thornton and Grist 2006). The volumetric and cross-sectional image analysis that is obtained with CTA and MRA allows vessels to be evaluated in multiple directions with a single scan, whereas several images and injections of contrast are necessary to gain the same information using fluoroscopy (Hellinger and Rubin 2006). Advances are being made that allow CTA and MRA to be performed simultaneously with interventional techniques, which may result in more efficient and accurate IO procedures in the future (Ladd et al. 2000; Hellinger and Rubin 2006; Thornton and Grist 2006; Kos et al. 2008). Ultrasound has many applications in IR. Many of the disease processes that may require IO treatments can be diagnosed by ultrasonography. In dogs, hepatocellular carcinoma, a tumor commonly treated by chemoembolization and radiofrequency ablation in humans (Okusaka et al. 2009; Hiraoka et al. 2010), is easily identified by abdominal ultrasound (Liptak, Dernell et al. 2004). In one study of dogs with hepatocellular carcinoma, diagnosis of a hepatic mass was made with ultrasound in 93.5% of cases (Liptak, Dernell et al. 2004). Other tumors that may require an IO treatment, such as urethral, colonic, and thyroid neoplasia, can also be evaluated by ultrasound (Hume et al. 2006; Weisse et al. 2006; Barber 2007). In a recent case series of dogs with masses obstructing hepatic venous outflow, masses were often identified by ultrasonography (Schlisksup et al. 2009). In addition to the diagnostic utility of ultrasound, this modality is regularly employed in humans to aid in obtaining vascular access (to initiate the Seldinger technique) during performance of IR procedures (Longo et al. 1994; Dodd et al. 1996; Ahmad et al. 2008; Arthurs et al. 2008). Ultrasound can also be used to perform procedures, including the placement of drainage catheters, percutaneous biopsies, percutaneous ethanol injection of hepatic neoplasia, and radiofrequency ablation (Longo et al. 1994; Dodd et al. 1996; Solbiati 1998). Contrast agents Contrast agents are required components of most intravascular procedures and many stenting procedures. The predominant contrast agents used for angiography include iodinated agents, gadolinium-based agents, and carbon dioxide (Ehrmann et al. 1994; Moresco et al. 2000; Spinosa et al. 2000; Spinosa et al. 2001; Brown et al. 2003; Namasivayam et al. 2006; Bui et al. 2007). Iodinated contrast agents are available in both ionic and nonionic forms; nonionic agents are less osmolar than

their ionic counterparts (Singh and Daftary 2008). Severe reactions are reported to occur with similar incidence among all iodinated contrast agents, but mild and moderate contrast reactions occur more commonly with the use of higher osmolality iodinated contrast agents (Singh and Daftary 2008). Nephrotoxicity is a major potential complication associated with the use of iodinated contrast agents and has become the third most common cause of acute renal failure in humans (Akgun et al. 2006). The most commonly used iodinated contrast agents are the nonionic monomers such as iohexol, iopromide, iopamidol, and ioversol (Dickinson and Kam 2008). Gadolinium-based contrast agents and CO2 are used most commonly in patients who have had a previous adverse reaction to an iodinated contrast agent and in those patients with an increased risk for development of nephrotoxicity (Moresco et al. 2000; Spinosa et al. 2000; Spinosa et al. 2001; Dickinson and Kam 2008), although some recent studies have reported nephrotoxicity in association with gadolinium contrast usage (Akgun et al. 2006; Ergün et al. 2006). Agents such as gadopentetate dimeglumine, gadodiamide, gadoteridol, and gadoversetamide are the most readily available gadolinium-based contrast agents (Akgun et al. 2006) and are used when previous CO2 usage has resulted in a suboptimal study due to bowel gas artifacts or as a supplement to CO2 angiography (Spinosa et al. 2000; Spinosa et al. 2001). Gadolinium-based contrast agents produce less detailed contrast studies as compared with iodinated agents and are therefore less useful for angiography during IR procedures (Spinosa et al. 2000). When using gadolinium-based contrast agents, digital subtraction angiography is recommended to compensate for the less detailed study that is otherwise obtained (Spinosa et al. 2000). To outline a hollow viscus such as the esophagus, urethra, and colon, substances such as barium and iodinated contrast agents have been used (Hume et al. 2006; Weisse et al. 2006; Farese et al. 2008). In a recent study of esophageal tumors in dogs, barium sulfate was found to be useful in identifying mass location (Farese et al. 2008). In dogs, iodinated contrast agents have been used to evaluate urethral obstructions prior to urethral stenting (Weisse et al. 2006). Additionally, an iodinated contrast agent was used prior to colonic stenting to delineate colonic obstructions secondary to adenocarcinoma in cats (Hume et al. 2006).

Instrumentation and Implants Access needles Traditional hypodermic needles or over-the-needle catheters (Figure 3.1) can be used to puncture vessels

Interventional Oncology  37

(a)

(b)

(c)

(d)

Figure 3.1.  Interventional oncology instrumentation. From left to right: (A) 18-gauge over-the-needle catheter (left), 22-gauge overthe-needle catheter (right). (B) 0.035-inch hydrophilic guidewire. (C) Dilator and vascular access sheath. (D) Catheter with angled-tip.

when obtaining vascular access using the Seldinger technique (Seldinger 1953). The size of the access needle used determines the wire size that can be introduced through the needle and into the vessel. The standard venous access needle is an 18-gauge needle, which accepts guidewires up to 0.38 inches in diameter (Braun 1997). A 21-gauge needle is considered to be a micropuncture needle and allows for introduction of guidewires up to 0.018 inches in diameter (Braun 1997; Valji 2006). Guidewires Selection of a particular guidewire (Figure 3.1) is dictated by the size of access needle that has been placed, the technique to be performed, and the vessel(s) to be selected. Most guidewires are available in three standard lengths: 150 cm, 180 cm, and 260 cm (Braun 1997). Alternative lengths of 60 cm, 125 cm, and 145 cm have been reported, but these are not readily available (Valji 2006; Kipling et al. 2009). The standard diameters of most guidewires are 0.035 and 0.038 inches. Smaller gauge wires of 0.014 and 0.018 inches are used when microcatheters and smaller (micropuncture) vascular

access needles are used (Braun 1997; Valji 2006; Kipling et al. 2009). There are a few primary principles that must be adhered to when using guidewires. First, most guidewires contain a hydrophilic coating made of polytetrafluoroethylene that needs to be primed with saline to allow for smooth passage through the lumen that has been selected (Braun 1997; Kipling et al. 2009). When sufficiently wet, the guidewire should pass easily through a catheter and allow an increased ability to perform vascular selection (Braun 1997; Kipling et al. 2009). It is essential that the guidewire remain wet during the procedure to improve guidewire function (Kipling et al. 2009). Second, the length of the selected guidewire should be at least twice the length of the catheter that is being used (Braun 1997). Third, if a guidewire is not passing easily through a vascular access needle, the needle may need to be repositioned. The wire should not be forced as the needle may be subintimal or against a sidewall (Valji 2006). Lastly, a torque device can be placed on the end of a guidewire (approximately 5–10 cm from a catheter hub that has been introduced over the guidewire) to better manipulate and steer the

38  Veterinary Surgical Oncology

guidewire (Kipling et al. 2009). These torque devices can be invaluable when passing a guidewire into vessels that are difficult to access and when crossing stenotic regions. Guidewires are also used for nonvascular stenting procedures (Hume et al. 2006; Weisse et al. 2006; Culp et al. 2007; Kipling et al. 2009). Stents that are placed through malignant obstructions are introduced over a guidewire, and the stent delivery system tapers down to the guidewire to allow for easier placement. In companion animals, 0.035-inch hydrophilic guidewires have been used to facilitate stent placement for tracheal, urethral, and colonic obstructions (Hume et al. 2006; Weisse et al. 2006; Culp et al. 2007). Sheaths The use of intravascular sheaths (Figure 3.1) is indicated when a procedure involves multiple exchanges into and out of a vessel. The sheath protects the vessel wall from damage and allows for easier passage of different catheter types. Additionally, sheaths protect the vessel from stiff intravascular devices, balloon catheters, and intravascular foreign bodies (Braun 1997; Valji 2006; Stavropoulos et al. 2006). A dilator that tapers down to a guidewire is usually present within a sheath and allows expansion of the previously made hole in the blood vessel. Sheaths contain a valve that prevents blood leakage while allowing entrance of specialized catheters, wires, stents, snares, and biopsy forceps (Snow and O’Connell 2000; Stavropoulos et al. 2006). Additionally, sheaths contain a sidearm that allows for injection of contrast, which can pass around wires and nonocclusive catheters (Snow and O’Connell 2000; Valji 2006). The French gauge of a sheath is determined by the largest gauge catheter that can fit through the sheath and represents the inner diameter of the sheath (Braun 1997; Snow and O’Connell 2000). The French size of a sheath is generally considered to be 2 French gauges smaller than the outer diameter (Valji 2006). In human medicine, sheaths have also been used for nonvascular procedures such as antegrade ureteric stenting, percutaneous transhepatic biliary drainage, and colonic stenting (Braun 1997; Snow and O’Connell 2000). The use of sheaths in urethral stenting has been described in veterinary medicine (Weisse et al. 2006; Newman et al. 2009). Some sheaths are designed with a peel-away component that is used during the placement of venous access devices and drainage catheters (Braun 1997). The peel-away component allows a device to be inserted through the sheath, and the sheath can then be removed while leaving the device in place.

Catheters When performing intravascular procedures, there are many catheter types that are available to the interventional radiologist. Decisions about which catheter is most optimal for a specific procedure is based on experience and the anatomy of the vessel that is to be selected. In human medicine, catheters with an outer diameter of 5 French (1 French = 3 mm) (Silberstein et al. 1992; Valji 2006) are chosen most commonly. (Braun 1997). This catheter size has the advantage of having good torque control and high flow rates when contrast is injected (Braun 1997). A larger catheter (6–7 French) affords the user increased control of torque (Wojtowycz 1990a; Braun 1997). Commonly used catheters are made of nylon, Teflon, or polyurethane (Wojtowycz 1990a; Valji 2006). Catheters are often categorized by the shape of the tip, with the basic catheter tip shapes being straight, pigtail, hook and angled (Braun 1997; Valji 2006). Straight catheters are commonly used for embolotherapy (delivery of a vascular occlusion agent) and for opacification of a vascular tree; however, care should be taken when injecting through a straight catheter with a single end-on hole as injury to the vessel is possible (Braun 1997). Pigtail catheters are superior for opacification, as they allow for the injection of a bolus of contrast through several small holes in the catheter, thus preventing the jet effect that may be seen with straight catheters (Wojtowycz 1990a; Braun 1997; Valji 2006). Pigtail catheters should be removed over a guidewire so that the catheter is straight during removal and therefore less likely to cause vascular damage (Wojtowycz 1990a). Hook catheters, such as the shepherd’s hook and cobra catheters, are used to catheterize vessels that have acute angled branches (Braun 1997). These visceral catheters are advanced over a guidewire to the desired location. The catheters are then allowed to reform (take on the original shape of the catheter) in a large vessel (generally aorta or vena cava) and are then gently pulled back into the lumen of the vessel selected (Braun 1997). Angled-tip catheters (Figure 3.1) are used in the selection of upwardbranching vessels; as with other catheters, a guidewire is often used to facilitate vessel selection and to maintain position once it has been established (Braun 1997). The Berenstein catheter is an example of an angled-tip catheter (Braun 1997). Similar to sheaths and guidewires, catheters can be used for nonvascular techniques. When stenting luminal obstructions, such as in the urethra or trachea, marker catheters are employed to determine appropriate stent size (Hume et al. 2006; Weisse et al. 2006; Culp et al. 2007; Newman et al. 2009). Another nonvascular

Interventional Oncology  39

oncologic use includes palliation of malignant thoracic and abdominal effusions by placing catheters for percutaneous drainage. This has been described in several human studies (Brooks and Herzog 2006; Fleming et al. 2009). Pigtail catheters are often used for draining malignant effusions since they have multiple fenestrations; a locking loop mechanism that maintains the catheter position in the desired location is also present on some pigtail catheters. When performing IO techniques, there are certain techniques and principles of catheter usage that are important to understand, namely the coaxial technique, the so-called rule of 110, and superselective catheterization with guidewire-microcatheter combination. As the size of a vessel that is being targeted for catheterization decreases, the catheter diameter must also decrease. Placement of a catheter that is too large can result in damage to the vessel wall, blood stasis with subsequent tissue ischemia, difficulty in removing the catheter, and vessel rupture (Stavropoulos et al. 2006). The coaxial technique is commonly employed to introduce progressively smaller catheters and wires to allow for superselection of vessels while minimizing the risk for development of complications. Simply stated, this technique involves placement of a smaller wire or catheter (depending on the stage of the procedure) into a catheter that is already within the blood vessel. The smaller wires and catheters share the same axis as the indwelling catheter, thus the term coaxial is used. This technique has been described in human patients undergoing selection of small vessels (Korogi et al. 1995; Tajima et al. 2008). A guide for vessel selection (particularly celiac and superior mesenteric arteries in humans) has been described and termed the “rule of 110” (Chuang 1981; Nemcek 1996). According to Chuang (1981), this rule states that “both the length and width of the catheter tip should be about 110% of the width of the aorta at the level of the artery (that is to be selected) as it branches”. The width refers to the distance between the tip of the catheter and the proximal straight limb of the catheter (Chuang 1981). If this technique is employed with a catheter of appropriate size and tip, the tip of the catheter should engage the branching vessel (i.e., celiac, cranial mesenteric, or renal artery in dogs and cats) as the catheter is pulled caudally and should subsequently enter the chosen vessel. A guidewire can be combined with a microcatheter (using the coaxial technique) to allow for superselective catheterization. To accomplish this, the guidewiremicrocatheter combination is passed through a guiding catheter to a vessel branch that is a few orders less than the desired branch (Stavropoulos et al. 2006). The guidewire-microcatheter combination is then advanced

toward the desired branch, and the guidewire is used to select the desired branch. The catheter is then immediately but gently advanced over the guidewire and into the vessel to prevent the guidewire from backing out. This should be performed in a series of steps that are slow and calculated (Stavropoulos et al. 2006). The guidewire can then be advanced further into the vessel or into a smaller branching vessel as needed. Vascular closure devices A vascular closure device is an instrument used to close the hole in a blood vessel that remains after removal of a vascular access sheath. The use of arteriotomy closure devices is well documented in human interventional cardiologic medicine; however, the use of these devices in veterinary medicine is uncommon (Meyerson et al. 2002; Koreny et al. 2004; Nikolsky et al. 2004). Historically, manual compression combined with bed rest was used to control bleeding at an arteriotomy site (Hoffer and Bloch et al. 2003). Concerns over the high rates of bleeding at these high-pressure sites spurred the development and use of devices that can be used to close vessels after a vascular IO procedure has been performed (Dauerman et al. 2007). Benefits seen with the use of closure devices have included decreased time to hemostasis and earlier ambulation (Meyerson et al. 2002; Tron et al. 2002; Hoffer and Bloch 2003; Dauerman et al. 2007). Closure devices are not universally used in human cases, however, and some studies have suggested similar or higher complication rates associated with the use of closure devices as compared to manual compression (Meyerson et al. 2002; Nikolsky et al. 2004; Dauerman et al. 2007). Vascular closure devices can be divided into two categories: passive and active (Silber 1998; Meyerson et al. 2002; Tron et al. 2002; Dauerman et al. 2007). Passive closure devices assist or enhance manual compression and do not provide immediate (75% tumor necrosis had significantly lower recurrence rates at 1 year (15%) versus dogs with 5 years), and no sex predisposition is observed (Murphy et al. 2004). Breeds reported to have a high incidence of cutaneous MCTs include boxers, Boston terriers, golden retrievers, Labrador retrievers, beagles, and schnauzers (Murphy et al. 2004; Gieger et al. 2003; Hahn et al. 2008; Hahn et al. 2004). The majority (65%–80%) of cutaneous MCTs are solitary. Approximately 50%–60% of canine cutaneous MCTs occur on the trunk, with 25% occurring on the limbs and the remainder on the head and neck areas. All MCTs are locally aggressive, but low- and intermediategrade MCTs have a lower metastatic potential than highgrade MCTs. The gross appearance of cutaneous MCTs is very variable, ranging from raised hairless masses to aggressive, invasive ulcerated lesions (Figure 4.4). Paraneoplastic effects can result from release of inflammatory mediators contained within the MCT cytoplasmic granules. Histamine release from cutaneous MCT can cause local effects of edema and erythema formation (Darier’s sign) that is associated with the release of vasoactive substances from MCTs (Figure 4.5). Wound healing can also be delayed due to the release of proteases. Histamine release can also cause gastrointestinal ulceration by stimulating H2 receptors on parietal cells of the stomach, which produce hydrogen chloride. Nonsteroidal anti-inflammatory drugs (NSAIDs) are contraindicated in the management of MCTs because of the increased potential for gastrointestinal ulceration. Diagnosis of MCTs Fine-needle aspiration cytology is sufficient to confirm the diagnosis of cutaneous MCTs in approximately 90% of cases but does not provide information on the tumor grade. Cytologically, MCT have large round cells with central nuclei and abundant cytoplasm. The cytoplasm contains blue to purple granules that stain with toluidine blue (Figure 4.6). Other inflammatory cells such as eosinophils and neutrophils are frequently seen mixed with the MCT cells on cytologic examination. An incisional biopsy is required to provide sufficient tissue to determine the histologic grade of cutaneous MCTs. An incisional biopsy to establish MCT grade is indicated if negative prognostic factors are present or the surgical site is not amenable to wide surgical

62  Veterinary Surgical Oncology

(b)

(a)

(c)

(e)

(d)

Figure 4.3.  Surgical en bloc resection of cutaneous tumor with planned reconstructive skin fold transposition flap. (A) Preoperative skin marking of proposed en bloc tumor excision and axial skin fold transposition flap. (B) En bloc tumor excision. (C) En bloc tumor excision completed with deep margins. (D) Transposition flap raised to close tumor resection site. (E) Completed tumor resection and reconstructive procedure closure. Images courtesy of Dr. Julius Liptak.

resection (e.g., distal extremity) to determine if a more conservative marginal excision is appropriate (e.g., for a low-grade tumor) or if or more radical surgery (e.g., amputation) or other therapies (e.g., radiation and/or chemotherapy) are indicated for high-grade MCTs. Drugs used for the treatment of an acute anaphylactic reaction, including histamine antagonists, corticosteroids, and epinephrine, should be available when performing an incisional biopsy of a known MST.

Clinical staging A modified version of the WHO clinical staging scheme is used to stage canine cutaneous MCTs into one of four stages (Table 4.2) (Turrel et al. 1988). Preoperative staging is important to determine prognosis and appropriate treatment options, including surgical dose, radiation, and chemotherapy. A minimum database of a complete blood count, serum biochemistry,

Skin and Subcutaneous Tumors  63

(a)

(b)

Figure 4.4.  Range of cutaneous mast cell tumor appearances. (A) Ulcerated grade III MCT. (B) Grade II MCT on muzzle.

Figure 4.5.  Darier’s sign secondary to vasoactive substance release.

Figure 4.6.  Cytological appearance of canine cutaneous MCT.

and urinalysis is indicated as part of a presurgical workup for any patient with cancer. The typical path of metastatic spread is initially to the regional lymph node and then to distant sites of the spleen, liver, or bone marrow. Complete staging for cutaneous MCT includes FNA cytology of the regional lymph node, abdominal ultrasound, and thoracic radiographs. FNA cytologic examination of the regional lymph node should be done in all cases of cytologically confirmed cutaneous MCT, regardless of whether the lymph node is enlarged or not, to detect early metastasis (Langenbach et al. 2001). Dogs with metastasis to the regional lymph node are classified as clinical stage II tumors and have a poor prognosis (Krick et al. 2009). Lymph node aspirates from normal dogs can contain mast cells (Bookbinder et al. 1992), so diagnosis of lymph node involvement based on FNA cytology should be if greater than 3% of the cell population are mast cells

(Dobson et al. 2004). No standardized cytologic criteria exist for differentiating reactive and metastatic MCT in lymph nodes. A recent study described cytologic criteria for metastatic mast cell disease in lymph nodes and found that dogs with stage II disease had a significantly shorter survival time than dogs with stage I disease independent of grade and that dogs with grade III primary MCTs were more likely to have stage II disease (Krick et al. 2009). Abdominal ultrasonography is recommended as part of complete staging to assess the liver, spleen, and abdominal lymph nodes for metastasis. The routine aspiration cytology of normal livers and spleens is controversial. One study frequently identified mast cells in normal livers and spleens and concluded that routine aspiration cytology could not be recommended (Finora et al. 2006). Another study found that dogs with cutaneous MCTs that had cytologic evidence of

64  Veterinary Surgical Oncology Table 4.2.  WHO Clinical staging system for canine mast cell tumors. Clinical Stage

Description

0

Single tumor, incompletely excised from dermis Single tumor, confined to dermis without regional lymph node involvement Single tumor, confined to dermis with regional lymph node involvement Multiple dermal tumors or large infiltrating tumors, with or without regional lymph node involvement Any tumor with distant metastases or recurrence with metastases (including blood or bone marrow involvement)

I II III

IV

Substage a: No systemic signs of disease; Substage b: Signs of systemic disease. Adapted from: London, C.A. and B. Seguin. 2003. Mast cell tumors in the dog. Vet Clin N Am Small Anim Pract 33(3):473–489.

mast cell metastases, using criteria of clustering of mast cells and atypical mast cell morphology, had a decreased survival time compared to dogs without cytologic evidence of distant disease (Stefanello et al. 2009). All dogs that had cytologic evidence of mast cell metastasis in the liver or spleen had abnormal ultrasonographic findings. The authors concluded that cytology of the liver and spleen was indicated as part of clinical staging for dogs with cutaneous MCTs regardless of the ultrasonographic appearance of the liver or spleen. Thoracic radiographs are included as part of the staging process to assess for thoracic lymph node enlargement or evidence of concurrent thoracic disease. Cutaneous MCTs rarely metastasize to lungs. The routine use of buffy coat smears and bone marrow evaluation are not advocated for routine staging of cutaneous MCTs due to the rarity of bone marrow involvement and the common finding of mastocytemia in dogs with disease other than cutaneous MCTs (McManus 1999; LaDue et al. 1998; Endicott et al. 2007). The presence of neoplastic mast cell infiltration in the bone marrow is rare and is generally more common in dogs with grade III primary cutaneous tumors (O’Keefe et al. 1987). Reported indications for bone marrow sampling as part of the clinical staging process of dogs with cutaneous MCTs include abnormal hemogram findings or presentation for tumor regrowth, progression, or new occurrence (Endicott et al. 2007).

Preoperative imaging of the cutaneous MCT with ultrasonography, computer tomography, or MRI can facilitate definition of anatomical margins, especially deep margins, prior to surgery and assist with surgical planning if reconstructive procedures are planned. Assessment of size and shape of intracavitary lymph nodes can also be obtained from this imaging. Prognostic factors Histologic parameters Grade The most widely used grading system for canine cutaneous MCTs, developed by Patnaik and colleagues, is based on histomorphologic features, including cellularity, cell morphology, invasiveness, mitotic activity, and stromal reaction and is prognostic for survival (Table 4.3) (Patnaik et al. 1984). Well-differentiated (grade I) MCTs account for 26%–55% of all MCTs; intermediate differentiated (grade II) MCTs account for 25%–59% of MCTs; and poorly differentiated (grade III) MCTs account for 16%–40% of MCTs (Murphy et al. 2006). Tumor grade is the most consistent prognostic indicator for biological behavior and survival time in cutaneous MCTs across multiple studies (Turrel et al. 1988; Patnaik et al. 1984; Thamm et al. 1999). Higher tumor grade is associated with higher risk of metastasis, lower local control rates, and shorter survival times. Grade II MCTs are the most common grade identified and have the widest range of biological behavior compared to the other two grades. Grade III MCTs have an aggressive clinical behavior and poor survival time compared to grade I or II MCTs, with a reported median survival time for dogs with grade III MCTs of 224 days and a metastatic rate of 55%–96% (Hume et al. 2007; Bostock 1986). There is significant variation in grading of MCTs between pathologists despite existence of a well-defined grading scheme. In one study in which 10 veterinary pathologists independently graded the same 60 cutaneous MCTs using the Patnaik grading system, agreement was 62.1% (Northrup et al. 2005). Most variation in classification was between grade I and grade II and grade II and grade III tumors. Surgeons should be aware that variation in histologic grading of MCTs exists when planning appropriate primary or adjuvant therapy. The majority of dogs diagnosed with grade II MCT will have a good prognosis; however, there is a subset of these patients that will develop metastases and have decreased survival time. Recently, mitotic index, argyrophylic nucleolar organizer regions (AgNOR), and Ki67 proliferation markers have been used to help differentiate between grade II MCTs with a poor and good prognosis to guide which patients may require close monitoring

Skin and Subcutaneous Tumors  65 Table 4.3.  Patnaik scheme for grading canine mast cell tumors. Grade

Patnaik Grade

Well differentiated

I

Intermediately differentiated

II

Anaplastic undifferentiated

III

Microscopic features • Well differentiated mast cells with clearly defined cytoplasmic borders with regular spherical or ovoid nuclei • Granules are large, deep staining, and plentiful • Cells confined to the dermis and interfollicular spaces • Cells closely packed with indistinct cytoplasmic boundaries • Nuclear/cytoplasmic ratio lower than anaplastic • Mitotic figures infrequent • More granules than anaplastic • Neoplastic cells infiltrate or replace the lower dermal and subcutaneous tissues • Highly cellular, undifferentiated cytoplasmic boundaries • Irregular size and shape of nuclei • Frequent mitotic figures • Low number of cytoplasmic granules • Neoplastic tissue replaces the subcutaneous and deep tissues

Adapted from: London, C.A. and B. Seguin. 2003. Mast cell tumors in the dog. Vet Clin N Am Small Anim Pract 33(3):473–489.

and possible adjuvant therapy (Maglennon et al. 2008; Romansik et al. 2007; Scase et al. 2006).

Mitotic index (MI) is an indirect measure of cellular proliferation indices and directly correlates with histologic grade; it is a strong predictor for overall survival. Dogs with cutaneous MCTs with a MI of 5 or less had a significantly longer survival time than those with a MI greater than 5, regardless of histologic grade (Romansik et al. 2007).

expression are well established. Increased expression of KIT in the cytoplasm of neoplastic mast cells is a strong indicator of increased risk of local tumor recurrence and a decreased survival time (Kiupel et al. 2004). Histologic panels that employ multiple markers are now available at various veterinary pathology laboratories to facilitate prognostication, especially for grade II MCTs. In particular, identification of KIT mutations from the proliferative panel may be helpful to determine if the patient will respond to tyrosine kinase inhibitor therapy.

Proliferation indices

Size

Indicators of cellular proliferation can provide prognostic information about the likelihood of MCTs recurring locally and help differentiate the prognoses of grade II MCTs (Seguin et al. 2006). The three most commonly used cellular proliferation indices are AgNORs, proliferating cell nuclear antigen (PCNA), and number of Ki67–positive nuclei. Increasing AgNOR and PCNA scores were significantly associated with a shorter progressionfree interval (Gill et al. 2007). A Ki67 index of greater than 1.8% is a significantly prognostic indicator for poorer survival for grade II MCTs (Scase et al. 2006; Maglennon et al. 2008). Ki67 index is also a prognostic factor, independent of grade. The KIT protein is a tyrosine kinase receptor that is a product of the c-kit proto-oncogene. Mutations in c-KIT result in aberrant cytoplasmic expression of KIT in 9%–30% of canine MCTs, with high-grade tumors more likely to have a mutation. Immunohistochemistry protocols for the detection of the KIT receptor

Cutaneous MCT size is predictive for survival time. Dogs with grade III MCT primary tumors greater than 3 cm in maximum diameter have a shorter median survival time than those with tumors less than 3 cm diameter (Hahn et al. 2004).

Mitotic index

Breed Boxers, golden retrievers and Labrador retrievers have been identified as breeds at increased risk of developing multiple cutaneous MCTs (Thamm et al. 1999). Anatomical location MCTs located in the inguinal, perineal, or scrotal regions were previously reported to have a poorer prognosis compared to other cutaneous locations (Turrel et al. 1988). Recent studies have refuted this finding and concluded that when MCTs in these locations are treated appropriately, survival times and tumor-free intervals are equivalent to other cutaneous locations (Cahalane

66  Veterinary Surgical Oncology

et al. 2004; Sfiligoi et al. 2005). A retrospective study of 24 dogs with MCTs located in the muzzle region identified this location as a site for biologically aggressive tumors with higher regional metastatic rates than previously reported for MCTs in other sites (Gieger et al. 2003). Another study found that dogs with tumors located on the extremities had a longer tumor-free interval than dogs with MCT located on the trunk (Turrel et al. 1988). Multiple cutaneous MCT Presentation with multiple cutaneous MCTs occurs in 9–21% cases (Murphy et al. 2004; Murphy et al. 2006; Mullins et al. 2006). Boxers and older dogs are more likely to likely to present with multiple cutaneous MCTs (Kiupel et al. 2005). Multiple cutaneous MCT presentation was originally considered a poor prognostic sign. Recent studies, however have found no difference in survival times for single versus multiple cutaneous MCT (Murphy et al. 2006; Thamm et al. 1999). Dogs with multiple cutaneous MCTs have a low rate of metastasis and a good prognosis for long-term survival with adequate excision of all MCTs (Mullins et al. 2006). Clinical stage The WHO staging system for cutaneous MCT (Table 4.2) has been reported to be prognostic for tumor-free time and survival time with surgical MCT treatment (Turrel et al. 1988). Dogs with clinical stage 0 (i.e., single, incompletely excised local dermal tumor without regional lymph node involvement) have longer tumorfree and survival times than dogs with more extensive disease (Turrel et al. 1988). The presence of lymph node metastasis (stage II disease) carries a poorer prognosis compared to stage I disease (Murphy et al. 2006; Hume et al. 2007; Turrel et al. 1988; Krick et al. 2009). Dogs with grade III primary cutaneous MCTs are more likely to have metastatic MCT in regional lymph nodes (Krick et al. 2009). Dogs with lymph nodes affected by metastatic disease that are treated with either surgery or radiation have prolonged survival time compared to untreated lymph node-positive dogs (Hume et al. 2007). The significance of clinical stage has been questioned as a reliable prognostic indicator as animals with multiple cutaneous tumors are assigned a higher clinical stage, even though these multiple tumors do not necessarily indicate metastatic disease and are not associated with a poorer outcome compared to single cutaneous MCTs (Thamm et al. 1999). Local tumor recurrence Local tumor recurrence after surgical excision is associated with a decreased overall survival time (Seguin et al. 2006).

Treatment of canine cutaneous MCTs The optimal treatment for an individual patient with MCT disease depends on the tumor grade, anatomical site, clinical stage, and surgical and radiation therapy facilities available. Available treatment options for cutaneous MCTs include surgical excision, radiation therapy, and chemotherapy. If MCT disease is confined to the local cutaneous site, surgical excision is the treatment of choice. Surgery Perioperative management Perioperative surgical complications may be encountered related to the release of vasoactive substances from mast cell granules secondary to tumor manipulation. MCTs should therefore not be manipulated extensively during the perioperative period to avoid the risk of a degranulation reaction. Preoperative treatment with H1 blocker (diphenhydramine) and H2 blockers (cimetidine or ranitidine) and corticosteroids is strongly recommended in dogs with cutaneous MCTs that will be surgically manipulated, including biopsy, and those that show evidence of degranulation (Darier’s sign) or melena or hemoptysis associated with gastrointestinal ulcerations secondary to histamine release. Epinephrine should be available in the case of a potential anaphylactic reaction. Hypotension during surgery can be caused by mast cell degranulation and histamine release. Perioperative and intraoperative intravenous fluid therapy is indicated for circulatory support. Invasive or noninvasive blood pressure monitoring is strongly recommended. Coagulation abnormalities can occur locally at the surgical site related to heparin release, causing bleeding at the time of surgery and bruising postoperatively. Delayed wound healing can be observed after MCT excision as a result of proteolytic enzyme release and vasoactive amines from the MCT. Neoadjuvant prednisone treatment may facilitate resection when adequate surgical margins cannot be confidently attained because of mass location or size or both (Stanclift and Gilson 2008; Dobson et al. 2004). Mean reduction in MCT volume was 80.6% in 70% cases treated with neoadjuvant prednisolone. Reduction in tumor size may be related to the anti-inflammatory effect of prednisolone, reducing tumor-related inflammation and edema secondary to tumor cytokine release. There was no difference in response rate between a high dose (2.2 mg/kg) and a low dose (1.0 mg/kg) prednisone protocol. The determination of appropriate surgical margins should be based on tumor dimensions at initial

Skin and Subcutaneous Tumors  67

presentation rather than at post prednisolone treatment tumor size. Margins Surgical margins Wide surgical excision with adequate lateral and deep margins is the treatment of choice for most MCTs. The deep margin in particular needs to be a good-quality margin rather than a quantity margin. Fascia and collagen-dense tissues are good barriers to tumor infiltration. The deep margin should include a fascial plane deep to the tumor that has not been invaded by tumor. This margin should be removed en bloc with the tumor so that the tumor matrix is not encountered during the surgery. The surgical dogma of 3 cm lateral and one fascialplane deep margins for MCTs has been challenged recently, especially for low-grade and smaller-sized MCTs. Simpson et al. (2004) reported that a 2 cm lateral margin and a deep margin of one fascial plane appeared to be adequate for complete excision of grade I and II MCTs in dogs. In fact, a 1 cm lateral margin was able to obtain tumor-free margins in 75% of grade II and 100 of grade I cutaneous MCTs. A 2 cm lateral margin and one deep facial plane excision was successful in completely excising 91% of grade I and II MCTs (Fulcher et al. 2006). A similar local recurrence rate and de novo development rate was observed compared to previous reports with a 3 cm margin. Investigators concluded that excision of grade I and II MCTs with 2 cm margins might minimize complications associated with larger local tumor resection (Fulcher et al. 2006). Wide surgical margins are not a prerequisite for a successful long-term outcome in dogs with well-differentiated cutaneous MCTs (Murphy et al. 2004). Tumor depth has no prognostic significance. (Kiupel et al. 2005) There were no grade III tumors in these studies, so adequate margins for grade III MCTs have as yet not been determined; thus, margins 3 cm lateral and at least one fascial plane deep are recommended. The excised specimen should be submitted in toto and not in sections. The anatomical relationship between the deep fascial plane and lateral margins should be preserved with sutures to help orient the pathologist, and the deep and lateral margins should be inked (ideally with separate colors). Anatomical site considerations Options for distal extremity MCT Appropriate therapy for cutaneous MCTs located on an extremity is dictated by tumor grade. For low- and intermediate-grade MCTs, a combination of a marginal

surgical resection with planned external beam radiation therapy is a rational treatment option. Amputation may be indicated for grade III MCTs to achieve wide surgical margins. Palliative radiation therapy (4 × 8 Gy weekly) in combination with prednisolone has been reported to be useful in the management of measurable MCTs located on a distal extremity (Dobson et al. 2004). Treatment of high-grade MCTs Patients with histologically confirmed high-grade MCTs should have complete staging tests performed before surgical intervention. These includes FNA cytology of the mass, incisional biopsy, regional lymph node cytology (regardless of size), and abdominal ultrasound. Postoperative recommendations Completely excised MCTs Grade I or II MCTs excised with complete surgical margins do not require any adjuvant therapy as the risk for local recurrence (5%–11%) or metastasis is relatively low (Murphy et al. 2004; Seguin et al. 2001; Weisse et al. 2002; Michels et al. 2002). Patients should be evaluated regularly for signs of local recurrence and any new cutaneous masses should be thoroughly investigated. Grade III MCTs that are completely excised have a low chance for local recurrence but a high chance to develop metastatic disease. As such, these cases should receive adjunctive chemotherapy to delay or prevent metastatic spread. Incompletely excised MCTs Incompletely excised grade I or II MCTs have a low chance of local recurrence and low chance of metastatic spread (Seguin et al. 2006). The surgeon has several recommended treatment options in the case of incompletely excised grade I or II MCTs, including monitoring, additional surgery, and adjuvant chemotherapy or radiation therapy. Murphy et al. concluded that dogs with well-differentiated, incompletely excised tumors that did not receive adjuvant treatment did as well as those that did have additional therapy (Murphy et al. 2004). Seguin et al. found that dogs developing local recurrence had shorter survival times than dogs without local recurrence with incompletely excised grade II MCTs (Seguin et al. 2006). Another study showed histopathological tumor-free versus non-tumor-free margins were not associated with a different frequency of tumorrelated death; however, significantly more dogs in the non-tumor-free margin group relapsed by 12 and 24 months postoperatively compared to the tumor-free margin group (Michels et al. 2002).

68  Veterinary Surgical Oncology

The preferred treatment for incompletely excised MCTs, if possible, is excision of the surgical scar with a larger margin of normal tissue at least one fascial plane deep. If the anatomical location does not permit extensive resection, a more conservative re-resection and/or adjuvant radiation therapy is indicated. Incompletely excised grade III MCTs have a high chance of both local recurrence and metastatic spread. These cases should receive additional local therapy (either additional surgery or radiation therapy) as well as chemotherapy (Hahn et al. 2004). Radiation therapy for MCT Radiation therapy can be used as part of multimodal therapy either as an adjunctive therapy after incomplete surgical MCT excision or as a primary treatment modality of the primary tumor and/or regional lymph node(s). Radiation is most effective and most commonly used as an adjunctive therapy after surgical removal or tumor size reduction to a microscopic level (LaDue et al. 1998). When possible, surgery should be performed prior to radiation therapy to decrease the tumor volume, as dogs treated with adjuvant radiation therapy with smaller tumor volumes have longer disease-free intervals than those with larger tumor volumes (LaDue et al. 1998; Hahn et al. 2004). The risk of systemic effects as a result of MCT degranulation is present in tumors treated with radiation of macroscopic disease, so pretreatment with prednisolone is recommended (Dobson et al. 2004). Radiation therapy is very effective at eliminating residual microscopic disease after incomplete excision of grade I and II MCTs. Local tumor control rates of 86%– 94% at 2 and 3 years are reported after adjunctive radiation therapy for incompletely excised grade II MCTs (Frimberger et al. 1997; al-Sarraf et al. 1996; Poirier et al. 2006). These rates are similar to the local tumor control rate of 89% achieved with complete surgical excision of grade II MCTs (Weisse et al. 2002). Radiation therapy for incompletely excised grade III, stage 0 MCTs, is encouraging compared to untreated incompletely excised grade III MCTs with a 1-year local control rate of 65% and 1-year survival rate of 71% (Hahn et al. 2004). If the regional lymph node is positive for MCT disease (stage II disease), radiation therapy of the affected lymph node has been advocated by some investigators to improve survival time. A study by Poirier et. al. found no difference in overall survival rate, whether the regional lymph node was prophylactically irradiated or not (Poirier et al. 2006). Palliative radiation using 4 × 8 Gy fractions at weekly intervals has been reported as a treatment option for

unresectable MCTs alone or in combination with chemotherapy (Dobson et al. 2004). Chemotherapy for MCT The use of adjuvant chemotherapy is indicated after resection of grade III MCTs because of their high metastatic rate and for metastatic and unresectable MCTs. The use of chemotherapy after incomplete resection of grade I and II MCTs is not indicated based on their lower metastatic potential. Some of the chemotherapeutic agents that have reported activity against canine MCTs are prednisolone, vinblastine, CCNU (Lomustine), vinorelbine, and chlorambucil. The response rate of macroscopic cutaneous MCTs to oral prednisolone as a single agent is reported to be 20% (McCaw et al. 1994). Most of these responses were partial responses with remission times between 10 and 20 weeks. The reported response rates for single-agent vinblastine range from 12% to 27% (Rassnick et al. 2008). Vinblastine is commonly administered to dogs at a dosage of 2.0 mg/m2. The dose can be escalated to 3.0 to 4.0 mg/m2 with neutropenia as the dose-limiting toxicity (Vickery et al. 2008; Bailey et al. 2008). CCNU (Lomustine) is an antitumor alkylating agent in the nitrosourea family. Lomustine is administered orally at a dose of 50–90 mg/m2 every 21 days. A response rate of 47% for measurable cutaneous MCTs treated with single-agent CCNU (90 mg/m2) is reported (Rassnick et al. 1999). Acute toxicities include neutropenia and thrombocytopenia. CCNU can cause a delayed, cumulative dose-related, chronic hepatotoxicity that is irreversible and can be fatal (Kristal et al. 2004). Combination chemotherapy Combination chemotherapy protocols using prednisone or prednisolone and vinblastine (Davies et al. 2004; Thamm et al. 1999; Thamm et al. 2006), CCNU and vinblastine (Cooper et al. 2009), and CCNU and prednisone (Hosoya et al. 2009) have been reported as adjuvant chemotherapy for macroscopic and microscopic cutaneous MCT disease. The rationale for combination therapy protocols is to increase local recurrence-free intervals, metastasis-free intervals, and survival times over single agent protocols. Adjuvant therapy such as prednisone and vinblastine is best employed after the initial tumor resection, rather than at the time of recurrence (Thamm et al. 1999). Deionized or hypotonic water Conflicting results have been reported on the efficacy of deionized water as an adjunctive therapy after surgical

Skin and Subcutaneous Tumors  69

excision of cutaneous mast cell tumors (Brocks et al. 2008; Grier et al. 1995; Jaffe et al. 2000). A prospective, placebo-controlled, double-blinded and randomized clinical trial found that hypotonic water does not decrease the rate of local recurrence in dogs with solitary MCT after marginal surgical excision (Brocks et al. 2008). Tyrosine kinase inhibitors In dogs, 20%–30% of MCTs express a mutated form of KIT, a receptor tyrosine kinase involved in the development or progression of MCT growth and differentiation. Small-molecule tyrosine kinase inhibitors including imatinib mesylate (Gleevec), masitinib, and toceranib have shown efficacy against canine MCTs (London et al. 2009; Isotani et al. 2008; Hahn et al. 2008). Toceranib phosphate (Palladia) has recently been licensed for use in veterinary medicine for treatment of mast cell disease. The tumor should be positive for c-KIT mutations for Palladia to be potentially effective. A multicenter, placebo-controlled, randomized study recently demonstrated a 42% response rate to Palladia in dogs with grade II or III cutaneous MCTs (London et al. 2009). Ulcers of the stomach and intestine have been a common side effect of this medication (London et al. 2009). Feline cutaneous MCT Cutaneous MCT is the second most common feline skin tumor, after basal cell tumor (Miller et al. 1991). Feline cutaneous MCTs are most commonly located on the head and neck, followed by the trunk and extremities (Litster and Sorenmo 2006). Feline MCTs located on the head are less biologically active than in dogs. An increased breed incidence in Siamese cats for cutaneous MCT is reported compared to other breeds (Miller et al. 1991). Feline cutaneous MCTs have a benign biological behavior compared to canine cutaneous MCTs. The Patnaik histopathological grading scheme used for canine cutaneous MCTs is not prognostic in cats (MolanderMcCrary et al. 1998; Lepri et al. 2003). There are two forms of feline cutaneous mast cell disease, mastocytic and the less common histiocytic. The histiocytic form occurs in cats younger than 4 years old and is usually characterized by multiple nonpruritic, firm, hairless, pink subcutaneous nodules. Histiocytic MCTs generally regress spontaneously. Histologic classifications of feline cutaneous MCTs are well differentiated, poorly differentiated, or histiocytic. High mitotic activity (>4 mitoses/high-powered field) is reported as a negative prognostic indicator for feline cutaneous MCT (Lepri et al. 2003; Johnson et al. 2002).

Cats with cutaneous MCTs should be staged with an abdominal ultrasound to evaluate the spleen for evidence of MCTs that may be metastasizing to the cutaneous location. The prognosis for feline cutaneous MCT with surgical resection is good, with a 16%–36% local recurrence rate. Incomplete surgical excision is not associated with a higher rate of tumor recurrence in cats with cutaneous MCTs (Molander-McCrary et al. 1998; Litster and Sorenmo 2006). Radiation therapy using strontium-90 has recently been reported as an effective treatment for feline cutaneous MCT (Turrel et al. 2006). A distinct visceral form of mast cell tumor, which affects the spleen without cutaneous involvement, exists in cats and carries a poor prognosis (Litster and Sorenmo 2006). Systemic signs of chronic vomiting, anorexia, and weight loss can be associated with this form of MCT disease. Splenectomy is the recommended treatment for the visceral MCT form if disease is located in the spleen.

Mesenchymal Tumors and Melanoma Introduction This section will first evaluate the management of soft tissue sarcomas (STSs) and describe general adjunctive therapies. Specific types of STSs will then be discussed with tumor-specific treatment options. Soft tissue sarcomas Soft tissue sarcomas (STSs) are a heterogenous group of tumors that originate from connective tissues surrounding, supporting and bridging anatomical structures or tissues. STSs have similar biological behaviors, often displaying both benign and malignant characteristics. Although skin and subcutaneous tumors are the most commonly observed STSs, these sarcomas can, in principle, arise from any part of the body (Ehrhart 2005; Ettinger 2003; Kuntz et al. 1997). In general, STSs are slow-growing and locally invasive tumors, comprised mainly of spindle-shaped cells, with a low tendency for metastatic spread. The group of STSs includes wellknown tumor types such as fibrosarcoma, peripheral nerve sheath tumors (PNSTs), hemangiopericytoma, liposarcoma, myxosarcoma, and undifferentiated sarcomas (Ehrhart 2005; Gaitero et al. 2008; Liptak 2007). STSs are grouped together because histologic classification and differentiation is often complicated. The nomenclature follows classification of human STSs, based on patterns of cellular proliferation and individual cell morphology without conclusive identification of the cells of origin and is poorly standardized for animals. Some pathologists therefore prefer the term spindle cell tumors of canine soft tissue (McColl Williamson and

70  Veterinary Surgical Oncology

Middleton 1998). Further differentiation of histologic diagnosis is reached using immunohistochemistry (Gaitero et al. 2008; Ettinger et al. 2006). The discussion pertaining to the exact histologic differentiation is not one of major clinical importance because the overall biological behavior of the STS is similar. Several important features of biological behavior that are common to all STSs include the following:

• STSs form a pseudocapsule, with the tumor cells infil• • • •

trating through the outer borders of the capsule into surrounding tissues. Local recurrence after conservative surgical excision is common. STSs metastasize in up to 20% of all cases, mainly via a hematogenous route. Macroscopic or bulky (>5 cm in diameter) STSs respond poorly to chemotherapy and radiation therapy. Histopathological grade is predictive of metastasis, and resected tumor margins predict local recurrence (Liptak et al. 2007).

Prognosis of soft tissue sarcomas The prognosis of STSs depends on tumor size, histologic grade, site, fixation to underlying structures, presence of metastasis, and completeness of removal (i.e., the surgical margins) (Ettinger 2003; Kuntz et al. 1997). The incidence of metastases in STSs is low (16 months) was significantly longer compared to incomplete excisions (9 months) (Davidson et al. 1997). Forrest et al. (2000) treated hemangiopericytoma, fibrosarcoma, and other STSs with radiation therapy after tumors were excised to microscopic disease, with a dose that ranged from 42 to 57 Gy given in 3 to 4.2 Gy daily fractions on a Monday through Friday schedule. Median time to local recurrence was more than 798 days. STSs tumors at oral sites had a statistically significant lower median survival (540 days) as compared to other tumor sites (2,270 days). Lawrence et al. (2008) reported that coarsely frac­ tionated radiation therapy may be a reasonable pallia­ tive option for the management of canine STSs. The

74  Veterinary Surgical Oncology

treatment protocol used single parallel opposed fields with a 3 cm margin surrounding the palpable edge of the tumor, if possible, using a cobalt teletherapy unit. CT scans were used, when available, to help estimate field size and depth of treatment, but true image-based computer planning was not performed. The total dose of radiation applied to the tumor was 32 Gy to the isocenter, delivered as one 8 Gy fraction on days 0, 7, 14, and 21. The overall objective response rate was 50% and included seven partial and one complete response (a cutaneous hemangiosarcoma on the left ventral thorax of a dog that was also treated with chemotherapy). The median progression-free interval was 155 days, with a range of 72–460 days. Radiotherapy in combination with chemotherapy can be used for STSs that have metastasized or have a high risk of metastases, such as high-grade STS, feline vaccineassociated sarcoma, and oral melanoma. Acute side effects of radiation therapy on the skin include moist desquamation and alopecia. Late effects of radiation therapy on the skin include fibrosis, contraction, nonhealing ulcer, and leukotrichia. The higher the dose per fraction, the higher the probability of late effects (Forrest et al. 2000; McEntee 2006; Moore 2002). Chemotherapy STSs are a heterogeneous group of tumors. Because of this heterogeneity, it is hard to obtain sufficient data to set up a solid treatment protocol based on adequately proven clinical trials. The effectiveness of chemotherapy as an adjuvant therapy after resection of STSs is thus unclear. Chemotherapy may be beneficial in cases of metastasis, incomplete resection of high-grade tumors, and tumors not treatable with surgery or radiation therapy. Several chemotherapy protocols are used either as single agent or in combination. Metastases are uncommon in STSs, however, and are reported to be 15% for low-grade malignant to 41% for high-grade malignant types of cutaneous STSs. Single-agent doxorubicin, mitoxantrone, or combination protocols using vincristine, doxorubicin, and cyclophosphamide have been reported to be effective for STS (Thornton 2008). Elmslie et al. (2008) treated 30 dogs after incomplete removal of soft tissue sarcoma with continuously lowdose cyclophosphamide (10 mg/m2) and standard-dose piroxicam (0.3 mg/kg) therapy. Disease-free interval (DFI) was 410 days and 211 days, respectively, for all STSs at all sites (trunk, extremities) in treated dogs compared with 55 untreated controls. Although the median DFI was not reached for the treated dogs, it seems that the DFI was significantly prolonged (Elmslie et al. 2008; Kuntz et al. 1997; Rassnick 2003; Schlieman et al. 2006).

Immunotherapy The expression of genes encoding for immunostimulatory cytokines or tumor-associated antigens that may negatively influence tumor viability are being used more frequently. Several attenuated poxvirus vector systems have been developed, for example NYVAC (Copenhagen vaccinia virus), TROVAC (Fowl pox virus), and ALVAC (Canary pox virus) viral vectors (Paoletti et al. 1995). These recombinant viruses have been administered without any major side effects to animals and humans (Fries et al. 1996). To prevent the recurrence of fibrosarcoma, ALVAC- or NYVAC-based recombinants expressing feline or human IL2, respectively, were administered to domestic cats. In the absence of immunotherapy, recurrence was observed in 61% of animals within a 12-month follow-up period after treatment with surgery and iridium-based radiotherapy. Only 39% of the cats receiving NYVAC-human interleukin-2 (IL-2) and 28% of the cats receiving ALVAC-feline IL-2 exhibited tumor recurrences (Jourdier et al. 2003). Additionally, intratumoral administration of histoincompatible cells expressing human IL-2 in spontaneous canine melanoma and feline fibrosarcoma, in combination with surgery and radiotherapy, has been shown to increase the diseasefree period and survival time (Quintin-Colonna et al. 1996). Angiogenesis plays an essential role in tumor growth, invasion, and metastasis. Vascular endothelial cell growth factor (VEGF) is one of the key growth factors regulating the process of angiogenesis. Kamstock et al. (2007) evaluated the effect of xenogeneic VEGF vaccination in dogs with cutaneous STS. A total of six immunizations with a human VEGF vaccine were administered intradermally to dogs once every other week for three immunizations, then once every 4 weeks for three additional immunizations. Eventually four out of nine dogs remained long enough in the study to receive five or more immunizations. The five dogs that failed to receive greater than three immunizations were removed from the study due to progressive tumor growth. A decrease in plasma VEGF concentration was observed in three of the four dogs that received five or more VEGF immunizations. Tumor microvessel density (MVD) was evaluated in biopsy specimens on week 6 and 16 of the study from these four dogs. Two of the four multiplyvaccinated dogs demonstrated a significant (>50%) decrease in tumor MVD at one or more time points. It should be noted that in one of these dogs, tumor MVD increased at a later time point coincident with progressive growth. In the other two dogs, tumor MVD remained relatively constant after immunization of the tumor. Based on these results, it appeared that repeated VEGF

Skin and Subcutaneous Tumors  75

immunization was capable of inhibiting tumor angiogenesis in at least half of the dogs. At this point, more research is needed to fully understand the VEGF pathway and when and how to use it. Palliative procedures Surgery, radiotherapy, or chemotherapy (doxorubicin/ cyclophosphamide) may be used for palliation, to slow the progression of disease, and to alleviate presumed discomfort (Lawrence et al. 2008). Plavec et al. treated 15 dogs with unresectable STSs (2006). Total tumor radiation dose was 24 Gy, given in three 8 Gy fractions on days 0, 7, 21, or weekly. Tumor responses in 15 dogs included one partial remission (liposarcoma), 13 tumors with stable, disease and one progressive disease; median time to progression and median survival time were 263 and 332 days, respectively. None of the treated dogs developed serious complications, even though brachial plexus (once) and bones were in the radiation field. The only side effects of radiation therapy were slowed hair growth rate or change of the color of growing hair. It is important to note, however, and to communicate with the owner, that palliative care cannot prolong life for years, but rather aims to improve the quality of life for several weeks to months (Plavec et al. 2006). Specific soft tissue sarcomas Fibrosarcomas Fibrosarcomas (FSAs) are tumors derived from mesenchymal cells or fibroblasts. FSAs infiltrate surrounding tissues, are locally aggressive, and metastasize hematogenously to distant sites including the lungs, liver, bone, brain, and skin (Powers et al. 1995). Tumor grade is an important determinant in the histologic assessment of soft tissue-origin FSA. Grade I or II FSAs of the skin are unlikely to metastasize. Aggressive and complete surgical excision is the treatment of choice for these FSAs, and long-term control or cure is likely with aggressive surgery with or without radiation therapy. A grade III, or highgrade, FSA is more likely to metastasize, and adjuvant chemotherapy is warranted (Dernell et al. 1998; Davis et al. 2007). Radiation therapy or chemotherapy may also be indicated in the case of unresectable FSA. For microscopic local residual disease, radiation therapy seems to be a highly effective treatment option (Chun 2005; Forrest et al. 2000; Little and Goldschmidt 2007; Mikaelian and Gross 2002). In contrast to FSAs of the skin, FSAs originating from the oral cavity generally behave in a more malignant way and carry a poor prognosis due to an invasive growth pattern and frequent inability of complete removal.

Peripheral nerve sheath tumors Peripheral nerve sheath tumors (PNSTs) include a variety of neoplasms including (malignant) schwannoma, neurofibroma, and neurofibrosarcoma. Reported incidence is 0.5%–2% of all skin tumors in dogs (Goldschmidt and Shofer 1992; Kaldrymidou et al. 2002; Pakhrin et al. 2007), although reported data vary considerably, depending on varying classification of these tumors. They are locally aggressive and metastasize rarely (1,460 days) than cats that did not undergo surgery (60 days). In a study of 53 cats, subcutaneous tumors were associated with longer survival than visceral tumors, and cutaneous tumors were associated with longer survival than subcutaneous tumors. Completely excised tumors were associated with longer survival than incompletely excised tumors, and cats with incompletely excised tumors had longer survival times than those for which surgical resection was not attempted (Johannes et al. 2007; McAbee et al. 2005). In general, surgical excision is the therapy of choice in cats and dogs with cutaneous HSA. Chemotherapy

Skin and Subcutaneous Tumors  77

may be a viable adjunctive therapy, especially for incompletely resected tumors. Five dogs with subcutaneous HSA treated with surgery, doxorubicin, and cyclophosphamide had a median survival time of 211 days (Sorenmo et al. 1993). Seven dogs treated with surgery, vincristine, doxorubicin, and cyclophosphamide had a median survival of 425 days (Hammer et al. 1991). A more recent study of 17 dogs with subcutaneous HSA and 4 dogs with intramuscular HSA, receiving adequate local control and doxorubicin chemotherapy, reported a median DFI and survival of 1,553 and 1,189 days for subcutaneous HSA and 266 and 273 days for intramuscular HSA, respectively. Younger age (16 versus 9 months) than those with incomplete excisions (Davidson et al. 1997). Marginal excision is the major reason that high recurrence rates (up to 70%) have been reported (Davidson et al. 1997; Hershey et al. 2000). Advanced imaging of the tumor is recommended (McEntee and Samii 2000; Morrison and Star 2001) for

appropriate treatment planning. Median time to recurrence was significantly longer when a cat was operated by a specialist surgeon (274 days) compared to a referring veterinarian (66 days) (Hershey et al. 2000). Other prognostic factors for survival time that are significant include local recurrence, presence of distant metastasis, and the number of surgeries (Cohen et al. 2001; Eckstein et al. 2009; Romanelli et al. 2008). The most important prognostic factor for local recurrence, and subsequent survival time, is the achievement of clean surgical margins (Banerji and Kanjilal 2006; Cronin et al. 1998; Hershey et al. 2000; Kobayashi et al. 2002). Cats undergoing limb amputation for FISAS did better than local excision anywhere else on the body (Hershey et al. 2000). Size of the tumor has been reported to influence survival time after surgery (Cohen et al. 2001; Dillon et al. 2005; Spugnini et al. 2007). To achieve wide tumor resection, resection of the dorsal portion of interscapular vertebral spinous processes, partial scapulectomy (Trout et al. 1995), lateral body wall resection (Lidbetter et al. 2002), and hemipelvectomy (Straw et al. 1992) may be necessary (Davidson et al. 1997; Davis et al. 2007; Hershey et al. 2000; Romanelli et al. 2008). Some surgeons promote using wider surgical margins than the commonly recommended 2–3 cm lateral margins with one tissue plane in depth, because of the high recurrence rate of FISAS. Recently, 57 cats with FISASs were treated by wide resection using 4–5 cm lateral margins and one fascial plane deep to the tumor, including partial scapulectomy and removal of dorsal spinal processes if indicated. Histologically complete resections were reported for 95% of the tumors; 5% had tumor cells in the margins. Local tumor recurrence developed in 39%, with distant metastasis in 21%. Fiftyone percent of the cats were alive at an overall median follow-up period of 366 days (median follow-up period for the alive cats was 600 days; Romanelli et al. 2008). A recent study of 99 cats with FISASs treated by widemargin resection, using 5 cm lateral margins and two fascial planes beneath the tumor, reported 15% recurrence and 18% distant metastasis. Adjuvant radiation therapy may improve outcome. Median DFIs after complete resection combined with radiation therapy were 405–1,110 days. Survival times after complete resection combined with radiation therapy were 476–1,290 days. DFIs after resection with contaminated margins combined with radiation therapy were 112–600 days. Median survival times for contaminated margins combined with radiotherapy were 502– 900 days (Cohen et al. 2001; Cronin et al. 1998; Eckstein et al. 2009). According to a recent study, radiation therapy of residual microscopic tumor improved median DFI and survival time (20 and 30 months, respectively)

80  Veterinary Surgical Oncology

compared to residual macroscopic tumor (4 and 7 months, respectively; Eckstein et al. 2009). In a study of 76 cats with vaccine-associated sarcomas, 26 cats were treated with chemotherapy in addition to surgery and radiotherapy. Neither recurrence rates, rate of metastasis, nor survival times were improved in the chemotherapy group. These results suggest that the benefit of chemotherapy is limited in the treatment of vaccine-associated sarcomas (Cohen et al. 2001).

References Aitken, M.L. and A.K. Patnaik. 2000. Comparison of needle-core (Trucut) biopsy and surgical biopsy for the diagnosis of cutaneous and subcutaneous masses: a prospective study of 51 cases (November 1997–August 1998). J Am Anim Hosp Assoc 36(2):153–157. Alberdein, D., J.S. Munday, C.B. Dyer, et al. 2007. Comparison of the histology and immunohistochemistry of vaccination-site and nonvaccination-site sarcomas from cats in New Zealand. NZ Vet J 55(5):203–207. al-Sarraf, R., G.N. Mauldin, A.K. Patnaik, et al. 1996. A prospective study of radiation therapy for the treatment of grade 2 mast cell tumors in 32 dogs. J Vet Intern Med 10(6):376–378. Anderson, G.M., A. Dallaire, L.M Miller, et al. 1999. Peripheral nerve sheath tumor of the diaphragm with osseous differentiation in a one-year-old dog. J Am Anim Hosp Assoc 35(4):319–322. Avallone, G., P. Helmbold, M. Caniatti, et al. 2007. The spectrum of canine cutaneous perivascular wall tumors: Morphologic, phenotypic and clinical characterization. Vet Pathol 44(5): 607–620. Bacon, N.J., W.S. Dernell, N. Ehrhart, et al. 2007. Evaluation of primary re-excision after recent inadequate resection of soft tissue sarcomas in dogs: 41 cases (1999–2004). J Am Vet Med Assoc 230(4):548–554. Baez, J.L., M.J. Hendrick, F.S. Shofer, et al. 2004. Liposarcomas in dogs: 56 cases (1989–2000). J Am Vet Med Assoc 224(6):887–891. Bagley, R.S., S.J. Wheeler, L. Klopp, et al. 1998. Clinical features of trigeminal nerve-sheath tumor in 10 dogs. J Am Anim Hosp Assoc 34(1):19–25. Baker-Gabb, M., G.B. Hunt, and M.P. France. 2003. Soft tissue sarcomas and mast cell tumours in dogs: Clinical behaviour and response to surgery. Aust Vet J 81(12):732–738 Bailey, D.B., K.M. Rassnick, O. Kristal, et al. 2008. Phase I dose escalation of single-agent vinblastine in dogs. J Vet Intern Med 22(6):1397–1402. Baldi, A. and E.P. Spugnini. 2006. Thoracic haemangiopericytoma in a cat. Vet Rec 159(18):598–600. Banerji, N. and S. Kanjilal. 2006. Somatic alterations of the p53 tumor suppressor gene in vaccine-associated feline sarcoma. Am J Vet Res 67(10):1766–1772 Ben-Amotz, R., G.W. Ellison, M.S. Thompson, et al. 2007. Pericardial lipoma in a geriatric dog with an incidentally discovered thoracic mass. J Small Anim Pract 48(10):596–599. Bergman, P.J. 2007. Canine oral melanoma. Clin Tech Small Anim Pract 22(2):55–60. Bergman, P.J., M.A. Camps-Palau, J.A. McKnight, et al. 2006. Development of a xenogeneic DNA vaccine program for canine malignant melanoma at the Animal Medical Center. Vaccine 24(21): 4582–4585. Bergman, P.J., J. McKnight, A. Novosad, et al. 2003. Long-term survival of dogs with advanced malignant melanoma after DNA vac-

cination with xenogeneic human tyrosinase: A phase I trial. Clin Cancer Res 9(4):1284–1290. Bergman, P.J., S.J. Withrow, R.C. Straw, et al. 1994. Infiltrative lipoma in dogs: 16 cases (1981–1992). J Am Vet Med Assoc 205(2): 322–324. Bernstein, J.A., E.C. Hodgin, H.W. Holloway, et al. 2006. Mohs micrographic surgery: a technique for total margin assessment in veterinary cutaneous oncologic surgery. Vet Comp Oncol 4(3):151–160. Bookbinder, P.F., M.T. Butt, and H.J. Harvey. 1992. Determination of the number of mast cells in lymph node, bone marrow, and buffy coat cytologic specimens from dogs. J Am Vet Med Assoc 200(11):1648–1650. Borges, A.F. 1982. Dog-ear repair. Plast Reconstr Surg 69(4):707–713. Bostock, D.E. 1979. Prognosis after surgical excision of canine melanomas. Vet Pathol 16(1):32–40. Bostock, D.E. 1986. Neoplasms of the skin and subcutaneous tissues in dogs and cats. Br Vet J 142(1):1–19. Bostock, D.E. and M.T. Dye. 1980. Prognosis after surgical excision of canine fibrous connective tissue sarcomas. Vet Pathol 17(5): 581–588. Brambilla, P.G., P. Roccabianca, C. Locatelli, et al. 2006. Primary cardiac lipoma in a dog. J Vet Intern Med 20(3):691–693. Brehm, D.M., C.H. Vite, H.S. Steinberg, et al. 1995. A retrospective evaluation of 51 cases of peripheral nerve sheath tumors in the dog. J Am Anim Hosp Assoc 31(4):349–359. Brocks, B.A., I.J. Neyens, E. Teske, et al. 2008. Hypotonic water as adjuvant therapy for incompletely resected canine mast cell tumors: A randomized, double-blind, placebo-controlled study. Vet Surg 37(5):472–478. Brown, N.O., A.K. Patnaik, and E.G. MacEwen. 1985. Canine hemangiosarcoma: Retrospective analysis of 104 cases. J Am Vet Med Assoc 186(1):56–58. Bulakowski, E.J., J.C. Philibert, S. Siegel, et al. 2008. Evaluation of outcome associated with subcutaneous and intramuscular hemangiosarcoma treated with adjuvant doxorubicin in dogs: 21 cases (2001–2006). J Am Vet Med Assoc 233(1):122–128. Cahalane, A.K., S. Payne, L.G. Barber, et al. 2004. Prognostic factors for survival of dogs with inguinal and perineal mast cell tumors treated surgically with or without adjunctive treatment: 68 cases (1994–2002). J Am Vet Med Assoc 225(3):401–408. Chijiwa, K., K. Uchida, and S. Tateyama. 2004. Immunohistochemical evaluation of canine peripheral nerve sheath tumors and other soft tissue sarcomas. Vet Pathol 41(4):307–318. Chun, R. 2005. Common malignant musculoskeletal neoplasms of dogs and cats. Vet Clin North Am Small Anim Pract 35(5):1155– 1167, vi. Cohen, M., G.S. Post, and J.C. Wright. 2003. Gastrointestinal leiomyosarcoma in 14 dogs. J Vet Intern Med 17(1):107–110. Cohen, P.R., P.T. Martinelli, K.E. Schulze, et al. 2007. The cuticular purse string suture: A modified purse string suture for the partial closure of round postoperative wounds. Int J Dermatol. 46(7):746–753. Cohen, M., J.C. Wright, W.R. Brawner, et al. 2001. Use of surgery and electron beam irradiation, with or without chemotherapy, for treatment of vaccine-associated sarcomas in cats: 78 cases (1996– 2000). J Am Vet Med Assoc 219(11):1582–1589. Connery, N.A. and C.R. Bellenger. 2002. Surgical management of haemangiopericytoma involving the biceps femoris muscle in four dogs. J Small Anim Pract 43(11):497–500. Cooper, M., X. Tsai, and P. Bennett. 2009. Combination CCNU and vinblastine chemotherapy for canine mast cell tumours: 57 cases. Vet Comp Oncol 7(3):196–206.

Skin and Subcutaneous Tumors  81 Cooper, B.J. and B.A. Valentine. 2002. Tumors of muscle. In Tumors in Domestic Animals, pp. 319–363. D.J. Meuten, editor. Iowa State Press: Ames, IA. Couto, S.S., S.M. Griffey, P.C. Duarte, et al. 2002. Feline vaccineassociated fibrosarcoma: Morphologic distinctions. Vet Pathol 39(1):33–41. Cronin, K., R.L. Page, G. Spodnick, et al. 1998. Radiation therapy and surgery for fibrosarcoma in 33 cats. Vet Radiol Ultrasound 39(1):51–56. Daly, M.K., C.F. Saba, S.S. Crochik, et al. 2008. Fibrosarcoma adjacent to the site of microchip implantation in a cat. J Feline Med Surg 10(2):202–205. Davidson, E.B., C.R. Gregory, and P.H. Kass. 1997. Surgical excision of soft tissue fibrosarcomas in cats. Vet Surg 26(4):265–269. Davies, D.R., K.M. Wyatt, J.E. Jardine, et al. 2004. Vinblastine and prednisolone as adjunctive therapy for canine cutaneous mast cell tumors. J Am Anim Hosp Assoc 40(2):124–130. Davis, K.M., E.M. Hardie, B.D. Lascelles, et al. 2007. Feline fibrosarcoma: Perioperative management. Compend Contin Educ Vet 29(12):712–714. Dennis, R. 2008. Imaging features of orbital myxosarcoma in dogs. Vet Radiol Ultrasound 49(3):256–263. Dernell, W.S., S.J. Withrow, C.A. Kuntz, et al. 1998. Principles of treatment for soft tissue sarcoma. Clin Tech Small Anim Pract 13(1):59–64. De Queiroz, G.F., J.M. Matera, and M.L. Zaidan Dagli. 2008. Clinical study of cryosurgery efficacy in the treatment of skin and subcutaneous tumors in dogs and cats. Vet Surg 37(5):438–443. Dillon, C.J., G.N. Mauldin, and K.E. Baer. 2005. Outcome following surgical removal of nonvisceral soft tissue sarcomas in cats: 42 cases (1992–2000). J Am Vet Med Assoc 227(12):1955–1957. Dobson, J., S. Cohen, and S. Gould. 2004. Treatment of canine mast cell tumours with prednisolone and radiotherapy. Vet Comp Oncol 2(3):132–141. Dobson, J.M., S. Samuel, H. Milstein, et al. 2002. Canine neoplasia in the UK: Estimates of incidence rates from a population of insured dogs. J Small Anim Pract 43(6):240–246. Eckstein, C., F. Guscetti, M. Roos, et al. 2009. A retrospective analysis of radiation therapy for the treatment of feline vaccine-associated sarcoma. Vet Comp Oncol 7(1):54–68. Ehrhart, N. 2005. Soft-tissue sarcomas in dogs: A review. J Am Anim Hosp Assoc 41(4):241–246. Elmslie, R.E., P. Glawe, and S.W. Dow. 2008. Metronomic therapy with cyclophosphamide and piroxicam effectively delays tumor recurrence in dogs with incompletely resected soft tissue sarcomas. J Vet Intern Med 22(6):1373–1379. Endicott, M.M., S.C. Charney, J.A. McKnight, et al. 2007. Clinicopathological findings and results of bone marrow aspiration in dogs with cutaneous mast cell tumours: 157 cases (1999–2002). Vet Comp Oncol 5(1):31–37. Enneking, W.F. and G.E. Maale. 1988. The effect of inadvertent tumor contamination of wounds during the surgical resection of musculoskeletal neoplasms. Cancer 62(7):1251–1256. Enneking, W.F., S.S. Spanier, and M.A. Goodman. 1980. A system for the surgical staging of musculoskeletal sarcoma. Clin Orthop Relat Res (153):106–120. Essman, S.C., J.P. Hoover, R.J. Bahr, et al. 2002. An intrathoracic malignant peripheral nerve sheath tumor in a dog. Vet Radiol Ultrasound 43(3):255–259. Ettinger, S.N. 2003. Principles of treatment for soft-tissue sarcomas in the dog. Clin Tech Small Anim Pract 18(2):118–122. Ettinger, S.N., T.J. Scase, K.T. Oberthaler, et al. 2006. Association of argyrophilic nucleolar organizing regions, Ki-67, and proliferating

cell nuclear antigen scores with histologic grade and survival in dogs with soft tissue sarcomas: 60 cases (1996–2002). J Am Vet Med Assoc 228(7):1053–1062. Finora, K., N.F. Leibman, M.J. Fettman, et al. 2006. Cytological comparison of fine-needle aspirates of liver and spleen of normal dogs and of dogs with cutaneous mast cell tumours and an ultrasonographically normal appearing liver and spleen. Vet Comp Oncol 4(3):178–183. Foale, R.D., R.A. White, R. Harley, et al. 2003. Left ventricular myxosarcoma in a dog. J Small Anim Pract 44(11):503–507. Forrest, L.J., R. Chun, W.M. Adams, et al. 2000. Postoperative radiotherapy for canine soft tissue sarcoma. J Vet Intern Med 14(6):578–582. Fossum, T.W., C.G. Couto, W.D. DeHoff, et al. 1988. Treatment of hemangiopericytoma in a dog, using surgical excision, radiation, and a thoracic pedicle skin graft. J Am Vet Med Assoc 193(11):1440–1442. Fries, L.F., J. Tartaglia, J. Taylor, et al. 1996. Human safety and immunogenicity of a canarypox-rabies glycoprotein recombinant vaccine: An alternative poxvirus vector system. Vaccine 14(5): 428–434. Frimberger, A.E., A.S. Moore, S.M. LaRue, et al. 1997. Radiotherapy of incompletely resected, moderately differentiated mast cell tumors in the dog: 37 cases (1989–1993). J Am Anim Hosp Assoc 33(4):320–324. Fulcher, R.P., L.L. Ludwig, P.J. Bergman, et al. 2006. Evaluation of a two-centimeter lateral surgical margin for excision of grade I and grade II cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 228(2):210–215. Gaitero, L., S. Anor, D. Fondevila, et al. 2008. Canine cutaneous spindle cell tumours with features of peripheral nerve sheath tumours: A histopathological and immunohistochemical study. J Comp Pathol 139(1):16–23. Galeotti, F., F. Barzagli, A. Vercelli, et al. 2004. Feline lymphangiosarcoma—Definitive identification using a lymphatic vascular marker. Vet Dermatol 15(1):13–18. Ghisleni, G., P. Roccabianca, R. Ceruti, et al. 2006. Correlation between fine-needle aspiration cytology and histopathology in the evaluation of cutaneous and subcutaneous masses from dogs and cats. Vet Clin Pathol 35(1):24–30. Gieger, T.L., A.P. Theon, J.A. Werner, et al. 2003. Biologic behavior and prognostic factors for mast cell tumors of the canine muzzle: 24 cases (1990–2001). J Vet Intern Med 17(5):687–692. Gill, V., N.F. Leibman, S. Monette, et al. 2007. Histologic grade, prolifereation indices and clinical outcome in canine subcutane­ ous mast cell tumors. In Veterinary Cancer Society. Fort Lauderdale, FL. Gilson, S.D. and E.A. Stone. 1990. Surgically induced tumor seeding in eight dogs and two cats. J Am Vet Med Assoc 196(11):1811–15. Goldschmidt, M.H. and M.J. Hendrick. 2002. Tumors of the skin and soft tissues. In Tumors in Domestic Animals. 4th edition, p. 45. D.J. Meuten, editor. Iowa State Press: Iowa City, IA. Goldschmidt, M.H. and F.S. Shofer. 1992. Skin Tumors of the Dog and Cat, 1st edition. Pergamon Press: Oxford. Goldschmidt, M.H. and F.S. Shofer. 1998. Uncommon skin tumors. In Skin Tumors of the Dog and Cat, p. 291–295. Butterworth Heinemann: New York. Graves, G.M., D.E. Bjorling, and Mahaffey E. 1988. Canine hemangiopericytoma: 23 cases (1967–1984). J Am Vet Med Assoc 192(1):99–102. Greene, F.L. 2002. AJCC Cancer Staging Manual. American Joint Committee on Cancer, Report No.: 0387952713. American Cancer Society: New York.

82  Veterinary Surgical Oncology Grier, R.L., G. Di Guardo, R. Myers, et al. 1995. Mast cell tumour destruction in dogs by hypotonic solution. J Small Anim Pract 36(9):385–388. Gross, T.L., P.J. Ihrke, E.J. Walder, et al. 2005. Skin Diseases of the Dog and Cat: Clinical Histopathologic Diagnosis, 2nd edition. Blackwell Science: Oxford. Hahn, K.A., G.K. King, and J.K. Carreras. 2004. Efficacy of radiation therapy for incompletely resected grade-III mast cell tumors in dogs: 31 cases (1987–1998). J Am Vet Med Assoc 224(1):79–82. Hahn, K.A., G. Ogilvie, T. Rusk, et al. 2008. Masitinib is safe and effective for the treatment of canine mast cell tumors. J Vet Intern Med 22(6):1301–1309. Hammer, AS, C.G. Couto, J. Filppi, et al. 1991. Efficacy and toxicity of VAC chemotherapy (vincristine, doxorubicin, and cyclophosphamide) in dogs with hemangiosarcoma. J Vet Intern Med 5(3):160–166. Hanna, S.L. and B.D. Fletcher. 1995. MR imaging of malignant softtissue tumors. Magn Reson Imaging Clin N Am 3(4):629–650. Harcourt-Brown, T.R., N. Granger, P.M. Smith, et al. 2009. Use of a lateral surgical approach to the femoral nerve in the management of two primary femoral nerve sheath tumours. Vet Comp Orthop Traumatol 22(3):229–232. Hargis, A.M., P.J. Ihrke, W.L. Spangler, et al. 1992. A retrospective clinicopathologic study of 212 dogs with cutaneous hemangiomas and hemangiosarcomas. Vet Pathol 29(4):316–328. Hendrick, M.J. and M.H. Goldschmidt. 1991. Do injection site reactions induce fibrosarcomas in cats? J Am Vet Med Assoc 199(8): 968. Hershey, A.E., K.U. Sorenmo, M.J. Hendrick, et al. 2000. Prognosis for presumed feline vaccine—associated sarcoma after excision: 61 cases (1986—1996). J Am Vet Med Assoc 216(1):58–61. Hosoya, K., W.C. Kisseberth, F.J. Alvarez, et al. 2009. Adjuvant CCNU (lomustine) and prednisone chemotherapy for dogs with incompletely excised grade 2 mast cell tumors. J Am Anim Hosp Assoc 45(1):14–18. Hudson-Peacock, M.J. and C.M. Lawrence. 1995. Comparison of wound closure by means of dog ear repair and elliptical excision. J Am Acad Dermatol 32(4):627–630. Hume, C., M. Kiupel, L. Rigatti, et al. 2007. Outcome of dogs with grade III mast cell tumors. In Veterinary Cancer Society. Fort Lauderdale, FL. Isotani, M., N. Ishida, M. Tominaga, et al. 2008. Effect of tyrosine kinase inhibition by imatinib mesylate on mast cell tumors in dogs. J Vet Intern Med 22(4):985–988. Jaffe, M.H., G. Hosgood, S.C. Kerwin, et al. 2000. Deionised water as an adjunct to surgery for the treatment of canine cutaneous mast cell tumours. J Small Anim Pract 41(1):7–11. Johannes, C.M., C.J. Henry, S.E. Turnquist, et al. 2007. Hemangiosarcoma in cats: 53 cases (1992–2002). J Am Vet Med Assoc 231(12):1851–1856. Johnson, T.O., F.Y. Schulman, T.P. Lipscomb, et al. 2002. Histopathology and biologic behavior of pleomorphic cutaneous mast cell tumors in fifteen cats. Vet Pathol 39(4):452–457. Jourdier, T.M., C. Moste, M.C. Bonnet, et al. 2003. Local immunotherapy of spontaneous feline fibrosarcomas using recombinant poxviruses expressing interleukin 2 (IL2). Gene Ther 10(26): 2126–2132. Kaldrymidou, H., L. Leontides, A.F. Koutinas, et al. 2002. Prevalence, distribution and factors associated with the presence and the potential for malignancy of cutaneous neoplasms in 174 dogs admitted to a clinic in northern Greece. J Vet Med A Physiol Pathol Clin Med 49(2):87–91.

Kamstock, D., R. Elmslie, D. Thamm, et al. 2007. Evaluation of a xenogeneic VEGF vaccine in dogs with soft tissue sarcoma. Cancer Immunol Immunother 56(8):1299–1309. Kapatkin, A.S., H.S. Mullen, D.T. Matthiesen, et al. 1992. Leiomyosarcoma in dogs: 44 cases (1983–1988). J Am Vet Med Assoc 201(7):1077–1079. Kass, P.H., W.L. Spangler, M.J. Hendrick, et al. 2003. Multicenter casecontrol study of risk factors associated with development of vaccine-associated sarcomas in cats. J Am Vet Med Assoc 223(9): 1283–1292. Kim, H.J., H.S. Chang, C.B. Choi, et al. 2005. Infiltrative lipoma in cervical bones in a dog. J Vet Med Sci 67(10):1043–1046. Kim, D.Y., D.Y. Cho, D.Y. Kim, et al. 2003. Malignant peripheral nerve sheath tumor with divergent mesenchymal differentiations in a dog. J Vet Diagn Invest 15(2):174–178. Kiupel, M., J.D. Webster, J.B. Kaneene, et al. 2004. The use of KIT and tryptase expression patterns as prognostic tools for canine cutaneous mast cell tumors. Vet Pathol 41 (4):371–377. Kiupel, M., J.D. Webster, R.A. Miller, et al. 2005. Impact of tumour depth, tumour location and multiple synchronous masses on the prognosis of canine cutaneous mast cell tumours. J Vet Med A Physiol Pathol Clin Med 52(6):280–286. Kobayashi, T, M.L. Hauck, R. Dodge, et al. 2002. Preoperative radiotherapy for vaccine associated sarcoma in 92 cats. Vet Radiol Ultrasound 43(5):473–479. Kolm, U.S., M. Kleiter, A. Kosztolich, et al. 2002. Benign intrapericardial lipoma in a dog. J Vet Cardiol 4(1):25–29. Kotilingam, D., D.C. Lev, A.J. Lazar, et al. 2006. Staging soft tissue sarcoma: Evolution and change. CA Cancer J Clin 56(5):282–291. Krick, E.L., A.P. Billings, F.S. Shofer, et al. 2009. Cytological lymph node evaluation in dogs with mast cell tumours: Association with grade and survival. Vet Com Oncol 7(2):130–138. Kristal, O., K.M. Rassnick, J.M. Gliatto, et al. 2004. Hepatotoxicity associated with CCNU (lomustine) chemotherapy in dogs. J Vet Intern Med 18(1):75–80. Kuntz, C.A., W.S. Dernell, B.E. Powers, et al. 1997. Prognostic factors for surgical treatment of soft-tissue sarcomas in dogs: 75 cases (1986–1996). J Am Vet Med Assoc 211(9):1147–1151. LaDue, T., G.S. Price, R. Dodge, et al. 1998. Radiation therapy for incompletely resected canine mast cell tumors. Vet Radiol Ultrasound 39(1):57–62. Langenbach, A., P.M. Mcmanus, M.J. Hendrick, et al. 2001. Sensitivity and specificity of methods of assessing the regional lymph nodes for evidence of metastasis in dogs and cats with solid tumors. J Am Vet Med Assoc 218(9):1424–1428. Lawrence, J., L. Forrest, W. Adams, et al. 2008. Four-fraction radiation therapy for macroscopic soft tissue sarcomas in 16 dogs. J Am Anim Hosp Assoc 44(3):100–108. LeCouteur, R.A. 2001. Tumors of the nervous system. In Small Animal Clinical Oncology, pp. 381–455. S.J. Withrow and E.G. MacEwen, editors, WB Saunders: Philadelphia. Leiter, U. and C. Garbe. 2008. Epidemiology of melanoma and nonmelanoma skin cancer—The role of sunlight. Adv Exp Med Biol. 624:89–103. Lenard, Z.M., S.F. Foster, A.J. Tebb, et al. 2007. Lymphangiosarcoma in two cats. J Feline Med Surg 9(2):161–167. Lepri, E., G. Ricci, L. Leonardi, et al. 2003. Diagnostic and prognostic features of feline cutaneous mast cell tumours: A retrospective analysis of 40 cases. Vet Res Commun 27(Suppl 1):707–709. Levy, M.S., A.S. Kapatkin, A.K. Patnaik, et al. 1997. Spinal tumors in 37 dogs: Clinical outcome and long-term survival (1987–1994). J Am Anim Hosp Assoc 33(4):307–312.

Skin and Subcutaneous Tumors  83 Lidbetter, D.A., F.A. Williams, Jr., D.J. Krahwinkel, et al. 2002. Radical lateral body-wall resection for fibrosarcoma with reconstruction using polypropylene mesh and a caudal superficial epigastric axial pattern flap: a prospective clinical study of the technique and results in 6 cats. Vet Surg 31(1):57–64. Liptak, J.M. 2007. Soft tissue sarcomas. In Withrow & MacEwen’s Small Animal Clinical Oncology, pp. 425–454. S.J. Withrow and E.G. MacEwen, editors. WB Saunders: St. Louis. Little, L.K. and M. Goldschmidt. 2007. Cytologic appearance of a keloidal fibrosarcoma in a dog. Vet Clin Pathol 36(4): 364–367. Litster, A.L. and K.U. Sorenmo. 2006. Characterisation of the signalment, clinical and survival characteristics of 41 cats with mast cell neoplasia. J Feline Med Surg 8(3):177–183. Liu, S.M. and I. Mikaelian. 2003. Cutaneous smooth muscle tumors in the dog and cat. Vet Pathol 40(6):685–692. London, C.A., P.B. Malpas, S.L. Wood-Follis, et al. 2009. Multi-center, placebo-controlled, double-blind, randomized study of oral toceranib phosphate (SU11654), a receptor tyrosine kinase inhibitor, for the treatment of dogs with recurrent (either local or distant) mast cell tumor following surgical excision. Clin Cancer Res 15(11):3856–3865. Luna, L.D., M.L. Higginbotham, C.J. Henry, et al. 2000. Feline nonocular melanoma: A retrospective study of 23 cases (1991–1999). J Feline Med Surg 2(4):173–181. MacEwen, E.G., I.D, Kurzman, D.M. Vail, et al. 1999. Adjuvant therapy for melanoma in dogs: Results of randomized clinical trials using surgery, liposome-encapsulated muramyl tripeptide, and granulocyte macrophage colony-stimulating factor. Clin Cancer Res 5(12):4249–4258. MacEwan, E.G., B.E. Powers, and D. Macy. 2001. Soft tissue sarcomas. In Small Animal Clinical Oncology, 3rd edition. S.J. Withrow, editor. WB Saunders: Philadelphia. Maglennon, G.A., S. Murphy, V. Adams, et al. 2008. Association of Ki67 index with prognosis for intermediate-grade canine cutaneous mast cell tumours. Vet Comp Oncol 6(4):268– 274. Mayhew, P.D. and D.J. Brockman. 2002. Body cavity lipomas in six dogs. J Small Anim Pract 43(4):177–181. Mayr, B., W. Swidersky, W. Schleger et al. 1990. Cytogenetic characterization of a canine haemangiopericytoma. Br Vet J 146(3): 260–263. Mazzei, M., F. Millanta, S. Citi, et al. 2002. Haemangiopericytoma: Histological spectrum, immunohistochemical characterization and prognosis. Vet Dermatol 13(1):15–21. McAbee, K.P., L.L. Ludwig, P.J. Bergman, et al. 2005. Feline cutaneous hemangiosarcoma: A retrospective study of 18 cases (1998–2003). J Am Anim Hosp Assoc 41(2):110–116. McCaw, D.L., M.A. Miller, G.K. Ogilvie, et al. 1994. Response of canine mast cell tumors to treatment with oral prednisone. J Vet Intern Med 8(6):406–408. McColl Williamson, M. and D.J. Middleton. 1998. Cutaneous soft tissue tumours in dogs: Classification, differentiation, and histogenesis. Vet Dermatol 9(1):43–48 McEntee, M.C. 2006. Veterinary radiation therapy: Review and current state of the art. J Am Anim Hosp Assoc 42(2):94–109. McEntee, M.C., R.L. Page, G.N. Mauldin, et al. 2000. Results of irradiation of infiltrative lipoma in 13 dogs. Vet Radiol Ultrasound 41(6):554–556. McEntee, M.C. and V.F, Samii. 2000. The utility of contrast enhanced computed tomography in feline vaccine associated sarcomas: 35 cases [abstract]. Vet Radiol Ultrasound 41:575.

McEntee, M.C. and D.E. Thrall. 2001. Computed tomographic imaging of infiltrative lipoma in 22 dogs. Vet Radiol Ultrasound 42(3):221–225. McKnight, J.A., G.N. Mauldin, M.C. McEntee, et al. 2000. Radiation treatment for incompletely resected soft-tissue sarcomas in dogs. J Am Vet Med Assoc 217(2):205–210. McLaughlin, R. Jr. and A.B. Kuzma. 1991. Intestinal strangulation caused by intra-abdominal lipomas in a dog. J Am Vet Med Assoc 199(11):1610–1611. McManus, P.M. 1999. Frequency and severity of mastocytemia in dogs with and without mast cell tumors: 120 cases (1995–1997). J Am Vet Med Assoc 215(3):355–357. McSporran, K. 2009. Histologic grade predicts recurrence for marginally excised canine subcutaneous soft tissue sarcomas. Vet Pathol 46(5):928–933. Michels, G.M., D.W. Knapp, D.B. DeNicola, et al. 2002. Prognosis following surgical excision of canine cutaneous mast cell tumors with histopathologically tumor-free versus nontumor-free margins: A retrospective study of 31 cases. J Am Anim Hosp Assoc 38(5): 458–466. Mikaelian, I. and T.L. Gross. 2002. Keloidal fibromas and fibrosarcomas in the dog. Vet Pathol 39(1):149–153. Miles, J. and D. Clarke. 2001. Intrathoracic lipoma in a Labrador retriever. J Small Anim Pract 42(1):26–28. Miller, M.A., S.L. Nelson, J.R. Turk, et al. 1991. Cutaneous neoplasia in 340 cats. Vet Pathol 28(5):389–395. Molander-McCrary, H., C.J. Henry, K. Potter, et al. 1998. Cutaneous mast cell tumors in cats: 32 cases (1991–1994). J Am Anim Hosp Assoc 34(4):281–284. Moore, A.S. 2002. Radiation therapy for the treatment of tumours in small companion animals. Vet J 164(3):176–187. Morrison, W.B. and R.M. Starr. 2001. Vaccine-associated feline sarcomas. J Am Vet Med Assoc 218(5):697–702. Mukaratirwa, S., J. Chipunza, S. Chitanga, et al. 2005. Canine cutaneous neoplasms: Prevalence and influence of age, sex and site on the presence and potential malignancy of cutaneous neoplasms in dogs from Zimbabwe. J S Afr Vet Assoc 76(2):59–62. Mullins, M.N., W.S. Dernell, S.J. Withrow, et al. 2006. Evaluation of prognostic factors associated with outcome in dogs with multiple cutaneous mast cell tumors treated with surgery with and without adjuvant treatment: 54 cases (1998–2004). J Am Vet Med Assoc 228(1):91–95. Murphy, S., A.H. Sparkes, A.S. Blunden, et al. 2006. Effects of stage and number of tumours on prognosis of dogs with cutaneous mast cell tumours. Vet Rec 158 (9):287–291. Murphy, S., A.H. Sparkes, K.C. Smith, et al. 2004. Relationships between the histological grade of cutaneous mast cell tumours in dogs, their survival and the efficacy of surgical resection. Vet Rec 154 (24):743–746. Northrup, N.C., E.W. Howerth, B.G. Harmon, et al. 2005. Variation among pathologists in the histologic grading of canine cutaneous mast cell tumors with uniform use of a single grading reference. J Vet Diagn Invest 17(6):561–564. O’Brien, M.G. 2003. Skin and subcutis. In Textbook of Small Animal Surgery, 3rd edition, pp. 2359–2368. D. Slatter, editor. Philadelphia: Saunders. O’Keefe, D.A., C.G. Couto, C. Burke-Schwartz, et al. 1987. Systemic mastocytosis in 16 dogs. J Vet Intern Med 1(2):75–80. Owen L. 1980. TNM Classification of Tumors in Domestic Animals, 1980. Geneva: World Health Organization, Geneva. Pakhrin, B., M.S. Kang, I.H. Bae, et al. 2007. Retrospective study of canine cutaneous tumors in Korea. J Vet Sci 8(3):229–236.

84  Veterinary Surgical Oncology Paoletti, E., J. Taylor, B. Meignier, et al. 1995. Highly attenuated poxvirus vectors: NYVAC, ALVAC and TROVAC. Dev Biol Stand 84:159–163. Patnaik, A.K., W.J. Ehler, and E.G. MacEwen. 1984. Canine cutaneous mast cell tumor: Morphologic grading and survival time in 83 dogs. Vet Pathol 21(5):469–474. Patterson, C.C., R.L. Perry, and B. Steficek. 2008. Malignant peripheral nerve sheath tumor of the diaphragm in a dog. J Am Anim Hosp Assoc 44(1):36–40. Pavletic, M.M. 2000. Use of an external skin-stretching device for wound closure in dogs and cats. J Am Vet Med Assoc 217(3): 350–354. Pavletic, M.M. 2003. Chpt 23—Pedicle Grafts. In: Slatter, D.H. (ed.) Textbook of Small Animal Surgery. Third Edition. Philadelphia, PA: Saunders. Plavec, T., M. Kessler, B. Kandel, et al. 2006. Palliative radiotherapy as treatment for non-resectable soft tissue sarcomas in the dog—A report of 15 cases. Vet Comp Oncol 4(2):98–103. Poirier, V.J., W.M. Adams, L.J. Forrest, et al. 2006. Radiation therapy for incompletely excised grade II canine mast cell tumors. J Am Anim Hosp Assoc 42(6):430–434. Powers, B.E., P.J. Hoopes, and E.J. Ehrhart. 1995. Tumor diagnosis, grading, and staging. Semin Vet Med Surg (Small Anim) 10(3): 158–167. Quintin-Colonna, F., P. Devauchelle, D. Fradelizi, et al. 1996. Gene therapy of spontaneous canine melanoma and feline fibrosarcoma by intratumoral administration of histoincompatible cells expressing human interleukin-2. Gene Ther 3(12):1104–1112. Rassnick, K.M. 2003. Medical management of soft tissue sarcomas. Vet Clin North Am Small Anim Pract 33(3):517–531. Rassnick, K.M., D.B. Bailey, A.B. Flory, et al. 2008. Efficacy of vinblastine for treatment of canine mast cell tumors. J Vet Intern Med 22(6):1390–1396. Rassnick, K.M., A.S. Moore, L.E. Williams, et al. 1999. Treatment of canine mast cell tumors with CCNU (lomustine). J Vet Intern Med 13(6):601–605. Rassnick, K.M., D.M. Ruslander, S.M. Cotter, et al. 2001. Use of carboplatin for treatment of dogs with malignant melanoma: 27 cases (1989–2000). J Am Vet Med Assoc 218(9):1444–1448. Rohrborn, A. and H.D. Roher. 1998. Surgical aspects in the multidisciplinary treatment of soft tissue sarcomas. Praxis (Bern 1994) 87(34):1050–1060. Romanelli, G., L. Marconato, D. Olivero, et al. 2008. Analysis of prognostic factors associated with injection-site sarcomas in cats: 57 cases (2001–2007). J Am Vet Med Assoc 232(8):1193– 1199. Romansik, E.M., C.M. Reilly, P.H. Kass, et al. 2007. Mitotic index is predictive for survival for canine cutaneous mast cell tumors. Vet Pathol 44(3):335–341. Sawamoto, O., J. Yamate, M. Kuwamura, et al. 1999. A canine peripheral nerve sheath tumor including peripheral nerve fibers. J Vet Med Sci 61(12):1335–1338. Scase, T.J., D. Edwards, J. Miller, et al. 2006. Canine mast cell tumors: Correlation of apoptosis and proliferation markers with prognosis. J Vet Intern Med 20(1):151–158. Scavelli, T.D., A.K. Patnaik, C.J. Mehlhaff, et al. 1985. Hemangiosarcoma in the cat: Retrospective evaluation of 31 surgical cases. J Am Vet Med Assoc 187(8):817–819. Schlieman, M., R. Smith, and W.G. Kraybill. 2006. Adjuvant therapy for extremity sarcomas. Curr Treat Options Oncol 7(6):456– 463. Schulman, F.Y., T.O. Johnson, P.R. Facemire, et al. 2009. Feline peripheral nerve sheath tumors: Histologic, immunohistochemical, and

clinicopathologic correlation (59 tumors in 53 cats). Vet Pathol 46(6):1166–1180. Schultheiss, P.C. 2004. A retrospective study of visceral and nonvisceral hemangiosarcoma and hemangiomas in domestic animals. J Vet Diagn Invest 16(6):522–526. Seguin, B., M.F. Besancon, J.L. McCallan, et al. 2006. Recurrence rate, clinical outcome, and cellular proliferation indices as prognostic indicators after incomplete surgical excision of cutaneous grade II mast cell tumors: 28 dogs (1994–2002). J Vet Intern Med 20(4):933–940. Seguin, B., N.F. Leibman, V.S. Bregazzi, et al. 2001. Clinical outcome of dogs with grade-II mast cell tumors treated with surgery alone: 55 cases (1996–1999). J Am Vet Med Assoc 218(7):1120–1123. Seguin, B., D.E. Mcdonald, M.S. Kent, et al. 2005. Tolerance of cutaneous or mucosal flaps placed into a radiation therapy field in dogs. Vet Surg 34(3):214–222. Selting, K.A., B.E. Powers, L.J. Thompson, et al. 2005. Outcome of dogs with high-grade soft tissue sarcomas treated with and without adjuvant doxorubicin chemotherapy: 39 cases (1996–2004). J Am Vet Med Assoc 227(9):1442–1448. Sfiligoi, G., K.M. Rassnick, J.M. Scarlett, et al. 2005. Outcome of dogs with mast cell tumors in the inguinal or perineal region versus other cutaneous locations: 124 cases (1990–2001). J Am Vet Med Assoc 226(8):1368–1374. Shaw, S.C., M.S. Kent, I.K. Gordon, et al. 2009. Temporal changes in characteristics of injection-site sarcomas in cats: 392 cases (1990– 2006). J Am Vet Med Assoc 234(3):376–380. Shelly, S.M. 2003. Cutaneous lesions. Vet Clin North Am Small Anim Pract 33(1):1–46. Simon, D., D.M. Ruslander, K.M. Rassnick, et al. 2007. Orthovoltage radiation and weekly low dose of doxorubicin for the treatment of incompletely excised soft-tissue sarcomas in 39 dogs. Vet Rec 160(10):321–326. Simpson, A.M., L.L. Ludwig, S.J. Newman, et al. 2004. Evaluation of surgical margins required for complete excision of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 224(2):236–240. Sorenmo, K.U., K.A. Jeglum, and S.C. Helfand. 1993. Chemotherapy of canine hemangiosarcoma with doxorubicin and cyclophosphamide. J Vet Intern Med 7(6):370–376. Spugnini, E.P., A. Baldi, B. Vincenzi, et al. 2007. Intraoperative versus postoperative electrochemotherapy in high grade soft tissue sarcomas: A preliminary study in a spontaneous feline model. Cancer Chemother Pharmacol 59(3):375–381. Stanclift, R.M. and S.D. Gilson. 2008. Evaluation of neoadjuvant prednisone administration and surgical excision in treatment of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 232(1): 53–62. Stefanello, D., E. Morello, P. Roccabianca, et al. 2008. Marginal excision of low-grade spindle cell sarcoma of canine extremities: 35 dogs (1996–2006). Vet Surg 37(5):461–465. Stefanello, D., P. Valenti, S. Faverzani, et al. 2009. Ultrasound-guided cytology of spleen and liver: A prognostic tool in canine cutaneous mast cell tumor. J Vet Intern Med 23(5):1051–1057. Straw, R.C., S.J. Withrow, and B.E. Powers. 1992. Partial or total hemipelvectomy in the management of sarcomas in nine dogs and two cats. Vet Surg 21(3):183–188. Sugiyama, A., T. Morita, A. Shimada, et al. 2008. Primary malignant peripheral nerve sheath tumor with eosinophilic cytoplasmic globules arising from the greater omentum in a dog. J Vet Med Sci 70(7):739–742. Swaim, S.F. 2003. Chpt 24—Skin Grafts. In: Slatter, D.H. (ed.) Textbook of Small Animal Surgery. Third Edition. Philadelphia, PA: Saunders.

Skin and Subcutaneous Tumors  85 Thamm, D.H., E.A. Mauldin, and D.M. Vail. 1999. Prednisone and vinblastine chemotherapy for canine mast cell tumor—41 cases (1992–1997). J Vet Intern Med 13 (5):491–497. Thamm, D.H., M.M. Turek, and D.M. Vail. 2006. Outcome and prognostic factors following adjuvant prednisone/vinblastine chemotherapy for high-risk canine mast cell tumour: 61 cases. J Vet Med Sci 68(6):581–587. Theilen, G.H. and B.R. Madewell. 1979. Veterinary Cancer Medicine. Lea & Febiger: Philadelphia. Thornton, K. 2008. Chemotherapeutic management of soft tissue sarcoma. Surg Clin North Am 88(3):647–660, viii. Trout, N.J., M.M. Pavletic, and K.H. Kraus. 1995. Partial scapulectomy for management of sarcomas in three dogs and two cats. J Am Vet Med Assoc 207(5):585–587. Trout, N.J. 2003. Chpt 22—Principles of Plastic and Reconstructive Surgery. In: Slatter, D.H. (ed.) Textbook of Small Animal Surgery. Third Edition. Philadelphia, PA: Saunders. Turrel, J.M., J. Farrelly, R.L. Page, et al. 2006. Evaluation of stron­ tium 90 irradiation in treatment of cutaneous mast cell tumors in cats: 35 cases (1992–2002). J Am Vet Med Assoc 228(6): 898–901.

Turrel, J.M., B.E. Kitchell, L.M. Miller, et al. 1988. Prognostic factors for radiation treatment of mast cell tumor in 85 dogs. J Am Vet Med Assoc 193(8):936–940. Vail, D.M. and S.J. Withrow. 2007. Tumors of the skin and subcutaneous tissues. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 375–401. S.J. Withrow and E.G. MacEwen, editors. St. Louis, MO: Saunders Elsevier. Vickery, K.R., H. Wilson, D.M. Vail, et al. 2008. Dose-escalating vinblastine for the treatment of canine mast cell tumour. Vet Comp Oncol 6(2):111–119. Ward, H., L.E. Fox, M.B. Calderwood-Mays, et al. 1994. Cutaneous hemangiosarcoma in 25 dogs: A retrospective study. J Vet Intern Med 8(5):345–348. Weisse, C., F.S. Shofer, and K. Sorenmo. 2002. Recurrence rates and sites for grade II canine cutaneous mast cell tumors following complete surgical excision. J Am Anim Hosp Assoc 38(1):71–73. Williams, J.H. 2005. Lymphangiosarcoma of dogs: A review. J S Afr Vet Assoc 76(3):127–131. Withrow, S.J. and D.M. Vail, editors. 2007. Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition. St. Louis, MO: Saunders Elsevier.

5 Head and neck tumors Sara A. Ayres, Julius M. Liptak

Lymph Node Staging The regional lymph nodes for head and neck cancers include the mandibular, parotid, and retropharyngeal lymphocentrums. These lymph nodes should be carefully palpated for enlargement or asymmetry. This may be inaccurate, however, because it has been demonstrated that lymph node size is not an accurate predictor of metastasis in dogs with oral melanoma (Williams and Packer 2003). Furthermore, of the three regional lymphocentrums, only the mandibular lymph nodes are externally palpable (Smith 1995) and only 55% of cats and dogs with metastatic oral and maxillofacial tumors have metastasis to the mandibular lymph nodes (Smith 2002). Given these limitations, it may be prudent to biopsy these lymph nodes in all dogs with a known malignancy of the head; however this still remains controversial in veterinary medicine. Currently, lymph node aspirates are recommended for all animals with head and neck tumors, regardless of the size or degree of fixation of the lymph nodes (Williams and Packer 2003; Smith 2002). It is hoped that sentinel lymph node assessment in the near future will become more widely accepted and practiced as this may permit the preoperative diagnosis of metastatic lymph nodes, without more aggressive en bloc surgical excisions of the regional lymph nodes. Methods to detect sentinel lymph nodes in people with head and neck cancer include lymphoscintigraphy, intraoperative blue dyes, and intraoperative gamma probes (Lurie et al. 2006). Lymphoscintigraphy, intraoperative dyes, and contrast-enhanced ultrasonography have been described in dogs with various tumors, including head and neck cancer (Nieweg et al. 2001; Nyman et al. 2005; Worley et al. 2007).

En bloc resection of the regional lymph nodes has been described, and although the therapeutic benefit of this approach is unknown, it may provide valuable staging information (Smith 1995, 2002). The skin is incised from the rostral and proximal aspect of the vertical ear canal, ventral to the caudal aspect of the zygomatic arch, to the bifurcation of the external jugular vein (Smith 1995). The platysma and parotidoauricularis muscles are incised to reveal fascia and loose areolar tissue covering the vertical ear canal and masseter muscle. Incision of the areolar tissue over the ventral aspect of the zygomatic arch exposes the parotid lymphocentrum, which has one to three lymph nodes, along the rostral edge of the parotid salivary gland (Smith 1995). The mandibular lymphocentrum, which contains one to five lymph nodes, is located between the bifurcation of the jugular vein and division of the lingofacial vein into its lingual and facial branches (Smith 1995). The medial retropharyngeal lymphocentrum, which usually consists of one elongated lymph node on the lateral aspect of the thyropharyngeus muscle, is exposed by incising the adventitia along the caudal aspect of the mandibular salivary gland and retracting the mandibular salivary gland rostrally and the brachiocephalicus and sternocephalicus muscles dorsally (Smith 1995).

Nasal Planum Tumors Nasal planum tumors are relatively common in cats, but rare in dogs. Squamous cell carcinoma (SCC) is the most common tumor of the nasal planum in both cats and dogs, with SCC originating from the cornified external surface of the nasal planum in cats and the mucous membrane of the nostril or nasal planum in dogs (Withrow 2007). Other common sites for SCC in

Veterinary Surgical Oncology, First Edition. Edited by Simon T. Kudnig, Bernard Séguin. © 2012 by John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.

87

88  Veterinary Surgical Oncology

Figure 5.2.  Preoperative appearance of a nasal planum SCC in a cat. The deeply ulcerated and erosive morphology of this lesion is characteristic of SCC of the nasal planum in cats.

Figure 5.1.  Squamous cell carcinoma of the nasal planum in a golden retriever. Note the marked ulceration and erosion combined with some areas of proliferation.

cats involve poorly haired regions of the head and neck, such as the eyelids, pinnae, and preauricular areas (Thomson 2007; Lana et al. 1997). Predisposing causes for the development of head and neck SCC in cats and dogs include light pigmentation (i.e., white hair coat), short hair coat, and chronic exposure to ultraviolet light (Withrow 2007; Thomson 2007; Dorn et al. 1971; Lana et al. 1997). White-haired cats have 13.4 times the risk of developing head and neck SCC compared to darker colored cats (Dorn et al. 1971). Older animals are typically affected with a mean age at presentation of 8 years for dogs and 12 years for cats (Withrow 2007; Ruslander et al. 1997; Lana et al. 1997). There is no breed predisposition in cats, but there is a high incidence of nasal planum SCC in male golden and Labrador retrievers (Figure 5.1; Lascelles et al. 2000). Other tumor types include lymphoma, fibrosarcoma, mast cell tumors, malignant melanoma, hemangioma, and fibroma (Withrow 2007). Eosinophilic granulomas and immune-mediated diseases are non-neoplastic conditions involving the nasal planum that can have a similar erosive to proliferative appearance mimicking benign and malignant tumors (Withrow 2007). History and clinical signs Head and neck SCC are often chronic cancers progressing from actinic changes (crusting and erythema) to

Figure 5.3.  The typical appearance of a SCC of the pinna in a cat. Note the spectrum of changes typical of feline cutaneous SCC, with regions of erythema (actinic changes), superficial ulceration and erosion (carcinoma in situ), and multifocal areas of deep ulceration and erosion along the apex and medial border of the helix.

carcinoma in situ lesions (noninvasive carcinoma confined to the epidermis characterized by superficial erosions and ulcers) to invasive carcinomas (deep erosive lesions; Figures 5.2 and 5.3) (Withrow 2007; Thomson 2007; Lana et al. 1997). Occasionally, nasal planum SCC may have a proliferative appearance (Thomson 2007).

Head and Neck Tumors  89

The eyelids, preauricular areas, and pinnae should also be examined carefully as up to 30% of cats have SCC in multiple head and neck locations (Withrow 2007; Vail and Withrow 2007). Diagnosis and clinical staging A deep wedge or incisional biopsy is required for definitive diagnosis of nasal planum tumors (Withrow 2007). Cytologic examination of fine-needle aspirates, impression smears, or superficial biopsies is often unrewarding because they usually reveal inflammation, ulceration, and hemorrhage, which can be present in both neoplastic and nonneoplastic lesions (Withrow 2007; Thomson 2007). Biopsies may not be required in cats with erosive or ulcerated lesions of the eyelids, preauricular area, and/or pinnae, and possibly even the nasal planum, when there is a high index of suspicion for SCC (Figure 5.2) (Thomson 2007). A deep incisional biopsy is recommended in cats with proliferative lesions, and it can be justified in all dogs and perhaps cats with any nasal planum lesion to determine histology and the depth of invasion (Withrow 2007; Thomson 2007). Advanced imaging, with either computed tomography (CT) or magnetic resonance imaging (MRI), is recommended for local staging of nasal planum tumors in dogs to determine the caudal extent of the tumor and to assist in determining margins for surgical planning (Withrow 2007). Metastasis to the regional lymph nodes and/or lungs is rare, but it is occasionally noted in cats with advanced or poorly differentiated lesions (Thomson 2007; Withrow 2007). The regional lymph nodes should be palpated and aspirated or ideally excised (see previous section on lymph node staging). Three-view radiographs of the thoracic cavity should be considered, but invariably show no evidence of pulmonary metastasis (Lana et al. 1997; Withrow 2007).

procedures are described in detail in the chapter on oral tumors. Surgical technique For nasal planum resection, animals are positioned in sternal recumbency with their head elevated and symmetrically positioned. Systemic analgesia with perioperative nonsteroidal anti-inflammatory drugs and opioids can be supplemented with bilateral infraorbital nerve blocks using bupivacaine (Thomson 2007). The hair (with or without whiskers) is clipped around the nasal planum and lips. Following surgical preparation and draping, the surgical margins should be demarcated with a sterile marker pen. Minimum lateral margins from the visible or palpable extent of the tumor are 5 mm in cats and, preferably, 2 cm in dogs (Thomson 2007). The deep margin preferred is the junction between the cartilaginous and bony nasal tissue. A full-thickness incision is performed with a scalpel blade along the marked lateral margins and deeply through the cartilaginous turbinates, at the level of the nasal and incisive bones. These incisions will result in brisk hemorrhage, which can be controlled with digital pressure and the judicious use of cautery. Following hemostasis, the alar folds can be excised to increase the diameter of the nasal airways and improve the ability to breathe postoperatively. The defect following resection of the nasal planum can be closed with a purse-string suture (Withrow and Straw 1990; Withrow 2007) or the preferred skin-tonasal mucosa closure using a simple interrupted suture pattern of 4-0 or 5-0 nonabsorbable suture material (Figure 5.4) (Thomson 2007).

Nasal planum resection Nasal planum resection is recommended for invasive SCC in cats and dogs (Withrow and Straw 1990; Kirpensteijn et al. 1994; Lana et al. 1997; Thomson 2007; Withrow 2007). Surgery provides excellent local tumor control and has several advantages compared to other treatment options, including the ability to examine surgical margins, wide availability as compared to radiation therapy and photodynamic therapy, affordability, less treatment time, and an acceptable cosmetic outcome (Lana et al. 1997). More aggressive surgical procedures have also been described in dogs with more extensive nasal planum tumors (i.e., nasal planum resection combined with either premaxillectomy or bilateral maxillectomy) (Evans et al. 1985; Lascelles et al. 2004). These

Figure 5.4.  Immediate postoperative appearance following nasal planum resection in the cat from Figure 5.2. Skin-to-nasal mucosa closure is preferred to purse-string suture closure. Also note that excision of the alar folds bilaterally results in widening of the nasal airways.

90  Veterinary Surgical Oncology

Skin-to-mucosa closure is preferred because the incidence of postoperative complications, such as stenosis of the nasal aperture, is decreased. Should the pursestring technique be used, it is recommended to leave the nasal opening larger than planned to allow some further closing of the opening by second intention healing. The deep and lateral margins of the excised nasal planum should be inked and the sample submitted for histopathological assessment of tumor type and completeness of excision. Postoperative management In the immediate postoperative period, analgesia and intravenous fluids should be continued until the animal is eating and drinking voluntarily. Cats will often not want to eat for 1–4 days after nasal planum resection. The return to voluntary eating can be improved with the perioperative use of infraorbital nerve blocks, use of aromatic, warmed, and/or favorite foods, removal of the scab over the surgery site under sedation, and the use of appetite stimulants. Supplemental nutrition through feeding tubes is rarely required. Animals can be discharged when they start to eat voluntarily. A short course of nonsteroidal anti-inflammatory drugs is recommended after discharge for analgesic purposes. A scab or crust usually forms over the surgical site, and this should be carefully removed at suture removal 10–14 days after surgery. Sedation may be required for suture removal.

Figure 5.5.  Cosmetic appearance 6 weeks after nasal planum resection in a cat. Cosmetic results are usually good following nasal planum resection in cats. Note this cat has also had a bilateral pinnectomy for SCC. Up to 30% of cats have multifocal head and neck squamous cell carcinoma. (Image courtesy of Dr. Maurine J. Thomson)

Complications Complications are uncommon. The most significant complication is stenosis of the nasal aperture. This is more common following purse-string closure of the defect. A skin-to-nasal mucosa closure is recommended to minimize the risk of nasal stenosis. If stenosis occurs, then management options include surgical reconstruction with wide skin excision and resection of the rostral nasal septum, laser ablation, insertion of rubber stents, or permanent placement of stainless steel intraluminal expansile stents (Withrow 2007). The prognosis is guarded to good for cats that develop nasal stenosis. Other complications include poor appetite in the initial postoperative period, mild serous nasal discharge, and increased incidence of sneezing (Thomson 2007). The cosmetic appearance is good in cats (Figure 5.5) and fair to good in dogs (Figure 5.6). Adjunctive management Adjunctive treatment is rarely required for cats and dogs with completely excised nasal planum SCC, and most of these patients are considered cured (Lascelles et al. 2000; Thomson 2007). Some authors recommend

Figure 5.6.  Cosmetic appearance following nasal planum resection in a golden retriever. Note that the cosmetic appearance is acceptable but not as good as cats following nasal planum resection.

systemic chemotherapy using either carboplatin or doxorubicin for cats with poorly differentiated SCC or SCC lesions with evidence of lymphatic invasion, regardless of whether they have been completely excised (Thomson 2007). Radiation therapy is recommended

Head and Neck Tumors  91

for incompletely excised nasal planum SCC (Lascelles et al. 2000; Thomson 2007; Withrow 2007). Prognosis The prognosis following nasal planum resection is good to excellent. In cats, local tumor recurrence was reported in two of seven cats (29%) with incompletely excised tumors and no cats with completely resected nasal planum SCC (Lana et al. 1997). In this study, two cats with progressive local disease were euthanatized, and the remainder were cured (Lana et al. 1997). Local tumor control is also excellent in dogs following nasal planum resection, either alone or in combination with either premaxillectomy or bilateral maxillectomy, with no local tumor recurrence or metastasis in dogs with completely excised SCC (Kirpensteijn et al. 1994; Lascelles et al. 2000, 2004). Other treatment options Prevention of precancerous (actinic) lesions The risk of cutaneous SCC can be minimized in animals with poor pigmentation and hair coats by limiting exposure to sunlight. Topical sunscreens are usually ineffective because animals will quickly lick them off (Withrow 2007). Tattooing has limited efficacy as it only protects the deeper dermal layers and not the superficial epidermis. Synthetic vitamin A derivatives, such as isotretinoin or etretinate, increase epithelial differentiation and may reverse or limit the progression of precancerous lesions (Vail and Withrow 2007). However, in one study only 1 of 15 cats with precancerous or SCC lesions responded to isotretinoin therapy (Evans et al. 1985). Cryosurgery Cryosurgery causes tissue destruction with the controlled use of freezing and thawing. Liquid nitrogen and nitrous oxide are the two most commonly used cryogens. Spray freezing is preferred to contact freezing because colder temperatures can be achieved (Thomson 2007). To maximize tissue destruction, the tumor and a minimum 5 mm periphery around the tumor should be rapidly frozen to −20°C and then allowed to thaw slowly. This process should be performed three times in total. The response rate to cryosurgery is dependent on tumor size and location. Response rates are better with small, superficial, and noninvasive lesions (less than 5 mm), and tumors involving the eyelids and pinnae (Clarke 1991; Lana et al. 1997). Multiple treatments are often required for nasal planum lesions. Local tumor control can be good with 84% of 90 cats treated with cryosurgery tumor-free at 12 months and 81% tumor-free at 36 months (Clarke 1991). However, other researchers have

reported local recurrence in 73% of cats following cryosurgery, with a median time to recurrence of 184 days (Lana et al. 1997). Cryosurgery is a readily available and cost-effective treatment modality for small lesions, but disadvantages include the fact that margins cannot be determined and the risk of local recurrence is higher than either nasal planum resection or radiation therapy. Radiation therapy Radiation can be delivered either as local or external beam therapy (Lana et al. 1997). Local radiation therapy with 90strontium plesiotherapy is indicated for cats with superficial but not deep lesions, because strontium only penetrates to a depth of 2 mm (Van Vechten and Theon 1993; Goodfellow et al. 2006). 90Strontium is administered by the local application in five fractions of 10 Gy over a 10-day period for a total dose of 50 Gy (Goodfellow et al. 2006). For appropriate lesions, tumor control is very good, with 85% of 15 cats achieving a complete response after either one (11 cats) or two (2 cats) cycles of radiation therapy. None of these cats had evidence of local recurrence after a median follow-up of 652 days (Goodfellow et al. 2006). Full-course external beam radiation therapy can be used for superficial and deep nasal planum lesions. Orthovoltage, megavoltage, and proton beam irradiation have been described (Theon et al. 1995; Lana et al. 1997; Fidel et al. 2001). Smaller and superficial lesions are more responsive and can be cured with external beam radiation therapy, but response rates and tumor control are decreased for larger and more invasive lesions (Theon et al. 1995). Cure rates for small lesions are 56%, with median disease-free intervals of 12–16 months and median survival times of 361–946 days (Theon et al. 1995; Lana et al. 1997; Fidel et al. 2001). Photodynamic therapy Photodynamic therapy involves the local or systemic administration of a photosensitizer that is preferentially retained by tumor tissue. The subsequent irradiation of the tumor with a light of a wavelength that is absorbed by the photosensitizer results in the formation of oxygenfree radicals and tissue death (Roberts et al. 1991; Peaston et al. 1993; Stell et al. 2001; Buchholz et al. 2007). Similar to other nonsurgical treatment options, photodynamic therapy is recommended only for superficial lesions because the penetration of the wavelength of light used to activate the photosensitizer is limited to 3–4 mm (Buchholz et al. 2007). Complete responses are reported in up to 100% of cats with tumors less than 5 mm, but in less than 30% of cats with larger tumors (Roberts et al. 1991; Peaston et al. 1993; Stell et al. 2001; Buchholz et al. 2007). Complications include local

92  Veterinary Surgical Oncology

tumor recurrence in up to 64% of cats, facial edema, erythema, and necrosis, which resolves slowly over 3–6 weeks (Roberts et al. 1991; Peaston et al. 1993; Stell et al. 2001; Buchholz et al. 2007). Intralesional chemotherapy Intralesional carboplatin has been investigated in cats with SCC of the nasal planum. Twenty-three cats with advanced lesions were treated with intralesional carboplatin (100 mg/m2) resulting in a complete response rate of 73%, 1-year disease-free survival rate of 55%, and local recurrence rate of 30% (Theon et al. 1996). The combination of carboplatin with purified sesame oil reduced systemic absorption and toxicity (Theon et al. 1996). External beam radiation therapy has been combined with intralesional carboplatin in six cats with advanced nasal planum SCC resulting in a complete response in all cats and duration of response for a minimum of 8 months in four of these cats (de Vos et al. 2004).

Tumors of the Pinna Tumors of the pinna are uncommon in both cats and dogs. In dogs, tumors involving the pinna are similar to cutaneous and subcutaneous tumors of other sites, such as mast cell tumors (MCTs) and soft tissue sarcomas (STSs) (Bostock 1986; Vail and Withrow 2007). SCC is the most common tumor of the pinna in cats (Bostock 1986; Vail and Withrow 2007). The gross appearance of these tumors is highly variable, ranging from an ulcerative and erosive appearance of feline SCC lesions (Figure 5.3) to solid masses in many canine pinna tumors (Figure 5.7) (Lana et al. 1997; Ruslander et al. 1997; Vail and Withrow 2007). Diagnosis and clinical staging Definitive preoperative diagnosis is often not required for ulcerative or erosive pinna lesions in cats because of the high likelihood of SCC. However, pinna lesions in dogs should be diagnosed prior to definitive therapy because this will provide valuable information on tumor type, histologic grade, and treatment options. Biopsy options include fine-needle aspiration and incisional or wedge biopsy. Excisional biopsy is not recommended in dogs because of the risk of incomplete resection and tumor recurrence. Staging for metastatic disease is dependent on tumor type. Pinnectomy Partial or total pinnectomy is the recommended treatment for ulcerative and solid tumors of the pinna. Animals are positioned in either sternal or lateral

Figure 5.7.  A MCT along the medial border of the helix of the pinna. Incisional biopsy is recommended for MCTs of the pinna because the extent of surgical margins can be determined by histologic grade. In this case, the tumor was excised with 2 cm margins, as indicated by the sterile marker pen.

recumbency, depending on tumor location and surgeon preference. The medial and lateral aspects of the pinna are clipped and aseptically prepared. Chlorhexidine preparations should be used with caution, especially in cats, because of the risk of ototoxicity (Igarashi and Suzuki 1985). Following draping, the surgical margins can be marked with a sterile marking pen (see Figure 5.7). The surgical margins are dependent on tumor type and, for canine MCT, histologic grade: 1 cm margins are recommended for SCC in cats (Figure 5.8A), benign tumors in dogs, and grade I MCT; 2 cm margins for grade II MCT; and 3 cm margins for canine STS and grade III MCT (Lana et al. 1997; Ruslander et al. 1997; Simpson et al. 2004; Vail and Withrow 2007). The medial and lateral skin and auricular cartilage are incised along the marked margins with a scalpel blade. For tumors involving the apex of the pinna, or both the lateral and medial borders of the helix of the pinna, the pinna is amputated. Partial pinnectomy without amputation is possible for lesions involving either the medial or lateral border of the helix of the pinna. Depending on the location of the tumor and extent of excision, ligation or cauterization of the greater auricular arteries and veins may be required. The auricular cartilage should be trimmed to permit closure of the skin edges over the cartilage without tension. The skin edges are closed in

Head and Neck Tumors  93

(a)

Figure 5.9.  Immediate postoperative appearance of the reconstructed pinna following MCT resection in Figure 5.7. Pinna reconstruction following partial pinnectomy for tumors involving the medial or lateral borders of the helix of the pinna can be inventive. Cosmetic results are often good.

(b)

Figure 5.8.  (A) Pinnectomy in a cat with multifocal head and neck SCC. Note that preauricular SCC lesions have also been excised. (B) Closure of the pinna involves suturing of the skin over the auricular cartilage. The cartilage should be trimmed if it increases tension on the wound closure. (Images courtesy of Dr. Maurine J. Thomson)

one or two layers with the subcutaneous tissue closed in a simple continuous pattern using absorbable monofilament suture material and the skin in either a simple interrupted or continuous pattern with nonabsorbable monofilament suture material (Figure 5.8B). Following partial pinnectomy of tumors involving the lateral or medial borders of helix of the pinna, reconstruction of the pinna depends on the defect size and location. This reconstruction is often inventive but needs to be planned prior to surgery to ensure sufficient skin has been clipped and prepared to allow reconstruction (Figure 5.9). The margins should be inked and the tumor submitted for histopathologic assessment of tumor type and completeness of surgical excision (Figure 5.10). Complications following pinnectomy are uncommon and include hemorrhage, wound dehiscence, and local

Figure 5.10.  Partial pinnectomy of the affected pinna in Figure 5.3. Minimum margins of 1 cm are recommended to minimize the risk of incomplete excision and tumor recurrence.

tumor recurrence. Local tumor control is good following complete excision of feline pinna SCC, with tumor recurrence or de novo tumor development in 23% of cats (Atwater et al. 1991). However, repeat excision or adjunctive treatment is recommended following incomplete resection to decrease the risk of local recurrence (Atwater et al. 1991). Cosmetic results are usually acceptable (Withrow and Straw 1990). Adjunctive treatment Adjunctive treatment options for cats with incompletely excised pinna SCC include cryosurgery, photodynamic

94  Veterinary Surgical Oncology

therapy, 90strontium plesiotherapy, and external beam radiation therapy (Clarke 1991; Peaston et al. 1993; Lana et al. 1997; Stell et al. 2001; Vail and Withrow 2007). Cryosurgery can also be used as the primary treatment for small and superficial lesions of the feline pinna (Clarke 1991). Adjunctive treatment options in dogs depend on the completeness of excision and tumor type. External beam radiation therapy should be considered for incompletely excised MCT and STS, and systemic chemotherapy is recommended for high-grade or metastatic MCT (Vail and Withrow 2007).

Tumors of the External Ear Canal Tumors of the external ear canal are uncommon in the dog and cat. Patients typically present with clinical signs of otitis externa that are poorly responsive to medical management. Clinical signs include otic discharge, head shaking, scratching at the ear, and the presence of a mass. Hemorrhagic discharge can be indicative of trauma or neoplasia (Lanz and Wood 2004). Pain on opening the mouth and the presence of neurologic deficits, including facial nerve paresis, Horner’s syndrome, and vestibular disease (head tilt, ataxia, nystagmus), is consistent with middle ear involvement. Ten percent of dogs with malignant tumors and 25% of cats with either benign polyps or malignant tumors have neurologic deficits (ter Haar 2006). Clinical signs may be present for weeks to years prior to presentation. Otitis externa may develop secondary to obstruction of the ear canal; however, an association has also been found between chronic otitis and the secondary development of neoplastic lesions (Rogers 1988; London et al. 1996; Moisan and Watson 1996; Zur 2005). If a patient no longer responds to medical management, the presence of a mass (neoplastic, nonneoplastic, or a foreign body) must be considered (Rogers 1988). Bilateral involvement does not exclude the possibility of neoplasia. There are reports of cats and dogs bilaterally affected with ceruminous gland adenocarcinoma or squamous cell carcinoma (Theon et al. 1994; Bacon et al. 2003; Zur 2005). Most patients are middle-aged and older (Rogers 1988; ter Haar 2006). The mean age of dogs with malignant tumors is 9.9 years (range 4–18 years) and of dogs with benign tumors is 9.4 years (range 4–18 years). The mean age of cats with malignant tumors is 11 years (range 3–20 years), and for benign tumors it is 6.9 years (0.5–15 years) (London et al. 1996). In another study, the mean age for malignant tumors was 9.8 years and for benign tumors was 7.7 years (Bacon et al. 2003). Cocker spaniels are overrepresented for both benign and

malignant tumors, likely due to their propensity for otitis externa (London et al. 1996). Historically, it has been reported that the majority of canine tumors are benign and the majority of feline tumors are malignant (Rogers 1988; Theon et al. 1994). However, there have been reports with 50% to the majority of canine tumors also being malignant (London et al. 1996; Moisan and Watson 1996). In dogs, malignant tumors are locally invasive with a low incidence of metastatic disease at the time of presentation (London et al. 1996). Cats tend to have more aggressive malignant tumors than the dog. They may invade the middle ear (Rogers 1988) and often invade the wall of the ear canal extending into the adjacent soft tissue. Metastatic rates of 0%–50% have been reported with metastases to regional lymphatics, lungs, and distant viscera (London et al. 1996; Bacon et al. 2003; Moisan and Watson 1996). Most cats die due to tumor recurrence or progression of local disease. Once local disease is better controlled, it is possible that metastatic disease may become more prevalent (London et al. 1996). Ceruminous gland tumors are the most common tumor of the canine and feline external ear canal (Carlotti 1991; Rogers 1988; Bacon et al. 2003). Malignant tumors of dogs and cats include ceruminous gland adenocarcinoma, SCC, and carcinoma of undetermined origin (Figure 5.11). Other malignant tumors affecting dogs include round cell tumor, sarcoma, malignant melanoma, fibrosarcoma, mast cell tumor, leiomyosarcoma, plasmacytoma, and hemangiosarcoma. Benign tumors of dogs and cats include inflammatory polyps, ceruminous gland adenoma, basal cell tumor, and papillomas. Histiocytoma, plasmacytoma, benign melanoma, and fibroma have also been reported in the dog (London et al. 1996; Lucke 1987; Rogers 1988; Bacon et al. 2003; ter Haar 2006). SCC is rare in the external ear canal; however, when present it typically involves the external auditory meatus and has particularly aggressive behavior with local invasion, regional lymph node involvement and, later, pulmonary metastases (Rogers 1988; Salvadori et al. 2004). Intraorbital and orbital metastases have also been documented (Hayden 1976). In one case report, two cats that had previous bilateral pinnal amputations for SCC developed SCC within the external ear canal and metastatic meningeal carcinomatosis (Salvadori et al. 2004). Although not neoplastic, inflammatory polyps are the most common mass lesion affecting the external ear canal of the cat (Lanz and Wood 2004). They affect the external ear canal by extension from the middle ear (see the section on tumors of the middle ear, below). Ceruminous gland cysts are nonneoplastic masses that can involve the external ear canal of the cat (Carlotti

Head and Neck Tumors  95

(a)

(b)

Figure 5.11.  (A) Ceruminous gland adenocarcinoma in the horizontal ear canal of a cat, following total ear canal ablation. (B) Intraoperative picture of concurrent lymph node metastasis.

(a)

(b)

Figure 5.12.  (A) Ceruminous gland cysts affecting the pinna and external ear canal of a 13-year-old spayed domestic short-haired cat. (B) Pinnectomy and total ear canal ablation and lateral bulla osteotomy were performed. Note the extension of the cysts from the vertical ear canal to the pinna with secondary hyperplasia and exudate within the vertical ear canal. The horizontal ear canal was also resected, but not included in this image.

1991; Rogers 1988). Occasionally, they will extend on to the pinna (Figure 5.12). Patients are most commonly 2–15 years of age, with multiple masses 1–5 mm in size and containing black fluid. They are black fluid-filled cysts 1–5 mm in size that can extend from the ear canal on to the pinna. They have been reported in cats 2–15 years of age. Both ears may be affected. They are best treated with surgical resection. Diagnosis Cytology is useful to evaluate concurrent infection with bacteria or yeast; however, it rarely yields neoplastic

cells. Tumors do not exfoliate cells readily into the ear canal, and they are masked by inflammatory debris (De Lorenzi et al. 2005; Zur 2005). To look for a mass lesion, the ear is gently lavaged with warm sterile saline and aspirated with a soft, rubber catheter (De Lorenzi et al. 2005). In severe cases, sedation or general anesthesia may be necessary (Lanz and Wood 2004). Otoscopy is then performed. The masses can be firm or friable, pedunculated or invasive. Pedunculated masses are typically benign, and malignant masses are typically broadbased (Rogers 1988; London et al. 1996). A raised or ulcerated appearance does not differentiate benign from malignant (London et al. 1996). Ceruminous adenocarcinoma and SCC are more likely to be invasive, and there may be a hemorrhagic otic discharge. Following identification of the mass, a fine-needle biopsy (FNB) of the mass should be performed with a 22- or 23-gauge hypodermic or spinal needle (De Lorenzi et al. 2005). In one study, looking at 27 cats with tumors of the external ear canal, cytology from FNB correctly differentiated inflammatory polyps and mast cell tumors from ceruminous gland adenoma and adenocarcinoma. This differentiation is useful as inflammatory polyps may be treated with traction-avulsion or a ventral bulla osteotomy, whereas the other tumor types would require a total ear canal ablation and lateral bulla osteotomy. The needle is placed into the mass and then withdrawn; syringe aspiration is not necessary. Scraping and impression smears are other techniques for cytologic assessment if the yield is inadequate with FNB alone (Rogers 1988). An alternative is to obtain a small tissue sample (grab biopsy; Bacon et al. 2003) with

96  Veterinary Surgical Oncology

alligator forceps or a snare (Rogers 1988). For deep masses, an incision into the vertical ear canal can be performed (ter Haar 2006). With this technique, however, there is risk of contaminating adjacent normal tissue, and with the availability of video otoscopes, this technique is rarely necessary. Following definitive mass removal, histopathology should be performed in all cases to confirm the diagnosis and assess invasion into surrounding structures (De Lorenzi et al. 2005). In patients with malignant tumors, evaluation of regional lymph nodes and thoracic radiographs are also recommended (Theon et al. 1994). Historically, bulla radiographs (oblique lateral and open-mouth views) have been performed when there is suspicion of middle or inner ear involvement. However, abnormal findings, such as a soft-tissue density within the tympanic cavity, osseus thickening of the bulla, and sclerosis of the petrous temporal bone, are not specific for tumor ingrowth. They could also be consistent with secondary infection or inflammation (Rogers 1988; Marino et al. 1993). In one study, the presence of these radiographic changes in dogs with ceruminous gland adenocarcinoma did not affect outcome (Marino et al. 1993). CT or MRI is preferred to rule out destruction of the bulla and invasion into adjacent soft tissues, as this negatively affects prognosis and may highlight the need for more aggressive surgery and/or further treatment such as radiation therapy. CT has been shown to be more accurate than radiographs in patients with moderate to severe ear disease; however, in patients with milder disease, the accuracy of both CT and bulla radiographs is more variable (Rohleder et al. 2006). Treatment Surgical excision is the treatment of choice (Rogers 1988). Surgical procedures include lateral ear canal resection, vertical ear canal resection, and total ear canal resection with lateral bulla osteotomy (TECA and LBO). Lateral and vertical ear canal resections have limited application and should be restricted to benign tumors confined to the lateral or vertical ear canal (Figure 5.13). A CT or MRI should be done to ensure that the horizontal ear canal is not involved (Lanz and Wood 2004). TECA and LBO are recommended for any patient with a malignant tumor of the ear canal (Moisan and Watson 1996; ter Haar 2006). In one study of 11 dogs with ceruminous gland adenocarcinoma, four patients with tumors of the vertical ear canal treated with lateral ear canal resection had a 75% recurrence rate compared to no recurrence in the seven patients treated with TECA and LBO (Marino et al. 1993). Similarly, in one study of 22 cats with ceruminous gland adenocarcinoma, cats

(a)

(b)

Figure 5.13.  (A) Vertical ear canal resection for ceruminous adenoma of the vertical ear canal of a dog. (B) Postoperative specimen. Note the focal involvement of the vertical ear canal, allowing this more limited surgical procedure.

with tumors involving the lateral aspect of the vertical ear canal had a lateral ear canal resection (Marino et al. 1994). Sixty-six percent of the tumors recurred, with only 2 of the 6 cats alive at 1 year. The median diseasefree interval was 10 months (range 1–14). This is compared to cats with a TECA and LBO with a 42-month median survival (range 4–60), 25% recurrence rate, and 12 of 16 alive at 1 year. As there is little indication for a lateral and vertical ear canal resection, the discussion will be limited to TECA and LBO. A complete blood count and biochemical profile should be performed prior to surgery. Patients with chronic otitis externa may be hypergammaglobulinemic with a compensatory hypoalbuminemia. In rare cases, fresh-frozen plasma or synthetic colloids, such as pentastarch, may be needed to help support blood pressure. Patients with chronic disease may also have significant intraoperative hemorrhage. For patients with bilateral disease, it may be preferable to stage the

Head and Neck Tumors  97

procedures several days to weeks apart. This reduces the risk of postoperative ventilatory compromise, secondary to edema of the surgery site, when both ears have surgery simultaneously. The hematocrit and total solids should be assessed before proceeding with surgery of the second ear. Patients should have an opiate such as hydromorphone or morphine included in their premedication. It may also be administered during surgery for additional analgesia. The entire pinna and lateral aspect of the head is clipped and prepared for aseptic surgery. Prophylactic intravenous antibiotics are initiated prior to making the skin incision. The patient is placed in lateral recumbency with a towel placed under the head. A circular incision is made with a scalpel blade around the opening of the vertical ear canal encompassing all hypertrophied tissue. Be cautious with the rostral and caudal aspects of the incision to ensure that the blood supply to the pinna is not compromised. The incision is extended through the auricular cartilage. The incised auricular cartilage is grasped with a towel clamp and retracted toward the surgeon, exposing the external wall of the medial aspect of the vertical ear canal. The muscle is closely dissected from the medial aspect of the ear canal down to the level of the horizontal ear canal. Dissection around the vertical ear canal is then initiated from the exposed medial aspect and continued laterally, alternating in a rostral and caudal direction. Exposure is improved by applying tension on the ear canal away from the area of dissection (Lanz and Wood 2004). The periauricular tissues should be closely dissected, with Metzenbaum and iris scissors, from the perichondrium to prevent trauma to the facial nerve. The facial nerve exits the skull through the stylomastoid foramen, caudal to the external acoustic meatus, and courses caudal and ventral to the horizontal ear canal near its junction with the vertical ear canal. Gentle tissue retraction should be employed, with retractors placed superficial to the facial nerve to prevent iatrogenic trauma, as it may be entrapped in adjacent fibrous tissue (Smeak and Inpanbutr 2005). Bipolar cautery should be used for hemostasis. The ear canal is transected at the junction of the horizontal and osseus external ear canal. Transection can be performed by a stab incision with a scalpel blade, scissors, and by twisting the ear canal. Rongeurs may be needed if the ear canal is ossified (Smeak and Inpanbutr 2005). Instruments should be directed in a caudal-tocranial direction away from the facial nerve. Following removal of the ear canal, the osseus ear canal should be removed with rongeurs. Care is taken to avoid trauma to the branches of the external carotid artery and the retroglenoid vein ventral to the bulla (Lanz and Wood 2004). If hemorrhage occurs, the surgery site should be

filled with saline and packed with a gauze square for 5 minutes. The alternative is to apply focal pressure to the origin of the bleeding with a cotton-tipped applicator. Bone wax can then be placed (Smeak and Inpanbutr 2005). The tissues on the lateral aspect of the bulla are then gently elevated with a periosteal elevator. The lateral and ventral aspects of the bulla are removed, ensuring good access to the caudal aspect of the tympanic cavity (Smeak and Inpanbutr 2005). When very thickened, this can be facilitated by orienting the ronguers in a caudodorsal-to-cranioventral direction. Do not use ronguers on the rostral aspect of the bulla to avoid trauma to the epitympanic recess (Smeak and Inpanbutr 2005). It is imperative that all of the debris and epithelium of the osseus external ear canal and bulla be completely removed to prevent chronic abscessation/fistulation. The lining of the bulla is gently elevated with a curette and, when thickened, can be grasped with a rongeur for removal. This will expose the white, shiny medial wall of the bulla. The bulla of the cat is divided into a ventromedial and a dorsolateral compartment. It is imperative that both compartments be evaluated. Be careful with dorsal and dorsomedial curettage to avoid damage to the sympathetic trunk and round window. Abnormal tissue hanging dorsally can be gently grasped with a small curved hemostat (Smeak and Inpanbutr 2005) and teased off. The tympanum may have rolled craniodorsally with the malleus and can be removed by gently grasping craniodorsally into the epitympanic recess (McAnulty et al. 1995). A swab of the epithelium and contents of the bulla should be submitted for aerobic and anaerobic culture and sensitivity. Samples from the bulla and external ear canal are submitted for histopathology. The surgical site is gently lavaged with sterile saline. Some authors indicate that a drain is not necessary if there has been meticulous hemostasis, debridement of devitalized tissue, and good tissue apposition (Devitt et al. 1997). Furthermore, if there is any risk that excision is incomplete, a passive drain is contraindicated as it would allow tumor seeding into adjacent tissue. The subcutaneous tissues are closed with 2-0 or 3-0 absorbable suture in a simple interrupted pattern. Care is taken to ensure that the facial nerve and drain are not entrapped in the closure. The superficial subcutaneous tissues are sutured over the auricular cartilage with 3-0 or 4-0 absorbable suture in a simple continuous pattern. The skin is apposed with nonabsorbable suture (3-0 or 4-0) in a simple interrupted or cruciate pattern. It is important that the cartilage is covered to prevent granulation tissue formation. A soft, padded bandage can be placed following surgery. The affected pinna is placed over the dorsum of

98  Veterinary Surgical Oncology

the head with several gauze squares placed between the pinna and the head. A secondary layer of gauze roll is applied loosely alternating cranial and caudal to the opposite ear. This is followed with a tertiary layer such as Vetrap. The location of the pinna should be drawn on the bandage to prevent inadvertent trauma during bandage removal. The bandage is changed daily and removed 2 days after surgery. An Elizabethan collar is indicated if the patient is traumatizing the incision. A modification of the above technique is recommended for animals with erect ear carriage. Using new instruments and gloves, an advancement flap is made rostrolateral to the original circular incision. Two parallel incisions are extended rostrally from the circular incision. The skin is elevated and apposed to the caudal aspect of the circular incision. Following surgery, the ear is bandaged in an erect position with rolled gauze placed on the concave aspect of the pinna.

morphine constant rate infusion with lower sedation scores. Filters are placed in the line, and the catheter must be handled aseptically to prevent the introduction of bacteria. Further studies are needed to evaluate acceptable administration rates that would increase the likelihood of analgesia, without risk of systemic toxicity and with minimal wound complications (Wolfe et al. 2006). Bolus injections of bupivacaine through the catheter every 6 hours may be more successful. There is no evidence that the local administration of anesthetic agents caused an increased incidence of facial nerve deficits or ototoxicity (Wolfe et al. 2006). If there is evidence of postoperative facial nerve paresis or paralysis, eye lubricant should be administered for several days following surgery. This is not typically needed long term as tear production and third eyelid function are usually unaffected (White and Pomeroy 1990). Patients with concurrent keratoconjunctivitis sicca will require permanent treatment.

Postoperative management

Complications

Patients should be treated empirically with antibiotics until culture and sensitivity results are obtained. An appropriate antibiotic is then administered for 2–4 weeks following surgery. Mixed bacterial populations are frequent (Devitt et al. 1997). Common isolates include Staphylococcus intermedius, Pseudomonas aeruginosa, β-hemolytic Streptococcus, Proteus spp., Streptococcus canis, and Eschirichia coli (Devitt et al. 1997; Lanz and Wood 2004). Analgesia consists of injectable opiates throughout surgery and immediately postoperatively. Hydromorphone or morphine may be administered every 4–6 hours or as a continuous rate infusion for 24–48 hours following surgery (Lanz and Wood 2004). A fentanyl patch may be placed 12–24 hours prior to surgery (Lanz and Wood 2004). Nonsteroidal anti-inflammatory medication and tramadol or codeine should be administered for 5 days following surgery (Lanz and Wood 2004). Local infusion of anesthetic agents, such as intraoperative splash block, nerve blocks, and by continuous infusion into the wound, have been investigated. They are inexpensive, reduce the risk of break-through pain, and provide local pain control without systemic side effects such as sedation. In one study, patients were easier to manage under anesthesia with preoperative local nerve blocks with bupivacaine (Buback et al. 1996); however, their postoperative evaluation did not differ from patients receiving perioperative opiates or opiates and a bupivacaine splash block. In another study, continuous local infusion of lidocaine via an indwelling wound catheter provided comparable pain relief to a

In the immediate postoperative period, there can be swelling of the surgery site. This, in combination with the bandage, can cause difficulty breathing, particularly in brachycephalic patients. Patients should be monitored closely and the bandage loosened or removed if there is any concern. The risk of this is higher in patients that have had concurrent bilateral TECA and LBO. Other complications include infection of the surgery site, necrosis of the pinna (Matthiesen and Scavelli 1990; Lanz and Wood 2004), hemorrhage, and neurologic dysfunction, which can include Horner’s syndrome, facial nerve paresis/paralysis, and vestibular disease. Necrosis of the pinna typically occurs along the caudal pinna margin. It is treated with debridement and open wound management (Matthiesen and Scavelli 1990; Lanz and Wood 2004). Preexisting neurologic deficits are often permanent. Following surgery, the incidence of facial nerve paresis/ paralysis in dogs is 5%–50%, with most cases resolving within 4 weeks. The cat has a higher incidence of neurologic dysfunction when compared to the dog and a higher incidence of permanent facial nerve deficits despite meticulous dissection (Matthiesen and Scavelli 1990; Bacon et al. 2003; Lanz and Wood 2004). Horner’s syndrome tends to occur only in the feline patient and is less likely to resolve (Matthiesen and Scavelli 1990; Lanz and Wood 2004). It has been reported that in cats undergoing TECA and LBO, 50% developed facial nerve paralysis and 42% developed Horner’s syndrome following surgery (Bacon et al. 2003). The majority of cases resolved within 1 month.

Head and Neck Tumors  99

Hearing loss is expected in all patients following surgery. Patients with significant ear disease may have been deaf prior to surgery. Owners should be warned of this in patients with bilateral disease. When evaluated by brain stem–evoked audiometry (BERA), patients undergoing TECA and LBO had complete hearing loss. The only patients in which hearing was maintained were patients that had the tympanic membrane and ossicles retained (McAnulty et al. 1995). This is rarely clinically acceptable as the retention of the tympanic membrane risks the development of otitis media and the late development of fistulous tracts (ter Haar 2006). A significant complication following surgery is the development of otitis media and a fistulous tract. This has been reported in 5%–10% of cases (Lanz and Wood 2004). This may appear 3–12 months after surgery and is attributed to epithelium being left in the bulla or external osseus ear canal (Matthiesen and Scavelli 1990). Other postulated causes include osteomyelitis of the auditory ossicles and inadequate drainage of the middle ear through the auditory tube (Lanz and Wood 2004). Risk is reduced by removing the tympanic membrane and the majority of the lateral and ventral walls of the bulla, allowing ingrowth of vascularized soft tissue (McAnulty et al. 1995). It is the removal of the bony wall of the lateral bulla that is more useful, rather than relying solely on aggressive curettage. There may be transient improvement with long-term antibiotics; however, a lateral or ventral bulla osteotomy to remove the epithelium is preferable. Prognosis In dogs, most tumors are characterized by local invasion and a low incidence of metastasis. Patients with tumors confined to the ear canal have a good prognosis with TECA and LBO. A median survival time of greater than 58 months has been reported in dogs with malignant aural tumors (London et al. 1996). In a study of 7 dogs with ceruminous gland adenocarcinoma, with a median follow-up of 36 months (8–72 months), there were no reports of local recurrence or metastatic disease (Marino et al. 1993). Patients with SCC have a poorer prognosis. In another study, 3 of 23 dogs with ceruminous gland adenocarcinoma died from their tumor, compared to 4 out of 8 with SCC (London et al. 1996). In a third study, local tumor control was obtained with ear canal ablation in 6 dogs, however, 2 (1 with ceruminous gland adenocarcinoma and 1 with SCC) developed pulmonary metastases 10 months and 6.5 months, respectively, following surgery (Matthiesen and Scavelli 1990). The mean follow-up time was 18 months (12–44 months).

Bulla involvement is a negative prognostic indicator in dogs. Dogs with tumor invasion into the bulla had a significantly reduced median survival time of 5.3 months, whereas those with tumors confined to the vertical or horizontal ear canals survived longer than 30 months (London et al. 1996). Extension into the bulla did not appear to affect outcome in another study, however, histopathology was not done to differentiate secondary infection from tumor ingrowth (Marino et al. 1993). Cats tend to have more aggressive tumors than dogs (London et al. 1996). In one study, cats with malignant aural tumors had a median survival time of 11.7 months (London et al. 1996). Negative prognostic indicators included preexisting neurologic deficits, histologic evidence of invasion, and a histologic diagnosis of SCC or carcinoma of undetermined origin versus adenocarcinoma (London et al. 1996; Bacon et al. 2003). Cats with neurologic deficits at presentation had a significantly shorter median survival time of 1.5 months compared to 15.5 months in cats without neurologic deficits. Cats with histologic evidence of invasion had a median survival time of 4 months versus 21.7 months for cats that did not have invasion. Cats with ceruminous gland adenocarcinoma lived significantly longer than cats with SCC (median of 49 months versus 3.8 months) and carcinoma of undetermined origin (5.7 months). Cats with ceruminous gland adenocarcinoma, treated with TECA and LBO, had a median disease-free interval of 42 months, survival time of 50 months, and a recurrence rate of 25% with 75% alive at 1 year (Marino et al. 1994; Bacon et al. 2003). In another study of 15 cats, 7 had ceruminous gland adenocarcinoma; of those 7 cats, 3 were dead by 6 months (Williams and White 1992). The remaining patients had no evidence of recurrence 6 months after surgery. High mitotic index (≥3 mitotic figures per highpower field) is a negative prognostic indicator in cats with ceruminous gland adenocarcinoma (Bacon et al. 2003). Histopathologic grade (well, intermediately, and poorly differentiated) did not correlate with outcome. Adjunctive treatment In people, total ear canal ablation is most likely to be curative when combined with external beam radiation therapy, particularly with aggressive or very invasive tumors (Zur 2005). There is little information in the veterinary literature regarding outcome from adjuvant therapies. Radiation therapy is recommended for patients where surgical resection is incomplete or where surgical resection is declined (Theon et al. 1994). CT or MRI is recommended for radiation therapy planning (Theon et al. 1994). Megavoltage radiation is

100  Veterinary Surgical Oncology

recommended in preference to orthovoltage as there is better delivery of radiation to the deeper structures (bulla and vestibular apparatus), reducing the rate of recurrence (Theon et al. 1994). Acute reactions are mild in most patients, and in those with later recurrence, further intervention, including surgery and a second course of radiation therapy, is well-tolerated (Theon et al. 1994). In one study, three patients developed neurologic signs after radiation therapy. This was associated with tumor recurrence in two. In one patient, the development of Horner’s syndrome may have been due to radiation injury; however, the dog had long-term tumor control. Cranial and peripheral nerves are not normally considered major dose-limiting tissues, whereas the brain and spinal cord are. This study used a MondayWednesday-Friday schedule (Rogers 1988; Theon et al. 1994). New protocols with more frequent dosing (smaller individual doses) and a higher overall dose may reduce the rate of recurrence. There is little information in the literature regarding chemotherapy. As local control continues to improve, metastatic disease may become more prevalent.

Tumors of the Middle Ear Primary tumors of the middle ear in small animals are rare, with local extension of tumors from the external ear canal being more common (Rogers 1988; Little et al. 1989; ter Haar 2006). Tumors of the middle ear are typically found in older patients, with some reports supporting a female sex predisposition (Lane and Hall 1992; Lucroy et al. 2004). In one study, the median age was 10.25 years (7–14 years) (Trevor and Martin 1993). Clinical signs include hemorrhagic or serosanguinous otic discharge and pain on opening the mouth (Indrieri and Taylor 1984; Pentlarge 1984; ter Haar 2006). Pain on opening the mouth has been attributed to extension and/or involvement of the temporomandibular joint (Lane and Hall 1992). Neurologic deficits are common, including facial nerve paralysis, Horner’s syndrome, head tilt, nystagmus, and ataxia. Patients with tumor extension into the nasopharynx may present with increased chronic nasal discharge, respiratory noise or compromise, dysphagia, and exercise intolerance (Trevor and Martin 1993; Bradley 1984; Lanz and Wood 2004; ter Haar 2006). A similar mass may also be identified in the external ear canal. Clinical signs may be present for weeks or months. Para-aural abscessation has also been described. Metastases to the eye, brain, and larynx have been reported (Lane and Hall 1992). Papillary adenomas have been documented in the middle ear of dogs (Little et al. 1989). Aggressive

neoplasia is rare in dogs. Cholesteatoma, an epidermoid cyst, can be misinterpreted as an aggressive tumor. It is an accumulation of keratinized debris in the middle ear of dogs (Hardie et al. 2008). It is typically associated with chronic otitis externa/media. Patients with more advanced disease can present with pain opening their mouth and neurologic signs, including head tilt, facial nerve paresis, and ataxia. With CT, some patients will have marked expansion of the bulla and lysis of the squamous or petrosal portions of the temporal bone. Contrast medium enhancement has also been observed that can be confused with a neoplastic process. Preoperative biopsies may be considered (transpalatal aspirate from the expanded bulla or a ventral approach). A ventral bulla osteotomy (VBO) or TECA and LBO should be performed, with tissue from the middle ear submitted for histopathology and culture and sensitivity. Alternatively, a caudal auricular approach to the tympanic bulla for removal of a cholesteotoma has also been described in a dog, allowing preservation of conduction potentials on BERA response tests (Hardie et al. 2008). Early surgery can be curative; however, recurrence has been reported. Complete removal of the epithelium can be more difficult in these patients as the epithelium is adhered to the invaginated bone. SCC is the most common tumor type in cats (Stone et al. 1983; Rogers 1988; Lane and Hall 1992). Fibrosarcoma, adenocarcinoma, and lymphosarcoma have also been documented in the middle ears of cats (Rogers 1988; Lane and Hall 1992; Trevor and Martin 1993). In one review of 11 cats, 54% of the tumors were SCC; 10 of 11 cats had Horner’s syndrome and/or vestibular disease; and 45% had oral signs, including pain on opening their mouths and/or dysphagia (Lucroy et al. 2004). In one report of 5 cats, all had facial nerve deficits and most demonstrated pain on opening the mouth. Other clinical signs included ataxia, head tilt, nystagmus, and Horner’s syndrome (Rogers 1988; Trevor and Martin 1993). Nasopharyngeal polyps (NPs) are the most common mass lesion in the feline middle ear (Rogers 1988; Lanz and Wood 2004) and should be differentiated from neoplastic conditions. They typically affect younger patients (less than 2 years of age) (Bradley 1984; Lanz and Wood 2004); however, they have been reported in patients up to 15 years of age (Rogers 1988). In one study, the mean age of cats with NP was 1.5 years (6 months–5 years) (Trevor and Martin 1993). Patients can be bilaterally affected (Trevor and Martin 1993). It has been proposed that they are secondary to respiratory tract infections with subsequent otitis media (Rogers 1988; Trevor and Martin 1993). Feline calicivirus has been thought to be an inciting agent (Lanz and Wood 2004). A congenital

Head and Neck Tumors  101

MRI should be interpreted with caution following surgery as scar tissue has been shown to enhance contrast and may be misinterpreted as residual disease. In people, follow-up studies, to assess completeness of surgical excision, are performed within 5–7 days of surgery to reduce this risk (Lucroy et al. 2004). It is important to do a thorough neurologic evaluation prior to surgery and to warn owners that preexisting neurologic deficits may not resolve with surgery. Treatment

Figure 5.14.  Traction-avulsion of a nasopharyngeal polyp extending into the external ear canal of a 4-year-old male neutered domestic short-haired cat. A ventral bulla osteotomy was performed to remove the polyp from the middle ear. Nine months previously, a nasopharyngeal polyp had been removed from the pharynx with traction-avulsion.

cause has also been postulated in some cases (Lanz and Wood 2004). NPs can arise in the middle ear, auditory tube, or nasopharynx with extension into the nasopharynx and/or external ear canal (Lanz and Wood 2004). Clinical signs include otic discharge, head shaking, head tilt, or a visible mass. Patients with involvement of the nasopharynx may present with increased chronic nasal discharge, respiratory noise or compromise, dysphagia, and exercise intolerance (Bradley 1984; Trevor and Martin 1993; Lanz and Wood 2004). NPs are typically pedunculated and can be pale gray, white, or pink (Lanz and Wood 2004) (Figure 5.14). Diagnosis Skull radiographs may demonstrate a soft-tissue density in the bulla and/or nasopharynx (Bradley 1984; Trevor and Martin 1993; Lanz and Wood 2004). Lateral oblique and open-mouth views should be performed to evaluate the bulla (Bradley 1984; Lanz and Wood 2004). The bulla is typically thickened with a fluid density within it (Trevor and Martin 1993). Soft-tissue swelling adjacent to the bulla and marked bony destruction of the tympanic bulla, petrous temporal bone, zygomatic arch, and temporomandibular joint have been described (Indrieri and Taylor 1984; Pentlarge 1984; Lane and Hall 1992; Trevor and Martin 1993). Advanced imaging, such as CT or MRI, is particularly important in patients with neurologic deficits (Horner’s or peripheral vestibular syndrome) to assess invasion into adjacent structures. This allows for appropriate surgical planning. CT or

VBO is recommended for diseases confined to the middle and inner ear. There is good exposure of the tympanic cavity, and gravity assists drainage (Trevor and Martin 1993). Total ear canal ablation is recommended in patients with involvement of the external, middle, and inner ear (Trevor and Martin 1993). In patients with middle ear neoplasia that has extended beyond the bulla, surgery can be combined with a rostrotentorial craniectomy, which appears to improve outcome (Lucroy et al. 2004). Mortality in people is typically due to intracranial extension of the tumor (Stone et al. 1983). Follow-up with radiation therapy is recommended, although an increase in survival time is not always achieved (Stone et al. 1983; Rogers 1988). Some patients may also benefit from chemotherapy (ter Haar 2006). Ventral bulla osteotomy A rostral-to-caudal incision is made over the bulla, which can be palpated in the cat, caudal and medial to the vertical ramus of the mandible. The platysma muscle is incised, and the linguofacial vein is retracted. The digastricus muscle is dissected from the hyoglossal and styloglossal muscles. The hypoglossal nerve is on the lateral aspect of the hypoglossal muscle and should be protected. The muscles are distracted with self-retaining retractors, exposing the ventral aspect of the bulla. An osteostomy is performed with a Steinmann pin or burr. The opening is enlarged with rongeurs. Samples are removed from the bulla for histopathology, cytology, and culture and sensitivity (aerobic and anaerobic). In one study, four of seven cats with NPs and one of four cats with tumors of the middle ear had positive cultures (Trevor and Martin 1993). The bulla of the cat is divided into a ventromedial and a dorsolateral compartment by an incomplete bone shelf. It is imperative that both compartments be evaluated. It is recommended to be cautious with dorsal and dorsomedial curettage to reduce trauma to the sympathetic nerve fibers as they run along the promontory along the dorsomedial aspect of the bulla and the round window. The bulla is lavaged with warm sterile saline. A drain should not be placed in the

102  Veterinary Surgical Oncology

Figure 5.15.  Traction-avulsion of a large nasopharyngeal polyp in a cat. Note the intratracheal urinary catheter used to insufflate oxygen during polyp removal. The size of the polyp had precluded intubation with an endotracheal tube. Following polyp removal the cat was intubated and recovered without incident.

case of malignant tumors as it may track neoplastic cells outside of the surgical field. The subcutaneous tissues and skin are closed routinely. If a drain is placed, a soft padded bandage should be placed to protect the drain and to qualify and quantify wound drainage. The surgical approach is the same in dogs; however, the bulla cannot be palpated externally. Following surgery, animals should receive hydromorphone or morphine as needed for pain control. Nonsteroidal anti-inflammatory medication may be given for 5 days following surgery. Thirteen percent to 83% of patients undergoing VBO have positive middle ear cultures, and antibiotics are administered for 1–4 weeks based on culture and sensitivity results (Trevor and Martin 1993). The drain is removed 1–3 days following surgery. Treatment options for NPs include traction-avulsion and VBO (Figure 5.15). Traction-avulsion has historically been associated with a 40%–50% recurrence rate. In one study, there was no recurrence in patients treated with prednisolone following traction-avulsion (Anderson et al. 2000). The patient should be preoxygenated prior to anesthesia. The anesthetist should be prepared for a more difficult intubation if there is a large mass in the nasopharynx. Although it is preferable to avoid it, a temporary tracheotomy may be performed. Following induction, oxygen may be administered by a small urinary catheter placed into the airway (Figure 5.15). This can also be used as a guide for endotracheal tube placement. Ventral deviation of the soft palate may be noted in patients with a nasopharyngeal mass (Anderson et al. 2000). Pressure can be placed on the soft palate pushing the polyp into view (Anderson et al. 2000). Alternatively, the soft palate can be retracted with a spay

hook to expose the polyp. Rarely, the soft palate is incised to expose the mass. The base of the mass is grasped with hemostats (Figure 5.15). Traction and rotation are applied to ensure complete removal of the stalk. Blood should be removed from the nasopharynx with cottontipped applicators. A mirror should be used to ensure that no gross disease is left behind. Following surgery, prednisolone may be administered at 1–2 mg/kg daily for 2 weeks, 0.5–1 mg/kg for 7 days, then every other day for 7–10 days. Masses are removed similarly from the external ear canal (see Figure 5.14). If they are difficult to reach, a vertical incision can be made in the lateral wall of the vertical ear canal (ter Haar 2006). An incision is made in the overlying skin. The subcutaneous tissues and parotid gland are dissected from the cartilage of the lateral aspect of the vertical ear canal. A vertical stab incision is made. Stay sutures are placed cranial and caudal to the incision to facilitate exposure. Small closed hemostats are introduced into the horizontal ear canal. The polyp is then grasped as closely as possible to the osseus meatus and avulsed with rotation and traction. It should not be simply excised as it is important to remove the stalk. The middle ear is then gently lavaged with warm saline. The osseus meatus and lateral aspect of the tympanic cavity can be gently palpated with a curette to remove any additional tissue. The cartilage of the ear canal is then closed with 4-0 monofilament suture in a simple interrupted suture pattern. The subcutaneous tissues are closed similarly with a simple continuous pattern. Some authors recommend that a VBO be performed in all patients with NPs that have radiographic changes of the bulla or neurologic deficits. Excellent results have been reported (Bradley 1984). Complications The patient’s breathing should be monitored closely following surgery, particularly if a bandage was placed. Cats are more likely to be dyspneic following surgery, necessitating bandage removal. Horner’s syndrome is the most common neurologic complication following surgery, occurring in up to 80% of cases (Bradley 1984; Trevor and Martin 1993; Lanz and Wood 2004), but normally resolving within 4 weeks of surgery. The cause of Horner’s syndrome is trauma to the postganglionic sympathetic axons as they course through the middle ear (Bradley 1984; Trevor and Martin 1993). Facial nerve paralysis and peripheral vestibular syndrome (secondary to aggressive curettage) may also occur (Trevor and Martin 1993; Lanz and Wood 2004). Hypoglossal nerve deficits have also been reported following VBO, due to

Head and Neck Tumors  103

excessive retraction (Lanz and Wood 2004). The primary nonneurologic complication is recurrence of the inflammatory polyp (6% recurrence rate has been reported following VBO). Recurrent otitis media/interna, pharyngeal swelling, and incisional drainage have also been reported (Lanz and Wood 2004). It is interesting to note that VBO does not appear to have much effect on hearing if the ossicles are intact and the tympanum is preserved. In one study, 11 out of 12 dogs had normal hearing following VBO despite the development of subperiosteal new bone from the inner surface of the bulla with some obliteration of the bulla (ter Haar 2006). Damage to the tympanic membrane does cause conductive hearing impairment, however. This appears to resolve once the tympanic membrane has healed. Prognosis There are few published reports of outcome following surgery for neoplasia of the middle ear. A dog with papillary adenoma had a ventral bulla osteotomy and partial curettage of the middle ear. He was euthanized 2 years after surgery for chronic otorrhea (Little et al. 1989). Cats most commonly have aggressive tumors, with the majority being euthanized at the time of diagnosis (Stone et al. 1983; Indrieri and Taylor 1984; Pentlarge 1984; Lane and Hall 1992). In one study of four cats with middle ear neoplasia with tumor types including anaplastic carcinoma, SCC, lymphosarcoma, and ceruminous gland adenocarcinoma, the mean survival time was 1.25 months (4 days to 3 months), with surgery not improving the clinical course (Trevor and Martin 1993). The patients were euthanized due to progressive or recurrent clinical signs. Survival may be improved with advanced imaging prior to surgery and a more aggressive surgical approach. Follow-up with radiation therapy would be ideal. In patients with extension beyond the bulla, the VBO can be combined with a rostrotentorial craniectomy (Lucroy et al. 2004). In one 15 year-old cat with a papillary adenoma there was no evidence of recurrence 840 days following surgery (Lucroy et al. 2004). In a second 13 year-old cat, an adenocarcinoma was incompletely excised, due to firm attachments to the brainstem. Further surgery and radiation therapy were declined. The cat was euthanized 630 days following surgery due to progressive neurologic signs (Lucroy et al. 2004). In patients with NPs, a decreased recurrence rate is expected with VBO (Lanz and Wood 2004). In one study, 7 cats had no recurrence, with a mean follow-up of 17 months following VBO (range 5–36 months following surgery). In another study of 37 cats with NP,

follow-up was available for 22 cats. Recurrence occurred in 9 cats (41%) after 1–9 months (median 3.5 months) following traction-avulsion (Anderson et al. 2000). Recurrence was more likely when NPs extended into the external ear canal (50% recurred) and in patients where traction-avulsion was performed without prednisolone. Patients with NPs in the nasopharynx and patients treated with prednisolone had no evidence of recurrence. Historically, it has been recommended that any patient with changes on bulla radiographs should have a ventral bulla osteotomy performed. According to Anderson and colleagues (2000), 30% of the cats with radiographic changes of the bullae had resolution following traction-avulsion. One cat with recurrence had resolution following a second traction-avulsion followed by prednisolone.

Salivary Gland Tumors Salivary gland tumors are uncommon in the dog and rare in the cat. Most patients are elderly. In a review of the literature, 79 dogs and 72 cats had malignant salivary gland disease (Wells and Robinson 1975; Evans and Thrall 1983; Louw and Van Schouwenburg 1984; Carberry et al. 1987; Brunnert and Altman 1990, 1991; Spangler and Culbertson 1991; Burek et al. 1994; Habin and Else 1995; Thomsen and Myers 1999; Pérez-Martínez et al. 2000; Hammer et al. 2001; Sozmen et al. 2002, 2003; Wiedmeyer 2003; Mazzullo et al. 2005; Militerno et al. 2005; Smrkovski et al. 2006; Faustino and Dias Pereira 2007; Oyamada et al. 2007; Kim et al. 2008; Psalla et al. 2008). The median age for dogs was 9 years (3–14 years) (Habin and Else 1995; Evans and Thrall 1983; Louw and Van Schouwenburg 1984; Carberry et al. 1987; Brunnert and Altman 1990; Thomsen and Myers 1999; Pérez-Martínez et al. 2000; Hammer et al. 2001; Sozmen et al. 2003; Militerno et al. 2005; Smrkovski et al. 2006; Faustino and Dias Pereira 2007), and the median age for cats was 10 years (6–16 years) (Wells and Robinson 1975; Carpenter and Bernstein 1991; Burek et al. 1994; Sozmen et al. 2002, 2003; Mazzullo et al. 2005; Oyamada et al. 2007; Kim et al. 2008; Psalla et al. 2008). In another retrospective study of 24 dogs and 30 cats, the median age for dogs was 10 years (range of 3–14 years), and the median age for cats was 12 years (range 7–22 years) (Hammer et al. 2001). In dogs, there is no breed or sex predilection. In cats, however, Siamese or Siamese-cross cats may be at increased risk, with male cats being affected twice as often as female cats (Hammer et al. 2001). The tumors are typically malignant, locally invasive, and of epithelial origin. Simple adenocarcinoma is the most common tumor type (Carberry et al. 1988;

104  Veterinary Surgical Oncology

Hammer et al. 2001). Of the 79 dogs reported in the literature with malignant salivary tumors, 66 (84%) were adenocarcinoma (including simple, complex, cyst, and basal cell). The remaining 13 (16%) included malignant mixed tumor (n = 3), acinic cell carcinoma (n = 2) and one of each of the following: mucoepidermoid carcinoma, SCC, carcinoma in a pleomorphic adenoma, solid anaplastic carcinoma, malignant fibrous histiocytoma, malignant myoepithelioma, extraskeletal osteosarcoma, and mast cell tumor (Evans and Thrall 1983; Louw and Van Schouwenburg 1984; Carberry et al. 1987; Brunnert and Altman 1990; Spangler and Culbertson 1991; Habin and Else 1995; Thomsen and Myers 1999; Pérez-Martínez et al. 2000; Hammer et al. 2001; Sozmen et al. 2003; Militerno et al. 2005; Smrkovski et al. 2006; Faustino and Dias Pereira 2007). Of the 72 cats reported in the literature with malignant salivary tumors, 62 (86%) were adenocarcinoma (including simple, complex, basal cell, acinar and ductular). Other tumor types include malignant mixed (n = 4), acinic cell carcinoma (n = 2), and one each of the following: SCC, solid carcinoma, sebaceous carcinoma, mucoepidermoid carcinoma (Wells and Robinson 1975; Carpenter and Bernstein 1991; Spangler and Culbertson 1991; Burek et al. 1994; Hammer et al. 2001; Sozmen et al. 2002, 2003; Mazzullo et al. 2005; Oyamada et al. 2007; Kim et al. 2008; Psalla et al. 2008). Secondary invasion of the salivary glands has been reported with fibrosarcoma and lymphosarcoma (Spangler and Culbertson 1991). Benign tumors are rare. Adenomas and lipomatous infiltration of the mandibular and parotid salivary glands have been reported in the dog (Carberry et al. 1988; Bindseil and Madsen 1997; Brown et al. 1997). These masses are not fixed to adjacent tissues and are cured with surgical resection. There is one report of a teratoma associated with the mandibular salivary gland in a boxer (Lambrechts and Pearson 2001). There was no evidence of recurrence 7 months following surgery. Adenoma, papillary adenomas, and cystadenoma have been reported in the cat (Spangler and Culbertson 1991; Carberry et al. 1988) (Figure 5.16). Necrotizing sialometaplasia, a rare disease of the salivary glands, has been described in the dog and the cat (Brooks et al. 1995; Brown et al. 2004). It can be mistaken as a neoplastic condition when evaluated by fineneedle aspirate alone. Histopathology is needed to make a definitive diagnosis. In most cases, animals present with unilateral enlargement of the mandibular salivary gland. One case of bilateral involvement has been reported. Terrier breeds are predisposed (six out of seven reported cases) (Brooks et al. 1995). Dogs have more dramatic clinical signs with a history of ptyalism, nausea, pain on opening the mouth, frequent vomiting,

(a)

(b)

(c)

Figure 5.16.  (A, B) Salivary adenoma involving the lip of a cat that was surgically excised with wide margins. (C) Postoperative appearance.

and anorexia. To our knowledge, these clinical signs have not been reported in the cat. Treatment involves excision of the affected glands and, in dogs, a short course of anticonvulsant medication following surgery. In one study, four dogs treated with surgical excision alone had continued pain and vomiting and were euthanized. The remaining three had phenobarbital initiated following surgery, with resolution or improvement in clinical signs

Head and Neck Tumors  105

(Brooks et al. 1995). In two reported feline cases treated with surgical excision alone, the patients were diseasefree 6 months and 2 years following surgery (Brown et al. 2004). Phenobarbital-responsive salivary gland enlargement and hypersialosis have also been reported in dogs with no cytologic or histopathologic evidence of salivary gland pathology (Stonehewer et al. 2000). The majority present with bilateral enlargement of the mandibular salivary glands; however, unilateral involvement and parotid salivary gland involvement have also been reported. The glands may be painful on palpation. It has been postulated that the salivary gland enlargement and hypersialosis may be indicative of limbic epilepsy. Three patients in one report responded well to phenobarbital, and one was also maintained with potassium bromide. None required surgery. Sialoceles also present with a swelling in the ventral neck, intermandibular space, pharynx, or under the tongue and are not fixed to underlying structures. They have a characteristic clinical appearance with the presence of saliva on cytology and no evidence of neoplasia. With salivary gland neoplasms, the most common presenting complaint is the recent appearance of a mass. In one study, dogs had a median duration of clinical signs of 8 weeks and cats had a median duration of 4 weeks (Hammer et al. 2001). Of the 79 dogs and 72 cats reviewed in the literature, the medical records of 5 dogs and 4 cats had information regarding the duration of clinical signs. The median duration for cats was 5 weeks (3 weeks–6 months), and the median duration for dogs was 2 weeks (4 days–6 weeks) (Carpenter and Bernstein 1991; Habin and Else 1995; Wells and Robinson 1975; Carberry et al. 1987; Thomsen and Myers 1999; PérezMartínez et al. 2000; Lambrechts and Pearson 2001; Kim et al. 2008; Psalla et al. 2008). Tumors of the parotid salivary gland are located at the base of the ear, whereas tumors of the mandibular salivary gland involve the upper neck and those of the sublingual salivary gland may extend to the floor of the mouth. Tumors of the zygomatic salivary gland involve the lip and maxilla and may cause ocular signs such as exophthalmos, epiphora, and divergent strabismus (Carberry et al. 1988; Militerno et al. 2005). Other clinical signs include halitosis, dysphagia, weight loss, anorexia, neurologic deficits (facial nerve paresis, Horner’s syndrome), and sneezing (Hammer et al. 2001; Mazzullo et al. 2005). The mass is typically unilateral, firm, nonpainful, and fixed to adjacent structures (Militerno et al. 2005). There have been reports of bilateral involvement (Mazzullo et al. 2005). The mandibular gland, followed by the parotid gland, are the most commonly affected salivary glands (Wells and Robinson 1975; Evans and Thrall 1983; Carberry

et al. 1987; Carpenter and Bernstein 1991; Spangler and Culbertson 1991; Habin and Else 1995; Thomsen and Myers 1999; Pérez-Martínez et al. 2000; Hammer et al. 2001; Sozmen et al. 2002, 2003; Mazzullo et al. 2005; Militerno et al. 2005; Smrkovski et al. 2006; Oyamada et al. 2007; Kim et al. 2008). There are reports of tumors affecting the sublingual, zygomatic, and minor salivary glands; however, they are much less common (Louw and Van Schouwenburg 1984; Brunnert and Altman 1990; Spangler and Culbertson 1991; Burek et al. 1994; Hammer et al. 2001; Sozmen et al. 2003; Faustino and Dias Pereira 2007; Psalla et al. 2008). Most tumors are characterized by a rapid infiltrative growth at the time of diagnosis. Metastatic disease typically involves the regional lymph nodes and lung. In one study, 25% of dogs and 55% of cats had metastasis at the time of diagnosis (Hammer et al. 2001). Seventeen percent of the dogs and 39% of the cats had metastasis to regional lymph nodes, and 8% of dogs and 16% of cats had distant metastatic disease (Hammer et al. 2001). Metastasis to the kidney, bone, eyes, and the brain have been reported (Habin and Else 1995). Cats have more aggressive disease, with a higher incidence of local and distant metastasis at the time of presentation (Hammer et al. 2001). Dogs have a better outcome when diagnosed early, but this has not been demonstrated in the cat. Diagnosis In most cases, cytology can differentiate neoplastic from nonneoplastic diseases. The cytologic findings consistent with salivary gland adenocarcinoma are well described in the literature (Militerno et al. 2005). Cytology may be difficult to interpret in cases of necrotizing sialometaplasia (Brooks et al. 1995). Incisional biopsy will yield a more definitive diagnosis. At a minimum, the mass and associated lymph nodes should be evaluated by fine-needle aspiration, and thoracic radiographs should be performed. The medial retropharyngeal lymph node has been shown to be a draining lymph node for mandibular salivary adenocarcinoma. Metastases to the submandibular, prescapular, axillary, and bronchial lymph nodes have also been reported (Habin and Else 1995; Mazzullo et al. 2005). Radiographs of the cervical region may also be performed to evaluate displacement of structures by a soft tissue mass (the primary tumor and/or lymph nodes) and to evaluate adjacent bone for osseus changes. Ultrasound, CT or MRI can be used to evaluate tumor and lymph node size and to look for evidence of invasion into adjacent soft tissue or bone. MRI of head and neck lymph nodes, in clinically normal dogs, has been described (Kneissl and Probst 2006). CT images can also be used to evaluate the thorax for metastases. Ocular

106  Veterinary Surgical Oncology

examination is also recommended as metastases to the eye have been reported. In most cases, there will be other overt signs of metastatic or primary disease; however, patients with choroidal metastases may only have reduced vision. Routine ocular examination may allow earlier detection of metastatic disease (Habin and Else 1995). Diagnostic tests include routine hematology and biochemical profile (±T4). Hypoglycemia has been reported in cases of salivary adenocarcinoma (Morrison 2002). Treatment Surgical resection of the salivary tumor is the treatment of choice. An incision is made over the salivary gland and extracapsular resection is performed. Incisional biopsy tracts and any fixed skin should be excised with the tumor. The gland is exposed and removed with a combination of sharp and blunt dissection. Cautery and hemoclips are useful to provide hemostasis. Extirpation of the ipsilateral neck can be performed with a good outcome (Withrow 2007). The sequelae may be an inability to blink the ipsilateral eyelids due to facial nerve paresis. Some authors recommend tarsorrhaphy and the use of eye drops; however, if tear production is normal, they should not be necessary. The patient is placed in dorsolateral recumbency with a rolled towel under the neck. The mandibular salivary gland lies in the bifurcation of the external jugular vein. The skin, subcutaneous tissue, and platysma muscle are incised in a longitudinal direction over the salivary tumor to expose the mass. Dissection is performed around the mass, with excision of adjacent invaded tissue. The proximity to the external jugular vein, carotid and lingual arteries, and vagosympathetic trunk should be noted. The sublingual salivary gland is closely associated with the mandibular salivary duct, and the caudal portion will be removed concurrently. It should be ligated on the rostral aspect of the mass with suture or a hemoclip. Following excision, the surgical site is lavaged with sterile saline. Dead space is closed with sutures, and the wound is closed routinely. Passive drains should not be placed because microscopic disease often remains and the drains could track neoplastic cells into adjacent normal tissue. Hemoclips may be placed within the surgical field to mark the site for radiation therapy. The parotid salivary gland is removed similarly. The patient is placed in lateral recumbency (Dunning 2003). A vertical incision is made over the mass, and the underlying platysma and parotidoauricularis muscles are incised. The tumor is excised from the adjacent tissue. The facial nerve courses ventral to the base of the vertical ear canal and should be preserved if possible. For closure,

the parotidoauricularis muscle is apposed and the superficial layers closed routinely. If there is significant dead space, a light, padded bandage may be placed for 2–5 days following surgery. Advanced imaging is recommended for tumors of the zygomatic salivary gland to determine if the globe can be preserved (Gilger et al. 1994). Invasive tumors of the zygomatic salivary gland are removed via en bloc resection of the affected tissue. Noninvasive tumors can be removed via a lateral orbitotomy with preservation of the globe (Gilger et al. 1994; Hedlund 2002; Bartoe et al. 2007). Generous amounts of lubricant should be placed on the ipsilateral eye and a temporary tarsorrhaphy is performed. The skin is incised over the dorsal rim of the zygomatic arch starting ventral to the lateral canthus and extending caudally to the base of the ear. The palpebral nerve courses in the subcutaneous fat over the temporalis muscle toward the lateral canthus of the eye. It should be isolated and retracted dorsally with moistened umbilical tape. The aponeurosis of the temporalis muscle is incised along the dorsal border of the zygomatic arch. If the arch is to be replaced, holes are predrilled on either side of the proposed osteotomy sites with a 0.35 mm drill bit or K-wire. The rostral osteotomy site is just caudal to the orbital ligament. The second site is extended caudally such that it gives maximum exposure to the tumor. The osteotomy is then performed between the holes with an oscillating or sagittal saw. The orbital ligament is transected midbody, and the arch is reflected ventrally. The lateral canthus and globe are retracted rostrally to expose the orbit. The mass is removed with sharp and blunt dissection. The site should be lavaged with sterile saline. The arch can be replaced and fixed with 22-gauge orthopedic wire. If radiation therapy is planned following surgery, the arch should not be replaced to reduce the risk of the development of a bony sequestrum. The temporalis fascia is sutured to the zygomatic arch. The orbital ligament is sutured with 3-0 absorbable suture, such as polydioxanone (PDS; Ethicon Inc., Cornelia, GA, USA) in a horizontal mattress pattern and the superficial tissues are closed routinely. An Elizabethan collar is indicated following surgery to prevent self-trauma. Complications include seroma formation, infection, facial nerve paresis, and tumor recurrence. As most of these tumors are invasive, follow-up with radiation therapy is recommended (Carberry et al. 1988). Patients have received 37–57 Gy. In one report, six dogs and four cats were treated with a cobalt-60 source and one dog and one cat were treated with a linear accelerator (Hammer et al. 2001). In another report, three dogs with parotid gland adenocarcinoma received

Head and Neck Tumors  107

10 doses from an orthovoltage unit totaling 45 Gy (Evans and Thrall 1983). Side effects include transient moist dermatitis, which can be severe, and permanent hair loss and color change. If the oral cavity is within the field, the patients may experience transient mucositis. If the eye is in the field, keratoconjunctivitis sicca, and cataracts may develop months following radiation therapy. Prognosis There are a few reports of prolonged survival following surgical resection with and without radiation therapy. In one study of 24 dogs and 30 cats treated with surgical resection with and without radiation therapy, the median survival was 550 and 516 days, respectively (Hammer et al. 2001). Although follow-up with radiation therapy is preferred, this study documented prolonged survival in patients that had repeat surgical resection of recurrent tumors (Hammer et al. 2001). This is dissimilar from other studies that describe aggressive tumors with the rapid development of metastatic disease. Histologic grade was not prognostic; however, advanced stage was a negative prognostic indicator. Of 79 dogs reported in the literature with malignant salivary tumors, 12 had information regarding treatment and outcome available (Evans and Thrall 1983; Louw and Van Schouwenburg 1984; Carberry et al. 1987; Brunnert and Altman 1990; Habin and Else 1995; Sozmen et al. 2003; Militerno et al. 2005; Smrkovski et al. 2006; Faustino and Dias Pereira 2007). Five were treated with surgery alone (lingual acinic cell carcinoma, extraskeletal osteosarcoma, mandibular adenocarcinoma, carcinoma in an adenoma, and malignant myoepithelioma), three (parotid gland adenocarcinoma) had surgery and radiation therapy, one (mast cell tumor) had surgery and chemotherapy, and three (metastatic parotid adenocarcinoma, invasive solid anaplastic carcinoma and mandibular basal cell adenocarcinoma) were not treated. Two untreated patients were euthanized at presentation due to extensive disease (Louw and Van Schouwenburg 1984; Habin and Else 1995), and one was euthanized 8 months after diagnosis with a massive tumor (Sozmen et al. 2003). Of the five patients with surgical excision of the mass, one died shortly after (malignant myoepithelioma) (Faustino and Dias Pereira 2007) and one (osteosarcoma) had recurrence and metastatic disease 1 month after surgery (Thomsen and Myers 1999). Three (lingual acinic cell carcinoma, mandibular adenocarcinoma, carcinoma in an adenoma) were disease-free 7, 8, and 12 months after surgery (Brunnert and Altman 1990; Militerno et al. 2005; Smrkovski et al. 2006). Of the 3 patients treated with

surgery and radiation therapy (parotid adenocarcinoma), 1 died 40 months after treatment due to a possible abdominal tumor and the other two were disease-free at 12 and 25 months after surgery (Evans and Thrall 1983). Of the 7 cats reported in the literature, 1 cat had surgery (adenocarcinoma of a minor salivary gland) (Burek et al. 1994) with tumor recurrence at 3 months, 1 had surgical excision and radiation therapy (malignant mixed tumor of the mandibular salivary gland) (Kim et al. 2008) with pulmonary metastases at 7 weeks, 4 (mixed tumor, basal cell adenocarcinoma, adenocarcinoma, and nasal acinic cell carcinoma) (Carpenter and Bernstein 1991; Sozmen et al. 2003; Mazzullo et al. 2005; Psalla et al. 2008) were euthanized at presentation, and 1 (poorly differentiated tumor of the mandibular salivary gland) (Wells and Robinson 1975) was euthanized 5 months following presentation. Unfortunately, there is little data evaluating the outcome with chemotherapy. An array of protocols have been implemented; however, there are too few numbers to assess outcome. In one study, dogs treated with surgery and chemotherapy had a poorer prognosis than dogs treated with surgery alone and dogs treated with surgery and radiation therapy (Hammer et al. 2001). This may be due to selection bias as patients with more aggressive disease are more likely to receive chemotherapy. In people, a correlation with environmental carcinogens and the development of salivary gland tumors has been identified. It has been proposed that salivary neoplasia in animals may also have a such a relationship, with one author proposing that it may be secondary to exposure to environmental carcinogens during grooming (Hammer et al. 2001).

Tumors of the Lip Any soft tissue tumor can affect the canine lip, with malignant melanoma being the most common (Todoroff and Brodey 1979; Vos and van der Gaag 1987). In one study of 35 lip tumors, 71% were malignant melanoma and the remaining were SCC (20%) and fibrosarcoma (9%) (Todoroff and Brodey 1979). Mast cell tumor and extramedullary plasmacytoma (with and without amyloid deposits) have also been reported (Lucke 1987; Brunnert and Altman 1991; Rowland et al. 1991). There is one report of four aged dogs with round cell sarcomas of possible myelomonocytic origin. A granular cell tumor of the lip has been reported (Turk et al. 1983). Most patients are older than 10 years of age. Male patients and small breed dogs, particularly cocker spaniels are predisposed to malignant melanoma (Vos and

108  Veterinary Surgical Oncology

van der Gaag 1987; Ramos-Vara et al. 2000; Schultheiss 2006). Malignant melanoma is characterized by local infiltration and metastasis to regional lymph nodes and less frequently to lungs and other organs (Bostock 1979; Ramos-Vara et al. 2000). The majority of melanomas on the canine lip affect the mucous membrane and have a malignant histologic appearance (Schultheiss 2006). Immunohistochemistry may be indicated to confirm a diagnosis of malignant melanoma as they have a variable degree of pigmentation and may be completely unpigmented (Ramos-Vara et al. 2000). SCC is the most common tumor affecting the lip of cats (Vos and van der Gaag 1987). Other tumor types include fibrosarcoma, lymphosarcoma, and mastocytoma (Bradley 1984; Vos and van der Gaag 1987). Malignant melanoma is rare (Bradley 1984). Most patients are middle-aged or older (Bradley 1984; Vos and van der Gaag 1987).

(a)

Treatment Treatment options include surgical resection, cryosurgery, radiation therapy, and chemotherapy. Curative interstitial brachytherapy has been reported in people (Jha et al. 2006). Electrochemotherapy has also been reported (Aminkov and Manov 2004). Resection of the lip can be combined with a partial maxillectomy/ mandibulectomy when there is bony involvement (Salisbury et al. 1986). The upper labial mucosa and cheek are supplied by the lateral nasal artery (a branch of the infraorbital artery) and the angular artery of the mouth and superior labial artery (branches of the facial artery) (Salisbury et al. 1986). The lower lip is supplied by the caudal, middle, and rostral mental arteries (Pavletic 1999). Local reconstruction techniques include simple wedge resection, advancement and buccal rotation flaps, and axial pattern flaps (Yates et al. 2007). Goals of surgery include separating the nasal from the oral cavity, reconstructing the buccal mucosal surface to prevent excessive scarring, and providing a tension-free closure to the haired skin. Small mucosal defects can heal by second intention; however, large mucosal defects need to be reconstructed. Mucosa is preferred for reconstruction, but when not available, haired skin can be used. Small and benign tumors can be removed with simple techniques. Wedge, rectangular, or pentagonal fullthickness skin incisions are made with a scalpel blade to remove the neoplasm, ensuring that adequate margins are obtained (Pavletic 1999) (Figure 5.17). The resection margins can be marked preoperatively with a sterile marker to ensure that the margins are not altered with resection. Moist gauze squares can be placed

(b)

(c)

Figure 5.17.  (A) Mast cell tumor affecting the cheek of a 7-yearold female spayed Beagle mix. (B) Full-thickness resection of the lip. (C) The incision was closed in a Y with no residual disturbance to eyelid or lip function. Appearance 2 weeks following surgery.

Head and Neck Tumors  109

deep to the lip to apply tension against which to cut. Alternatively, skin hooks, towel clamps, or tension sutures may be placed at the margins to facilitate making the incision (Swaim and Henderson 1997). For closure, the first suture aligns the labial margins. Starting away from the labial margin, the oral mucosa is apposed with 3-0 simple interrupted sutures and includes the submucosa. The skin is then apposed with simple interrupted 3-0 nonabsorbable sutures. Wedge incisions are closed in a linear fashion, and rectangular incisions are closed in a Y. The mouth should be manipulated to ensure a tension-free closure. When combined with a maxillectomy, the upper mucosal margin is sutured to the palate, closing the oronasal defect. Pavletic has also described the lower labial lift-up and upper labial pull-down techniques for lesions that allow preservation of the lip margin. This is reserved for small or benign tumors to ensure that adequate tissue margins are obtained. This technique is also useful in patients with invasive maxillary or gingival tumors that have focal or more dorsal lip involvement. The primary tumor is removed with full-thickness excision of the labial lesion and preservation of the lip margin. It is important to ensure that adequate tissue margins are obtained. The labial mucosa is then sutured to the gingival margin or hard palate. The skin is closed with simple interrupted sutures. In cases where this causes dorsal deviation of the lip, a V-incision is made at the lip margin. The caudal lip margin is then advanced cranially to close the defect (Figure 5.18). A full-thickness advancement flap can be used for larger defects of the rostral one-third of the upper lip and can be combined with a partial maxillectomy for more extensive lesions (Figure 5.19). This technique does not work as well in cats as they have less free lip margin and a smaller commissure. The mass is resected with a rectangular or pentagonal incision (Pavletic 1999; Swaim and Henderson 1997). When possible, a 0.5 cm strip of mucosa should be left along the gingival border to facilitate closure. An incision is made at the dorsal extent of the resection and extended caudally, typically beyond the level of the commissure to allow advancement without tension (Swaim and Henderson 1997). The flap is elevated and advanced rostrally to close the defect. Areas in the flap that are under tension, including the infraorbital nerve, artery, and vein, can be incised (Swaim and Henderson 1997). A small wedge is resected from the dorsorostral aspect of the flap to ensure that the tissue has an adequate blood supply, preventing dehiscence. The mucosal surface is apposed with 3-0 absorbable suture in a simple interrupted pattern. Alternatively, the submucosa is apposed allowing the mucosa to evert into the oral cavity (Swaim and Henderson

(a)

(b)

(c)

Figure 5.18.  (A) Local recurrence of a SCC on the dorsolateral lip of an 11-month-old male neutered Yorkshire terrier. A maxillectomy had been performed 7 weeks previously. (B) Full-thickness resection of the mass in (A), including maxilla. (C) The defect is closed with an upper labial pull-down technique. The labial mucosa is sutured to the palatine mucosa to close the oral cavity. The ventral skin margin is then apposed to the dorsal skin margin. To prevent upward deviation of the lip edge, the commissure is advanced rostrally. A strip of lip margin is resected, and the middle of the resected lip becomes the dorsal extent of a vertical suture line. The rostral and caudal portions of the resected lip are then apposed to each other in a vertical incision. For this patient, surgery was followed with radiation therapy and carboplatin. There was no evidence of recurrence 3 years following surgery. (Images courtesy of Dr. Karen Tobias)

110  Veterinary Surgical Oncology

(a)

(b)

(c)

(d)

(e)

Figure 5.19.  (A) A mast cell tumor involving the rostral lip of an adult mixed breed dog. (B) Following full-thickness resection of the lip. (C) The defect was closed with a full-thickness advancement flap. Note the gingival mucosa left to facilitate closure. An incision was made at the dorsal extent of the resection and extended caudally. The flap was then advanced rostrally to close the defect. Sutures are placed first at the labial margin and rostrodorsal corner to ensure that the tissues are aligned. (D) The mucosal surface is apposed with simple interrupted absorbable sutures. The skin is then apposed with nonabsorbable sutures. Jaw mobility should be assessed intraoperatively. (E) Postoperative appearance with mild deviation of the planum nasale. This will resolve in 2–3 weeks.

1997). When combined with a partial maxillectomy, holes are drilled in the maxilla to anchor the suture and reduce the risk of dehiscence. To ensure alignment, the first sutures are placed at the rostrodorsal corner and at the labial margin. Sutures are then continued caudally and ventrally from the rostrodorsal border. The skin is closed with 3-0 nonabsorbable sutures. Jaw mobility

should be assessed intraoperatively. Use of this flap may deviate the planum to the side of the surgery; however, this typically resolves within 2–3 weeks. A similar flap can be elevated for rostral lesions of the lower lip (Pavletic 1999). The lower lip is easier to mobilize and can be advanced with a shorter skin incision. The mass is resected, and when possible, a 0.5 cm strip

Head and Neck Tumors  111

of mucosa is left along the gingival border to facilitate closure. A skin incision is made at the ventral aspect of the wound and extended caudally parallel to the mandible. The skin incision is often shorter than the mucosal incision due to its inherent elasticity. The flap is drawn cranially and sutured into the wound bed. The mucosa is apposed with 3-0 absorbable suture material in a simple interrupted pattern. At the space between the canine tooth and first premolar, the lip should be attached dorsally to prevent sagging of the lip. The skin is apposed with 3-0 nonabsorbable simple interrupted sutures. For larger defects, a buccal rotation flap may be created (Figure 5.20). The cheek margin is advanced rostrally to fill the defect. The caudal labial margin is brought to the dorsorostral corner of the defect. An appropriate strip of mucosa is removed from the labial margin to allow apposition of the lip to the rostral vertical incision. The mucosal surface is apposed with 3-0 absorbable suture pattern, and the skin is apposed with 3-0 nonabsorbable suture in a simple interrupted suture pattern. Vertical mattress sutures may be placed in any areas under tension. This procedure advances the commissure of the lips rostrally. The commissure can be extended caudally by incising the commissure and suturing the skin to mucosa dorsally and ventrally, however, this is rarely indicated. There are several axial pattern flaps that can be used in lip reconstruction. The caudal auricular axial pattern flap has been used to reconstruct a chin defect (Aber et al. 2002). The lateral aspect of the wing of the atlas is the base of the flap. The dorsal midline is the dorsal border of the flap, and the ventral border runs parallel to it, originating at the depression between the ear and the wing of the atlas. The incisions are connected with a transverse incision at the level of the scapula. The flap is rotated rostrally below the ear to close the mandibular defect. It can be connected via a tubed flap or a bridging incision. If a tube is used, it is transected 3–4 weeks after surgery. If there is concern with the integrity of the blood supply, it can be performed in two stages, with an initial incision made halfway between the tube and the donor site and the incision resutured. The tube is then resected 2–3 days later. A tube flap yields a better cosmetic appearance than a bridging incision; however, there is risk of tube trauma by the patient, and a second procedure is needed for removal. The superficial temporal artery axial pattern flap can be used to reconstruct the upper lip and can be combined with a maxillectomy if necessary (Lester and Pratschke 2003; Fahie and Smith 1997). The flap is based on the dorsal aspect of the zygomatic arch. The rostral incision extends dorsally from the lateral orbital rim,

(a)

(b)

Figure 5.20.  (A) A buccal rotation flap was used to repair the lip of a 7-year-old male neutered schnauzer. The lip was traumatized in a dog fight. The rotation flap is outlined with surgical marker. (B) The flap has been rotated into the defect. (Images courtesy of Dr. Geraldine Hunt)

and the caudal skin incision extends dorsally from the caudal aspect of the zygomatic arch. Parallel skin incisions are continued to the dorsal orbital rim of the contralateral eye. Dissection is performed deep to the frontalis muscle, carefully elevating the flap toward its base. Atraumatic tissue handling is important, and the flap must be kept moist. The flap can then be rotated into the defect. A closed active suction drain can be placed deep to the flap. In one case report, porcine

112  Veterinary Surgical Oncology

intestinal submucosa was used to close the mucosal defect (Lester and Pratschke 2003). It was removed 6 days after surgery, but was felt to be a useful scaffold for tissue regeneration. The angularis oris axial pattern cutaneous flap is oriented caudally from the commissure of the mouth (Yates et al. 2007). Parallel incisions are extended caudally along the ventral aspect of the zygomatic arch and the ventral aspect of the ramus of the mandible to the horizontal ear canal. The flap is elevated from caudal to rostral, deep to the platysma muscle. Careful dissection is performed at the flap base to preserve cutaneous vasculature. The donor site is closed primarily without tension. The flap can be used to reconstruct both the buccal surface and haired lip when the width of the flap is at least one half the width of the defect. The flap is flipped 180 degrees into the defect with hair facing the oral cavity. The upper border of the flap is sutured to the mucosal margin with 3-0 absorbable suture in a simple interrupted suture pattern. The flap is folded onto itself and sutured to the skin with 3-0 non-absorbable suture in a simple interrupted suture pattern. In patients with larger defects, the flap is used to reconstruct the buccal mucosa, and adjacent skin is mobilized over the flap to close the cutaneous defect (Figure 5.21). Lip margin that can be apposed without tension is sutured. The base of the flap is inverted, and the flap is placed into the defect with the haired portion facing the oral cavity. The flap is sutured dorsally to the gingival or hard palate and ventrally to the mandibular gingiva, with 3-0 absorbable sutures in a simple interrupted pattern. The skin is advanced over the flap and inverse tube segment. This leaves a narrow opening to an epithelium-lined skin tube. In 4–6 weeks after surgery, once the surgery site has healed and neovascularization has occurred, the short tube segment can be excised. For patients where the upper lip margin is preserved, the buccal mucosa of the lip is sutured to the gingiva or hard palate and the flap is rotated into the defect dorsal to the lip margin via a bridging incision. It can reach the philtrum without tension. With this technique, the redundant skin folds can accumulate food and saliva. Once the flap has healed, redundant tissue can be resected. As an alternative, a tubed flap can be used to reach the lesion in lieu of a bridging incision. The tube can be resected in 3–4 weeks after the flap has healed. Postoperative care An Elizabethan collar should be placed for 1–2 weeks after surgery to prevent trauma to the surgery site. In

patients with a caudal auricular axial pattern flap, a light padded bandage may be placed over the donor site to prevent irritation from the collar. Moist, cool compresses may be placed over the surgery site for 72 hours after surgery to keep the surgery site clean and reduce swelling (Swaim and Henderson 1997). Soft food should be fed for 4 weeks after surgery, and playing with hard toys must be prevented. Antibiotics may be given in the perioperative period. Complications Billowing of the flap is seen with expiration when the flap is placed over the exposed nasal cavity. This typically resolves within 10 days of surgery. Periodic hair trimming may be needed in patients where the flap extends rostrally to the planum nasale to prevent irritation and sneezing (Yates et al. 2007). Likewise, halitosis may occur when haired skin is used to reconstruct the buccal mucosa. Periodic sedation may be needed to trim hair in the mouth. Clients should be warned that buccal advancement flaps and commissure rotation flaps will move the commissure rostrally, changing the shape of the face, and may cause deviation of the planum (Swaim and Henderson 1997). Rotation or axial pattern flaps also move hair with different lengths or thickness, which will change facial appearance. Ear carriage may also be altered. Prognosis Outcome for tumors of the lip are often combined with those reported for tumors of the oral cavity (Todoroff and Brodey 1979; Vos and van der Gaag 1987; RamosVara et al. 2000). In one study, no difference was found in outcome between tumors of the lip and oral cavity, when evaluated separately, and the results were combined (Bostock 1979). Melanomas of the lip or oral cavity should be considered behaviorally malignant, despite histologic appearance (Bostock 1979). Death is due to local recurrence and/or metastatic disease. In one study, the median survival time was 15 weeks, with 10% of the dogs alive 2 years after diagnosis (Bostock 1979). In another study, one-third of the patients with histologically malignant tumors were alive and tumor free more than 1 year after surgery (Schultheiss 2006). Overall, soft tissue sarcomas have a low incidence of metastasis (Vos and van der Gaag 1987). In a report of one dog with a mast cell tumor, there was no evidence of recurrence 9 months after surgery (Yates et al. 2007). Plasmacytomas affecting the lip can respond well to surgical excision (Lucke 1987; Brunnert and Altman 1991). In one case report of four dogs, there was no

Head and Neck Tumors  113

(a)

(b)

(d)

(g)

(e)

(h)

(c)

(f)

(i)

Figure 5.21.  (A) Buccal SCC in an 11-year-old male neutered west highland white terrier. (B) Full-thickness resection of the right cheek, part of the mandible, and part of the maxilla of the patient in (A). (C) The rostral lip margins are apposed. In this patient, a local transposition flap is elevated. Alternatively, the angularis oris axial pattern flap can be mobilized into the defect. The skin flap will be used to reconstruct the buccal surface of the lip. The base of the flap is folded to make an inverse tube (as indicated by the forceps). (D) The skin flap is rotated into the defect with the skin surface facing the mouth. (E) The skin is sutured to the mucosa of the palate. (F) The skin flap is then sutured to the mandible, reconstructing the buccal aspect of the cheek. (G) Intraoral view of the completed reconstruction. (H) Adjacent skin is then advanced over the tubed flap. The triangular-shaped dimple is the opening of the epitheliumlined skin tube. This will become less obvious with hair regrowth. Mild discharge can be treated with clipping and gentle flushing of the tube and possible antimicrobial therapy. This is usually sufficient. However, if the discharge is more frequent, the tube can be resected 4–6 weeks after surgery, once revascularization of the flap has occurred. (I) Intraoral view of the flap 15 days after surgery with hair regrowth. Halitosis was managed with dental mouthwashes, intermittent antimicrobial therapy, and periodic trimming of the hair. The patient was euthanized 15 months after surgery due to renal failure. There was no evidence of local recurrence or overt metastatic disease at that time. (Images courtesy of Dr. Doug Huber)

evidence of recurrence 3 and 26 months following surgery, and follow-up was not available for two dogs (Lucke 1987). Of the dogs with round cell sarcoma of possible myelomonocytic origin, patients had local recurrence

and metastatic disease (regional lymph node and lung) 10 weeks to 1 year after resection (Kipar et al. 1995). Granular cell tumors of the lip are behaviorally benign. In one case, there was no evidence of recurrence 28 months after surgical resection (Turk et al. 1983).

114  Veterinary Surgical Oncology

To our knowledge, there are no case studies that evaluate tumors of the feline lip, exclusively. All are combined with other oral tumors. In general, the majority of oral tumors are described as being malignant with an aggressive course. There is one report of a cat with a well-differentiated sebaceous adenocarcinoma of the chin reconstructed with a caudal auricular axial pattern flap (Aber et al. 2002). Lymph node evaluation was not performed prior to surgery, and the tumor was removed with 0.25 cm margins. The cat was euthanized 5 months after surgery with probable metastatic disease to the regional lymph nodes. There is a report of a cat with a large periodontal fibromatous epulis involving the upper lip reconstructed with a superficial temporal axial pattern flap (Lester and Pratschke 2003). Seven months after surgery there was no evidence of recurrence.

References Aber, S.L., T. Amalsadvala, J. Brown, and S.F. Swaim. 2002. Using a caudal auricular axial pattern flap to close a mandibular skin defect in a cat. Vet Med 97:666–671. Aminkov, B. and V. Manov. 2004. Electrochemotherapy—A novel method of treatment of malignant tumours in the dog. Bulgarian J Vet Med 7:209–213. Anderson, D.M., R.K. Robinson, and R.A.S. White. 2000. Management of inflammatory polyps in 37 cats. Vet Rec 147(24):684– 687. Atwater, S.W., B.E. Powers, and R.C. Straw. 1991. Squamous cell carcinoma of the pinna and nasal planum: 54 cats (1980–1991). Proceed Vet Cancer Soc 11:35–36. Bacon, N.J., R.L. Gilbert, D.E. Bostock, et al. 2003. Total ear canal ablation in the cat: Indications, morbidity, and long-term survival. J Small Anim Prac 44(10):430–434. Bartoe, J.T., A.H. Brightman, and H.J. Davidson. 2007. Modified lateral orbitotomy for vision-sparing excision of a zygomatic mucocele in a dog. Vet Ophthalmol 10(2):127–131. Bindseil, E. and J.S. Madsen. 1997. Lipomatosis causing tumour-like swelling of a mandibular salivary gland in a dog. Vet Rec 140(22):583–584. Bostock, D.E. 1979. Prognosis after surgical excision of canine melanomas. Vet Pathol 16(1):32–40. Bostock, D.E. 1986. Neoplasms of the skin and subcutaneous tissue in dogs and cats. Brit Vet J 142:1–19. Bradley, R.L. 1984. Selected oral, pharyngeal, and upper respiratory conditions in the cat. Oral tumors, nasopharyngeal and middle ear polyps, and chronic rhinitis and sinusitis. Vet Clin North Am Small Anim Prac 14(6):1173–184. Brooks, D.G., H.A. Hottinger, and R.W. Dunstan. 1995. Canine necrotizing sialometaplasia: A case report and review of the literature. J Am Anim Hosp Assoc 31(1):21–25. Brown, P.J., J.M. Bradshaw, M. Sozmen, et al. 2004. Feline necrotising sialometaplasia: A report of two cases. J Feline Med Sur 6(4):279–281. Brown, P.J., V.M. Lucke, M. Sozmen, et al. 1997. Lipomatous infiltration of the canine salivary gland. J Small Anim Prac 38(6): 234–236.

Brunnert, S.R. and N.H. Altman. 1990. Canine lingual acinic cell carcinoma (clear cell variant) of minor salivary gland. Vet Pathol 27(3):203–205. Brunnert, S.R. and N.H. Altman. 1991. Identification of immunoglobulin light chains in canine extramedullary plasmacytomas by thioflavine T and immunohistochemistry. J Vet Diagn Invest 3(3):245–251. Buback, J.L., H.W. Boothe, G.L. Carroll, et al. 1996. Comparison of three methods for relief of pain after ear canal ablation in dogs. Vet Surg 25:380–385. Buchholz, J., M. Wergin, H. Walt, et al. 2007. Photodynamic therapy of feline cutaneous squamous cell carcinoma using a newly developed liposomal photosensitizer: Preliminary results concerning drug safety and efficacy. J Vet Intern Med 21(4):770–775. Burek, K.A., R.J. Munn, and B.R. Madewell. 1994. Metastatic adenocarcinoma of a minor salivary gland in a cat. J Vet Intern Med Series A 41(6):485–490. Carberry, C.A., J.A. Flanders, W.I. Anderson, et al. 1987. Mast cell tumor in the mandibular salivary gland in a dog. Cornell Vet 77(4):362–366. Carberry, C.A., J.A. Flanders, H.J. Harvey, et al. 1988. Salivary gland tumors in dogs and cats: A literature and case review. J Am Anim Hosp Assoc 24:561–567. Carlotti, D.N. 1991. Diagnosis and medical treatment of otitis externa in dogs and cats. J Small Anim Pract 32:394–400. Carpenter, J.L. and M. Bernstein. 1991. Malignant mixed (pleomorphic) mandibular salivary gland tumors in a cat. J Am Anim Hosp Assoc 27:581–583. Clarke, R.E. 1991. Cryosurgical treatment of cutaneous squamous cell carcinoma. Australian Vet Pract 21:148–153. De Lorenzi, D., U. Bonfanti, C. Masserdotti, et al. 2005. Fine-needle biopsy of external ear canal masses in the cat: Cytologic results and histologic correlations in 27 cases. Vet Clin Pathol 34(2):100–105. de Vos, J.P., A.G.O. Burm, and B.P. Focker. 2004. Results from the treatment of advanced stage squamous cell carcinoma of the nasal planum in cats, using a combination of intalesional carboplatin and superficial radiotherapy: A pilot study. Vet Comp Oncol 2:75–81. Devitt, C.M., H.B. Seim, III, R. Willer, et al. 1997. Passive drainage versus primary closure after total ear canal ablation-lateral bulla osteotomy in dogs: 59 dogs (1985–1995). Vet Surg 26(3):210– 216. Dorn, C.R., D.O. Taylor, and R. Schneider. 1971. Sunlight exposure and risk of developing cutaneous and oral squamous cell carcinomas in white cats. J Natl Cancer Inst 46(5):1073–1078. Dunning, D. 2003. Oral cavity. In Textbook of Small Animal Surgery, third edition. pp. 553–572. D. Slatter, editor. Saunders: Philadelphia, PA. Evans, A.G., B.R. Madewell, and A.A. Stannard. 1985. A trial of 13-cisretinoic acid for treatment of squamous cell carcinoma and preneoplastic lesions of the head in cats. Am J Vet Res 46(12): 2553–2557. Evans, S.M. and D.E. Thrall. 1983. Postoperative orthovoltage radiation therapy of parotid salivary gland adenocarcinoma in three dogs. J Am Vet Med Assoc 182(9):993–994. Fahie, M.A. and M.M. Smith. 1997. Axial pattern flap based on the superficial temporal artery in cats: An experimental study. Vet Surg 26(2):86–89. Faustino, A.M. and P. Dias Pereira. 2007. A salivary malignant myoepithelioma in a dog. Vet J 173(1):223–226. Fidel, J.L., E. Egger, H. Blattmann, et al. 2001. Proton irradiation of feline nasal planum squamous cell carcinomas using an accelerated protocol. Vet Radiol Ultrasound 42(6):569–575.

Head and Neck Tumors  115 Gilger, B.C., R.D. Whitley, and S.A. McLaughlin. 1994. Modified lateral orbitotomy for removal of orbital neoplasms in two dogs. Vet Surg 23(1):53–58. Goodfellow, M., A. Hayes, S. Murphy, et al. 2006. A retrospective study of strontium-90 plesiotherapy for feline squamous cell carcinoma of the nasal planum. J Feline Med Surg 8(3):169– 176. Habin, D.J. and R.W. Else. 1995. Parotid salivary gland adenocarcinoma with bilateral ocular and osseous metastases in a dog. J Small Anim Pract 36(10):445–449. Hammer, A., D. Getzy, G. Ogilvie, et al. 2001. Salivary gland neoplasia in the dog and cat: Survival times and prognostic factors. J Am Anim Hosp Assoc 37(5):478–482. Hardie, E.M., K.E. Linder, and A.P. Pease. 2008. Aural cholesteatoma in twenty dogs. Vet Surg 37(8):763–770. Hayden, D.W. 1976. Squamous cell carcinoma in a cat with intraocular and orbital metastases. Vet Pathol 13(5):332–336. Hedlund, C.S. 2002. Surgery of the digestive system: Salivary mucoceles. In Small Animal Surgery, second edition. pp. 302–307. T.W. Fossum, editor. Mosby: St. Louis. Igarashi, Y. and J. Suzuki. 1985. Cochlear ototoxicity of chlorhexi­ dine gluconate in cats. Arch Otorhinolaryngol 242(2):167– 176. Indrieri, R.J. and R.F. Taylor. 1984. Vestibular dysfunction caused by squamous cell carcinoma involving the middle ear and inner ear in two cats. J Am Vet Med Assoc 184(4):471–473. Jha, A.K., G. Prasiko, H. Mod, et al. 2006. Curative interstitial brachytherapy for early stage carcinoma lip. JNMA J Nepal Med Assoc 45(162):252–257. Kim, H., M. Nakaichi, K. Itamoto, et al. 2008. Malignant mixed tumor in the salivary gland of a cat. J Vet Sci 9(3):331–333. Kipar, A., W. Baumgärtner, and E. Burkhardt. 1995. Round cell sarcomas of possible myelomonocytic origin localized at the lip of aged dogs. J Vet Intern Med Series A 42(3):185–200. Kirpensteijn, J., S.J. Withrow, and R.C. Straw. 1994. Combined resection of the nasal planum and premaxilla in three dogs. Vet Surg 23(5):341–346. Kneissl, S. and A. Probst. 2006. Magnetic resonance imaging features of presumed normal head and neck lymph nodes in dogs. Vet Radiol Ultrasound 47(6):538–541. Lambrechts, N.E. and J. Pearson. 2001. Cervical teratoma in a dog. J S Afr Vet Assoc 72(1):49–51. Lana, S.E., G.K. Ogilvie, S.J. Withrow, et al. 1997. Feline cutaneous squamous cell carcinoma of the nasal planum and the pinnae: 61 cases. J Am Anim Hosp Assoc 33(4):329–332. Lane, I.F. and D.G. Hall. 1992. Adenocarcinoma of the middle ear with osteolysis of the tympanic bulla in a cat. J Am Vet Med Assoc 201(3):463–465. Lanz, O.I. and B.C. Wood. 2004. Surgery of the ear and pinna. Vet Clin North Am Small Anim Pract 34(2):567–599. Lascelles, B.D., R.A. Henderson, B. Seguin, et al. 2004. Bilateral rostral maxillectomy and nasal planectomy for large rostral maxillofacial neoplasms in six dogs and one cat. J Am Anim Hosp Assoc 40(2):137–146. Lascelles, B.D., A.T. Parry, M.F. Stidworthy, et al. 2000. Squamous cell carcinoma of the nasal planum in 17 dogs. Vet Rec 147(17): 473–476. Lester, S. and K. Pratschke. 2003. Central hemimaxillectomy and reconstruction using a superficial temporal artery axial pattern flap in a domestic short hair cat. J Feline Med Surg 5(4):241– 244. Little, C.J.L., G.R. Pearson, and J.G. Lane. 1989. Neoplasia involving the middle ear cavity of dogs. Vet Rec 124(3):54–57.

London, C.A., R.R. Dubilzeig, D.M. Vail, et al. 1996. Evaluation of dogs and cats with tumors of the ear canal: 145 cases (1978–1992). J Am Vet Med Assoc 208(9):1413–1418. Louw, G.J. and S.J.E.M. Van Schouwenburg. 1984. A case of a highly invasive carcinoma of a salivary gland in a crossbred dog. J S Afr Vet Assoc 55: 131–132. Lucke, V.M. 1987. Primary cutaneous plasmacytomas in the dog and cat. J Small Anim Pract 28:49–55. Lucroy, M.D., K.M. Vernau, V.F. Samii, et al. 2004. Middle ear tumours with brainstem extension treated by ventral bulla osteotomy and craniectomy in two cats. Vet Comp Oncol 2:234–242. Lurie, D.M., B. Seguin, P.D. Schneider, et al. 2006. Contrast-assisted ultrasound for sentinel lymph node detection in spontaneously arising canine head and neck tumors. Invest Radiol 41(4): 415–421. Marino, D.J., J.M. MacDonald, D.T. Matthiesen, et al. 1994. Results of surgery in cats with ceruminous gland adenocarcinoma. J Am Anim Hosp Assoc 30:54–58. Marino, D.J., J.M. MacDonald, D.T. Matthiesen, et al. 1993. Results of surgery and long-term follow-up in dogs with ceruminous gland adenocarcinoma. J Am Anim Hosp Assoc 29:560–563. Matthiesen, D.T. and T. Scavelli. 1990. Total ear canal ablation and lateral bulla osteotomy in 38 dogs. J Am Anim Hosp Assoc 26:257–267. Mazzullo, G., A. Sfacteria, N. Iannelli, et al. 2005. Carcinoma of the submandibular salivary glands with multiple metastases in a cat. Vet Clin Pathol 34(1):61–64. McAnulty, J.F., A. Hattel, and C.E. Harvey. 1995. Wound healing and brain stem auditory evoked potentials after experimental total ear canal ablation with lateral tympanic bulla osteotomy in dogs. Vet Surg 24(1):1–8. Militerno, G., R. Bazzo, and P.S. Marcato. 2005. Cytological diagnosis of mandibular salivary gland adenocarcinoma in a dog. J Vet Intern Med. Series A 52(10):514–516. Moisan, P.G. and G.L. Watson. 1996. Ceruminous gland tumors in dogs and cats: A review of 124 cases. J Am Anim Hosp Assoc 32(5):448–452. Morrison, W.B. 2002. Paraneoplastic syndromes and the tumors that cause them. In Cancer in Dogs and Cats: Medical and Surgical Management, second edition. pp. 731–744. W.B. Morrison, editor. Teton NewMedia: Jackson, WY. Nieweg, O.E., P.J. Tanis, and B.B. Kroon. 2001. The definition of a sentinel node. Ann Surg Oncol 8(6):538–541. Nyman, H.T., A.T. Kristensen, I.M. Skovgaard, et al. 2005. Characterization of normal and abnormal canine superficial lymph nodes using gray-scale B-mode, color flow mapping, power, and spectral Doppler ultrasonography: A multivariate study. Vet Radiol Ultrasound 46(5):404–410. Oyamada, T., H. Okujima, R. Ando, et al. 2007. A case of malignant mixed salivary tumor composed of squamous cell carcinoma and osteosarcoma in a cat. J Japan Vet Med Assoc 60:724–728. Pavletic, M.M. 1999. Facial reconstruction. In Atlas of Small Animal Reconstructive Surgery, second edition. pp. 297–327. Saunders: Philadelphia, PA. Peaston, A.E., M.W. Leach, and R.J. Higgins. 1993. Photodynamic therapy for nasal and aural squamous cell carcinoma in cats. J Am Vet Med Assoc 202(8):1261–1265. Pentlarge, V.W. 1984. Peripheral vestibular disease in a cat with middle and inner ear squamous cell carcinoma. Compendium on Continuing Education for the Practicing Veterinarian 6:731– 734.

116  Veterinary Surgical Oncology Pérez-Martínez, C., R.A. García-Fernández, L.E. Reyes Avila et al. 2000. Malignant fibrous histiocytoma (giant cell type) associated with a malignant mixed tumor in the salivary gland of a dog. Vet Pathol 37(4):350–353. Psalla, D., C. Geigy, M. Konar, et al. 2008. Nasal acinic cell carcinoma in a cat. Vet Pathol 45(3):365–368. Ramos-Vara, J.A., M.E. Beissenherz, M.A. Miller, et al. 2000. Retrospective study of 338 canine oral melanomas with clinical, histologic, and immunohistochemical review of 129 cases. Vet Pathol 37(6):597–608. Roberts, W.G., M.K. Klein, M. Loomis, et al. 1991. Photodynamic therapy of spontaneous cancers in felines, canines, and snakes with chloro-aluminum sulfonated phthalocyanine. J Natl Cancer Inst 83(1):18–23. Rogers, K.S. 1988. Tumors of the ear canal. Vet Clin North Am Small Anim Prac. 18(4):859–868. Rohleder, J.J., J.C. Jones, R.B. Duncan, et al. 2006. Comparative performance of radiography and computed tomography in the diagnosis of middle ear disease in 31 dogs. Vet Radiol Ultrasound 47(1):45–52. Rowland, P.H., B.A. Valentine, K.E. Stebbins, et al. 1991. Cutaneous plasmacytomas with amyloid in six dogs. Vet Pathol 28(2): 125–130. Ruslander, D., B. Kaser-Hotz, and J.C. Sardinas. 1997. Cutaneous squamous cell carcinoma in cats. Compendium on Continuing Education for the Practicing Veterinarian 19:1119–1129. Salisbury, S.K., D.C. Richardson, and G.C. Lantz. 1986. Partial maxillectomy and premaxillectomy in the treatment of oral neoplasia in the dog and cat. Vet Surg 15:16–26. Salvadori, C., C. Cantile, and M. Arispici. 2004. Meningeal carcinomatosis in two cats. J Comp Pathol 131(2–3):246–251. Schultheiss, P.C. 2006. Histologic features and clinical outcomes of melanomas of lip, haired skin, and nail bed locations of dogs. J Vet Diagn Invest 18(4):422–425. Simpson, A.M., L.L. Ludwig, S.J. Newman, et al. 2004. Evaluation of surgical margins required for complete excision of cutaneous mast cell tumors in dogs. J Am Vet Med Assoc 224(2):236– 240. Smeak, D.D. and N. Inpanbutr 2005. Lateral approach to subtotal bulla osteotomy in dogs: Pertinent anatomy and procedural details. Compendium on Continuing Education for the Practicing Veterinarian 27:377–384. Smith, M.M. 1995. Surgical approach for lymph node staging of oral and maxillofacial neoplasms in dogs. J Am Anim Hosp Assoc 31(6):514–518. Smith, M.M. 2002. Surgical approach for lymph node staging of oral and maxillofacial neoplasms in dogs. J Vet Dent 19(3): 170–174. Smrkovski, O.A., A.K. LeBlanc, S.H. Smith, et al. 2006. Carcinoma ex pleomorphic adenoma with sebaceous differentiation in the mandibular salivary gland of a dog. Vet Pathol 43(3):374–377. Sozmen, M., P.J. Brown, and J.W. Eveson. 2002. Sebaceous carcinoma of the salivary gland in a cat. J Vet Intern Med Series A 49(8):425–427. Sozmen, M., P.J. Brown, and J.W. Eveson. 2003. Salivary gland basal cell adenocarcinoma: A report of cases in a cat and two dogs. J Vet Intern Med Series A 50(8):399–401. Spangler, W.L. and M.R. Culbertson. 1991. Salivary gland disease in dogs and cats: 245 cases (1985–1988). J Am Vet Med Assoc 198(3):465–469. Stell, A.J., J.M. Dobson, and K. Langmack. 2001. Photodynamic therapy of feline superficial squamous cell carcinoma using topical 5-aminolaevulinic acid. J Small Anim Pract 42(4):164–169.

Stone, E.A., M.H. Goldschmidt, and M.P. Littman. 1983. Squamous cell carcinoma of the middle ear in a cat. J Small Anim Pract 24:647–651. Stonehewer, J., A.J. Mackin, S. Tasker, et al. 2000. Idiopathic phenobarbital-responsive hypersialosis in the dog: An unusual form of limbic epilepsy. J Small Anim Pract 41:416–421. Swaim, S.F. and R.A. Henderson Jr. 1997. Wounds of the Head. In Small Animal Wound Management, second edition. pp. 191–233. Williams & Wilkins: Baltimore, MD. ter Haar, G. 2006. Inner ear dysfunction related to ear disease in dogs and cats. Eur J Companion Anim Prac 16:127–136. Theon, A.P., P.Y. Barthez, B.R. Madewell, et al. 1994. Radiation therapy of ceruminous gland carcinomas in dogs and cats. J Am Vet Med Assoc 205(4):566–569. Theon, A.P., B.R. Madewell, V.I. Shern, et al. 1995. Prognostic factors associated with radiotherapy of squamous cell carcinoma of the nasal plane in cats. J Am Vet Med Assoc 206(7):991–996. Theon, A.P., M.K. VanVechten, and B.R. Madewell. 1996. Intratumoral administration of carboplatin for treatment of squamous cell carcinomas of the nasal plane in cats. Am J Vet Res 57(2):205–210. Thomsen, B.V. and R.K. Myers. 1999. Extraskeletal osteosarcoma of the mandibular salivary gland in a dog. Vet Pathol 36(1):71–73. Thomson, M. 2007. Squamous cell carcinoma of the nasal planum in cats and dogs. Clin Tech in Small Anim Pract 22(2):42–45. Todoroff, R.J. and R.S. Brodey. 1979. Oral and pharyngeal neoplasia in the dog: A retrospective survey of 361 cases. J Am Vet Med Assoc 175(6):567–571. Trevor, P.B. and R.A. Martin. 1993. Tympanic bulla osteotomy for treatment of middle-ear disease in cats: 19 cases (1984–1991). J Am Vet Med Assoc 202(1):123–128. Turk, M.A.M., G.C. Johnson, A.M. Gallina, et al. 1983. Canine granular cell tumour (myoblastoma): A report of four cases and review of the literature. J Small Anim Pract 24:637–645. Vail, D.M. and S.J. Withrow. 2007. Tumors of the skin and subcutaneous tissues. In Withrow & MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 375–401. S.J. Withrow and D.M. Vail, editors. Saunders Elsevier: St. Louis, MO. Van Vechten, M.K. and A.P. Theon. 1993. Strontium-90 plesiotherapy for treatment of early squamous cell carcinoma of the nasal planum in 25 cats. Proceed Vet Cancer Soc p. 107. Vos, J.H. and I. van der Gaag. 1987. Canine and feline oral-pharyngeal tumours. J Vet Intern Med Series A 34(6):420–427. Wells, G.A.H. and M. Robinson. 1975. Mixed tumour of salivary gland showing histological evidence of malignancy in a cat. J Comp Pathol 85(1):77–85. White, R.A.S. and C.J. Pomeroy. 1990. Total ear canal ablation and lateral bulla osteotomy in the dog. J Small Anim Pract 31:547–553. Wiedmeyer, C.E., M.S. Whitney, L.D. Dvorak, et al. 2003. Mass in the laryngeal region of a dog. Vet Clin Pathol 32(1):37–39. Williams, L.E. and R.A. Packer. 2003. Association between lymph node size and metastasis in dogs with oral malignant melanoma: 100 cases (1987–2001). J Am Vet Med Assoc 222(9):1234–1236. Williams, J.M. and R.A.S. White. 1992. Total ear canal ablation combined with lateral bulla osteotomy in the cat. J Small Anim Pract 33:225–227. Withrow, S.J. 2007. Cancer of the gastrointestinal tract: Salivary gland cancer. In Withrow & MacEwen’s Small Animal Clinical Oncology, fourth edition. S.J. Withrow and D.M. Vail, editors p. 476. Saunders Elsevier: St. Louis, MO. Withrow, S.J. 2007. Tumors of the respiratory system: Cancer of the nasal planum. In Withrow & MacEwen’s Small Animal Clinical

Head and Neck Tumors  117 Oncology, fourth edition. pp. 511–515. S.J. Withrow and D.M. Vail, editors. Saunders Elsevier: St. Louis, MO. Withrow, S.J. and R.C. Straw. 1990. Resection of the nasal planum in nine cats and five dogs. J Am Anim Hosp Assoc 26:219–223. Wolfe, T.M., S.W. Bateman, L.K. Cole, et al. 2006. Evaluation of a local anesthetic delivery system for the postoperative analgesic management of canine total ear canal ablation—A randomized, controlled, double-blinded study. Vet Anaesth Analg 33(5):328–339.

Worley, D., T. Scanlon, A. Reiter, et al. 2007. Lymphatic staging and sentinel lymph node identification in canine oral malignant melanoma. Proceed Vet Cancer Soc 27:34. Yates, G., B. Landon, and G. Edwards. 2007. Investigation and clinical application of a novel axial pattern flap for nasal and facial reconstruction in the dog. Aust Vet J 85(3):113–118. Zur, G. 2005. Bilateral ear canal neoplasia in three dogs. Vet Derm 16(4):276–280.

6 Oral tumors Julius M. Liptak, B. Duncan X. Lascelles

Introduction Oral tumors are common in both cats and dogs, with cancers of the oral cavity accounting for 3%–12% and 6% of all tumors in these species, respectively (Patnaik et al. 1975; Dorn and Priester 1976; Hoyt and Withrow 1984; Vos and van der Gaag 1987; Stebbins et al. 1989). The surgical oncologist plays a pivotal role in the diagnosis, staging, and treatment of cancer in cats and dogs with oral tumors. For instance, an incisional biopsy is often required for definitive diagnosis of an oral tumor, and this biopsy needs to be planned appropriately so that the biopsy site will not have a negative impact on the curative-intent treatment plan. Also, clinical staging requires excision of the regional lymph nodes, which is important for determining postoperative treatment plans and prognosis. And the most common definitive treatment of oral tumors is surgical excision. For the purposes of this chapter, oral tumors will include tumors involving the mandible, maxilla, palate, and tongue.

Diagnosis and Clinical Staging History and clinical signs Most cats and dogs with oral cancer present with a mass in the mouth that is noticed by the owner. Cancer in the caudal pharynx, however, is rarely seen by the owner, and the animal may present with signs of hypersalivation, exophthalmos or facial swelling, epistaxis, weight loss, halitosis, bloody oral discharge, dysphagia or pain on opening the mouth, or occasionally cervical lymphadenopathy (especially squamous cell carcinoma [SCC] of the tonsil) (Kosovsky et al. 1991; Schwarz et al. 1991a, 1991b; Wallace et al. 1992; Reeves et al. 1993). Loose teeth, especially in an animal with generally good

dentition, may be indicative of underlying neoplastic bone lysis, especially in cats (Madewell et al. 1976). A complete examination of the oral cavity during annual health checks is recommended to screen for oropharyngeal masses as this may permit earlier diagnosis, better treatment, and improved prognosis. If a mass is noted during examination of the oral cavity, then accurate notes should be written in the medical records of the size and anatomic location of the mass in relation to adjacent dentition. Furthermore, a photo of the mass can assist the surgeon in determining the best approach for biopsy and definitive surgery. Diagnosis and clinical staging The diagnosis and clinical staging of animals with oropharyngeal masses is imperative before definitive surgical excision. A biopsy is required for definitive diagnosis, and this will assist the clinician in determining biologic behavior and prognosis. Clinical staging consists of evaluating the extent of the local tumor and the presence of metastatic disease. The regional lymph nodes and lungs are the two most common sites of metastasis in cats and dogs with oral tumors (Liptak and Withrow 2007). The procedures required for the diagnosis and clinical staging of animals with oral cancer can usually be performed under a short general anesthesia. Diagnosis A large incisional biopsy is often required for a definitive diagnosis. Cytologic touch or aspiration preparations are usually not rewarding and can result in an incorrect diagnosis because many oral tumors are associated with a high degree of necrosis and inflammation (Liptak and Withrow 2007). Dogs with exophytic or ulcerated masses

Veterinary Surgical Oncology, First Edition. Edited by Simon T. Kudnig, Bernard Séguin. © 2012 by John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.

119

120  Veterinary Surgical Oncology

will generally tolerate a deep wedge or core punch biopsy without general anesthesia. Biopsy is recommended in the diagnostic workup of cats and dogs with an oral mass. Biopsy is recommended to differentiate benign from malignant disease, for owners basing their treatment options on prognosis, and when other treatment modalities, such as radiation therapy, may be preferable. Oral cancers are commonly infected, inflamed, or necrotic, and it is important to obtain a large specimen. Cautery may distort the specimen and should be used only for hemostasis after blade incision or punch biopsy. Large samples of healthy tissue at the edge and center of the lesion will increase the diagnostic yield, but care must be taken not to contaminate normal tissue, which cannot be removed with surgery or included in the radiation field. Biopsies should always be performed from within the oral cavity and not through the lip to avoid seeding tumor cells in normal skin and compromising curative-intent surgical resection. For small lesions (e.g., peripheral odontogenic fibromas, papillomas, or small labial mucosal melanoma), curative-intent resection (excisional biopsy) may be undertaken at the time of initial evaluation. However, accurate notes should be included in the medical records and/or there should be photographic evidence, to detail the size and anatomical location of the mass if excision is incomplete and further treatment is required. For more extensive disease, waiting for biopsy results is recommended so that appropriate treatment plans can be formulated.

Figure 6.1.  CT scan of a dog with a zygomatic squamous cell carcinoma extending into the inferior orbit and caudal maxilla. CT scans provide superior detail on the extent of the tumor and tumor invasion and is the preferred imaging modality for planning of surgical resection of tumors involving the maxilla, orbit, and hard palate.

Clinical staging—local tumor imaging Cancers that are adherent to or arising from bones of the mandible, maxilla, or palate should be imaged under general anesthesia to determine the presence of bone lysis and the extent of local disease. Regional radiographs include open mouth, intraoral, oblique lateral, and ventrodorsal or dorsoventral projections (Dhaliwal et al. 1998). Bone lysis is not radiographically evident until 40% or more of the cortex is destroyed, and hence apparently normal radiographs do not exclude bone invasion (Liptak and Withrow 2007). Advanced imaging modalities are now widely available, and these are recommended for imaging of oral tumors, particularly tumors arising from the maxilla, palate, and caudal mandible (Figures 6.1 and 6.2). Computed tomography (CT) scans are generally preferred to magnetic resonance imaging (MRI) because of superior bone detail, but both CT or MRI scans will provide more information on the local extent of the tumor than will regional radiographs. This information is important for planning the definitive surgical procedure (or radiation therapy if indicated).

Figure 6.2.  CT scan of a dog with a multilobular osteochondrosarcoma arising from the caudal mandible. CT scans provide superior detail on the extent of the tumor and tumor invasion and is the preferred imaging modality for planning of surgical resection of tumors involving the caudal mandible.

Oral Tumors  121

Clinical staging—regional lymph nodes Regional lymph nodes should be carefully palpated for enlargement or asymmetry. However, caution should be exercised when making clinical judgments based on palpation alone as lymph node size is not an accurate predictor of metastasis. In one study of 100 dogs with oral melanoma, 40% of dogs with normal-sized lymph nodes had metastasis and 49% of dogs with enlarged lymph nodes did not have metastasis (Williams and Packer 2003). Furthermore, the regional lymph nodes include the mandibular, parotid, and medial retropharyngeal lymph nodes; however, the parotid and medial retropharyngeal lymph nodes are not externally palpable (Smith 1995). Additionally, only 55% of 31 cats and dogs with metastasis to the regional lymph nodes had metastasis to the mandibular lymph nodes (Herring et al. 2002). Preoperative assessment of the regional lymph nodes is difficult. Currently, lymph node aspirates are recommended for all animals with oral tumors, regardless of the size or degree of fixation of the lymph nodes (Herring et al. 2002; Williams and Packer 2003). It is hoped that in the near future sentinel lymph node assessment will become more widely accepted and practiced as this may permit the preoperative diagnosis of metastatic lymph nodes without more aggressive en bloc surgical excisions of the regional lymph nodes. Methods to detect sentinel lymph nodes in people with head and neck cancer include lymphoscintigraphy, intraoperative blue dyes, and intraoperative gamma probes (Balogh et al. 2002). Lymphoscintigraphy, intraoperative dyes, and contrastenhanced ultrasonography have been described in dogs with various tumors, including head and neck cancer (Balogh et al. 2002; Lurie et al. 2006). En bloc resection of the regional lymph nodes has been described and, although the therapeutic benefit of this approach is unknown, it may provide valuable staging information (Smith 1995; Herring et al. 2002). The skin is incised from the rostral and proximal aspect of the vertical ear canal, ventral to the caudal aspect of the zygomatic arch, to the bifurcation of the external jugular vein (Smith 1995). The platysma and parotidoauricularis muscles are incised to reveal fascia and loose areolar tissue covering the vertical ear canal and masseter muscle. Incision of the areolar tissue over the ventral aspect of the zygomatic arch exposes the parotid lymphocentrum, which has one to three lymph nodes, along the rostral edge of the parotid salivary gland (Smith 1995). The mandibular lymphocentrum, which contains one to five lymph nodes, is located between the bifurcation of the jugular vein and division of the lingofacial vein into its lingual and facial branches (Smith 1995). The medial retropharyngeal lymphocentrum,

which usually consists of one elongated lymph node on the lateral aspect of the thyropharyngeus muscle, is exposed by incising the adventitia along the caudal aspect of the mandibular salivary gland and retracting the manidbular salivary gland rostrally and the bra­ chiocephalicus and sternocephalicus muscles dorsally (Smith 1995). Clinical staging–distant metastasis The final step in the clinical staging of animals with oral tumors is imaging of the thoracic cavity for metastasis to the lungs. Three-view thoracic radiographs (right and left lateral projections and either dorsoventral or ventrodorsal projection) are generally recommended. Helical CT scans should be considered for animals with highly metastatic tumor types, such as oral malignant melanoma, as CT scans are significantly more sensitive in detecting pulmonary metastatic lesions compared to radiographs (Nemanic et al. 2006). Based on these diagnostic steps, oral tumors are then clinically staged according to the World Health Organization (WHO) staging scheme (Table 6.1) (Owen 1980).

General Surgical Considerations Surgical excision is the most commonly used modality for treatment of the local oral tumor. The surgical approach depends on the type and location of the oral tumor. Except for peripheral odontogenic fibroma, the majority of tumors involving the mandible, maxilla, and hard palate have some underlying bone involvement, and surgical resection of these should include bony margins to increase the likelihood of complete excision. Radical surgeries such as mandibulectomy and maxillectomy are well tolerated by cats and dogs. These procedures are indicated for oral tumors involving the mandible and maxilla, particularly lesions with extensive bone invasion and tumor types that have poor sensitivity to radiation therapy (Withrow and Holmberg 1983; Bradley et al. 1984; White et al. 1985; Withrow et al. 1985; Salisbury et al. 1986; Salisbury and Lantz 1988; Kosovsky et al. 1991; Schwarz et al. 1991a, 1991b; White 1991; Wallace et al. 1992; Kirpensteijn et al. 1994; Lascelles et al. 2003, 2004). Minimum margins of at least 2 cm, and preferably 3 cm, are recommended for malignant cancers such as SCC, malignant melanoma, and fibrosarcoma in the dog. If possible, SCC in the cat should be treated with surgical margins greater than 2 cm because of high local recurrence rates. However, these margins may not be possible without significant morbidity because of the extent of the tumor. Moreover, margins required for resection of benign and malignant oral tumors have not been investigated, and lesser

122  Veterinary Surgical Oncology Table 6.1.  Clinical staging (TNM) of oral tumors in dogs and cats. Primary Tumor (T)   Tis Tumor in situ   T1 Tumor 4 cm in diameter at greatest dimension    T3a Without evidence of bone invasion    T3b With evidence of bone invasion Regional Lymph Nodes (N)   N0 No regional lymph node metastasis   N1 Movable ipsilateral lymph nodes    N1a No evidence of lymph node metastasis    N1b Evidence of lymph node metastasis   N2 Movable contralateral lymph nodes    N2a No evidence of lymph node metastasis    N2b Evidence of lymph node metastasis   N3 Fixed lymph nodes Distant Metastasis (M)   M0 No distant metastasis   M1 Distant metastasis [specify site(s)] Stage Grouping

Tumor (T)

Nodes (N)

Metastasis (M)

I II III IV

T1 T2 T3 Any T Any T Any T

N0, N1a, N2a N0, N1a, N2a N0, N1a, N2a N1b N2b, N3 Any N

M0 M0 M0 M0 M0 M1

margins may result in acceptable rates of local tumor control (Syrcle et al. 2008). Where 3 cm margins are not possible without significant risk of morbidity, 1 cm margins of normal tissue beyond either the grossly visible tumor or the extent of the tumor as determined by imaging, whichever is greater, may be acceptable for malignant oral tumors and 0.5 cm margins for benign tumors (Syrcle et al. 2008). Greater margins are recommended if possible because these guidelines have not been validated and there is little difference in functional and cosmetic outcome with more extensive surgery.

Rostral and segmental bony resections may be sufficient for benign lesions and rostral SCC in dogs. Larger resections, such as hemimandibulectomy, hemimaxillectomy, orbitectomy, and radical maxillectomy, are necessary for more aggressive malignant tumors, especially fibrosarcoma, and any tumor in a more caudal location (Withrow and Holmberg 1983; Bradley et al. 1984; White et al. 1985; Withrow et al. 1985; Salisbury et al. 1986; Salisbury and Lantz 1988; Kosovsky et al. 1991; Salisbury 1991; Schwarz et al. 1991a, 1991b; White 1991; Wallace et al. 1992; Kirpensteijn et al. 1994; Lascelles et al. 2003, 2004; Verstraete 2005). Reconstruction following mandibular resection has been described, but is rarely necessary because of good postoperative function and cosmetic appearance (White et al. 1985; Boudrieau et al. 1994, 2004; Bracker and Trout 2000; Spector et al. 2007). The surgical techniques for oral tumor resection and postoperative management are described in detail below. A detailed knowledge of the regional anatomy is important for a successful outcome and to minimize the risk of complications. The anatomy should be reviewed prior to surgery, in combination with either CT or MRI images of the patient, to plan the surgical approach, resection, and reconstruction. Cat and dog skulls should be available for intraoperative orientation and planning. Antibiotics In general, prophylactic antibiotics are usually not necessary for intraoral procedures (i.e., those not involving a skin incision) because the risk of infection is low due short surgical times and excellent vascular supply to the oral cavity. However, some surgeons prefer the use of prophylactic antibiotics (such as a first-generation cephalosporin, clavulanate-potentiated amoxicillin, or ampicillin potentiated with sulbactam) administered prior to surgery, every 90 minutes during surgery, and every 6 hours for the first 24 hours after surgery (Dernell et al. 1998a). Analgesia Regional nerve blocks and systemic nonsteroidal antiinflammatory drugs and narcotics are recommended for perioperative analgesia to minimize morbidity and improve postoperative comfort (Beckman and Legendre 2002). Recent research also suggests minimizing perioperative pain may help decrease the rate of metastasis following oncologic surgery (Exadaktylos et al. 2006). See Table 6.2 for an outline of suggested analgesic protocols for cats and dogs undergoing maxillofacial resection and reconstruction.

Oral Tumors  123 Table 6.2.  Suggested analgesic protocols for oral surgery in cats and dogs. Analgesic Protocols

Cats

Dogs

Preoperative

Fentanyl patch placed the day before surgery (3–4 mcg/kg/hr) Intravenous NSAID once at induction or following full recovery if hemorrhage expected (e.g., carprofen 1–2 mg/kg [healthy cats] or meloxicam 0.1 mg/kg) Bupivicaine regional nerve blocks at induction Continuous rate IV infusion of • hydromorphone (0.005–0.01 mg/kg/hr), fentanyl (2–4 mcg/kg/hr), or morphine (0.05 mg/kg/hr) • medetomidine (1–4 mcg/kg/hr) • ketamine (0.1–0.5 mg/kg/hr) • Oral NSAID for 5–14 days at approved dose. Suggest meloxicam 0.1 mg/kg once on day 1 (perioperative dose), 0.05 mg/kg q 24 hrs for 4 days, 0.025 mg/kg q 24 hrs for 4 days, then 0.025 mg/kg q 48 hrs. Do not use if the cat is stressed or has any risk factors for gastrointestinal ulceration. • Oral opioid (buprenorphine via the oral transmucosal route, 10 mcg/kg q 8–12 hrs) for 7–10 days, or fentanyl patch replaced as required. • Oral opioid derivative (e.g., tramadol 4 mg/kg PO q 12 hrs) for 7–10 days, or fentanyl patch replaced as required.

Fentanyl patch placed the day before surgery (3–4 mcg/kg/hr) Intravenous NSAID once at induction or following full recovery if hemorrhage expected (e.g., Carprofen 4 mg/kg or meloxicam 0.1 mg/kg)

Perioperative

Discharge

Bupivicaine regional nerve blocks at induction Continuous rate IV infusion of • hydromorphone (0.005–0.02 mg/kg/hr), fentanyl (2–4 mcg/kg/hr), or morphine (0.1 mg/kg/hr) • medetomidine (1–2 mcg/kg/hr) • lidocaine (25–30 mcg/kg/min) • ketamine (2 mcg/kg/min) • Oral NSAID for 10–14 days at approved dose. Do not use if the dog is stressed or has any risk factors for gastrointestinal ulceration. • Oral opioid derivative (e.g., tramadol 4 mg/kg PO q 6–12 hrs) for 7–10 days, or fentanyl patch replaced as required.

Positioning and preparation

Surgical considerations

Animals are placed in the appropriate position and, if necessary, a mouth gag is placed on the lower or nonsurgical side. The exact positioning will vary from case to case. The surgeon should evaluate the positioning in each patient to ensure optimal access to tissues at all stages of the procedure. In some procedures (e.g., combined rostral maxillectomy and nasal planum resection), the animal may need to be repositioned during surgery to optimize access for resection and closure. Conforming vacuum packs and tape can be used to assist in positioning and maintaining position of the head. If necessary, the surgical site is clipped and surgically prepared. No clipping is required for intraoral procedures. For more caudal procedures, the periorbital hair is usually clipped, and the eye is included within the surgical field. The oral cavity is irrigated with 10% povidone-iodine solution, which is first diluted 1:10 with tap water (Dernell et al. 1998a). Towel clamps or staples are used to secure the drapes in such a manner as to allow for mobilization of the labial tissues.

For excision of malignant oral tumors, minimum margins include 3 cm of bone rostral and caudal to the tumor, based on either gross palpation or, preferably, imaging findings, and 1 cm of soft tissue (buccal, gingival, or palatine mucosa). Lesser margins can be used for benign oral tumors, and marginal excision, possibly combined with cryosurgery, is suitable for peripheral odontogenic fibroma (Dernell et al. 1998a; Liptak and Withrow 2007). The mandibular symphysis may act as a barrier for tumor invasion. If there is no evidence of rostral mandibular tumors crossing the mandibular symphysis, then excision of the symphysis, including the contralateral middle incisor, should be sufficient for bony margins. An oscillating saw is preferred for mandibular osteotomies, although pneumatic burrs, Gigli wire, and bone cutters can also be used. Osteotomes should be avoided for all nonsymphyseal mandibular osteotomies because of the risk of bone shattering. In the maxilla and hard palate, osteotomies can be performed with an oscillating saw, pneumatic burr, and/or

124  Veterinary Surgical Oncology

Figure 6.3.  Bone tunnels have been drilled into the hard palate to provide a secure two-layer closure (hard palate-to-labial submucosa and palatine mucoperiosteum-to-labial mucosa) following partial maxillectomy in a dog.

osteotome and mallet. Monofilament absorbable suture material (e.g., polydioxanone) is recommended for closure of oral defects because they maintain adequate tensile strength for prolonged periods and are relatively inert, which minimizes mucosal irritation and inflammation (Salisbury et al. 1986; Dernell et al. 1998a). A reverse-cutting swaged-on needle is preferred for suturing fibrous soft tissues of the oral cavity because of easier passing of the needle with minimal trauma and better suture purchase (Salisbury et al. 1986). Two-layer closures are preferred to one-layer closures because of a reduced risk of incisional dehiscence (Dernell et al. 1998a). Suture material should be passed through bone tunnels if possible, particularly in the hard palate and maxilla, because of increased holding power compared to soft tissue (Figure 6.3).

Surgical Approach to Tumors of the Mandible Anatomy The lower jaw consists of two hemimandibles that form a fibrocartilaginous symphysis rostrally and articulate caudally with the skull at the temporomandibular joint (Evans and Christensen 1979). The hemimandible consists of the body and ramus (Figure 6.4). Teeth erupt along the alveolar margin of the mandibular body. The mandibular canal is an important oncological consideration and extends along the body of the mandible. The inferior alveolar artery, vein, and nerve enter the mandibular canal caudally at the mandibular foramen, and the mental nerves innervating the lower lip and chin exit rostrally through three mental foramina (Evans and Christensen 1979). Tumors involv-

Figure 6.4.  Anatomy of the mandible. The inferior alveolar artery, vein, and nerve course through the mandibular foramen into the mandibular canal. The mandibular canal is an important consideration when resecting mandibular tumors as invasion into the mandibular canal necessitates a more aggressive subtotal or total mandibulectomy. (Reproduced with permission from Evans, H.E. and A. deLahunta, editors. 2000. The head. In Guide to the Dissection of the Dog, pp. 259–321. Saunders: Philadelphia.)

ing the mandibular body can theoretically extend along the mandibular canal and hence the caudal margins for mandibulectomy procedures of these tumors should extend caudal to the mandibular foramen to minimize the risk of incomplete tumor resection. The ramus consists of three prominent processes: the coronoid process on the dorsal aspect of the ramus, the condylar process on the caudal aspect of the ramus, and the angular process on the caudoventral aspect of the ramus (Figure 6.4) (Evans and Christensen 1979). The masseter muscle inserts on the lateral surface of the coronoid and angular processes, the temporal muscle on the medial aspect of the ramus, the pterygoid muscle on the medial and caudal aspects of the ramus and angular process, and the digastricus muscle along the ventral aspect of the mandibular body (Figure 6.5) (Evans and Christensen 1979). Mandibulectomy involves en bloc excision of a tumor of the lower jaw. Various mandibulectomy procedures have been described, including unilateral and bilateral rostral mandibulectomy, segmental mandibulectomy, caudal mandibulectomy, and subtotal and total hemimandibulectomy.

Oral Tumors  125

(a)

(c)

(b)

(c)

Figure 6.5.  Anatomy of the mandible showing the muscles of mastication. (A) Pterygoideus medialis and lateralis muscles. (B) Masseter and pterygoideus medialis muscles. (C) Origin of the temporalis and pterygoideus medialis and lateralis muscles. (D) Cutout to show the deep portion of the masseter muscle. Reproduced with permission from Evans, H.E. and G.C. Christensen, editors. 1979. Muscles. In Miller’s Anatomy of the Dog, pp. 269–410. Philadelphia: Saunders.

Rostral mandibulectomy—unilateral Unilateral rostral mandibulectomy is recommended for dogs with benign acanthomatous ameloblastoma or SCC lesions that are rostral to the second premolar tooth and do not cross the mandibular symphysis. Bilateral rostral mandibulectomy should be considered for these tumor types that cross the mandibular symphysis, and hemimandibulectomy is recommended for malignant tumors other than SCC in the rostral mandible or acanthomatous ameloblastoma or SCC lesions caudal to the second premolar tooth. For unilateral rostral mandibulectomy, the dog is positioned in lateral recumbency with affected side uppermost (Dernell et al. 1998a). The labial mucosa is incised with a minimum of 1 cm margins around the

mass (Figure 6.6A) (Dernell et al. 1998a). This incision is continued rostrally to the mandibular symphysis and caudally to the planned osteotomy site. The labial mucosa is reflected off the mandible with periosteal elevators to preserve the soft tissue of the lip that will be used later for reconstruction of the defect (Figure 6.6B). The sublingual and mandibular salivary gland ducts open at the sublingual caruncle in the frenulum of the tongue (Northrup et al. 2006). This should be preserved if possible (Dernell et al. 1998a), but excisional margins should not be compromised by this anatomic consideration and complications such as ranula formation are uncommon. The rostral osteotomy should include the mandibular symphysis, and hence the osteotomy is positioned eccentrically between the contralateral canine tooth and mandibular symphysis (Figure 6.6C). This

126  Veterinary Surgical Oncology

(a)

(b)

(c)

(e)

(d)

(f)

(g)

Figure 6.6.  Unilateral rostral mandibulectomy. (A) The labial and gingival mucosa are incised with a minimum of 1 cm margins from an acanthomatous ameloblastoma localized to the mandibular canine tooth. (B) The mucosa is then reflected off the underlying mandible (arrows) using periosteal elevators to expose the planned osteotomy sites and protect soft tissues from trauma during osteotomy of the mandible. (C) An eccentric rostral mandibular osteotomy should be performed with either an oscillating saw, osteotome and mallet, or bone cutters to include the mandibular symphysis. (D, E) The caudal osteotomy is performed with an oscillating saw with minimum margins of 1–2 cm for benign tumors, such as this acanthomatous ameloblastoma, and 2–3 cm for malignant tumors. (F) The resultant defect following removal of the unilateral rostral mandibular segment. (G) This defect is closed by suturing the sublingual mucosa to the labial mucosa in a single layer of either simple interrupted or simple continuous sutures using monofilament absorbable suture material.

osteotomy can be performed with an oscillating saw, biradial saw, or osteotome and mallet (Dernell et al. 1998a). A bone cutter may be sufficient in cats and small dogs. The position of the caudal osteotomy is based on tumor type, tumor dimensions, and imaging evidence of bone invasion. Margins of 1–2 cm for acanthomatous ameloblastoma and 2–3 cm for SCC caudal to the caudal limit of the mass or level of bone invasion should be sufficient. The caudal osteotomy should be performed with an oscillating saw and tapered at the occlusional

margin to minimize wound tension during closure (Figures 6.6D, E) (Dernell et al. 1998a). Gigli wire can also be used for the caudal osteotomy, but the mandible can shatter when an osteotome and mallet is used for this osteotomy; hence, this should be avoided. Following mandibulectomy, the sublingual mucosa is sutured to the labial mucosa in a single layer of a simple interrupted or simple continuous suture pattern using monofilament absorbable suture material (Figures 6.6F, G) (Dernell et al. 1998a).

Oral Tumors  127

Rostral mandibulectomy—bilateral

Segmental mandibulectomy

Bilateral rostral mandibulectomy is recommended for dogs with benign acanthomatous ameloblastoma or SCC lesions rostral to the first premolar tooth and across the mandibular symphysis. For larger lesions or malignant tumors other than SCC, the caudal osteotomies can be positioned as far caudally as the fourth premolar tooth. However, the risk of complications, particularly difficulty prehending food, is greater with more caudal osteotomies. For bilateral mandibulectomy, the dog can be positioned in either dorsal, lateral, or sternal recumbency depending on surgeon’s preference (Dernell et al. 1998a). Dorsal recumbency provides better exposure for dissection of the tumor and performing the osteotomies. Sternal recumbency provides superior exposure for wound closure. The surgical technique for bilateral mandibulectomy is similar to unilateral mandibulectomy. The labial mucosa is incised with a minimum of 1 cm margins around the mass and continued caudally to the planned osteotomy sites. The labial mucosa is reflected off the mandible with periosteal elevators to preserve the soft tissue of the lip, which will be used later for reconstruction of the defect (Figure 6.7A, B). The sublingual and mandibular salivary gland ducts open at the sublingual caruncle in the frenulum of the tongue (Northrup et al. 2006). This should be preserved if possible (Dernell et al. 1998a), but excisional margins should not be compromised by this anatomical consideration. The position of the caudal osteotomy depends on tumor type, tumor dimensions, and imaging evidence of bone invasion. Margins of 1–2 cm for acanthomatous ameloblastoma and 2–3 cm for SCC caudal to the caudal limit of the mass or level of bone invasion should be sufficient. The caudal osteotomy should be performed with an oscillating saw and tapered at the occlusional margin to minimize wound tension during closure (Figure 6.7C) (Dernell et al. 1998a). Stabilization of the remaining portions of the hemimandible has been described (Boudrieau et al. 1994; Bracker and Trout 2000; Boudrieau et al. 2004; Spector et al. 2007), but is not necessary because function and cosmetic appearance are not improved and the risk of complications is increased. For closure, V-shaped wedges of redundant skin can be removed either laterally or rostrally en bloc with the tumor or after tumor excision to create a soft tissue ridge or dam to minimize drooling and to improve cosmetic appearance (Dernell et al. 1998a). The sublingual mucosa is sutured to the labial mucosa in a single layer of simple interrupted sutures using monofilament absorbable suture material (Figure 6.7D) (Dernell et al. 1998a).

Segmental mandibulectomy is recommended for dogs with benign acanthomatous ameloblastoma or lowgrade malignant tumors, such as SCC, located in the midmandibular body. Furthermore, these tumors should not penetrate cortical bone. Segmental mandibulectomy is contraindicated for malignant tumors other than SCC. The dog is positioned in lateral recumbency. The labial and lingual mucosa are incised with a minimum of 1 cm margins around the mass (Figure 6.8A). The dissection is continued around the mandibular body with periosteal elevators until the mandibular body is fully exposed. Osteotomies are performed rostral and caudal to the mass with an oscillating saw or Gigli wire (Figure 6.8B). Margins of 1–2 cm for acanthomatous ameloblastoma and 2–3 cm for SCC should be sufficient for complete excision. For closure, the sublingual mucosa is sutured to the labial mucosa in a single layer of simple interrupted or simple continuous sutures using monofilament absorbable suture material (Figure 6.8C). Subtotal and total hemimandibulectomy Hemimandibulectomy is recommended for malignant tumors, particularly those with extensive involvement of the mandibular body. The only difference between subtotal and total hemimandibulectomy is the caudal margins. For subtotal hemimandibulectomy, the mandible is osteotomized caudal to the mandibular canal, and this is primarily indicated when the tumor can be excised with 3 cm caudal margins without sacrificing the ramus or temporomandibular joint. It should be noted, however, that preservation of these structures does not improve postoperative function or cosmesis; hence, subtotal hemimandibulectomy should not be performed if caudal margins will be compromised. Total hemimandibulectomy includes the temporo­ mandibular joint and ramus. This procedure is more aggressive and is indicated for tumors in which 3 cm caudal margins cannot be attained with subtotal hemimandibulectomy. Dogs are positioned in lateral recumbency for hemimandibulectomy (Dernell et al. 1998a). To improve exposure for total hemimandibulectomy, a full-thickness incision is performed extending from the commissure of the lip to the rostral aspect of the ramus (Dernell et al. 1998a). Branches of the facial artery and vein should be ligated or cauterized if they are encountered during this incision (Dernell et al. 1998a). The parotid salivary duct is usually dorsal to this incision but should be avoided if possible. This skin incision is not necessary for subtotal hemimandibulectomy.

(a)

(b)

(c)

(d)

Figure 6.7.  Bilateral rostral mandibulectomy. (A) A malignant melanoma involving the rostral mandible and crossing the symphyseal midline. (B) The labial mucosa is incised with minimum margins of 1 cm, and the mucosa is then reflected with periosteal elevators immediately caudal to the planned osteotomy site to protect soft tissues from trauma during osteotomy of the mandible. (C) The mandibular osteotomy is performed with an oscillating saw with minimum caudal margins of 1–2 cm for benign tumors and 2–3 cm for malignant tumors, such as this malignant melanoma. (D) The resultant defect can be closed primarily or V-shaped wedges of lip can be removed either laterally or rostrally to improve cosmetics and function. This defect is closed by suturing the sublingual mucosa to the labial mucosa in a single layer of either simple interrupted or simple continuous sutures using monofilament absorbable suture material.

(a)

(b)

(c)

Figure 6.8.  Segmental mandibulectomy. (A) The labial and gingival mucosa are incised with minimum margins of 1 cm from an acanthomatous ameloblastoma arising from the lateral alveolar ridge of the third premolar tooth, and the mucosa is then reflected with periosteal elevators immediately rostral and caudal to the planned osteotomy sites to protect soft tissues from trauma during osteotomy of the mandible. (B) The mandibular osteotomies are performed with either an oscillating saw or Gigli wire, with minimum margins of 1−2 cm. Note that this segmental mandibulectomy is not recommended for malignant tumors, and hence larger margins are not required. (C) The resultant defect is closed by suturing the sublingual mucosa to the labial mucosa in a single layer of either simple interrupted or simple continuous sutures using monofilament absorbable suture material.

128

Oral Tumors  129

For both subtotal and total hemimandibulectomy, the labial and buccal mucosa are incised with a minimum of 1 cm margins around the mass (Figure 6.9A) (Dernell et al. 1998a). The mucosal incisions are continued rostrally to the level of the planned ostoeotomy and caudally to the ramus (Figure 6.9B). The lateral border of the tongue is freed as the medial incision is continued rostrally into the sublingual mucosa. During this dissection, the mandibular and sublingual salivary ducts may be encountered. Some surgeons recommend identification and ligation of these ducts, but this is rarely necessary. The dissection is continued around the mandibular body with periosteal elevators until the mandibular body is fully exposed. The genioglossus, geniohyoideus, and mylohyoideus muscles are either transected (if within 1 cm margins of the tumor) or elevated from the medial aspect of the mandible if the tumor does not penetrate the medial cortex (Dernell et al. 1998a). Depending on the location of the tumor, the rostral osteotomy can either be performed through the mandibular symphysis or eccentrically between the contralateral canine tooth and mandibular symphysis. This osteotomy can be performed with either an oscillating saw or osteotome and mallet, whereas a bone cutter may be sufficient in cats and small dogs (Figure 6.9C). Separation of the mandibular symphysis permits lateral movement of the hemimandible and better exposure and visualization for caudal dissection. Importantly, the inferior alveolar artery and vein should be identified and ligated as they course over the lateral surface of the medial pterygoid muscle before entering the mandibular foramen (Figure 6.9D). For subtotal hemimandibulectomy, the caudal osteotomy is positioned at the rostral edge of the insertion of the masseter muscle (Figure 6.9E) (Dernell et al. 1998a). This osteotomy should be performed with an oscillating saw. For total hemimandibulectomy, the dissection is continued further caudally. Depending on the location of the tumor, the masseter muscle is either incised with 1 cm margins or elevated from the ventrolateral surface and ventral margin of ramus while retracting the hemimandible in a caudodorsal direction (Figure 6.9F, G); the digastricus muscle is elevated from its insertion along the caudoventral border of the mandibular body (Figure 6.9H); and the pterygoid muscles are elevated from their insertion on the medial aspect of the caudoventral surface of angle of mandible (Figure 6.9I) (Dernell et al. 1998a). The temporomandibular joint capsule is incised laterally and medially and then luxated (Figure 6.9J) (Dernell et al. 1998a). Finally, the tem­ poralis muscle is elevated from its insertion on the

coronoid process of the mandibular ramus (Dernell et al. 1998a). The defect is closed in three layers following total hemimandibulectomy and two layers following subtotal hemimandibulectomy (Dernell et al. 1998a). The deep layer is closed by suturing the pterygoid, masseter, and temporalis muscles. The submucosa is then closed in a simple continuous pattern. The mucosa and skin are closed with either a simple interrupted or simple continuous pattern using monofilament absorbable suture material. A simple interrupted pattern is recommended for wounds under tension, whereas a simple continuous pattern is sufficient for wounds with minimal or no tension (Figure 6.9K). The commissure of the lip can be advanced rostrally to the level of the first premolar or canine tooth to minimize hanging of the tongue on the resected side. A stented vertical mattress suture is recommended at the rostral extent of the lip advancement because of high tension at this point when the mouth is opened. Advancement of the commissure of the lip may improve postoperative cosmesis, but it may also increase the risk of wound complications. Caudal (vertical ramus) mandibulectomy Caudal mandibulectomy is indicated for benign or lowgrade malignant lesions confined to the mandibular ramus, such as osteoma or multilobular osteochondrosarcoma. More extensive hemimandibulectomy procedures are recommended for higher-grade malignant tumors. The temporomandibular joint can either be preserved or excised depending on the location of the tumor (Dernell et al. 1998a). Caudal mandibulectomy can also be combined with inferior orbitectomy for tumors with more extensive bony or soft tissue involvement. The dog is positioned in lateral recumbency for caudal mandibulectomy (Dernell et al. 1998a). A curved skin incision is performed over the ventral aspect of the zygomatic arch (Figure 6.10A). The temporalis muscle is elevated off the dorsal aspect of the zygomatic arch with periosteal elevators. The masseter muscle is then elevated off the medial aspect of the zygomatic arch (Figure 6.10B) (Dernell et al. 1998a). During this dissection, the infraorbital artery, vein, and nerve coursing along the medial aspect of the zygomatic arch should be identified and preserved. To expose the mandibular ramus, osteotomies of the zygomatic arch should be performed rostrally and caudally with an oscillating saw or Gigli wire (Figure 6.10C). Similar to the mandible, an osteotome and mallet should not be used for these osteotomies as the hard brittle bone of the zygomatic arch tends to shatter (Dernell et al. 1998a). Following removal of the zygomatic arch, the masseter muscle is elevated from

(a)

(b)

(c)

(d)

(f)

(g) (h)

(e)

(i)

130

(j)

(k)

Oral Tumors  131 Figure 6.9.  Subtotal and total hemimandibulectomy. (A) A skin incision may be required extending caudally from the commissure of the lips, and then the buccal and labial skin are dissected free from the masseter muscle (m) and mandible following an incision along the labial mucosa. (B) The labial and gingival mucosa are incised with minimum margins of 1 cm in a cat with a mandibular squamous cell carcinoma, and the mucosa is reflected with periosteal elevators to protect soft tissues from trauma during osteotomy of the mandible. (C) The rostral mandibular symphyseal separation can be performed with an oscillating saw, osteotome and mallet, or bone cutters (pictured). (D) The hemimandible is reflected laterally, and the inferior alveolar artery and vein are identified and ligated (arrow) caudal to their entry into the mandibular foramen. (E) For subtotal mandibulectomy, which is indicated for rostral malignant tumors invading into the medullary cavity or mandibular canal, the caudal osteotomy is positioned immediately rostral to the rostral attachment of the masseter muscle. (F, G) For total hemimandibulectomy, the dotted lines indicates where the masseter “F” and digastricus “G” muscles are incised from the caudal mandible. (H) The digastricus “D” muscle is reflected off the ventral aspects of the caudal mandible. (I) The pterygoid muscles “P” are reflected from the medial and ventral aspects of the caudal mandible. (J) The masseter muscle is reflected dorsally to expose and incise the temporomandibular joint (dotted line). (K) The resultant defect is closed in two to three layers. The first layer consists of suturing the master, digastricus, and pterygoid muscles. The submucosa is then closed with a simple continuous pattern, and then the sublingual mucosa is sutured to the labial mucosa using either simple interrupted or simple continuous sutures of monofilament absorbable suture material. (Line diagrams 6.9A, F, G, and J reproduced with permission from Withrow, S.J., and D.L. Holmberg. 1983. Mandibulectomy in the treatment of oral cancer. J Am Anim Hosp Assoc 19:277–278. Line diagram 6.9E reproduced with permission from Dernell, W.S., P.D. Schwartz and S.J. Withrow. 1998. Mandibulectomy. In Current Techniques in Small Animal Surgery, pp. 132–142. M.J. Bojrab, G.W. Ellison, and B. Slocum, editors. Baltimore: Williams & Wilkins.)

the ventrolateral surface and ventral margin of the ramus, and the temporalis muscle is elevated from its rostromedial insertion on the coronoid process of the mandible (Figure 6.10D) (Dernell et al. 1998a). For caudal mandibulectomies, which preserve the temporomandibular joint, the inferior alveolar artery and vein should be identified and preserved as it courses over the lateral surface of the medial pterygoid muscle before entering the mandibular foramen (Dernell et al. 1998a). The osteotomy of the mandibular ramus is positioned immediately dorsal to the temporomandibular joint in a coronal direction (Figures 6.10E, F). This osteotomy should be performed with an oscillating saw or pneumatic burr. When the temporomandibular joint is excised with the caudal mandibulectomy, the inferior alveolar artery and vein are ligated and transected as the medial pterygoid muscle is elevated from the ventromedial aspect of the mandibular angle (Dernell et al. 1998a). The osteotomy is positioned caudal to the first molar tooth for this more extensive caudal mandibulectomy (Figure 6.10E). This osteotomy should also be performed with an oscillating saw or pneumatic burr. Following caudal mandibulectomy, the defect is closed in three layers with apposition of the fascia of the temporalis and masseter muscles, then subcutaneous tissue, and finally skin (Dernell et al. 1998a). Surgical approach to tumors of the mandible in cats The surgical approach for management of benign and malignant tumors of the mandible is similar to that of

dogs. Mandibulectomy in cats is complicated by the small size of the mandible relative to the size of the oral tumor and the need for 1 cm margins for benign lesions and 3 cm margins for malignant tumors. As a result, less aggressive procedures such as unilateral rostral, segmental, and caudal mandibulectomy are rarely possible. Subtotal and total hemimandibulectomy are the most commonly indicated and performed procedures for the management of mandibular tumors in cats (Figure 6.9). The surgical technique is the same as for dogs; however, an esophageal or gastric feeding tube should be inserted as eating can be problematic following mandibulectomy in cats (Figure 6.11) (Northrup et al. 2006). Postoperative management Analgesia In the immediate postoperative period, intravenous fluids and analgesia are continued, and an Elizabethan collar should be placed as soon as the animal is sternally recumbent to prevent self-trauma (Dernell et al. 1998a). Analgesia should include a nonsteroidal antiinflammatory drug and an opioid. Cyclooxygenase-2 (COX-2) selective or specific nonsteroidal antiinflammatory drugs are preferred because of their safety index, efficacy, and possible anticancer effects (Umar et al. 2003). Nonsteroidal anti-inflammatory drugs should either not be administered or their dose decreased in animals with conditions such as renal failure, hypotension, or hepatic disease (Mathews 2000; Lascelles et al. 2007). Opioids, such as fentanyl or morphine, are preferably administered as a continuous rate infusion

132  Veterinary Surgical Oncology

(a)

(b)

(c)

(e)

(d)

(f)

Figure 6.10.  Vertical ramus or caudal mandibulectomy. (A) The dotted line represents the skin incision over the zygomatic arch. (B) The temporalis “T” and masseter “M” muscles are elevated subperiosteally from the zygomatic arch. (C, D) The rostral and caudal aspects of the zygomatic arch are then osteotomized (arrows) (C) to expose the vertical ramus of the mandible (D) following removal of the ostectomized segment of zygomatic arch. (E) The dotted lines represent the two options for osteotomy of the vertical ramus of the mandible. Depending on the location and type of the tumor, the vertical ramus mandibulectomy can preserve “a” or include “b” the temporomandibular joint. (F) In this dog with a multilobular osteochondrosarcoma, the tumor (arrow) has been excised with 2 cm margins (arrowheads) while preserving the temporomandibular joint. (Line diagrams A and E are reproduced with permission from Withrow, S.J., and D.L. Holmberg. 1983. Mandibulectomy in the treatment of oral cancer. J Am Anim Hosp Assoc 19:277–278.)

rather than intermittent intramuscular injections. Continuous rate infusions of opioids can be combined with ketamine and/or lidocaine for an enhanced analgesic effect. Animals can usually be weaned off continuous rate infusions over a 24- to 48-hour period. Cats and

dogs should be discharged with a nonsteroidal antiinflammatory drug and an oral opioid, such as codeine or tramadol. Nonsteroidal anti-inflammatory drugs should be used with care, particularly in cats (Lascelles et al. 2007). See Table 6.2 for suggested perioperative

Oral Tumors  133

Bradley et al. 1984; Withrow et al. 1985; Salisbury et al. 1986; Salisbury and Lantz 1988; Kosovsky et al. 1991; Schwarz et al. 1991a, 1991b; Hutson et al. 1992; Wallace et al. 1992; Lascelles et al. 2003). Enteral feeding tubes are not usually required following oral surgery in dogs, but they are recommended for cats treated with any type of mandibulectomy as eating can be difficult for 2–4 months following surgery (Hutson et al. 1992; Northrup et al. 2006). Cosmetic appearance

Figure 6.11.  It is important that a feeding tube be inserted following any mandibulectomy procedure in cats because voluntary intake is often poor for at least 2 weeks postoperatively.

analgesic protocols for cats and dogs undergoing mandibular resections. Nutrition Intravenous fluids should be continued until the dog eats voluntarily and is drinking sufficient quantities to maintain hydration. This is rarely a problem, and most dogs can be discharged within 24–48 hours. To prevent disruption of intraoral incisions, dogs should only be fed soft canned food and prevented from chewing on hard objects for 4 weeks (Dernell et al. 1998a). Supplemental nutrition is rarely required in dogs but is important in cats. Following various mandibulectomy procedures in 42 cats, 73% of cats were either dysphagic or inappetent in the immediate postoperative period, and five of these cats never ate voluntarily following surgery (Northrup et al. 2006). An esophageal or gastric feeding tube is strongly recommended in cats because of the high risk of inappetence (Figure 6.11). Supplemental tube feeding is well tolerated in cats, and complications associated with feeding tubes are infrequent (Marks 1998). The feeding tube should be removed when the cat begins to eat voluntarily and consistently. Complications Blood loss and hypotension are the most common intraoperative complications (Wallace et al. 1992; Lascelles et al. 2003). Postoperative complications are uncommon and include incisional dehiscence, epistaxis, increased salivation, mandibular drift and malocclusion, and difficulty prehending food (Withrow and Holmberg 1983;

The cosmetic appearance of cats and dogs following mandibulectomy is usually good to excellent. As recommended previously, owner acceptance of postoperative appearance and function is improved with a thorough discussion, including the use of pre- and postoperative images of the appropriate procedure, before surgery. Owner satisfaction with the cosmetic appearance and functional outcome following mandibulectomy is high, with 83% and 85% of owners satisfied following mandibulectomy in cats and dogs, respectively (Fox et al. 1997; Northrup et al. 2006). The cosmetic appearance of dogs is not altered after unilateral rostral, segmental, and caudal mandibulectomy (Figure 6.12A–D). These procedures are rarely performed in cats, although cosmetic appearance is also unchanged following caudal mandibulectomy. Bilateral rostral mandibulectomy is the most cosmetically challenging of the mandibulectomy procedures in both cats and dogs because of mandibular shorten­ ing, excessive drooling and cheilitis, and the tongue hanging out, especially when panting or excited (Figure 6.13). Subtotal and total hemimandibulectomy results in a mild concavity on the resected side, which is rarely appreciable, mandibular drift, and the tongue hanging out on the resected side (Figure 6.14) (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; Schwarz et al. 1991a; White 1991; Dernell et al. 1998a; Northrup et al. 2006). Eating difficulties Eating difficulties are reported in 44% dogs and 73% cats following mandibulectomy (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). In dogs, this is often related to difficulty in prehending food. Prehension difficulties are more common following bilateral rostral mandibulectomy with resection of both canine teeth, particularly the more aggressive bilateral procedures extending caudally to the second premolar teeth (Schwarz et al. 1991a). The majority of dogs with prehension

134  Veterinary Surgical Oncology

(a)

(b)

(c)

(d)

Figure 6.12.  (A, B) The typical postoperative appearance of a dog following unilateral rostral mandibulectomy. Note the saliva accumulation rostrally and minimal change in cosmesis. (C,D) The typical postoperative appearance of a dog following vertical ramus (or caudal) mandibulectomy. Note that the cosmetic appearance of the dog is unaltered.

difficulties adapt within 2 weeks. Supplemental feeding may be required during this period, such as force feeding, tube feeding, or feeding soft foods made into a ball. If prehension difficulties continue beyond 2 weeks, then other causes should be investigated. Injury to the hypoglossal nerve and mandibular drift can also occasionally result in difficulties in prehending and eating food (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). In cats, eating difficulties are common regardless of the mandibulectomy procedure. In the short term (≤4 weeks), inappetence is reported in 83% of cats following bilateral rostral mandibulectomy, 74% of cats following hemimandibulectomy, and 83% of cats following resection of more than 50% of the mandible (Northrup et al. 2006). Inappetence is also common in the long term, with 10% of cats with bilateral rostral mandibulectomy, 53% of cats with hemimandibulectomy, and 83% of cats with resection of greater than 50% of the mandible

experiencing eating difficulties (Northrup et al. 2006). A feeding tube should be inserted at the time of mandibulectomy in cats because of this high risk of inappetence. An esophageal or gastric feeding tube is preferred because of the ability to maintain and use these feeding tubes for a prolonged period. The feeding tube should be removed when the cat begins to eat voluntarily and consistently. Incisional swelling Swelling of the surgical site is common following nonrostral mandibulectomies (i.e., segmental and caudal mandibulectomy, and subtotal and total hemimandibulectomy) (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). Swelling resolves spontaneously in 5–7 days. Ice packing every 4 hours and the administration of nonsteroidal anti-inflammatory drugs may assist in decreasing the severity of postoperative swelling.

Oral Tumors  135

Figure 6.13.  The typical postoperative appearance of a dog following bilateral rostral mandibulectomy. Note the shortened mandible and the tongue hanging out.

Figure 6.15.  A ranula-like lesion (arrow) in a dog 1 day after subtotal hemimandibulectomy for an osteosarcoma. These may represent either a hematoma or accumulation of saliva. Treat­ ment is rarely required because these lesions often resolve spontaneously.

caused by trauma to the mandibular and sublingual salivary ducts, but hematoma or seroma formation are more likely. Ranula-like lesions will usually resolve spontaneously and treatment is rarely required. Wound dehiscence

Figure 6.14.  The typical postoperative appearance of a dog following subtotal (pictured) or total hemimandibulectomy. The mandible drifts toward the midline and the tongue hangs out on the resected side.

Ranula-like lesions Ranula-like lesions are uncommon and appear as soft, fluctuant, nonpainful swellings in the frenulum of the tongue ipsilateral to the mandibulectomy procedure (Figure 6.15) (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a). They are most frequently observed following either subtotal or total hemimandibulectomy. Ranula-like lesions may be

Wound dehiscence is reported in 13% of cats and 8%– 33% of dogs following mandibulectomy (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). Wound dehiscence most commonly occurs 3–7 days after surgery. The two most common sites for wound dehiscence are over the rostral end of the osteotomized mandible and at the commissure of the lips. Tension at these sites is the most likely cause of dehiscence, although the use of cautery, rapidly absorbing suture material (i.e., catgut or poliglecaprone 25), and poor wound-healing capabilities as a result of radiation therapy, chemotherapy, or debilitation may also contribute to wound dehiscence (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). Wound dehiscence can be managed with either second-intention healing, if the dehisced area is small and granulating, or debridement and resuturing, if the defect is large. Dehiscence of the commissure of the lip usually requires surgical revision to improve the cosmetic appearance and the use of tension-relieving sutures, such as a vertical mattress pattern (Dernell et al. 1998a). If the tongue is the cause of tension and dehiscence, then tube feeding may be required for 1–2 weeks.

136  Veterinary Surgical Oncology

Excessive drooling Excessive drooling is common, particularly following bilateral rostral mandibulectomy in cats and dogs and resection of more than 50% of the mandible in cats (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). Ptyalism will either resolve spontaneously or significantly reduce in volume after several weeks in the majority of animals. However, for cats and dogs with persistent drooling, cheilitis and facial dermatitis are common sequalae. The management options for these complications include daily washing with an antiseptic solution and surgical cheiloplasty. Mandibular drift and malocclusion Mandibular drift is common following mandibulectomy, especially the more aggressive mandibulectomy techniques, and is characterized by the mandible drifting toward the contralateral side (Figure 6.14) (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). Mandibular drift is caused by a loss of mandibular support at either the mandibular symphysis or temporomandibular joint. Mandibular drift results in malocclusion and can predispose to osteoarthritis of the temporomandibular joints. Drift of the lower canine tooth toward the midline can cause ulceration and trauma to the overlying hard palate. This is rarely a problem in dogs, but has been reported in up to 18% of cats, especially following segmental mandibulectomy, hemimandibulectomy, and resection of more than 50% of the mandible (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). Mandibular drift does not require treatment unless the drift results in complications. For animals with secondary hard palate trauma, the lower canine tooth should either be extracted or shortened with vital pulpotomy (Dernell et al. 1998a). Reconstruction of the mandible with either ulnar or rib autografts and promotion of osseous ingrowth with bone morphogenetic protein-2 have been described to prevent or treat mandibular drift (Boudrieau et al. 1994, 2004; Bracker and Trout 2000; Spector et al. 2007), but this is rarely required and is associated with increased expense and risk of complications. Miscellaneous complications Other complications reported following mandibulectomy include hemorrhage, infection, pain, difficulty

grooming, and osteoarthritis of the temporoman­ dibular joint (Withrow and Holmberg 1983; Bradley et al. 1984; Salisbury and Lantz 1988; Kosovsky et al. 1991; White 1991; Dernell et al. 1998a; Northrup et al. 2006). Intraoperative hemorrhage can be profuse if the inferior alveolar artery is not ligated prior to subtotal or total hemimandibulectomy. To avoid this complication, the inferior alveolar artery should be identified on the caudomedial aspect of the mandible as it courses over the temporomandibular joint and pterygoid muscles before entering the mandibular foramen. Hemorrhage can be controlled with ligation or cautery of the inferior alveolar artery and vein. Alternatively, products that either chemically or physically promote hemostasis, such as Gelfoam or bone wax, can be used if the inferior alveolar cannot be ligated. Hemorrhage from the inferior alveolar artery is rarely severe enough to warrant a whole blood transfusion, but blood loss should be carefully monitored and a crossmatched compatible blood transfusion considered if hematocrit acutely decreases below 15%–30%, particularly if this occurs in combination with hypotension, hypoxia, and/or clinical, biochemical, or echocardiographic evidence of anaerobic metabolism (Jutkowitz 2004). Infection is very rare following mandibulectomy because of the rich vascular supply to the oral cavity (Dernell et al. 1998a). Incisional abscesses are treated with debridement, copious lavage with an isotonic crystalloid solution, closure of dead space, drainage if possible, and culture-directed antibiotics. Osteoarthritis of the temporomandibular joint is a relatively common radiographic finding, particularly in any animal with mandibular drift, but it rarely manifests as a clinical problem (Dernell et al. 1998a). If degenerative joint disease of the temporomandibular joint causes pain on opening of the jaw or eating difficulties, then nonsteroidal anti-inflammatory drugs and chondroprotective agents should be administered. Grooming difficulties are a specific complication in cats, especially following hemimandibulectomy where 26% of cats have been reported to have long-term grooming problems (Northrup et al. 2006). The cause of these grooming difficulties may be related to ptyalism, tongue protrusion, mandibular drift, and possibly pain. They are difficult to manage, and in these cases, owners must groom their cats. In one retrospective review of mandibulectomy in cats, grooming difficulties were infrequent (18%) but had a major impact on the perceived quality of life for affected cats (Northrup et al. 2006).

Oral Tumors  137

Surgical Approach to Tumors of the Maxilla Anatomy For successful maxillofacial surgery, a detailed knowledge of the anatomy is required. This includes not only a fundamental knowledge of the bones comprising the maxilla and face (frontal, nasal, maxillary, incisive, and pterygoid) (Figures 6.16A, B) but also a detailed appreciation of the three-dimensional relationship between the bones and various anatomic features of the skull. The best way to appreciate this is to have appropriate skulls (brachycephalic, mesocephalic, and dolichocephalic dog skulls and cat skull) for review prior to and during surgery. During maxillary surgery, the close proximity of the nasal cavity and cranial cavity needs to be appreciated. A knowledge of the vascular anatomy of the maxilla and orbit is necessary because of the risk of hemorrhage and hypotension during maxillectomy procedures. This includes the major palatine and sphenopalatine arteries for rostral maxillectomies and the infraorbital artery and maxillary artery for caudal maxillectomies (Figures 6.16C, D) (Evans and Christensen 1979). Maxillectomy refers to the en bloc excision of a tumor on the upper jaw, which may involve parts the incisive, palatine, lacrimal, zygomatic, frontal, and vomer bones in addition to the maxilla (Evans and Christensen 1979). The resultant defects are closed using soft tissue flaps, particularly vestibular (e.g., alveolar and buccal) mucosal-submucosal flaps with or without palatal mucoperiosteal flaps. Various maxillectomy procedures have been described, including incisivectomy (previously known as premaxillectomy), unilateral and bilateral rostral maxillectomy, central maxillectomy, and caudal maxillectomy. Bilateral rostral maxillectomy can be combined with resection of the nasal planum if necessary or continue further caudally for a radical maxillectomy. Caudal maxillectomy can be combined with various orbitectomy procedures if necessary and can be performed either through an intraoral approach or combined intraoral-dorsolateral skin incision approach depending on the tumor location. Incisivectomy Premaxillectomy is often used in the veterinary literature to describe excisions confined to the incisive bone (Withrow et al. 1985). Premaxilla is not accepted veterinary anatomical nomenclature, and incisivectomy is therefore more appropriate to describe resection of the area rostral to the canines. Incisivectomy is recommended for dogs with peripheral odontogenic fibroma or small SCC lesions confined

to the incisive bone and associated incisors (Figure 6.17A). More aggressive rostral maxillectomy procedures should be considered for malignant tumors other than SCC and larger acanthomatous ameloblastoma and SCC lesions extending further caudally. For incisivectomy, the dog is positioned in dorsal recumbency. The labial mucosa is incised with a minimum of 1 cm margins around the mass. The labial mucosa is reflected off the incisive bone with periosteal elevators to preserve the soft tissue of the lip, which will be used later for reconstruction of the defect. The rostral hard palatine mucosa is also incised 1 cm caudal to the mass, but rostral to the canines (Withrow et al. 1985). Larger margins are used if the planned margins will damage existing tooth roots. An osteotomy is performed caudal to the mass along the hard palate mucosal incision. The extent of the bone resection will depend on whether or not, or how much, bone is involved in the neoplastic process. The nasal cavity is not usually entered in this procedure, but the ventrolateral nasal cartilages are exposed (Withrow et al. 1985). Bleeding is generally minor and is controlled by a combination of cautery, ligation, and direct pressure. The defect is closed with a labial mucosal-submucosal flap. The flap is created by undermining the labial mucosa to include the mucosa, submucosa, and as much subcutaneous tissue as possible, using Metzenbaum scissors (Withrow et al. 1985). Vertical releasing incisions are used if necessary. The flap is sutured into position with a two-layer closure, with the first layer consisting of simple interrupted sutures preplaced through holes predrilled in the bone of the hard palate. The labial and oral mucosa are opposed using simple interrupted sutures (Figure 6.17B) (Withrow et al. 1985). Rostral maxillectomy—unilateral Unilateral rostral maxillectomy is recommended for dogs with benign acanthomatous ameloblastoma or SCC lesions located around the canine tooth and do not extend caudal to the second premolar tooth (Dernell et al. 1998b). Bilateral rostral maxillectomy should be considered for these tumor types that cross the midline, and hemimaxillectomy is recommended for unilateral lesions extending caudal to the second premolar tooth. Dogs are positioned in either dorsal or dorsolateral recumbency because the majority of tumors can be resected via an intraoral approach. At this level, the sphenopalatine, major palatine, and infraorbital vessels are larger than at the incisive bone, and bleeding can be more substantial. The surgical technique for unilateral rostral maxillectomy is similar to incisivectomy. The

138  Veterinary Surgical Oncology

(a)

(b)

(c)

(d)

Figure 6.16.  Anatomy of the maxilla and skull. (A, B) A number of maxillectomy procedures have been described and these involve removal of part or all of the maxilla, but may also involve en bloc removal of the incisive, nasal, zygomatic, frontal, and palatine bone. (C, D) A knowledge of the vascular anatomy of the maxilla and orbit is essential, particularly for caudal maxillectomies, because of the potential for intraoperative hemorrhage. The major palatine and sphenopalatine arteries should be identified for rostral maxillectomies and the infraorbital and maxillary arteries for caudal maxillectomies. (Diagrams A and B reproduced with permission from Evans, H.E. and A. deLahunta, editors. 2000. The head. In Guide to the Dissection of the Dog, pp. 259–321. Philadelphia: Saunders. Diagrams C and D reproduced with permission from Evans, H.E. and G.C. Christensen, editors. 1979. Systemic arteries. In Miller’s Anatomy of the Dog, pp. 652–756. Philadelphia: Saunders.)

labial and gingival mucosa and palatine mucoperiosteum are incised with a minimum of 1 cm margins around the mass (Figure 6.18A) (Dernell et al. 1998b). Bleeding from the transected major palatine artery can be brisk following the palatine incision, but this can

usually be controlled with digital pressure, although occasionally ligation or cautery may be required (Dernell et al. 1998b). The mucosa is then reflected off the underlying bone with periosteal elevators to preserve the soft tissues that will be used later for reconstruction of the

Oral Tumors  139

(a)

(b)

Figure 6.17.  (A) A benign acanthomatous ameloblastoma arising from the periodontal ligament of the mandibular incisors. Note that this is a mandibular incisivectomy. (B) The incisivectomy has been performed with an oscillating saw. The labial and oral mucosa are closed in two layers.

defect. Osteotomies are performed in the maxilla, incisive bone, and hard palate with either a pneumatic burr, small oscillating saw, biradial saw (Figure 6.18B), or osteotome and mallet (Dernell et al. 1998b). The caudal bone cuts are performed last so bleeding can be quickly controlled after the tumor and bone segment are removed. The nasal cavity is exposed during unilateral rostral maxillectomy (Figure 6.18C). Bleeding can be significant from the turbinates, and a combination of tamponade and ice-cold saline or 0.05% oxymetazoline spray can be used for hemostasis. Cautery is rarely successful for turbinate bleeding. The defect is closed with a labial mucosal-submucosal flap as described for incisivectomy. The flap is created by undermining the labial mucosa to include the mucosa, submucosa, and as much subcutaneous tissue as possible (Figure 6.18D) (Dernell et al. 1998b). It is important to undermine sufficient tissue to prevent the overlying skin from being drawn medially, resulting in a poor cosmetic result. The flap is sutured into position with a two-layer closure, with the first layer consisting of simple interrupted sutures preplaced through holes predrilled in the bone of the hard palate (Figure 6.18E, F). The labial and oral mucosa are opposed using simple interrupted sutures (Figure 6.18G) (Dernell et al. 1998b). Rostral maxillectomy—bilateral Bilateral rostral maxillectomy is recommended for dogs with benign acanthomatous ameloblastoma or SCC lesions located rostral to the second premolar teeth and crossing the midline (Figure 6.19A) (Dernell et al.

1998b). Bilateral rostral maxillectomy can be combined with resection of the nasal planum for tumors involving both the nasal planum and rostral maxilla (Kirpensteijn et al. 1994). Radical maxillectomy should be considered for SCC lesions extending caudal to the second premolar teeth, malignant tumors other than SCC located rostral to the second premolar teeth and crossing the midline of the palate, and any tumor type invading into the nasal cavity and cartilages (Lascelles et al. 2004). Dogs are positioned in dorsal recumbency for bilateral rostral maxillectomy because resection and reconstruction are performed through an intraoral approach. The surgical technique for bilateral rostral maxillectomy is similar to unilateral rostral maxillectomy. The labial and gingival mucosa and palatine mucoperiosteum are incised with a minimum of 1 cm margins around the mass. Bleeding from the transected major palatine artery can be brisk following the palatine incision, but this can usually be controlled with digital pressure although occasionally ligation or cautery may be required (Dernell et al. 1998b). The mucosa is then reflected off the underlying bone with periosteal elevators to preserve the soft tissues that will be used later for reconstruction of the defect (Figure 6.19B). Osteotomies are performed in the maxilla and hard palate with an oscillating saw, but a pneumatic burr or osteotome and mallet can also be effective. The nasal cavity is exposed during bilateral rostral maxillectomy (Figure 6.19C). Bleeding can be significant from the turbinates, and a combination of tamponade and ice-cold saline or 0.05% oxymetazoline spray can be used for hemostasis. Cautery is rarely successful for turbinate bleeding. To minimize drooping of

(a)

(b)

(c)

(d)

(e)

(f)

(g)

Figure 6.18.  Unilateral rostral maxillectomy. (A) The labial and gingival mucosa and palatine mucoperiosteum are incised with minimum margins of 1 cm (dotted lines). Following these incisions, bleeding from the transected major palatine artery may be brisk and should be controlled with digital pressure. (B) A 30 mm biradial saw is being used to perform a unilateral rostral maxillectomy on a dog with an acanthomatous ameloblastoma. These osteotomies can also be performed with a pneumatic burr, oscillating saw, or an osteotome and mallet. (C) Following elevation of the incised mucosa “M” and lateral “LO,” rostral “RO,” caudal “C,” and medial “MO” osteotomies, the resected segment of maxilla is removed and the nasal cavity “NC” is exposed. (D) The submucosa-mucosa of the adjacent lip is undermined to create a labial mucosal flap for reconstruction of the intraoral defect and also minimize the lip being drawn medially, resulting in a poor cosmetic result. (E) The resultant defect is closed in two layers. Holes are predrilled into the bone of the hard palate. (F) The deep layer consists of simple interrupted sutures through these predrilled holes in the bone of the hard palate and labial submucosa. (G) Then the labial and oral mucosa are opposed using simple interrupted or simple continuous sutures of monofilament absorbable suture material. (Diagram A reproduced with permission from Dernell, W.S., P.D. Schwartz, and S.J. Withrow. 1998. Maxillectomy and premaxillectomy. In Current Techniques in Small Animal Surgery, pp. 124–132. M.J. Bojrab, G.W. Ellison, and B. Slocum, editors. Baltimore: Williams & Wilkins.)

140

Oral Tumors  141

(a)

(b)

(c)

(d)

Figure 6.19.  Bilateral rostral maxillectomy. (A) Bilateral rostral maxillectomy is indicated for dogs with benign invasive tumors, such as this dog with an acanthomatous ameloblastoma, or small SCC lesions rostral to the second premolar tooth and crossing the midline. (B) Following incisions in the labial and gingival mucosa and palatine mucoperiosteum, the mucosa and mucoperiosteum are then reflected with periosteal elevators to protect soft tissues from trauma during maxillectomy. (C) The bilateral rostral maxillectomy is performed with an oscillating saw using minimum caudal margins of 1–2 cm for benign tumors, such as this acanthomatous ameloblastoma, and 2–3 cm for malignant tumors. Bleeding can be brisk from the nasal cavity (arrow) and transection of the major palatine arteries (arrowhead). (D) The resultant defect is closed in two layers. The deep layer consists of simple interrupted sutures through predrilled holes in the bone of the hard palate and labial submucosa, and then the labial and mucoperiosteum of the hard palate are opposed, often in a T-shape, using simple interrupted or simple continuous sutures of monofilament absorbable suture material.

the nose, which is a common postoperative cosmetic defect following bilateral rostral maxillectomy because of loss of ventral support, the nasal bone should be preserved at the point where the nasal cartilages attach. Alternatively, a cantilever suture technique has been described in which a buried mattress suture is used to elevate the nasal cartilage (Figure 6.20) (Pavletic 1999).

The defect is closed with bilateral labial mucosalsubmucosal flaps. The flaps are created by undermining the left- and right-sided rostral labial mucosa to include the mucosa, submucosa, and as much subcutaneous tissue as possible. The flaps are sutured in a T-shape with a two-layer closure, with the first layer consisting of simple interrupted sutures preplaced through holes predrilled in the bone of the hard palate. The labial and oral

142  Veterinary Surgical Oncology

(a)

(b)

(c)

(d)

Figure 6.20.  (A) For the cantilever suture technique, a 3–5 cm skin incision is performed along the dorsal midline of the maxilla rostral to the medial canthus of each eye. The skin is undermined and retracted to expose the nasal and maxillary bones. A hole is drilled transversely across the maxilla immediately ventral to the nasal bone with a small Steinmann pin. (B) A large gauge (1 to 1-0) monofilament suture material is passed through the hole using either a straight swaged needle or a hypodermic needle as a guide. (C) A 1 cm incision is made through the epithelial surface of each lateral cartilage. The swaged needle is passed rostrally deep to the skin, exits through the lateral cartilage incision on one side of the nasal planum. It is redirected perpendicularly and passed transversely through the ipsilateral lateral cartilage incision across the nasal planum, exits through the contralateral lateral cartilage incision, and then is redirected perpendicularly through the second lateral cartilage incision and caudally deep to the skin. It exits at the dorsal muzzle incision. (D) The cantilever suture is then tightened resulting in elevation of the rostral nose. (Illustrations courtesy of Dave Carlson)

mucosa are opposed using simple interrupted sutures (Figure 6.19D) (Dernell et al. 1998b). Rostral maxillectomy—bilateral combined with nasal planum resection Bilateral rostral maxillectomy combined with nasal planum resection is indicated for SCC of the nasal planum

that has invaded the incisive area or rostral maxilla (Figure 6.21A) (Kirpensteijn et al. 1994). Dogs are positioned in sternal recumbency with the mouth held open with a gag. The head should be elevated to facilitate access to the oral cavity for resection and reconstruction. The pharynx should be packed with gauze sponges to minimize the risk of aspirating blood and lavage fluid.

(a)

(c)

(e)

(b)

(d)

(f)

Figure 6.21.  (A) Bilateral rostral maxillectomy combined with nasal planum resection is indicated in this dog with a nasal planum SCC invading into the incisive bone. (B) The labial and gingival mucosal and palatine mucoperiosteal incisions are continued along the same sagittal plane through the right and left rostral lips and dorsal to the nasal planum in a cat with an invasive nasal planum SCC. (C) The nasal planum is resected en bloc with either the incisive bone, as depicted in this intraoperative image, or rostral maxilla. (D) The resultant defect with free lip edges following en bloc removal of the resected nasal planum and incisive bone. (E, F) For closure of the defect, the labial submucosa is sutured to predrilled bone tunnels in the palatine bone, and the labial mucosa is sutured to the mucoperiosteum of the hard palate using absorbable monofilament suture material in either a simple interrupted or continuous suture pattern. The free edges of the lip are sutured in the rostral midline with a standard two-layer closure.

143

144  Veterinary Surgical Oncology Figure 6.22.  Radical maxillectomy. (A, B) The labial and gingival mucosal and palatine mucoperiosteal incisions are continued along the same sagittal plane through the right and left rostral lips and skin along the dorsal maxilla in a dog with a bilateral rostral oral fibrosarcoma (A, dorsal view) and a dog with an invasive nasal planum squamous cell carcinoma (arrow) (B, lateral view). (C) An osteotomy is performed with an oscillating saw from a dorsal approach perpendicular to the maxilla with a minimum of 2–3 cm margins caudal to the tumor. (D) A portion of the labial mucosa is removed (X) to establish a neolabiopalatine margin of 1 cm wide (A′ to a′ when sutured in place). A′ to a′ is the distance the new lip will hang over the palatine mucosa at the rostral end of the maxilla, and this creates the new palatobuccal recess rostrally. The flap will be transposed and sutured (A′ to A, a′ to a, b′ to b, and c′ to c). (E) Bone tunnels are drilled into the rostral bone of the hard palate so that the deep layer of the intraoral closure can be performed by suturing the labial submucosa to these bone tunnels (arrow). The intraoral closure is completed by suturing the labial mucosa to the mucoperiosteum of the hard palate in either a simple interrupted or simple continuous pattern using absorbable monofilament suture material. (F) Figure-eight suture pattern. The lip margins from the left and right sides are aligned with a figure-eight suture pattern (F, part A), and a horizontal mattress suture pattern is used to accurately align the lip margin with the knot tied away from the lip margin. The remainder of the skin is closed with simple interrupted sutures (F, part B). (G) Rolling figure-eight sutures are used to roll the skin around the edge of the maxillary bone to hasten mucocutaneous healing and to cover the exposed edges of the maxilla. (H) To reconstruct the nasal orifice, the skin edges are sutured to bone tunnels in the maxillary bone using either a rolling figure-eight (see G) or simple interrupted pattern, and the lip is reconstructed along the rostral aspect of the maxilla (see D). (Some images (C–E) and line diagrams (F and G) are reproduced with permission from Lascelles, B.D.X., R.A. Henderson, B. Seguin, B., et al. 2004. Bilateral rostral maxillectomy and nasal planectomy for large rostral maxillofacial neoplasms in six dogs and one cat. J Am Anim Hosp Assoc 40:137–146.)

The nasal planum, labial and gingival mucosa, and palatine mucooperiosteum are incised a minimum of 1 cm caudal to the extent of the tumor (Kirpensteijn et al. 1994). The nasal planum excision usually involves full-thickness incisions through the rostral lips (Figure 6.21B). The mucosal incisions extend transversely across either the incisive region or rostral maxilla to join the full-thickness incisions in the left and right rostral lips. The osteotomy is performed from a dorsal approach and preferably with an oscillating saw (Figure 6.21C). The resultant defect is reconstructed by drawing the free edges of the lip to the rostral maxilla so that the left and right lips meet in the midline of the rostral extent of the excised palate, thus recreating a continuous rostral lip (Figure 6.21D). The labial submucosa is sutured to bone tunnels drilled in the hard palate and the labial mucosa to the palatine mucosa. Depending on the level of resection, reconstruction of the new nasal orifice is achieved by suturing the skin to the edge of the nasal cartilages or the nasal bone. Simple continuous purse strings have been described for closure of the nasal opening to an appropriate size (Kirpensteijn et al. 1994), but this approach can result in stenosis of the nasal aperture and respiratory complications. If nasal planum resection has been combined with an incisivectomy, then the skin is sutured to the nasal cartilages in a single layer of simple interrupted sutures. For the combination of nasal planum resection with a bilateral rostral maxillectomy, a two-layer closure is preferred with subcuta­ neous tissue sutured to bone tunnels drilled in the maxillary bone and skin to the nasal mucosa using a

simple interrupted suture pattern (Figures 6.21E, F) (Kirpensteijn et al. 1994). Radical maxillectomy Radical bilateral maxillectomy is recommended for tumors of the rostral maxilla extending dorsally into the nasal cavity and malignant tumors caudal to the second premolar teeth and extending across the midline (Lascelles et al. 2004). The surgical approach is similar to nasal planum resection combined with bilateral maxillectomy. Dogs are positioned in sternal recumbency with the mouth held open with a gag. The head should be elevated to facilitate access to the oral cavity for resection and reconstruction. The pharynx should be packed with gauze sponges to minimize the risk of aspirating blood and lavage fluid. The first drape is placed in the mouth over the mandible, tongue, and endotracheal tube, but not pressed tightly against the commissures of the lip because these should be mobile to permit labial advancement for reconstruction of the nasal defect (Lascelles et al. 2004). Additional drapes are placed around the maxilla with the eyelids exposed for orientation. Skin, labial and gingival mucosa, and palatine mucoperiosteum are incised a minimum of 2 cm, and preferably 3 cm, caudal to the extent of the tumor as determined by advanced imaging and intraoperative palpation. The skin incision involves full-thickness incisions through the lips perpendicular to the labial margin and extending dorsally and transversely across the maxilla (Figure 6.22A, B) (Lascelles et al. 2004). As the skin incision is continued deeply

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

145

146  Veterinary Surgical Oncology

through the subcutis tissue and nasolabial muscles, the infraorbital neurovascular bundle should be ligated and transected. The mucosal incisions extend transversely across the alveolar margins and palate to join the fullthickness incisions in the left and right lips. The osteotomy is performed from a dorsal approach with an oscillating saw perpendicular to the maxilla. To facilitate reconstruction, the osteotomy should be performed slightly caudal to the level of the skin and mucosal incisions (Figure 6.22C). The resultant defect is reconstructed by drawing the free edges of the lip to the rostral maxilla so that the left and right lips meet in the midline of the rostral extent of the excised palate (Lascelles et al. 2004). This creates a continuous rostral lip and new nasal orifice and also divides the nasal and oral cavities. Recreation of the rostral lip requires either a unilateral or bilateral labial flap (Lascelles et al. 2004). The labiogingival reflection on each side is incised as necessary to mobilize the labial flap. The mucosa of the labial flaps is removed except for a 0.5–1.0 cm width adjacent to the labial margin (Figure 6.22D). This distance is determined by apposing the labial tissues to identify the contact point of the palatine mucosa and labium, and then assessing how much of the reconstructed lip will project ventrally from the palatine mucosa (Lascelles et al. 2004). The lip can interfere with food transfer into the oral cavity if this margin is excessive. Once the mucosa is excised, the labial submucosa is sutured to bone tunnels drilled in the hard palate and the labial mucosa to the palatine mucoperiosteum (Figure 6.22E) (Lascelles et al. 2004). Next, the labial skin margins are sutured together using a figure-eight suture pattern (Figure 6.22F). To reconstruct the nasal orifice, the skin edges are sutured to bone tunnels drilled in the maxillary bone using either a rolling figure-eight or simple interrupted suture pattern (Figure 6.22G). Suture tightening results in the skin covering the rostral edge of the maxilla (Figure 6.22H) (Lascelles et al. 2004). Caudal maxillectomy—intraoral approach Caudal maxillectomy using an intraoral approach is recommended for unilateral benign and malignant tumors located along the alveolar margins of the mid-to-caudal maxilla (Figure 6.23A) (Dernell et al. 1998b). A combined intraoral and dorsolateral approach is preferred for more extensive tumors, particularly those with involvement of the lateral maxilla and inferior orbit, because of better exposure, ability to achieve hemostasis, and superior ability to achieve surgical margins (Dernell et al. 1998b; Lascelles et al. 2003). Advanced imaging is recommended for tumors of the caudal maxilla and orbit to determine the extent and resectability of the

tumor and to plan the surgical approach. In some cases, particularly those with more extensive involvement of the orbit, enucleation may be required to improve exposure and likelihood of achieving complete excision of the tumor. The possibility of enucleation and adjuvant radiation therapy if the tumor is incompletely excised should be discussed with the owner prior to surgery. Dogs are positioned in lateral or dorsolateral recumbency with the mouth held open with a gag. The pharynx should be packed with gauze sponges to minimize the risk of aspirating blood and lavage fluid. The gingival mucosa and mucoperiosteum of the hard palate are incised with appropriate margins around the mass, depending on whether the tumor is benign or malignant (Figure 6.23B). Bleeding can be brisk following the palatine incision due to transection of the major palatine artery or its branches, but this can usually be controlled with digital pressure, cautery, or ligation (Dernell et al. 1998b). The mucosa and mucoperiosteum are then reflected off the underlying maxillary and palatine bone with periosteal elevators to preserve the soft tissues that will be used later for reconstruction of the defect (Figure 6.23C). Osteotomies are performed in the maxilla and hard palate with either a pneumatic burr, small oscillating saw, or osteotome and mallet. The rostral and lateral osteotomies are performed initially, followed by the palatine and caudal osteotomies (Dernell et al. 1998b). The caudal osteotomy can be difficult to complete because of poor access and visibility. This is preferably performed with an osteotome and mallet with the osteotome started at the junction between the inferior orbit and caudal maxilla, caudal to the molar teeth, and directed rostrally to connect the lateral maxilla and palatine osteotomies. The palatine and caudal osteotomies are performed last because of the greater potential for hemorrhage from the nasal turbinates and maxillary artery, respectively (Dernell et al. 1998b). If possible, the maxillary artery should be ligated prior to resection, although this can be difficult from an intraoral approach because of poor exposure and visibility. Following completion of the osteotomies, the free bone segment should be gently elevated and removed (Figure 6.23D). Bleeding from the surgical site is then controlled with either suture or metallic clip ligation or cautery for vessels and a combination of tamponade and ice-cold saline or 0.05% oxymetazoline spray for bleeding from the turbinates. Cautery is rarely successful for turbinate bleeding. The defect is closed with a labial mucosal-submucosal flap as described for other maxillectomy procedures (Dernell et al. 1998b). The flap is created by undermining the labial mucosa to include the mucosa and submucosa (Figure 6.23E). It is important to undermine

(a)

(b)

(c)

(e)

(d)

(f)

(g)

Figure 6.23.  (A) Caudal maxillectomy through an intraoral approach is recommended for benign and malignant tumors involving alveolar margins of the mid-to-caudal maxilla and not crossing the midline of the hard palate, such as this oral fibrosarcoma in a dog. (B) The labial and gingival mucosa and palatine mucoperiosteum are incised with appropriate margins around the tumor (dotted lines), and a gauze sponge “A” is been placed into the caudal oropharynx to prevent passive aspiration of blood and lavage fluid intraoperatively. (C) Following incisions using appropriate margins in the gingival and labial mucosa and palatine mucoperiosteum, the mucosa and mucoperiosteum are reflected off their underlying bone to protect the soft tissues from trauma during osteotomies and to preserve these soft tissues for reconstruction of the defect. (D) Rostral, lateral and caudal maxillary and medial palatine osteotomies have been performed, and the bone segment is then gently elevated to minimize damage to the underlying nasal turbinates. (E) The submucosa-mucosa of the adjacent lip is undermined to create a labial mucosal flap for reconstruction of the intraoral defect and also minimize the lip being drawn medially, resulting in a poor cosmetic result. (F, G) The resultant defect (F) with exposure of the nasal cavity “NC” is closed in two layers. The deep layer consists of simple interrupted sutures through predrilled holes in the bone of the hard palate and labial submucosa, and then the labial and oral mucosa are opposed using simple interrupted or simple continuous sutures of monofilament absorbable suture material (G). (Line diagrams B and E reproduced with permission from Dernell, W.S., P.D. Schwartz, and S.J. Withrow. 1998. Maxillectomy and premaxillectomy. In Current Techniques in Small Animal Surgery, pp. 124–132. M.J. Bojrab, G.W. Ellison, and B. Slocum, editors. Baltimore: Williams & Wilkins.)

147

148  Veterinary Surgical Oncology Figure 6.24.  (A) A caudal maxillectomy through a combined approach is recommended for tumors that arise or extend dorsolaterally from the caudal maxilla or inferior orbit, such as this squamous cell carcinoma (SCC) arising from the rostral inferior orbit and zygomatic arch. (B) A skin incision is performed along the dorsolateral aspect of the muzzle and extended ventral to the eye and along the zygomatic arch. (C) This incision is then continued through the subcutaneous tissues and to the level of the bone. Note the skin margins around the biopsy site (arrow) that will be excised en bloc with the maxillary bone segment. (D) If required, the masseter “M” and temporalis “T” muscles should be incised and elevated from the zygomatic arch. (E) A second incision is performed intraorally in the buccal mucosa with a minimum of 1 cm margins from the tumor (arrow). (F) During development of the bipedicle flap, the facial vein “FV” may require ligation in the caudal aspect of the skin incision “X,” but should be preserved if possible to facilitate venous drainage postoperatively (LNL, superficial levator nasolabialis muscle; DNV, dorsal nasal vein; and LNV, lateral nasal vein). (G) Following periosteal elevation of the muscles and appropriate hemostasis, the bipedicle flap can be retracted dorsally or ventrally to improve exposure for tumor excision (SCC, arrow). (H–K) The position of the osteotomies is dependent on the tumor type and location; however, in general, osteotomies are performed with an oscillating saw in the zygomatic arch “1,” with either an oscillating saw, osteotome and mallet, or (I) pneumatic burr in the dorsolateral (see number 3 in part H) and (J) rostral (see number 2 in part H) maxilla, with an oscillating saw or osteotome and mallet in the hard palate (K), and then through the inferior orbit (labeled “A” in part H) with an osteotome and mallet to connect the dorsolateral maxilla and hard palate osteotomies. (L) The osteotomized bone segment is then gently elevated from the surgical site and removed. (M) The deep layer of the intraoral closure is best performed through the skin incision with simple interrupted sutures between the buccal-labial submucosa and predrilled holes in the bone of the hard palate (arrowheads). (N) The buccal-labial mucosa is sutured to the mucoperiosteum of the hard palate using either a simple interrupted or continuous pattern of absorbable monofilament suture material. (O) The dorsolateral skin incision is closed routinely. (Images F and H reproduced with permission from Lascelles, B.D.X., M.L. Thomson, W.S. Dernell, et al. 2003. Combined dorsolateral and intraoral approach for the resections of tumors of the maxilla in the dog. J Am Anim Hosp Assoc 39:294–305.)

sufficient tissue to prevent the overlying skin being drawn medially, resulting in a poor cosmetic result. The flap is sutured into position with a two-layer closure, with the first layer consisting of simple interrupted sutures preplaced through holes predrilled in the bone of the hard palate. The labial and oral mucosa are opposed using a simple interrupted or continuous suture pattern (Figures 6.23F, G) (Dernell et al. 1998b). Caudal maxillectomy—combined approach Caudal maxillectomy via a combined intraoral and dorsolateral approach is recommended for tumors of the mid-to-caudal maxilla that either arise or extend dorsolaterally and/or caudally to the alveolar margin (Figure 6.24A) (Dernell et al. 1998b; Lascelles et al. 2003). The combined approach provides better exposure and thus improved ability to achieve hemostasis and completely excise the tumor compared to the intraoral approach for more extensive caudal maxillary tumors (O’Brien et al. 1996; Lascelles et al. 2003). As discussed previously, advanced imaging is recommended for tumors of the caudal maxilla and orbit to determine the extent of the tumor, resectability of the tumor, and to plan the surgical approach (see Figure 6.1). In some cases, particularly those with more extensive involvement of the orbit, enucleation may be required to improve exposure and likelihood of achieving complete excision of the tumor. The possibility of enucleation and adjuvant radiation

therapy if the tumor is incompletely excised should be discussed with the owner prior to surgery. Dogs are positioned in lateral recumbency with the mouth held open with a gag. The pharynx should be packed with gauze sponges to minimize the risk of aspirating blood and lavage fluid. The dorsolateral skin incision is created first with an incision lateral to the midline of the dorsal aspect of the nasal cavity and extending caudally and ventrally to the eye along the zygomatic bone (Figure 6.24B) (Lascelles et al. 2003). This incision is continued through the subcutaneous tissue, between the paired levator nasolabialis muscles, and down to bone (Figure 6.24C). Caudally, the ventral aspect of the globe is separated from the dorsal zygoma with a combination of sharp and blunt dissection, leaving the conjunctival sac intact, and the masseter muscle is elevated from the ventral aspect of the zygomatic arch using a combination of sharp and blunt dissection (Figure 6.24D) (Lascelles et al. 2003). The most common complication of the combined approach for caudal maxillectomy is blood loss from the maxillary artery (Lascelles et al. 2003). The maxillary artery should be ligated early in the procedure to prevent this complication. The maxillary and palatine arteries are exposed deep within the orbit by carefully retracting the globe dorsally and ligated with either suture material or metallic clips. Exposure may be limited and can be improved by resecting the zygomatic arch rostral to the orbital ligament (Lascelles et al. 2003); however, this is often difficult and

(a)

(b)

(d)

(g)

(c)

(e)

(f)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

149

150  Veterinary Surgical Oncology

ligation of the maxillary artery is frequently completed at the end of the procedure when the bone segment is freed. The tumor should be excised with appropriate margins depending on whether it is benign or malignant. The dorsal margins are prepared by reflecting the periosteum and associated soft tissues with a periosteal elevator. A second incision is made in the buccal musosa dorsal to the gingiva and with appropriate margins through an intraoral approach (Figure 6.24E). The mucosal incision is continued down to the bone and undermined with periosteal elevators and scissors to connect the mucosal and dorsal skin incisions and create a bipedicle flap (Lascelles et al. 2003). The infraorbital artery and vein, which will be encountered during this dissection along with the infraorbital nerve, should be individually ligated. The facial vein should also be ligated at the most dorsocaudal aspect of the incision (Figure 6.24F), but the facial vein should be preserved at the level of the dorsal nasal tributary if possible to facilitate venous drainage from the muzzle (Lascelles et al. 2003). The bipedicle flap can be retracted dorsally and ventrally to allow visualization of the lateral aspect of the maxilla and improve exposure for tumor excision (Figure 6.24G) (Lascelles et al. 2003). The dorsal and rostral maxillary osteotomies are performed at appropriate margins from the tumor with an oscillating saw, pneumatic burr, or osteotome (Figures 6.24H–J). An incision is then made through the mucoperiosteum of the hard palate medial to the alveolar margin or as dictated by the margins of the tumor. The major palatine artery or its branches may be transected during this incision, and bleeding from these vessels can be controlled with either digital pressure, cautery, or ligation. The mucoperiosteum is elevated with a periosteal elevator to preserve the soft tissues, which will be used later for reconstruction of the defect. An osteotomy is performed in the palatine bone from the ventral aspect of the rostral maxillary osteotomy and extending caudally to the planned caudal extent of the excision (Figure 6.24K). Finally, the caudal osteotomy is completed such that it connects the caudal aspects of the dorsal maxillary and palatine osteotomies (Lascelles et al. 2003). Both the palatine and caudal osteotomies can be performed with either an oscillating saw or osteotome and mallet. If the caudal osteotomy involves the inferior orbit, then this is preferably performed with an osteotome and mallet, with the osteotome started at the junction between the inferior orbit and caudal maxilla, caudal to the molar teeth, and directed rostrally to connect the lateral maxillary and palatine osteotomies. The maxillary artery courses deeply through the

caudoventral aspect of the inferior orbit. If the maxillary artery has not been previously ligated, then it should be ligated with either suture material or metallic clips at this stage (Lascelles et al. 2003). Following completion of the osteotomies, the free bone segment should be gently elevated and removed (Figure 6.24 L). Bleeding from the surgical site is then controlled with either suture or metallic clip ligation or cautery for vessels and a combination of tamponade and ice-cold saline or 0.05% oxymetazoline spray for bleeding from the tur­ binates. Cautery is rarely successful for turbinate bleeding. The intraoral incision is closed first. This is best approached from a dorsal direction through the bipedicle flap (Lascelles et al. 2003). The degree of tension on the lip should be assessed prior to closure. To minimize tension and to prevent the lip being drawn medially, resulting in a poor cosmetic result, particularly following extensive resection of the hard palate, a labial mucosal-submucosal flap may be required. The flap is sutured into position with a two-layer closure (Figure 6.24M,N). If this closure is performed through the dorsal skin incision, then the labial mucosa and mucoperiosteum of the hard palate are opposed using a simple interrupted or continuous suture pattern. The thick fibrovascular free edge of the lip is sutured to predrilled holes in the bone of the hard palate (Lascelles et al. 2003). If this closure is performed through an intraoral approach, which tends to be more awkward, then the lip submucosa-to-palatine bone closure is performed first, followed by the mucosa-mucoperiosteum closure. Finally, the dorsolateral skin incision is closed routinely (Figure 6.24O). No attempt is made to close the dead space between the oral and lateral nasal incisions. Drains are not required, as the surgical site opens into the nasal cavity and drainage occurs via the nasal orifice. Surgical approach to tumors of the maxilla in cats The surgical approach for management of benign and malignant tumors of the maxilla is similar to dogs. Maxillectomy in cats is complicated by the small size of the maxilla relative to the size of the oral tumor and the need for 1 cm margins for benign lesions and 3 cm margins for malignant tumors. As a result, less aggressive procedures such as rostral maxillectomy are rarely possible. Hemimaxillectomy, either via an intraoral or combined approach (Figures 6.25A–D), is the most commonly indicated and performed procedure for the management of maxillary tumors in cats. Temporary carotid artery occlusion should not be performed in cats

Oral Tumors  151

(a)

(c)

(b)

(d)

Figure 6.25.  Caudal maxillectomy through a combined approach in a cat with a maxillary osteosarcoma. (A) The dorsolateral skin and intraoral incisions have been performed to create a bipedicle flap. (B) The dorsolateral (arrows) and caudal osteotomies (arrowheads) have been performed with an osteotome and mallet through the skin incision to achieve adequate dorsal and caudal margins. (C) The hard palate osteotomy (arrows) has been performed with an oscillating saw to connect to the rostral and caudal maxillary osteotomies. (D) The osteotomized bone segment is gently removed from the surgical site.

to decrease intraoperative blood loss because it can result in fatal cerebral hypoxia (Holmes and Wolstencroft 1959; Gillian 1976; Holmberg 1996). Maxillectomy does not result in the same functional consequences as mandibulectomy in cats, with eating and grooming rarely affected in comparison. Although rarely necessary, an esophageal or gastric feeding tube could be considered for nutritional supplementation postoperatively. Postoperative management Analgesia In the immediate postoperative period, intravenous fluids and analgesia are continued, and an Elizabethan

collar should be placed as soon as the animal is ster­ nally recumbent to prevent self-trauma (Dernell et al. 1998b). Analgesia should include a nonsteroidal antiinflammatory drug and an opioid. COX-2 selective or specific nonsteroidal anti-inflammatory drugs are preferred because of their safety index, efficacy, and possible anticancer effects (Umar et al. 2003). Nonsteroidal antiinflammatory drugs should either not be administered or their dose decreased in animals with conditions such as renal failure, hypotension, or hepatic disease (Mathews 2000; Lascelles et al. 2007). Opioids, such as fentanyl or morphine, are preferably administered as a continuous rate infusion rather than intermittent intramuscular

152  Veterinary Surgical Oncology

(a)

(b)

(d)

(e)

(c)

Figure 6.26.  (A) The typical appearance of a dog following caudal maxillectomy through an intraoral approach 24 hours postoperatively (note the epistaxis). (B) The typical appearance of a dog following caudal maxillectomy through a combined approach 24 hours postoperatively. (C) The typical appearance of a cat following caudal maxillectomy through a combined approach 24 hours postoperatively. (D and E) Depending on the degree of resection of the hard palate, the lip may be drawn toward the midline of the muzzle. Although some medialization of the lip is to be expected, this can be minimized by undermining the labial submucosal-mucosal flap.

injections. Continuous rate infusions of opioids can be combined with ketamine and/or lidocaine for an enhanced analgesic effect. Animals can usually be weaned off continuous rate infusions over a 24–48 hour period. Cats and dogs should be discharged with a nonsteroidal anti-inflammatory drug and an oral opioid, such as codeine or tramadol. Nonsteroidal antiinflammatory drugs should be used with care, particularly in cats (Lascelles et al. 2007). See Table 6.2 for suggested perioperative analgesic protocols for cats and dogs undergoing mandibular resections. Nutrition Intravenous fluids should be continued until the dog eats voluntarily and is drinking sufficient quantities to maintain hydration. This is rarely a problem, and most dogs can be discharged within 24–48 hours. Dogs treated with radical maxillectomy may have difficulty eating dry food and may also require initial assistance in feeding (Lascelles et al. 2003). This includes manual feeding or feeding from an inclined bowl. However, the majority of dogs adapt to unassisted eating within 2–3 weeks (Lascelles et al. 2003). Supplemental nutrition is rarely required in either dogs or cats following most maxil-

lectomy procedures, but an esophagostomy or gastrostomy feeding tube is recommended in cats following bilateral rostral maxillectomy, with or without nasal planum resection, or radical maxillectomy. To prevent disruption of intraoral incisions, cats and dogs should only be fed soft canned food and prevented from chewing on hard objects or playing with toys for 4 weeks. Miscellaneous An Elizabethan collar is applied in the immediate postoperative period to minimize self-trauma. Bleeding from the ipsilateral nostril(s) is common for 1–3 days postoperatively (Figure 6.26A). This rarely requires treatment, but the volume should be monitored. Following radical maxillectomy, the surgical site oozes serosanguineous fluid and becomes crusty and contaminated with food material and saliva. Topical petrolatum-based antibiotic ointment is initially placed around the nasal orifice wounds and a topical misting of physiological saline administered by a conventional spray bottle can be useful to humidify and cleanse the nasal turbinates. Following discharge, owners should clean and maintain the patency of the new rostral orifice

Oral Tumors  153

(a)

(b)

Figure 6.27.  ( A) The postoperative appearance of a dog following unilateral rostral maxillectomy. The lesion was relatively small and the labial mucosal-submucosal flap has been sufficiently undermined to prevent excessive medialization of the lip. (B) For dogs with larger lesions, there may be tension on the labial mucosal-submucosal flap resulting in medialization of the lip such that the lip is positioned medial to the ipsilateral mandibular canine tooth. Medialization of the lip can be minimized by undermining the labial submucosal-mucosal flap, but it can be difficult to prevent this cosmetic result.

with saline-soaked cotton balls or cotton swabs for approximately 4 weeks. The surgical site does not tend to be contaminated with food material once healing is complete (by 8 weeks). However, dogs and cats will continue to have a mild, persistent, clear nasal discharge. This does not bother animals and does not result in dermatitis on the new rostral lip. Rhinitis, a concern because of the exposed turbinates, does not occur. Hair regrowth around the new nasal orifice does not cause any problems, and we have not seen any cases of selftrauma of the new orifice from the tongue. Tear staining can be expected due to disruption of the nasolacrimal duct. Complications Cosmetic appearance The cosmetic appearance of cats and dogs following various maxillectomy procedures is usually good to excellent. The major exception is radical maxillectomy. As discussed previously, owner acceptance of post­ operative appearance and function is improved with a thorough discussion, including the use of pre-and postoperative images of the appropriate procedure, before surgery. Owner satisfaction with the cosmetic appearance and functional outcome following maxillectomy is high, with 85% of dog owners satisfied following partial maxillectomy (Fox et al. 1997). Following unilateral rostral and caudal maxillectomies (Figures 6.26A–E), the skin and lip are often drawn medially toward the midline, and the extent of this will depend on the medial extent of the resection, resulting

Figure 6.28.  The typical appearance of a dog following bilateral rostral maxillectomy. Note the mild drooping of the nose as a result of loss of ventral palatine support.

in a dished-in appearance (Fox et al. 1997). Depending on the extent of this medialization of the lip, the ipsilateral mandibular canine tooth may protrude lateral to the upper lip following maxillectomy procedures involving the rostral maxilla (Figure 6.27A,B). Postoperative swelling can also be significant, particularly if venous drainage has been compromised during tumor excision, but this usually subsides within 3 weeks, resulting in an improved cosmetic appearance. The most common cosmetic defect following bilateral rostral maxillectomy is drooping of the nose because of loss of ventral support (Figure 6.28) (White et al. 1985;

154  Veterinary Surgical Oncology

(a)

(b)

Figure 6.29.  (A, B) The typical postoperative appearance following radical maxillectomy. Of all the surgical techniques used for resection of oral tumors, radical maxillectomy has the most challenging cosmetic results. Note the shortening of the muzzle and exposure of the mandible and mandibular teeth.

Withrow et al. 1985; Salisbury et al. 1986; Schwarz et al. 1991b; White 1991; Wallace et al. 1992; Lascelles et al. 2003). Another common cosmetic defect is protrusion of the mandibular canine teeth rostral to the resected maxilla following bilateral rostral maxillectomy combined with nasal planum resection and radical maxillectomy (Lascelles et al. 2003, 2004). The cosmetic appearance of dogs following radical maxillectomy is the most challenging of all maxillofacial resections (Figures 6.29A, B) (Lascelles et al. 2004). For owners that have elected to proceed with this procedure, children within the family and uninformed visitors have the greatest difficulty in accepting the altered cosmetic appearance. Owners should be thoroughly counseled prior to surgery regarding the expected appearance of their dog. As previously recommended, the use of postoperative images of dogs treated with radical maxillectomy facilitates this discussion. Despite the change in cosmetic appearance, it is important to note that the dog’s behavior is not altered and its function remains good to excellent. The most significant functional effects include difficulty or inability in retrieving or picking up items, difficulty in eating dry food, and messy eating and drinking. Eating difficulties Eating difficulties are uncommon following the majority of maxillectomy procedures in cats and dogs (White et al. 1985; Withrow et al. 1985; Salisbury et al. 1986; Schwarz et al. 1991b; White 1991; Wallace et al. 1992;

Fox et al. 1997; Dernell et al. 1998b; Lascelles et al. 2003). Dogs treated with radical maxillectomy may have difficulty eating dry food and may also require initial assistance in feeding (Lascelles et al. 2004). This includes manual feeding or feeding from an inclined bowl. However, the majority of dogs adapt to unassisted eating within 2–3 weeks. Supplemental feeding may be required for up to a week in cats following bilateral rostral maxillectomy, with or without nasal planum resection, or radical maxillectomy. Wound dehiscence and oronasal fistula Wound dehiscence is the most common complication following maxillectomy and is reported in 5%–33% of dogs and can result in the development of an oronasal fistula (White et al. 1985; Withrow et al. 1985; Harvey 1986; Salisbury et al. 1986; Schwarz et al. 1991b; White 1991; Wallace et al. 1992; Fox et al. 1997; Dernell et al. 1998b; Lascelles et al. 2003). Wound dehiscence most commonly occurs within 3–7 days of surgery and usually caudal to the canine teeth (Harvey 1986; Schwarz et al. 1991b). Tension at these sites is the most likely cause of dehiscence, although the use of cautery, rapidly absorbing suture material (i.e., catgut or poliglecaprone 25), and poor wound-healing capabilities as a result of radiation therapy, chemotherapy, or debilitation may also contribute to wound dehiscence (White et al. 1985; Withrow et al. 1985; Harvey 1986; Salisbury et al. 1986; Schwarz et al. 1991b; White 1991; Wallace et al. 1992;

Oral Tumors  155

Figure 6.31.  The oronasal fistula has been debrided and repaired with an transposition flap of adjacent skin. Mucosal or mucoperiosteal flaps are preferred for repair of oronasal defects, but these tissues are often not available because of previous tumor resection.

Figure 6.30.  Dehiscence of the intraoral incision following caudal maxillectomy through an intraoral approach with the subsequent development of an oronasal fistula (arrow).

Fox et al. 1997; Dernell et al. 1998b; Lascelles et al. 2003). To minimize the risk of dehiscence, maxillectomy defects should be closed in two layers, preferably using bone tunnels for the first layer, under minimal tension with long-lasting monofilament suture material. Tension can be decreased with careful planning and harvesting of the labial mucosal-submucosal flap. Cautery is frequently cited as the cause for dehiscence, however, it is rare if the maxillectomy defect is closed with tension-free closure techniques. If dehiscence occurs, then the full extent of dehiscence should be assessed. Dehiscence or failure of the labial mucosa-submucosal flap can result in the development of an oronasal fistula (Figure 6.30). A number of techniques have been described for the management of oronasal fistulae (Kirby 1990; Griffiths and Sullivan 2001; Lanz 2001; Niles and Birchard 2001; Bryant et al. 2003; Dundas et al. 2005). However, these may not be possible following maxillectomy because much of the available buccal mucosal or mucoperiosteal tissue typically used for these reconstructions has either been excised or used for reconstruction of the original maxillectomy defect. In these cases, angularis oris axial pattern buccal flaps (Bryant et al. 2003), advancement of skin flaps into the oral cavity (Figure 6.31) (Dundas et al. 2005), or free microvascular grafts of the rectus abdominis muscle may be useful (Lanz 2001). If the

dehiscence does not involve an oronasal fistula, then these can be managed with either second-intention healing if the dehisced area is small and granulating or debridement and resuturing if the defect is large. Incisional swelling Swelling of the surgical site is common following caudal maxillectomy, particularly with the combined intra­ oral-dorsolateral approach, and radical maxillectomy (Lascelles et al. 2003, 2004). Swelling usually resolves spontaneously within 3 weeks. Ice packing every 4 hours and the administration of nonsteroidal antiinflammatory drugs may assist in decreasing the severity of postoperative swelling. Ulcer formation secondary to trauma by teeth Ulcer formation due to trauma by teeth is a relatively common complication following maxillectomies involving the rostral maxilla (White et al. 1985; Withrow et al. 1985; Harvey 1986; Salisbury et al. 1986; Schwarz et al. 1991b; White 1991; Wallace et al. 1992; Fox et al. 1997; Dernell et al. 1998b; Lascelles et al. 2003). The ulcerated region is usually caused by the ipsilateral mandibular canine tooth (Figure 6.32), although any teeth can cause this trauma, and can occur either inside or outside the lip. Ulceration results from the lip being drawn medially into the occlusal plane of the teeth. Ulcer formation is potentiated by the loss of sensation to the upper lip following severance of the infraorbital nerve. Once the

156  Veterinary Surgical Oncology

Figure 6.32.  Ulceration of the upper lip in a cat following unilateral hemimaxillectomy. The upper lip has been drawn medially into the occlusal plane of the mandibular canine tooth resulting in ulceration of the lip margin (arrow). In this case, the ulcerated lip resolved following capping of the mandibular canine tooth.

source of the ulceration has been identified, the tooth or teeth are removed or capped. Miscellaneous complications Other complications reported following maxillectomy include hemorrhage, pain, subcutaneous emphysema, infection, and nasal discharge secondary to rhinitis. Intraoperative hemorrhage can be profuse if the major palatine or maxillary artery are not ligated prior to caudal hemimaxillectomy (Dernell et al. 1998b). To avoid this complication, the maxillary artery should be identified as it courses along the ventral aspect of the inferior orbit and ligated. Temporary occlusion of the carotid arteries may be effective in reducing bleeding from the major palatine artery, but this should only be performed in dogs as carotid artery occlusion can be fatal in cats (Holmes and Wolstencroft 1959; Gillian 1976; Hedlund et al. 1983; Holmberg 1996; Holmberg and Pettifer 1997). Hemorrhage from either the maxillary or major palatine artery can be severe enough to warrant a whole blood transfusion, so blood loss should be carefully monitored and a cross-matched compatible blood transfusion considered if hematocrit acutely decreases below 15%–30%, particularly if this occurs in combination with hypotension, hypoxia, and/or clinical, biochemical, or echocardiographic evidence of anaerobic metabolism (Jutkowitz 2004). Subcutaneous emphysema, with skin over the surgical site moving with respiration, is occasionally noted in the early postoperative period after maxillectomy proce-

dures involving exposure of the nasal cavity, particularly caudal maxillectomies (Dernell et al. 1998b). This is usually mild, nonprogressive, and resolves spontaneously within 7 days. Infection is very rare following maxillectomy because of the rich vascular supply to the oral cavity (Dernell et al. 1998b). Incisional abscesses are treated with debridement, copious lavage with an isotonic crystalloid solution, closure of dead space, drainage if possible, and culture-directed antibiotics. Rarely, a mild but persistent nasal discharge is observed following maxillectomy procedures in which the nasal turbinates have been exposed (White et al. 1985; Withrow et al. 1985; Harvey 1986; Salisbury et al. 1986; Schwarz et al. 1991b; White 1991; Wallace et al. 1992; Fox et al. 1997; Dernell et al. 1998b; Lascelles et al. 2003). This discharge is usually clear, but can occasionally be mucoid to mucopurulent. Treatment is rarely required, but culture-directed antibiotics may be necessary if an infected rhinitis is suspected.

Mandibular and Maxillary Tumors in Dogs The most common malignant tumors of the mandible and maxilla in dogs are, in descending order, malignant melanoma, SCC, and fibrosarcoma (Todoroff and Brodey 1979; Withrow and Holmberg 1983; Bradley et al. 1984; White et al. 1985; Withrow et al. 1985; Salisbury et al. 1986; Salisbury and Lantz 1988; Kosovsky et al. 1991; Schwarz et al. 1991a, 1991b; White 1991; Wallace et al. 1992). Other malignant oral tumors include osteosarcoma, chondrosarcoma, anaplastic sarcoma, multilobular osteochondrosarcoma, intraosseous carcinoma, myxosarcoma, hemangiosarcoma, lymphoma, and mast cell tumor (Madewell et al. 1976; Todoroff and Brodey 1979; Withrow and Holmberg 1983; Bradley et al. 1984; White et al. 1985; Withrow et al. 1985; Salisbury et al. 1986; Salisbury and Lantz 1988; Straw et al. 1989; Reeves et al. 1993; Smith 1995; Holmberg and Pettifer 1997; Dernell et al. 1998c; Dhaliwal et al. 1998; Williams and Packer 2003; Dennis et al. 2006; Liptak and Withrow 2007). SCC is the most common oropharyngeal cancer in cats and the most frequently diagnosed tumor in the tongue of dogs (Madewell et al. 1976; Todoroff and Brodey 1979; Withrow and Holmberg 1983; Bradley et al. 1984; White et al. 1985; Withrow et al. 1985; Salisbury et al. 1986; Salisbury and Lantz 1988; Straw et al. 1989; White 1991; Reeves et al. 1993; Smith 1995; Holmberg and Pettifer 1997; Dernell et al. 1998c; Dhaliwal et al. 1998; Williams and Packer 2003; Dennis et al. 2006; Liptak and Withrow 2007). A summary of the common oral tumors is found in Table 6.3.

Table 6.3.  Summary of common oral tumors in the dog and cat. Canine

Feline

Malignant Melanoma

Squamous Cell Carcinoma

Fibrosarcoma

Acanthomatous Ameloblastoma

Squamous Cell Carcinoma

Fibrosarcoma

Frequency Age (years) Sex predisposition Animal size Site predilection

30%–40% 12 None-Male

17%–25% 8–10 None

8%–25% 7–9 Male

5% 8 None

70%–80% 10–12 None

13%–17% 10 None

Smaller Gingiva, buccal, and labial mucosa

Larger Rostral mandible

None Rostral mandible

— Tongue, pharynx, and tonsils

— Gingiva

Lymph node metastasis

Common (41%– 74%)

None

Rare

Rare

Distant metastasis Gross appearance

Common (14%– 92%) Pigmented (67%) or amelanotic (33%), ulcerated Common (57%)

Rare (21.9 seconds) were more likely to have complications than patients with normal values. It is therefore recommended that routine prebiopsy coagulation profiles (i.e., one-stage prothrombin time, activated partial thromboplastin time, and platelet count) be performed prior to ultrasound-guided biopsy

198  Veterinary Surgical Oncology

procedures (Bigge et al. 2001; Rawlings and Howerth 2004). A greater concern with ultrasound guided tissue procurement of neoplastic lesions is the accuracy of these tests when compared to open surgical tissue biopsy. In one study, agreement between fine-needle liver aspirates and surgical wedge biopsies were found in only 30.3% (17 of 56) of canine cases and 51.2% (21 of 41) of feline cases, respectively (Wang et al. 2004). A second study showed that the morphological diagnosis assigned to a needle biopsy specimen concurred with the definitive histological diagnosis in only 40% (36 of 91) of dogs and cats with hepatic disease (Cole et al. 2002). These discrepancies should serve as a caution to clinicians considering a fine-needle aspirate to be a definitive diagnostic tool when characterizing infiltrative or neoplastic liver disease (Rawlings and Howerth 2004). Attempts at improving the correlation between cytology and histopathology have been investigated through the application of the immunohistochemical proliferation marker, Ki-67, to cytological liver specimens. In this study, it was identified that cytological specimens of dogs with liver tumors (n = 9) showed greater than 50% Ki-67-positive cells when compared to dogs (n = 21) with nonneoplastic liver disease having little or no Ki-67 positive cells (Neumann and Kaup 2005). Using Ki-67 proliferation indices, it was determined that the diagnostic accuracy of cytological evaluation was increased from 78% to 100% for malignant neoplasia (Neumann and Kaup 2005). Serological quantification of alphafetoprotein (AFP) is another novel technique that has been investigated as a means to confirm a diagnosis of malignant liver neoplasia in dogs. In people, abnormally high levels of circulating AFP have been demonstrated in 70%–80% of human hepatic tumors (Lowseth et al. 1991). In China, AFP monitoring has become a routine screening tool for detection of hepatocellular carcinoma (Lowseth et al. 1991). A single canine study demonstrated significantly higher serum levels of AFP in dogs with hepatocellular carcinoma and cholangiocarcinoma when compared to dogs with other types of liver neoplasia, nonneoplastic liver disease, and no liver disease (Lowseth et al. 1991). It should be noted that these techniques have not gained widespread acceptance, and additional large-scale studies are needed before their clinical utility is validated. Laparoscopic biopsy techniques Laparoscopy has become a routinely used adjunct to preoperative imaging examinations in humans with hepatic neoplasia (Lo et al. 2000; Montorsi et al. 2002). In particular, laparoscopy is an effective tool for staging the local extent of disease and for determining ideal

candidate selection in patients that will subsequently undergo laparotomy for hepatic resection. Importantly, the use of laparoscopy can help avoid unnecessary laparotomy in patients with unresectable disease (Lo et al. 2000). In dogs, laparoscopic biopsy of the liver using clamshell laparoscopic biopsy forceps has been shown to produce minimal immediate hemorrhage and results in adequate tissue samples for accurate histological evaluation (Vananjee et al. 2006). Samples can be obtained from regions of the liver (adjacent to biliary structures and large vessels) that are generally inaccessible with percutaneous techniques, and laparoscopy is not strictly contraindicated in patients with ascites or coagulation defects (Rawlings and Howerth 2004). Additionally, the 2D imaging achieved with laparoscopy is superior to that of traditional laparotomy, as images are magnified by the laparoscope (Rawlings and Howerth 2004). Laparoscopy is routinely used as an adjunct staging tool when preoperative imaging analysis has failed to instill confidence that the liver tumor is resectable or when primary hepatic origin of the tumor cannot be established. Detailed descriptions of operative laparoscopic biopsy techniques have been published and may be found elsewhere (Mayhew 2009). Incisional biopsy and partial lobectomy Liver biopsy via laparotomy is considered the gold standard technique for procuring representative tissue sufficient to achieve a definitive histological diagnosis. In humans, a biopsied liver sample must contain a minimum of six to eight portal triads in order to be considered adequate for an accurate histological diagnosis. In one veterinary study, 19% of 4 mm punch-biopsy samples and 42% of needle-biopsy (16-gauge needle) samples contained less than six to eight portal triads (Vananjee et al. 2006). Direct translation from human studies regarding the ideal number of portal triads necessary for an adequate tissue diagnosis cannot be made; however, concern exists over the small sample size obtained through minimally invasive techniques. Liver biopsy samples measuring 1 × 1 × 1 cm consistently result in the sufficient number of portal triads required for accurate histological assessment. Therefore, open surgical biopsies of the liver should result in procurement of at least this volume of tissue (Vananjee et al. 2006). Nodules that are centrally located within a liver lobe may preclude the removal of an ideal volume of liver tissue for biopsy. In general, use of a 6 mm (or larger) Baker or Keyes skin biopsy punch to procure tissue from a portion of the nodule is sufficient for accurate histological assessment. When performing this technique, it is important to avoid penetration of more than

Alimentary Tract  199

half the thickness of the liver lobe so as to prevent traumatic laceration of the large hepatic veins situated along the concave surface of the lobe (Martin et al. 2003). Once the biopsy tissue has been removed from the liver, hemorrhage from the biopsy defect can be controlled through placement of a premeasured (cut to fit) piece of Gelfoam dressing within the defect, followed by several minutes of gentle tamponade (Martin et al. 2003). With extremely vascular tumors (i.e., metastatic or primary hemangiosarcoma), uncontrolled hemorrhage may occur from the biopsy site despite tamponade with a hemostatic agent. Gentle placement of a mattress suture (the author prefers a small diameter, monofilament absorbable suture such as 3-0 to 4-0 polydioxinone) across the defect may aid in further controlling hemorrhage; however, the friable nature of the liver in these situations may result in tearing once the suture is tightened. As a failsafe mechanism, the surgeon should be prepared to perform a partial lobectomy to gain control of the hemorrhaging biopsy tract. This reiterates the need for careful planning of a proposed biopsy site. When multiple representative nodules are present, preferential selection of a peripherally located nodule in the left division may facilitate “damage control,” if a biopsyrelated complication occurs. Incisional biopsy and partial liver lobectomy techniques can be categorized based on the location of the tumor within the liver lobe. Wedge resections or encircling ligature (guillotine) placement techniques are used for peripherally located tumors, whereas midbody lobe dissection and ligation techniques are used for more centralized lesions. The encircling ligature technique is applicable for small nodules that are located at the extreme periphery of a liver lobe or for attainment of a section of diffusely infiltrated liver that is to be removed for staging purposes. Wedge resections can be used for larger lesions and for those that are not located at the extreme periphery of the lobe. With either procedure, a single crushing (guillotine) ligature (Figure 7.10a) or multiple overlapping mattress ligatures (wedge) (Figure 7.10b) are placed such that the liver parenchyma is compressed sufficiently to prevent the leakage of blood and bile (Rawlings and Howerth 2004). For partial lobectomy of tumors located more centrally within the liver lobe, a moderate amount of hepatic parenchymal dissection will be necessary. Initially, the proposed section of liver to be resected is outlined by sharply (scalpel blade or electrocautery) dividing the liver capsule along the ventral or convex surface of the lobe (Blass and Seim 1985). Following this, blunt dissection of the hepatic parenchyma can be accomplished through either digital compression (finger fracture

Figure 7.10.  (A) Illustration demonstrating the guillotine technique for procuring an incisional biopsy of a liver nodule located on the periphery of a lobe. (B, C) Illustration demonstrating the mattress suture (wedge) technique for procuring an incisional biopsy of a liver nodule located on the periphery of a lobe. (Illustration courtesy of Dave Carlson.)

technique) or separation using the blunt end of a Bard scalpel handle. This blunt dissection will allow isolation and ligation of vessels located within the hepatic parenchyma. As an alternative to blunt dissection with a scalpel handle or finger fracture, the inner component of a Poole suction tip can be used for intrahepatic dissection. This technique is particularly useful when dissecting through deeper or hard to access regions of the liver lobe since visualization is enhanced when the suction tip removes hemorrhage and portions of the crushed parenchyma (Martin et al. 2003). Once adequate dissection has been achieved, ligation of the remaining parenchyma and its associated vasculature may be achieved through application of an appropriately sized linear stapling device (see total lobectomy next) or by placement of an encircling ligature. Complete lobectomy A ventral midline celiotomy, extending from the xiphoid process to the caudal aspect of the abdomen, is the preferred surgical approach for most liver lobectomy procedures. A combination midline celiotomy and paracostal approach can be used if additional lateral exposure is necessary (Martin et al. 2003). A caudal sternotomy combined with incision into the diaphragm is routinely required to facilitate exposure of very large tumors, tumors originating from an entire liver division, tumors

200  Veterinary Surgical Oncology

with extensive visceral or omental adhesions, and for dissection of tumors originating within the right division (right lateral or caudate lobes). Incision of the diaphragm toward the caval foramen is reported to normalize intrathoracic pressure, thus relaxing the diaphragm, allowing for more caudal mobilization of the liver (Bjorling et al. 1985). Diaphragmatic incision without caudal sternotomy is commonly employed by many oncological surgeons and this technique can also dramatically facilitate exteriorization and visualization of difficult to access liver masses. To ensure adequate visualization of the liver, the falciform ligament must be removed in its entirety. Since the liver serves as a receptacle for metastasis 2.5 times more frequently than a source of primary liver tumor development, a complete exploratory laparotomy should be performed prior to addressing the liver tumor (Strombech 1978; Cullen and Popp 2002). It is not uncommon for liver tumors to have extensive omental adhesions or be adhered to surrounding visceral structures such as the diaphragmatic border of the stomach, the peritoneal wall, the surrounding normal liver lobes, gall bladder, pancreas, or the diaphragm itself. Initial dissection should concentrate on relieving these adhesions so that adequate visualization of the extent of the tumor may be accomplished. Care must be taken to identify large and important vessels (i.e., pancreatic branch of the splenic artery) within the omentum as significant displacement and anatomical derangement of this tissue can occur. To minimize hemorrhage during dissection of the omentum, the author prefers to use the LigaSure Tissue Sealing Device (Covidien [Valleylab], Boulder, CO) or the LDS stapling device (LDS-2 Reusable Instrument with DS-15W Stainless Steel Disposable Loading Unit; Covidien [Autosuture], Norwalk, CT) to alleviate omental adhesions. If rupture of the tumor capsule is evident, a margin of normal omentum should be left behind on the tumor capsule so that additional contamination of the abdomen does not occur when the omentum is returned to its normal anatomical position. The three commonly used surgical techniques for total liver lobectomy include mass suture ligation, dissection and ligation of individual vessels and ducts, and use of linear surgical stapling equipment (Martin et al. 2003; Blass and Seim 1985). Prior to implementation of any technique, the associated triangular ligament of the affected lobe must be severed, as this will allow improved access to the hilum of the lobe (Martin et al. 2003). Mass suture ligation should be reserved for cats and small dogs with tumors located distal to the hilum. To facilitate knot security, a tissue “crush zone” may need to be created within the liver parenchyma near the

hilum. A single encircling ligature is reported to be sufficient for mass ligation; however, authors recommend double ligation based on the demonstrated safety of this approach in research animals undergoing lobectomy (Martin et al. 2003; Lewis et al. 1990). In general, most oncological surgeons prefer to use surgical stapling equipment when performing liver lobectomies. A study in healthy research dogs demonstrated that staple-assisted lobectomies could be performed significantly faster (mean time of 173 seconds vs. 759 seconds with individual ligation) than blunt dissection and individual vessel ligation (Lewis et al. 1990). Importantly, stapled lobectomies also resulted in a more complete excision when compared to the individual suture ligation technique. Suture techniques consistently resulted in 5–15 mm of liver parenchyma remaining along the excision line distal to the ligatures, which could make the difference in achieving an adequate tumor tissue margin during lobectomy (Lewis et al. 1990). It is also reported that less hepatic vein length is required when performing a stapled lobectomy, which may be an important consideration if the tumor is encroaching on the hilum (Martin et al. 2003). The thoracoabdominal (TA-30 Premium Reusable Stapler) stapling device with V3 stapling cartridge (white color) is preferred for occlusion of hilar vessels during complete lobectomy (Covidien [Autosuture], Norwalk, CT). Both TA-55 and TA-90 stapling devices can be used for lobectomies when the hilar tissue exceeds the V3 cartridge length of 30 mm (Figures 7.11A, B) (Martin et al. 2003). It has been recommended that individual vessel ligation be completed when liver parenchymal vessels are 2 mm or more in diameter (Martin et al. 2003). Since the green-colored (TA-55 and TA-90) stapling cartridge achieves a closed staple height of only 2 mm after discharge, use of this cartridge should be avoided if possible (Martin et al. 2003; Tobias 2007). After application of the linear stapling device, it is not uncommon to have focal regions of hemorrhage at the lobectomy site. Minor hemorrhage can be addressed through the application of hemoclips or hemostatic agents such as Gelfoam absorbable gelatin powder (Pfizer [Pharmacia & Upjohn Co.], New York, NY). If generalized oozing from the lobectomy stoma is present, authors prefer to cover the stoma with contoured strips of the hemostatic agent Surgicel (Ethicon 360 [Johnson & Johnson Inc.], New Brunswick, NJ). Excessive parenchymal hemorrhage can be controlled by compression of the hepatic artery and portal vein within the epiploic foramen (Pringle maneuver) (Blass and Seim 1985). Digital compression is preferred over vascular clamp application since the common bile duct is often incorporated into the compressed tissue and

Alimentary Tract  201

(a)

(b)

Figure 7.11.  (A, B) Intraoperative image of TA-90 linear stapling device used for partial lobectomy of a hepatocelluar carcinoma located on the periphery of the lobe. Mild to moderate hemorrhage from the staple should be expected once the cartridge is released. This hemorrhage is generally controlled with tamonade, hemoclips, or application of a hemostatic agent.

posttraumatic bile duct stricture may be less likely with digital occlusion. Application of the Pringle maneuver can be sustained for periods up to 15 minutes without resultant deleterious effects to hepatic structure and function (Blass and Seim 1985). If hemorrhage persists despite compression of the hepatic artery and portal vein, bleeding from the hepatic venous system should be investigated (Blass and Seim 1985). With severe hemorrhage and life-threatening hypotension, the descending aorta can also be temporarily occluded at the aortic hiatus to increase blood pressure and flow to critical organs (Martin et al. 2003). Hepatic lobectomies of central (quadrate and right medial lobes) and right division-based tumors often pose technical challenges that require special consideration prior to attempting excision. Tumors arising from the lobes within the central division often envelop or form robust adhesions to the gallbladder, making its dissection from the hepatic fossae more difficult. Additionally, dissection of the gallbladder away from the tumor often results in laceration of the tumor capsule causing contamination. Based on the intimate association of the quadrate and right medial liver lobes, tumors within the central division or gallbladder often involve both lobes, particularly if its location is near the hilum. It is for these reasons that lobectomy of tumors in the central division often requires a combined cholecystectomy and removal of both of the lobes that constitute the division (Figure 7.12). Dissection and individual ligation of the cystic duct is preferred if cholecystectomy is required; however, this is often not possible. En bloc

Figure 7.12.  Postoperative image of a central division liver lobectomy after a mass effect within the gallbladder caused rupture and secondary bile peritonitis. Adhesions were present in the region of the hilus; however, with careful dissection, the cystic duct could be isolated and individually ligated. After this, both the quadrate and right medial liver lobes could be excised en bloc using a TA-90 (blue cartridge) stapling device.

ligation of the cystic duct and the hilar vessels of both lobes are routinely accomplished using TA stapling equipment, without an increased incidence of postligation hemorrhage or bile leakage. Preligation retrograde catheterization of the common bile duct is generally not required unless obvious pathology within this organ is present.

202  Veterinary Surgical Oncology

Tumor excision arising from lobes within the right division can be particularly challenging due to their intimate association with the caudal vena cava and because of difficulty visualizing and exteriorizing these lobes. A right paracostal incision should be considered to enhance visualization and to assist with precision placement of stapling equipment at the hilum (Lewis et al. 1987, 1990). Reflection of the hepatic parenchyma from the caudal vena cava is necessary and can be achieved using a combination of blunt and sharp dissection (Martin et al. 2003). If the hepatic tissue is friable, blunt dissection with the thin inner component of a Poole suction tip is also helpful (Martin et al. 2003). Authors have found the Poole dissection technique useful for creating a “safe zone” for staple application during liver lobectomies (in any division) where the tumor encroaches on the hilum and cannot be completely excised. This is especially useful during combined cholecystectomy and liver lobectomy of the entire central division, as the hilum is difficult to definitively isolate in this area. In people, use of the Cavitron Ultrasonic Surgical Aspirator, or CUSA (CUS Ultrasonic Surgical Aspiration System; Covidien [Valleylab], Boulder, CO), for hepatic dissection during lobectomies has consistently resulted in a reduction in perioperative blood loss, duration of operative procedures, and postoperative morbidity and mortality (Storck et al. 1991; Fasulo et al. 1992; Farid and O’Connell 1994). Recent reports in the human medical literature also support the use of an electrothermal bipolar vessel sealing system (LigaSure device) for hepatic lobectomy. Like the CUSA, the Ligasure device has been shown to reduce the amount of intraoperative hemorrhage and contributes to a significantly faster operative time when compared to traditional operative techniques (Campagnacci et al. 2007; Saiura et al. 2006; Romano et al. 2005). Application of the ligasure device in human patients with cirrhotic livers results in an increased incidence of hemorrhage and therefore use of the ligaSure in veterinary patients with extremely friable liver tissue should be avoided (Romano et al. 2005). Authors have used both the CUSA and the LigaSure with good success during dissection of hepatic parenchyma for total lobectomy in the dog. Regardless of the technique, during a right division-based liver lobectomy, it is recommended that precautionary measures be taken to control hemorrhage prior to parenchymal dissection around the vena cava. This is best accomplished through the placement of Rommel tourniquets around the caudal vena cava (cranial and caudal to the liver), the portal vein, and the celiac and cranial mesenteric arteries so that massive hemorrhage can be contained while adequate repair of the traumatized vessel can be accomplished (Blass and Seim 1985). In general, the

complication rate associated with complete liver lobectomy for the treatment of liver tumors is relatively low. In the most comprehensive study to date, intraoperative complications occurred in 28.6% of dogs (12 of 42) undergoing exploratory celiotomy and liver lobectomy for primary hepatocellular carcinoma (Liptak et al. 2004a). These complications consisted of traumatic laceration of the vena cava (3 of 42), mild to moderate hemorrhage after lobectomy (7 of 42), and vascular compromise to the left lateral liver lobe (necessitating removal) after excision of the left medial lobe (2 of 42). Two of the 3 dogs suffering from iatrogenic laceration of the vena cava died during the procedure, whereas, the remaining dogs with complications recovered uneventfully (Liptak et al. 2004a). The operative mortality rate in this study was 4.8% (2 of 42). When considering operative mortality rate based on tumor location, none of the dogs with left (28 of 42) or central division (8 of 42) tumors died intraoperatively. A total of 5 dogs were treated for right-sided liver lobe tumors and 2 of these dogs died during surgery, making the intraoperative mortality rate 40% for right division-based tumors (Liptak et al. 2004a). This distinction reiterates the importance of thorough client communication prior to attempting resection of tumors located within the right division of the liver. Recently, a novel surgical technique has been described for isolation and ligation of liver lobes at the level of the hilus (Covey et al. 2009). In this pilot anatomical study of seven cadaver dogs, a detailed description of the vascular and biliary supply to each hepatic lobe was provided. A predictable consistency of the hepatic vasculobiliary anatomy was identified within each lobe, and detailed descriptions for hilar lobectomy were provided. Although the clinical utility of this technique has not been evaluated, its proposed use is in cases where the tumor encroaches on the hilus, making lobectomy through traditional techniques unlikely to yield complete tumor excision (Covey et al. 2009). Aftercare The following variables dictate the intensity of care in the postoperative setting: the preoperative health status of the patient, the existence of underlying secondary or metastatic liver disease, and the type of liver resection that was performed (Blass and Seim 1985). Of primary concern is the potential for postoperative hemorrhage resulting from either technical errors in hemostasis or from the development of an acquired coagulopathy. Serial monitoring of the patient’s blood pressure, heart rate, packed cell volume (PCV), and total protein (TP) levels is essential for establishing trends related to bleeding complications. Commonly, a transient decline in the

Alimentary Tract  203

PCV and TS will be observed shortly after surgery, which is likely secondary to fluid dilution during surgery. Impaired synthesis of prothrombin and other coagulation factors may occur after extensive hepatic resection or if metastatic disease is present, resulting in postoperative coagulopathies (Blass and Seim 1985). Serial monitoring of blood coagulation parameters should be performed in at-risk patients and therapeutic intervention with parenteral vitamin K therapy, fresh whole blood, or fresh-frozen plasma should not be delayed if hemostatic derangement is identified. Functional outcome Dogs and cats tolerate hepatectomy well, presuming that the remaining volume of liver is healthy and that the patient’s general condition is not plagued by a multitude of nonhepatic diseases or systemic metastasis from their underlying hepatic neoplasia. Variables associated with a reduced tolerance to hepatectomy in dogs include serum albumin levels less than 2 g/100 mL, prolonged biliary obstruction, and liver cirrhosis (Martin et al. 2003). Creation of a side-to-side portocaval shunt for portal decompression has been shown to improve survival in research dogs after extensive hepatectomy (Mitsumoto et al. 1999; Martin et al. 2003). It is not uncommon for clients to question the regenerative capacity of the liver in their animals after hepatectomy. The liver has an immense capacity to regenerate, which begins within 24 hours after hepatectomy and peaks at the third postoperative day (Martin et al. 2003; Bjorling et al. 1985). It is reported that after 70% hepatectomy, the liver is capable of complete restoration of its mass by the sixth postoperative week. This is accomplished through both compensatory hypertrophy and hyperplasia of the remaining hepatocytes (MacKenzie et al. 1975; Martin et al. 2003; Bjorling et al. 1985). Up to 70% of the canine liver can be resected without resulting in gross metabolic abnormalities, whereas dogs receiving 84% partial hepatectomy resulted in uniform fatality (Ogata et al. 1997). Common tumors for which this procedure is performed Primary liver tumors are uncommonly reported in the dog and cat, comprising 0.6%–1.3% and 1.0%–2.9% of all canine and feline neoplasms, respectively (Liptak et al. 2004b, 2007). It should be reiterated that the liver serves as a receptacle for metastatic disease 2.5 times more frequently than it does for primary liver tumor development (Strombech 1978; Cullen and Popp 2002; Liptak et al. 2004b). Metastases to the liver commonly result from hematogenous spread; therefore, tumors arising from organs within the portal circulation would

be expected to show high rates of hepatic metastasis (Magne and Withrow 1985). Histologically, most hepatic metastases are carcinomas or variants thereof, with the most commonly reported sites of origin being from the mammary glands, spleen, adrenal glands, pancreas, bone, and lung (Magne and Withrow 1985). Primary liver tumors can be differentiated into four distinct categories based on the histological origin of the tumor: (1) Hepatocelluar, (2) bile duct, (3) carcinoid (neuroendocrine), and (4) mesenchymal (Patnaik et al. 1980; Liptak 2007). Benign liver tumors are more common than malignant tumors in cats, whereas dogs are more commonly diagnosed with malignant tumors (Patnaik et al. 1980; Liptak 2007; Liptak et al. 2004a; Post and Patnaik 1992; Lawrence et al. 1994). When determining candidacy for surgical intervention, the morphological distribution of the tumor (see above) is equally as important as the biology of the underlying disease. Since a preoperative tissue diagnosis is often not attained prior to surgical intervention, the morphologic distribution of the tumor trumps other variables as the single most important factor for determining whether or not a patient is treated surgically. Often in dogs this decision is based on the likelihood of the tumor being a massive hepatocellular carcinoma, and although this is a reasonable and appropriate assumption, it reinforces the necessity of comprehensive preoperative discussion with the animal’s owner and the necessity of a core understanding of the biology of all potential primary liver tumors. Hepatocellular Hepatocellular carcinoma (HCC) is the most common primary liver tumor in dogs and the second most common in cats (Patnaik et al. 1980; Liptak 2007; Liptak et al. 2004a; Post and Patnaik 1992; Lawrence et al. 1994). In dogs with HCC, the massive morphological subtype predominates, representing 61% (30 of 49) of all HCC cases (Figure 7.13). Nodular and diffuse subtypes tend to occur with relatively equal frequency, consisting of 29% (14 of 49) and 10% (5 of 49), respectively (Patnaik et al. 1981a). Additionally, these tumors have a predilection for developing within the left division of the liver, accounting for 67% (20 of 30) to 68.3% (28 of 42) of cases (Liptak et al. 2004b; Patnaik et al. 1981a). This is important because dogs with massive HCC treated for left-sided tumors tend to have less severe intraoperative complications, resulting in significantly longer survival times when compared to dogs with right-sided tumors (Liptak et al. 2004b). The metastatic rate for dogs with HCC is variable, with reported rates ranging from 4.8% (2 of 42) to 61% (35 of 57) (Liptak et al. 2004b; Patnaik et al. 1981a).

204  Veterinary Surgical Oncology

(a)

(b)

Figure 7.13.  (A, B) Postoperative image of canine (A) and feline (B) HCC, both of which were of the massive morphological distribution and were easily excised using en bloc staple ligation of the hilar vessels.

Metastatic potential seems to correlate with tumor morphology. In one study, 35.6% (11 of 30) of dogs with massive HCC had evidence of distant metastasis, whereas 93% (13 of 14) and 100% (5 of 5) of nodular and diffuse HCCs had metastasis, respectively (Liptak et al 2004a; Patnaik et al. 1981a). The most common sites of metastasis include lymph nodes, lungs, other lobes of the liver, and the peritoneum (Liptak et al. 2004b; Patnaik et al. 1981a). Minimal, large-scale veterinary studies assessing the efficacy of surgical treatment for HCC have been published. In the study by Liptak et al. (2004a) surgical treatment (via lobectomy) of 42 dogs with massive HCC yielded a median survival time (MST) of more than 1,460 days. In contrast, dogs that were not treated surgically had a MST of 270 days (range, 0–415 days), which was significantly less than that of surgically treated dogs. The calculated tumor-related mortality rate was 15.4 times higher in dogs treated medically compared to dogs treated with surgery. Interestingly, completeness of excision did not significantly impact long-term survival, and local recurrence was not reported in the 4 dogs with incompletely excised tumors (Liptak et al. 2004a). In the study by Kosovsky et al. (1989), local recurrence was only detected in 1 of 18 dogs after partial liver lobectomy for HCC. These data suggest that lobectomy should be considered in dogs with massive HCC even if excision is not likely to be complete. Likewise, a poor prognosis

should not necessarily be assigned to dogs when postoperative tumor margin assessment reveals that the HCC was incompletely excised. In two comprehensive, retrospective studies evaluating nonhematopoietic hepatobiliary neoplasms in cats, HCC was indentified in only 4% (3 of 62) of the study population (Lawrence et al. 1994; Post and Patnaik 1992). Therefore, specific prognostic criteria and ex­­ pected outcome statistics for feline HCC are not readily available. It is generally assumed that cats treated with complete liver lobectomy for solitary, massive HCCs without evidence of metastatic disease will enjoy prolonged survival times, similar to dogs in this predicament. HCCs in cats with nodular or diffuse distributions would be expected to respond less favorably to surgical therapy; however, this has yet to be substantiated with large-scale scientific studies. Bile duct Carcinoma of the bile duct is the most common primary liver tumor in the cat and the second most common in dogs (Patnaik et al. 1980, 1981b; Liptak 2007; Liptak et al. 2004a; Post and Patnaik 1992; Lawrence et al. 1994). Adenomas represent the second type of bile duct tumor identified in dogs and cats. Bile duct adenomas, also termed hepatobiliary cystadenomas (based on their gross cystic appearance), are common in cats, representing more than 50% of all feline hepatobiliary tumors

Alimentary Tract  205

(a)

(b)

Figure 7.14.  (A, B) Postmortem images of a cat with a massive cystadenoma originating from the left division of the liver. Note the characteristic fluid-filled cystic component of the mass.

(Figures 7.14A, B). Hepatobiliary cystadenomas are relatively uncommon in dogs (Liptak et al. 2004b; Post and Patnaik 1992; Lawrence et al. 1994; Adler and Wilson 1995; Trout et al. 1995). Anatomically, bile duct tumors can be found within any of the three segments of the biliary system: intrahepatic ducts, gallbladder, or extrahepatic ducts. Intrahepatic tumors tend to predominate in dogs with bile duct carcinoma, whereas in cats, a relatively equal distribution of intrahepatic and extrahepatic tumors has been reported. Bile duct carcinoma of the gallbladder is considered rare in both species, accounting for less than 5% of cases (Patnaik et al. 1980, 1981b; Patnaik 1992; Liptak et al. 2004b; Post and Patnaik 1992; Lawrence et al. 1994). Bile duct carcinomas are biologically aggressive tumors with high rates of systemic metastasis. In one study of dogs with bile duct carcinoma, systemic metastasis was found in 21 of 24 (87.5%) cases, with lymph nodes, lungs, and peritoneum representing the most common sites for metastasis (Liptak et al. 2004b; Patnaik et al. 1981b). Diffuse intraperitoneal metastasis and carcinomatosis is frequently reported in cats with bile duct carcinoma, representing 67%–80% of cases (Figure 7.15) (Patnaik 1992; Liptak et al. 2004b; Post and Patnaik 1992; Lawrence et al. 1994).

Figure 7.15.  Intraoperative image of a cat with bile duct carcinoma and secondary carcinomatosis.

206  Veterinary Surgical Oncology Table 7.2.  Literature summary of the frequency of morphologic classifications of malignant primary liver tumors in dogs. Tumor Type

Massive

Nodular

Diffuse

Hepatocellular carcinoma Bile duct carcinoma Neuroendocrine tumor Sarcoma (mesenchymal)

53%–84%

16%–25%

0%–9%

37%–46% 0%

0%–46% 33%

17%–54% 67%

36%

64%

0%

Source:  Reproduced with permission from Liptak, J.M., W.S. Dernell, and S.J. Withrow. 2004. Liver tumors in cats and dogs. Compend Contin Educ Pract Vet 26:50–57.

In dogs with bile duct carcinoma, massive and nodular morphological distribution subtypes tend to predominate, representing 37%–46% and 54%, respectively. Diffuse disease tends to occur less commonly with a reported incidence of 17%–54% (Table 7.2) (Liptak et al. 2004b; Patnaik et al. 1980, 1981b). Liver lobectomy for treatment of massive bile duct carcinoma has been recommended in dogs and cats. Unfortunately, largescale clinical studies are not available, and no prognostic factors have been identified. Clients should be cautioned, however, that most animals treated for bile duct carcinoma tend to succumb to the disease within 6 months of therapy secondary to local recurrence or distant metastasis (Liptak et al. 2004b). Hepatobiliary cystadenomas are slow-growing, benign tumors that tend to occur in older cats and are often discovered incidentally. When present, clinical signs are usually nonspecific and are related to the mass effect created by the tumor compressing surrounding abdominal organs (i.e., vomiting from gastric compression). Treatments for cystadenomas in people include percutaneous aspiration and drainage, marsupialization, and partial or total excision. Malignant transformation from cystadenoma to carcinoma has been reported in cats, and therefore total excision via lobectomy has been recommended as the treatment of choice (Liptak et al. 2004b; Adler and Wilson 1995; Trout et al. 1995). The morphological distribution of the tumor should be considered prior to lobectomy as multicentric disease has been reported with a frequency equal to massive, solitary lesions (Adler and Wilson 1995). Generally, cats with solitary lesions that are treated with lobectomy enjoy prolonged survival times. In a study of five cats undergoing lobectomy for treatment of

hepatobiliary cystadenomas, surgical complications were not observed and tumor recurrence or tumorrelated mortality was not observed (Trout et al. 1995). Carcinoid (neuroendocrine) Hepatic carcinoid tumors, derived from neuroectodermal tissue or from amine precursor uptake and decarboxylation (APUD) cells, are rarely reported in dogs and cats (Patnaik et al. 1981c; Liptak et al. 2004b). In a necropsy study of 12,245 dogs seen over a 15-year period, hepatic carcinoids represented only 14% (15 of 110) of all the primary liver tumors identified (Patnaik et al. 1980). More commonly, carcinoids develop as primary pancreatic or gastrointestinal tract tumors and metastasize to the liver (Patnaik et al. 1980). Since these tumors arise from dispersed cells of the neuroendocrine system, they can be functional in origin. Low-molecular-weight polypeptide or protein hormones such as secretin, cholecystokinin, and chromogranin can be synthesized from the tumors, and when derived from APUD cells, the tumors can secrete serotonin or adrenocorticotrophic hormone (Morrell et al. 2002). In dogs, hepatic carcinoids tend to affect younger animals, compared to other primary hepatic neoplasms. In one study, the median age was 8 years, with 71% of dogs being less than 10 years of age at the time of diagnosis (Patnaik et al. 1980). Although primary hepatic carcinoids have been reported in the gallbladder, they more commonly have intrahepatic origin in the dog (Patnaik et al. 1981c; Liptak et al. 2004b; Morrell et al. 2002; Willard et al. 1988). Unfortunately, due to the typical morphological distribution of these tumors within the liver, surgical therapy is not usually considered beneficial. A diffuse distribution is observed in 67% of cases, and nodular patterns are seen in 33% of cases. The massive morphological distribution (i.e., solitary lobe involvement) is generally not seen (Patnaik et al. 1980, 1981c; Liptak et al. 2004b). Additionally, metastatic disease is identified in 93% of dogs, with locoregional lymph nodes, lungs, and the peritoneum representing the most common sites of extrahepatic spread (Patnaik et al. 1980, 1981c; Liptak et al. 2004b). Mesenchymal Commonly reported types of primary hepatic sarcomas include leiomyosarcoma, hemangiosarcoma, and fibrosarcoma, representing 9%, 3%, and 1% of all primary hepatic malignancies, respectively. Other types of primary hepatic sarcomas include liposarcoma, rhabdomyosarcoma, osteosarcoma, and malignant mesenchyma. In general, however, malignant primary and nonhematopoietic sarcomas are rare in dogs and cats (Patnaik et al. 1980, 1981b; Patnaik 1992; Liptak et al.

Alimentary Tract  207

2004b; Post and Patnaik 1992; Lawrence et al. 1994). Benign primary hepatic mesenchymal tumors, such as hemangioma, have been reported; however, these are also uncommon. Large-scale veterinary studies evaluating prognostic criteria for the surgical management of hepatic sarcomas have not been performed. These tumors usually demonstrate an aggressive biological behavior, with metastasis being identified in 86%–100% of cases (Liptak et al. 2004b; Patnaik et al. 1980; Kapatkin et al. 1992). Likewise, a nodular morphological distribution is seen in 64% of cases, making candidacy for surgery unpredictable. A massive distribution is seen in 36% of cases, and fortunately, diffuse disease is not commonly observed (Liptak et al. 2004b; Patnaik et al. 1980). Surgical treatment for solitary massive hepatic sarcomas has been recommended; however, prognosis is usually poor unless metastatic disease is not a factor. Adjuvant therapies Controlled studies evaluating the efficacy of specific chemotherapeutic regimens for animals with malignant hepatic neoplasia have not been performed (Liptak 2007). Likewise, the therapeutic benefit of radiation therapy has not be investigated scientifically, however, it is unlikely to be efficacious based on the fact that the canine liver is unable to tolerate cumulative doses beyond 30 Gy (Liptak 2007). In humans with metastatic liver tumors or with persistent recurrent disease, chemotherapy is usually recommended; however, no single drug or combination of drugs given systemically leads to a reproducible response rate of more than 25% or adds any survival benefit (Kokudo et al. 2010). Transarterial chemoembolization (TACE) is commonly employed in the treatment of people with advanced-stage or nonresectable liver tumors (primarily HCC) (Liapi et al. 2007). TACE involves selective, percutaneous catheterization (using fluoroscopy) of the desired hepatic artery branch based on tumor location. Angiography is used to define the affected region, and once identified, the vessel targeting the specific tumor bed is accessed. A combination of cisplatin, doxorubicin, and mitomycin C mixed with ethiodol is injected until stasis of blood flow to the region is achieved. Mechanistically, TACE is effective based on the understanding that most malignant hepatic lesions receive their blood supply by the hepatic artery, thereby delivering highly concentrated doses of chemotherapy to the tumor bed and sparing the surrounding hepatic parenchyma (Liapi et al. 2007). A meta-analysis that included seven randomized trials of arterial embolization for unresectable HCC in people showed a statistically significant improvement in 2-year survival compared with control (either

conservative treatment or less favorable therapy, such as intravenous 5-fluorouracil). TACE-treated patients showed a median survival of more than 2 years compared to a median survival of 4 to 7 months in patients with inoperable HCC (Llovet et al. 2003; Liapi et al. 2007). The potential for therapeutic application of TACE in veterinary medicine is vast, and success in the palliation of four dogs with HCC has been reported (Weisse et al. 2002).

Pancreas Pancreatic anatomy The pancreas is divided into right and left lobes. The right lobe is located in the mesoduodenum between the duodenum and ascending colon. The left lobe is contained within the deep leaf of greater omentum between the left kidney, stomach, and transverse colon and is positioned dorsally to the great vessels including the portal vein. The anatomy of the pancreatic duct system differs significantly between the cat and the dog. Digestive secretions enter the duodenum as follows. 1. The accessory pancreatic duct (largest one in dogs; opens at the minor duodenal papilla). Only 20% of cats have an accessory pancreatic duct (Figure 7.16). 2. The pancreatic duct (relatively small and may be absent in dogs, but is the main or only duct in cats; opens at the major duodenal papilla). It most often drains the left lobe of the pancreas (Figure 7.16). Sixty-eight percent of dogs have both an accessory pancreatic duct and a pancreatic duct (Cornell and Fischer 2003). A third duct is present in 8% of dogs. Interlobular ducts converge and enter the duodenum at right angles to the duodenal wall without tunneling. The minor duodenal papilla opens approximately 2 cm aboral to major duodenal papilla, and the common bile duct enters the duodenum at the major duodenal papilla. Extrahepatic biliary obstruction may occur secondary to pancreatic swelling or masses due to impingement of the common bile duct as it enters the major duodenal papilla. When multiple pancreatic ducts exist in the dog, they all communicate with the parenchyma of the gland. It is therefore possible to maintain exocrine secretion from the entire gland with ligation of a single duct. Pancreatic vasculature Celiac arterial vasculature is generally the primary blood supply, with two to three direct branches from the celiac artery to the pancreas commonly found (Figure 7.16). Branches of the splenic artery enter the left limb. The

208  Veterinary Surgical Oncology

Figure 7.16.  Anatomy of the canine pancreatic duct system and vascular system. (Illustration courtesy of Dave Carlson)

celiac artery branches into the hepatic artery to supply the body of the pancreas. The hepatic artery receives blood from the right gastroepiploic artery and continues as the gastroduodenal artery, also supplying the body and the cranial right limb. The gastroduodenal artery receives blood from duodenal arterial branches to become the cranial pancreaticoduodenal artery, supplying the cranial half of the right limb of the pancreas. The cranial mesenteric artery supplies only the caudal portion of the right limb, as the caudal pancreaticoduodenal artery. Damage to the the right cranial and caudal pancreaticoduodenal arteries may lead to duodenal devitalization, because these vessels also supply the duodenum. Surgical procedures and principles Pancreatic biopsy Biopsies may be used to differentiate pancreatitis, fibrosis, pancreatic carcinoma or other tumors, pancreatic abscess, pancreatic pseudocyst, pancreatic nodular hyperplasia, or multiple cystic adenomas. Preoperative

biopsies are usually not performed when endocrine pancreatic tumors are suspected. Incisional biopsy options for pancreatic biopsy include ultrasound-guided needle-core biopsy (TruCut) and open surgical approaches such as wedge, suture fracture technique, or blunt dissection and ligation. These procedures are performed without significant concerns for morbidity as long as special attention is paid to protect the pancreatic vasculature and duct system. A small portion of the caudal aspect of right pancreatic limb is generally biopsied at surgery for diffuse disease, using a small scalpel blade. Crushing of tissues should be avoided where possible, and the pancreas should be handled gently. Laparoscopic biopsy is an additional option that is less invasive than full abdominal exploration and allows the acquisition of excellent tissue samples while minimizing operative morbidity (Barnes et al. 2006). Fine-needle aspiration Fine-needle aspiration (FNA) can also be performed for sample collection of pancreatic abnormalities and is

Alimentary Tract  209

creas should be minimized to avoid postoperative pancreatitis and because identifying abnormal areas of the pancreas such as small tumors can be more difficult, once the pancreas has been inflamed. Indications

Figure 7.17.  Enucleation of a pancreatic mass (insulinoma) using blunt dissection. (Image courtesy of Dr. Simon Kudnig)

generally performed with the assistance of ultrasound guidance. This can be very successful for the diagnosis of exocrine neoplasia (see below), but may be best avoided for cystic pancreatic structures such as abscesses. Partial pancreatectomy Partial pancreatectomy involves the removal of one section of the pancreas and can be performed using either a suture fracture technique or blunt dissection. With any surgical procedure involving the pancreas, it is essential to pay close attention to the pancreatic vasculature and the ductal system. Suture fracture technique is most appropriate for focal lesions near the extremity. Small tumors can be enucleated using blunt dissection (Figure 7.17). A wider margin of grossly normal tissue may be preferable for malignant disease, although this has not been conclusively proven to assist survival or recurrence with a small number of cases in one study (Mehlhaff et al. 1985). Many papers do not specify if surgical removal was via enucleation or partial pancreatectomy, and the use of adjunctive medical management is also inconsistent (see Table 7.5). However, partial pancreatectomy is recommended in preference to enucleation, where feasible. Blood vessels and ducts supplying the portion of pancreas to be removed are identified and ligated. Seventy-five percent to 90% of the pancreas can be removed without any impairment of endocrine or exocrine function if the duct system to remaining portion is left intact. The pancreas has significant regenerative capacity (Cornell and Fischer 2003). Monofilament, synthetic absorbable suture material is used; nonabsorbable, braided, or cat-gut sutures are avoided. Hemoclips or TA-30 or -55 staples can also be used to assist in removal (Bellah 1994). Handling of the pan-

Partial pancreatectomy may be used to treat (and diagnose) pancreatic insulinoma, gastrinoma, glucagonoma, abscess, or pseudocyst. Insulinoma is the most common endocrine tumor of the pancreas, and adenocarcinoma is the most common exocrine tumor of the pancreas (Jubb, 1993). Insulinoma is rare in cats, with only five cases reported (Greene and Bright 2008; Kraje 2003; McMillan et al. 1985, O’Brien et al. 1990; Hawks et al. 1992). Total pancreatectomy Total pancreatectomy is generally avoided in dogs with clinical pancreatic disease due to the difficulty in maintaining duodenal blood supply. Pancreaticoduodenectomy is also avoided due to a high morbidity and mortality. It has, however, been performed experimentally for research involving diabetes mellitus or exocrine pancreatic insufficiency. The primary difficulty associated with total pancreatectomy is maintaining the duodenal blood supply with removal of the right limb of the pancreas. Care must be taken to maintain the primary branch of the splenic artery with removal of the left pancreatic limb or splenectomy will be required (Cornell and Fischer 2003). Regardless of the technique used in clinical cases, generally the amount of edema, adhesion, or fibrosis prevents removal of the right limb of the pancreas while maintaining a viable duodenum (Cornell and Fischer 2003). Pancreaticoduodenectomy is similarly generally not performed clinically because of the high associated morbidity and mortality (Cornell and Fischer 2003). This procedure requires rerouting of the biliary tract and long-term pancreatic supplementation of both endocrine and exocrine functions. Aftercare Prevention of pancreatitis is assisted by intravenous fluid therapy, nothing by mouth for the first 36–48 hours, and withholding oral food for 3–5 days after surgery. Small amounts of oral water are tried first, from 36 to 48 hours after surgery, and if no vomiting is seen, a few teaspoons of low-fat bland food are tried on postoperative day 3. A jejunostomy tube may be needed to bypass the pancreas, and this is often placed preemptively at the initial surgery. Partial or total parenteral nutrition may also be required postoperatively in

210  Veterinary Surgical Oncology

individual animals. Antiemetics are generally given preemptively as part of postoperative management. Imaging of pancreatic neoplasia Specific indications for the examination of the pancreas include, but are not limited to, vomiting, anorexia, weight loss, abdominal pain, icterus, therapy-resistant diabetes mellitus, and hypoglycemia (Hecht and Henry 2007). Ultrasonographic examination of the pancreas Abdominal ultrasound does not always provide a clear image of the pancreas, for example, due to gas or other contents in the stomach, duodenum, or colon (Robben et al. 2005; Garden et al. 2005). Pancreatic tumors may also be small and poorly delineated from the surrounding parenchyma (Iseri et al. 2007). A neoplastic pancreas may appear to be normal on abdominal ultrasound or may mimic or be associated with abscessation, pancreatic necrosis, or pancreatitis (Hecht and Henry 2007). When a pancreatic tumor is identified, it generally appears as a pancreatic or peripancreatic nodule or mass lesion of variable size and echogenicity (Hecht et al. 2007; Lamb et al. 1995; Seaman 2004; Bennett et al. 2001). However, ultrasonographic features in feline pancreatic nodular hyperplasia include pancreatic nodules of up to 1 cm diameter, and there is significant overlap between the ultrasonographic findings for pancreatitis, pancreatic neoplasia, nodular hyperplasia (Hecht and Henry 2007), and multiple cystic adenomas. In a study of 19 cats, a pancreatic mass was detected using either radiography or ultrasonography in 50% of cats with pancreatic neoplasia. In the same study, a single pancreatic nodule or mass exceeding 2 cm in at least one dimension was the only imaging finding unique to pancreatic malignant neoplasms, but was only encountered in 4 of 14 cats. (Hecht et al. 2007). Pancreatic ultrasound should also include an evaluation of the entire peritoneal cavity (Hecht and Henry 2007), to identify abdominal effusion, extrahepatic biliary obstruction, and possible metastatic disease. There are numerous differential diagnoses for hepatic nodules, and cytology or histopathology is required before hepatic metastasis is assumed. In cats with primary pancreatic neoplasia, metastatic liver lesions tend to be single and large, whereas multiple smaller lesions tend to be hepatic nodular hyperplasia. (Hecht et al. 2007). Lamb et al (1995) reported that abdominal ultrasound identified 12 of 16 (75%) pancreatic neoplasms and 6 of 11 (55%) abdominal metastases in dogs.

Radiographic examination of the pancreas Abdominal mass effect and poor serosal detail can be observed in 78% of cats with malignant pancreatic tumors (Hecht et al. 2007) and in 100% of cats with pancreatic adenocarcinoma (Seaman 2004). Barium studies may reveal delayed transit time, duodenal narrowing, or duodenal invasion in cases of pancreatic carcinoma (Withrow 2007a). There are no reports of abdominal radiographs detecting abdominal metastasis of insulinomas (Steiner and Bruyette 1996). Imaging of endocrine pancreatic tumors Insulinoma If the clinical signs and serum glucose and insulin pair support a diagnosis of insulinoma, a transabdominal ultrasound is commonly performed preoperatively to (1) identify pancreatic mass(es) and (2) assess the liver and regional lymph nodes for any evidence of metastasis. Findings should not be overinterpreted, as multiple suspicious masses seen on abdominal ultrasound correlate poorly with metastatic disease at the time of exploratory surgery (Tobin et al. 1999). Abdominal ultrasound visualized a pancreatic nodule in only one of five dogs with insulinoma (due to gas in surrounding bowel), whereas CT diagnosed a pancreatic mass in two of three dogs (Garden et al. 2005). Ultrasound detected 5 of 14 (36%) primary insulinomas, CT detected 10 of 14 (71%) primary insulinomas, and single-photon emission computed tomography (SPECT) with radiolabeled octreotide (a specific form of somatostatin receptor scintigraphy) detected 6 of 14 (43%) primary insulinomas (Robben et al. 2005). Although conventional pre- and postcontrast CT was more sensitive than ultrasound or SPECT in this study, it significantly overestimated metastases (28 falsepositive lesions) (Robben et al. 2005). Despite the low sensitivity of abdominal ultrasound for detecting insulinomas, it is still useful as an initial screening test in the assessment of dogs with hypoglycemia and for the exclusion of differential diagnoses such as insulin-like growth factor II–like peptide-producing extrapancreatic tumors (Boari et al. 1995)(see Table 7.3). Dynamic CT (with contrast medium injection and images taken in arterial and pancreatic phases) clearly identified a pancreatic nodule in one dog with insulinoma, and tumor size and location on CT correlated with surgical findings. The difference between the CT values of the pancreatic mass and those of the normal pancreatic parenchyma was the highest at the arterial phase, similar to humans (Iseri et al. 2007).

Alimentary Tract  211 Table 7.3.  Differential diagnosis for hypoglycemia in dogs. Insulin (iatrogenic overdose, insulinoma—insulin-secreting tumor of pancreatic islet B cells—or other neoplasm secreting insulin-like factor, for example, leiomyosarcoma or lymphoma) Hunting dog hypoglycemia Idiopathic hypoglycemia of neonates and toy breeds Endocrinopathies (e.g., hypoadrenocorticism, growth hormone deficiency, glucagon deficiency (pancreatic disease) Hepatic disease (e.g., portosystemic shunts, necrosis due to toxins and infectious agents, cirrhosis, storage diseases) Chronic renal failure Starvation Pregnancy Sepsis Severe polycythemia Laboratory error

Source:  Adapted from Feldman and Nelson 2004; Cornell and Fischer 2003.

Dual-phase CT angiography (CTA) has been reported in three dogs with histopathologically confirmed pancreatic insulinoma. In all three dogs, there was agreement between the dual-phase CTA findings and the surgical findings, and dual-phase CTA findings identified lesions not seen with abdominal ultrasonography. The arterial and portal phases of the dual-phase study were critical for complete identification of all lesions present (Mai and Caceres 2008). Somatostatin receptor scintigraphy (using the radiolabeled somatostatin analog pentetreotide) showed abnormal foci of pentetreotide activity (attributed in each case to the presence of an insulinoma) in four of five dogs, but only defined the anatomical location of the primary tumor in one of four dogs (Garden et al. 2005). Lester et al. (1999) also reported successful use of somatostatin receptor scintigraphy to assist the presurgical diagnosis of an insulinoma in one dog, where abdominal ultrasound failed to image the pancreas. Nuclear scintigraphy using radioactive labeled octreotide is sensitive for masses as small as 3 mm (Robben et al. 1997). Histopathology is the gold standard for the diagnosis of pancreatic disease (Webb et al. 2008). Laparoscopy is a method of obtaining pancreatic biopsies; however, its use for the diagnosis and staging of canine insulinoma remains to be fully evaluated. Exploratory surgery, pancreatic biopsy, and histopathology are ultimately required in most cases for a definitive diagnosis and accurate staging. Visible metastases are found in

40%–50% of insulinomas at surgery (Leifer et al. 1986; Caywood et al. 1988; Trifonidou et al. 1998). Infusion of methylene blue (3 mg/kg diluted in 250 mL 0.9% sodium chloride and administered over 30–40 minutes), can potentially facilitate identification of nodules. However, this can be harmful through the production of Heinz body hemolytic anemia and can also cause pseudocyanosis and acute renal failure (Siliart and Stambouli 1996). When appropriately diluted to levels that are almost clear to the human eye, methylene blue becomes a moderate-strength fluorophore. After intravenous injection into insulinoma- bearing transgenic mice, primary and metastatic tumors, even when less than 1 mm, were visible intraoperatively under nearinfrared fluorescent light (Joshua et al. 2010). Pancreatic ultrasound has been used to identify small insulinomas intraoperatively (Robben et al. 2005). Endoscopic ultrasound, using a high-frequency (10 MHz) transducer to visualize the pancreas from the adjacent stomach or duodenum, is used in humans and has been investigated in dogs. Good sensitivity and visualization of pancreatic parenchyma (i.e., lobular structure, pancreatic duct, and vessels, except for the tips of the pancreatic limbs) was reported (Morita et al. 1998) and is likely to be useful for detecting small pancreatic lesions. Glucagonoma Most dogs with glucagonoma have metastatic disease at the time of diagnosis, which may be detected by abdominal ultrasound (Oberkirchner et al. 2009; Allenspach et al. 2000). Abdominal ultrasonography was conducted in eight cases of canine glucagonoma, but a pancreatic mass was visible in only one case (Gross et al. 1990; Allenspach et al. 2000; Torres, Caywood, et al. 1997; Torres, Johnson, et al. 1997; Bond et al. 1995; Miller et al. 1991) Contrast-enhanced CT diagnosed a primary pancreatic mass in one dog, where gas-filled loops of intestine had precluded its visualization on abdominal ultrasound (Langer et al. 2003). Gastrinoma Abdominal ultrasound and endoscopy may demonstrate the gastrointestinal ulceration commonly seen with gastrinoma (Zerbe and Washabau 2000). Abdominal ultrasound does not reliably show a pancreatic mass, as they are often small or microscopic (Roche et al. 1982). Somatostatin receptor scintigraphy (using the radiolabeled somatostatin analogs) may assist in diagnosis (Gibril et al. 1996; Schirmer et al. 1995; Altschul et al. 1997).

212  Veterinary Surgical Oncology

Imaging of exocrine neoplasia

Carcinoma

Carcinoma

Metastatic sites reported include the liver, small intestine, lungs, heart, diaphragm and regional lymph nodes in cats (Seaman 2004) and liver, regional lymph nodes, mesentery, intestine, spleen, adrenal gland, diaphragm, lung, and lumbar vertebrae in dogs (Bennett et al. 2001).

As previously mentioned, many nonmalignant diseases such as pancreatic nodular hyperplasia, which is a common incidental finding in old dogs and cats, can mimic pancreatic carcinoma ultrasonographically (Jubb 1993). Therefore, imaging findings need to be related to clinical signs, laboratory data, and ultimately cytology or histopathology to obtain a definitive diagnosis (Hecht and Henry 2007). Ultrasound may reveal metastatic spread to the mesentery (carcinomatosis), which manifests as numerous hypoechoic nodules associated with the connecting peritoneum, often with concurrent abdominal effusion (Hecht and Henry 2007). In one study, ultrasound-guided FNA correctly diagnosed pancreatic carcinoma in 92% of cases in the dog and cat (Bennett et al. 2001). A diagnosis may be obtained from cytology of ascitic fluid in cases where there is a malignant effusion. Radiographically, a lack of serosal detail (4 of 6 cats) and abdominal mass effect (6 of 6 cats) is commonly seen in cats with malignant pancreatic neoplasia (86% of malignancies were carcinomas) (Hecht et al. 2007). The Airedale terrier is reported to be predisposed to pancreatic adenocarcinoma (Priester, 1974). Thoracic imaging for staging purposes Insulinoma The presence of radiographically detectable thoracic metastatic disease has not been reported (Kruth et al. 1982; Caywood et al. 1988; Steiner and Bruyette 1996; Tobin et al. 1999; Polton and Brearley 2007). However, thoracic radiographs can be performed to rule out other causes of hypoglycemia. Glucagonoma Thoracic radiographs may demonstrate the presence of metastasis in cases of glucagonoma (Lurye and Behrend 2001; Brentjens and Saltz 2001; Bailey and Page 2007). Gastrinoma Metastasis is common, with 70%–75% documented in dogs and cats at presentation (Zerbe and Washabau 2000; Feldman and Nelson 2004). The sites for metastasis reported include the liver, regional lymph nodes, spleen, peritoneum, and mesentery (Feldman and Nelson 2004; Zerbe and Washabau 2000). Thoracic radiographs usually do not identify pulmonary metastatic disease; however, they may be used to reveal megaesophagus (secondary to severe esophagitis). Contrast fluoroscopy may show esophageal hypomotility (Kyles 2003).

Surgery for endocrine pancreatic neoplasia Insulinoma Medical management to stabilize blood glucose is required prior to and following surgery (see Chapter 13). For an in-depth discussion on medical management, see Feldman and Nelson 2004. Surgical exploration is recommended if there is persistently low blood glucose, appropriate clinical signs, and a high insulin level, even if a pancreatic mass is not identified with abdominal ultrasound (Tobin et al. 1999). The optimal treatment for insulinoma is partial pancreatectomy to remove the primary tumor, and if possible, metastatic lesions such as enlarged lymph nodes (Figure 7.18). Greater than 85% of primary insulinomas are single nodules (Steiner and Bruyette 1996). Multiple intrapancreatic masses have been reported in approximately 15% of cases (Mehlhaff et al. 1985; Caywood et al. 1988). The pancreas is gently palpated for small nodules (less than 1.0–1.3 cm diameter); however, up to 20% are not palpable (Steiner and Bruyette 1996). If a nodule is identified, it is removed, paying close attention to the pancreatic vasculature and ductal pattern (Figure 7.18). Diffusely infiltrative lesions diagnosed only by partial pancreatectomy and histopathology have also been described (Kruth et al. 1982; Feldman and Nelson 2004). Random removal of one limb of the pancreas is not recommended because unidentified tumors are most often located in the body and no single limb is more commonly affected than another (Siliart and Stambouli 1996). Feldman and Nelson (2004) recommended against removal of insulinomas from the body (central section) of the pancreas due to concerns for creating severe, life-threatening pancreatitis. Caywood et al. (1988), Tobin et al. (1999), and Trifonidou et al. (1998) did not find any correlation between tumor location and prognosis. Metastases are present in up to 51% of dogs at the time of initial surgery (Steiner and Bruyette 1996). Metastasis is most common to the liver and regional lymph nodes (see Table 7.4 and Figure 7.18), but also to the mesentery, omentum, and duodenum, and has not been reported in the lung (Mehlhaff et al. 1985; Caywood et al. 1988). Multiple suspicious masses seen on abdominal ultrasound correlate poorly with metastatic disease at time of exploratory surgery (Tobin et al. 1999).

Alimentary Tract  213 Table 7.4.  Staging for insulinoma. Stage I II III

(a)

(b)

Confined to pancreas Pancreas and regional lymph nodes Distant metastasis (i.e. liver)

Enlarged lymph nodes seen at surgery are also not always due to metastatic disease, and it is therefore not recommended to euthanize animals based on gross lymph node enlargement alone. In one study, only three of even dogs with ultrasound evidence of hepatic metastasis had confirmed β-cell metastasis (Mehlhaff et al. 1985). During surgery, regional lymph nodes should be removed for staging purposes if they are visible or enlarged, and a liver biopsy should also be taken for staging purposes. Pancreatic blood supply should be preserved, and the pancreas handled gently to minimize postoperative pancreatitis. In selected patients, debulking of gross disease by enucleation of larger liver nodules may be a feasible option, but in many dogs liver metastasis is diffuse and small, making surgical debulking very difficult. Studies have shown that survival can be prolonged in dogs that receive tumor debulking and medical management compared to dogs receiving medical treatment only (Polton and Brearley 2007; Tobin et al. 1999). Close monitoring of blood glucose is essential with this treatment regime (Siliart and Stambouli 1996). Euthanasia of dogs with stage III disease at surgery is potentially unwarranted, because even dogs with widespread metastasis have been managed up to 1 year with a combination of medical and surgical management (Polton and Brearley 2007). However, if postoperative clinical signs associated with hypoglycemia are not controlled, or other serious postoperative complications are not controlled with medical management, euthanasia is then appropriate. Potential complications of surgery for insulinoma

(c)

Figure 7.18.  (A) Insulinoma within the body of the pancreas (black arrow) and enlarged adjacent regional lymph node (white arrow). (B) Excised insulinoma via partial pancreatectomy (black arrow) and excised enlarged regional lymph node (white arrow). (C) Large insulinoma within the body of the pancreas (black arrows) and metastatic liver nodule (white arrow).

Intraoperative complications. The inability to find a pancreatic nodule is reported in up to 20% of insulinoma cases (Siliart and Stambouli 1996). The detection of a nodule can be facilitated by the use of intraoperative ultrasound, partial pancreatectomy, or methylene blue injection. Elie and Zebra (1995) reported that 3% of dogs will have diffuse pancreatic insulinoma without a specific mass, and no predisposition for tumor location within the pancreas (left limb, body, or right limb) has been identified (Caywood et al. 1988). Other intraoperative complications include the identification of

214  Veterinary Surgical Oncology

previously unseen metastatic disease, the identification of multiple pancreatic tumors (found in up to 15% of cases) (Tobin et al. 1999), and hemorrhage due to disruption of pancreatic or duodenal blood supply, with the potential to cause duodenal or pancreatic devitalization (see above). Postoperative complications. Postoperative complications following partial pancreatectomy for insulinoma are common. A 12% (3 of 26) perioperative mortality rate has been reported after partial pancreatectomy, due to pancreatitis or sepsis and diabetic ketoacidosis (Tobin et al. 1999). Mehlhaff et al. (1985) and Trifonidou et al. (1998) reported low (8.7% and 10%, respectively) postoperative mortality rates after partial pancreatectomy or enucleation. Leifer et al. 1986 reported a mortality rate of 3 of 40 dogs (7.5%) due to pancreatitis, cardiac arrest, and sepsis. Postoperative pancreatitis occurs in up to 10%–43% dogs, particularly following resection of tumors located in the head of the pancreas (Tobin et al. 1999; Feldman and Nelson 2004; Mehlhaff et al. 1985; Trifonidou et al. 1998). Approximately 15%–26% of dogs remain hypoglycemic following surgery (Mehlhaff et al. 1985; Caywood et al. 1988; Trifonidou et al. 1998; Tobin et al. 1999) due to inoperable disease or metastatic disease. Recurrent hypoglycemia and the associated recurrence of clinical signs occurs in 52%–100% dogs (Tobin et al. 1999; Kyles 2003) with the mean time to recurrence and initiation of prednisone treatment being 60 days (Tobin et al. 1999). When hypoglycemia returns, medical management, repeat exploratory surgery, or euthanasia are possible options (Mehlhaff et al. 1985). Diabetes mellitus occurs in 8%–35% dogs due to the prolonged hyperinsulinemia and hypoglycemia resulting in atrophy of normal islet tissue, exacerbated by partial pancreatectomy, further decreasing insulin reserves (Kyles 2003). Hyperglycemia is usually transient, but can be permanent, and these animals may require long-term insulin therapy. Pancreatic abscesses or pseudocyst formation can occur subsequent to postoperative pancreatitis. Fatal gastric dilation-volvulus has also been reported as a postoperative complication in three dogs (Leifer et al. 1986; Mehlhaff et al. 1985). As such, prophylactic gastropexy may be worth considering at the time of initial surgery in predisposed breeds. Recurrent severe hypoglycemia can result in irreversible brain damage and persistent seizures (Feldman and Nelson 2004; Mehlhaff et al. 1985), as well as a paraneoplastic peripheral neuropathy (see Chapter 13).

months for stage II and III disease treated with surgery. Steiner and Bruyette (1996), reported a mean survival time with surgical treatment of 11.5 months for 114 dogs. The overall prognosis is usually guarded as a surgical cure is unlikely to be achieved. However, when insulinoma dogs previously treated with partial pancreatectomy showed relapse of hypoglycemia, treatment with oral prednisolone as adjunctive medical management resulted in a MST of 1,316 days (Polton and Brearley 2007). Poor prognostic factors that are reported with insulinoma include the following.

• Conservative treatment. MST is 74 days with conser• • •

• •

• •

•

vative treatment versus 381 days for partial pancreatectomy (Tobin et al. 1999). Age. Survival time is significantly decreased in younger dogs (Caywood et al. 1988). Serum insulin levels. High preoperative serum insulin levels indicates poorer prognosis (Caywood et al. 1988). Higher stage of disease. Stage III insulinomas have MST less than 6 months versus 18 months for stage I and II disease. Eighty percent of dogs are alive at 14 months when the disease is confined to the pancreas, whereas less than 20% of dogs with metastasis are alive at 12 months (Caywood et al. 1988). Persistent postoperative hypoglycemia. MST is 90 days versus 680 days for normoglycemic dogs (Trifonidou et al. 1998). Clinical stage. Clinical stage influences the duration of normoglycemia following surgical resection. Dogs with stage I insulinomas maintain normoglycemia for a median of 14 months versus 1 month for dogs with stage II and III disease (Caywood et al. 1988). Tumors with a high mitotic count. These tumors are thought to carry a worse prognosis, although only 11 cases were reported (Dunn et al. 1993). Enucleation. Mehlhaff et al. (1985) reported that enucleation (10 dogs) was associated with a shorter mean survival time than partial pancreatectomy (15 dogs; 11.5 months versus 17.9 months); however, the number of cases reported was small (see Table 7.5). Paraneoplastic peripheral neuropathy. Dogs with concurrent paraneoplastic peripheral neuropathy or brain damage from chronic hypoglycemia have a guarded prognosis for neurological recovery (Mehlhaff et al. 1985; Kyles 2003), although improvement in peripheral neuropathy is possible with medical or surgical treatment (Kyles 2003).

Prognosis for insulinoma treated with surgery

Gastrinoma

Caywood et al. (1988) reported a median survival time (MST) of 18 months for stage I disease, and a MST of 6

Gastrointestinal ulceration (with or without perforation) is common and occurs in 80% of cats and dogs

Alimentary Tract  215 Table 7.5.  Reported survival times for dogs with insulinoma treated with surgery ± medical therapy Reference

Number of cases

Survival time (ST)

Surgical technique

Adjunctive medical managementa

Unspecified in 23 dogs, partial pancreatectomy specified in 2 dogs Local enucleation Partial pancreatectomy Unspecified

Yes

Partial pancreatectomy

Yes

Partial pancreatectomy

Yes

Partial pancreatectomy

Yes

Unspecified

No

Kruth et al. 1982

25

12.3 months (mean ST)

Mehlhaff et al. 1985 Mehlhaff et al. 1985 Leifer et al. 1986

10 15 18

Caywood et al. 1988

47

Tobin et al. 1999

26

Polton and Brearley 2007

28

Trifonidou et al. 1998

31

11.5 months (mean ST) 17.9 months (mean ST) 435 days (14.5 months) (median ST) 18 months (stage 1 data) (median ST) 381 days (12.7 months) (median ST) 547 days (18.2 months) (median ST) 258 days (8.6 months) (median ST)

Yes Yes No

a Medical management options reported included frequent feeding, oral glucose, oral glucocorticoids, diazoxide. ST, survival time.

with gastrinomas (Feldman and Nelson 2004; Zerbe and Washabau 2000; Altschul et al. 1997; Simpson and Dykes 1997; Green and Gartrell 1997; Brooks and Watson 1997). Part of the surgical management should involve resection of gastrointestinal ulceration/perforation with treatment of the resultant intra-abdominal sepsis. Gastrinomas are usually solitary, with 60% in the right lobe, 40% in the pancreatic body, and rare involvement of the left lobe (Zerbe and Washabau 2000). Partial pancreatectomy of the right limb can be performed if the tumor is not found due to a high percentage of right limb involvement (Feldman and Nelson 2004). Debulking of disease can be palliative (Zerbe and Washabau 2000; Kyles 2003). Glucagonoma Skin biopsies are required to confirm superficial necrolytic dermatitis (SND), although the diagnosis of SND does not confirm the presence of glucagonoma (Langer et al. 2003). Complete surgical resection of the glucagonsecreting tumor is the treatment of choice. Biopsies of lymph nodes and the liver are taken for staging purposes (as for insulinoma). Palliation of clinical signs can be achieved by surgical debulking of tumor burden (Langer et al. 2003). Surgery for exocrine tumors Carcinoma is the most common tumor of the exocrine pancreas in dogs and cats (Jubb 1993) and is more

common than insulinomas. Overall, however, they are very uncommon (less than 0.5% of all tumors) (Withrow 2007a). These tumors originate from either acinar cells or ductal cells (Withrow 2007a). The clinical features can be difficult to differentiate from pancreatitis in many cases. A palpable abdominal mass is common in cats but uncommon in dogs (Withrow 2007a). These tumors have often metastasized before the appearance of clinical signs and are therefore usually at an advanced stage when diagnosed (very extensive disease locally and with metastasis). Abdominal effusion is common, and carcinomatosis (metastatic spread to the mesentery) is seen occasionally (Hecht et al. 2007; Hecht and Henry 2007). The prognosis is extremely poor because of their aggressive nature and resistance to chemotherapy. Complete pancreatectomy has been reported to be successful in 78 of 80 normal experimental dogs (Cobb and Merrell 1984). However, the preservation of duodenal blood flow in the presence of pancreatic disease is much more difficult (Cornell and Fischer 2003). In humans, complete pancreatectomy and pancreati­ coduodenectomy (Whipple procedure) have a 5%–30% operative mortality rate (Withrow 2007a). Gastrojejunostomy can be performed as a palliative procedure bypassing an obstructed duodenum (Withrow 2007a). Cholecystoduodenosotomy/cholecystojejunostomy may also be a palliative option if extramural biliary obstruction exists. However, heroic surgical procedures are generally not recommended due to the universally poor

216  Veterinary Surgical Oncology

prognosis (due to a metastatic rate and the anatomical location of the tumor) and associated high morbidity and mortality (Cornell and Fischer 2003). One-year survival has not been reported, and radiation or chemotherapy has been shown to be of limited benefit in people and animals (Withrow 2007a). Other exocrine tumor types include adenoma, lymphosarcoma, squamous cell carcinoma, lymphangiosarcoma, and spindle cell sarcoma (Andrews 1987; Münster and Reusch 1988; Hecht et al. 2007). Panniculitis, polyarthritis, and osteomyelitis have been reported in association with exocrine pancreatic tumors in two dogs (Gear et al. 2006). In both cases, there was no response to medical management, resulting in euthanasia. Postmortem examination revealed a pancreatic exocrine adenoma in one dog and a pancreatic adenocarcinoma with widespread metastases in the other.

Small Intestine Diagnostic workup and biopsy techniques Radiography, contrast studies, and ultrasonography are commonly employed imaging modalities used in the diagnosis and staging of animals with suspected intestinal neoplasms (Paoloni et al. 2003; Penninck et al. 2003). Based on the high incidence of systemic metastasis, radiographic screening of the thorax via a three-view metastatic series is indicated in any animal suspected of an intestinal neoplasm. Standard two-view abdominal radiographs may identify intestinal gas dilation oral to a soft-tissue density, supporting a diagnosis of an intestinal mass that is causing obstruction of the bowel. Additional radiographic findings can include the presence of free gas or fluid within the peritoneum, indicating a potential perforation of the intestinal tract. In one study, 32% (8 of 25) of animals with spontaneous (nontraumatic) pneumoperitoneum had gastrointestinal tract rupture attributable to neoplasia (Saunders and Tobias 2003). Contrast radiographic studies may provide additional information in the detection of annular or intraluminal lesions within the intestinal tract because they more precisely outline narrowing of the lumen at the site of tumor development (Paoloni et al. 2003). Contrast studies may be contraindicated if there is concern for a ruptured intestinal mass. Ultrasonography has been described as the most effective and least invasive diagnostic modality available in small animals to detect gastrointestinal tumors (Paoloni et al. 2003). Abdominal ultrasound can be particularly useful in differentiating neoplasia from other commonly encountered intestinal diseases such an

enteritis (Penninck et al. 2003). In the study by Penninck et al., loss of ultrasonographically detectable intestinal wall layering was predictive for dogs with intestinal tumors. When loss of wall layering was documented, dogs were 50.9 times more likely to have an intestinal tumor compared to inflammatory conditions within the bowel (Penninck et al. 2003). Another distinct benefit of ultrasound is that it allows for sampling of an intestinal mass with FNA, as well as visualization of the abdomen as a whole in the evaluation for regional metastasis. Advanced imaging with CT and MRI can be used in the staging of dogs with intestinal neoplasia; however, studies evaluating the clinical efficacy of this modality in veterinary medicine are lacking. Three tissue sampling options exist for small intestinal neoplasms prior to full surgical abdominal exploration. These include FNA, which generally requires ultrasound guidance, endoscopic mucosal biopsies (limited to the duodenum), and laparoscopic assisted biopsies. FNA cytology has an approximate 95% specificity and 70% sensitivity when compared with histological diagnoses for intestinal lesions (Bonfanti et al. 2006). Intraoperative cytological impression smears are more likely to agree with histology than FNA. Surgical impression smears are able to facilitate intraoperative decision making but may not be able to provide a definitive diagnosis prior to an invasive surgical procedure. FNA has a 71% sensitivity for GI lymphoma and 44% sensitivity for smooth muscle tumors as compared to histopathology (Bonfanti et al. 2006). Endoscopy is an important diagnostic tool used for staging the dog or cat with suspected intestinal neoplasia and has been used in clinical practice for the past 20 years (Willard et al. 2001). Despite this, limitations of this modality must be recognized as deficiencies in technique, instrumentation, and variability between pathological interpretations that can yield inaccurate or uninterpretable results. In two studies, the range of nondiagnostic samples submitted for histopathological assessment was 19%–46% in dogs and 21%–50% in cats, respectively (Willard et al. 2001; Van der Gaag and Happe 1990). As a result, it is generally recommended that a minimum of eight individual tissue pieces be submitted when performing endoscopic biopsies of the duodenum in dogs and cats (Willard, et al. 2001). Differentiation between inflammatory bowel disease and intestinal lymphoma (LSA) is essential prior to the initiation of therapy for cats and dogs. Endoscopic biopsy samples have a limited ability to differentiate between these two disease processes in cats. In the study by Evans et al. (2006), endoscopic duodenal biopsies confirmed a diagnosis of LSA in only 33% (3 of 9) of

Alimentary Tract  217

cats that had biopsy-confirmed LSA in that region of the bowel. Since the majority of gastrointestinal LSA cases are associated with the small intestine (jejunum and ileum) in cats, full-thickness biopsies are recommended when clinical signs consistent with GI lymphoma are present within this species (Evans et al. 2006). Laparoscopic-assisted biopsy is another useful technique for the diagnosis of small intestinal neoplasia prior to a full abdominal exploration (Evans et al. 2006; Barnes et al. 2006; Freeman 2009). A standard threeportal laparoscopic technique is used, with one portal placed caudal to the umbilicus and two additional portals placed in paramedian locations, lateral to the third mammary glands. Laparoscopy allows good visualization for full abdominal exploration. Laparoscopic samples can be collected from parenchymal organs such as the liver, spleen, adrenal glands, and pancreas. Laparoscopic-assisted samples can be obtained from the stomach, small intestine, and urinary bladder. Small intestinal and gastric samples are obtained by exteriorizing the tissues to be biopsied through a 4 cm ventral midline incision just cranial to the umbilicus. Samples can then be collected using standard antimesenteric sampling techniques or with the aid of a harmonic scalpel. There is no significant difference in the sample quality between tissues collected with the standard open and laparoscopic-assisted techniques (Barnes et al. 2006; Freeman 2009). Technical aspects of surgical procedure Anatomy A significant amount of redundancy exists within the small intestine. The overall intestinal length is relative to animal size and is approximately five times the trunk length in dogs and cats. The small intestine is about four times the length of the large intestine (Grandage 2003). The jejunum is freely moveable and easily manipulated for resection procedures. The duodenum, on the other hand, has several anatomical constraints that potentially complicate surgical procedures associated with this section of the small intestine. Surgical resections associated with the proximal duodenum are especially challenging because of the association of the pancreatic and biliary systems and particularly the entry of their respective duct systems into the proximal duodenum. The presence of the body and right limb of the pancreas in the mesoduodenum also significantly complicates any proposed surgical resections involving the duodenum. The descending duodenum is also relatively immobile because of regional anatomical constraints such as the duodenocolic ligament (Grandage 2003).

The arterial blood supply to the small intestine is primarily derived from branches of the cranial mesenteric artery. Branches of this main aortic branch anastomose with the celiac and caudal mesenteric arterial braches in the most oral and aboral portions of the small intestine, respectively. The jejunum is entirely supplied by cranial mesenteric arterial branches (Grandage 2003). The venous return from the small intestine is through the portal vein. The portal vein is composed of four visceral branches in the dog. These include the cranial mesenteric, splenic, caudal mesenteric, and the gastroduodenal branches (Grandage 2003). The majority of the small intestine is drained by the cranial mesenteric branch. The pylorus, duodenum, and right limb of the pancreas are drained by the gastroduodenal and the large intestine is drained by the caudal mesenteric branch of the portal vein (Grandage 2003). Lymphatics from the duodenum and jejunum drain into paired hepatic lymph nodes (adjacent to the portal vein) and the pancreaticoduodenal nodes at the origin of jejunal arteries (Grandage 2003). Lymphatics from the ileum drain into multiple colic lymph nodes and jejunal nodes (Bezuidenhout 1993). For staging purposes, mesenteric lymph nodes should be biopsied with any lesion of the small intestine, regardless of size (Crawshaw, et al. 1998). For any surgical procedure involving the gastrointestinal tract it is highly advised to follow the Halstedian principles of uncomplicated tissue healing (Webster 1955). These include minimization of tension, maintenance of blood supply to the surgical site, minimization of contamination, and gentle tissue handling (Webster 1955). Repeated moistening of visceral tissues is necessary to minimize desiccation and resultant inflammation and adhesion formation. Tension is generally not a concern with surgical procedures involving the jejunum because it is freely moveable. Tension, however, can be an issue with procedures associated with the duodenum because of its relatively fixed location in the right paralumbar region. The area of the proximal duodenum must be respected at all times. This is an extremely difficult area for resection procedures for several reasons. First, special attention must be paid to the location of the major duodenal papilla because of the entry of the bile duct and pancreatic ducts in both dogs and cats. In cats, the pancreatic duct entering the major duodenal papilla is the only source of pancreatic drainage in 80% of animals (Grandage 2003). It is therefore essential to prevent obstruction with surgical procedures in this species. If the major duodenal papilla is compromised in cats, then long-term exocrine pancreatic enzyme supplementation

218  Veterinary Surgical Oncology

is required. In dogs, the minor duodenal papilla is most often the primary entry point for pancreatic secretions; therefore, the major duodenal papilla can be excised and normal exocrine pancreatic function will be maintained in most dogs (Grandage 2003). Because of the close association of the proximal duodenum with the body and right limb of the pancreas and their shared blood supply, excision of the duodenum without significant compromise of the pancreas is tedious and risky. Technical options Segmental resection of the small intestine is performed in a standard manner, whereas there are several surgical options for intestinal anastomosis. If the affected area of small intestine involves a loop of jejunum, then there are no anatomical constraints associated with the procedure. The affected area of small intestine is isolated and packed off from the remainder of the abdomen with laparotomy sponges. The vasculature to the area of intestine to be removed is identified and ligated using hemostatic clips, electrothermal coagulation, or encircling ligatures. The mesentery is then incised approximately perpendicular to the long axis of the bowel, leaving as much mesentery intact as possible to facilitate closure of the mesenteric defect at the conclusion of the procedure. Intestinal contents are then milked away from the area of transection. Noncrushing (i.e., Doyen) tissue clamps are placed approximately 1–2 cm away from the incision line on the tissue that will remain following anastomosis. The gross margins of the neoplasm are identified, and transection is performed with 5 cm margins of normal intestine on either end of the mass (Crawshaw et al. 1998). The cut edges of the remaining intestine are then placed in close proximity to one another and the anastomosis performed. Several options exist for anastomosis of the intestine following the resection procedure. Hand-sewn anastomoses are most commonly performed. They are generally straightforward and require no specialized instrumentation. One-layer appositional suture patterns are recommended for small intestinal anastomoses. This can be performed using either a simple interrupted or continuous pattern. Simple interrupted intestinal anastomosis For the simple interrupted hand-sewn technique, the mesenteric suture is placed first, followed by a single suture placed at 180 degrees from the first suture (placed along the antimesenteric border). Two additional sutures are then placed at the midpoint between the mesenteric and antimesenteric sutures. Additional sutures are then equally spaced out along the circumference of the intestine at 2–3 mm intervals and 2–3 mm from the cut edge

of the intestine. Once all sutures have been placed, the intestinal lumen can be occluded with the opposed fingers of an assistant 4–5 cm away from the anastomosis, and the surgical site is leak tested using sterile saline. Saline is injected into the antimesenteric surface of the intestinal lumen using a 25-ga needle and an appropriately sized syringe. The injection is continued until modest pressure is created at the anastomosis. If additional pressure is needed, the area between the finger occlusion can be carefully pressed between the index finger and thumb of the surgeon. Special attention is paid to the mesenteric border as this is the most common area of leakage because of poor visualization associated with mesenteric fat deposits. Careful dissection of the mesenteric fat along the mesentery can be performed in order to improve the precision of suture placement in this region. It should be noted, however, that excessive intralumenal distension, beyond the bursting strength of the anastomosis, can result in leakage. Once the anastomosis is completed, the mesenteric defect is then closed using a monofilament absorbable suture material. During mesenteric closure, iatrogenic trauma to the segmental intestinal vasculature must be avoided. For the simple continuous sutured anastomosis, sutures are placed at the mesenteric and antimesenteric margins as described for the simple interrupted pattern. The free suture tag is, however, left long so that it can be tied to the advancing continuous pattern from the suture placed at 180 degrees from its location. Full-thickness suture bites are then taken at 2–3 mm intervals until one-half of the intestinal lumen has been closed. Once the continuous pattern has advanced to the level of the previously placed suture at 180 degrees from its origin, the suture is tied to the long suture tag. This process is continued for both sides of the intestinal lumen. Following completion of the anastomosis, leak testing can be performed as described for the simple interrupted pattern (Weisman et al. 1999). As an alternative to the hand-sewn technique, the intestinal anastomosis can be performed using 4.8 × 3.4 mm surgical skin staples. The primary advantage to this stapling technique is the rapidity with which the procedure can be performed. Stapled small intestinal anastomoses are performed in approximately oneseventh the time as simple interrupted hand-sewn anastomoses (Coolman et al. 2000b). The technique begins with the placement of three stay sutures placed at equal distances, beginning at the antimesenteric margin. An assistant creates tension between adjacent stay sutures, and staples are placed at 3 mm intervals for a total of 12–15 staples. No differences are reported in complication rates between hand-sewn and skin staple anastomoses (Coolman et al. 2000b).

Alimentary Tract  219

(a)

(b)

(c)

(d)

Figure 7.19.  (A) Intraoperative image of a dog undergoing a functional end-to-end anastomosis of the jejunum secondary to an intestinal adenocarcinoma. The mass has been resected with 5 cm margins, and the luminal ends have been occluded with atraumatic tissue clamps. The antimesenteric intestinal borders are apposed in preparation for insertion of the GIA stapling device. (B) The GIA stapling device has been inserted into the apposed intestinal segments. Preplaced stay sutures facilitate proper alignment of the bowel in preparation for discharging the stapling device. (C) A TA stapling device is then positioned across the open stoma and fired to complete the functional end-to-end anastomosis. Take care to ensure that all tissue layers of both of the terminal ends of the stoma re included in the staple line. (D) Intraoperative image of the completed stapled anastomosis. Anchoring suture(s) must be placed across the base of the anastomosis (black arrowheads) in order to reinforce the staple line. This region is prone to tension, which can result in disruption of the staple line. (Images courtesy of Dr. Pam Schwartz)

An additional alternative to the end-to-end appositional anastomosis is a functional side-to-side anastomosis, which is performed using GIA and TA stapling devices. The resection procedure is performed as for any small intestinal anastomosis. Once resection is complete, stay sutures are placed in the mesenteric border of the transected intestinal segments and the antimesenteric

borders of the two cut ends of the remaining intestines are placed adjacent to one another in a parallel position (Figure 7.19A). The GIA stapler is advanced into the lumen of each small intestinal orifice (Figure 7.19B). The stapler is then fired, which creates a large stapled side-to-side anastomosis between the two lumens. At this point, a common lumen has been created between

220  Veterinary Surgical Oncology

the two intestinal loops with the free end remaining open. The GIA stapler is then removed and an appropriately sized TA stapler (TA-55 with blue cartridge usually suffices) is used to close the free end of the common lumen (Figure 7.19C). An anchoring suture is always placed at the end of the anastomotic staple line because this region is under the greatest amount of tension (Figure 7.19D). The mesenteric defect is then closed. Leak testing can then be performed using sterile saline as previously described (Ullman et al. 1991). Occasionally, significant luminal size disparity is created between the oral and aboral portions of the small intestine, especially if the intestinal mass has created a chronic partial obstruction or an anastomosis is being performed between the small intestine and the colon. If this is the case, then several surgical maneuvers are available to address luminal disparity. If small disparities exist, then differential spacing of sutures can be performed between the larger and smaller lumens. Smaller bites are taken between sutures of the smaller lumen than between bites for the larger lumen. Transecting the smaller lumen bowel at an angle away from the mesenteric margin can also be used to rectify luminal disparities. More tissue is removed from the antimesenteric border than the mesenteric border. If larger disparities exist, then differential suturing can be combined with spatulation of the smaller lumen. Spatulation is accomplished by making a longitudinal incision along the antimesenteric border of intestinal lumen. Finally, luminal disparities can be eliminated by partial closure of the antimesenteric border of the larger intestinal segment, followed by routine completion of the handsewn anastomosis. Anastomosis augmentation techniques Surgical site reenforcement is regularly performed to promote rapid and uncomplicated healing. The two most commonly performed procedures are omental and serosal patching. The omentum promotes healing by providing a source of additional blood flow and lymphatic drainage (Hosgood 1990). Increase in blood flow to the surgical site helps to control infection if compromised blood flow results from the disease process or surgical procedure. Following completion of the intestinal anastomosis, the omentum is wrapped around the surgical site. Several partial thickness sutures may be used to facilitate adherence of the omentum to affected region of bowel. Lengthening of the omentum can be performed based on the right or left gastroepiploic artery/vein (Hosgood 1990). Alternatively, an omental pedicle extension technique can be used; however, this is rarely indicated

during routine intestinal procedures for neoplasia (Ross and Pardo 1993). Serosal patching is another useful technique for reinforcement of surgical incisions involving the gastrointestinal tract. Serosal patching involves the mobilization of a portion of the freely moveable jejunum. The procedure generally involves placement of two separate loops of jejunum directly over the surgical site. Either a simple interrupted or continuous suture pattern using 3-0 or 4-0 absorbable monofilament suture material is placed between the seromuscular layers of the two pieces of intestine directly over the surgical site. This is repeated until two sections of normal jejunum have been sutured over the surgical site (Crowe 1984). Aftercare The intensity of postoperative management will vary greatly based on the preoperative status and intraoperative (i.e., septic peritonitis secondary to ruptured intestinal mass) findings within the clinical patient. Aftercare involves attention to several different facets of the patient’s status. Consideration of the patient’s preoperative nutritional status should be used to determine whether or not supplemental enteral nutrition would be required in the postoperative setting. Indications for immediate assisted (enteral) feeding because of malnutrition include prolonged anorexia (i.e., longer than 48 hours), weight loss of greater than 10% of body weight, inadequate muscle mass, and low albumin concentration (i.e., less than 2.5 g/dL in cats or less than 2.1 g/dL in dogs) (Rasmussen 2003). Animals suffering from septic peritonitis as a result of a ruptured intestinal mass should also be considered ideal candidates for intraoperative feeding tube placement. This author generally prefers gastrojejunostomy (GJ) feeding tubes in cases of extensive bowel resection or in patients with peritonitis. The benefit of GJ tube feeding includes dual lumen access to both the stomach and small intestine, yet operative morbidity is minimized since only a gastric incision is required for tube placement (Cavanaugh et al. 2008). Gastroparesis is common after intestinal surgery, and gastric decompression in the early postoperative period improves patient comfort and facilitates a reduction in nausea and vomiting, which if not alleviated can result in aspiration pneumonia in severely debilitated animals. If significant impairment in gastric motility is present, GJ tubes allow feeding to commence through the jejunostomy component (J tube) of the tube system while the gastric dysfunction is monitored (gastric residual fluid volume is quantified every 4–6 hours by aspirating the gastric component of the tube system) and managed (motility

Alimentary Tract  221

medication is titrated based on trends in calculated residual volumes). With the GJ-tube system, J-tube feedings are generally initiated within 12 hours of surgery using a continuous rate infusion of a commercially available liquefied diet. The required energy requirements (resting energy requirement [RER] = 30 × B.W. [kg] + 70) are calculated and then initiated at 25% of the patient’s need. Supplementation rates are generally doubled every 24 hours if the patient is tolerating feeding. We caution exceeding 75% of the RER through the J tube as this will commonly precipitate vomiting due to an inability to handle this volume of fluid through the small intestine. Solitary placement of gastric or enterostomy feeding tubes can also be extremely useful in the management of animals recovering from treatment of an intestinal neoplasm. Clinical variables such as the procedure to be performed, anticipated outcome, and preoperative status should be assessed to decide which type of feeding tube would ultimately meet the needs of the individual patient. A comprehensive description of the surgical technique for placement of these feeding tubes can be found in the gastric neoplasia section of this text. Postoperative monitoring of serum biochemical and objective clinical parameters is important to ensure that recovery is progressing uneventfully. Animals undergoing elective treatment of an intestinal mass will generally need nothing more than supportive care (i.e., maintenance fluid therapy and appropriate analgesia) for 24–48 hours after surgery. On the contrary, animals presenting with septic peritonitis or with signs consistent with intestinal obstruction may need intensive monitoring and treatment in the postoperative period. Daily evaluation of serum electrolytes and renal values allow objective interpretation of the animal’s hydration status. Monitoring of blood albumin and total protein levels will be useful in assessing the need for colloidal supplementation in the form of intravenous hydroxyethyl starch therapy (Hespan; B. Braun Medical Inc. Melsungen, Germany). Severe protein deficiencies may require transfusions of human or canine albumin in order to control clinical signs and facilitate wound healing while protein deficiencies are replaced through enteral nutritional supplementation. Although controversial, clinical studies have found that systolic blood pressure, serum albumin, and total solid levels are significantly increased with albumin administration and that human albumin is safe to be administered to dogs (Mathews and Barry 2005; Trow et al. 2008). Owners should always be counseled of the risks of albumin transfusion, however, as serious hypersensitivity (acute and delayed) reactions have been

reported in both healthy and hypoalbuminemic dogs (Francis et al. 2007; Cohn et al. 2007; Yamaya et al. 2004). As an alternate to albumin, some clinicians will use plasma transfusions to combat hypoperfusion and hypoalbuminemia. It should be known, however, that the effects of plasma administration on serum albumin and colloid osmotic pressure in critically ill dogs has not been reported, and the cost to effectively increase serum albumin in the clinical patient may be cost prohibitive, as a plasma dosage of 22.5 mL/kg may be required to increase serum albumin by 0.5 g/dL (Trow et al. 2008; Mazzaferro et al. 2002). For most intestinal procedures, a first- or secondgeneration cephalosporin is prescribed perioperatively; however, antibiotics are not indicated in the postoperative period unless active infection is identified at the time of the surgical procedure. Daily monitoring of blood glucose is useful to establish trends toward hypoglycemia. If hypoglycemia is present, concern for intestinal dehiscence should be raised as septicemia commonly results in this finding. Trends in the patient’s body temperature, abdominal comfort, and intraabdominal fluid volume are also useful to assess healing because pyrexia, pain, and an increase in fluid volume could be indicative of intestinal-mediated surgical complication. Shifts in the immature white blood cell lines (development of a degenerative left shift) within the first 2–5 days after surgery can also be an early indicator of intestinal dehiscence precipitating septic peritonitis. Functional outcome—potential complications Dehiscence Dehiscence, resulting in septic peritonitis, is the most significant postoperative complication associated with small intestinal resection procedures. Dehiscence generally occurs within 3–5 days of the surgical procedure. This is based on the lag phase of wound healing when the strength of the anastomosis is at its lowest point. Strength of intestinal healing approaches 100% of normal by 10–17 days (Coolman et al. 2000a). Reported rates of intestinal dehiscence range between 7% and 16% (Brown 2003). The rate of dehiscence associated with small intestinal surgical procedures for neoplasia has been reported to be approximately 12% (Allen et al. 1992). Reported clinical factors contributing to an increased rate of dehiscence include poor preoperative nutritional status, generalized peritonitis at the time of the surgical procedure, and advanced age of the patient (Coolman et al. 2000a). Surgical factors affecting healing include poor intestinal blood supply or iatrogenic trauma to the anastomosed segment of bowel, excessive

222  Veterinary Surgical Oncology

tension on the anastomosis, and inappropriate choice of suture material (Allen et al. 1992; Coolman et al. 2000a). In the study by Ralphs et al. (2003), dogs with two or more of the following factors were predicted to be at high risk for developing anastomotic leakage: preoperative peritonitis, intestinal foreign body, and serum albumin concentration of 2.5 g/dL or less. Two other clinical studies have not substantiated hypoalbuminemia as a risk factor for wound healing after intestinal surgery (Harvey 1990; Shales et al. 2005). Mortality rates up to 74% are reported following generalized peritonitis from intestinal dehiscence (Allen et al. 1992; Brown 2003). In the more recent veterinary literature, survival rates of 70% and 71% have been reported in dogs with septic peritonitis when either active peritoneal or open peritoneal drainage is incorporated into the management of these animals, respectively (Mueller et al. 2001; Staatz et al. 2002). Stricture Dysfunctional stricture following small intestinal resection-anastomosis is rarely a problem with appositional suture patterns. The primary argument for appositional suture pattern techniques is avoidance of stricture. Apposition of the submucosal vascular plexus promotes wound healing without stricture formation. Inverting suture patterns are associated with the greatest decrease in lumen size at the anastomosis site (Ellison 1989). Everting suture patterns increase adhesion formation, lead to mucosal ischemia and prolongation of the inflammatory (lag) phase and therefore should be avoided during the generation of an intestinal anastomosis (Ellison 1989). Recurrence Large-scale studies evaluating the local recurrence rate of gastrointestinal tumors have not been performed. In general, the rate of recurrence is dependent on the tumor type and its location within the intestinal tract. Local recurrence is generally not a problem with jejunal masses because wide surgical margins (i.e., 5 cm) are generally easily attained. Local recurrence is more likely with ductal sparing (i.e., bile and pancreatic ducts) procedures associated with the proximal duodenum; however, this can be circumvented with the use of more aggressive excisions (see gastrojejunostomy procedure [Billroth II] earlier in this chapter). Mechanical and physiological complications Short bowel syndrome (SBS) can occur with aggressive small intestinal resections. Clinical signs include diarrhea, steatorrhea, malnutrition, and weight loss. Gut adaptation occurs over time through increases in entero-

cyte number and size, increased villus height and crypt depth, and intestinal diameter (Brown 2003). Intestinal villi increase the surface area of the intestinal lumen by approximately 8 times in dogs and 15 times in cats (Brown 2003). Experimentally, several surgical procedures to treat SBS have been described. These include construction of intestinal valves, interposition of reversed intestinal segments, colonic interposition, and reversed electrical intestinal pacing (Brown 2003). None of these have been successfully used in a clinical setting. Maintenance of the ileocolic component of the ileocecal valve is thought to be important to minimize clinical signs associated with SBS in dogs or cats with extensive intestinal resections. In the study by Gorman et al. (2006), risk factors (age of patient, percentage of intestine removed, underlying disease) for the development of complications after extensive small bowel resection were evaluated. No risk factors were identified and 12 of 15 (80%) of the dogs and cats that underwent extensive small bowel resection were reported to have good longterm outcomes (Gorman et al. 2006). Postoperative ileus is common after intestinal surgery and generally occurs within the first 24 hours of the postoperative period. Factors that promote the development of ileus include extensive intestinal manipulation, long operative time, and extensive tissue resection (Brown 2003). When present, ileus should aggressively managed using injectable motility modifying agents and antiemetics such as metoclopramide, maropitant, and ondansetron. Length limits to intestinal resection The minimum length of small intestine required for survival with oral nutrition alone is unknown in dogs, however, up to 75% of the small intestine can often be removed without resultant clinical signs. Jejunal resection is better tolerated than removal of the ileum, and preservation of the ileocolic valve is important in prevention of bacterial overgrowth. Experimentally, 20% of dogs have survived for over 1 year with 70%–90% of their intestines removed (Yanoff et al. 1992). Unfortunately, risk factors for the development of SBS in dogs and cats undergoing extensive intestinal resection have not been identified. As a result, owners should be warned of the potential for development of this syndrome when greater than 50% of the small intestine is planned to be removed (Gorman et al. 2006). Common tumors for which this procedure is performed Four general categories of tumors occur within the small intestine. These include epithelial, smooth muscle (mesenchymal), neuroendocrine, and round cell neoplasms

Alimentary Tract  223

(Selting 2007). Lymphoma is generally reported as the most common intestinal neoplasm, comprising approximately 30% of all feline tumors and 6% of all canine tumors (Selting 2007). Nonobstructive lymphomatous intestinal disease is generally managed medically, and prognosis and treatment recommendations vary depending on the tumor subtype. Although controversial, solitary obstructive lymphomatous lesions can be effectively managed with surgical excision. If obstruction is suspected but not confirmed clinically (i.e., patient is not showing severe intestinal signs), trial treatment with chemotherapy can be attempted and an objective clinical response can be measured with serial abdominal ultrasound exams. If the tumor does not respond to chemotherapy or if progressive disease is documented, then surgical excision is warranted. When a large tumor burden is present, owners should be counseled about the possibility of intestinal perforation if a rapid response to chemotherapy is elicited. Adenocarcinoma (ACA) is the most commonly treated intestinal tumor in dogs and cats, but is reported to be the second most common tumor type to occur within the bowel (Selting 2007). Approximately 40% of adenocarcinomas have metastasized at the time of surgical exploration. These tumors primarily metastasize to regional lymph nodes; however, other commonly reported sites of metastasis include the liver, peritoneum, mesentery, and omentum. In dogs, the overall prognosis for surgically treated small intestinal ACA is 7–15 months with localized disease and 3 months with metastatic disease (Paoloni et al. 2003; Crawshaw et al. 1998; Birchard et al. 1986). Untreated dogs have been reported to have a mean survival time of 12 days (Selting 2007). In the study by Crawshaw et al. (1998), when lymph nodes were negative for metastasis, the 1-year survival rate was reported to be 66.7% compared to 20% if documented lymph node metastasis had occurred. It is therefore recommended to sample regional lymph nodes (if they are accessible) prior to surgery with ultrasoundguided needle aspirates. Regional lymph nodes should also be biopsied at the time of definitive surgery for disease prognostication, and sampling should be performed even if the lymph nodes appear grossly normal. Currently, the therapeutic value of regional lymphadenectomy at the time of primary tumor excision is unknown, and this procedure is generally not recommended due to its potential for associated morbidity (Crawshaw et al. 1998). In surgically treated cats with intestinal ACA, reported survival times range from 20 weeks (median) to 15 months (mean) (Turk et al. 1981; Kosovsky et al. 1998). The reported survival time of untreated cats with intestinal ACA is approximately 2 weeks (Kosovsky et al. 1998; Birchard et al. 1986).

Mesenchymal neoplasms of the gastrointestinal tract have recently been divided into gastrointestinal stromal tumors (GISTs) and leiomyosarcomas (GILMs). GISTs are most commonly identified in the cecum and large intestine, and GILMs occur more commonly in the stomach and small intestine. GISTs have a higher rate of intestinal perforation but increased long-term survival times when compared to GILMs (Maas et al. 2007). No perforations have been reported with GILMs. The higher perforation rate for GISTs is thought to be a result of their typical cecal location, which leads to delayed development of clinical signs as compared to more oral positioned neoplasms. GISTs are composed of a high percentage of interstitial cells of Cajal, which regulate intestinal motility. These tumor types are differentiated based on immunohistochemistry. In one study, a median survival time of 37.4 months was reported for dogs with GISTs (cecum and large intestine) and 7.8 months for dogs with GILMs that survived perioperative period (Russel et al. 2007). Another study did not show a significant difference in survival rates between GILMs or GISTs and with both tumor types; approximately 80% of dogs were tumor free at 1 year and 65% at 2 years (Maas et al. 2007). Adjuvant therapies There is no evidence of benefit associated with the administration of chemotherapy in animals suffering from epithelial- or mesenchymal-based neoplasms. Che­­­ motherapy is generally also not thought to be helpful in humans with small intestinal neoplasia (Stanclift and Gilson 2004). Medical treatment with tyrosine kinase inhibitors and imatinib mesylate (Gleevec) is being explored in people with GIST. A partial response is observed in approximately 50% of human cases treated with Gleevec (Maas et al. 2007). Tyrosine kinase receptor inhibitor therapy has recently received a great deal of attention in veterinary medicine, allowing for the development of two commercially available therapeutics (Palladia [toceranid phosphate, SU11654] and Kinavet [mastinib]), which may be appropriate for use in dogs with documented advanced-stage GISTs (London 2009; Russel et al. 2007). At this time, however, only anecdotal reports of successful use of these agents has been reported and use of these agents for GIST is considered to be off-label.

Colorectal Tumors Clinical workup and biopsy principles During a complete physical examination, abdominal palpation may identify a palpable abdominal mass in some cases. In a digital rectal examination, the tumor is

224  Veterinary Surgical Oncology

(a)

(b)

Figure 7.20.  Phases of fine-needle aspiration of enlarged sublumbar lymph nodes. (A) After the lymph node has been penetrated by the needle, the stylet is removed and aspiration is performed. (B) The finger is extracted from the anus, and the material is sprayed on a slide for cytological examination.

often palpable as, mainly in dogs, most rectal tumors are located 3–8 cm from the anus; stenosis may be felt if the tumor growth is circumferential. Anorectal stricture may be very rarely idiopathic as a result of anorectal spastic contraction. The latter condition is usually seen in German shepherd dogs and disappears under general or epidural anesthesia (Niebauer 1993). It is important during palpation to identify infiltrative tumors. As the normal rectum is freely movable, in case of full-thickness wall infiltration the rectal tube may be variably fixed to the surrounding tissues. Infiltration may be caused by the primary rectal tumor (more often an adenocarcinoma) that, after its growth into the full thickness of the intestinal wall, has invaded surrounding tissues; as an alternative, the rectum may be secondarily invaded by tumors arising from pelvic organs (mainly the prostate gland; see Chapter 10). During digital rectal examination, the so-called sublumbar lymph nodes are palpated. Colorectal tumors can primarily spread to sublumbar and colic lymph nodes, the latter located in the mesocolon. The term sublumbar refers to all lymph centers present in the sublumbar region: the iliosacral lymph center (which includes the medial iliac, hypogastric, and sacral lymph nodes) and the iliofemoral lymph nodes (Bezuidenhout 1993). The sublumbar lymph nodes draining the anus, rectum, and colon are the hypogastric (ventral to the sixth and seventh lumbar vertebrae and adherent to the external and internal iliac arteries) and

the medial iliac lymph nodes (ventral to the fifth and sixth lumbar vertebrae, between the deep circumflex and external iliac arteries). These lymph nodes, if enlarged, may be variably felt, depending on both their size and the dog’s size. Cytology samples from these lymph nodes may be collected transrectally using a long 22-gauge spinal needle coaxial to the index finger and interposed between a surgical rubber glove and the finger part of another glove on the same index finger. The animal may need to be sedated for the procedure. After manual evacuation of any feces from the caudal rectum, the lubricated index finger is introduced into the rectum and advanced until it reaches the caudal pole of one of the enlarged lymph nodes. The needle is introduced after passing it through the needle-protecting finger-glove first and the entire thickness of the intestinal wall second. The stylet is removed, a 2–5 mL syringe is connected, and aspiration is performed (Figure 7.20). Slides are prepared in a routine manner. Complications of this procedure are very rare and are generally related to small-sized lymph nodes and inadvertent puncture of blood vessels. In larger dogs, these lymph nodes may be not reached if they are even slightly enlarged. Alternatively, a transabdominal ultrasoundguided fine-needle aspiration can be performed. During a complete laboratory workup (blood, urine), possible changes seen are anemia (usually microcytic hypochromic due to chronic blood loss or normocytic

Alimentary Tract  225

Figure 7.21.  The rectal mass is exposed by traction on four stay sutures applied 1–2 cm cranially to the anorectal line. Another stay suture has been applied on the rectal wall close to the caudal margin of the mass in order to facilitate exteriorization. In this particular case, an excisional biopsy was performed (see Figures 7.28F, G) after incision of the mucosal layer, since the tumor was intramural (histology of the biopsy revealed a leiomyosarcoma). After the histological result was known, a colorectal resection was performed. (From Buracco P. 2007. Tumori colorettali. In Oncologia del cane e del gatto G. Romanelli, editor. London: Elsevier. Used by permission.)

normochromic in chronic disease), thrombocytopenia, hypoproteinemia (chronic bleeding) or hyperproteinemia (paraneoplastic) (Trevor et al. 1993), hypoglycemia (Bagley et al. 1996), and erythrocytosis (Sato et al. 2002). Paraneoplastic leukocytosis (left shift neutrophilia, monocytosis, and eosinophilia) has been reported in dogs in association with a rectal adenomatous polyp (Thompson et al. 1992; Knottenbelt et al. 2000a) Biopsy samples may be collected using one or more of these procedures. 1. Biopsies may be collected directly, without anesthesia, if the mass is prolapsed through the anus. This technique is not advised as only superficial samples are collected and bleeding may be a concern. 2. A needle core biopsy (Tru-Cut needle) may be done through the anus after rectal eversion by traction on four stay sutures applied 1–2 cm cranially to the anorectal line (Figure 7.21). 3. As an alternative, an incisional biopsy may be performed. This procedure requires anesthesia (including an epidural) and does not yield any information regarding the large intestine cranial to the lesion.

4. Endoscopic examination (proctocolonoscopy) may also be done. Preparation for endoscopy may be provided by fasting the animal 1.5–2 days before the procedure; drinking is allowed until 8 hours before the examination. The day before the examination, warm water enemas (10–20 mL/kg twice daily) are performed; finally, the evening before, an osmotic laxative is administered orally to the patient. Proctocolonoscopy is necessary to characterize the tumor (single or multiple lesions, position, size, length, and circumferential extension) (Figure 7.22A–C) and to obtain a biopsy. Due to the procurement of only superficial samples, endoscopic biopsies can sometimes be inadequate (Morello et al. 2008). The large intestine cranial to the lesion is also inspected for multiple tumors. The complication rate of flexible colonoscopy is very low (mortality rate of 0.28%), and major complications such as fatal aspiration of colon electrolyte solution (used for bowel cleansing prior to colonoscopy), colonic perforation, and excessive hemorrhage after biopsy have been rarely reported (Leib et al. 2004). 5. Another procedure, the colotomy (occasionaly performed for tumor biopsy), will be not described here. Imaging techniques Radiography and contrast studies At present, radiography and contrast studies have been largely superseded by ultrasonography as the latter is much more effective for evaluating intestinal intramural lesions. However, survey radiographs may identify an abdominal mass (Slawienski et al. 1997) and suggest an obstructive condition (Figure 7.23A) whereas contrast studies may outline the site of obstruction (Figure 7.23B) Ultrasound Ultrasound examination of the abdomen (Myers and Penninck 1994; Rivers et al. 1997; Slawienski et al. 1997; Paoloni et al. 2003; Llbrés-Diaz 2004) is considered to be the most appropriate for intestinal malignancies, even though the precise site (small or large intestine) may be not distinguished. Abdominal ultrasonography should always be performed with colorectal tumors, despite the fact that the bone of the pubis prevents imaging of the rectum in the pelvic canal, therefore potentially resulting in an unremarkable study. Ultrasound may identify areas of localized, irregular thickening of the intestinal wall (more than 4 mm) with loss of delineation of the normal intestinal wall layers, fluid and/or fecal material accumulation (obstructive lesion),

(a)

(c1)

(b)

(c2)

Figure 7.22.  Endoscopic view of (A) a rectal leiomyosarcoma (the same as in Figure 7.21); (B) a rectal adenocarcinoma; and (C) a 35 cm long 360-degree colorectal adenocarcinoma (C1: CT view; C2: endoscopic view). (Photos in (A) and (B) from Buracco P. 2007. Tumori colorettali. 2007. In Oncologia del cane e del gatto G. Romanelli, editor. London: Elsevier. Used by permission. Endoscopic pictures courtesy of Dr. Caccamo Roberta.)

(a)

(b)

Figure 7.23.  (A) megacolon caused a colorectal adenocarcinoma in a dog; (B) barium enema in a case of colonic lymphoma in a dog.

226

Alimentary Tract  227

and/or abdominal lymphadenomegaly (colic, sublumbar), as well as other abdominal abnormalities (mainly on the omentum, mesentery, and mesenteric lymph nodes). It has been reported that the different tumor histotypes should have distinct ultrasonographic appearance, and diffuse echogenicity is described in the case of adenocarcinoma in most dogs. Ultrasound-guided fine needle aspiration (with a 22- or 20-gauge spinal needle) or biopsy (with a 18-gauge Tru-Cut biopsy needle) of these lymph nodes, as well as of any intestinal lesion, may be attempted transcutaneously. Thorax radiography Radiographic evaluation of the thorax (three views: two lateral, one dorsoventral) is needed to evaluate the presence of lung metastases, although they are extremely rare in these diseases. Contrast-enhanced CT and MRI Contrast-enhanced CT and MRI are indicated to detect intrapelvic infiltration of the rectal tumor, determine the extent of the disease, and confirm intrabdominal/ sublumbar lymphadenomegaly (Figure 7.24; see also 7.22, 7.33, and the section on perianal tumors). Laparoscopy If laparoscopy is available and in the hands of an experienced operator, it may be useful to inspect the entire abdomen and the sublumbar space, as well as to take biopsies. Surgical techniques Preparation for colorectal surgery A canine experimental study has confirmed the role of preoperative mechanical bowel preparation (started 24 hours before surgery with fasting and ingestion of 20 mL of magnesium hydroxide plus 15 mL/kg 10% mannitol orally) in decreasing the early mortality rate due to dehiscence and peritonitis after segmental colectomy and end-to-end anastomosis (Feres et al. 2001). Present recommendations to decrease the risk of intraoperative contamination vary slightly among clinicians, but in general they include, when possible, a low-residue diet started from 2–3 days (Hedlund and Fossum 2007b) to 1 week before surgery and no warm water enema or laxative within the preoperative hours, from 3–72 hours (Hedlund and Fossum 2007b; Holt and Brockman 2003), if a standard surgical excision is to be performed. The author of this section of the chapter prescribes 4–5 days of low-residue diet before surgery and avoids performing an enema during the preoperative 72 hours. In the case of obstruction and suspected perforation (the

latter mainly in cats with lymphoma), enemas are avoided. In preoperatively debilitated patients, enteral or parenteral nutrition, plasma, or blood transfusion may be considered. Animals are prepared by withdrawing food and water for 12–24 hours and 8 hours before surgery, respectively. Some surgeons, however, allow preoperative water access (Hedlund and Fossum 2007b). Preoperative antibiotic therapy started 24 hours before surgery is controversial, but given that the risk of infection is high, antibiotics against anaerobes and gram-negative aerobes may be used. Drugs that decrease colorectal anaerobic bacteria include third–generation cephalosporins, neomycin or kanamycin, cefazolin and metronidazole, and neomycin and erythromycin (Holt and Brockman 2003; Hedlund and Fossum 2007b). Cephalosporins are used at induction of anesthesia and then every 2 hours of operative time. Positioning of the patient varies depending on the surgical procedure: dorsal recumbency for ventral midline celiotomy (typhlectomy, colectomy, ventral approach to rectum-colon) or sternal recumbency for anal, transanal, and para-anal approaches. For sternal recumbency, the perineal area and the proximal onethird to one-half of the tail are clipped, and manual evacuation of rectum and anal sac contents is performed. Animals are positioned with their perineal area elevated, with the tail bandaged and secured dorsally and cranially, with the hind legs hanging over the packed end of the surgical table (to avoid pressure lesions to both skin and femoral nerves). After tumor resection has been completed, the resection margins are identified with China ink staining or other tissue inking systems (Davidson Marking Systems, Bradley Products Inc., Bloomington, MN) and/or with one or more sutures usually placed orad or aborad to the lesion, depending on the resection performed. The specimen is then immersed in neutral-buffered 10% formalin. Surgical instruments and gloves are changed at this stage. Surgical procedures Typhlectomy This procedure is indicated for tumors confined to the cecum. The cecum can be removed from the colon only or in conjunction with the distal ileum and proximal colon, depending on the extent of resection required to achieve excisional margins. In the former case, after dissection of the ileocecal fold, the cecum is isolated and two Doyen clamps, one of which is placed at its base, are applied. After double ligation of the appropriate vessels (cecal branches of the ileocecal artery running in the ileocecal fold), the cecum is resected by incising between

228  Veterinary Surgical Oncology

(b) (a)

(d)

(c)

Figure 7.24.  (A) intraoperative view of an infiltrative colorectal adenocarcinoma. This should be recognized before surgery as it represents a negative prognostic factor both for surgery and metastatic spread to regional lymph nodes (in this picture colic lymph nodes—arrow—appear enlarged; at histology they were metastatic). (B,C): CT scan of the same lesion showing both the infiltration and obstruction caused by the tumor (B) and an enlarged sublumbar lymph node (arrow) (C). (D) The tumor is resected through a combined approach, abdominal first and then transanal (see also Fig. 7.33 and 7.35). The two enlarged colic lymph nodes are visible and resected with the tract of the intestine that is removed. (A and B reprinted with permission from Buracco P., Tumori colorettali. In Oncologia del cane e del gatto, edited by Romanelli G, 2007, Elsevier)

the two clamps with a scalpel. A Parker-Kerr suture is usually used to close the defect (3-0/4-0 polydioxanone, polyglyconate, poliglecaprone 25) (Figure 7.25A). As an alternative, the intestine may be closed with simple interrupted sutures or by using a TA or GIA stapler, resulting in the eversion of the cut edge of all layers. Further manual oversewing of the staple line is optional (Holt and Brockman 2003; Tobias 2007). In the case of larger tumors, after double ligation of the corresponding ileocolic arterial branches, the cecum is removed together with both the distal ileum and proximal colon (Figure 7.25B). Care is taken to ensure that

the diameters of the two intestinal stumps are the same using one of the following techniques: spatulating the small intestine on the antimesenteric border in order to increase its diameter; incising the smaller intestine at an oblique angle (45–60 degrees, with the antimesenteric border shorter than the mesenteric one; or partially oversewing the colon to reduce its diameter. An ileocolonic end-to-end anastomosis is then performed. Manual anastomosis is accomplished using full-thickness appositional simple interrupted sutures with 3-0 or 4-0 absorbable monofilament material (polydioxanone, polyglyconate, or poliglecaprone 25) with the knots in

Alimentary Tract  229

(a)

(b)

Figure 7.25.  (A) A typhletomy for a small cecal leiomyosarcoma has been performed, including a Parker-Kerr suture for closure. (B) A typhectomy has been performed in another dog for a larger leiomyosarcoma, and the cecum has been removed together with a portion of both small and large intestine. (Photographs courtesy of Dr. Romanelli Giorgio)

an extraluminal position. As an alternative, end-to-end anastomosis may be performed with a circular EEA stapling device (see below) (Kudisch and Pavletic 1993; Holt and Brockman 2003; Hedlund and Fossum 2007b; Tobias 2007; Banz et al. 2008). After checking that there is no gross leakage from sutures (by carefully inspecting the anastomosis or by gently distending the affected intestinal segment with sterile saline as either side of the anastomosis is digitally occluded), an abdominal lavage with warm sterile saline and suction precede both wrapping of the anastomotic site with omentum and standard closure of the abdomen. For postoperative care see the section on the colectomy, below. Functional outcome is usually good. Potential complications Complications may include dehiscence, infection, stricture (caused by inappropriate surgical technique and/or suture material; it may require surgical exploration if it causes obstruction), recurrence, and metastasis. The removal of the ileocolic junction in cats (Sweet et al. 1994) and dogs may result in clinical signs, such as increase in the frequency of defecation and looser stool for a variable period of time (from weeks to months). This should be communicated to owners prior to surgery. For the extent of resection, see above. Colectomy Indications for colectomy are tumors confined to the colon only. Although this situation is rare, particularly in dogs where large intestinal tumors are more frequently colorectal (Selting 2007), subtotal colectomy is a common consideration in cats as feline colonic adenocarcinomas are often amenable to surgical removal by colectomy alone (Slawienski et al. 1997). One unusual

report of a pure canine colonic adenocarcinoma involved the entire colon, although this tumor was not resected (Prater et al. 2000). Colectomy may be total or subtotal, depending on the extent of resection required to achieve excisional margins. In the latter case the ileocecocolic valve is preserved, taking care to transect the ascending colon 3–5 cm from the cecum to obtain a final tensionfree end-to-end anastomosis. The colon is exteriorized, and the area to be removed is identified. If possible, feces are digitally milked away from the portion of colon to be resected. The colon is isolated with moistened sponges, and two Carmalt clamps are applied at the extremities of the section to be removed; two noncrushing clamps (Doyen) are each applied cranial and caudal to the two Carmalt clamps on the section of intestine to be preserved; as an alternative, the assistant may use his or her fingers (index and middle fingers) as a noncrushing digital clamp. In applying the two Carmalt clamps, close attention is paid to the margins of resection (see later). Vasa recta (coming from both the ileocolic artery, a branch of the cranial mesenteric artery, and the caudal mesenteric artery for the distal half of the descending colon) are double ligated. For segmental colectomy, care is taken to ligate only vasa recta and to spare vessels that run parallel to the bowel (Holt and Brockman 2003; Hedlund and Fossum 2007b). If most or all the colon and the cranial part of the rectum requires removal, the caudal mesenteric artery (which emerges from the abdominal aorta at the level of the caudal aspect of the 5th lumbar vertebra and from which the cranial rectal artery is derived) may require ligation, even though this may potentially decrease the blood supply at the anastomotic site and increase the risk of postoperative dehiscence and stricture. Every effort, therefore, is made to save major colic vessels when the extent of the disease allows this to occur. Despite this recommendation, no

230  Veterinary Surgical Oncology

(a)

(c)

experience, preference, and costs. The circular stapling device may be introduced through the anus or through a small cecal incision (Kudisch and Pavletic 1993; Banz et al. 2008) (Figure 7.26B, C). Stapling may be a problem in cats and small dogs because of size limitations (Holt and Brockman 2003; Hedlund and Fossum 2007b) but the equipment that is now available (21, 25, 28, and 31 mm diameter sizes) may be used in most cats and small dogs (Banz et al. 2008). Recently, a successful endto-end colocolic anastomosis technique with biofragmentable rings after subtotal colectomy in cats has been reported (Ryan et al. 2006) (Figure 7.26D). After checking that there is no gross leakage from sutures (by carefully inspecting the anastomosis or by gently distending the affected intestinal segment with sterile saline as either side of the anastomosis is digitally occluded), an abdominal lavage with warm sterile saline and aspiration precede both wrapping the anastomotic site with omentum and standard closure of the abdomen. Postoperative care Important postoperative considerations for postoperative care include the provision of analgesia for 24–48 hours, fluid and electrolyte therapy, antibiotics in the case of established postoperative infection and peritonitis, and Elizabethan collars. A small amount of water should be provided 8–12 hours after surgery and a light feeding (e.g., Hill’s i/d) 12–24 hours after surgery. Return to normal feeding is usually within 3 days (Holt and Brockman 2003). Functional outcome is usually good.

(b)

Figure 7.26.  (A) Colocolonic end-to-end anastomosis obtained in a cat using a manual suture technique. (B) Colocolonic end-toend anastomosis obtained in a dog with a circular EEA stapling device. (C) Sutureless colocolonic end-to-end anastomosis with Valtract biofragmentable anastomosis ring (BAR) device in one cat (Ryan et al. 2006). (Photographs B and C courtesy of Dr. Eric Monnet)

postoperative complications were seen in a recent paper in which ligation of the caudal mesenteric artery was required in two dogs to enable adequate tumor resection (Sarathchandra et al. 2009). Finally, the colon is transected between the clamps with a scalpel or Metzenbaum scissors and is opened to confirm that sufficient macroscopic margins have been achieved. Further resection, using a new set of surgical instruments and gloves, can be considered in the event that excisional margins are questionable. End-to-end anastomosis (ileocolonic or colocolonic) may be manual (see typhlectomy and Figure 7.26A) or by stapling depending on the surgeon’s

Potential complications Complications may include short-term rectal bleeding, loose feces, tenesmus, stricture (that can require a second surgery if it causes obstruction), dehiscence and peritonitis (that require patient stabilization and surgical exploration), and potential loss of reservoir continence if most of the colon is removed (see also later). Loss of reservoir continence results in more frequent conscious defecation (Guilford 1990; Dean and Bojrab 1993). Cats tolerate more removal than dogs (90%–95% of the colon) (Bertoy et al. 1989); however, no long-term adverse effects is seen in dogs after removal of up to 70% of the entire colon (Bertoy et al. 1989; Jimba et al. 2002; Hedlund and Fossum 2007b). In a recent paper, subtotal colectomy with preservation of the ileocolic junction in dogs resulted in elimination of liquid stools 10–12 times a day; normalization of fecal output (normal fecal consistency and two to three daily defecations without tenesmus) was achieved within 5–10 weeks (median 7 weeks) (Nemeth et al. 2008). For the extent of resection, see above.

(a)

(d)

(b)

(c)

(e)

(f)

(g)

Figure 7.27.  Pull-out procedure. (A) Application of a stay suture 1–2 cm cranially to the anorectal line. (B) Eversion of the rectum by traction on four of these stay sutures and exposure of two rectal polyps that are resected locally (C, D). In this case the rectal incision was closed with a one-layer continuous suture (4-0 poliglecaprone 25) (E). (F, G) the lesion (see Figures. 7.21 and 7.22A) is exposed. Excisional biopsy is performed by incising the rectal mucosa at the periphery of the lesion (F), and the mass is excised by blunt dissection and traction. Finally, the rectal wall is sutured with a simple interrupted suture pattern using absorbable monofilament material (G). (Photographs F and G are from Buracco P. 2007. Tumori colorettali. 2007. In Oncologia del cane e del gatto G. Romanelli, editor. London: Elsevier. Used by permission.)

231

232  Veterinary Surgical Oncology

Figure 7.28.  Endoscopic polypectomy. The polypectomy snare is opened and the polyp is surrounded at the base. The snare is then gently closed around the base of the polyp until a mild color change in the polyp head is observed. The snare is now tightly closed, the current is activated and the polyp excised. (Photograph courtesy of Prof. Gualtieri Massimo)

Simple excision This procedure should be reserved for small, single and superficial benign tumors (e.g., polyps) located in the caudal-midrectum (Figure 7.27B–D). In general, simple excision should be considered as a biopsy procedure and further surgeries should be considered based on the histology results (Morello et al. 2008). Excision may be performed using sharp instruments (Figure 7.27C) or with an electrosurgical snare or cautery tip used in conjunction with proctocolonoscopy (Palminteri 1966; Holt and Durdey 1999) (Figure 7.28). As an alternative, cryosurgical, laser, or TA stapler devices may be used (Valerius et al. 1997; Shelley 2002; Tobias 2007; Swiderski and Withrow 2009). Standard surgical excision relies on prolapse of the rectum (pull out procedure) through a transanal approach (Figure 7.27A, B, F). After standard surgical preparation of the area, the rectal wall is everted through the anus via traction on four stay sutures applied 1–2 cm cranially to the rectocutaneous line. The lesion is exposed externally, which may require the sequential placement of further stay sutures as necessary (see Figure 7.21), and excision of the mass can begin. If the lesion is attached to the wall through a stalk, the latter is simply ligated and transected. If the lesion has a sessile attachment, excision is performed by incising the normal mucosa along the periphery of the lesion or deeper (Figure 7.27C). Closure is performed in one to two layers, depending on the depth of the incision, with a simple interrupted or simple continuous suture pattern with absorbable monofilament material (3-0 or 4-0 polydioxanone, polyglyconate, or poliglecaprone 25) (Figure 7.27E–G). As an alternative, a linear stapling device may be used.

In a recent paper, the use of a 30 mm, vascular thoracoabdominal (TA) stapling device applied in a transverse or longitudinal orientation at the base of the mass allowed tumor resection, leaving three rows of staggered staples with a minimum of 0.5 cm surgical margins (Swiderski and Withrow 2009). The procedure is indicated for superficial tumors located in the distal third of the rectum, with a base of attachment of less than 3 cm; the reported complication rate was low, operative time was short (around 15 minutes), deep resection was beyond the mucosal layer but full-thickness resection was avoided (to prevent infection), and the final closure line was inverted into the rectal lumen. Regardless of the procedure performed, the stay sutures are then removed. Postoperative care is described below. Functional outcome is usually good. Potential complications Possible complications are rectal bleeding and tenesmus for 1–4 or more days. Cryosurgery has been associated with complications such as stricture, rectal prolapse, and the development of perineal hernia secondary to tenesmus (Church et al. 1987). Extent of resection This is a conservative procedure. Colorectal resection (variable portions of both descending colon and rectum) This technique is mainly applicable to dogs, but it can also be used in cats. The colorectal junction is about at the level of the pelvic inlet near the caudal peritoneal reflection; the latter is at the level of the second caudal vertebra. The rectococcygeus muscles (that attach the rectum to the ventral fifth caudal vertebra) are caudal to this reflection. Each side of the rectum is supported laterally by the levator ani and coccygeus muscles. On each side, at the level of the peritoneal reflection, the pelvic plexus provides innervation to the rectum via the parasympathetic pelvic and sympathetic hypogastric nerves. The rectal blood supply is derived from essentially three arteries: the cranial rectal artery (from the caudal mesenteric artery) and the middle and caudal rectal arteries (from the internal pudendal artery). If the cranial rectal artery is ligated, most of the intrapelvic rectum should be resected to ensure adequate blood supply to the anastomosis; in fact, it has been shown that the cranial rectal artery is the most important vessel for both the terminal colon and rectum (Goldsmid et al. 1993). Resection may be performed using a number of different approaches. Any colorectal resection may be problematic or even contraindicated when infiltration of the extrarectal tissues is evident (see Figures 7.22 and 7.24A, B).

Alimentary Tract  233

(a)

(d)

(b)

(e)

(c)

(f)

Figure 7.29.  (See also Figures 7.24D and 7.32.) Transanal pull-through procedure. (A) Eversion of the rectum by traction on stay sutures. (B) The rectococcygeus muscles have been isolated before transection. (C) The rectum is bluntly dissected and isolated. (D) The rectal section to be removed is opened longitudinally, and (E) progressive resection and suturing are accomplished sequentially without touching the tumor. (F) Interrupted suturing (one layer) with absorbable material is close to completion, and the stay sutures have been released.

Dorsal inverted (dorsal perineal) approach The technique described here reflects that originally reported in two reports (Mckeown et al. 1984; Anderson et al. 1987). Indications for this procedure include rectal resections for small malignant tumors located in the caudal-midrectum. The anal sac content is evacuated prior to the placement of the purse-string suture in the anus. Urethral catheterization is advised if extensive dissection is expected (Holt et al. 1991). An inverted U-shaped incision is made over the dorsal aspect of the anus, terminating on both sites just medial to the tuber ischium. Meticulous hemostasis and dissection are performed to the level of the dorsal rectum and several muscles such as the levator ani (just lateral to rectum), coccygeus (lateral to the levator ani), external sphincter (around the caudal end of rectum), and the rectococcygeus (dorsally, located on the midline from the coccygeal vertebrae and dividing into two bundles that surround the rectum laterally) are identified. The

rectococcygeus muscle is severed between the coccygeal vertebrae and rectum or more proximally to free the rectum (Figure 7.29B). Blunt dissection is performed bilaterally between the rectum and the external sphincter and levator ani muscles, taking care not to damage the pudendal nerve and its termination as the caudal rectal nerve, which innervates the external sphincter muscle. If necessary, both levator ani muscles are transected to increase surgical exposure. The rectum is then further freed by circumferential blunt dissection in a cranial direction until the caudal peritoneal reflection is identified. The rectum is then pulled caudally, and stay sutures using 2-0 suture material are applied, as needed, both on the section to be removed and on the cranial rectal extremity that will be spared. This helps manipulation, avoids leakage of fecal material into the pelvic canal after incision, and prevents cranial retraction of the proximal intestinal segment; transection is completed both cranially and caudally (1–2 cm cranial to the external sphincter muscle) with a scalpel or Metzenbaum scissors. The

234  Veterinary Surgical Oncology

part of the rectum bearing the tumor is removed and opened to confirm that the macroscopic normal margins are wide enough; if not, further resection is performed using a new set of surgical instruments and new gloves. Moistened gauze is inserted into the cranial intestine as needed in order to avoid spillage of fecal material into the surgical field. End-to-end approximating anastomosis is performed using a simple interrupted suture pattern with absorbable monofilament material (3-0 or 4-0 polydioxanone, polyglyconate, or poliglecaprone 25). In order to triangulate the lumen and to facilitate the application of sutures 3 mm apart, three full-thickness appositional interrupted stay sutures are applied at the 12, 4, and 8 o’clock positions around the rectal circumference. Care is taken during suturing not to incorporate the opposite wall and inadvertently close the rectal lumen. Ventral sutures are applied after having rotated the bowel by 120 degrees. Knots are placed on the serosal surface in order to decrease postoperative rectal irritation (Holt et al. 1991). As an alternative, end-to-end anastomosis may be performed with a circular EEA stapling device inserted in the rectum (Hedlund and Fossum 2007b; Tobias 2007; Banz et al. 2008). Both levator ani muscles are reattached with mattress sutures if previously severed; the rectococcygeous may also be sutured to the intestine. The use of postoperative drainage tubes depends on the degree of fecal contamination that occurred during surgery; however, it has been claimed that healing at the anastomotic site could be disturbed if the drain is against the site (Hedlund and Fossum 2007b). Interposition of fat between the anastomotic site and the soft latex drain has been proposed to avoid this (Holt et al. 1991). After routine closure of both subcutaneous tissues and skin, the anal purse string suture is removed, and the gauze sponges in the rectum are extracted. Postoperative care, functional outcome, potential complications, and extent of resection are described below. Lateral (perineal) approach This approach is usually not advisable for rectal tumor resection since only one side of the rectal tube is exposed. Rectal pull-through procedure This procedure is indicated for confirmed malignant tumors and for recurrences of previously excised benign tumors located in the midcaudal rectum. Incision is started circumferentially over the anal skin (Figure 7.30A); bilateral anal sac removal is often concomitantly performed if the procedure involves their openings (Aronson 2003). The rectum is progressively dissected and pulled caudally out of the body with the help of stay sutures or grasping forceps; the rectococcygeus muscle is

also transected (Figure 7.30B). Care is taken, if feasible, to save the external sphincter muscle, undermining inside the circumference of this muscle. Cranial rectal transection is performed according to the extent of resection required to achieve excisional margins, and stay sutures are applied to prevent the cranial rectal segment from retracting into the abdominal cavity after resection. As an alternative, the rectum is partially transected and the closure is started (Figure 7.30C). The removed tract is then opened longitudinally to check that there is adequate macroscopically healthy tissue excised at the resection margins. As an alternative, the rectal section to be removed is opened longitudinally, and progressive resection and suturing are accomplished sequentially without touching the tumor (Figure 7.29D). Moistened gauzes are packed into the cranial rectum in order to avoid the spillage of fecal material in the pelvic canal. Closure is performed in one to two layers with simple interrupted sutures (3-0 or 4-0 polydioxanone, polyglyconate, poliglecaprone 25) (Figures 7.29E and 7.30E). The single-layer suture pattern would be preferable because the double-layer suture pattern has been associated with a higher risk of dehiscence when performed outside of the abdomen (Everett 1975). In the doublelayer suture pattern, the deeper one approximates the intestinal serosa (or adventitia)/muscularis to the perianal subcutaneous tissues, and the superficial layer approximates the submucosa/mucosa to the skin (Aronson 2003). This author usually performs a singlelayer closure. Postoperative care is discussed below. Functional outcome is usually good. Potential complications Fecal incontinence is an undesirable complication of this procedure (discussed below). Extent of resection The only limit to resection is the maintenance of a tension-free end-to-end anastomosis. Resection with this approach usually involves only the rectum; however, it can be extended beyond the caudal peritoneal reflection for a small portion of the distal descending colon. Transanal pull-through procedure This procedure is indicated for confirmed malignant tumors and for recurrences of previously excised benign tumors located in the midcranial rectum (Aronson 2003; Hedlund and Fossum 2007b; Morello et al. 2008). The rectal wall is prolapsed through the anus with four stay sutures (see Figures 7.21, 7.27A, and 7.29A). A fullthickness circumferential incision is made through the rectal wall. When feasible, a minimum of 1–1.5 cm of

Alimentary Tract  235

(a)

(d)

(b)

(c)

(e)

Figure 7.30.  Rectal pull-through procedure. (A) Circumferential skin incision. (B) The rectum is exteriorised by blunt dissection and traction (the arrow indicates where the rectococcygeus muscle has been severed). (C) Partial transection and suturing (one layer) are accomplished sequentially. (D) The part of the rectum that has been removed and (E) the final image of the area are shown. (Photographs from Buracco P. 2007. Tumori colorettali. 2007. In Oncologia del cane e del gatto G. Romanelli, editor. London: Elsevier. Used by permission.).

distal rectum is spared in order to preserve fecal continence (Morello et al. 2008). The rectum is mobilized following transection of the rectococcygeal muscles (Figures 7.29B and 7.30B.), and blunt dissection is performed along the external surface of the bowel (Figures 7.29C and 7.30B). The mobilized rectum is pulled caudally out of the body, and stay sutures are applied to prevent the cranial segment of the rectum from retracting into the abdominal cavity after resection. Cranial rectal transection is then performed. To establish the point of cranial resection, the rectum is either externally palpated or longitudinally opened, the latter is avoided in the case of a 360-degree circumferential tumor. If a longitudinal opening is used, progressive resection and suturing are accomplished sequentially without touching the tumor (Figures 7.29D). If not opened previously, this is done after resection to confirm that resection margins are adequate. Moistened gauze sponges are packed in the cranial rectum to avoid the spillage of fecal material in the pelvic canal. Finally, the normal cranial rectum or descending colon is manually anastomosed with the preserved distal rectal stump with

one (preferably full-thickness sutures) or two layers (sero-muscular and mucosal-submucosal layers) in the situation where there is tension at the anastomotic site, with simple appositional interrupted sutures (3-0 or 4-0 polydioxanone, polyglyconate, poliglecaprone 25) (Figure 7.29E). The release of the stay sutures allows the rectal anastomotic site to return into the pelvic canal (Figure 7.29F). As an alternative, an end-to-end anastomosis is performed with a circular EEA stapling device introduced through the anus (Tobias 2007; Banz et al. 2008). This is discussed later. Postoperative care, potential complications, and extent of resection are all described later. Functional outcome is usually good. Caudal abdominal approach with either sagittal pubic symphyseal separation or osteotomy This approach is indicated for tumors located in the cranial rectum and for those tumors with further extension into the distal colon (Davies and Read 1990; Allen and Crowell 1991; Aronson 2003; Hedlund and Fossum

236  Veterinary Surgical Oncology

(a)

(b)

Figure 7.31.  (A) Both the pubis and ischium are spread apart with a Finocchietto retractor. (Photo courtesy of Dr. Julius Liptak) (B) After bilateral ischial and pubic osteotomy, both a colorectal resection and an end-to-end anastomosis have been completed.

2007b). The urethra is catheterized, and the skin incision is ventral (from the xiphoid process of the sternum to the cranial vulva in females or from the xiphoid, parapreputially to the scrotum in males). The subcutis is then bluntly undermined and the external obturator muscles are separated to expose the pubic symphysis. Both the pubis and ischium are separated exactly on the midline with an osteotome and mallet or an oscillating saw and spread apart with a Finochietto retractor (Figure 7.31A). To increase exposure, osteotomy of both the pubis and ischium has been recently described (Yoon and Mann 2008). In detail, the various steps are as follows. 1. Subperiosteal elevation of adductor muscles up to two-thirds of the obturator foramina. 2. Incision of the prepubic tendon along the pubis rim as needed.

3. Predrilling of a hole using a pin and Jacob’s chuck on each side of the four planned longitudinal osteotomies (two pubic and two ischial) to facilitate subsequent closure. 4. Pubic and ischial osteotomy (with an oscillating saw), taking care to protect the two obturator nerves and vessels using a malleable retractor (osteotomies are medial to the lateral border of the both obturator foramina, bilaterally). 5. Subperiosteal elevation of only one of the two internal obturator muscles from the pubis/ischium to enable reflection of the osteotomized bone plate on the other side. 6. Excision of the affected segment of bowel (this is isolated by careful undermining and accurate hemostasis and placement of moistened laparotomy sponges with or without stay sutures. Clamps are then placed to allow transection and removal of the diseased segment at the appropriate level (see colectomy section). The specimen is then opened to confirm that enough macroscopically healthy tissue has been maintained at resection margins; manual or stapling end-to-end anastomosis are then performed as previously described (Figure 7.31B). 7. Preplacement of sutures (orthopedic wire or 0 polydioxanone, the latter in small dogs and cats only) in the predrilled holes and reduction of the segment of bone plate by tightening the sutures. 8. Reapposition of the two adductor muscles with 3-0 polydioxanone in a simple interrupted cruciate pattern. 9. Drilling of four holes along the pubic brim to allow apposition of the prepubic tendon with 3-0 polydioxanone interrupted sutures through the bone tunnels. 10. Thorough lavage of the area prior to routine skin closure. If a symphysiotomy has been performed, the symphysis is repaired by passing a 0.8 mm stainless steel wire through a wire passer and then through both the obturator foraminae; Davies and Read (1990) suggested that the wire pass along the borders of the symphysis rather than through predrilled holes given that the bone at this level can collapse when the wire is tightened. Care is taken not to incorporate any vessel or nerve in the wire suture loop. The wire is then tightened and its ends trimmed and bent; the two obturator muscles are then apposed with interrupted sutures. The exposure provided by sagittal pubic symphyseal separation to perform both resection and anastomosis can be limited (Williams and Niles 2005). The osteotomy procedure reported by Yoon and Mann (2008)

Alimentary Tract  237

appears to offer several advantages, including good operative space as well as the option to bluntly dissect both the cranial peritoneal reflection and pelvic nerves (on both sides of the rectum) and to ligate only specific vessels as needed (cranial rectal artery or individual vasa recta, depending on the bowel segment to resect). Postoperative care Effective analgesia (constant-rate infusion of lidocaine, opioids) is required for at least 2–3 days, as recommended by many surgeons (Davies and Read 1990; Williams and Niles 2005). In the case of symphyseal distraction with a Finochietto retractor, sharp pain may be caused by sacroiliac subluxation. Functional outcome Limping may be evident for some days after surgery (mainly after symphyseal distraction), but prolonged analgesia with nonsteroidal anti-inflammatories may improve the clinical signs. For complications related to colorectal resection, see below. Specific complications Possible complications of this surgical approach may be a sinus tract related to the wire cerclage suture (Davies and Read 1990), nonunion at the site of the pubic symphyseal separation. with the need to restrict activity for a minimum of 4 months (Allen and Crowell 1991), and risk of sacroiliac subluxation. No major complications (e.g., hemorrhage, postoperative lameness, avascular bone necrosis, infection, urinary or fecal incontinence, etc.) were reported after pubic/ischial osteotomy (Yoon and Mann 2008). All animals had exercise restricted for 4 weeks postoperatively, and the authors argue that absorbable suture material is adequate in small dogs and cats as an alternative to orthopedic wire, given that neither the pubis nor the ischium are weight-bearing segments of the pelvis. The extent of resection is described below. The so-called Swenson’s pull-through and modifications This procedure is indicated for tumors of the midcranial rectum extending to the distal colon, and it is an alternative to sagittal pubic symphyseal separation or osteotomy. The animal is positioned in dorsal recumbency, and both the ventral abdomen and the perineal area are prepared for surgery. A caudal celiotomy is performed first. The colon is isolated by double ligation of the corresponding vasa recta, two Doyen clamps are applied cranially and caudally to the proposed point of

transection (Figure 7.32A), and the colon is divided. The length of the rectocolonic tract to be removed should be determined preoperatively, based on a combination of colonoscopic, ultrasonographic, and CT findings. Due to the risk of intraperitoneal contamination, it is not recommended to open the resected intestinal segment intraoperatively to assess excisional margins. After division, each colonic stump is oversewn with a continuous Parker-Kerr inverting suture (3-0/4-0 PDS); as an alternative, the two colonic stumps may be closed by stapling. The two stumps are then connected, leaving about 1 cm between them, with two to four 2-0 sutures (if four, two are diagonal and two are straight; Figure 7.32B). At this point two different techniques may be used 1. The first technique is the so called Swenson’s pull through (Swenson and Bill 1948; White and Gorman 1987; Holt and Brockman 2003; Hedlund and Fossum 2007b). The dog is positioned in dorsal recumbency. A second surgeon introduces an Allis tissue forceps into the rectum and both grasps and everts the distal rectum through the anus. This everted portion is resected circumferentially over 360 degrees (or the Parker-Kerr suture is removed) until the suture material connected to the proximal segment is exposed. This suture material is grasped and used as a stay suture; further stay sutures may be applied in the proximal colorectal portion as needed. Finally, the distal segment (bearing the tumor) is resected, and after removal of the Parker-Kerr suture from the proximal segment, an end-to-end anastomosis (in one or two layers, see above) is performed manually or with an EEA stapling device. 2. In a recently reported modified technique (Morello et al. 2008), steps are taken in order to achieve a further tension-free end-to-end rectocolonic anastomosis. After ligation of the appropriate vessels, the colon is divided according to the required excisional margins, the two stumps are oversewn and connected with sutures, and the abdomen is closed routinely. The dog is then positioned in sternal recumbency. A transanal rectal pull-through amputation is performed as previously described (see Figures 7.24D, 7.29C, and 7.32C). After excision of the distal colorectal stump bearing the tumor, the suture used to oversew the cranial stump is removed and the surgical procedure is concluded as previously described. The release of the stay sutures placed in the rectal cuff attached to the anus allows the anastomotic site to return to the pelvic canal. As an alternative, a stapling device may be used to perform an end-to-end anastomosis (as discussed below).

238  Veterinary Surgical Oncology

(a)

(b)

(c)

Figure 7.32.  (See also Figures 7.24D and 7.29.) (A) Two Doyen clamps are applied on the colon cranially and caudally to the proposed point of division. (B) After division, the two colonic stumps are closed with a continuous Parker-Kerr inverting suture and then connected with four 2-0 sutures, two diagonal and two straight, leaving a space of about 2 cm between the two intestinal stumps. (C) A transanal pull-through procedure is performed until the proximal stump emerges. At this point the procedure is terminated as in the transanal pull-through procedure. (Photograph C is from Buracco P. 2007. Tumori colorettali. 2007. In Oncologia del cane e del gatto G. Romanelli, editor. London: Elsevier. Used by permission.).

Postoperative care, potential complications, and extent of resection are described later Functional outcome is usually good, but complications may develop. Complicatons are described below. Metastasis resection Metastatic lesions are excised through a midline celiotomy (see also the section on perianal tumors). Care is taken to evaluate all potential sites where metastasis can be found, including the sublumbar (medial iliac, see Figures 7.24 and 7.53; hypogastric, Figures 7.33A and B, and see also Figure 7.54); colic (see Figures 7.24A, D); and mesenteric lymph nodes, liver, spleen, and omentum. If feasible, metastatic lesions are excised and submitted for histological examination, or alternatively, biopsy samples are taken when the lesions are infiltrative and/ or multiple. Intraoperative radiation, if available, may be performed in selected cases at the site of metastatic sublumbar lymphadenopathy. Surgical palliation The use of an incontinent end-on-colostomy as a fecal diversion technique for inoperable and obstructive

colorectal malignancies, or when there is failure of a pull-through colorectal amputation, has been reported (Hardie and Gilson 1997; Kumagai et al. 2003; Hedlund and Fossum 2007b). In this author’s experience, such procedures are rarely accepted by owners due to the difficultly in managing the animal. The permanent end-on-colostomy (Hedlund and Fossum 2007b) is performed after resection of the distal colon by creating a stoma at the level of the left abdominal wall and suturing the serosa of the proximal colon to the abdominal musculature with a 3-0 monofilament absorbable suture and a full-thickness suture between the colon and skin. A colopexy is then performed adjacent to the stoma to avoid herniation, and the abdomen is closed routinely. A fecal storage device is then attached to the stoma. A temporary fecal diversion may be performed as a “loop colonostomy” at the level of the left flank (Hedlund and Fossum 2007b). In this case, the descending colon is not resected but used for the colonostomy. For this technique, after applying the loop ostomy rod to stabilize the intestine at the level of the abdominal wall incision, the stoma is created by suturing the colonic seromuscular layers to the subcutis of the abdominal incision; the

Alimentary Tract  239

(a)

(b)

Figure 7.33.  (See also Figure 7.24.) (A) Enlarged metastatic sublumbar lymph nodes (hypogastric lymph nodes), one of which is pointed out with the tip of a cotton swab. (B) Excision of one of these lymph nodes.

colon is then incised longitudinally and the stoma is completed by suturing the colonic seromuscular layer to the skin. A fecal storage device is then attached to the stoma and the stoma closed when it is no longer required. To achieve palliation for extensive benign colorectal tumors and as an alternative to radical and full-thickness excision, a transanal endoscopic treatment has been recently proposed based on a retrospective study involving 13 dogs (Holt 2007). Tumor resection, even during recut procedures, was achieved by varied combinations of an electrode-cutting loop, an electrocautery with a ball-end electrode, and traditional surgery (the latter was only needed in a single dog after two endoscopic treatments). Results included a cure in 5 dogs, palliation in 3, and were poor in 5 dogs. Complications of the technique (arising usually 4–5 days after the treatment) included rectal perforation with peritonitis and death. Nonsurgical treatments Local radiotherapy has been used for small nonmetastatic rectal adenocarcinomas less than 3 cm in size and localized to the distal half of rectum and anal canal (Turrel and Theon 1986). The radiation was applied as a single high dose from an orthovoltage machine in one report (Turrel and Theon 1986). Perforation with subsequent peritonitis was reported in a second paper (Church et al. 1987). According to the first report, tumor control and survival rates at 1 year were 46% and 67%, respectively, with a median and mean tumor-free period

of 6 and 9.7 months, respectively and a median and mean survival times of 7 and 11.3, respectively. Complications reported included occasional tenesmus within 1–2 days of the procedure that usually resolved within 2 weeks (Turrel and Theon 1986). Perforation with subsequent peritonitis was reported in the second paper (Church et al. 1987), and no further use of this procedure has been reported. Radiation should only be recommended if the tumor can be exposed out of the body, is small-sized, and is able to be completely irradiated. However, it is this author’s opinion that surgical excision may be a faster option in many of these cases. Finally, in an experimental model using normal dogs undergoing proctectomy and stapling anastomosis, the possibility of treating colorectal recurrence with photodynamic therapy with motexafin lutetium was evaluated. The procedure is still under investigation; however, the preliminary results of this study show that the technique could be proposed as an adjuvant treatment together with chemotherapy and radiotherapy (Ross et al. 2006). Additional technical details for colorectal anastomosis The discussion below addresses all colorectal resections. For more information, see also specific sections. For manual end-to-end anastomosis trim the exuberant mucosa with Metzenbaum scissors, being sure to include all the intestinal layers in the sutures. More importantly, ensure that all the layers at least as deep as the submucosa, which is the holding layer of the suture,

240  Veterinary Surgical Oncology

are included. Engage slightly more serosa than mucosa (Hedlund and Fossum 2007b). Surgical stapling The stapling equipment used in the large intestine surgery include thoracoabdominal (TA), gastrointestinal (GIA), and circular staplers (EEA, CEEA) (Holt and Brockman 2003; Hedlund and Fossum 2007b; Tobias 2007, Banz et al. 2008). TA staplers (or even skin staplers, which are less expensive) may be used for an end-to-end intestinal anastomosis using a triangulating technique. The result is an everting suture. A TA linear stapler may also be used for typhlectomy and, transanally, for removal of polyps of the distal rectum with a base of attachment of less than 3 cm (Swiderski and Withrow 2009). GIA staplers are linear devices that are usually employed in small intestinal surgery. However, they may also be used for typhlectomy. The final result is an everting suture. Circular staplers (EEA, CEEA) may be used for both colonic and rectal end-to-end anastomosis and may be inserted in the anus (Banz et al. 2008) or through a separate small incision as appropriate (e.g., the cecum) (Kudisch and Pavletic 1993). Two purse-string sutures secure the two intestinal ends to be anastomosed to the anvil of the EEA device. After firing, a double row of circumferential full-thickness B-shaped titanium staples are applied. This results in a two-layer inverting anastomosis that may reduce the intestinal lumen; this technique has not resulted in any reported complications (Kudisch and Pavletic 1993). The incision site used to insert the instrument is sutured manually or with a TA stapler. All reported descriptions of large intestinal anastomosis in veterinary medicine using a circular EEA stapling device (Kudisch and Pavletic 1993; Holt and Brockman 2003; Hedlund and Fossum 2007b; Tobias 2007; Banz et al. 2008) use a double approach (abdominal–transcecal, transcolonic, etc., and abdominal or transanal) (see Figure 7.26B,C). A recent paper using a porcine model proposed a technique whereby transanal pull-through is achieved via sigmoidectomy using pneumoperitoneum, gastrotomy access to work intraperitoneally, a circular stapler inserted through the anus, a linear stapler introduced in the colon through a colotomy, excision of the sygmoid, and a stapled colocolonic anastomosis (Leroy et al. 2009). Advantages of intestinal stapling procedures include decreased surgical time, good approximation of the two intestinal stumps, good hemostasis without compromising vascularization, higher bursting pressure during the early stage of healing and higher tensile strength at 7 days postsurgery compared to hand-

sutured anastomosis, minimal inflammation and necrosis, and good hemostasis. Potential complications include stricture, adhesion, dehiscence and peritonitis (caused by excessive tension, poor blood supply, or inappropriate staple size), mucosal ulcerations (more frequent than after manual anastomosis), hemorrhage, transient anal dysfunction (mainly in cats), and rectovaginal fistula (Klein et al. 2006; Tobias 2007; Banz et al. 2008). Postoperative care Care requires analgesia (opioids and nonsteroidal antiinflammatory drugs), fluid, and electrolyte therapy, according to acid-base status of the animal, until the animal eats spontaneously (usually after 2–3 days). Antibiotics are administered only when there is established infection. The animal should be monitored for disseminated intravascular coagulation (DIC) and treated as appropriate. Elizabethan collars are used as needed. A small amount of water is offered 8–12 hours after surgery, and a small amount of food (e.g., Hill’s i/d) 12–24 hours after surgery if there is no vomiting and then 3–4 times a day, with return to normal feeding after 2–3 days. Stool softeners (e.g., lactulose) should be started when the animal starts to eat, and should be mixed with food. Functional outcome is usually good if complications do not occur. Potential complications Complications may include postoperative hematochezia and dyschezia (from 1–2 days up to 1–2 weeks after surgery), and tenesmus (up to 1–2 months; see next below). The duration may depend on the amount of the colorectal resection performed. These complications are usually self-limiting. Stricture formation and tenesmus may be present. Postoperative stricture may usually be felt by digital rectal exploration. It may be caused by excessive colorectal resection, excessive inflammation and/or inadequate blood supply (Goldsmid et al. 1993), improper anastomosis and/or suture material, and localized infection. Tenesmus can persist longer, and fecal softeners may be used for several weeks unless colitis and diarrhea are present (in the latter case, reservoir incontinence may be observed even though sphincteric continence has been preserved; see below). Sometimes the use of excessively stiff monofilament material for suturing may be the cause of a persistent colorectal inflammation and pain. If the problem persists, a colonoscopy is warranted in order to evaluate the severity of the lesion, to take a

Alimentary Tract  241

biopsy, and to decide if balloon dilation (bougienage) with or without a second surgery is required Infection may be present due to manipulation of the rectum with intraoperative spillage of fecal material. This should be avoided with appropriate surgical technique, the intraoperative use of antibiotics, abundant lavage and fluid aspiration, a change in surgical gloves and instruments if they become contaminated, and the use of surgical drains. Surgical drainage devices have been advocated when fecal contamination has occurred during surgery, despite the adverse effect they can have on anastomotic healing (Hedlund and Fossum 2007b); the interposition of fat between the anastomotic site and the soft latex drain has been proposed to minimize this effect (Holt and Durdey 1999). This author avoids the use of drains in surgical procedures of the colon and rectum and prefers to pay particular attention to minimizing intraoperative contamination combined with adequate preoperative patient preparation. The use of sponges packed in the intestinal stump helps to limit contamination that can occur after removal of the more controllable hard feces adjacent to a site of obstruction. In selected cases, such as when there is a substantial amount of fecal material demonstrated radiographically, colorectal resection may be preceded by colotomy and fecal emptying, particularly when abdominal exploration is performed for the inspection and removal of other masses (such as sublumbar lymph nodes) and/or simple colectomy. Postoperative dehiscence and infection may occur. The most common time for postoperative dehiscence is 3–5 days after surgery (Holt and Lucke 1985; Anderson et al. 1987). This is a very serious complication that can be fatal, particularly if it occurs within the pelvic canal. Potential causes include excessive tension at the anastomotic site, inadequate blood supply, improper technique, and inappropriate selection of suture material. It has been reported that the risk of dehiscence is greater with resections greater than 6 cm (Anderson et al. 1987; Phillips 2001). Fecal incontinence may also occur. The pathogenesis of fecal incontinence after rectal pull-through surgery is still debated. The two major factors contributing to fecal continence are external anal sphincter function (provided by the integrity of both the muscular component and the caudal rectal branch of the pudendal nerve) and reservoir continence (the ability of the descending colon to distend and store feces before voluntary expulsion; reservoir continence is provided by the integrity of both colonic length and motility) (Dean and Bojrab 1993). Other factors that seemingly contribute to continence include the length of the distal rectum preserved after

rectal resection and sparing of the rectal cranial peritoneal reflection (Anderson et al. 1987; Anson et al. 1988; Swenson and Bill 1948; Gaston 1948a, 1948b, Gaston 1961; Karlan et al. 1959). A minimum of 1–1.5 cm of distal rectum has been recommended to preserve fecal continence. This explains why the transanal approach may be a preferred option, provided that the extent of resection required to achieve excisional margins is guaranteed. In one recent study using a transanal approach in dogs, the distal rectum was spared. The dogs were all clinically continent, except one dog that had a resection at the anorectal junction and subsequently became clinically continent at 5 months postoperatively for no identifiable reason (Morello et al. 2008). With the rectal pull-through procedure or when the distal rectal resection is at the level of the rectocutaneous junction, fecal incontinence remains a risk due to a combination of the distal rectum being removed, iatrogenic neurological trauma to the caudal rectal branch of the pudendal nerve, and damage to the external anal sphincter muscle during dissection (Figure 7.34; see also 7.30 ). Fecal incontinence has also been associated with amputations involving more than 6 cm of rectum in combination with transection of the caudal peritoneal reflection (Anderson et al. 1987). In a recent paper, colorectal amputation of greater than 6 cm in length always included the peritoneal reflection; however, permanent fecal incontinence was not observed (Morello et al. 2008). Similar results also have been reported after anastomosis of the colon to the distal 1.5 cm of rectum (Swenson and Bill 1948). Furthermore, colorectal resections including limited parts of the descending colon and rectum with preservation of its distal part should have little influence on the fecal reservoir continence that may be lost after more extensive colonic resections make it impossible to store feces (Dean and Bojrab 1993; Gaston 1948a, 1948b, 1951, 1961; Karlan et al. 1959; Peck and Hallenbeck 1964). Clinical signs of fecal reservoir incontinence include more frequent conscious defecation (Guilford 1990). In general, this author believes that, provided the distal rectum has been spared, transient loss of fecal continence is generally associated with complications of wound healing, including inflammation/ colitis with diarrhea and stricture with tenesmus, particularly in small dogs that seem more prone to develop such problems. In small dogs, therefore, the aggressive colorectal surgical procedures should either be avoided or clearly discussed with owners when a “salvage” procedure is warranted. A short- to long-term follow-up is needed to thoroughly assess changes in continence (Sapin et al. 2006; Morello et al. 2008). In this author’s opinion, the key factors to increasing the probability of

242  Veterinary Surgical Oncology

(a)

(b)

Figure 7.34.  In this case the pull-through procedure was performed for an inflammatory disease. The last 1 cm of rectum was not spared (A). (B) The final result. This dog was permanently incontinent. Another dog with a similar surgery regained clinical fecal continence after 5 months for unknown reasons.

maintaining fecal continence include sparing the distal rectum, appropriate proper surgical technique and choice of suture material, absence of infection, avoidance of tension at the anastomotic site, and careful preservation of blood supply. The veterinary literature has not fully addressed the effect on fecal continence of anastomosis between the distal ileum and the distal rectum; in humans, fecal continence is often preserved in these cases, even though reservoir continence may be impaired or lost (Günther et al. 2003; van Laarhoven et al. 2004). With ileoanal anastomosis, however, fecal continence in dogs is lost (Karlan et al. 1959; Peck and Hallenbeck 1964). One study has suggested that in dogs, as in humans, there is an increased tendency to develop pigment gallstones after proctocolectomy, which is associated with an increase in the concentration of unconjugated bilirubin in gallbladder bile (Noshiro et al. 1996). It is not known if this is clinically relevant. Extent of resection There is apparently no limit to resectioning the large intestine beyond the caudal peritoneal reflection. Care is taken to ligate only those vessels that are definitively required to be ligated to maximize the preservation of blood supply to the remaining segment of bowel. The final goal should always be to achieve a tension-free endto-end anastomosis to avoid complications, including dehiscence, sepsis, and stricture formation. The combined procedures (the so-called Swenson’s pull-through and modifications described previously in this chapter),

carry a high complication rate and should be reserved for very extensive malignant tumors located at the mid and/or cranial third of the rectum requiring extension into the descending colon to achieve adequate excisional margins (Morello et al. 2008). In these cases, pubic/ ischial osteotomy may be an excellent alternative (Yoon and Mann 2008). The pubic/ischial osteotomy approach may offer more advantages for resection of malignant tumors over the combined procedures due to a more accurate vessel ligation, given that even a transient vascular insufficiency may result in healing and clinical complications of varying severity. Posttreatment prognosis Dogs Intestinal tumors account for less than 10% of all tumors (Selting 2007). The large intestine is more frequently affected and represents 36%–60% of all canine intestinal neoplasia. Colorectal tumors are more prevalent in male dogs (Holt and Lucke 1985; Birchard et al. 1986; Church et al. 1987; Patnaik et al. 1976). The reported mean age is around 7–8 years, with a mean weight of 30 kg (range, 3.7–57) (Seiler 1979; Holt and Lucke 1985; Church et al. 1987; Phillips 2001). More than half of colorectal cancers are malignant, with adenocarcinoma being the most prevalent malignant tumor (Cotchin 1959; Church et al. 1987; Patnaik et al. 1978). Other less common colorectal neoplasia include leiomyoma (McPherron et al. 1992); leiomyosarcoma (Brueker and Withrow 1988; Kapatkin et al. 1992; Bagley et al. 1996; Cohen et al. 2003); plas-

Alimentary Tract  243

macytomas (Kupanoff et al. 2006; Rannou et al. 2009); and carcinoids (Patnaik et al. 1980; Sykes and Cooper 1982; Selting 2007). Adenomatous polyps account for the majority (up to 50%) of benign rectal tumors (Holt and Lucke 1985; Birchard et al. 1986). In situ carcinoma (Tis), a transition between adenomatous polyp and invasive carcinoma, has been reported to have histological evidence of atypia that may progress to malignancy in 17%–50% of cases (Seiler 1979; Patnaik et al. 1980; Holt and Lucke 1985; Birchard et al. 1986; Valerius et al. 1997; Danova et al. 2006). After marginal resection, the recurrence rate has been reported to be as high as 55% for carcinoma in situ and 17% for adenomatous polyps (Valerius et al. 1997; Phillips 2001). Local resection yields a survival of 5–24 months (Seiler 1979). A minimum 2 cm margin of resection is recommended for excision of both these lesions because polyps excised marginally may recur and progress to malignancy (Morello et al. 2008). In most clinical situations, these lesions are often first marginally resected (with excisonal biopsy; see the discussion on simple excision above), and if the histopathology identifies an adenomatous polyp, periodical monitoring is warranted. If there is recurrence of the mass, a more radical surgery is indicated when a biopsy confirms malignant progression (Morello et al. 2008). It should be remembered, however, that endoscopic biopsies may be too superficial and fail to demonstrate malignant progression such as a carcinoma in situ. For this reason a resection margin of a minimum of 2 cm (see also discussion on adenocarcinoma) is recommended for excision of both polyps and carcinoma in situ previously excised marginally if recurrence occurs (Morello et al. 2008). Canine adenocarcinoma is described as nodular (single or multiple), pedunculated, or annularconstrictive (Patnaik et al. 1980; Church et al. 1987; Phillips 2001). It has been reported that pedunculated or polypoid lesions have a good prognosis after surgical resection, whereas the annular colorectal adenocarcinoma is characterized by the worst prognosis. The reported metastatic rate for rectal adenocarcinoma ranges from 0% to 80% (Patnaik et al. 1980; Church et al. 1987). The reported mean survival time for malignant colorectal carcinomas varies from 6 to 14 months (White and Gorman 1987; Williams and Niles 2005), to 22 months after surgical resection, and 24 months after cryosurgery (Church et al. 1987). Recommended margins of resection vary among authors: from 1 to 2 cm, either for rectal polyps or malignancy, to a minimum of 2 cm to 4 to 8 cm for malignancies of both the small and large intestine (Palminteri 1966; Crawshaw et al. 1998; Phillips 2001; Aronson 2003; Williams

and Niles 2005; Danova et al. 2006). A modified TNM system for canine colorectal adenocarcinoma has been reported (Turrel and Theon 1986), and a further modification of this TNM system, with respect to the T grading, has been recently proposed by Morello et al. (2008). The newer scheme include is as follows: T0 (no evidence of tumor); Tis (in situ carcinoma—mucosal; intraepithelial or invasion of the lamina propria); T1 (tumor in mucosa and submucosa only); T2 (tumor extending to muscularis and serosa); and T3 (tumor extending to a contiguous structure). This modification has been tentatively correlated to margins of resection and survival: eight adenocarcinomas and two Tis were removed with 3–6 cm of grossly normal tissue on both sides of the resection, and one Tis was removed with 2 cm of grossly normal tissue on both sites of the resection. Postoperative recurrence and metastatic rates for adenocarcinoma were 18.2% and 0%, respectively. Median disease-free interval and survival times were not reached. Mean disease-free and overall survival times were 44.3 and 44.6 months (range, 0–75 months), respectively. Applying the modified staging system, long survival times are therefore expected after complete surgical excision of TisN0M0 and T1N0M0 adenocarcinoma. Resection margins of a minimum of 5 cm are recommended given that in one case incomplete margins were detected at 5 cm (Morello et al. 2008). This principle should be applied for any intestinal malignancy; however, for colorectal tumors this degree of excision is not always feasible. A recent paper compared canine and human colorectal cancers and demonstrated a strong degree of genetic homology in terms of copy number alternatives (CNAs), suggesting a high probability of these genetic alterations being cancer causative rather than passenger changes (Tang et al. 2010). A recent preliminary study on the characterization of the expression pattern of claudin tight junction proteins (implicated in epithelial cellular adhesion and therefore in colorectal carcinogenesis) in canine colorectal cancer has been published (Jakab et al. 2010). Leiomyoma and leiomyosarcoma are more frequent in medium to large canine breeds; the age of affected dogs normally ranges from 9 to 11 years (the latter mainly in the case of cecal leiomyosarcoma) (Maas et al. 2007); however, dogs younger than 2 years have been reported with these tumors (Holt and Lucke 1985). Even if benign, leiomyoma may cause obstruction depending on its size and location (Katamoto et al. 2003) (Figure 7.35A,B). Rarely, leiomyosarcoma may be associated with paraneoplastic hypoglycemia (Bagley et al. 1996). Leiomyosarcomas (see Figures 7.21 and 7.22A) are invasive but slow to metastasize. They may occur both in the small

244  Veterinary Surgical Oncology

(a)

(b)

Figure 7.35  (A) Blunt undermining of a leiomyoma in a dog through a dorsal approach to rectum. The tumor was dorsal to the rectum and caused marked rectal obstruction. (B) Postexcision aspect of the tumor.

and large intestine (most commonly in the cecum). Gastrointestinal stromal tumors (GISTs) have been reported in dogs and are differentiated from leiomyoma and leiomyosarcoma only after immunohistochemistry (LaRock and Ginn 1997; Frost et al. 2003; Selting 2007; Maas et al. 2007). Histological malignancy in dogs is more often observed in cecal GISTs; in these cases, perforation and peritonitis are also more frequent (Maas et al. 2007). In humans, GISTs are the most common mesenchymal tumor of the gastrointestinal tract, and their origin from the interstitial cell of Cajal and distinctiveness from smooth muscle tumors have been recently documented. At present, their treatment is based on surgery (when feasible) and specific inhibitors of KIT tyrosine kinase function (e.g., imatinib mesylate) due to a positive immunohistochemical staining for KIT protooncogene mutations. These inhibitors may be used either as adjuvant treatment or palliation for unresectable and metastatic tumors (Gold and Dematteo 2006). The same positive immunohistochemical staining for KIT proto-oncogene mutations in dogs (Frost et al. 2003; Maas et al. 2007) should justify, in selected cases, the use of such KIT tyrosine kinase inhibitors in this species. Surgical margins for leiomyosarcoma and GIST carry the same rules as those applied to adenocarcinoma; however, leiomyoma may be resected marginally. Long disease-free intervals are expected after surgery for smooth muscle neoplasms. Reported median survival (with or without metastasis) is about 21 months; the reported 1- and 2-year survival rate is 75% and 66%, respectively (Brueker and Withrow 1988; Kapatkin et al. 1992; Cohen et al. 2003). No statistical difference was

found when GISTs were compared with non-GIST smooth muscle tumors. In a recent publication, the 1and 2-year recurrence-free periods for cecal tumors was 83.3% and 61.9%, respectively. Interestingly, both castrated and spayed dogs showed a longer survival (Maas et al. 2007). Plasmacytomas may be secretory, resulting in hyperproteinemia and a monoclonal gammopathy (Trevor et al. 1993). They are characterized by slow growth and lack of recurrence after complete excision (Kupanoff et al. 2006). Whereas the definitive margin of resection is not known, a 1–2 cm margin of macroscopically healthy tissue is generally recommended to surround the tumor. Carcinoids are rarely reported and may metastasize (Sykes and Cooper 1982; Patnaik et al. 1980; Selting 2007). If complete excision is attempted, the same recommendations applied to any colorectal malignancy are followed. One case of rectal ganglioneuroma has been reported in a dog that was still alive 2.5 years after excision (Reimer et al. 1999). Cats Most reported information for colonic neoplasms in cats comes from the publication by Slawienski et al. (1997). Colonic neoplasms account for less than 1% of all feline tumors and 10%–15% of all alimentary tumors. In this species the small intestine is more frequently affected than the large intestine (Selting 2007). The reported mean age is 12.5 years, and a predisposition has been reported in Siamese cats. The most prevalent neoplasms include adenocarcinoma, lymphoma (many of

Alimentary Tract  245

which are localized at the ileocecal valve and can result in ileocolonic intussusception and rectal prolapse) (Demetriou and Welsh 1999); mast cell tumors; and carcinoids (neuroendocrine carcinoma). Hemangiosarcomas have also been described in cats (Sharpe et al. 2000). Surgery is recommended whenever possible to increase survival times, and resection margins should comprise 2.5–5 cm of macroscopic healthy tissue and include excision of any metastatic lesion. Prognosis is poor, as all the four cats presented in one paper died or were euthanased within 1 week (Sharpe et al. 2000). For feline colonic adenocarcinoma, the reported median survival after complete excision is about 138 days (range, 119–314). Metastasis has been associated with poorer prognosis. Chemotherapy with doxorubicin may increase survival times (Slawienski et al. 1997). Large intestinal mast cell tumors are rare in cats. Metastasis at presentation to the colic and mesenteric lymph nodes and the liver is frequent. Complete surgical excision should be attempted for isolated lesions. The reported median survival is 199 days (range 69–412 days) (Slawienski et al. 1997). Feline large intestinal lymphoma may be metastatic at presentation to colic and mesenteric lymph nodes, the liver, and the kidney. The role of both surgery and chemotherapy is unknown (Slawienski et al. 1997), but surgery may be indicated when the lesion is isolated. Carcinoid tumors (neuroendocrine carcinoma) are very rarely reported. In the two feline cases reported by Slawienski et al. (1997), both were metastatic at presentation to the colic and mesenteric lymph nodes, liver, and the peritoneum. If complete excision is to be performed, the same recommendations applied to other colorectal malignancies are followed. Adjuvant therapies The administration of cyclooxygenase-2 (COX-2) in­­ hibitors (e.g., piroxicam) has been suggested in dogs with polyps and malignant epithelial colorectal cancers (Knottenbelt et al. 2000a, 2000b, 2006). In humans, their prolonged administration has shown promising results in preventing the progression of epithelial colorectal cancer, but serious cardiovascular side effects may arise from these treatments (Bertagnolli 2007). Adjuvant doxorubicin for feline colonic adenocarcinoma may lead to a median survival time that is longer (280 days; range, 210–354) in comparison with cats that did not receive doxorubicin (56 days; range, 2–259) (Slawienski et al. 1997). Chemotherapy (doxorubicin or mitoxantrone) and radiotherapy have been suggested for both canine leiomyosarcoma and adenocarcinoma, but their efficacy in

large clinical trials has not been demonstrated (Ogilvie et al. 1991; Cohen et al. 2003). Intraoperative radiotherapy, if available, may be recommended for the excision sites of metastatic sublumbar lymph nodes. The use of specific inhibitors of KIT tyrosine kinase function can be considered in cases of inoperable and/or metastatic GISTs, which are positive to KIT proto-oncogene mutations.

Perianal Tumors Surgical procedures Surgical procedures for perianal tumors include castration only, cytoreduction or marginal excision of perianal hepatoid adenoma in conjunction with castration, en bloc excision of any perianal malignancy with or without anoplasty (rectocutaneous suture), mono- or bi-lateral anal sac removal, and metastasis resection. Clinical workup and biopsy procedures The clinical work-up and the biopsy principles for perianal tumors include a complete physical examination as well as the following. Digital rectal examination. Perianal hepatoid tumors originate from the circumanal modified sebaceous glands that are called hepatoid due to their resemblance to liver cells on cytology and histology. They are usually visible and palpable (Figures 7.36 and 7.37), and in the case of malignancy, circumferential growth can result in palpable stenosis (Figure 7.38). Stricture in this area, however, is not pathognomonic for neoplasia as it may occur secondary to trauma and inflammatory diseases such as perianal fistulas (particularly in German shepherd dogs) and infection. Anal stricture can also be idiopathic as a result of anorectal spastic contraction. The latter condition, observed also by this author, is very rare and seen more frequently in German shepherd dogs; it usually disappears under general or epidural anesthesia (Niebauer 1993). Tumors of anal sacs are only sometimes visible and can be palpated during rectal examination at the 4 o’clock and 8 o’clock positions (Figure 7.39), commonly as an incidental finding (Williams et al. 2003). This maneuver may facilitate performing a fine-needle aspiration biopsy (FNAB). During digital rectal examination, it is also important to gain information on the degree of fixation of the tumor to the surrounding tissues. Finally, it may be useful to detect any so-called sublumbar lymphadenomegaly (see also the discussion on “colorectal tumors”).

246  Veterinary Surgical Oncology

(a)

(b)

(c)

Figure 7.36.  Three different clinical presentations of canine hepatoid adenoma: (A) Single perianal adenoma; (B) “invasive” perianal adenoma; and (C) multiple perianal adenoma. These lesions may only be correctly diagnosed by biopsy and histological examination.

(a)

(b)

Figure 7.37.  (A, B) Two different presentations of canine perianal adenocarcinoma. (B) Adenocarcinoma metastatic to the sublumbar lymph nodes.

This includes the sacral (these do not actually drain the perianal region and are present in only 50% of dogs), hypogastric, and medial iliac lymph nodes (Bezuidenhout 1993). Complete laboratory workup (blood, urine). Possible changes observed include hypercalcemia, in 25%–50% and up to 90% of cases with anal sac adenocarcinoma and is occasionally seen in cases of perianal adenocarcinoma (White and Gorman 1987; Berrocal et al. 1989;

Rosol et al. 1990; Ross et al. 1991;, Bennett et al. 2002; Williams et al. 2003). Secondary hypophosphatemia and increases in renal parameters can be seen after prolonged hypercalcemia. In anal sac adenocarcinoma, hypercalcemia is caused by a parathyroid hormone-related protein (PTHrp) produced by the tumor (Rosol et al. 1990; Gröne et al. 1994); in perianal adenocarcinoma, the cause of hypercalcemia is uncertain and the author has not observed this finding in any clinical cases.

Alimentary Tract  247

Figure 7.38.  Hepatoid perianal adenocarcinoma in an 11-yearold male German shepherd causing anal stenosis.

Figure 7.40.  Lateral abdominal radiograph which shows an enlargement of the sublumbar lymph nodes secondary to an anal sac adenocarcinoma in a dog. The ventral aspect of both the sacrum and 7th lumbar vertebra show some productive changes. (Thanks to Dr. Sheldon Padgett for this picture).

adenocarcinoma). The clinical differentiation between hepatoid adenoma and adenocarcinoma (as well as perianal fistulas) may also be difficult. For this reason, biopsy is mandatory. Indeed, diagnostic information is often provided by cytology for other tumor types found in this region.

• FNAB from enlarged sublumbar lymph nodes can be

aspirated during a digital rectal exploration (see also section on colorectal tumors; see also Figure 7.20) or via transabdominal ultrasonographic-guided aspiration (Llabrés-Diaz 2004) Biopsy samples from perianal tumors can be obtained • by using a Tru-cut needle (needle core biopsy), which may require sedation, punch biopsy (also done without anesthesia if the lesion is already ulcerated), incisional biopsy, or excisional biopsy. Imaging techniques Figure 7.39.  Right anal sac adenocarcinoma in a dog.

Cardiologic examination. This examination is particularly important in hypercalcemic dogs Tumor biopsy. FNAB and cytological examination can confirm the hepatoid nature of the tumor without, in many cases, giving an exact diagnosis (adenoma vs.

There are several imaging techniques used to stage perianal tumors. Lateral radiographic evaluation of the abdomen may be useful to assess for sublumbar lymphadenomegaly (Figure 7.40). In selected cases, urethrocystography may be useful to outline urethral compression by an exceedingly large metastatic sublumbar lymph node (Hoelzler et al. 2001).

248  Veterinary Surgical Oncology

Ultrasound examination of the abdomen is more useful than radiology to evaluate the sublumbar region (Llabrés-Diaz 2004), liver, spleen, and all the other abdominal organs. During the procedure, transabdominal ultrasound guided FNAB (using a 22- or 20-gauge spinal needle) or biopsy (using a 18-gauge Tru-Cut biopsy needle) may be performed. Interestingly, even though not demonstrated histologically, all sublumbar lymph nodes greater that 1 cm have shown neoplastic progression in one report (Polton and Brearley 2007). The finding of large vessel infiltration in the sublumbar area from metastatic lymph nodes may be a contraindication for surgery; in this case palliative procedures such as radiotherapy with or without chemotherapy should be considered (Polton and Brearley 2007). Ultrasound examination of both testicles can be used to evaluate the presence of lesions that are not palpable

clinically (frequently, this would be an interstitial cell tumor concomitant with a hepatoid adenoma). Radiographic evaluation of the thorax is performed with the standard three views (two lateral, one dorsoventral) to assess for lung metastasis, which are more commonly seen with anal sac adenocarcinoma and malignancies other than those of hepatoid origin (Figure 7.41). A case of lung metastases caused by an anal sac adenocarcinoma associated with paraneoplastic hypertrophic osteopathy in a dog has been recently reported (Hammond et al. 2009). Laparoscopy, if available and in the hands of an experienced operator, can be useful to inspect the entire abdomen and the sublumbar space, allowing for biopsy collection. CT can be used to assess the lungs, presence of lymphadenomegaly, and abdominal organs. In the case of sublumbar lymph node involvement, CT provides valuable information regarding tumoral involvement of great vessels and surgical resectability of nodes (Figures 7.42 and 7.43). Bone scintigraphy may be used in selected cases of anal sac adenocarcinoma to assess for bone metastases. Surgical techniques and procedures

Figure 7.41.  Multiple lung metastases from an anal sac adenocarcinoma in a dog. Metastasis is also present in a humerus.

(a)

Surgery of the perianal region may be performed traditionally, with a CO2 laser, or only for small lesions using cryosurgery (Dow et al. 1988; Liska and Withrow 1978; Shelley 2002). Histology should be performed on all surgical specimens. The evaluation of the excisional margin is not possible if laser or cryosurgery are used, and this is considered acceptable only for benign lesions (Turek and Withrow 2007); testicles and lymph nodes that are excised are also submitted for histology. Gloves

(b)

Figure 7.42.  (A) CT scan of a large sublumbar lymph node that caused a significant obstruction of the colorectum at the pelvic inlet in a dog. Metastatic lymphadenomegaly originated from a perianal adenocarcinoma. (B) At necropsy, the tumor surrounded the blood vessels; a thrombus is also evident in the caudal vena cava (arrow).

Alimentary Tract  249

(a)

(b)

(c)

Figure 7.43.  (A) CT scan of a large medial iliac lymph node secondary to a perianal adenocarcinoma. The great vessels are peripheral. (B) Intraoperative view after node resection (top of the picture is cranial). (C) Postoperative macroscopic view of the node.

and surgical instruments are changed after the tumor has been excised, and at any time contamination from the anus, anal sacs, and/or rectum has occurred.

and Hayes 1979). It should also be noted that a testicular interstitial cell tumor may be present in combination with a perineal hernia.

Castration

En bloc excision with or without anoplasty

Castration may be the sole procedure required in intact males for small, nonulcerated, and histologically confirmed hepatoid adenomas (perianal and/or tail gland) as it induces progressive tumor shrinking. As an alternative, an incisional or excisional biopsy is performed, with histology as a guide to plan further steps (clinical observation over a period of 2 months, marginal or en bloc excision with or without castration, depending on the histological result).

A minimum of 1–3 cm of macroscopically normal tissue should be included around the tumor, depending on the tumor type; reconstruction often requires suturing the rectum to the skin (Figure 7.46). En bloc excision in this region includes removal of variable portions of perianal/perineal skin, part or all of the external sphincter muscle, one or both anal sacs (Figure 7.47), part or all of the anal canal up to the distal rectum (pull-through), and in some instances amputation of the tail (Figure 7.48). The reconstruction may warrant the use of local skin advancement or transposition flaps (Figure 7.49). Moistened gauze sponges are packed into the rectum to avoid the spillage of fecal material into the pelvic canal. In a standard anoplasty, closure may be performed in one or two layers with simple interrupted sutures. The single-layer suture pattern approximates the submucosa/ mucosa to the skin (using 3-0 or 4-0 absorbable braided, e.g., polyglactin 910, or monofilament material); in the two-layer suture pattern, absorbable monofilament material (3-0 or 4-0 polydioxanone, polyglyconate, poliglecaprone 25) is used for the first layer (adventitiamuscularis of the rectum/subcutaneous tissue), and in the second layer the epithelial lining (skin/rectal submucosa-mucosa or skin/skin depending on the

Cytoreduction or marginal excision This technique is used for histologically confirmed hepatoid adenoma (perianal and/or tail gland) together with castration in intact males (Figures 7.44 and 7.45). In females, marginal excision of perianal adenoma, usually of small size, may be sufficient. In this author’s opinion, the use of either a braided or monofilament absorbable suture material is suitable in the perianal region; however, the latter is somewhat stiffer and can cause discomfort to the animal. Correction of perineal hernia Correction of perineal hernia may be required as it is observed in 10% of dogs with perianal adenoma (Wilson

250  Veterinary Surgical Oncology

(a)

(c)

(b)

(d)

Figure 7.44.  (A–D) Marginal excision of a large perianal adenoma. (Photos from Buracco P. 2007. Tumori perianali. 2007. In Oncologia del cane e del gatto G. Romanelli, editor. London: Elsevier. Used by permission.).

excision performed) can be approximated with both 3-0 or 4-0 absorbable braided (e.g., polyglactin 910) or monofilament material. The sutures must be applied with as minimal tension as possible (Aronson 2003). This author generally uses a single-layer closure and reserves the two-layer closure for cases with excessive tension. Anal sac removal If hypercalcemia is present, this should be treated before surgery using parenteral fluid therapy, furosemide, prednisone, bisphosphonate, and calcitonin. Anal sacculectomy for neoplasia is usually performed unilaterally, but it should be noted that anal sac adenocarcinoma may also be bilateral in rare cases (Ross et al. 1991; Emms

2005; Turek and Withrow 2007). As fecal continence is a concern, marginal excision may be the preferred choice in the case of bilateral anal sac tumors (Emms 2005). Prolonged survival can be achieved even when concurrent excision of metastatic regional lymph nodes is performed (Hobson et al. 2006; Polton et al. 2007). Anal sac adenocarcinoma occurs rarely in cats (Chun et al. 1997; Mellanby et al. 2002; Parry 2006) and excision is performed in a similar manner to anal sac tumors in dogs (Figures 7.50 and 7.51). Anal sac adenocarcinoma may also develop at the level of the controlateral anal sac some time subsequent to the removal of an anal sac tumor on one side (Turek and Withrow 2007; Emms 2005). Excision is often performed marginally, but en bloc excision (see previous section) is attempted in all the cases where this is

Alimentary Tract  251

(a)

(b)

Figure 7.45.  (A) Multiple perianal adenomas. (B) postoperative view after marginal excision.

(a)

(b)

Figure 7.46.  Phases of anoplasty. (A) The anal region has been excised, and (B) the rectum has been sutured to the skin. As the external sphincter muscle has been removed, this dog will be permanently incontinent. This issue should be discussed with the owner before surgery.

possible and is usually dependent on the size of the primary tumor. Resection of metastases Resection of metastases refers mainly to the sublumbar lymphadenectomy through a midline celiotomy and is performed principally in cases of anal sac adenocarci-

noma (Hobson et al. 2006; Polton and Brearley 2007), but less commonly in cases of perianal adenocarcinoma or other malignancies such as squamous cell carcinoma, malignant melanoma, and various sarcomas. More than one sublumbar lymphadenectomy is often performed (Hause et al. 1981; Hobson et al. 2006) (Figure 7.52).

252  Veterinary Surgical Oncology

(a)

(b)

(c)

Figure 7.47.  Phases of the excision of a soft tissue sarcoma in a dog. (A) Clinical appearance of the tumor. (B) The surgical excision has been completed, and variable amounts of skin, fascia, and muscles (including part of the external sphincter) have been removed. (C) Final appearance of the wound (the use of skin flaps was not required in this case). The dog experienced transitory fecal incontinence.

(a)

(b)

Figure 7.48.  (A) The clinical presentation of a recurrent perianal adenocarcinoma. (B) To achieve sufficient margins of macroscopically normal tissue, the tail was amputated. For reconstruction, the dorsal skin of the tail was spared and used as a skin flap.

Identification of enlarged sublumbar lymph nodes to be removed (see also the section on colorectal tumors) is facilitated by retracting the descending colon on one side and palpating the pulse of the descending aorta. The medial iliac lymph nodes are identified between the deep circumflex and external iliac arteries at the level of the fifth and sixth lumbar vertebrae (Figure 7.53), and

the hypogastric lymph nodes are identified between and in close association with the external and internal iliac arteries ventral to the sixth and seventh lumbar vertebrae (Bezuidenhout 1993) (see also Figures 7.33 and 7.54). Lymph node removal is performed with careful undermining and dissection and in some cases firm

Alimentary Tract  253

Postoperative care

adhesion of the nodes to either the sublumbar musculature or a vessel wall makes excision very risky due to hemorrhage (Ross et al. 1991) or even be impossible (see Figure 7.42). One single report has described an unresectable metastatic cystic iliac lymph node that was omentalized with palliation for 18 months (Hoelzler et al. 2001). Removal of metastatic nodules in the liver, spleen, and lung is occasionally performed for these tumors.

Systemic analgesics are recommended for 12–72 hours and food and water given within 8–12 hours, depending on the surgical procedure performed. An Elizabethan collar is recommended to prevent self-trauma, and ionized calcium levels should be monitored once a day during the first 48 hours when preoperative hypercalcemia has been documented. Normalization of ionized calcium levels usually occurs within 24 hours when adequate resection has been performed. In the case of anoplasty, the perianal region should be lubricated with petrolatum until the sutures are removed. Cosmetic and functional outcome Poor cosmesis may be a concern in some cases of anoplasty with or without tail amputation. Functionally, the main concern is fecal incontinence (see below). Potential complications Hypocalcemia, while rare (Williams et al. 2003), is treated with 10% calcium gluconate intravenously with monitoring of cardiac rhythm during administration (Hedlund and Fossum 2007a). Hematochezia can be seen for 8–48 hours or longer, depending on the surgical procedure performed. This complication is usually self-limiting. Transient tenesmus is usually self-limiting within 1–5 days, depending on the surgical procedure performed. It may be associated with poor pain control. Dehiscence may occur when an anoplasty is performed with excessive tension or infection is present secondary to the manipulation of the rectum and anus.

Figure 7.49.  Advancement flap in a rectocutaneous plasty. These flaps have the tendency to dehisce when one of its borders is adjacent to the base of the tail due to movement. In this case, there was dehiscence and the wound was allowed to heal by second-intention healing.

(a)

(b)

Figure 7.50.  (A) marginal excision of an anal sac adenocarcinoma in a dog. (B) Macroscopic appearance of the tumor.

254  Veterinary Surgical Oncology

(a)

(d)

(b)

(c)

(e)

(f)

Figure 7.51.  Clinical (A) and CT view (B) of a large anal sac adenocarcinoma in an 11-year-old spayed bitch that was hypercalcemic at presentation. (C,D) Intraoperative pictures of the marginal resection of the primary tumor. (E) The surgical site after closure of the wound. (F) The macroscopic view of the tumor.

If possible, the wound should be resutured to avoid stricture, otherwise second-intention healing can be used, which necessitates frequent wound management, including warm hydrotherapy two to four times per day (Figures 7.55 and 7.56). Skin flap dehiscence is more likely if one border of the flap is adjacent to the to the base of the tail, as tail movement retards flap healing (see Figure 7.49). In dogs undergoing sublumbar lymphadenectomy, septic shock has been reported (Ross et al. 1991; Williams et al. 2003). Temporary urinary incontinence may be a complication of sublumbar lymphadenectomy because of damage to the innervation of the bladder during resection (Ross et al. 1991). Stenosis and strictures occur mainly after anorectal surgery, including in many cases also anal sac removal.

Figure 7.52.  Macroscopic view of several sublumbar lymph nodes after excision. One of these was removed piecemeal. Removal of all nodes (together with the primary tumor) is essential if the dog is hypercalcemic in order to normalize calcemia.

Alimentary Tract  255

Figure 7.53.  Intraoperative appearance of the excision of an enlarged medial iliac node in a dog with anal sac adenocarcinoma. (Photos from Buracco P. 2007. Tumori colorettali. 2007. In Oncologia del cane e del gatto G. Romanelli, editor. London: Elsevier. Used by permission.).

(a)

Figure 7.54:  Region of the hypogastric lymph nodes after excision.

(b)

Figure 7.55.  (A) This male German shepherd dog experienced dehiscence of the closure. (B) The wound was not sutured and was managed conservatively for 2 weeks.

Anal stricture may be the result of a poor epithelial apposition during surgery or excessive tension and infection with further dehiscence followed by a secondintention healing. It results in protracted tenesmus. Treatment options include bougienage, a simple incision, or a wedge resection. If the problem persists, en bloc excision should be performed Fecal incontinence can occur following anoplasty (Ross et al. 1991). If only half of the circumference of the sphincter is removed, incontinence may be transient

(see Figure 7.47B) (Turek and Withrow 2007). If resection involves a complete 360-degree circumference (Figure 7.57; see also Figures 7.46, 7.49, 7.55, 7.56), fecal incontinence is likely and should be discussed with the owner before surgery. Provided that tissue undermining is performed as close as possible to the anorectal wall, continence may be partially preserved (Figure 7.58), especially if a solid stool consistency is maintained with a low-residue diet.

256  Veterinary Surgical Oncology

(a)

(b)

Figure 7.56.  (A) Anoplasty, (B) following dehiscence healing, has occurred via second-intention healing.

Figure 7.57.  Incontinent dog 2 years after surgery (compare to Figure 7.46).

This may be acceptable for many owners but this clinical result is difficult to predict preoperatively. A possible explanation of preserved and progressively acquired partial fecal continence relies on the fact that the healing process provides adhesions between the rectum and adjacent tissues and muscles of the pelvic diaphragm, including the levator ani, coccygeus, retractor penis or constrictor vulvae muscles, and coccygeal fascia (Lewis 1968). It should also be noted that in a normal dog, all

Figure 7.58.  Anoplasty in a German shepherd dog 10 days after surgery. In this case, the external sphincter muscle was spared, and after a transient period of fecal incontinence, the dog was able to consciously retain feces.

of these structures are anatomically connected to some extent with the external sphincter muscle. Recurrence of the perianal tumor rarely occurs when an adenoma is treated with castration with or without excision of the primary tumor. Recurrence in a castrated

Alimentary Tract  257

(a)

(b)

Figure 7.59.  (A) Ulcerated hepatoid adenoma of the “tail gland” in a dog. The tumor was marginally resected and castration was performed. (B) Hepatoid adenocarcinoma of the tail in an 11-year-old male German shepherd. No metastasis was found, and the tail was amputated.

male dog suggests possible malignancy, and further diagnostic investigation is warranted (Turek and Withrow 2007). Recurrences are likely in cases of incompletely excised perianal and anal sac adenocarcinoma. As noted previously, a second anal sac adenocarcinoma may rarely develop at the level of the contralateral sac subsequent to the removal of an anal sac adenocarcinoma on one side (Emms 2005; Turek and Withrow 2007). Additional metastases are more likely in anal sac adenocarcinoma and occur less frequently with perianal adenocarcinoma. With malignant perianal tumors, clinical, laboratory, and diagnostic imaging evaluations are recommended every 3 months during the first year following surgery and every year thereafter. In cases of anal sac adenocarcinoma, disease progression may be in form of local recurrence and/or further metastasis and hypercalcemia. Common perianal tumors and prognosis This discussion refers only to dogs, as cats have only anal sacs and not perianal hepatoid glands. Most tumors originate from the modified sebaceous hepatoid glands located in the perianal or circumanal region (see Figures 7.36, 7.37, 7.38), tail (Figure 7.59), prepuce, trunk, and hind leg. These glands are also present in bovines (Blazquez et al. 1988).

Their function, both in carnivores and bovines, is unknown; however, it has been suggested the perianal or hepatoid glands may be odor-producing glands via proteins that may act as an olfactory marker (Shabadash and Zelikina 1995; Martins et al. 2008). Immunohistochemistry for growth hormone was found to be positive in 23 of 24 canine perianal adenomas and in 5 of 5 perianal adenocarcinomas (Petterino et al. 2004). Canine benign tumors are referred to as hormonally dependent perianal adenomas as they can be stimulated by androgens or inhibited by estrogens (Turek and Withrow 2007). Hence, they regress in male dogs after castration. Perianal adenomas occur more frequently in elderly intact males and less frequently in spayed than intact bitches (Figure 7.60). Hyperadrenocorticism can be a concurrent finding in female dogs, and the adrenal gland is the source of androgenic stimulation (Dow et al. 1988; Hill et al. 2005). The vast majority of perianal adenomas are welldifferentiated tumors, with only about one quarter being moderately or poorly differentiated neoplasms (Berrocal et al. 1989; Vail et al. 1990; Turek and Withrow 2007). More recently, in a review of 240 perianal tumors (Martins et al. 2008), the Goldschmidt et al. (1998) histological classification was applied and compared to the older classification produced by Berrocal et al (1989):

258  Veterinary Surgical Oncology

Perianal adenocarcinoma

Figure 7.60.  Perianal adenoma in a spayed bitch. This followed a similar excision 9 months previously for the same lesion.

hyperplasia was diagnosed in 4% of cases, well and moderately differentiated adenoma in 54% of cases (20% and 34%, respectively), poorly differentiated adenoma (or hepatoid gland epithelioma) in 19% of cases, carcinoma in 20% of cases, and other tumors (of mesenchymal/vascular origin) in 13% of cases (Martins et al. 2008). Independently of the histological classification adopted, gland differentiation (that also implies the presence of androgenic receptors on the surface of cells) is associated with both hormone-dependence and favorable prognosis after surgery. Martins et al (2008) suggest that immunohistochemistry for PCNA (proliferating cell nuclear antigen) may be used, together with histomorphology, to distinguish benign and malignant tumors of the perianal gland (the cut-off value of the PCNA index suggested is about 0.60%); the apoptosis index should follow a similar trend given that whereas carcinoma cells divide more frequently than adenoma cells, they also have a higher death rate. In this study, the net growth was expressed as a net growth index, correlating with both the PCNA and apoptosis indices. Another recent paper has proposed a differentiation between hepatoid adenoma versus carcinoma based on the use of mouse monoclonal antibodies (Ganguly and Wolfe 2006). Marginal excision and castration is the treatment of choice for perianal adenomas (Turek and Withrow 2007), and recurrence is rare, particularly in male dogs (Wilson and Hayes 1979).

These tumors are relatively slow growing, locally invasive, and often have a protracted clinical history that may include multiple previous excisions. The tumor metastasizes to the sublumbar lymph nodes in 15% of cases, and distant metastasis (lungs, liver, spleen, and sublumbar lymph nodes) is also reported (Wilson and Hayes 1979; Vail et al. 1990). Concurrent hypercalcemia is occasionally seen, and the etiology is unknown at present. These tumors are most prevalent in elderly male or female dogs (Berrocal et al. 1989; Vail et al. 1990) and do not demonstrate hormonal dependence despite the fact that hepatoid gland carcinomas still express androgen receptors (Wilson and Hayes 1979; Vail et al. 1990; Pisani et al. 2006). In one study, 8 of 16 canine perianal adenocarcinomas overexpressed mutated p53 tumor suppressor protein (Gamblin et al. 1997). Predisposition is seen in large male dogs such as Arctic breeds and German shepherds (Vail et al. 1990). Surgery, when feasible, is the first option for this tumor. The only prognostic factor identified is clinical stage (Owen 1980), and dogs with tumors over 5 cm have an 11 times higher risk of death compared to dogs with smaller tumors. Dogs with tumor staging T1N0M0 (T1 = tumor of less than 2 cm, superficial or proliferative) and T2N0M0 (T2 = tumor of 2–5 cm or, with minimal invasion independent of size) had a 2-year disease free interval of about 75% and 60%, respectively and dogs with tumor stages beyond T2 (T3 = tumor of more than 5 cm or invasive tumors independent of size; T4 = invasive tumor) had median survival of 6–12.5 months (Vail et al. 1990). The reported median survival time in dogs with confirmed metastasis (15% of cases at pre­ sentation) was 7 months (Vail et al. 1990). Recently, the evaluation of computer-assisted nuclear cytological morphometric parameters has been proposed and has demonstrated that mean nuclear area, mean nuclear perimeter, maximum nuclear diameter, and minimum nuclear diameter may be used as prognostic indicators for canine perianal adenocarcinoma. Significant differences in survival were seen in all of these parameters between metastatic tumors with positive regional lymph nodes and nonmetastatic tumors (Simeonov and Simeonova 2008b). Anal sac adenocarcinoma These neoplasms are less common than perianal hepatoid tumors and generally affect elderly dogs; dogs as young as 5 years, however, have also been reported (White and Gorman 1987; Ross et al. 1991; Williams et al. 2003; Turek and Withrow 2007). At present, no

Alimentary Tract  259

Figure 7.61.  Ventral vertebral changes caused by an adjacent metastatic sublumbar lymphadenomegaly originating from an anal sac adenocarcinoma in a dog. (Image courtesy of Dr. Sarah Boston)

Figure 7.62.  Humeral metastasis secondary to an anal sac adenocarcinoma in a 9-year-old spayed female Schnauzer. The dog presented with lameness; abdominal ultrasound did not reveal any lesion. Bone biopsy was consistent with a metastatic adenocarcinoma. The primary anal sac adenocarcinoma was unilateral and 2 mm small.

gender predisposition has been documented (Ross et al. 1991; Straw et al. 1994; Bennett et al. 2002; Williams et al. 2003; Polten et al. 2006; Polton and Brearley 2007). The mean relative risk estimate associated with being neutered was 1.4 in one study, and the effect of neutering appeared to be more significant in male dogs compared to female dogs (Polton and Brearley 2007). Breed predisposition is reported in spaniels (English cocker spaniels, mean relative risk estimate of 7.3, and springer and Cavalier King Charles spaniels) (Polton and Brearley 2007), and heritability in these dogs is suspected (Polton 2009). Specific computer-assisted cytological nuclear morphometric parameters have been suggested to facilitate the differentiation between anal sac adenoma (indeed very rare) and anal sac adenocarcinoma (Simeonov and Simeonova 2008a). Metastasis is evident at presentation in 46%–96% of dogs, with the main sites including regional lymph nodes (sublumbar and only rarely inguinal), lungs, liver, spleen, other abdominal organs, and the skeleton, including lumbar vertebrae (Goldschmidt and Zoltowski 1981; Ross et al. 1991; Bennett et al. 2002; Turek et al. 2003; Williams et al. 2003; Brisson et al. 2004; Turek and Withrow 2007; Hammond et al. 2009) (Figure 7.61; see also Figures 7.40 and 7.41).

In rare cases, lung or bone metastasis may develop without evidence of regional lymphadenopathy (Turek and Withrow 2007) (Figure 7.62). Hypercalcemia is present in 25%–90% of cases, and it is less common in male dogs than in spayed bitches (White and Gorman 1987; Rosol et al. 1990; Ross et al. 1991; Bennett et al. 2002; Williams et al. 2003). When present, hypercalcemia caused by the PTH-like compound produced by the tumor (Rosol et al. 1990; Gröne et al. 1994) can be used as a marker for both metastasis and recurrent lesions. The primary tumor rarely involves both anal sacs and may be very small (or occult) even in the presence of large metastatic lymphadenopathy (Ross et al. 1991; Bertazzolo et al. 2003; Turek et al. 2003; Emms 2005). This tumor is not commonly reported in cats (Chun et al. 1997; Mellanby et al. 2002; Parry 2006). Surgery plays a fundamental role in the treatment of these tumors, independent of any adjuvant therapeutic modality (Williams et al. 2003). When feasible, an aggressive resection should be performed; however, only marginal excision is achieved in most cases (see Figures 7.50 and 7.51). After surgical excision, survival has been reported to be about 1 year, with a wide range (few days up to 96 months) (White and Gorman 1987; Ross et al. 1991; Bennett et al. 2002; Williams et al. 2003; Hobson

260  Veterinary Surgical Oncology

et al. 2006). Concurrent excision of regional metastatic lymph nodes usually results in increased survival times (Hobson et al. 2006; Polton and Brearley 2007). Hypercalcemia, which together with regional metastatic lymph node involvement has been described as a negative prognostic factor, may normalize only after gross cytoreduction. The role of hypercalcemia as a negative prognostic factor (decreased survival) is unclear (Ross et al. 1991; Bennett et al. 2002; Turek et al. 2003; Williams et al. 2003), unless secondary renal failure has already occurred. Additionally, the prognostic significance of regional metastatic lymphadenopathy is unclear since it will likely develop, even if it is not present at the initial presentation, or it may be actually present despite not being clinically detectable (Ross et al. 1991; Turek et al. 2003; Williams et al. 2003). Lung metastasis and primary tumors greater than 10 cm in size are associated with a decreased survival (Williams et al. 2003). Death is usually the consequence of recurrence and metastasis; the latter develops in almost all the patients after a variable amount of time. TNM clinical staging is the same as that used for skin tumors. One prognostic parameter for survival is the diameter of the primary tumor (improved survival with tumor size 10 cm or smaller) (Williams et al. 2003). A modified and more simplified clinical staging system that correlates with survival time has been proposed recently (Polton and Brearley 2007). According to this study, stage 1 and 2 is when the primary tumor is less than or greater than 2.5 cm without evidence of metastasis, respectively. In stages 3a, 3b, and 4, the sizes of the primary tumor are irrelevant, with regional lymph nodes less than 4.5 cm of diameter in stage 3a, more than 4.5 cm in stage 3b, and with distant metastases in stage 4. Dogs in this report were divided into two groups: one was evaluated retrospectively (80 dogs) and one prospectively (50 dogs). As expected, median survival was longer in stage 1 and 2 dogs (40 and 24 months, respectively), with a significant gap between stage 3a and 3b dogs in both the retrospective group (16 and 11 months, respectively) and in the prospective group (15 and 10 months, respectively). The shortest survival (less than 3 months) was documented in dogs with stage 4 disease. Despite the fact that chemotherapy did not influence outcome, the authors of this report still recommended its use in stage 3b lesions. Other perianal tumors Squamous cell carcinoma is a rare tumor that may develop from the anal sac (Esplin et al. 2003), anal canal, and perianal skin (Figure 7.63). These tumors demonstrate malignant behavior, both in terms of local infiltration and regional and systemic spread.

Figure 7.63.  Squamous cell carcinoma of the anal region in an 11-year old male German shepherd.

Malignant melanoma is another rare lesion that may develop from the perianal region, although it has also been thought to derive from the anal sac (Kim et al. 2005; Young et al. 2005). Its behavior is unknown, but it is likely to be highly malignant, as are its more typical oral counterparts (Kim et al. 2005). One report has dealt with a transient local palliation by electrochemotherapy with cisplatin in a dog (Spugnini, Filipponi, et al. 2007). Lymphoma (Figure 7.64), mast cell tumors, benign tumors (lipoma, leiomyoma, hemangioma; Figure 7.65), and soft tissue sarcomas (see Figure 7.47) are also seen in the perianal region (Ueno et al. 2002; Brønden et al. 2010). Adjuvant therapies For perianal adenoma, no adjuvant treatment is needed. Alternative treatments are the administration of diethylstilbestrol (which should be avoided due to the potential for bone marrow suppression), radiation, and hyperthermia (Grier et al. 1980; Gillette 1970). The excellent results achieved with surgical treatment render these procedures unattractive (Turek and Withrow 2007). A potential treatment is the use of drugs for chemical castration (Turek and Withrow 2007). For perianal adenocarcinoma, radiation may be an option. This tumor type is often considered radioresistant (Vail et al. 1990). Radiation can be applied both locally and to the site of the excised metastatic sublumbar lymph nodes (Straw et al. 1994; Bley et al. 2003).

Alimentary Tract  261

Figure 7.64.  Lymphoma of the anal region in a 7-year-old female mongrel dog.

Figure 7.65.  Large hemangioma located between the anus and the tail base in a 5-year-old male Leonberger.

There is, however, only scant information regarding outcome available in the literature. At present, there is no evidence that adjuvant chemotherapy (doxorubicin with or without cyclophosphamide, cisplatin, or actinomycin D) is useful, and remission, if any, is usually

partial and transient (Vail et al. 1990; Hammer et al. 1994). Finally, perianal adenoma and adenocarcinoma have also been treated with chemotherapy using bleomycin or cisplatin combined with electroporation (Tozon et al. 2005; Spugnini, Dotsinsky, et al. 2007). For anal sac adenocarcinoma, reported chemotherapeutic agents include cisplatin, carboplatin, doxorubicin with or without cyclophosphamide, mitoxantrone, epirubicin, melphalan, actinomycin D, mithramycin, chlorambucil, vincristine, L-asparaginase, gemcitabine, and piroxicam (Ross et al. 1991; Hammer et al. 1994; Bennett et al. 2002; Turek et al. 2003; Williams et al. 2003; Emms 2005; Polton and Brearley 2007). The role of chemotherapy is unclear. It has been reported that the adjuvant use of platinum compounds may allow a longer survival (Bennett et al. 2002), whereas in other reports no difference was found independently of the protocol used (Williams et al. 2003; Polton and Brearley 2007). Mitoxantrone administration associated with radiotherapy has a reported overall median survival time of 31 months, with improved local and regional tumor control. Complications of radiation therapy were observed in more than 50% of cases; however, control of these complications was possible in the majority of instances (Turek et al. 2003). In cats, doxorubicin or carboplatin may be used after surgery as the clinical behavior of this tumor is expected to be similar to that of dogs. With regard to radiotherapy, given that the surgical removal of anal sac adenocarcinoma is often marginal, adjuvant or intraoperative radiation both for local (48–50 Gy delivered in small fractions) and metastatic disease (Straw et al. 1994; Turek et al. 2003; Williams et al. 2003; Polton and Brearley 2007; Turek and Withrow 2007) may be justified. For the sublumbar metastatic sites, radiation is usually applied once intraoperatively with 15–19 Gy. Acute adverse effects of adjuvant radiation may be skin desquamation, colitis (that may be refractory to medical treatment), and tenesmus, and these signs are usually self-limiting in approximately 1 month. Possible chronic adverse radiation-induced effects may include tenesmus secondary to rectal stricture, intestinal perforation, diarrhea, and fecal incontinence (Turek et al. 2003). In order to avoid adverse effects to the colon, smaller fractions (2.7–2.9 Gy) and avoidance of radiation potentiators are recommended for pelvic irradiation (Anderson et al. 2002; Arthur et al. 2008). Hypofractionated radiotherapy (900 cGy a week), as both a neoadjuvant and adjuvant treatment, or as a salvage procedure in inoperable cases, has been also used (Polton and Brearley 2007). Finally, electrochemotherapy with cisplatin has been proposed in a dog as an adjuvant treatment for an

262  Veterinary Surgical Oncology

incompletely excised anal sac adenocarcinoma (Spugnini et al. 2008).

References Adler, R. and D.W. Wilson. 1995. Biliary cystadenoma of cats. Vet Pathol 32:415–418. Allen, S.W. and S.W. Crowell. 1991. Ventral approach to the pelvic canal in the female dog. Vet Surg 20(2):118–121. Allen, D.A., D.D. Smeak, and R. Schertel. 1992. Prevalence of small intestinal dehiscence and associated clinical factors: A retrospective study of 121 dogs. J Am Anim Hosp Assoc 28:70–76. Allenspach, K., P. Arnold, T. Glaus, et al. 2000. Glucagon-producing neuroendocrine tumor associated with hypoaminoacidaemia and skin lesions. J Sm Anim Pract 41:402–406. Altschul, M., K.W. Simpson, N.L. Dykes, et al. 1997. Evaluation of somatostatin analogues for the detection and treatment of gastrinoma in a dog. J Small Anim Pract 38:286–291. Anderson, C.R., E.A. McNiel, E.L. Gillette, et al. 2002. Late complications of pelvic irradiation in 16 dogs. Vet Radiol Ultrasound 43(2):187–192. Anderson, G.I., D.B. McKeown, G.D. Partlow, et al. 1987. Rectal resection in the dog. A new surgical approach and the evaluation of its effect on fecal continence. Vet Surg 16(2):119–125. Andrews, L.K. 1987. Tumors of the exocrine pancreas. In Diseases of the Cat, pp. 505–507. J. Holzworth, editor. Philadelphia: Saunders. Anson, L.W., C.W. Betts, and E.A. Stone. 1988. A retrospective evaluation of the rectal pull-through technique. Procedure and postoperative complications. Vet Surg 17(3):141–146. Aronson, L.R. 2003. Rectum and anus. In Textbook of Small Animal Surgery, 3rd edition, pp. 682–708. D. Slatter, editor. Philadelphia: Saunders. Arthur, J.J., M.M. Kleiter, D.E. Thrall, et al. 2008. Characterization of normal tissue complications in 51 dogs undergoing defini­ tive pelvic region irradiation. Vet Radiol Ultrasound 49(1):85– 89. Attum, A.A., J.R. Hankins, J. Ngangana, et al. 1979. Circular myotomy as an aid to resection and end-to-end anastomosis of the esophagus. Ann Thorac Surg 28:126–132. Bagley, R.S., J.K. Levy, and D.E. Malarkey. 1996. Hypoglycemia associated with intra-abdominal leiomyoma and leiomyosarcoma in six dogs. J Am Vet Med Assoc 208:69–71. Bailey, D.B. and R.L. Page. 2007. Tumors of the endocrine system. In Withrow and MacEwen’s Small Animal Clinical Oncology, 4th editon, pp. 583–609. S.J. Withrow and D.M. Vail, editors. Philadelphia: Saunders. Banz, W.J., D.J. Jackson, K. Richter, et al. 2008. Transrectal stapling for colonic resection and anastomosis (10 cases). J Am Anim Hosp Assoc 44:198–204. Barnes, R.F., C.L. Greenfield, D.J. Schaeffer, et al. 2006. Comparison of biopsy samples obtained using standard endoscopic instruments and the harmonic scalpel during laparoscopic and laparoscopicassisted surgery in normal dogs. Vet Surg 35:243–251. Beaudry, D., D.W. Knapp, T. Montgomery, et al. 1995. Hypoglycemia in four dogs with smooth muscle tumors. J Vet Intern Med 9:415–418. Bellah, J.R. Surgical stapling of the spleen, pancreas, liver, and uro­ genital tract. 1994. Vet Clin North Am Small Anim Pract 24:375– 394. Bellah, J.R. and P.E. Ginn. 1996. Gastric leiomyosarcoma associated with hypoglycemia in a dog. J Am Anim Hosp Assoc 32:283–286.

Bennett, P.F., D.B. DeNicola, P. Booney, et al. 2002. Canine anal sac adenocarcinomas: Clinical presentation and response to therapy. J Vet Inter Med 16(1):100–104. Bennett, P.E., K.A. Hahn, R.L. Toal, et al. 2001. Ultrasonographic and cytopathological diagnosis of exocrine pancreatic carcinoma in the dog and cat. J Am Anim Hosp Assoc 37:466–473. Bergman, J.R. 1985. Nodular hyperplasia in the liver of the dog: An association with changes in the Ito cell population. Vet Pathol 22:427–438. Berrocal, A., J.H. Vos, T.S. van den Ingh, et al. 1989. Canine perineal tumours. Zentralblatt fur Veterinarmedizin. Reihe A 36(10): 739–749. Bertagnolli, M.M. 2007. Chemoprevention of colorectal cancer with cyclooxygenase-2 inhibitors: Two steps forward, one step back. Lancet Oncol 8(5):439–443. Bertazzolo, W., S. Comazzi, P. Roccabianca, et al 2003. Hypercalcaemia associated with a retroperitoneal apocrine gland adenocarcinoma in a dog. J Small Anim Pract 44(5):221–224. Bertoy, R.W., D.M. MacCoy, L.G. Wheaton, et al. 1989. Total colectomy with ileorectal anastomosis in the cat. Vet Surg 18(3): 204–210. Bezuidenhout, A.J. 1993. The lymphatic system. In Miller’s Anatomy of the Dog, 3rd edition, pp. 739–748. H.E. Evans, editor. Philadelphia: Saunders. Bigge, L.A., D.J. Brown, and D.G. Penninck. 2001. Correlation between coagulation profile findings and bleeding complications after ultrasound-guided biopsies: 434 cases (1993–1996). J Am Anim Hosp Assoc 37:228–233. Birchard, S.J., C.G. Couto, and S. Johnson. 1986. Nonlymphoid intestinal neoplasia in 32 dogs and 14 cats. J Am Anim Hosp Assoc 22:533–537. Bjorling, D.E., K.W. Prasse, and R.A. Holmes. 1985. Partial hepatectomy in dogs. Compend Contin Educ Pract Vet 3:257–259. Blass, C.E. and H.B. Seim. 1985. Surgical techniques for the liver and biliary tract. Vet Clin North Am Small Anim Pract 15(1):257–275. Blazquez, N.B., J.M. French, S.E. Long, et al. 1988. A pheromonal function for the perineal skin glands in the cow. Vet Rec 123:49–50. Bley, C.R., S. Stankeova, A. Sumova, et al. 2003. Metastases of perianal gland carcinoma in a dog: Palliative tumor therapy. Schweiz Arch Tierheilkd 145(2):89–94. Boari, A., A. Barreca, G.E. Bestetti, et al. 1995. Hypoglycemia in a dog with leiomyoma of the gastric wall producing an insulin-like growth factor II–like peptide. Eur J Endocrinol 143:744–750. Bond, R., P.E. McNeil, H. Evans, et al. 1995. Metabolic epidermal necrosis in two dogs with different underlying diseases. Vet Rec 136:466–471. Bonfanti, U., W. Bertazzolo, E. Bottero, et al. 2006. Diagnostic value of cytologic examination of gastrointestinal tract tumor in dogs and cats: 83 cases (2001–2004). J Am Vet Med Assoc 229:1130–1133. Brentjens, R. and L. Saltz. 2001. Islet cell tumors of the pancreas: The medical oncologist’s perspective. Surg Clin North Am 81(3): 527–542. Brisson, B.A., D.P. Whiteside, and D.L. Holmberg. 2004. Metastatic anal sac adenocarcinoma in a dog presenting for acute paralysis. Can Vet J 45(8):678–681. Brønden, L.B., T. Eriksen, and A.T. Kristensen. 2010. Mast cell tumours and other skin neoplasia in Danish dogs— data from the Danish Veterinary Cancer Registry. Acta Veterinaria Scandinavica 52:6. Brooks, D. and G.L. Watson. 1997. Omeprazole in a dog with gastrinoma. J Vet Intern Med 11:379.

Alimentary Tract  263 Brown, D. 2003. Small intestines. In Textbook of Small Animal Surgery, 3rd edition, pp. 644–664. D. Slatter, editor. Philadelphia: Elsevier Science. Brueker, K.A. and S.J. Withrow. 1988. Intestinal leiomyosarcoma in six dogs. J Am Anim Hosp Assoc 24(3):281–284. Campagnacci, R., A. De Sanctis, M. Baldarelli, et al. 2007. Hepatic resections by means of electrothermal bipolar vessel device (EBVS) ligaSure V: Early experience. Surg Endosc 21(12):2280–2284. Campbell, J.R. and H.M. Pirie. 1965. Leiomyoma of the oesophagus in a dog. Vet Rec 77:624–626. Cavanaugh, R.P., J.R. Kovak, A.J. Fischetti, et al. 2008. Evaluation of surgically placed gastrojejunostomy feeding tubes in critically ill dogs. J Am Vet Med Assoc 232:380–388. Caywood, D.D., J.S. Klausner, T.P. O’Leary, et al. 1988. Pancreatic insulin-secreting neoplasms: Clinical, diagnostic, and prognostic features in 73 Dogs. J Am Anim Hosp Assoc 24:577–584. Chun, R., S. Jakovljevic, W.B. Morrison, et al. 1997. Apocrine gland adenocarcinoma and pheochromocytoma in a cat. J Am Anim Hosp Assoc 33(1):610–612. Church, E.M., C.J. Mehlhaff, and A.K. Patnaik. 1987. Colorectal adenocarcinoma in dogs: 78 cases (1973–1984). J Am Vet Med Assoc 191(6):727–730. Clark, G.N. 1994. Gastric surgery with surgical stapling instruments. Vet Clin N Am Small Anim Pract 24:279–304. Clifford, C.A., E.S. Pretorius, C. Weisse, et al. 2004. Magnetic resonance imaging of focal splenic and hepatic lesions in the dog. J Vet Intern Med 18(3):330–338. Cobb, L.F. and R.C. Merrell. 1984. Total pancreatectomy in dogs. J Surg Res 37:235–240. Cohen, M., G.S. Post, and J.C. Wright. 2003. Gastrointestinal leiomyosarcoma in 14 dogs. J Vet Intern Med 17(1):107–110. Cohn, L.A., M.E. Kerl, C.E. Lenox, et al. 2007. Response of healthy dogs to infusions of human serum albumin. Am J Vet Res 68:657–663. Cole, T.L., S.A. Center, S.N. Flood, et al. 2002. Diagnostic comparison of needle and wedge biopsy specimens of the liver in dogs and cats. J Am Vet Med Assoc 220(10):1483–1490. Collins, J.D., M.L. Shaver, P. Batra et al. 1989. Anatomy of the abdomen, back, and pelvis as displayed by magnetic resonance imaging: Part one. J Nat Med Assoc 81:680–684. Coolman, B.R., N. Ehrhart, and S.M. Marretta. 2000a. Healing of intestinal anastomoses. Comp Contin Educ Pract Vet 22(4):363–371. Coolman, B.R., N. Ehrhart, G. Pijanowski, et al. 2000b. Comparison of skin staples with sutures for anastomosis of the small intestine in dogs. Vet Surg 29:293–302. Cornell, K. and J. Fischer. 2003. Surgery of the exocrine pancreas. In Textbook of Small Animal Surgery, 3rd edition, pp. 752–762. D. Slatter, editor. Philadelphia: Saunders. Cotchin, E. 1959. Some tumors in dogs and cats of comparative veterinary and human interest. Vet Rec 71:1040–1050. Covey, J., D.A. Degner, A.H. Jackson, et al. 2009. Hilar liver resection technique in dogs. Vet Surg 38(1):104–111. Crawshaw, J., J. Berg, J.C. Sardinas, et al. 1998. Prognosis for dogs with nonlymphomatous, small intestinal tumors treated by surgical excision. J Am Anim Hosp Assoc 34:451–456. Crowe, D.T. 1984. The serosal patch: Clinical use in 12 animals. Vet Surg 13:29–38. Crowe, D.T. and J.J. Devey. 1997. Clinical experience with jejunostomy feeding tubes in 47 small animal patients. J Vet Emerg Crit Care 7:7–19. Culbertson, R., J.E. Branam, and L.S. Rosenblatt. 1983. Oesophageal/ gastric leiomyoma in the laboratory beagle. J Am Vet Med Assoc 183:1168–1171.

Cullen, J.M. and J.A. Popp. 2002. Tumours of the liver and gall bladder. In Tumors in Domestic Animals, 4th edition, pp. 483–508. D.J. Meuten, editor. Ames, IA: Iowa State Press. Dallman, M.J. 1988. Functional suture-holding layer of the esophagus in the dog. J Am Vet Med Assoc 192:638–640. Danova, N.A., J.C. Robles-Emanuelli, and D.E. Bjorling. 2006. Surgical excision of primary canine rectal tumors by an anal approach in twenty-three dogs. Vet Surg 35(4):337–340. Davies, J.V. and H.M. Read. 1990. Sagittal pubic osteotomy in the investigation and treatment of intrapelvic neoplasia in the dog. J Small Anim Pract 31:123–130. Daye, R.M., M.L. Huber, and R.A. Henderson. 1999. Interlocking box jejunostomy: A new technique for enteral feeding. J Am Anim Hosp Assoc 35:129–134. Dean, P.W. and J.M. Bojrab. 1993. Defecation and fecal continence. In Disease Mechanisms in Small Animal Surgery, 2nd edition, pp. 287–291. Joseph M. Bojrab, editor. Philadelphia: Lea & Febiger. Demetriou, J.L. and E.M. Welsh. 1999. Rectal prolapse of an ileocaecal neoplasm associated with intussusception in a cat. J Feline Med Surg 1(4):253–256. Dietrich, C.F., A. Ignee, J. Trojan, et al. 2004. Improved characterisation of histologically proven liver tumors by contrast enhanced ultrasonography during the portal venous and specific late phase of SHU 508A. Gut 53:401–405. Dow, S.W., P.N. Olson, R.A.W. Rosychuck, et al. 1988. Perianal adenomas and hypertestosteronemia in a spayed bitch with pituitarydependent hyperadrenocorticism. J Am Vet Med Assoc 192(10): 1439–1441. Dunn, J.K., D.E. Bostock, M.E. Herrtage, et al. 1993. Insulin-secreting tumours of the canine pancreas: Clinical and pathological features of 11 cases. J Small Anim Pract 34: 325–331. Dvir, E., R.M. Kirberger, and D. Malleczek. 2001. Radiographic and computed tomographic changes and clinical presentation of spirocercosis in dogs. Vet Radiol Ultrasound 42:119–129. Dvir, E., R.M. Kirberger, V. Mukorera, et al. 2008. Clinical differentation between dogs with benign and malignant spirocercosis. Vet Parasitol 155:80–88. Dyce, K.M., W.O. Sack, and C.J.G. Wensing. 1996. The digestive apparatus. In Textbook of Veterinary Anatomy, 2nd edition, pp. 120– 122. K.M. Dyce, W.O. Sack, C.J. et al., editors. Philadelphia: Saunders. Easton, S. 2001. A retrospective study into the effects of operator experience on the accuracy of ultrasound in the diagnosis of gastric neoplasia in dogs. Vet Radiol Ultrasound 42:47–50. Eisele, J., J.R. Kovak-McClaran, J. Runge, et al. 2010. Evaluation of risk factors for morbidity and mortality after pylorectomy and gastroduodenostomy in dogs. Vet Surg 39:261–267. Elie, M.S. and C.A. Zebra. 1995. Insulinoma in dogs, cats, and ferrets. Compend Contin Educ Pract Vet 17:51–59. Ellison, G.W. 1989. Wound helaing in the gastrointestinal tract. Semin Vet Med Surg (Small Anim) 4:287–293. Emms, S.G. 2005. Anal sac tumours of the dog and their response to cytoreductive surgery and chemotherapy. Aust Vet J 83(6): 340–343. Esplin, D.G., S.R. Wilson, and G.A. Hullinger. 2003. Squamous cell carcinoma of the anal sac in five dogs. Vet Pathol 40(3):332–334. Evans, H.E. 1993. The alimentary canal. In Miller’s Anatomy of the Dog, 3rd edition, pp. 422–425. H.E. Evans, editor. Philadelphia: Saunders. Evans, H.E. 2010. The neck, thorax, and thoracic limb. In Guide to the Dissection of the Dog, 7th edition, p. 108. H.E. Evans and A. deLahunta, editors. Saunders: St. Louis.

264  Veterinary Surgical Oncology Evans, S.E., J.J. Bonczynski, J.D. Broussard, et al. 2006. Comparison of endoscopic and full-thickness biopsy specimens for diagnosis of inflammatory bowel disease and alimentary tract lymphoma in cats. J Am Vet Med Assoc 229:1447–1450. Everett, W.G. 1975. A comparison of one layer and two layer techniques for colorectal anastomosis. Br J Surg 62(2):135–140. Farese, J.P., N.J. Bacon, N.P. Ehrhart, et al. 2008. Oesophageal leiomyosarcoma in dogs: Surgical management and clinical outcome of four cases. Vet Comp Oncol 6:31–38. Farid, H. and T. O’Connell. 1994. Hepatic resections: Changing mortality and morbidity. Am Surg 60:748–752. Fasulo, F., A. Giori, S. Fissi, et al. 1992. Cavitron Ultrasonic Surgical Aspirator (CUSA) in liver resection. Int Surg 77:64–66. Feldman, E.C. and R.W. Nelson. 2004. Canine and Feline Endocrinology and Reproduction, 3rd edition, pp. 616–644. St. Louis: Saunders. Feres, O., J.C. Monteiro dos Santos Jr, and J.I. Andrade. 2001. The role of mechanical bowel preparation for colonic resection and anastomosis: An experimental study. Int J Colorectal Dis 16(6):353– 356. Flanders, J.A. 1989. Problems and complications associated with esophageal surgery. Prob Vet Med 1:183–194. Fourtanier, G., F. Prevost, and F. Lacaine. 1987. Nutritional status of patients with digestive system cancer: Preoperative prognostic significance. Gastroenterol Clin Biol 11:748–752. Foy, D.S. and J.F. Bach. 2010. Endoscopic polypectomy using endocautery in three dogs and one cat. J Am Anim Hosp Assoc 46:168–173. Francis, A.H., L.G. Martin, G.J. Haldorson, et al. 2007. Adverse reactions suggestive of type III hypersensitivity in six healthy dogs given human albumin. J Am Vet Med Assoc 230:873–879. Freeman, L.J. 2009. Gastrointestinal laparoscopy in small animals. Vet Clin North Am Small Admin Pract 39:903–924. Frost, D., J. Lasota, and M. Miettinen. 2003. Gastrointestinal stromal tumors and leiomyomas in the dog: A histopathologic, immunohistochemical, and molecular genetic study of 50 cases. Vet Pathol 40:42–54. Gamblin, R.M., Sagartz, J.E., and Couto C.G. 1997. Overexpression of p53 tumor suppressor protein in spontaneously arising neoplasms of dogs. Am J Vet Res 58(8):857–863. Ganguly, A. and Wolfe, L.G. 2006. Canine perianal gland carcinomaassociated antigens defined by monoclonal antibodies. Hybridoma (Larchmt) 25(1):10–14. Garden, O.A., J.C. Reubi, N.L. Dykes, et al. 2005. Somatostatin receptor imaging in vivo by planar scintigraphy facilitates the diagnosis of canine insulinomas. J Vet Intern Med 19:168–176. Gaston, E.A. 1948a. The physiology of fecal continence. Surg Gynecol Obstet 87:280–290. Gaston, E.A. 1948b. Fecal continence following resections of various portions of the rectum with preservation of the anal sphincter. Surg Gynecol Obstet 87:669–678. Gaston, E.A. 1951. Physiological basis for preservation of fecal continence after resection of rectum. JAMA 146:1486–1489. Gaston, E.A. 1961. Fecal continence following sphincter-preserving operations for rectal cancers. Am J Proctol, Gastroenterol Colon Rectal Surg 12:169–175. Gaynor, J.S. 2002. Cancer pain management. In Handbook of Veterinary Pain Management, pp. 405–419. J.S. Gaynor and W. Muir, editors. St. Louis: Mosby. Gear, R.N., N.J. Bacon, S. Langley-Hobbs, et al. 2006. Panniculitis, polyarthritis, and osteomyelitis associated with pancreatic neoplasia in two dogs. J Small Anim Pract 47(7):400–404. Gibril, F., J.C. Reynolds, J.L. Doppman, et al. 1996. Somatostatin receptor scintigraphy: Its sensitivity compared with that of other

imaging methods in detecting primary and metastatic gastrinomas. A prospective study. Ann Inter Med 125:26–34. Gibson, C.J., N.M.A. Parry, R.M. Jakowski, et al. 2010. Adenomatous polyp with intestinal metaplasia of the esophagus (Barret esophagus) in a dog. Vet Pathol 47:116–119. Gillette, E.L. 1970. Veterinary radiotherapy. J Am Vet Med Assoc 157:1707–1712. Glazer, A. and P. Walters. 2008. Esophagitis and esophageal strictures. Comp Cont Educ Pract 30:281–292. Gold, J.S. and R.P. Dematteo. 2006. Combined surgical and molecular therapy: The gastrointestinal stromal tumor model. Ann Surg 244:176–184. Goldschmidt, M.H., R.W. Dunstan, A.A. Stannard, et al. 1998. Melanocytic tumors and tumor-like lesions. In Histological Classification of Epithelial and Melanocytic Tumors of Skin of Domestic Animals, 3rd edition, pp. 38–40. M.H. Goldschmidt, R.W. Dunstan, A.A. Stannard, et al., editors. Washington DC: Armed Forces Institute of Pathology. Goldschmidt, M.H. and C. Zoltowski. 1981. Anal sac gland adenocarcinoma in the dog: 14 cases. J Small Anim Pract 22:119–128. Goldsmid, S.E., C.R. Bellenger, P.R. Hopwood, et al. 1993. Colorectal blood supply in dogs. Am J Vet Res 54(11):1948–1953. Gorman, S.C., L.M. Freeman, S.L. Mitchell, et al. 2006. Extensive small bowel resection in dogs and cats: 20 cases (1998–2004). J Am Vet Med Assoc 228:403–407. Grandage, J. 2003. Functional anatomy of the digestive system. In Textbook of Small Animal Surgery, 3rd edition, pp. 505–521. D. Slatter, editor. Philadelphia: Elsevier Science. Green, R.A. and C.L. Gartrell. 1997. Gastrinoma: A retrospective study of four cases (1985–1995). J Am Anim Hosp Assoc 33:524. Greene, S.N. and R.M. Bright. 2008. Insulinoma in a cat. J Small Anim Pract 49:38–40. Gregory, C.R., I.M. Gourley, D.S. Bruyet, et al. 1988. Free jejunal segment for treatment of cervical esophageal stricture in a dog. J Am Vet Med Assoc 193:230–232. Grier, R.L., W.G. Brewer, and G.H. Theilen. 1980. Hyperthermic treatment of superficial tumors in cats and dogs. J Am Vet Med Assoc 177:227–233. Gröne, A., J.R. Werkmeister, C.L. Steinmeyer, et al. 1994. Parathyroid hormone-related protein in normal and neoplastic canine tissues: Immunohistochemical localization and biochemical extraction. Vet Pathol 31(3):308–315 Gross, T.L., T.D. O’Brien, and A.P. Davies. 1990. Glucagon-producing pancreatic endocrine tumors in dogs with superficial necrolytic dermatitis. J Am Vet Med Assoc 197:1619–1622. Gualtieri, M. 2001. Esophagoscopy. Vet Clin N Am Small 31:605–630. Guilford, G.W. 1990. Fecal incontince in dogs and cats. Comp Contin Educ Vet 12(3):313–326. Günther, K., G. Braunrieder, B.R. Bittorf, et al. 2003. Patients with familial adenomatous polyposis experience better bowel function and quality of life after ileorectal anastomosis than after ileoanal pouch. Colorectal Disease 5(1):38–44. Haers, H. and J.H. Saunders. 2009. Review of clinical characteristics and applications of contrast-enhanced ultrasonography in dogs. J Am Vet Med Assoc 234(4):460–470. Hamilton, T.A. and J.L. Carpenter. 1994. Esophageal plasmacytoma in a dog. J Am Vet Med Assoc 204:1210–1211. Hammer, A.S., C.G. Couto, R.D. Ayl, et al. 1994. Treatment of tumorbearing dogs with actinomycin D. J Vet Intern Med 8:36–239. Hammond, T.N., M.M. Turek, and J. Regan. 2009. What is your diagnosis? Metastatic anal sac adenocarcinoma with paraneoplastic hypertrophic osteopathy. J Am Vet Med Assoc 235(3):267–268.

Alimentary Tract  265 Hardie, E.M. and S.D. Gilson. 1997. Use of colostomy to manage rectal disease in dogs. Vet Surg 26(4):270–274. Harvey, H.J. 1990. Complications of small intestinal biopsy in hypoalbuminaemic dogs. Vet Surg 19:289–292. Hause, W.R., S. Stevenson, D.J. Meuten, et al. 1981. Pseudohyperparathyroidism associated with adenocarcinomas of anal sac origin in four dogs. J Am Anim Hosp Assoc 17(3):373–379. Hawks, D., M.E. Peterson, K.L. Hawkins, et al. 1992. Insulin-secreting pancreatic (Islet Cell) carcinoma in a cat. J Am Vet Med Assoc 6:193–196. Hayari, L., D.D. Hershko, H. Shoshani, et al. 2004. J Pediatr Surg 39:540–544. Hecht, S. and G. Henry. 2007. Sonographic evaluation of the normal and abnormal pancreas. Clin Tech Small Anim Pract 22:115– 121. Hecht, S., D.G. Penninck, and J.H. Keating. 2007. Imaging findings in pancreatic neoplasia and nodular hyperplasia in 19 cats. Vet Radiol Ultrasound 48:45–50. Hedlund, C.S. and T.W. Fossum. 2007a. Anal neoplasia. In Small Animal Surgery, 3rd edition, pp. 507–511. T.W. Fossum, C.S. Hedlund, A.L. Johnson, et al., editors. St. Louis: Mosby. Hedlund, C.S. and T.W. Fossum. 2007b. Surgery of the perineum, rectum and anus; Surgery of the anus. In Small Animal Surgery, 3rd edition, pp. 480–507. T.W. Fossum, C.S. Hedlund, A.L. Johnson et al., editors. St. Louis: Mosby. Henderson, A.K. and C.R.L. Webster. 2006. The use of gastroprotectants in treating gastric ulceration in dogs. Compend Contin Educ Pract Vet 28:358–370. Hill, K.E., J.C. Scott-Moncrieff, M.A. Koshko, et al. 2005. Secretion of sex hormones in dogs with adrenal dysfunction. J Am Vet Med Assoc 226(4):556-–561. Hobson, H.P., M.R. Brown, and K.S. Rogers. 2006. Surgery of metastatic anal sac adenocarcinoma in five dogs. Vet Surg 35(3):267–270. Hoelzler, M.G., J.R. Bellah, and M.C. Donofro. 2001. Omentalization of cystic sublumbar lymph node metastases for long term palliation of tenesmus and dysuria in a dog with anal sac adenocarcinoma. J Am Vet Med Assoc 219(12):1729–1731. Holt, P.E. 2007. Evaluation of transanal endoscopic treatment of benign canine rectal neoplasia. J Small Anim Pract 48:17–25. Holt D.E. and D. Brockman. 2003. Large intestine. In Textbook of Small Animal Surgery, Slatter D. ed., 3rd edition, pp. 665–682. Philadelphia: Saunders. Holt, P.E. and P. Durdey. 1999. Transanal endoscopic treatment of benign canine rectal tumours: Preliminary results in six cases (1992 to 1996). J Small Anim Pract 40(9):423–427. Holt, D.E., D.E. Johnston, R. Orsher, et al. 1991. Clinical use of a dorsal surgical approach to the rectum. Comp Contin Educ Vet 13(10):1519–1528. Holt, P.E. and V.M. Lucke. 1985. Rectal neoplasia in the dog: A clinicopathological review of 31 cases. Vet Rec 116(15):400–405. Hosgood, G. 1990. The omentum—The forgotten organ: Physiology and potential surgical applications in dogs and cats. Comp Cont Ed Pract Vet 12:45–51. Iseri, T., K. Yamada, K. Chijiwa, et al. 2007. Dynamic computed tomography of the pancreas in normal dogs and in a dog with pancreatic insulinoma. Vet Radiol Ultrasound 48(4):328–331. Ivanc˘ić, M., F. Long, and G.S. Seiler. 2009. Contrast harmonic ultrasonography of splenic masses and associated liver nodules in dogs. J Am Vet Med Assoc 234:88–94. Jakab, C., M. Rusvai, P. Gálfi, et al. 2010. Expression of claudin-1, -3, -4, -5 and -7 proteins in low grade colorectal carcinoma of canines. Histology and Histopathology 25:55–62.

Jamil, J.H., K.R.S. Gill, and M.B. Wallace. 2008. Staging and restaging of advanced esophageal cancer. Curr Opin Gastroen 24:530– 534. Jimba, Y., J. Nagao, and Y. Sumiyama. 2002. Changes in gastrointestinal motility after subtotal colectomy in dogs. Surgery Today 32(12):1048–1057. Joshua, H., M.D. Winer, H.S. Choi, et al. 2010. Intraoperative localization of insulinoma and normal pancreas using invisible nearinfrared fluorescent light. Ann Surg Oncol 17:1094–1100. Jubb, K. 1993. The pancreas. In Pathology of Domestic Animals, 4th edition, volume 2, pp. 407–424. K. Jubb, P. Kennedy, and N. Palmer, editors. San Diego: Academic Press. Kapatkin, A.S., H.S. Mullen, D.T. Matthiesen et al. 1992. Leiomyosarcoma in dogs: 44 cases (1983–1988). J Am Vet Med Assoc 201:1077–1079. Karlan, M., R.C. McPherson, and R.N. Watman. 1959. An experimental evauation of fecal continence—Sphincter and reservoir in the dog. Surg Gynecol Obstet 108:469–475. Katamoto, H., D. Kumagai, N. Kouzai, et al. 2003. Space-occupying leiomyoma in the pelvic canal of a dog. J Small Anim Pract 44(6):277–279. Kato, H., Y. Tachimori, H. Watanabe, et al. 1998. Anastomotic recurrence of oesophageal squamous cell carcinoma after transthoracic oesophagectomy. Eur J Surg 164:759–764. Kerpsack, S.J. and S.J. Birchard. 1994. Removal of leiomyomas and other noninvasive masses from the cardiac region of the stomach. J Am Anim Hosp Assoc 30:500. Kim, D.Y., G.E. Mauldin, G. Hosgood et al. 2005. Perianal malignant melanoma in a dog. J Vet Intern Med 19(4):610–612 Kirberger, R.M., E. Dvir, and L.L. van der Merwe. 2009. The effect of positioning on the radiographic appearance of caudodorsal mediastinal masses in the dog. Vet Radiol Ultrasoun 50:630– 634. Klein, A., S. Scotti, A. Hidalgo, et al. 2006. Rectovaginal fistula following colectomy with an end-to-end anastomosis stapler for a colorectal adenocarcinoma. J Small Anim Pract 47(12):751– 753. Knottenbelt, C., D. Mellor, C. Noxon, et al. 2006. Cohort study of COX-1 and COX-2 expression in canine rectal and bladder tumor. J Small Anim Pract 47(4):196–200. Knottenbelt, C., J.W. Simpson, and M.L. Chandler. 2000a. Neutrophilic leucocytosis in a dog with a rectal tumour. J Small Anim Pract 41(10):457–460. Knottenbelt, C., J.W. Simpson, S. Tasker, et al. 2000b. Preliminary clinical observation on the use of piroxicam in the management of rectal tubulopapillary polyps. J Small Anim Pract 41(9):393– 397. Kokudo, N., S. Tamura, and M. Makuuchi. 2010. Liver tumors in Asia. In Malignant Liver Tumors: Current and Emerging Therapies, 4th edition, pp. 487–499. S. Breitenstein, editor. Hoboken, NJ: Blackwell. Kosovsky, J.E., S. Manfra-Marretta, and D.T. Matthiesen. 1989. Results of partial hepatectomy in 18 dogs with hepatocellular carcinoma. J Am Anim Hosp Assoc 25:203–206. Kosovsky, J.E., D.T. Matthiesen, and A.K. Patnaik. 1998. Small intestinal adenocarcinoma in cats: 32 cases (1978–1985). J Am Vet Med Assoc 192:233–235. Kraje, A.C. 2003. Hypoglycemia and irreversible neurologic compli­ cations in a cat with insulinoma. J Am Vet Med Assoc 223: 812–814. Kruth, S.A., E.C. Feldman, and P.C. Kennedy. 1982. Insulin-secreting islet cell tumors: Establishing a diagnosis and the clinical course for 25 dogs. J Am Vet Med Assoc 181(1):54–58.

266  Veterinary Surgical Oncology Kudisch, M. and M.M. Pavletic. 1993. Subtotal colectomy with surgical stapling instruments via a transcecal approach for treatment of acquired megacolon in cats. Vet Surg 22:457–463. Kumagai, D., T. Shimada, J. Yamate, et al. 2003. Use of an incontinent end-on-colostomy in a dog with annular rectal adenocarcinoma. J Small Anim Pract 44(8):263–366. Kupanoff, P.A., C.A. Popovitch, and M.H. Goldschmidt. 2006. Colorectal plasmacytomas: A retrospective study of nine dogs. J Am Anim Hosp Assoc 42(1):37–43. Kuzma, A.B., D.L. Holmberg, C.W. Miller, et al. 1989. Esophageal replacement in the dog by microvascular colon transfer. Vet Surg 18:439–445. Kyles, A.E. 2003. Exocrine pancreas. In Textbook of Small Animal Surgery, 3rd edition, pp. 1724–1736. D. Slatter, editor. Philadelphia: Saunders. Lamb, C.R. and J. Grierson. 1999. Ultrasonographic appearance of primary gastric neoplasia in 21 dogs. J Small Anim Pract 40:211–215. Lamb, C.R., K.W. Simpson, A. Boswood, et al. 1995. Ultrasonography of pancreatic neoplasia in the dog: A retrospective review of 16 cases. Vet Rec 137(3):65–68. Langer, N.B., A.E. Jergens, and K.G. Miles. 2003. Canine glucagonoma. Comp Contin Educ Pract Vet 25(2):56–63. LaRock, G. and P.E. Ginn. 1997. Immunohistochemical staining characteristics of canine gastrointestinal stromal tumors. Vet Pathol 34:303–311. Lawrence, H.J., H.N. Erb, and H.J. Harvey. 1994. Nonlymphomatous hepatobiliary masses in cats: 41 cases (1972–1991). Vet Surg 23:365–368. Leib, M.S., M.S. Baechtel, and W.E. Monroe. 2004. Complications associated with 355 flexible colonoscopic procedures in dogs. J Vet Intern Med 18(5):642–646. Leib, M.S., H. Dinnel, D.L. Ward, et al. 2001. Endoscopic balloon dilation of benign esophageal strictures in dogs and cats. J Vet Intern Med 15:547–552. Leifer, C.E., M.E. Peterson, and R.E. Matus. 1986. Insulin-secreting tumor: Diagnosis and medical and surgical management in 55 dogs. J Am Vet Med Assoc 188:60–64. Leroy, J., R.A. Cahill, S. Perretta, et al. 2009. Natural orifice translumenal endoscopic surgery (NOTES) applied totally to sigmoidectomy: An original technique with survival in a porcine model. Surg Endosc 23(1):24–30. Lester, N., S. Newell, R. Hill, et al. 1999. Scintigraphic diagnosis of insulinoma in a dog. Vet Radiol Ultrasound 40(2):174–178. Leveille, R., B.P. Partington, D.S. Biller, et al. 1983. Complications after ultrasound-guided biopsy of abdominal structures in dogs and cats: 246 cases (1984–1991). J Am Vet Med Assoc 203:413–415. Levine, M.S., A.R. Fisher, and S.E. Rubesin. 1991. Complications after total gastrectomy and esophagojejunostomy: Radiologic evaluation. AJR 157:1189–1195. Lewis, D.D., C.R. Bellenger, D.T. Lewis, et al. 1990. Hepatic lobectomy in the dog: A comparison of stapling and ligation techniques. Vet Surg 19:221–222. Lewis, D.D., G.W. Ellison, and J.R. Bellah. 1987. Partial hepatectomy using stapling instruments. J Am Anim Hosp Assoc 23:597–602. Lewis, D.G. 1968. Symposium on canine recto-anal disorders III: Clinical management. J Small Anim Pract 9:329–336. Liapi, E., C.C. Georgiades, K. Hong, et al. 2007. Transcatheter arterial chemoembolization: Current technique and future promise. Tech Vasc Interventional Rad 10:2–11. Liptak, J.M. 2007. Intestinal tumors. In Small Animal Clinical Oncology, 4th edition, pp. 491–503. S.J. Withrow and D.M. Vail, editors. St. Louis: Elsevier Science.

Liptak, J.M., W.S. Dernell, E. Monnet, et al. 2004a. Massive hepatocellular carcinoma in dogs: 48 cases (1992–2002). J Am Vet Med Assoc 225:1225–1230. Liptak, J.M., W.S. Dernell, and S.J. Withrow. 2004b. Liver tumors in cats and dogs. Compend Contin Educ Pract Vet 26:50–57. Liska, W.D. and S.J. Withrow. 1978. Cryosurgical treatment of perianal adenomas in the dog. 219(12):1729–1731. J Am Anim Hosp Assoc 14:457–463. Llabrés-Diaz, F.J. 2004. Ultrasonography of the medial iliac lymph nodes in the dog. Vet Radiol Ultrasound 45(2):156–165. Llovet, J.M., A. Burroughs, and J. Bruix. 2003. Hepatocellular carcinoma. Lancet 362:1907–1917. Lo, C.M., S.T. Fan, C.L. Liu, et al. 2000. Determining respectability for hepatocellular carcinoma: The role of laparoscopy and laparoscopic ultrasonography. J Hepatobiliary Pancreat Surg 7:260–264. London, C.A. 2009. Tyrosine kinase inhibitors in veterinary medicine. Top Companion Anim Med 24:106–112. Lowseth, L.A., N.A. Gillett, I.Y. Chang, et al. 1991. Detection of serum α-fetoprotein in dogs with hepatic tumors. J Am Vet Med Assoc 199:735–741. Lurye, J.C. and E.N. Behrend. 2001. Endocrine tumors. Vet Clin North Am Small Anim Pract 31(5):1083–1110. Maas, C.P., G.T. Haar, I. Van Der Gaag, et al. 2007. Reclassification of small intestinal and cecal smooth muscle tumors in 72 dogs: Clinical, histologic and immunohistochemical evaluation. Vet Surg 36:302–313. MacKenzie, R.J., C.M. Furnival, M.A. O’Keane, et al. 1975. The effect of hepatic ischaemia on liver function and the restoration of liver mass after 70% partial hepatectomy in the dog. Br J Surg 62:431–437. Macmanus, J.E., J.T. Dameron, and J.R. Paine. 1950. The extent to which one may interfere with the blood supply of the esophagus and obtain healing on anastomosis. Surgery 28:11–23. Magne, M.L. and S.J. Withrow. 1985. Hepatic neoplasia. Vet Clin North Am Small Anim Pract 15:243–256. Mai, W. and A.V. Caceres. 2008. Dual-phase computed tomographic angiography in three dogs with pancreatic insulinoma. Vet Radiol Ultrasound 49(2):141–148. Mariette, C., B. Castel, H. Toursel, et al. 2002. Surgical management of and long-term survival after adenocarcinoma of the cardia. Brit J Surg 89:1156–1163. Marik, P.E. and G.P. Zaloga. 2001. Early enteral nutrition in acutely ill patients: A systematic overview. Crit Care Med 29:2264– 2270. Markovitz, J., A. Rappaport, and A.C. Scott. 1949. The function of the hepatic artery in the dog. Am J Dig Dis 16:344–348. Martin, R.A., O.L. Lanz, and K.M. Tobias. 2003. Liver and biliary system. In Textbook of Small Animal Surgery, 3rd edition, pp. 708– 726. D. Slatter, editor. Philadelphia: Elsevier Science. Martins, A.M.C.R.P.F., A. Vasques-Peyser, L.N. Torres, et al. 2008. Retrospective—Systematic study and quantitative analysis of cellular proliferation and apoptosis in normal, hyperplastic, and neoplastic perianal glands in dogs. Vet Comp Oncol, 6(2):71– 79. Mathews, K.A. and M. Barry. 2005. The use of 25% human serum albumin: Outcome and efficacy in raising serum albumin and systemic blood pressure in critically ill dogs and cats. J Vet Emerg Crit Care 15:110–118. Mayhew, P. 2009. Surgical views—Techniques for laparoscopic and laparoscopic-assisted biopsy of abdominal organs. Compendium 31:170–177. Mayhew, P.D., R.W. Richardson, S.J. Mehler, et al. 2006. Choledochal tube stenting for decompression of the extrahepatic portion of the

Alimentary Tract  267 biliary tract in dogs: 13 cases (2002–2005). J Am Vet Med Assoc 228:1209–1214. Mayhew, P.D. and C.W. Weisse. 2008. Treatment of pancreatitisassociated extrahepatic biliary tract obstruction by choledochal stenting in seven cats. J Small Anim Pract 49:133–138. Mazzaferro, E.M., E. Rudloff, and R. Kirby. 2002. The role of albumin replacement in the critically ill veterinary patient. J Vet Emerg Crit Care 12:113–128. McCaw, D., M. Pratt, and R. Walshaw. 1980. Squamous cell carcinoma of the oesophagus in a dog. J Am Anim Hosp Assoc 16:561–563. McKeown, D.B., J.R. Cockshutt, G.D. Partlow, et al. 1984. Dorsal approach to the caudal pelvic canal and rectum. Effect of normal dogs. Vet Surg 13:181–184. McMillan, F.D., T. Barr, and H.C. Feldman. 1985. Functional pancreatic islet cell tumor in a cat. J Amer Anim Hosp Assoc 21:741– 746. McPherron, M.A., S.J. Withrow, H.B. Seim III, et al. 1992. Colorectal leiomyoma in seven dogs. J Am Anim Hosp Assoc 28(1):43–46. Mehlhaff, C.J., M.E. Peterson, A.K. Patnaik, et al. 1985. Insulin producing islet cell neoplasms: Surgical considerations and general management in 35 dogs. J Amer Anim Hosp Assoc 21:607–612. Mellanby, R.J., R. Foale, E. Friend, et al. 2002. Anal sac adenocarcinoma in a Siamese cat. J Feline Med Surg 4(4):205–207. Miller, W.H., W.I. Anderson, and J. McCann. 1991. Necrolytic migratory erythema in a dog with a glucagon-secreting endocrine tumor. Vet Dermatol 2:179–182. Mitsumoto, Y., D.K. Dhar, L. Yu, et al. 1999. FK506 with portal decompression exerts beneficial effects following extended hepatectomy in dogs. Eur Surg Res 31:48–56. Montorsi, M., R. Santambrogio, P. Bianchi, et al. 2002. Perspectives and drawbacks of minimally invasive surgery for hepatocellular carcinoma. Hepatogastroenterology 49(43):56–61. Morello, E., M. Martano, C. Squassino, et al. 2008.Transanal pullthrough rectal amputation for treatment of colorectal carcinoma in 11 dogs. Vet Surg 37:420–426. Morita, Y., M. Takiguchi, J. Yasuda, et al. 1998. Endoscopic ultrasonography of the pancreas in the dog. Vet Radiol Ultrasound 39(6):552–556. Morrell, C.N., M.V. Volk, and J.L. Mankowski. 2002. A carcinoid tumor in the gallbladder of a dog. Vet Pathol 39:756–758. Moss, G., A. Greenstein, S. Levy, et al. 1980. Maintenance of GI function after bowel surgery and immediate enteral full nutrition. I. Doubling of canine colorectal anastomotic bursting pressure and intestinal wound mature collagen content. J Parenter Enteral Nutr 4:535–538. Mueller, M.G., L.L. Ludwig, and L.J. Barton. 2001. Use of closedsuction drains to treat generalized peritonitis in dogs and cats: 40 cases (1997–1999). J Am Vet Med Assoc 219:789–794. Münster, M. and C. Reusch. 1988. Tumoren des exokrinen Pankreas der Katze. Tierärztl Prax 16:317–320. Myers, N.C., III and D.G. Penninck. 1994. Ultrasonographic diagnosis of gastrointestinal smooth muscle tumors in the dog. Vet Radiol Ultrasound 35(5):391–397. Nemeth, T., N. Solymosi, and G. Balka. 2008. Long-term results of subtotal colectomy for acquired hypertrophic megacolon in eight dogs. J Small Anim Pract 49(12):618–624. Neumann, S. and F. Kaup. 2005. Usefulness of Ki-67 proliferation marker in the cytologic identification of liver tumors in dogs. Vet Clin Path 34(2):132–136. Niebauer, G.W. 1993. Rectoanal disease. In Disease Metchanisms in Small Animal Surgery, 2nd edition, pp. 271–284. J.M. Bojrab, editor. Philadelphia: Lea & Febiger.

Noshiro, H., M. Hotokezaka, H. Higashijima, et al. 1996. Gallstone formation and gallbladder bile composition after colectomy in dogs. Dig Dis Sci 41(2):2423–2432. O’Brien, R.T., M. Lani, J. Matheson, et al. 2004. Contrast harmonic ultrasound of spontaneous liver nodules in 32 dogs. Vet Radiol Ultrasound 45(6):547–553. O’Brien, T.D., F. Norton, T.M. Turner, et al. 1990. Pancreatic endocrine tumor in a cat: Clinical, pathological and immunohistochemical evaluation. J Amer Anim Hosp Assoc 26:453–457. Oakes, M.G., G. Hosgood, T.G. Snider, et al. 1993. Esophagotomy closure in the dog: A comparison of a double-layer appositional and two single-layer appositional techniques. Vet Surg 22:451– 456. Oberkirchner, U., K.E. Linder, L. Zadrozn, et al. 2009. Successful treatment of canine necrolytic migratory erythema (superficial necrolytic dermatitis) due to metastatic glucagonoma with octreotide. Vet Dermatol 10:1365–3164. Ogata, A., M. Miyazaki, S. Ohtawa, et al. 1997. Short-term effect of portal vein arterialization on hepatic synthesis and endotoxaemia after extended hepatectomy in dogs. J Gastroenterol Hepatol 12:633–638. Ogilvie, G.K., J.E. Obradovich, R.E. Elmslie, et al. 1991. Efficacy of mitoxantrone against various neoplasms in dogs. J Am Vet Med Assoc 198(9):1618–1621. Owen, L.N. 1980. TNM classification of tumors in domestic animals. World Health Organization: Geneva. Ozmen, M.M., F. Ozmen, and B. Zulfikaroglu. 2008. Lymph nodes in gastric cancer. J Surg Oncol 98:476–481. Pacelli, F., R. Bellantone, G.B. Doglietto, et al. 1991. Risk factors in relation to postoperative complications and mortality after total gastrectomy in aged patients. Am Surg 57:341–345. Palminteri, A. 1966. The surgical management of polyps of the rectum and colon of the dog. J Am Vet Med Assoc 148(7):771–777. Paoletti, X., K. Oba, and T. Burzykowski. 2010. Benefit of adjuvant chemotherapy for resectable gastric cancer—A meta-analysis. JAMA 303:1729–1737. Paoloni, M., D.G. Penninck, and A.S. Moore. 2003. Ultrasonographic and clinicopathologic findings in 21 dogs with intestinal adenocarcinoma. Vet Radiol Ultrasound 43:562–567. Park, D.J., H.J. Lee, H.H. Kim, et al. 2005. Predictors of operative of morbidity and mortality in gastric cancer surgery. Br J Surg 92:1099–1102. Parker, N.R. and D.D. Caywood. 1987. Surgical diseases of the esophagus. Vet Clin N Am Small 17:333–358. Parry, N.M. 2006. Anal sac gland carcinoma in a cat. Vet Pathol 43(6):1008–1009. Parthasarathy, K.R. and K.P. Chandrasekharan. 1966. Fibrosarcoma associated with Spirocerca lupi infection in a dog. Indian Vet J 43:580–582. Patnaik, A.K. 1992. A morphologic and immunohistochemical study of hepatic neoplasms in cats. Vet Pathol 29:405–415. Patnaik, A.K., A.I. Hurvitz, and G.F. Johnson. 1978. Canine gastric adenocarcinoma. Vet Pathol 15:600–607. Patnaik, A.K., A.I. Hurvitz, and P.H. Leiberman. 1980. Canine hepatic neoplasms: A clinicopathologic study. Vet Pathol 17:553– 564. Patnaik, A.K., A.I. Hurvitz, P.H. Leiberman, et al. 1981a. Canine hepatocellular carcinoma. Vet Pathol 18:427–438. Patnaik, A.K., A.I. Hurvitz, P.H. Liberman, et al. 1981b. Canine bile duct carcinoma. Vet Pathol 18:439–444. Patnaik, A.K., P.H. Lieberman, A.I. Hurvitz, et al. 1981c. Canine hepatic carcinoids. Vet Pathol 18:445–453.

268  Veterinary Surgical Oncology Patnaik, A.K., S.K. Liu, and G.F. Johnson. 1976. Feline intestinal adenocarcinoma. A clinicopathologic study of 22 cases. Vet Pathol 13(1):1–10. Pavletic, M.M. 1990. Esophageal reconstruction techniques. In Current Techniques in Small Animal Surgery, 3rd edition, pp. 205– 213. M.J. Bojrab, editor. Philadelphia: Lea & Febiger. Pavletic, M.M. 1994. Stapling in esophageal surgery. Vet Clin N Am Small 24:395–412. Peck, S.A. and Hallenbeck, G.A. 1964. Fecal continence in the dog after replacement of rectal mucosa with ileal mucosa. Surg Gynecol Obstet 119:1312–1320. Penninck, D.G., A.S. Moore, and J. Gliatto. 1998. Ultrasonography of canine gastric epithelial neoplasia. Vet Radiol Ultrasound 39: 342–348. Penninck, D., B. Smyers, C.R.L. Webster, et al. 2003. Diagnostic value of ultrasound in differentiating enteritis from intestinal neoplasia in dogs. Vet Radiol Ultrasound 44:570–575. Petterino, C., M. Martini, and M. Castagnaro. 2004. Immunohistochemical detection of growth hormone (GH) in canine hepatoid gland tumors. J Vet Med Sci 66:569–572. Phillips, B.S. 2001. Tumors of the intestinal tract. In Small Animal Oncology, 3rd edition, pp. 335–346. S.J. Withrow and G.E. MacEwen, editors. Philadelphia: Saunders. Pisani, G., F. Millanta, D. Lorenzi, et al. 2006. Androgen receptor expression in normal, hyperplastic and neoplastic hepatoid glands in the dog. Res Vet Sci 81(2):231–236 Polton, G.A. 2009. Examining the heritability of anal sac gland carcinoma in cocker spaniels. J Small Anim Pract 50(1):57. Polton, G.A. and M.J. Brearley. 2007. Clinical stage, therapy, and prognosis in canine anal sac gland carcinoma. J Vet Intern Med 21(2):274–280. Polton, G.A., V. Mowat, H.C. Lee, et al. 2006. Breed, gender, and neutering status of British dogs with anal sac gland carcinoma. Vet Comp Oncol 4:125–131. Polton, G.A., R.N. White, M.J. Brearly, et al. 2007. Improved survival in a retrospective cohort of 28 dogs with insulinoma. J Small Anim Pract 48:151–156. Post, G. and A.K. Patnaik. 1992. Nonhematopoietic hepatic neoplasms in cats: 21 cases (1983–1988). J Am Vet Med Assoc 201:1080–1082. Prater, R.M., B. Flatland, S.J. Newman, et al. 2000. Diffuse annular fusiform adenorcarcinoma in a dog. J Am Anim Hosp Assoc 36(2):169–173. Prather, A.B., C.R. Berry, and D.E. Thrall. 2005. Use of radiography in combination with computed tomography for the assessment of noncardiac thoracic disease in the dog and cat. Vet Radiol Ultrasoun 46:114–121. Priester, W.A. 1974. Data from eleven United States and Canadian colleges of veterinary medicine on pancreatic carcinoma in domestic animals. Cancer Res 34:1372–1375. Ralphs, S.C., C.R. Jessen, and A.J. Lipowitz. 2003. Risk factors for leakage following intestinal anastomosis in dogs and cats: 115 cases (1991–2000). J Am Vet Med Assoc 223:73–77. Ranen, E., G. Dank, E. Lavy, et al. 2008. Oesophageal sarcomas in dogs: Histological and clinical evaluation. Vet J 178:78–84. Ranen, E., E. Lavy, I. Aizenberg, et al. 2004. Spirocercosis-associated esophageal sarcomas in dogs: A retrospective study of 17 cases (1997–2003). Vet Parasitol 119:209–221. Ranen, E., M.H. Shamir, R. Shahar, et al. 2004. Partial esophagectomy with single layer closure for treatment of oesophageal sarcomas in 6 dogs. Vet Surg 33:428–434. Rannou, B., P. Hélie, and C. Bédard. 2009. Rectal plasmacytoma with intracellular hemosiderin in a dog. Vet Pathol 46(6):1181– 1184.

Rasmussen, L. 2003. Stomach. In Textbook of Small Animal Surgery, 3rd edition, pp. 592–640. D. Slatter, editor. Philadelphia: Elsevier Science. Rawlings, C.A. and E.W. Howerth. 2004. Obtaining quality biopsies of the liver and kidney. J Am Anim Hosp Assoc 40:352–358. Reimer, M.E., M.S. Reimer, G.K. Saunders, et al. 1999. Rectal ganglioneuroma in a dog. J Am Anim Hosp Assoc 35(2):107–110. Ribelin, W.E. and W.S. Bailey. 1958. Oesophageal sarcomas associ­ ated with Spirocerca lupi infection in the dog. Cancer 6:1242–1246. Ridgway, R.L. and P.F. Suter. 1979. Clinical and radiographic signs in primary and metastatic esophageal neoplasms of the dog. J Am Vet Med Assoc 174:700–704. Rivers, B.J., P.A. Walter, D.A. Feeney, et al. 1997. Ultrasonographic features of intestinal adenocarcinoma in five cats. Vet Radiol Ultrasound 38(4):300–306. Robben, J.H., Y. Pollak, J. Kirpensteijn, et al. 2005. Comparison of ultrasonography, computed tomography, and single-photon emission computed tomography for the detection and localization of canine insulinoma. J Vet Intern Med 19:15–22 Robben, J.H., H.A. Visser-Wisselaar, G.R. Rutteman, et al. 1997. Vitro and in vivo detection of functional somatostatin receptors in canine insulinomas. J Nucl Med 38:1036–1042. Roche, A., A. Raisonnier, and M.C. Gillon-Savouret. 1982. Pancreatic venous sampling and arteriography in localizing insulinomas and gastrinomas. Procedure and results in 55 cases. Radiology 145:621–627. Rolfe, D.S., D.C. Twedt, and H.B. Seim. 1994. Chronic regurgitation or vomiting caused by oesophageal leiomyoma in three dogs. J Am Anim Hosp Assoc 30:425–430. Romano, F., C. Franciosi, R. Caprotti, et al. 2005. Hepatic surgery using the Ligasure vessel sealing system. World J Surg 29(1): 110–112. Rosin, E. 1975. Surgery of the esophagus. Vet Clin N Am 5:557–564. Rosol, T.J., C.C.A. Capen, J.A. Danks, et al. 1990. Identification of parathyroid hormone-related protein in canine apocrine adenocarcinoma of the anal sac. Vet Pathol 27(2):89–95. Ross, T.J., T.D. Scavelli, D.T. Matthiesen, et al. 1991. Adenocarcinoma of the apocrine glands of the anal sac in dogs: A review of 32 cases. J Am Anim Hosp Assoc 27(3):349–355. Ross, H.M., J.A. Smelstoys, A.S. Davis, et al. 2006. Photodynamic therapy with motexafin lutetium for rectal cancer: A preclinical model in the dog. J Surg Res 135:323–330. Ross, W.E. and A.D. Pardo. 1993. Evaluation of an omental pedicle extension technique in the dog. Vet Surg 22:37–43. Rossmeisl, J.H., S.D. Forrester, J.L. Robertson, et al. 2002. Chronic vomiting associated with a gastric carcinoid in a cat. J Am Anim Hosp Assoc 38:61–68. Russel, K.N., S.J. Mehler, K.A. Skorupski, et al. 2007. Clinical and immunohistochemical differentiation of gastrointestinal stromal tumors from leiomyosarcomas in dogs: 42 cases (1990–2003). J Am Vet Med Assoc 230:1329–1333. Ryan, S., H. Seim III, C. Macphail, et al. 2006 Comparison of biofragmentable anastomosis ring and sutured anastomoses for subtotal colectomy in cats with idiopathic megacolon. Vet Surg 35(8): 740–748. Saini, S. 1997. Imaging of the hepatobiliary tract. N Engl J Med 336:1889–1894. Saint, J.H. and F.C. Mann. 1929. Experimental surgery of the esophagus. Arch Surg 18:2324–2338. Saiura, A., J. Yamamoto, R. Koga, et al. 2006. Usefulness of LigaSure for liver resection: Analysis by randomized clinical trial. Am J Surg 192(1):41–45.

Alimentary Tract  269 Sakurada, A., T. Takahara, T.C. Kwee, et al. 2009. Eur Radiol 19:1461–1469. Samii, V.F., D.S. Biller, and P.D. Koblik. 1998. Normal cross-sectional antomy of the feline thorax and abdomen: Comparison of computed tomography and cadaver anatomy. Vet Radiol Ultrasound 39:504–511. Samii, V.F., D.S. Biller, and P.D. Koblik. 2005. Magnetic resonance imaging of the normal feline abdomen: An anatomic reference. Radiol Ultrasound 40:486–490. Sapin, E., A. Centonze, R. Mooq, et al. 2006. Transanal coloanal anastomosis for Hirschsprung’s disease: Comparison between endorectal and perirectal pull through procedures. Eur J Pediatr Surg 16:312–317. Sarathchandra, S.K., J.A. Lunn, and G.B. Hunt. 2009. Ligation of the caudal mesenteric artery during resection and anastomosis of the colorectal junction for annular adenocarcinoma in two dogs. Aust Vet J 87(9):356–359. Sato, K., Y. Hikasa, T. Morita, et al. 2002. Secondary erythrocytosis associated with high plasma erythropoietin concentrations in a dog with cecal leiomyosarcoma. J Am Vet Med Assoc 220(4):486–490. Saunders, B.W. and K.M. Tobias. 2003. Pneumoperitoneum in dogs and cats: 39 cases (1983–2002). J Am Vet Med Assoc 223:462– 468. Schirmer, W.J., W.S. Melvin, R.M. Rush, et al. 1995. Indium-111pentetreotide scanning versus conventional imaging techniques for localization of gastrinoma. Surgery 118:1105–1113. Schunk, C. 1990. Esophagus. In Current Techniques in Small Animal Surgery, 3rd edition, pp. 201–205. M.J. Bojrab, editor. Philadelphia: Lea & Febiger. Seaman, R.L. 2004. Exocrine pancreatic neoplasia in the cat: A case series. J Am Anim Hosp Assoc 40(3):238–245. Seibold, H.R., W.S. Bailey, B.F. Hoerlein, et al. 1955. Observations on possible relation of malignant esophageal tumors and Spirocerca lupi lesions in dog. Am J Vet Res 16:5–14. Seiler, R.J. 1979. Colorectal polyps of the dog: A clinicopathologic study of 17 cases. 1979. J Am Vet Med Assoc 174(1):72–75. Sellon, R.K., K. Bissonnette, and S.E. Bunch. 1996. Long-term survival after total gastrectomy for gastric adenocarcinoma in a dog. J Vet Intern Med 10:333–335. Selting, K.A. 2007. Intestinal tumors. In Withrow & MacEwewn’s Small Animal Oncology, 4th edition, pp. 491–503. S.J. Withrow and D.M. Vail, editors. Philadelphia: Saunders. Shabadash, S.A. and Zelikina, T.I. 1995. The sex dimorphism of the hepatoid circumanal glands in the dog and the dynamics of its development. Izvestiia AkademiiNauk Seriia Biologiccheskaia 5:590–605. Shales, C.J., J. Warren, D.M. Anderson, et al. 2005. Complications following full-thickness small intestinal biopsy in 66 dogs: A retrospective study. J Small Anim Pract 46:317–321. Shamir, M.H., R. Shahar, D.E. Johnston, et al. 1999. Approaches to esophageal sutures. Comp Cont Educ Pract 21:414–421. Sharpe, A., M.J. Cannon, V.M. Lucke, et al. 2000. Intestinal haemangiosarcoma in the cat: Clinical and pathological features of four cases. J Small Anim Pract 41(9):411–415. Shelley, B.A. 2002. Use of the carbon dioxide laser for perianal and rectal surgery. Vet Clin N Am Sm Anim Pract 32(3):621– 637. Siliart, B. and F. Stambouli. 1996. Laboratory diagnosis of insulinoma in the dog: A retrospective study and a new diagnostic procedure. J Small Anim Pract 37(8):367–370. Simeonov, R. and Simeonova, G. 2008a. Computer-assisted nuclear morphometry in the cytological evaluation of canine perianal adenocarcinomas. J Comp Pathol 139(4):226–230.

Simeonov, R. and Simeonova, G. 2008b. Quantitative analysis in spontaneous canine anal sac gland adenomas and carcinomas. Res Vet Sci 85(3):559–562. Simpson, K.W. and N.L. Dykes. 1997. Diagnosis and treatment of gastrinoma. Semin Vet Med Surg Sm Anim 12:274. Slawienski, M.J., G.E. Mauldin, N.G. Mauldin, et al. 1997. Malignant colonic neoplasia in cats: 46 cases (1990–1996). J Am Vet Med Assoc 211(7):878–881. Smallwood, J.E. and T.F. George. 1993a. Anatomic atlas for computed tomography in the mesaticephalic dog: Thorax and cranial abdomen. Vet Radiol 34:65–84. Smallwood, J.E. and T.F. George. 1993b. Anatomic atlas for computed tomography in the mesaticephalic dog: Caudal abdomen and pelvis. Vet Radiol 34:143–167. Spugnini, E.P., I. Dotsinsky, N. Mudrov, et al. 2007. Biphasic pulses enhance bleomycin efficacy in a spontaneous canine perianal tumors model. J Experimental Clin Cancer Res 26(4):483–487. Spugnini, E.P., I. Dotsinkky, N. Mudrov, et al. 2008. Adjuvant electrochemotherapy for incompletely excised anal sac carcinoma in a dog. In Vivo 22(1):47–49. Spugnini, E.P., M. Filipponi, L. Romani, et al. 2007. Local control and distant metastasis after electrochemotherapy of a canine anal melanoma. In Vivo 21(5):897–899. Staatz, A.J., E. Monnet, and H.B. Seim. 2002. Open peritoneal drainage versus primary closure for the treatment of septic peritonitis in dogs and cats: 42 Cases (1993–1999). Vet Surg 31:174– 180. Stanclift, R.M. and S.D. Gilson. 2004. Use of cisplatin, 5-fluorouracil, and second-look laparotomy for the management of gastrointestinal adenocarcinoma in three dogs. J Am Vet Med Assoc 225:1412–1417. Steiner, J.M. and D.S. Bruyette. 1996. Canine insulinoma. Comp Cont Ed Pract Vet 18:13–25. Storck, B.H., E.J. Rutgers, E. Gortzak, et al. 1991. The impact of the CUSA ultrasonic dissection device on major liver resections. Neth J Surg. 43:99–101. Straw, R.C., J.L. Tomlinson, G. Constantinescu, et al. 1987. Use of a vascular skeletal muscle graft for canine esophageal reconstruction. Vet Surg 16:155–156. Straw, R.C., S.J. Withrow, H.B. Seim III, et al. 1994. Intraoperative radiation for management of metastatic carcinoma to the lumbar lymph nodes. Proceedings of the 14th Annual Conference of the Veterinary Cancer Society, pp. 125–126. October 23–25, Townsend, TN. Strombech, D.R. 1978. Clinicopathologic features of primary and metastatic neoplastic disease of the liver in dogs. J Am Vet Med Assoc 173:267–269. Swann, H.M. and D.E. Holt. 2002. Canine gastric adenocarcinoma and leiomyosarcoma: A retrospective study of 21 cases (1986– 1999) and literature review. J Am Anim Hosp Assoc 38:157– 164. Sweet, D.C., E.M. Hardie, and E.A. Stone. 1994. Preservation versus excision of the ileocolic junction during colectomy for megacolon: A study of 22 cats. J Small Anim Pract 35(7):358–363. Swenson, O. and A.H. Bill. 1948. Resection of rectum and rectosigmoid with preservation of the sphincter for benign spastic lesions producing megacolon: An experimental study. Surgery 24: 212–220. Swiderski, J. and S.J. Withrow. 2009. A novel surgical stapling technique for rectal mass removal: A retrospective analysis. J Am Anim Hosp Assoc 45:67–71. Sykes, G.P. and B.J. Cooper. 1982. Canine intestinal carcinoids. Vet Pathol 19(2):120–131.

270  Veterinary Surgical Oncology Takiguchi, M., J. Yasuda, A. Hashimoto, et al. 1997. Oesophageal/ gastric adenocarcinoma in a dog. J Am Anim Hosp Assoc 33: 42–44. Tang, J., S. Le, L. Sun, et al. 2010. Copy number abnormalities in sporadic canine colorectal cancers. Genome Research 20(3): 341–350. Thompson, J.P., M.M. Cristopher, and G.W. Ellison. 1992. Paraneoplastic leukocytosis associated with a rectal adenomatous polyp in a dog. J Am Vet Med Assoc 201(5):737–738. Tobias, K.M. 2007. Surgical stapling devices in veterinary medicine: A review. Vet Surg 36:341–349. Tobin, R.L., R.W. Nelson, M.D. Lucroy, et al. 1999. Outcome of surgical versus medical treatment of with beta cell neoplasia: 39 cases (1990–1997). J Am Vet Med Assoc 215:226–230. Torres, S., D.D. Caywood, T.D. O’Brien, et al. 1997. Resolution of superficial necrolytic dermatitis following excision of a glucagon secreting pancreatic neoplasm in a dog. J Am Anim Hosp Assoc 33:313–319. Torres, S., K. Johnson, P. McKeever, et al. 1997. Superficial necrolytic dermatitis and a pancreatic endocrine tumor in a dog. J Small Anim Pract 38:246–250. Tozon, N., V. Kodre, G. Sersa, et al. 2005. Effective treatment of perianal tumors in dogs with electrochemotherapy. Anticancer Research 25(2A):839–845. Trevor, P.B., G.K. Saunders, D.R. Waldron, et al. 1993. Metastatic extramedullary plasmacytoma of the colon and rectum in a dog. J Am Vet Med Assoc 203(3):406–409. Trifonidou, M.A., J. Kirpensteijn, and J.H. Robben. 1998. A retro­ spective evaluation of 51 dogs with insulinoma. Vet Q 20: S114–S115. Trout, N.J., J.R. Berg, M.C. McMillan, et al. 1995. Surgical treatment of hepatobiliary cystadenomas in cats: Five cases (1988–1993). J Am Vet Med Assoc 206:505–507. Trow, A.V., E.A. Rozanski, A.M. Delaforcade, et al. 2008. Evaluation of use of human albumin in critically ill dogs: 73 cases (2003– 2006). J Am Vet Med Assoc 233:607–612. Turek, M.M., L.J. Forrest, W.M. Adams, et al. 2003. Postoperative radiotherapy and mitoxantrone for anal sac adenocarcinoma in the dog: 15 cases (1991–2001). Vet Comp Oncol 1(2):94–104. Turek, M.M. and S.J. Withrow. 2007. Perianal tumors. In Withrow & MacEwewn’s Small Animal Oncology, 4th edition, pp. 503–510. S.J. Withrow and D.M. Vail, editors. Philadelphia: Saunders. Turk, M.A.M., A.M. Gallina, and T.S. Russel. 1981. Nonhematopoietic gastrointestinal neoplasia in cats: A retrospective study of 44 cases. Vet Pathol 18:614–620. Turrel, J.M. and A.P. Theon. 1986. Single high dose irradiation for selected canine rectal carcinomas. Vet Radiol Ultrasound 27(5):141–145. Ueno, H., T. Kadosawa, H. Isomura, et al. 2002. Perianal rhabdomyosarcoma in a dog. J Small Anim Pract 43(5):217–220. Ullman, S.L., M.M. Pavletic, and G.N. Clark. 1991. Open intestinal anastomosis with surgical stapling equipment in 24 dogs and cats. Vet Surg 20:385–391. Vail, D.M., S.J. Withrow, P.D. Schwarz, et al. 1990. Perianal adenocarcinoma in the canine male: A retrospective study of 41 cases. J Am Anim Hosp Assoc 26(3):329–334. Valerius, K.D., B.E. Powers, M.A. McPherron, et al. 1997. Adeno­ matous polyps and carcinoma in situ of the canine colon and rectum: 34 cases (1982–1994). J Am Anim Hosp Assoc 33(2): 156–160. Van der Gaag, I. and R.P. Happe. 1990. The histological appearance of peroral small intestinal biopsies in clinically healthy dogs and dogs with chronic diarrhea. Zentrabl Veterinarmed 37:401–416.

van Laarhoven, C.J., W.E. Hueting, M.E. Schipper, et al. 2004. Ileoneorectal anastomosis: Medium- and long-term follow-up of 37 patients. Digestive Surgery 21(5–6):371–378. Vananjee, S.C., L.J. Bubenik, G. Hosgood, et al. 2006. Evaluation of hemorrhage, sample size, and collateral damage for 5 hepatic biopsy methods in dogs. Vet Surg 35(1):86–93. Wang, K.Y., D.L. Panciera, R.K. Al-Rukibat et al. 2004. Accuracy of ultrasound-guided fine-needle aspiration of the liver and cytologic findings in dogs and cats: 97 cases (1990–2000). J Am Vet Med Assoc 224(1):75–78. Webb, C.B., and C. Trott. 2008. Laparoscopic diagnosis of pancreatic disease in dogs and cats. J Vet Intern Med 22:1263–1266. Webster, G.V. 1955. Halstedian principles in the practice of plastic and reconstructive surgery. Stanford Med Bull 13:315–316. Weisman, D.L., D.D. Smeak, and S.J. Birchard. 1999. Comparison of a continuous suture pattern with a simple interrupted pattern for enteric closure in dogs and cats: 83 cases (1991–1997). J Am Vet Med Assoc 214(10):1507–1510. Weisse, C., C.A. Clifford, D. Holt, et al. 2002. Percutaneous arterial embolization and chemoembolization for treatment of benign and malignant tumors in three dogs and a goat. J Am Vet Med Assoc 221(10):1430–1436. White, R.A.S. and N.T. Gorman. 1987. The clinical diagnosis and management of rectal and pararectal tumours in the dog. J Small Anim Pract 28:87–107. Willard, M.D., R.W. Dunstan, and J. Faulkner. 1988. Neuroendocrine carcinoma of the gallbladder in a dog. J Am Vet Med Assoc 192:926–928. Willard, M.D., S.L. Lovering, N.D. Cohen, et al. 2001. Quality of tissue specimens obtained endoscopically from the duodenum of dogs and cats. J Am Vet Med Assoc 219:474–479. Williams, L.E., J.M. Gliatto, R.K. Dodge, et al. 2003. Carcinoma of the apocrine glands of the anal sacs in dogs: 113 cases (1985–1995). J Am Anim Hosp Assoc 223(6):825–831. Williams, J.M. and J.D. Niles. 1999. Use of omentum as a physiologic drain for treatment of chylothorax in a dog. Vet Surg 28:61–65. Williams, J. and J. Niles. 2005. The large intestine and rectum. In BSAVA Manual of Canine and Feline Abdominal Surgery, pp. 130– 132. J. Williams and J. Niles, editors. Quedgeley, Gloucester, UK: British Small Animal Veterinary Association. Wilson, G.P. and H.M. Hayes Jr. 1979. Castration for treatment of perianal gland neoplasms in the dog. J Am Vet Med Assoc 174(12):1301–1303. Withrow, S.J. 2007a. Esophageal cancer. In Small Animal Clinical Oncology, 4th edition, pp. 477–478. S.J. Withrow and D.M. Vail, editors. St. Louis: Saunders. Withrow, S.J. 2007b. Gastric cancer. In Small Animal Clinical Oncology, 4th edition, pp. 480–483. S.J. Withrow and D.M. Vail, editors. St. Louis: Elsevier Science. Withrow, S.J. 2007c. Exocrine cancer of the pancreas. In Withrow and MacEwen’s Small Animal Clinical Oncology, 4th edition, pp. 479– 480. S.J. Withrow and D.M. Vail, editors. Philadelphia: Saunders. Yagil-Kelmer, E., C. Wagner-Mann, and F.A. Mann. 2006. Postoperative complications associated with jejunostomy tube placement using the interlocking box technique compared with other jejunopexy methods in dogs and cats: 76 cases (1999–2003). J Vet Emerg Crit Care 16:S14–S20. Yamanaka, H., M. Nishi, T. Kanemaki, et al. 1989. Preoperative nutritional assessment to predict postoperative complication in gastric cancer patients. J Paren Enteral Nutr 13:286–291. Yamaya, Y., K. Niizeki, J. Kim, et al. 2004. Anaphylactoid response to Optison(R) and its effects on pulmonary function in two dogs. J Vet Med Sci 66:1429–1432.

Alimentary Tract  271 Yanoff, S.R., M.D. Willard, H.W. Boothe, et al. 1992. Short bowel syndrome in four dogs. Vet Surg 21:217–222. Yasuda, K., N. Shiraishi, Y. Adachi, et al. 2001. Risk factors for complications following resection of large gastric cancer. Br J Surg 88:873–877. Yoon, H.Y. and F.A. Mann. 2008. Bilateral pubic and ischial osteotomy for surgical management of caudal colonic and rectal masses in six dogs and a cat. J Am Vet Med Assoc 232:1016–1020. Young, K.D., G.E. Mauldin, G. Hosgood, et al. 2005. Perianal malignant melanoma in a dog. J Vet Intern Med 19(4):610–612.

Zaloga, G.P., L. Bortenschlager, K. Ward-Black et al. 1992. Immediate postoperative enteral feeding decreases weight loss and improves wound healing after abdominal surgery in rats. Crit Care Med 20:115–118. Zerbe, C.A. and R.J. Washabau. 2000. Gastrointestinal endocrine disease. In Textbook of Veterinary Internal Medicine, 5th edition, p. 1500. S.J. Ettinger and E.C. Feldman, editors. Philadelphia: Saunders.

8 Respiratory tract and thorax Marina Martano, Sarah Boston, Emanuela Morello, Stephen J. Withrow

Rhinotomy The nasal cavities can be surgically approached by a dorsal, ventral, or lateral rhinotomy. The latter procedure gives access only to the nasal vestibule, and it is therefore rarely indicated in oncologic surgery. Clinical workup and biopsy principles Diagnosis of nasal tumors can be made by combining clinical findings with imaging and biopsy results. The most frequent clinical sign is unilateral epistaxis; however, it can become bilateral once the tumor has invaded into the contralateral nasal cavity. Other signs include nasal and/or ocular discharge, occasional sneezing, signs of occlusion of the nasal cavities, facial deformities (Figure 8.1), and exophthalmos. Symptoms may be present for approximately 1–6 months prior to diagnosis. Clinical suspicion is confirmed by biopsy, which can also rule out other nasal diseases, such as fungal or bacterial infections, nasal foreign bodies, parasites, or systemic diseases (ehrlichiosis, leishmaniasis, coagulopathies, etc.). In cats, nasopharyngeal polyps should also be considered. Blood testing, while usually unremarkable, is necessary to evaluate the overall condition of the patient, and a coagulation profile is always recommended to evaluate the risk of hemorrhage from biopsy procedures or rhinotomy. Biopsy specimens can be obtained either blindly or with endoscopic guidance. With the latter, a rigid endoscope is introduced through the nares and, after the procedure has been completed, a small biopsy sample is withdrawn with forceps. With this procedure, a falsenegative result (i.e., inflammatory rhinitis) may be

obtained because of inadequate and/or superficial sampling. Better results can be achieved using a large-bore cannula or tube biopsy technique, applicable mainly in medium- to large-sized dogs. The animal is placed in sternal recumbency, under general inhalation anesthesia, with the nose parallel to the operating table (as for endoscopy). A cuffed endotracheal tube is always placed and the pharynx occluded with gauze sponges to avoid inhalation of debris or secretions. A rigid plastic cannula, such as the outer sleeve of a metal catheter or a rigid plastic urinary catheter, attached to a 12 mL syringe is used for collection of the biopsy sample (Turek and Lana 2007). To avoid entering the cribriform plate and the brain, the distance from the tip of the nose to the medial canthus of the eye is measured on the instrument and identified using a piece of tape or a pen mark (Figure 8.2A). The plastic tube attached to the syringe is then introduced past the wing of the nostril, no further than the mark on the tube (Figure 8.2B), and moved in and out repeatedly while suctioning to collect neoplastic material (Figure 8.2C). The moment the tumor is entered can be perceived as an increased resistance to the passage of the cannula. After this procedure, bleeding often occurs; however, it is usually moderate and selflimiting. In cases of copious bleeding, the nasal cavity can be filled with a gauze sponge soaked in diluted epinephrine (1:100,000); alternatively, oxymetazoline 0.05% can be used. The withdrawn material is evacuated from the tube by using the syringe filled with air and rolled onto a dry gauze to eliminate the blood. It is then placed in 10% buffered formalin and sent for histologic examination.

Veterinary Surgical Oncology, First Edition. Edited by Simon T. Kudnig, Bernard Séguin. © 2012 by John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.

273

274  Veterinary Surgical Oncology

Figure 8.1.  (A) Gross appearance of a cat with a nasal tumor. (B) Gross appearance of a dog with a nasal tumor. The nasal profile is disfigured to this degree only late in the course of the disease.

(a)

(b)

(c)

Figure 8.2.  Nasal biopsy with cannula. (A) The dog is positioned in sternal recumbency, with the nose parallel to the table; the distance between the medial canthus of the eye and the tip of the nose of the dog is measured on a large-bore urinary catheter. (B) The catheter is cut and a mark or a piece of tape is applied at the measured distance. (C) The catheter is connected to a 12 mL syringe and introduced into the nostril, applying a moderate amount of pressure to the plunger until the mass is penetrated. Suction is applied to the syringe, the cannula is withdrawn, and the material collected.

The same procedure can be performed with a small curette introduced via the nostril, especially in small dogs and cats; the material is scooped out and the nasal cavity flushed with cold saline in order to stop the bleeding. This core biopsy usually allows a good specimen to be obtained and can be superior to those collected via rigid endoscopy (Ogilvie and LaRue 1992). Brush cytology is diagnostic only in 50% of cases and is not usually recommended. Imaging techniques Each imaging technique should be performed before the biopsy is taken, as the blood and exudates can create artifacts that are difficult to interpret.

Survey radiographs of the skull can aid in the diagnosis, since they can reveal increases in soft tissue density inside the nasal cavities and frontal sinuses, bone erosion or new bone formation, or the presence of radiodense foreign bodies. The best views are the open mouth ventrodorsal and dorsoventral intra-oral projections (Figure 8.3B); oblique and rostrocaudal projections should be used for evaluation of frontal sinuses (Figure 8.3A). The open-mouth ventrodorsal view provides the best information on the entire nasal cavity, since it avoids superimposition with the mandible, and should always be performed. While observed radiographic changes may be suggestive of neoplasia, they are not pathognomonic, since severe fungal or inflammatory diseases can have a

Respiratory Tract and Thorax  275

Figure 8.3.  (A) Rostrocaudal view of the frontal sinuses in a dog. A radiodense material is present in the left (L) sinus; a differential between mucus and neoplastic material is not possible with radiography. (B) Open mouth ventrodorsal view of the nasal cavities in a cat. The tissue density is increased in the right nasal cavity. Care must be taken in interpreting the overlapping effect of oral tissues (i.e., the tongue).

(a)

(b)

(c)

Figure 8.4.  Contrast-enhanced CT imaging of the nasal cavity. (A) Some hyperdense material referable to neoplastic infiltrate (lymphoma) is evident in the left nasal cavity. (B) Neoplastic invasion of the left nasal cavity with massive erosion of the turbinates and lysis of the nasal septum in a dog. (C) Neoplastic invasion of the nasal cavity with lysis of the right orbital bone and invasion of the retroorbital region in a cat.

similar radiographic appearance (O’Brien et al. 1996). Increased soft tissue density of the frontal sinuses without bone involvement should be considered with caution, since it can be due only to obstruction to the outflow of mucous secretion (“frontal mucocele”) and not to the tumor itself. Three-view thoracic radiographs complete tumor staging. Survey or contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI) techniques are superior to radiography, as they allow more accurate assessment of the extent of the tumor and are more sensitive to early changes due to existing pathology

(Figure 8.4A–C) (Thrall et al. 1989). They are also necessary for precise 3D planning of radiation therapy. Rhinoscopy is performed after the radiographs are taken. This technique allows direct visualization of the nasal cavities and is usually performed both with a rostral and caudal approach. The caudal approach is generally performed first, with a flexible endoscope introduced from the oral cavity, over the free edge of the soft palate, to visualize the caudal nasal passage, nasopharynx, and choanae. If a mass is present at this level, a biopsy forceps is introduced in the endoscope and a sample is taken. Rostral rhinoscopy is performed using a rigid scope introduced from the nostril, and it provides a view of

276  Veterinary Surgical Oncology

Figure 8.5.  Endoscopic view of nasal masses in dog. This imaging technique permits collection of a sample for histology. (A) Melanoma in the rhinopharynx. (B) Polyp in the rhinopharynx. A ventral rhinotomy was performed in this dog. (Image courtesy of Dr. Roberta Caccamo)

the nasal meatus, ventral concha, alar cartilage, and the ventral and common nasal passages (Figure 8.5A, B). A rigid cystoscope is the best tool for this procedure, but it may be too large for use in cats and small dogs. The endoscope-guided biopsy is taken once the endoscopy is completed. Surgical techniques Rhinotomy, both via a dorsal or ventral approach, is indicated to collect biopsy samples from the nasal cavity when other procedures have failed to obtain a diagnosis and for debulking purposes. Surgery alone is not curative for nasal tumors and must always be associated with other treatment protocols, both in an adjuvant or neoadjuvant setting. In both procedures, after the animal is anesthetized and intubated, the pharynx is packed with sponge gauzes to avoid inhalation of blood or secretions during surgery. At the time of extubation, the tracheal tube is extracted with the cuff slightly inflated to facilitate the removal of accumulated blood or secretions. Dorsal rhinotomy The dorasal rhinotomy is the standard approach in dogs, but it is rarely performed in cats. It gives access to the entire nasal cavity and frontal sinus. The animal is positioned in ventral recumbency, with the nose parallel to the operating table (Hedlund 1998; Nelson 2003c). A midline skin incision is made from caudal to the nasal planum to the nasal canthus of the eyes. If the frontal sinuses are to be explored, the incision is extended caudally to a line connecting the zygomatic process of the frontal bones. The incision is deepened to the subcutaneous tissue and periosteum (Figure 8.6A). This latter is then elevated and reflected laterally on both sides of the

incision to expose the underlying bone. A unilateral or bilateral bone flap is then created using an osteotome or an oscillating saw (Figure 8.6B) and reflected rostrally (Figure 8.6C), preserving its attachment to the nasal ligaments. Alternatively, it can be completely detached and kept moist with sponges if it is to be repositioned, or it can be discarded. Once the nasal cavity is exposed, it is flushed with ice-cold sterile saline solution and suctioned to remove debris and blood clots before it is inspected. If the indication for rhinotomy is collection of a biopsy, this is performed with scissors or a curette. If the surgery has curative intent, the turbinates of one or both sides, depending on tumor extension and as previously evaluated by the imaging studies, are removed (Figure 8.6D). The frontal sinuses are inspected, mucous secretions are eliminated, and tumor debulking is also completed in this area if necessary. Bleeding can be copious, but can be controlled by flushing with ice-cold saline, applying digital pressure with gauzes soaked in diluted epinephrine (1:100,000), or careful use of electrocautery. Temporary carotid artery occlusion can facilitate the control of intraoperative bleeding in dogs (Hedlund et al. 1983). This procedure, however, should be avoided in cats. The nasal cavity is flushed and suctioned before closure to eliminate the majority of blood clots and reduce obstruction during the postoperative period. The bone flap can be repositioned, if not invaded by the tumor, or discarded (particularly if adjuvant orthovoltage radiation will follow). If preserved, the flap is sutured in place with nonabsorbable monofilament sutures prepositioned in three to four holes drilled on the sides of the bone and tied after the flap is in place. Metal wires should not be used if follow-up radiotherapy is planned. If the bone is discarded, the soft tissues are

Respiratory Tract and Thorax  277

Figure 8.6.  Dorsal rhinotomy. (A) After a midline incision, the skin and subcutaneous tissues are reflected laterally and the nasal bone exposed. (B) A bone flap is created using a high speed burr or an oscillating saw. (C) The bone flap is elevated with an osteotome and discarded (or kept moist in sterile gauze sponges). (D) The nasal cavity is inspected and the turbinates removed. The cavity is then washed with cold sterile saline solution. (E) The subcutaneous tissues and skin are closed using the standard technique.

closed with absorbable monofilament sutures in a simple continuous pattern (Figure 8.6E). The skin is sutured routinely. To avoid subcutaneous emphysema, which always occurs if the bone flap is not replaced, a small (1 cm) gap can be left in the caudal (frontal) end of the suture line (Birchard 1986), or a tube drain is positioned in the frontal sinuses and nasal cavity. The gap will naturally close within a few days. Packing the nasal cavities with sponges can help to control severe hemorrhage, but it obstructs nasal air-flow, thus creating discomfort to the animal, and subsequent removal 48 hours after surgery can be painful. Ventral rhinotomy The ventral approach provides access to the nasal cavities, nasopharynx, and the rostral half of the frontal sinuses. It has a better cosmetic result as compared to the dorsal approach, the risk of subcutaneous emphysema is limited, and the dog is already in the correct position for temporary carotid artery occlusion, if this is required. One disadvantage is the risk for oronasal

fistula formation and the fact that the frontal sinus is not completely visible. Holmberg et al. (1989) found that animals treated with this approach seemed to experience less discomfort and returned to their preoperative eating and grooming habits faster than those treated with the dorsal procedure. If this approach to the nasal cavity is chosen, adjuvant radiotherapy can be delivered only by megavoltage machines. The animal is positioned in dorsal recumbency, with the front legs secured caudally, parallel to the chest (Hedlund 1998; Nelson 2003c). The mouth is kept wide open by securing the mandible dorsally with a tape. A midline mucoperiosteal incision is made from the canine tooth to the level of the fourth premolar. The palate periosteum is undermined and elevated laterally (Figures 8.7A, 8.8A). Alternatively, the incision can be U-shaped, parallel to the dental arcade, with the U open caudally; the mucoperiosteal flap is then reflected caudoventrally. Care is taken to avoid damage to the major palatine arteries that emerge from the major palatine foramen at the level of the fourth maxillary premolar tooth and run

278  Veterinary Surgical Oncology

Figure 8.7.  Ventral rhinotomy in a dog. (A) A midline incision is made through the mucoperiosteal tissues of the hard palate with two shorter perpendicular incisions made at the extremities of the first incision. (B) The tissues are elevated and kept retracted with four stay sutures. A flap of the underlying bone is created with an oscillating saw and discarded. The nasal cavity is inspected, and the neoplastic mass and turbinates are removed. (C) After lavage of the nasal cavity with cold sterile saline, the soft tissues are closed in one or two layers with simple interrupted absorbable sutures. This is the same dog as in Figure 8.5B.

rostrally, midway between the midline and the dental arcade. A rectangular bone flap is created with a bone saw, air drill, or osteotome (Figure 8.8B) and discarded. Hemorrhage is controlled as previously described, but electrocautery is avoided if possible. Once the nasal cavities are entered, the gross tumor is removed with a curette (Figures 8.7B, 8.8C), or a biopsy is taken as described for the dorsal approach. The nasal cavity is then flushed with cool sterile saline solution. The mucoperiosteal tissue is closed in a one- or twolayer pattern with simple interrupted sutures using 3-0 to 5-0 monofilament absorbable material (Figures 8.7C, 8.8D). If the nasopharynx is to be inspected, the midline incision is extended caudally to the level 5–10 mm rostral to the edge of the soft palate. The incisional margins are kept open with stay sutures, and the exploration performed. The soft palate defect is closed in two or three layers with simple interrupted or continuous absorbable monofilament sutures. The hard palate is closed as previously described. The end of the soft palate is then reflected rostrally and the nasopharynx flushed with saline and aspirated to remove clots. Temporary carotid artery occlusion Temporary carotid artery occusion can be performed without problems in dogs due to collateral brain perfusion coming from the vertebral arteries. In cats, despite ongoing debate regarding the risks associated with carotid artery occlusion, the serious risks of brain hypoperfusion make this procedure, even with only temporary occlusion, contraindicated (Holmberg 1996).

Temporary carotid artery occlusion in dogs is performed with the dog positioned as for ventral rhinotomy and the neck positioned over a pad; the skin is incised on the midline from the larynx to the midtrachea. The paired sternohyoideus muscles are separated and retracted to expose the trachea. The external carotid artery is then palpated dorsolaterally to the trachea and exteriorized after blunt dissection of its sheath. After separation from the other neurovascular structures, it is occluded with a bulldog clamp (Figure 8.9), umbilical tape, or a vascular tie. The procedure is then repeated on the other side (Hedlund 1998). The skin incision is temporarily sutured in a continuous pattern or with staples, and the rhinotomy is continued. At the end of the procedure, using new surgical gloves and instruments, the skin suture and vascular clamps are removed. The carotid arteries are repositioned, the surgical field lavaged with sterile saline solution prior to closure of the sternohyoideus muscles. The wound is closed in a routine manner. An alternative technique to achieve temporary carotid artery ligation is to place a Rumel tourniquet around each common carotid artery and to leave the ends of the tourniquets exiting the incision when it is closed (Figure 8.10). Once the rhinotomy procedure is completed, the tourniquets can be removed without reopening the neck incision. This shortens surgery time and avoids the necessity to reprepare the cervical site for reexploration and removal of the vascular clamps after completion of the rhinotomy procedure. In most cases, release of the vessel occlusion at the completion of the procedure does not initiate profuse nasal bleeding.

Figure 8.8.  Ventral rhinotomy in a cat. The procedure is similar to that described in the dog. The bone flap can be created either with an osteotome or an oscillating saw.

Figure 8.9.  Temporary carotid artery occlusion in a dog. After a blunt dissection of the cervical muscles, the carotid arteries are exteriorized and clamped with bulldog vascular forceps. The skin is temporarily sutured over the clamps and a rhinotomy performed. At the end of the nasal surgery the skin over the vessels is reopened and the clamps removed.

Figure 8.10.  Intraoperative view con the Rumel tourniquet placed around the carotid artery after the isolation from adjacent structures. The muscular planes, the subcutis and the skin are sutured in place, leaving the mobile part of the tourniquet protruding out of the suture. At the end of the rhinotomy the tourniquet is released and the carotid artery freed again without opening the wound.

279

280  Veterinary Surgical Oncology

Aftercare The oro- and nasopharynx are cleaned of fluids and blood clots before extubation. The animal’s head is kept slightly down to avoid aspiration. Sneezing and bleeding are expected after surgery and can last for several days. Serous discharge can persist for days to weeks after surgery. Rhinotomy is a painful procedure, and analgesia must be provided for 3–5 days. An antibiotic (e.g., cefazolin) is administered intravenously at induction of anesthesia, but it is necessary in the postoperative period only if an avascular bone flap is repositioned or if the tumor was associated with infection. The nares should be kept clear from blood clots and secretions, especially in cats. Appetite is usually restored in a few days in dogs after both procedures, and a feeding tube is rarely required. In cats, appetite stimulants such as diazepam, oxazepram, and mirtazapine may be indicated in some cases. If the ventral approach is performed, the animal is offered only soft food for the first 10 days, and gradually switched to canned food that will be offered for the next 4–6 months. Chewing hard objects is not recommended for the same period of time. Cosmetic and functional outcome Functional outcome is good with both procedures. The elimination of the dorsal bone flap does not preclude good respiratory function once the soft tissues have healed. Serous nasal discharge and occasional sneezing can be present for an extended period of time after rhinotomy, and it should not be a concern for either owners or the veterinarian. The cosmetic appearance is superior with the ventral approach. Potential complications Hemorrhage can be copious with both approaches and a blood transfusion may be needed in some cases. Temporary carotid artery occlusion can limit this complication in dogs. Packing the nasal cavity with sponges for several days is another option to avoid hemorrhage, however it is not well tolerated. Subcutaneous emphysema with movement of the skin over the suture line, in association with respiration, usually occurs with the dorsal rhinotomy approach when the bone flap is removed, unless a small rhinostomy is left or a rhinostomy tube is placed. This is, however, usually self-limiting and resolves in 1–2 weeks. Airway obstruction by blood clots is another potential complication that can be limited by abundant flushing

of the nasal cavity before wound closure and by keeping the nares clean afterward. If the cribriform plate is eroded by the tumor or if the biopsy or the curettage procedure breaches the cribiform plate, a brain lesion may occur. With the ventral approach, an oronasal fistula can develop if hard food is offered prematurely or can be secondary to self-trauma. While anorexia is not frequent, it is more commonly seen in cats. Tumor recurrence is always expected since wide excisional margins cannot be achieved. Common nasal tumors Nasal tumors are rare in both dogs and cats, representing about 1%–2% of all tumors (Rassinick et al. 2006; Turek and Lana 2007). Eighty percent are malignant, and 60%–75% are of epithelial origin in dogs (Rassinick et al. 2006), with carcinoma and adenocarcinoma being the most common histotypes (Turek and Lana 2007). Squamous cell carcinoma and undifferentiated carcinoma are the other types frequently encountered, together with chondrosarcoma, osteosarcoma, and other mesenchymal tumors (Turek and Lana 2007). Nasal lymphoma is one of the more frequent nasal tumors in cats (Turek and Lana 2007). Animals affected with nasal tumors are generally older (>10 years), and there appears to be a predilection for male cats (Ogilvie and LaRue 1992). Dolichocephalic and mesocephalic dogs seem to be more often affected (Bukowski et al. 1998). Regardless of histotype, distant metastases are not frequent at diagnosis and occur late in the disease, developing in about 40%–50% of animals at the time of death (Turek and Lana 2007). Therapy is therefore directed principally against the primary lesion. Untreated dogs with nasal carcinomas have a life expectancy of about 3 months (Rassnick et al. 2006). Rassnick reported that the only factor that seemed to have a negative prognostic value was the presence of epistaxis at presentation (median survival 88 days vs. 224 days for dogs without epistaxis). Medical treatment with steroids or COX-2 inhibitors did not improve prognosis, even when, at least in epithelial tumors, COX-2 expression has been detected (Kleiter et al. 2004). The standard of care for nasal tumors is megavoltage radiation therapy. Surgical debulking is indicated preoperatively if an orthovoltage machine is used, since its penetration ability is inferior to that of megavoltage radiation. Median survival of dogs with nasal adenocarcinoma treated with megavoltage radiation is about 14–21 months (Adams et al. 2005; Henry et al. 1998), which is significantly longer than that of animals treated

Respiratory Tract and Thorax  281

with surgery alone (median 126 days). The response to radiotherapy of mesenchymal tumors, such as chondrosarcoma, is not as good as for epithelial tumors, as reported by Popovitch et al. (1994); however, surgery can achieve good palliation of symptoms (median 270 days). Even though surgery alone is not considered curative in nasal tumors, its adjuvant use combined with megavoltage radiotherapy has been shown to improve the quality of life in dogs affected with nasal neoplasia (Adams et al. 2005). In cats with nasal lymphoma, radiation therapy alone (Straw et al. 1986) or combined with chemotherapy (Haney et al. 2009; Sfiligoi et al. 2009) can achieve longterm survival and is better tolerated than surgery. Adjuvant treatments Surgery itself can be used in an adjuvant setting for nasal tumors, especially in dogs. Chemotherapy is considered after rhinotomy, if radiation is not available, to treat both epithelial and mesenchymal nasal tumors in dogs. A combination of intravenous doxorubicin and carboplatin, together with oral piroxicam, was associated with a long survival time in one report (Langova et al. 2004). Mitoxantrone and carboplatin have also been used in dogs, but without promising results. Chemotherapy as a radiosensitizer has been recently proposed (Nadeau et al. 2004). Chemotherapy can be used to treat nasal lymphoma in both dogs and cats, but the overall results are not as good as with radiation alone. Surgery is not indicated with nasal lymphoma.

Laryngeal Tumors

Figure 8.11.  Endoscopic image of an arytenoid chondrosarcoma in a 9-year-old Doberman that was treated with arytenoidectomy. (Image courtesy of Dr. Richard White)

Figure 8.12.  Oral examination of a dog with a laryngeal rhabdomyosarcoma. (Image courtesy of Dr. Richard White)

Biopsy principles If there is a suspicion of a laryngeal mass, general anesthesia for examination by direct laryngeal examination and laryngoscopy is usually the next diagnostic step (Figures 8.11, 8.12). Patients often present for dyspnea, and it may not be possible to pass an endotracheal tube past a laryngeal mass. The patient should be prepared for a tracheostomy, and instruments to perform a tracheostomy should be on hand. Tracheostomy may need to be performed to maintain general anesthesia for examination and biopsy or may be required after biopsy due to swelling of the laryngeal mucosa. Depending on the amount of time required for examination, it also may be possible to perform a direct laryngeal examination under short-acting injectable anesthetic drugs. An incisional biopsy of the mass should be performed for histopathology. This is performed orally if the mass

is accessible. A small wedge of tissue is excised. Closure of the biopsy site should be attempted, but may not be possible. Hemostasis can be assisted with pressure on the biopsy site. Cytology alone has been shown to be inaccurate in the diagnosis of rhabdomyosarcoma in dogs and can cause further swelling (O’Hara et al. 2001; Henderson et al. 1991; Jakubiak et al. 2005). It may be useful in cats to distinguish between lymphoma and squamous cell carcinoma. Ideally, a more definitive diagnosis should be achieved prior to initiating therapy. Incisional biopsy was found to be a reliable method for the definitive diagnosis of laryngeal masses in cats. In some cases, however, a diagnosis of lymphoid hyperplasia was found to be inaccurate on subsequent biopsies. The diagnosis of lymphoid hyperplasia is relatively rare, and a second biopsy may need to be considered if the diagnosis does

282  Veterinary Surgical Oncology

Figure 8.13.  Lateral radiograph of a dog with a laryngeal rhabdomyosarcoma. (Image courtesy of Dr. Paolo Buracco)

not correlate with the rest of the clinical picture (Jakubiak et al. 2005). If a tracheostomy is performed, it should be done with instruments different from those used for an incisional biopsy to prevent seeding of tumor cells at the tracheostomy site. Cytology or biopsy of the submandibular lymph nodes should be performed to evaluate for metastatic disease. It is currently unknown whether surgical removal and evaluation of the lymph nodes in deeper locations (i.e., retropharyngeal and parotid lymph nodes) is clinically important. One noninvasive method of evaluating the deeper lymph nodes of the head and neck is by advanced imaging. Imaging tests Radiographs of the larynx are helpful to localize disease. Common findings with laryngeal tumors include soft tissue opacity in the lumen and decreased margination of laryngeal structures (Figure 8.13) (Jakubiak et al. 2005). In one study, only 10% of cases with laryngeal neoplasia had no radiographic lesions. Ultrasound has also been reported as a method of diagnosing laryngeal masses (Rudorf et al. 1997, 2002). The advantage of this technique is that it allows examination in awake patients in sternal recumbency, which may be more feasible in patients with respiratory compromise. Ultrasoundguided fine-needle aspiration (FNA) has also been reported with this technique in cats. Three-view thoracic radiographs should be performed to evaluate for pulmonary metastasis and aspiration pneumonia. For cases where a surgical resection is being considered, a CT scan is necessary for surgical planning. The CT scan will help to determine the extent of disease and the feasibility of a curative-intent surgery (Figures 8.14, 8.15). A mass that is amenable to a partial or complete laryngectomy is one relatively confined within the larynx and that has not invaded the pharynx

Figure 8.14.  CT examination of a dog with a laryngeal rhabdomyosarcoma. (Image courtesy of Dr. Paolo Buracco)

or esophagus. The CT scan will also help to determine the origin of the mass (within the larynx or a mass invading from outside the larynx). The lymph nodes and thorax can also be assessed for evidence of metastasis by CT scan. Ideally, a CT scan would be performed prior to biopsy if they are done under the same general anesthetic. This is because the edema, inflammation, and hemorrhage caused by the biopsy may distort the CT image. A tracheostomy tube may be required to maintain general anesthesia if an endotracheal tube cannot be placed orally. Surgical techniques Vocal cordectomy laryngectomy Vocal cordectomy laryngectomy is limited to small benign tumors of the vocal folds. It is similar to vocal cordectomy performed for devocalization in dogs. Similar to devocalization in dogs, a transoral approach is not recommended due to the increased risk of the formation of granulation tissue and webbing of mucosa in this region that may lead to upper airway obstruction. The patient is positioned in dorsal recumbency with the head extended. A ventral midline approach is made from the basihyoid to the third tracheal ring. The ventral midline incision continues through the cricothyroid ligament and thyroid cartilage. Self-retaining retractors are used to retract the thyroid cartilage. The affected vocal cord is resected. The mucosa is sutured in a simple continuous pattern with a small-gauge, monofilament, absorbable suture material. Care must be taken to ensure accurate closure of the mucosa to prevent the development of granulation tissue and scarring of the larynx. The cricothyroid ligament and thyroid cartilage are closed using simple interrupted sutures. The site is

Respiratory Tract and Thorax  283

Figure 8.15.  CT reconstruction of a dog with laryngeal chondrosarcoma. Same dog as in Figures 8.1, 8.10. (Image courtesy of Dr. Charles Kuntz)

(a)

(b)

(c)

Figure 8.16.  Ten-year-old malamute with a grade II mast cell tumor of the arytenoid. (A) Tracheostomy and ventral midline approach to the larynx. (B) Ventral midline approach to the larynx. (C) Hemilaryngectomy. (Images courtesy of Dr. Bart Van Goethem)

lavaged with sterile saline. The subcutaneous tissue and skin are closed routinely. Hemilaryngectomy The approach to the larynx is the same as for cordectomy (Figure 8.16). The tumor is excised full thickness through the mucosa and thyroid or cricoid cartilages. The soft tissue attachments to the segment of larynx are excised along the larynx with blunt or sharp dissection. The resultant defect is closed primarily, if possible. If feasible, the mucosa is closed separately with simple continuous or simple interrupted sutures. The remaining laryngeal cartilage is closed using simple interrupted sutures. If it is not possible to close the site primarily, there are several techniques to close the defect. A myocutaneous flap can be elevated based on the sternohyoid muscle (Nelson 2003a). Preplanning is necessary if this flap is going to be used because the muscle and skin are not separated from one another during the approach. An island flap of the appropriate size is planned. The medial edge of the flap is the ventral midline incision. The rest of the island flap is harvested. The vessels supplying the flap are branches of the cranial thyroid

artery. The flap is depilated by shaving the epidermis down to dermis to prevent hair growth. The flap is rotated so the dermis is facing into the airway. The dermis is sutured to the mucosa in a simple interrupted suture pattern. The site is lavaged. The rest of the closure is routine. Free tissue transfer has also been reported to replace laryngeal defects but are not routinely used (Nelson 2003a). Temporary tracheostomy should be placed in these patients to prevent obstruction due to postoperative laryngeal edema and inflammation. A gastric feeding tube should be considered in these patients as they may not eat normally after surgery. Epiglottectomy Epiglottectomy has also been reported anecdotally. There are currently no peer-reviewed case reports or papers that evaluate this technique and outcome in dogs. Epiglottectomy can be performed via a transoral approach (Figure 8.17). The patient is placed in sternal recumbency and the head is suspended with the mouth open. The epiglottis is grasped and resected. If possible the, mucosa at the base of the epiglottis is closed using a simple interrupted pattern of monofilament,

284  Veterinary Surgical Oncology

(a)

(b)

(c)

Figure 8.17.  Images from a dog with fibrosarcoma of the epiglottis. (A) Oral approach for epiglottectomy/partial laryngectomy. (B) Closure of the epiglottectomy site. (C) Specimen photo. (Images courtesy of Dr. Laurent Findji)

absorbable suture material. Alternately, a laser can be used for resection of the epiglottis. Total laryngectomy The patient is positioned in dorsal recumbency with the head extended. A ventral midline approach is made that extends several centimeters cranial and caudal to the larynx. The paired sternohyoid muscles are bluntly dissected at midline, they are elevated from the basihyoid bone, and this attachment is incised bilaterally. The muscles are retracted laterally. A patent airway is maintained by either a preplaced tracheostomy that is attached to the anesthetic machine during surgery or by placing a sterile endotracheal tube through the transected end of the trachea at the time of surgery. If a tube is placed intraoperatively, the fascial attachments are bluntly dissected from the cranial trachea at the level of the first tracheal ring. The orally placed endotracheal tube is retracted. The trachea is transected at the level of the first tracheal ring. A sterile endotracheal tube is placed in the trachea and the orally placed tube is removed. If a tracheostomy tube has been placed, the trachea is resected at the same level. The larynx is removed en bloc by transection of all of its attachments to the skull by the hyoid apparatus and to the pharynx, tongue, hyoid bone, and sternum by the extrinsic muscles of the larynx (Figure 8.18). The hyoid apparatus is disarticulated bilaterally at the level of the thyrohyoid bone and the basihyoid and ceratohyoid bones. The thyropharyngeus, cricopharyngeus, sternothyroideus, and thyrohyoideus muscles are transected at their attachments to the larynx. Care must be taken to avoid damage to the nerves in this area, which provide

Figure 8.18.  Total laryngectomy specimen from a Sheltie with a laryngeal squamous cell carcinoma. (Image courtesy of Dr. Ralph Henderson)

sensory and motor function to the pharynx and esophagus and are necessary for normal swallowing. The larynx is retracted from caudal to cranial. The remaining soft tissue attachments are sharply and bluntly dissected from the larynx. The pharyngeal mucosa is then transected cranial to the larynx and the larynx is removed en bloc (Figures 8.19, 8.20). The area is lavaged with sterile saline. The pharyngeal mucosa is closed with a simple continuous inverting suture pattern. The paired thyropharyngeal and cricopharyngeal muscles are closed ventral to the pharyngeal mucosa. A permanent tracheostomy must be created. This will involve either converting the previously placed temporary tracheostomy into a permanent one or creating a tracheostoma with the severed end of the trachea (Figure

Respiratory Tract and Thorax  285

Figure 8.19.  Laryngeal chondrosarcoma in a 10-year-old dog. The trachea distal to the mass is being palpated with fingers. (Image courtesy of Dr. Charles Kuntz)

Figure 8.20.  Same dog as in Figure 8.19. En bloc resection of the larynx including the hyoid apparatus.

brane. The dorsal tracheal membrane is rotated ventrally and closed to the proximal end of the trachea. With either technique, an airtight seal should be created in the proximal trachea. The existing stoma site is enlarged and is oval or rectangular in shape. The sternohyoideus muscles are sutured dorsal to the trachea to hold it ventrally. Excessive skin folds are excised to prevent interference with the tracheostomy. The subcutaneous tissue is sutured to the tracheal wall around the stoma. The skin is sutured to the tracheal mucosa using a simple interrupted suture pattern. The remaining skin and subcutaneous tissue is closed routinely. To create a tracheostoma, the proximal trachea is directed ventrally between the sternohyoideus muscles. The sternohyoideus muscles are sutured to the trachea dorsally to hold the trachea in a ventral position. The tracheal orifice is trimmed to fit the contour of the exit site. The excess skin folds and a circular portion of skin are removed. The subcutaneous tissue is sutured to the tracheal wall. The tracheal mucosa is sutured to the skin in a simple interrupted suture pattern. The advantages of the creation of the tracheostoma are that the opening is wider and has less risk of stricture as it heals. Because of the larger stoma, this will facilitate endotracheal intubation if required in the future. A stomach tube should be placed postoperatively until the patient can eat and swallow normally. In some cases, the stomach tube feeding may need to be permanent. For information about surgical techniques, consult Fossum et al. 2007a, Nelson 2003a, Block et al. 1995, Dyce et al. 1987a, Crowe et al. 1986. Aftercare

Figure 8.21.  Photograph of a Sheltie 7 days after surgery with a laryngeal squamous cell carcinoma treated with total laryngectomy and permanent tracheostomy (as depicted in Figure 8.18).

8.21). To convert a temporary tracheostomy to a permanent one, the severed proximal end of the trachea is flattened and closed using simple interrupted sutures. Alternately, the first two remaining tracheal rings are excised without removing the dorsal tracheal mem-

After cordectomy, patients will need to be monitored in an intensive care unit for dyspnea. There is a risk of laryngeal edema due to cordectomy, and patients should be treated with dexamethasone perioperatively at an anti-inflammatory dose. After hemilaryngectomy patients will need to monitored in an intensive care unit for care of their temporary tracheostomy tubes and for pain management. The tracheostomy tubes can be removed 3–4 days after surgery. This can be done by either covering the end of the tube, if there is room for the air to flow by, or by removing the tube, and assessing respiratory function. This should be done in a quiet, controlled environment, and the clinician should be prepared to sedate or anesthetize the patient to replace the tube quickly if needed. Potential complications of cordectomy and hemilaryngectomy include scar formation and stenosis of the upper airway, incomplete margins of tumor excision, and aspiration of either blood in the

286  Veterinary Surgical Oncology

immediate postoperative period or food due to laryngeal dysfunction. After total laryngectomy and permanent tracheostomy, patients will require intensive care and monitoring. The tracheostomy site will need to be kept clean. The patient is fed via the gastric feeding tube until the pharyngeal mucosa has healed. The first oral feeding should be done under supervision to ensure normal swallowing function. Pain control should be provided for 5–7 days. For long-term management of these patients, the owners will need to keep the stoma site clean and trim the hair around it. No neck leads can be used, and the patients cannot swim. After this procedure, dogs will be unable to pant. Unrestricted activity should be avoided. As well, dogs will not be able to thermoregulate by panting and they should not be outdoors in hot weather (Henderson et al. 1991). Reported complications include dehiscence of the pharyngeal mucosa and iatrogenic hypoparathyroidism (Henderson et al. 1991). It is also possible that these patients will not be able to swallow normally, and the owners should be prepared for a permanent gastrostomy tube. Cosmetic and functional outcome Successful cordectomies and laryngectomies have been reported for the treatment of benign and malignant neoplasia of the larynx (Meuten et al. 1985; Block et al. 1995; Henderson et al. 1991). Descriptions of hemilaryngectomy are largely based on human reports. This is likely because the surgical cases are going to be either benign diseases such as rhabdomyoma, that is amenable to local resection, or malignant neoplasia that is not confined to a small area of the larynx. An article that discusses two cases of laryngeal mast cell tumors in dogs reports that in two cases a partial laryngectomy was attempted that resulted in recurrence in both dogs (Crowe et al.1986). This is because it is difficult to achieve adequate surgical margins for a malignant neoplasm of the larynx with a partial resection. In general, by the time that they are diagnosed, most malignant tumors of the larynx will require total laryngectomy for successful local treatment of the tumor. The entire larynx has been reported to be successfully removed with a good long-term outcome (Block et al.1995). However, this procedure is uncommon in veterinary medicine, and most of the information on this procedure is anecdotal. Most common tumors—Prognosis and decision making In cats, lymphoma and squamous cell carcinoma are the most commonly reported tumors (Jakubiak et al. 2005;

Carlisle et al. 1991; Saik et al.1986). In dogs, there are many different tumor types reported, with rhabdomyoma and rhabdomyosarcoma being reported relatively commonly (Block et al. 1995; Carlisle et al. 1991; Meuten et al. 1985; O’Hara et al. 2001; Henderson et al. 1991; Clercx et al. 1998). Other reported laryngeal tumors in dogs include carcinoma, squamous cell carcinoma, mast cell tumor, osteosarcoma, melanoma, lipoma, adenocarcinoma, chondrosarcoma, leiomyoma, fibropapilloma, fibrosarcoma, myxochondroma, invasive thyroid carcinoma, granular cell tumor, and plasmacytoma (Carlisle et al. 1991; Saik et al.1986; Rossi et al. 2007; Hayes et al. 2007). Oncocytoma is another tumor type that has been reported. It is thought to be likely that the tumors that have been diagnosed previously as oncocytomas are in fact rhabdomyomas (Meuten et al. 1985). Oncocytomas, rhabdomyomas, and granular cell tumors can all have similar appearances under standard light microscopy, and immunohistochemistry may be necessary for a definitive diagnosis. Adjunctive therapy Laryngeal lymphoma in cats is a nonsurgical disease that should be managed with chemotherapy and/or radiation. Chemotherapy should be considered in cases of malignant tumors with a high chance of systemic spread. Radiation was reported in a case of a solitary plasma cell tumor of the larynx in a dog with no effect. This was a surprising finding. However, the dog went into a complete and long-standing remission when treated with melphalan and prednisone. Radiation is not commonly used for tumors of the larynx but should be considered in cases of incomplete resection or when a surgical resection is not thought possible. With the advent of stereotactic radiosurgery, local control of malignant tumors in this area may become possible. One example for which radiation therapy would be useful is an invasive thyroid carcinoma. In cases of laryngeal tumors where a surgical resection is not thought to be possible, a permanent tracheostomy can be placed to relieve upper airway obstruction and the tumor site can be treated with radiation. Tracheostomy alone can also be considered as a palliative measure to relieve the upper airway obstruction.

Thoracotomy Thoracotomy is performed to explore the thoracic cavity and to take surgical biopsies in situations where other diagnostic tools have failed, or to excise either primary or metastatic intrathoracic tumors. Since the preoperative diagnostic rate is low for intrathoracic neoplasia (Tattersall and Welsh 2006), thoracotomy is almost

Respiratory Tract and Thorax  287

(b)

(a)

(c)

Figure 8.22.  Chest radiographs. (A) Ventrodorsal view of the same mass, which appears to be pulmonary (arrow). (B) Laterolateral view of the thorax of a dog with an intrathoracic mass. This view alone does not allow the determination of whether the lesion is pulmonary or mediastinal. (C) Feline thymoma. In this case the ventrodorsal or dorsoventral view is needed to correctly locate the lesion.

always performed with both diagnostic and curative intent. The most commonly employed techniques are lateral thoracotomy and median sternotomy; both procedures require assisted ventilation. For biopsy purposes, a thoracoscopy is usually preferred because it is less invasive (Kovak et al. 2002). For the lungs, pericardium, pleura, and for mediastinal masses, ectopic thyroid tumors, and other spaceoccupying masses, biopsies can be performed by tho­ racoscopy. This technique, however, does not allow the excision of large intrathoracic masses, whereas small lung tumors located away from the hilus, especially in the left caudal lung lobe, can be excised with this technique (Lansdowne et al. 2005). Imaging To choose the best approach to access the thorax (intercostal thoracotomy versus median sternotomy) and to decide which side and intercostal space to use, left and right lateral and dorsoventral or ventrodorsal radio-

graphs of the thorax must be performed (Figure 8.22). The use of contrast-enhanced CT allows better visualization of small intrathoracic masses that may not be visible with radiography (Figure 8.23A); however, MRI is not usually employed because it requires more complicated respiratory-gated techniques to image the chest. Surgical techniques Intercostal thoracotomy Generally, a limited area of one side of the thorax is explored if a solitary lesion has been previously identified. Usually one-third of one thoracic cavity and the corresponding mediastinal area are fully visualized with this approach. The intercostal space (3rd to 10th) is chosen according to radiographs or CT scans taken before surgery (see Table 8.1). If a lung tumor is to be resected, the 4th or 5th intercostal space should always be chosen, regardless of the lobe affected, because the

288  Veterinary Surgical Oncology

Figure 8.23.  CT images of a canine thorax. (A) Contrast-enhanced CT is a more accurate method compared to radiography for the detection of small lung metastases (arrows). (B) The use of CT allows evaluation of the full extent of thoracic wall masses that tend to grow in a centripetal direction. In this picture, a rib metastasis of a previously resected contralateral humeral osteosarcoma is evident.

Table 8.1.  Access to the main thoracic structures for tumor excision. Organ Cardiac structures Lung lobes Cranial esophagus Caudal esophagus

Intercostal Space 4–5 4–5 3–4 7–8

Side Both Right or left Left Both

Note:  Thymomas are usually excised by median sternotomy.

lung lobe hilus is accessible only by this approach (Kuntz 1998). The animal is positioned in lateral recumbency and an incision in the skin, subcutaneous tissue, and cutaneous trunci muscle is made with a scalpel, parallel to the desired intercostal space. The incision is extended from the costovertebral junction, past the costochondral arch to the sternum. The latissimus dorsi and pectoralis muscles are incised, or alternatively, can be retracted in small dogs and cats. At this point it is easy to count the ribs by passing a finger cranially underneath the latissimus dorsi muscle to identify the first rib. The fifth rib is easily identified by the caudal insertion of the scalenus and the cranial insertion of the external abdominal oblique muscles. One of these two muscles is incised, depending on the intercostal space chosen, and the incision is continued by separating the bellies of the serratus ventralis muscle; the intercostal muscles are then incised in the middle, avoiding the nerves and

vessels that run parallel to the caudal aspect of each rib, and the parietal pleura is exposed (Figure 8.24A). The thorax is entered by bluntly incising the pleura with scissors and continuing dorsally and ventrally, paying attention to the internal thoracic arteries and veins that run lateral to the inner side of the sternum (Figure 8.24B). A Finochietto retractor is applied to spread the ribs, with moistened sponges placed between the retractor and the ribs (Figure 8.24C). The rib cranial to the incision is usually easier to retract than the caudal one. At this point, the thorax may be inspected. Following lung lobectomy for a lung mass, the remaining lung in the exposed hemithorax is carefully palpated to detect the presence of other neoplastic nodules, and the hilar and sternal lymph nodes are evaluated and resected if needed. Prior to thoracotomy closure, blood clots are carefully removed, and a thoracostomy tube is inserted by tunneling it in the subcutis for at least three ribs in a caudocranial direction, before its entrance in the thorax. The tube must not enter the thorax through the thoracotomy site (Figure 8.25A). The thoracostomy tube is left open until airtight closure is accomplished, to avoid a tension pneumothorax. Closure of the thoracotomy is achieved by preplacing four to eight sutures around the rib immediately cranial and caudal to the incision, using absorbable or nonabsorbable heavy-gauge monofilament (2-0 to 0 USP, polydioxanone or polypropylene in small to medium size dogs and cats; 0 to 2 USP in large dogs), and having an assistant approximating them while tying (Figure 8.24D). To avoid damaging the underlying lungs and to decrease the risk of causing

Respiratory Tract and Thorax  289

(a)

(c)

(b)

(d)

Figure 8.24.  Intercostal thoracotomy. (A) After the thorax is entered, blunt incision is made with scissors in the pleura and continued dorsally and ventrally. (B) An incision in the skin, subcutaneous tissue, and cutaneous trunci muscle, extending from the costovertebral junction to the sternum, is made with a scalpel, parallel to the desired intercostal space. (C) A Finochietto retractor is applied to spread the ribs, protecting them with a moistened gauze sponge. (D) After the completion of the thoracic surgery, the chest wall is closed by preplacing four to eight large-gauge sutures around the rib immediately cranial and caudal to the incision and having an assistant approximating them while tying. (Image 8.24D courtesy Dr. R. Bussadori)

trauma to the intercostal vessels and nerves, the needle is inserted through the tissues with its blunt end first. The suture can be passed through small holes drilled in the rib itself, instead of surrounding it, to reduce postoperative pain and to avoid damaging the nerve and vessels on the caudal aspect of the rib caudal the incision (Rooney et al. 2004). The serratus ventralis and scalenus or external abdominal oblique muscles are sutured in a simple continuous pattern; the latissimus dorsi, cutaneous trunci, and subcutis are closed individually in simple

continuous suture patterns, and the skin is closed routinely. At this point the pleural space is evacuated and the thoracostomy tube is closed. Before completing the closure, a selective intercostal nerve block should be performed with 0.75% bupivacaine injected dorsally in the intercostal spaces one or two ribs cranial and caudal to the incision, or directly injecting the local anesthetic in the thoracostomy tube (2 mg/kg body weight [0.9 mg/ lb]) while the animal is still anesthetized (given that bupivacaine is painful) and keeping the animal laterally recumbent with operated side down for 20 minutes, to

290  Veterinary Surgical Oncology

(a)

(b)

(c)

(d)

(e)

Figure 8.25.  (A) The thoracic drain is applied making a stab incision two to three ribs caudal to the thoracotomy and tunneling the tube cranially in the subcutis. (B) The chest tube is secured to the skin with a Chinese finger suture. (C) A Christmas tree adaptor is applied to close it, using a metal wire to achieve an air-thigh. (D) Chest radiographs after the application of the drain in order to determine the tube’s correct or incorrect (E) placement.

allow the diffusion of the local anesthetic over the incision site. The area of the thorax that can be evaluated with intercostal thoracotomy can be extended using the rib pivot technique described by Schulman and Lippincott et al. (1988). In this technique, a rib cranial to the incision is rotated (pivoted) out of the surgical field. After the muscular incision, a transverse osteotomy is performed at the level of the costochondral junction of the cranial rib with an oscillating saw. The rib is grasped and rotated cranially, pivoting on the costovertebral

junction. At the end of the surgical procedure, the rib is pivoted back into its standard position, and a smallgauge orthopedic wire is used to stabilize the osteotomy site after having drilled two small holes just dorsal and ventral to the osteotomy. The remainder of the thoracic incision is closed as previously described. Another variation used to extend the thoracotomy exposure, which is rarely performed in small animals, is rib resection thoracotomy. The advantage of this procedure is that, compared to the intercostal technique, it results in fewer adhesions between the lungs and the

Respiratory Tract and Thorax  291

incision and provides better exposure of the thorax. A secure closure is more difficult to achieve however, especially in large breed dogs, because the suture pullout strength is weaker compared to what can be achieved by anchoring the sutures to the ribs. After the muscular layers are incised, as for the intercostal thoracotomy, the periosteum of the rib to be resected is incised and bluntly elevated from its lateral and medial surface. The rib is then excised with a bone cutter and removed. The parietal pleura is bluntly opened and the incision extended as previously described. Closure is accomplished by preplacing interrupted mattress sutures in the medial and lateral periosteal edges and tying them. The wound closure is completed as described above. Median sternotomy The median sternotomy is the only approach that allows complete access to both sides of the thoracic cavity. It is therefore indicated when a complete thoracic exploratory is necessary, as in the cases of multiple metastasis

(a)

(c)

excision, thymoma removal, etc. The only structures that cannot be easily reached with this approach are the great vessels, the bronchial bifurcation, and the thoracic duct. As demonstrated in different studies (Ringwald and Birchard 1989; Williams and White 1993; Burton and White 1996), median sternotomy is not associated with a greater number of complications than the intercostal approach, nor is it a more painful technique. It should not, therefore, be considered an inferior approach. The animal is placed in dorsal recumbency, with the front legs extended and secured cranially (Orton 1995b). A skin incision is performed over the midline of the sternum, from the level of the manubrium to the xyphoid process. The incision is deepened through the subcutis and pectoral musculature until the sternebrae are exposed. The sternum is then cut exactly on its midline with an oscillating saw, sternal splitter, or an osteotome (or scissors can be used in small-sized young animals) (Figure 8.26), paying careful attention to avoid

(b)

(d)

Figure 8.26.  Median sternotomy. (A) The skin and subcutaneous tissues are incised and separated until the sternum is reached. (B) The sternebrae are incised in their midline using an osteotome or an oscillating saw (C). (D) A Finochietto retractor is applied, and the thoracic surgery performed. (Image 8.26B courtesy of Dr. R. Bussadori)

292  Veterinary Surgical Oncology

Figure 8.27.  (A) The closure of the sternum is achieved by preplacing some figure-eight wire sutures and tying them (B). Soft tissues are then sutured routinely. (Image 8.27A courtesy of Dr. R. Bussadori)

damaging the underlying structures, such as the internal thoracic vessels. Staying as close as possible to the midline also allows improved wound healing. Depending on the part of the thorax to be exposed, every effort should be made to leave either the manubrium or the xyphoid process intact to achieve stable closure of the sternum and avoid dehiscence. A Finochietto retractor is applied to gently spread the incision (Figure 8.26D), and the desired surgical procedure is performed. Before closure, a thoracostomy tube is inserted by tunneling it under the subcutis, lateral to the midline, from three intercostal spaces caudal to its proposed entrance point in the chest. The best way to obtain a stable closure of the sternum is to place figure eight orthopedic wires around each sternebra, incorporating the costosternal junction (Figure 8.27) (Davis et al. 2006). In small dogs and cats, heavy-gauge monofilament sutures can be used as an alternative to wire; however, they have been shown to result in delayed healing in large-breed dogs (Pelsue et al. 2002). The pectoral muscles, subcutis, and the skin are closed in separate simple continuous layers. Median sternotomy may be extended cranially with a ventral midline cervical approach or caudally with a ventral midline celiotomy, if access to these regions is required. Aftercare Thoracic surgery is a painful procedure, and postoperative pain may result in respiratory impairment. Therefore, a good analgesic regimen, initiated before recovery of anesthesia, is essential. Parenteral or epidural opioids, selective intercostal nerve blocks (only for intercostal thoracotomy), or intrapleural analgesia may be used.

Morphine, oxymorphone, fentanyl, butorphanol, or buprenorphine can be administered by a parenteral route. While the latter two have less detrimental effect on ventilation and are less likely to decrease the heart rate, they are not considered to be strong analgesics (in particular butorphanol) and may not provide adequate analgesia in some patients. Morphine can also be injected in the epidural space, with minimal cardiopulmonary effects and 6- to 12-hour duration of action. Intrapleural administration of bupivacaine through the chest tube in conscious animals is painful and should be preceded by intrapleural administration of lidocaine. Monitoring of ventilation after surgery is essential, since it may be depressed by a number of factors including anesthetic agents, pain, postoperative complications such as pneumothorax, hemorrhage and pulmonary edema, or tight bandages. Blood gas analysis should always be performed to assess the ventilation and acidbase status of the patient. Supplemental oxygen is generally recommended, guided by the results of blood gas analysis, especially if lung lobectomy or pneumonectomy has been performed. Opening the chest usually results in a decrease in body temperature, therefore the patient should be actively warmed by circulating water or air blankets or warm water bottles. After chest-tube placement it is recommended to assess its position by radiography (both lateral and dorsoventral views) and its patency by flushing with sterile warm saline solution. The thoracostomy tube is aspirated at least every hour for the first 4 hours after surgery, then every 2 to 4 hours until removal. The presence of air, fluid, or blood is

Respiratory Tract and Thorax  293

evaluated and recorded, and the tube can be removed as early as 2 hours after surgery if it is nonproductive or when the fluid production is between 2.2 mL/kg/day (expected amount to be induced by the presence of the tube itself) and less than 8 mL/kg/day (1–3.6 mL/lb/day) (Kuntz 1998, Fossum et al. 2007c). If the tube is nonproductive after 2 hours, however, the position and patency of the tube should be checked by thoracic radiographs and flushing, respectively. A thoracic bandage, which is applied loosely enough to avoid the restriction of breathing, helps reduce subcutaneous emphysema by sealing the thoracotomy incision. It may also make the animal more comfortable in the immediate postoperative period. Antibiotics (e.g. cefazolin) are administered in the perioperative period, but they are discontinued after 12 to 24 hours if infection was not present preoperatively (Fossum et al. 2007c). Complications and outcome Complications after both intercostal thoracotomy and median sternotomy are not frequent, and the incidence is similar for both procedures (Ringwald and Birchard 1989). Even if severe, they are rarely fatal if the animal is closely monitored in the immediate postoperative period and during follow-up. Suture dehiscence due to unstable closure, neurological deficits on the front leg, and sternal osteomyelitis can develop after median sternotomy. Subcutaneous emphysema due to a nonairtight closure, wound edema, and discharge are sometimes seen after intercostal thoracotomy. Hemorrhage, pain, and swelling may occur after either procedure. Because of the position of the animal during surgery, respiratory deficits are more likely following the intercostal procedure, whereas circulatory and cardiac problems are more likely during median sternotomy. To avoid inadvertent thoracostomy tube removal and consequent pneumothorax, the animal should not be left unattended. Depending on the type of intrathoracic surgery performed, pneumothorax, hemothorax, pulmonary edema, circulatory shock, or other specific complications may occur. The functional outcome of the surgical procedure itself is usually excellent. Oncologic outcome depends on the tumor removed. Thymomas in both dogs and cats have good prognosis if complete removal can be achieved; the prognosis for lung tumors is related to the extension of the disease, histologic grade, and the presence of metastatic lymph nodes at presentation (Zitz et al. 2008; McNiel et al. 1997). Heart base, esophageal, and metastatic tumors usually have a guarded prognosis.

Tracheal Tumors Biopsy principles If the patient is stable and time allows, tracheoscopy is an important step in the diagnosis and treatment of tracheal. Tracheoscopy allows visualization of the mass and the ability to evaluate the trachea for the presence of more than one mass. Often when clinical signs are apparent, the mass is too large for an endotracheal tube or the bronchoscope to pass caudal to the mass. Tracheoscopy is also an important step because it allows for biopsy of the mass to determine tissue type before definitive surgery. This can be achieved by both brush cytology and bronchoscopic biopsy of the mass for histopathology. Imaging Tracheal masses can often be diagnosed by plain radiographs (Figure 8.28). A mass within the tracheal lumen is well-visualized due to the contrast with the surrounding air. The lateral projection is usually the best view to evaluate for a tracheal mass because there are fewer overlying structures. The most common findings with tracheal masses are either soft tissue opacity within the trachea or stenosis of the trachea (Jakubiak et al. 2005; Carlisle et al. 1991). Masses are usually not mineralized except for osteochondromas (Carlisle et al. 1991). Three-view thoracic radiography would usually be performed as part of the workup for a dyspneic patient. They become an important tool for staging the patient once a tracheal mass has been identified to assess for pulmonary metastasis. CT is also a useful tool to more accurately evaluate the size, location, and extent of a

Figure 8.28.  Lateral radiograph of a dog with a tracheal chondrosarcoma. (Image courtesy of Dr. J. Liptak)

294  Veterinary Surgical Oncology

(a)

(b)

(c)

Figure 8.29.  (A, B) CT scan of a dog with a tracheal chondrosarcoma. (C) Reconstruction of CT images to allow sagittal view. (Images courtesy of J. Liptak)

tracheal mass for surgical planning (Figure 8.29). The CT is also a useful tool to evaluate for metastasis to lymph nodes and lungs. Surgical techniques Resection and anastomosis Resection and anastomosis is the treatment of choice for most tracheal tumors. One notable exception is lymphoma. The surgical approach depends on the location of the tracheal tumor. For the entire length of the trachea, there are several important principles to keep in mind. The trachea is surrounded by adventitia. Within this connective tissue are the lateral pedicles. These lateral pedicles contain the segmental blood supply to the trachea. The blood supply to the trachea is made up of branches of the cranial and caudal thyroid arteries and veins, bronchoesophageal arteries and veins, and the internal jugular veins (Hedlund 1987, 1991). The recurrent laryngeal nerves are also contained within the lateral pedicles and must be protected during resection. The adventitia must be bluntly dissected from the trachea to prepare for resection and anastomosis. This is performed by starting this dissection plane directly against the tracheal wall. Dissection must be kept to the area to be resected to preserve the blood supply unless there is excessive tension on the trachea and a need for mobilization. It is recommended to start with a conservative dissection of the surrounding fascia and increase it if needed to decrease tension on the anastomosis site. Tracheal resection requires thoughtful preservation of the airway at all times during the surgery and good communication with the anesthetist. There are two options for maintaining the airway during tracheal resection. One method is to retract the endotracheal tube cranially

so that it is not in the resection site. Stay sutures are placed in the proximal and distal segments of the trachea proximal and distal to the proposed resection site. The resection is carried out quickly, and the stay sutures are then used to pull the remaining segments of trachea together. The endotracheal tube is then advanced into the distal segment (Figure 8.30). The other technique involves transection at the distal site first and the placement of a sterile endotracheal tube, elbow, and extension into the distal trachea to maintain anesthesia. The orally placed endotracheal tube is retracted but not removed. The segment of trachea to be removed is excised. Stay sutures are used to approximate the two segments, the sterile endotracheal tube is removed, and the orally placed endotracheal tube is advanced across the anastomosis site. The technique used depends on the surgeon and anesthetist’s preference and also on the length of time the resection may take. In larger resections, the sterile placement of an endotracheal tube is recommended as this will allow for more time to approximate the two segments of the trachea and perform tension-relieving techniques if necessary. A sterile endotracheal tube should always be available at the time of resection to allow quick intubation of the distal segment, if needed. The inner surface of the trachea is a clean-contaminated environment. There is also potential for contamination of the surgical site with manipulation and/or exposure of an orally placed endotracheal tube. Prophylactic antibiotics should be used intraoperatively. After resection, the surgical site should be cultured and lavaged. The goal of the anastomosis is perfect apposition of the mucosa with minimal tension. The tracheal epithelium will heal quickly by epithelialization if the mucosal edges are in apposition. However, if there are gaps at the anastomosis site, it will heal with the development of

Respiratory Tract and Thorax  295

(a)

(b)

Figure 8.30.  (A) Intraoperative photograph of a cat with a tracheal squamous cell carcinoma that has undergone resection. The orally placed endotracheal tube is retracted slightly during resection. A sterile endotracheal tube is placed in the caudal segment of the trachea while the sutures are preplaced. (B) The endotracheal tube that was in the caudal segment is removed, and the endotracheal tube that was place orally is advanced into the caudal segment. The anastomosis is completed with the preplaced sutures. (Images courtesy of Dr. J. Liptak)

granulation tissue and stenosis (Hedlund 1991). The two methods that are commonly reported are the splitcartilage technique (Hedlund 1987, 1991; Fingland et al. 1995; Urschel 1996) and the annular ligament-cartilage technique (Hedlund 1987, 1991; Vasseur 1979; Demetriou et al. 2006). The split-cartilage technique involves planning the proximal and distal incisions in the trachea so that they bisect a tracheal ring proximally and distally. When the two segments are apposed, the two portions of the transected rings are aligned. The annular ligamentcartilage technique involves placing the incisions in the annular ligament between each ring. The two segments are apposed, and suture is placed around the adjacent cartilage rings. The split-cartilage technique is preferred because it leads to more accurate apposition with a higher chance of first-intention healing and a decreased risk of postoperative stenosis (Hedlund 1984; Fossum et al. 2007a). It is recommended that the stay sutures be placed in the proximal and distal tracheal segments prior to resection to prevent retraction of the segments (Figure 8.30). This is particularly important when working in the caudal cervical trachea or thoracic inlet. The stay sutures are also used to manipulate the proximal and distal segments. The optimal type of suture pattern for tracheal anastomosis is controversial. There are two recent reports in dogs that evaluate suture patterns in vivo and in vitro. Fingland et al. (1995) evaluated large-segment tracheal resection in vivo and compared simple continuous versus a simple interrupted suture patterns. Lateral tension sutures were placed in both groups. The authors

concluded that there was significantly less luminal stenosis and more precise histologic apposition using simple interrupted sutures (Fingland et al. 1995). An in vitro study by Demetriou et al. (2006) evaluated pullout strength in vitro in the canine trachea. For this technique, the annular ligament-cartilage resection technique was employed. Pullout strength was compared for a simple interrupted, simple continuous, and simple interrupted pattern with tension-relieving horizontal mattress suture. The authors concluded that the simple continuous and simple interrupted reinforced with horizontal mattress patterns were significantly stronger than the simple-interrupted suture pattern. In this study, the constructs failed at the annular ligament, so it is possible that resection technique played a role in mode of failure (Demetriou et al. 2006). Urschell et al. (1996) compared the pullout strength of a simple-interrupted versus a horizontal mattress pattern in vivo in rat tracheas. The resection technique was a split-cartilage technique. This author found no difference in the strength of these two suture patterns. Based on these studies, a simple interrupted suture pattern is recommended. This may need to be reinforced with tension-relieving sutures placed laterally and ventrally if there is tension on the anastomosis site (Figure 8.31). The sutures in the dorsal tracheal membrane are placed first. The ventral sutures are placed next, and the rest of the sutures are filled in to ensure that the trachea is precisely approximated. The sutures should encircle the remaining portion of the proximal and distal cartilages. The sutures should be placed so that the knots are extraluminal (Hedlund 1991).

296  Veterinary Surgical Oncology

(a)

(b)

(c)

Figure 8.31.  (A) Intraoperative photograph of a dog with a tracheal chondrosarcoma undergoing resection. (B) Specimen photograph of a tracheal chondrosarcoma removed from a dog. (C) Resection and anastomosis of the tracheal chondrosarcoma using simple interrupted sutures and reinforcement with mattress sutures. (Images courtesy of Dr. J. Liptak)

A monofilament, nonreactive, absorbable or nonabsorbable suture material should be used (Hedlund 1991). A recent study evaluating polyglactin, polydiaoxanone, and polypropylene in an experimental model for tracheal resection and anastomosis in sheep found no difference in suture type on the outcome of anastomosis (Behrend and Klempnauer 2001a). After resection of the tracheal segment with the tumor, the edges of resection should be inked to evaluate the margins of resection. Surgical approaches A ventral midline approach is used for the cervical trachea. The patient is placed in dorsal recumbency with the forelimbs affixed caudally. The head is affixed in an extended position. However, for extensive resections, the head may have to be released to decrease tension on the

anastomosis site. The sternohyoideus muscles are dissected along the midline. The peritracheal fascia is bluntly dissected from the trachea, taking care to prevent damage to the lateral pedicles and recurrent laryngeal nerves. The tracheal mass is located by visualization and/or palpation. A mass in the cranial thoracic trachea or within the thoracic inlet can also be resected using this approach. The sternum should be included in the surgical preparation because a cranial sternotomy may be necessary (Hedlund 1987, 1991). The thoracic trachea is approached by a right-sided third to fifth intercostal space thoracotomy. The third or fourth intercostal space will allow access to the caudal trachea. The fifth intercostal space will allow access to the carina. At the third intercostal space, a standard right lateral thoracotomy is performed, and the right cranial lung lobe is packed off cranially to allow visualization.

Respiratory Tract and Thorax  297

Figure 8.33.  Bronchial mass has been resected, and an anastomosis of the bronchus has been performed. (Image courtesy of Dr. G. Romanelli)

Figure 8.32.  Surgical approach to the bronchus by lateral thoracotomy. (Image courtesy of Dr. G. Romanelli)

The mediastinal pleura overlying the trachea is incised dorsal to the vagus nerve. The vagus nerve lies between the trachea and cranial vena cava. A second incision is made in the pleura just ventral to the vagus nerve to allow ventral retraction of the nerve (Smith and Waldron 1993). The soft tissue is dissected circumferentially around the trachea, taking care to stay directly against the tracheal wall. The tracheal resection and anastomosis at this site is performed similarly to the cervical tracheal resection. At the fifth intercostal space, a standard lateral thoracotomy is performed, and the right middle and cranial lung lobes are packed off (Figure 8.32). The azygous vein will be visualized as it crosses the trachea and enters the cranial vena cava. The azygous vein is ligated and transected. The pleura is incised, and the vagus nerve is protected and retracted as above (Smith and Waldron 1993). For tumors of the caudalmost extent of the trachea, the mainstem bronchi are transected at the level of the carina. Sterile endotracheal tubes are placed in each bronchus to allow for continued ventilation. The blunt dissection is performed to free the carina/caudal trachea from the surrounding soft tissues. The caudal segment is transected. The right bronchus is anastomosed to the caudal trachea as an end-to-end anastomosis (Figure 8.33). The left bronchus is anastomosed to the caudal trachea as an end-to-side anastomosis (Nelson 2003b). Incomplete cartilage rings and the size discrepancy in

this area make the anastomoses more challenging at this site (Nelson 2003b; Hedlund 1991). Knots are positioned away from blood vessels to prevent erosion of the blood vessels with movement of the intrathoracic structures (Hedlund 1991). A pleural or pericardial patch can be sutured over the anastomosis site to reinforce the anastomosis (Nelson 2003b; Hedlund 1987, 1991). Within the thoracic cavity, the anastomosis site can be assessed for leakage intraoperatively by filling the thoracic cavity with warm saline and monitoring leakage. A chest tube is placed prior to closure of the thorax to allow monitoring for a pneumothorax. Tension-relieving techniques There are several techniques that have been reported to decrease the tension on the anastomosis site for large segment resections. The decision to use these techniques is based on the surgeon’s judgment intraoperatively. Intraoperatively, simple techniques involve the use of tension-relieving sutures and an increase in the amount of dissection of the adventitia and fascia surrounding the trachea. The tension-relieving sutures are placed one to two cartilage rings from the anastomosis site. The sutures should encircle a tracheal ring. The sutures can be either a horizontal or vertical mattress sutures. Three tension-relieving sutures should be placed, one ventral and two lateral (Hedlund 1984, 1991; Vasseur 1979; Nelson 2003b). Care must be taken not to buckle or deform the trachea with these sutures (Dallman and Bojrab 1982). If the amount of dissection around the trachea is extended cranially and caudally, care must be taken not to disrupt the blood supply to the trachea or the recurrent laryngeal nerves. More complex methods to relieve tension intraoperatively include tracheal

298  Veterinary Surgical Oncology

stretch by incisions in the annular ligaments and laryngeal release. Tracheal stretch is achieved by incision of the annular ligaments cranial and caudal to the anastomosis site to relieve tension. Care must be taken not to penetrate the mucosa with these incisions. This will result in a weakening of the trachea, which could lead to complications such as tracheal disruption (Hedlund 1991; Nelson 2003b). Laryngeal release is reported in human patients when there is excessive tension on the anastomosis site (Wright et al. 2004). It has been suggested that this technique be used in humans when more than 4 cm of trachea are resected (Wright et al. 2004). With this technique, the attachments of the hyoid apparatus are dissected free from the thyroid cartilage (Hedlund et al. 1991). This technique is not used commonly in canine patients. In experimental large segment resections in dogs where laryngeal release was not performed, the larynx was found to be caudally displaced

Figure 8.34.  A postoperative neck flexion harness is being used to prevent neck extension and stress after tracheal resection and anastomosis. (Image courtesy of Dr. J. Liptak)

(a)

when pre- and postoperative radiographs were compared (Fingland et al. 1995; Dallman and Bojrab 1982). This indicates that there is significant tension on the larynx with large segment resections, and laryngeal release may be helpful in some cases to decrease tension on the anastomosis site. Postoperatively, immobilizing the head in a flexed position can also be used to relieve tension. This can be done by suturing the skin on the chin to the manubrium or by attaching a muzzle to a body harness to hold the neck flexed (Nelson 2003b; Hedlund 1984, 1991) (Figure 8.34). Tracheal stenting For tumors that are not suitable for surgical resection, either due to the length of resection that is required, diffuse disease of the trachea, or evidence of metastasis, palliation with tracheal stenting should be considered. The self-expanding, metallic stents are placed under fluoroscopic guidance in the same manner as for tracheal collapse in dogs (Figure 8.35). The maximal trachea diameter is measured radiographically, and a stent that is 10%–15% larger than this in diameter is selected to ensure that the stent is held in position under tension. The stent selected is 2 cm or longer than the tracheal segment that requires stenting to ensure that the stent will span 1 cm of normal-diameter trachea cranial and caudal to the stenosed area (Culp et al. 2007). This technique has been reported to successfully palliate a cat with a tracheal carcinoma. The cat succumbed to metastatic disease 6 weeks after stent placement, but no stent associated complications were noted (Culp et al. 2007). This technique has been reported for use in humans with malignant tracheobronchial stenosis (Miyazawa et al. 2000). Aftercare After tracheal resection, patients should be monitored in an intensive care unit for dyspnea and pain

(b)

Figure 8.35.  (A) Preoperative radiograph of a tracheal carcinoma causing tracheal obstruction caudal to the thoracic inlet. (B) Postoperative radiograph after intraluminal stent placement. (Images courtesy of Dr. W. Culp)

Respiratory Tract and Thorax  299

management. Patients with a cervical tracheal resection and anastomosis should be monitored for subcutaneous emphysema. Patients with a thoracic tracheal resection and anastomosis should have a thoracic tube placed intraoperatively so that they can be monitored for pneumothorax. Both would indicate separation of the anastomosis site and air leakage. Dogs should be walked with a body harness postoperatively to minimize tension and pressure on the trachea. Patients need to be kept confined postoperatively. Ideally, they should have strict cage rest for a week to prevent disruption of the tracheal anastomosis site. Cough suppressants should be considered if the patient is coughing after surgery to prevent excessive pressure on the anastomotic site. Tracheoscopy can be performed as necessary to monitor patients postoperatively for stenosis, airway leakage, or to evaluate healing. Lateral radiographs of the trachea can also be used to evaluate for stenosis. However, this technique has been shown to be relatively insensitive as a tool for this purpose (Fingland et al. 1995). Complications The most common complication of tracheal resection and anastomosis is stenosis. There are two causes of postoperative stenosis: excessive tension, leading to separation of the anastomosic site and healing by second intention, and poor mucosal apposition, with a similar consequence. The amount of stenosis has been shown to correlate to the amount of tension on the anastomotic site (Behrend and Klempnauer 2001a). A second common complication is breakdown of the anastomotic site and air leakage. This will generally require a second surgery unless the leakage is very mild. Quantity of trachea that can be safely removed The amount of trachea that can be safely removed is controversial. Many recommendations are based on reports in human thoracic surgery. These, in turn, are often based on dog and sheep models. There is a difference in the amount of tension that the trachea can withstand in puppies compared with adult dogs (Hedlund 1991; Fossum et al. 2007a). The most commonly cited references for the amount of trachea that can be resected is 25% in puppies and 20%–60% in adult dogs (Hedlund 1991; Fossum et al. 2007a). In sheep, which are used for a model of tracheal resection for humans, up to 9 cm have been resected with good results in vivo (Behrend and Klempnauer 2001b). There are no case reports of successful massive resections of the trachea in dogs. The resections that are reported are generally small and

successfully managed with primary anastomosis and minimal tension-relieving techniques. Two experimental reports of canine tracheal resection and anastomosis exist. The aim of both studies was to create a hightension anastomosis. This was performed by removing 8 (Fingland et al. 1995) and 15–17 (Dallman and Bojrab 1982) rings of the trachea. Minimal tension-relieving techniques were used, and most of these experimental resections were successful. Given that there are 35–45 rings in the canine trachea (Hedlund 1991), this would correspond to 20%–40% of the trachea resected successfully in experimental dogs. In humans, there are similar discrepancies in the amount of trachea that can be safely removed. A limit of 4.5 cm has been cited as a limit for tracheal resection in humans to prevent excessive tension and failure of the anastomotic site (Grillo et al. 1964; Mulliken and Grillo 1968). A recent retrospective study that evaluated complications associated with tracheal resection and anastomoses in humans found that there was an increase in the complication rate when 4 cm or more of trachea was resected (Wright et al. 2004). Although 4 cm was not considered an absolute cut off in this article, it was suggested that if more than 4 cm was to be resected, a laryngeal release procedure should be performed. Other citations are more liberal with the amount of trachea that can be resected, with 50% being cited as a limit for the amount of trachea that can be safely resected (D’Cunha and Maddau 2003). Another source suggests that 30% of the trachea can be resected without any tension-relieving techniques, with up to 70% resection possible with major tension-relieving techniques and organ displacement (Halsband 1987). As with puppies, the trachea in pediatric humans is not as amenable to resection. Thirty percent is the recommended limit in human pediatric tracheal resection (Wright et al. 2002). Patients that were 17 years old or younger were found to be more likely to have anastomotic complications in a retrospective study of human patients with tracheal anastomosis (Wright et al. 2004). The margins that are required depend on the tumor type. Even with a carcinoma or sarcoma of the trachea, 3 cm margins are unlikely to be achievable. Margins of 1 cm in each direction are recommended (Fossum et al. 2007a). Common tumor types In dogs, the reported tracheal tumor types include osteochondroma, chondroma, osteosarcoma, chondrosarcoma, adenocarcinoma, carcinoma, lieomyoma, mast cell tumor, plasmacytoma, fibrosarcoma, and rhabdomyosarcoma (Morrison 1980; Beck et al. 1999; Carlisle

300  Veterinary Surgical Oncology

et al. 1991; Brodey et al. 1969; Black et al. 1981; Hill et al. 1987; Chaffin et al. 1988; Mahler et al. 2006; Yanoff et al. 1996). Tracheal tumors are rare. There are several reports in the literature of tracheal osteochondromas (Morrison 1980; Beck et al. 1999; Dubielzig and Dickey 1978; Gourley et al. 1970; Hough et al. 1977). These are benign tumors of dogs less than 1 year of age that can be successfully treated with surgical resection (Morrison 1980; Beck et al.1999; Dubielzig and Dickey 1978; Gourley et al. 1970; Hough et al.1977; Carlisle et al. 1991). Reported tracheal tumor types in cats include lymphoma (Brown et al. 2003; Kim et al. 1996; Schneider et al. 1979), adenocarcinoma (Culp et al. 2007; Cain and Manley 1983; Evers et al. 1994; Veith 1974), disseminated histiocytic sarcoma (Bell et al. 2006), inflammatory polyp (Sheaffer and Dillon 1996), and carcinoma (Brown et al. 2003). The distinction between lymphoma and the other tumor types is very important because the treatment of lymphoma is nonsurgical, with chemotherapy and/or radiation being the treatment of choice. Generally, the treatment of excision will be the same for the other tumor types. However, the margins of resection for an inflammatory polyp will be much less than for a malignant tumor.

Figure 8.36.  Anatomy of the canine lungs. (Illustration courtesy of Dave Carlson)

Adjunctive therapy Adjunctive therapy will depend on the tumor type. For benign tumors or low-grade tumors with clean margins of resection, no adjunctive therapy is necessary. For tumors with a more aggressive course, adjunctive chemotherapy may be recommended. Radiation therapy has been reported for the treatment of tracheal lymphoma in cats (Brown et al. 2003).

Lung Surgical procedures Oncologic surgical procedures performed on the lungs of cats and dogs include the collection of biopsies to diagnose local or diffuse disease and partial or complete lobectomy or complete pneumonectomy for primary or metastatic primary or metastatic lung tumors (Mehlhaff and Monney et al. 1985; Miles et al. 1988; Ogilvie et al. 1989; O’Brien et al. 1993; McNiel et al. 1997; Hahn and McEntee 1998; Kuntz 1998; Liptak et al. 2004a, 2004b). Surgical anatomy The trachea of cats and dogs bifurcates into two mainstem bronchi, which in turn divide into lobar bronchi, one for each lung lobe. The two lungs are separated by the mediastinum (Dunning and Orton et al. 1998).

Deep fissures divide the lungs of dog and cat in lobes allowing for the change in shape when the diaphragm moves or the spine bends (Figure 8.36). The left lung consists of the cranial lobe, which is further divided by an incomplete fissure into a cranial (formerly called the apical lobe) and a caudal part (formerly called the cardiac lobe) (Figure 8.36). The caudal lobe of the left lung (formerly called the diaphragmatic lobe) is completely separated from the adjacent caudal part of the cranial lobe by a fissure (Evans 1993). The right lung is larger than the left, and it is divided into cranial (formerly called the apical lobe), middle, accessory (also called the intermediate lobe), and caudal lobes (Figure 8.36). Ventrally, the caudal portion of the cranial right lobe and the cranial part of the middle lobe fail to cover the heart surface. This area corresponds to the cardiac notch of the right lung, and it is usually located at the ventral aspect of the fourth intercostal space. At this site, about 5 cm2 of sternocostal surface of the heart is exposed to the thoracic wall, thereby creating a window for cardiac puncture and ultrasonographic cardiac imaging (Evans 1993). The notch for the caudal vena cava is located between the dorsal and the right lateral processes of the accessory lobe. At this point, in close

Respiratory Tract and Thorax  301

association with the vena cava, the right phrenic nerve is located. The pulmonary trunk from the right ventricle bifurcates into the left and right pulmonary arteries, which ramify into branches that enter the lung lobes. The pulmonary veins from the left lung lobes usually maintain their separate identity and enter the left atrium. The veins from right cranial and middle lobes join to form a single vein that returns to the left atrium; the veins from the caudal and accessory lobes join before entering the heart. Variations in vascular anatomy, however, are common (Evans 1993). The pulmonary vessels closely follow the lobar distribution of the bronchi. Pulmonary arteries are located on the craniodorsal aspect of each bronchi, whereas pulmonary veins are situated caudoventrally (Dunning and Orton et al. 1998). The lobar bronchi are positioned between the arteries (lateral) and the veins (medial) (Grandage 2003). For the most part, the pulmonary lymphatics drain into the three groups of the tracheobronchial lymph nodes around the tracheal bifurcation. In a few dogs, pulmonary lymph nodes are also found on the dorsal surfaces of the lobar bronchi at the edge of the lung parenchyma (Evans 1993). Biopsy procedures Definitive diagnosis of lung lesions requires evaluation of tissue samples or cytologic specimens. The nature of a pulmonary nodule can be investigated by performing a transthoracic fine-needle aspiration (FNA) by a blind collection technique or under the guide of ultrasound, CT, or fluoroscopy. The technique used depends on the accessibility of the lesion. Diagnoses obtained by FNA cytopathology accurately reflected the diagnosis obtained on histopathological examination in 82% of cases. The agreement between cytopathological and histopathological interpretation was higher in samples collected with ultrasound guidance than in those collected in a blind fashion (DeBerry et al. 2002). Pulmonary lesions located in the periphery of the lobe can be more easily aspirated. The role of FNA and the requirement for preoperative diagnosis in the management of focal lung lesions is controversial. Most authors agree that FNA of pulmonary parenchymal lesions represents a useful, accurate, and safe tool for diagnosis of lung neoplasia and should always be performed. The reported accuracy of this technique for diagnosis in dogs and cats with pulmonary lesions varies between 82% and 91% (Teske et al. 1991; Wood et al. 1998; Reichle and Wisner et al. 2000; DeBerry et al. 2002; Zekas et al. 2005). The incidence of complications (i.e., pneumothorax, hemorrhage) ranges from 0% to 31% (Teske et al. 1991; Wood

Figure 8.37.  Postoperative view of lung cancer. Note the presence of necrotic and purulent material in the tumor that may confuse cytological interpretation of transthoracic FNA.

et al. 1998; Reichle and Wisner et al. 2000; DeBerry et al. 2002). In contrast, other authors cite that FNA of pulmonary lesions is diagnostic for neoplasia in only 37%– 50% of cases (Mehlhaff and Monney et al. 1985; McNiel et al. 1997). Focal lung tumors often have a necrotic and purulent center that may confuse interpretation (Figure 8.37) (Withrow 2007). For this reason, and due to the fact that FNA or other biopsy-obtained diagnoses rarely change the course of the treatment, it is appropriate to recommend surgical excision of a lung mass without attempting a preoperative FNA or other biopsy procedure. This decision is also influenced by the size and location of the nodule, the results of a CT examination that exclude other pulmonary lesions or lymph node involvement, the general health status of a patient, and the owner’s decision making. Tracheal wash, bronchoalveolar lavage, or brush cytopathology have poor sensitivity for diagnosis of lung neoplasia unless the cancer invades the tracheobronchial tree (Ogilvie et al. 1989; Hahn and McEntee et al. 1997; McNiel et al. 1997). These techniques provide the most information when there is interstitial or alveolar disease rather than a focal pulmonary lesion (Ogilvie et al. 1989). The utility of cytological analysis of pleural effusions in animals with lung tumors is controversial. In one study the diagnosis was confirmed based on the cytology of pleural effusion in 12 of 13 cats that had thoracocentesis performed (Hahn and McEntee 1997). In another report, however, a diagnosis was obtained in only 1 of 8 cats (Barr 1987). Samples for histopathological analysis can be obtained by percutaneous lung biopsy with a cutting needle

302  Veterinary Surgical Oncology

Figure 8.38.  Thoracic radiographs taken in (A) right lateral, (B), left lateral, (C) and dorsoventral projections of a 11-year-old dog with a well-circumscribed mass in the left cranial lung lobe. After surgical excision, the lung mass was diagnosed as an adenocarcinoma.

(Reichle and Wisner et al. 2000), or via transbronchial biopsy via a fiber-optic bronchoscope (Kuehn and Hess et al. 2004). Neither of these techniques is routinely used in cats and dogs. Biopsy samples of the pleural surface, lymph nodes, pericardium, and lung can be obtained via thoracoscopy using a transdiaphragmatic or intercostal approach (Griffin 2004) or via a lateral thoracotomy or median sternotomy. Imaging tests Conscious chest radiographs are a key step in the diagnosis of lung tumors (Mehlhaff and Monney et al. 1985; Kuntz 1998). Left and right lateral and ventrodorsal views of the thorax are always recommended (Figure 8.38). Radiographic appearance of primary lung tumors varies from a single, discrete mass in one lobe to multiple lesions and/or the diffuse involvement of entire

lung lobes (Mehlhaff and Monney et al. 1985; Miles 1988). Occasionally, pneumothorax or pleural effusion may complicate the radiographic diagnosis. Chest radiographs, however, can have limited diagnostic accuracy in detecting pulmonary metastases, pulmonary carcinosis, and tracheobronchial lymphadenopathy (Johnson et al. 2004; Paoloni et al. 2006). CT has been shown to be more efficacious than conventional radiography for detection of lung metastases (Nemanic et al. 2006) and for evaluation of tracheobronchial lymphadenopathy in the staging process (Paoloni et al. 2006) (Figure 8.39). MRI can also be used for the evaluation of lymph nodes. CT also allowed a clear identification of pleural nodules in case of malignant pleural mesothelioma (Echandi et al. 2007). In one study, thoracic radiography failed to reveal approximately 90% of pulmonary nodules that were subsequently detected on CT examination. The size of nodule detection is approximately 1 mm in diameter for CT, and that for thoracic radiography approaches

Respiratory Tract and Thorax  303

(a)

(c)

(b)

Figure 8.39.  (A) Right lateral and (B) left lateral (projections of the thorax of an 11-year-old German shepherd with mammary tumors and mandibular fibrosarcoma). Metastatic lesions were not detected on radiographs. CT scan of the thorax of the same dog (C) obtained at the level of the cranial lobes: a single subpleural nodule (arrow) was clearly identified in the left cranial lobe. (Image courtesy of Dr. Mauro Di Giancamillo)

7–9 mm for reliable detection (Nemanic et al. 2006). The high sensitivity but the lower specificity of CT screening can result in the diagnosis of a large number of falsepositive nodules (Li et al. 2004). In human medicine, therefore, the false-positive rate has highlighted the need to differentiate benign from malignant nodules. Higher accuracy in the radiologist’s evaluation of lung nodule morphology (size, lobulation, presence of coarse speculation, heterogeneous central attenuation) facilitated by, if needed, the use of automated computerized schemes can mitigate the problem of false-positive nodules (Li et al. 2004; Markowitz et al. 2007). If pulmonary metastases are suspected on throacic radiographs, it is the author’s recommendation that if the owner declines CT or MR, or if they are not available, thoracic radiographs should be repeated after 4–6 weeks to assess for the development of pulmonary nodules. Ultrasonography is not superior to radiography for the detection of pulmonary lesions. It is, however, indicated to guide diagnostic transthoracic aspiration or biopsy, thereby improving the accuracy of this procedure, particularly when pleural fluid is present. Pleural fluid may obscure details of thoracic structures on radiographs; however, it can enhance the diagnostic potential of thoracic ultrasound (Moore and Ogilvie et al. 2006). Isolated, discrete pulmonary lesions are unlikely to benefit from bronchoscopy. This procedure is, however, useful to visualize and biopsy intrabronchial lesions (Ogilvie et al. 1989; Kuehn and Hess 2004). Surgical procedures The thorax can be approached by a lateral intercostal thoracotomy or by a median sternotomy (for more details refer to section on thoracotomy).

Figure 8.40.  Intraoperative view of an enlarged tracheobronchial lymph node. (Image courtesy of Dr. Giorgio Romanelli)

Lung lobectomy Partial or complete lung lobectomy can be performed. Because the cranial and caudal parts of the cranial left lobe share common bronchi and vessels, it is difficult to excise one without removing the other. They are, therefore, usually removed together. On the right side, the accessory lobe divides incompletely from the caudal lobe, and it is generally resected with the caudal lobe. The caudal left and right lobes, the middle right lobe, and the cranial right lobe can, however, be separately removed (Orton 1995a). When lung tumor excision is performed, tracheobronchial lymph nodes (Figure 8.40) should always be palpated and biopsied if enlarged. If the pleural wall or mediastinum adheres to the tumor, these structures should be biopsied to evaluate

304  Veterinary Surgical Oncology

the completeness of surgical excision. Surgical margins should be inked and submitted for histologic evaluation in addition to the primary tumor mass (Kuntz 1998). Partial lobectomy The peripheral two-thirds or less of the lung lobe can be removed to obtain a biopsy sample or to treat peripheral focal lesions such as small tumors (Orton 1995a; Dunning and Orton et al. 1998; Nelson and Monnet 2003; Fossum et al. 2007b). Partial lobectomy can be performed through a left or right fourth-fifth intercostal thoracotomy or via median sternotomy. The affected lobe is identified, carefully exteriorized from the thoracic cavity, and gently palpated to assess the extent of lung excision. A moist laparotomy sponge can be placed under the lobe. To avoid spillage of neoplastic or infected cells, a pair of crushing forceps is placed across the lobe proximal to the lesion, at the resection site. Continuous overlapping sutures (2-0 to 4-0 absorbable suture) are placed 3–4 mm proximal to the forceps (Figure 8.41). If the partial lobectomy is performed at the proximal third of the lobe, larger bronchi and blood vessels may be encountered. To decrease the risk of hemorrhage or air leaks after resection, these structures can be individually occluded before performing the previously described suture pattern or a second row of continuous overlapping suture can be placed (Figure 8.41). At this point, the lung is transected with a scissor or a surgical blade proximal to the forceps, leaving a small portion of lung

Figure 8.41.  Partial lung lobectomy with two continuous overlapping sutures oversewn with a simple continuous suture. (Illustration courtesy of Dave Carlson)

distal to the suture line. The incision is oversewn with a simple continuous pattern of absorbable suture (3-0 to 5-0) (Figure 8.42). The remaining portion of the lobe is replaced in the thoracic cavity. Surgical gloves and instruments should be changed before closing the thorax. At this point, the chest cavity is filled with warm sterile saline solution to cover the excision site and positive pressure ventilation is maintained for a few seconds to evaluate for air leaks at the suture line. Simple interrupted sutures or hemoclips are placed to close air leaks. The fluid is removed, and a thoracostomy tube placed prior to closing the thorax. Thoracic radiographs (laterolateral and dorsoventral views) can be taken after surgery to evaluate pneumothorax and to check the position of the thoracostomy tube (Figure 8.43). Partial lobectomy can also be performed using surgical staples (LaRue et al. 1987; Walshaw 1994; Liptack et al. 2004a; Tobias 2007). A thoracoabdominal (TA) surgical stapler is adequate for this purpose. This instrument places two or three staggered rows of stainless steel staples into the tissue, which assume a B shape when compressed. TA devices are available as a reusable reloadable stainless steel stapler or as a disposable stapler that can be reloaded up to seven times. The reusable staplers are available in 30, 55, and 90 mm widths (TA 30, TA 55, TA 90), with cartridges having staples of 2.0, 3.5, or 4.8 mm. The disposable reloadable staplers are available in 30, 45, 60, 90 mm widths, with cartridges having a staple size of 2.5, 3.5, or 4.8 mm. Staple cartridges are available in different colors depending on staple size. The blue cartridges contain two rows of staples 4.0 mm wide, with a leg length of 3.5 mm and a closed height of 1.5 mm. The green cartridges have two rows of staples with a crown width of 4.0 mm, leg length of 4.8 mm, and closed height of 2.0 mm. The white ones, known as V3, contain three rows of staples 3.0 mm wide, with legs of 2.0 or 2.5 mm and a closed height of 1.0 mm. They come in 30 mm widths only. The V3 cartridges are preferred due to the added security provided by the additional row of staples and the smaller closed height. The selection of staple length and the height depends on the compressed width of the lung or bronchus at the resection site and on the compressed thickness of the tissue at the same point, respectively. It is important that all tissues to be ligated are comfortably placed within the staple line. For most partial lobectomies in dogs and cats, either the TA 55 or TA 90 stapler is used, with 3.5 mm staples. The stapler is placed across the lobe proximal to the lesion and fired (Figure 8.44). The lobe is transected, using the TA edge as a cutting guide. Before cutting the tissue from the stapler, the lung

Respiratory Tract and Thorax  305

(a)

(b)

(c)

(d)

Figure 8.42.  Operative view of partial lung lobectomy. The affected lobe is grasped with a dry laparotomy sponge (A). A crushing forceps is placed across the lobe proximal to the lesion, and a continuous overlapping suture is placed 4 mm proximal to the forceps (B). The lung is transected proximal to the forceps, leaving a small portion of lung distal to the suture line (C). The incision is oversewn with a simple continuous pattern of absorbable suture (D).

is clamped with a hemostat to avoid the spread of infected and/or tumor cells into the pleural space. The staple line is inspected for hemorrhage and air leakage. When performing a stapled partial lobectomy, there is no indication to routinely oversew staple lines. If persistent points of blood or air leakage are noted, they can be independently occluded either with individual sutures or vascular clips. It can be more expedient, with less risk of hemorrhage and pneumothorax to perform a complete rather than a partial lobectomy, particularly when dealing with a neoplastic process for which it is important to achieve adequate margins. Complete lobectomy The preferred surgical approach for complete lung lobectomy is via a left or right intercostal thoracotomy

over the affected lobe and its hilum (Dunning and Orton et al. 1998; Nelson and Monnet 2003; Fossum et al. 2007b). Median sternotomy does not allow easy visualization of pulmonary vessels and bronchi. When there are large amounts of infected material in the lobe to be removed, the bronchus should be clamped to prevent the passage of this material into the proximal bronchi and trachea (Fossum et al. 2007a). To isolate the lobe, moist laparotomy sponges are used. To isolate the hilus of the lobe, the pulmonary ligaments that attach the lobe to the mediastinum are transected while the visceral pleura is incised to visualize the pulmonary vessels. Careful attention is paid to correctly identify the vasculature and bronchus associated with the lobe to be resected. Using blunt dissection for the nonvisible deep side of the vessel, the pulmonary artery is exposed on its

306  Veterinary Surgical Oncology

entire circumference. To facilitate vessel ligation and to have a temporary emergency suture available for hemostasis in the event of vessel rupture, a small moistened umbilical tape can be passed around the artery for traction. The artery is ligated with nonabsorbable or absorb-

Figure 8.43.  Postoperative thoracic radiograph to evaluate pneumothorax and to check the position of the thoracostomy tube.

able suture material (2-0 to 3-0) (Figure 8.45A). A simple encircling ligature is first placed at the proximal end of the vessel near its bifurcation. Care is taken not to compromise the lumen of the parent vessel from which this artery arises. A second ligature is applied distal to the point where the artery is to be transected (Figure 8.45A). For safety, a transfixing suture should be placed proximal to the transection site. The artery is severed between the two distal sutures. The veins are approached on the ventral side of the bronchus, after retraction of the lobe dorsally. The pulmonary veins are ligated in a manner similar as for the arteries, with the exception that the distal suture should be placed first (Figure 8.45A). Care is taken not to lacerate the thin-walled veins during dissection and to avoid incorporating an adjacent vein in the suture. The main bronchus of the lobe is first dissected and then clamped with a pair of crushing forceps or Satinsky forceps, placed proximal and distal to the point where the transection is to be performed, close to the lobe (Figure 8.45B). The bronchus is transected between the two clamps and sutured proximal to the remaining clamp with a continuous or interrupted horizontal mattress suture (2-0 to 3-0 nonabsorbable monofilament suture) (Figure 8.45c). In small dogs and cats, a transfixing

Figure 8.44.  Operative view of a partial lung lobectomy performed with surgical staples. The lung lobe is lifted up (A), and the stapler is placed across the lobe proximal to the lesion and fired (B). (C) The lobe is transected, using the TA edge as a cutting guide.

Respiratory Tract and Thorax  307

surgical gloves and instruments should be changed. The thoracic cavity is filled with warmed sterile saline, and positive pressure ventilation is applied to check for air leaks. Before closure, fluid and sponges are removed and the packed off lungs are assessed for appropriate reinflation. A thoracostomy tube is routinely placed. Complete lobectomy can be performed with absorbable or nonabsorbable suture material. In the presence of an infectious process, braided, multifilament, and/or nonabsorbable material should be avoided (Fossum et al. 2007b). Complete lobectomy can be accomplished using an en bloc hilar stapling procedure (Walshaw 1994; LaRue et al. 1987). The hilus of the lobe to be excised is isolated sufficiently from the surrounding structures so that it fits easily into the cartridge of the TA stapler. With this technique, it is not necessary to individually isolate, ligate, and transect the pulmonary vessels. Only the surrounding adventitia needs to be dissected from the point of resection. Care is taken not to include the phrenic or vagus nerve in the staple line. A TA 30 stapler with 2.5 mm (white, V or V3) staple cartridges is preferred for complete lung lobectomy (Walshaw 1994). Before cutting distal to the staple line, large hemostats are placed to clamp the tissue distal to the staples to avoid spillage of material (infected or neoplastic) into the pleural cavity. There is no indication to routinely oversew staple lines after either partial or complete lobectomy (Walshaw 1994) (Figure 8.46). Pneumonectomy

Figure 8.45.  Hand suture technique for lung lobectomy. (A) Ligation of the veins and arteries, (B) division of the mainstem bronchus between the clamps, (c) mattress sutures, (D) simple continuous oversew of the bronchus. (Illustration courtesy of Dave Carlson)

suture can be used. A simple continuous suture pattern (3-0 to 4-0 absorbable suture) is placed on the distal end of the severed bronchus to oversew it (Figure 8.45D). To decrease the risk of adhesions and air leaks, some sutures can be placed in the surrounding pleura to cover the stumps of the vessels and bronchus. At this point,

A pneumonectomy refers to the resection of all lung lobes on either the right or left side (Figure 8.47), and it can be performed in both dogs and cats (Nelson and Monnet 2003; Liptak et al. 2004a). Removal of more than 75% of the lung is fatal (Dunning and Orton et al. 1998). The left and right lungs account for 42% and 58% of lung volume, respectively. Dogs tolerate a left pneumonectomy or resection of less than 50% of the lung volume, provided that the right lung is healthy. Compensation occurs by existing mechanisms only, such as the distention of the remaining lung and increased pulmonary blood flow, resulting in the recruitment of existing physiological reserves of diffusion capacity and the remodeling of the existing alveolar–capillary network. After right pneumonectomy (55%–58% of lung tissue is excised), compensation is due to the same mechanisms as left pneumonectomy as well as new or regenerative alveolar-capillary growth (Hsia et al. 1993, 1994; Nelson and Monnet 2003; Liptak et al. 2004a). Neoplasia involving all lobes of one lung is an indication for pneumonectomy (Nelson and Monnet 2003).

308  Veterinary Surgical Oncology

Figure 8.47.  Postoperative view of the left lung after pneumonectomy. (Image courtesy of Dr. Giorgio Romanelli)

Figure 8.46.  Intraoperative view of complete lung lobectomy performed with staples. (A) The hilus of the lobe to be excised is isolated sufficiently from the surrounding structures to fit easily into the staple cartridge. (B) There is no indication to routinely oversew staple lines after complete lobectomy. (Image courtesy of Dr. Giorgio Romanelli)

To improve visibility and operative space in the thorax, the affected lobes can be double-clamped at their pedicle and resected before performing vessel ligation. Pneumonectomy can be accomplished by en bloc resection by ligating the pulmonary artery, vein, and mainstem bronchus (Nelson and Monnet 2003; Clements et al. 2004; Liptak et al. 2004a) or by individual lobectomies for each lobe of one lung (Liptak et al. 2004a). It is wise to have a temporary salvage suture placed around the pulmonary artery of the lung to be resected, close to the bifurcation of the pulmonary trunk, in the event of hemorrhage and to provide traction. The pulmonary artery to be transected can be ligated by the placement of two simple encircling sutures (for vessels smaller than 5 mm) with or without a transfixation suture or oversewn (vessels greater than 5 mm) with a double layer of

4-0 or 5-0 simple continuous monofilament nonabsorbable suture. Before being transected and oversewn, two noncrushing vascular clamps are placed: the proximal clamp on the pulmonary artery adjacent to the main pulmonary trunk and the second clamp 1 cm further distally. The artery is then divided, leaving no more than 5 mm of artery distal to the proximal clump to decrease the length of the “blind end” where thrombi can potentially form. The artery is then oversewn. The veins are always double-ligated with a transfixing suture, depending on their size. Before approaching the bronchus, it is important to withdraw the endotracheal tube proximal to the carina or advance it in the contralateral bronchus. The mainstem bronchus is then clamped distal to the transection site. A monofilament nonabsorbable suture can be placed through the tracheal wall just adjacent to the bronchus for traction. A simple interrupted or continuous horizontal mattress suture (3-0 or 4-0 monofilament nonabsorbable suture material) is placed on the bronchus, 5–10 mm distal to the carina. The transection is made 3 mm distal to the suture, and the proximal cut edge of the bronchus is oversewn with a simple continuous suture. Since the mainstem bronchus has tracheatype cartilage rings, air leaks can occur as it does not easily collapse after suturing and the sutures can cut through the cartilage. The closure, therefore, can be reinforced by suturing a piece of pleura, pericardium, or fascia over the stump (Nelson and Monnet 2003). As an alternative to en bloc removal, individual complete lobectomies can be performed whereby pulmonary vessels are sutured and transected as previously described for a complete lung lobectomy. The mainstem bronchus is closed with interrupted or continuous mattress sutures and oversewn with a simple continuous suture pattern,

Respiratory Tract and Thorax  309

as for complete lobe resection. An en bloc hilar stapling technique can also be performed for each lobe (Walshaw 1994). The excision site is checked for bleeding, and the thorax is filled with saline, as previously described, to evaluate for air leaks. A thoracostomy tube should be always placed. As an alternative to pneumonectomy, an experimental study on extracorporeal lung resection in dogs has been described (Matsumoto et al. 2004) whereby the unilateral lung is extirpated, the pulmonary lobe with cancer is removed, and the residual pulmonary lobe is reimplanted. The technique is proposed for cases in which tumor removal is difficult, even by complete lobectomy or pneumonectomy. The extensive lymph node dissection required in some cases, however, has severe adverse effects on bronchial anastomotic healing (Matsumoto et al. 2004). Thoracoscopy Complete or partial lobectomy can be performed via thoracoscopy using an endoscopic stapling device, thereby using a minimally invasive approach (Brissot et al. 2003; Lansdowne et al. 2005) (Figure 8.48). This procedure is particularly useful for small neoplastic lung masses, located away from the hilus. Among all lobes, the caudal left lung lobe appears to be the easiest to remove successfully without intraoperative complications (Lansdowne et al. 2005). Biopsies from the tips or margins of lung lobes can also be performed. Biopsy samples are obtained via thoracoscopy by inserting a commercially available or selftied loop ligature (Roeder knot), through an instrument portal (Fossum et al. 2007b). A minimally invasive technique can also use a keyhole thoracotomy technique under thoracoscopic guidance (thoracoscopic-assisted keyhole thoracotomy) (Figure 8.49). A small intercostal incision is made, and Babcock forceps are used to grasp and exteriorize the lung tip to be biopsied. A ligature is placed around the margin or a thoracoabdominal stapler is used to resect the lung (Fossum et al. 2007b). Biopsy samples from peribronchial lymph nodes can also be obtained via thoracoscopy to stage the neoplastic disease for prognostic purposes. The advantages of thoracoscopy include decreased morbidity, decreased postoperative pain, and a more rapid recovery. Disadvantages include the cost of specialized equipment, the expertise required to perform the procedure, decreased gas exchange, and the possibility of cardiovascular compromise, both of which also occur with thoracotomy. A limitation of thoracoscopy is the ability to removal large tumors or tumors of any size located close to the hilus. Thoracoscopic lobectomy is best performed when the lungs are completely deflated, using a one-lung

Figure 8.48.  Thoracoscopic view of (A) a total lung lobectomy in a cat before placement of staples and of (B) a partial lung lobectomy in a dog after placement of staples. (Image courtesy of Dr. Jilles Duprè)

Figure 8.49.  Biopsy samples are obtained through a keyhole procedure via thoracoscopy. The lesion is localized via thoracoscopy, and a small intercostal incision is made to exteriorize the lung tip to be biopsied. A thoracoabdominal stapler is used to resect the lung. (Image courtesy of Dr. Giorgio Romanelli)

310  Veterinary Surgical Oncology

ventilation technique. Partial lung collapse due to the pneumothorax created by the introduction of trocar cannulas into the thoracic cavity is usually sufficient for exploratory thoracoscopy for biopsy of pleura or lung. Complications associated with thoracoscopy requiring conversion to a thoracotomy include hemorrhage from a lacerated intercostal vessel causing blood loss and poor intraoperative visibility, failure to maintain the lung deflation, and poor surgical access. The latter usually occurs with right middle lung lobe resection due to the fact that, after the placement of the cannulae, there is insufficient working space for safe application of staples at the hilus (Lansdowne et al. 2005). Thoracoscopy in humans is contraindicated in the case of endobronchial tumors, evidence of thoracic wall or mediastinal involvement, presence of concurrent lung disease making one-lung ventilation poorly tolerated, or enlarged mediastinal lymph nodes (Lansdowne et al. 2005). Similar contraindications are likely to exist in small animals. To decrease the risk of tumour seeding when removing lung tumors via thoracoscopy, the use of an endoscopic retrieval bag is preferred (Lansdowne et al. 2005). Readers are referred to endoscopic texts for more detailed information regarding thoracoscopic techniques. Metastasectomy Metastatic lesions in the lung parenchyma can be excised either by partial or complete lobectomy through lateral thoracotomy or via sternotomy when there is bilateral lung lobe involvement. When nodules up to 5 mm are located in the visceral subpleural tissues, they can be elevated with hemostats and excised, following the placement of a purse-string suture (3-0 or 4-0 monofilament absorbable suture material), to close the surrounding lung parenchyma. The placement of a low-pressure chest tube should prevent pneumothorax if the suture falls off. For subpleural lesions over 5 mm, a stapled wedge excision should be used (O’Brien et al. 1993). The number of metastatic lesions is one of the key factors that influences the success of lung metastasectomy; a more sensitive imaging technique than radiography, therefore, should be used to screen dogs for multifocal disease prior to metastasectomy (Liptak et al. 2004b). Postoperative care Animals should be hospitalized and closely monitored during the postoperative period for approximately 2–3 days. Respiratory function should be evaluated by blood gas analysis. Pain management should be addressed: provision of analgesia after thoracotomy is extremely important to facilitate recovery and to improve ventilation (Kuntz 1998).

Complications of thoracic surgery Perioperative mortality is rare after partial or complete lung lobectomy. Complications are related to the surgical technique but are uncommon (LaRue et al. 1987). The use of staples of inadequate length or height increases the chance of air leakage or hemorrhage due to insufficient tissue compression (Walshaw 1994). The major complications after partial and complete lobectomy or pneumonectomy include pneumothorax and/ or hemorrhage. Minor air leaks usually seal spontaneously and are self-limiting. Massive leaks (i.e., copious hemorrhage, sustained pneumothorax) require reexploration of the chest. Animals should be closely monitored in the postoperative period. Lung lobectomy should not be attempted where necrotic or tumor tissue is included in the staple or suture line, as it can result in suture dehiscence or in staple line disruption with subsequent hemothorax or pneumothorax (LaRue et al. 1987). High rates of morbidity and mortality are reported after pneumonectomy in human medicine. Acute and chronic respiratory, cardiac, and gastrointestinal complications are relatively common in both humans and animals (Liptak et al. 2004a). They include respiratory insufficiency or failure, pneumonia, pulmonary edema, thromboembolism and hypertension, chylo- and hemothorax, pleural effusion, congestive heart failure, supraventricular arrhythmia, myocardial ischemia, esophageal dysmotility and dilation, esophagopleural fistula, and delayed gastric emptying. The entrapment of the endobronchial blocker tip used for selective lung ventilation by a surgical staple has been reported in a dog (Levionnois et al. 2006). Types of lung tumors Primary lung tumors in the dog (Mehlhaff and Monney et al. 1985) and cat (Hahn and McEntee 1997) are rare; however, their prevalence appears to be increasing (Moulton et al. 1981). This trend may be due to a number of reasons, including longer companion animal life expectancy, the development of more accurate diagnostic techniques, an increased number of necropsies and biopsies being performed, and the constant improvement in standards of pet care (Moore and Ogilvie et al. 2006). Although most affected dogs come from urban areas, there is no apparent association between the environment and development of lung tumors (Ogilvie et al. 1989; Reif et al. 1992). Recently, however, an increased risk of lung cancer was observed in dogs with higher amounts of anthracosis (Bettini et al. 2010). Primary lung tumors are described by their site of origin (i.e., bronchial, bronchoalveolar, alveolar) or by their

Respiratory Tract and Thorax  311

(a)

(b)

(c)

Figure 8.50.  Eleven-year-old domestic shorthair cat affected by digit lung syndrome. Lung masses are evident on a lateral thoracic radiograph (A). It is characterized by metastatic spread of lung tumors to digits with evidence of digit ulceration (B) and phalangeal bone lysis on radiographs (C).

histopathological appearance (i.e., adenocarcinoma, squamous cell carcinoma) (Moulton et al. 1981). However, it is commonly difficult to classify lung tumors based on location due to their advanced status at the time of diagnosis (Moulton et al. 1981). Adenocarcinomas are the most common primary lung tumor reported in both species (Ogilvie et al. 1989, Hahn and McEntee 1997). Squamous cell carcinomas, anaplastic carcinoma, and other type of carcinomas are less frequently seen (Mehlhaff and Monney et al. 1985). Benign lung adenomas and primary mesenchymal lung neoplasms (osteosarcoma, hemangiosarcoma, and fibrosarcoma) are infrequently reported in the dog and cat. Lymphomatoid granulomatosis is a rare neoplasm of the lung in young to middle-aged dogs. Lobar consolidation or a large granuloma is evident on radiographs, and hilar

lymphadenopathy, circulating basophilia, and leukocytosis are common findings (Berry et al. 1990; Fitzgerald et al. 1991). Malignant histocytosis is a lung tumor that is highly metastatic. It is most commonly described in Bernese mountain dogs, but has also been described in rottweilers and golden retrievers (Rosin et al. 1986; Hayden et al. 1993; Ramsey et al. 1996). Metastatic pulmonary tumors are more common than primary lung tumors in dogs and cats (Moulton et al. 1981). Primary tumors that are most frequently associated with metastatic lung lesions on thoracic radiographs include transitional cell carcinoma, thyroid carcinoma, hemangiosarcoma, melanoma, and osteosarcoma (Miles et al. 1988). Distinguishing primary lung carcinomas from metastases is often a challenging task. The thyroid transcription factor-1 (TTF-1), a nuclear protein expressed in follicular cells of the thyroid gland and pneumocytes, was found to be 100% specific and 85% sensitive for primary lung carcinomas. Antibodies to TTF-1 have been found to be a useful marker for distinguishing between primary and metastatic canine epithelial tumors (Bettini et al. 2009). Primary lung neoplasms are highly aggressive and tend to metastatize to the lung, regional lymph nodes, pleural space (Hahn and McEntee 1997), skeletal musculature (Hahn and McEntee 1997; Langlais et al. 2006), bone (Hahn and McEntee 1997; Dhaliwal, KufuorMensah et al. 2007), heart, liver, kidney, etc (Dhaliwal, Kufuor-Mensah et al. 2007). The median age in dogs at diagnosis varies from 10 to 11 years (Miles et al. 1988; Ogilvie et al. 1989). For cats, the median age at diagnosis is approximately 12 years (Mehlhaff, Monney et al. 1985; Hahn, McEntee et al. 1997). Sex and breed predisposition have not been reported (Mehlhaff et al. 1984), although in an earlier study, a higher incidence of lung carcinomas was found in the boxer (Brodey and Craig 1965). Primary pulmonary masses are single in 54% and multiple in 37% of dogs (Ogilvie et al. 1989). Clinical signs may be absent in both dogs and cats in the early phases of the disease (Ogilvie et al. 1989); therefore, animals with pulmonary neoplasia can appear normal on physical examination. The finding of primary pulmonary neoplasia may be incidental when thoracic radiographs are taken for other reasons. Symptoms appear late in the course of the disease and may depend on the amount of lung involved, the invasiveness of the tumor, and the presence of metastatic disease (Mehlhaff et al. 1984). Clinical signs can be directly or indirectly associated with pulmonary tumors. With effusion secondary to the tumor, commonly encountered clinical signs include a nonproductive cough of a few weeks’ to several months’ duration (Mehlhaff et al. 1984) and/or

312  Veterinary Surgical Oncology

Figure 8.51.  (A) Lateral thoracic projection of an 11-year-old mixed dog with a pulmonary mass in the caudal lung lobe. The same dog showed lameness due to hypertrophic osteopathy. (B) Diffuse periosteal reaction is evident on the radiograph of the affected limb.

dyspnea. Decreased bronchovesicular sounds occasionally may be seen with large tumors, secondary spontaneous pneumothorax, or with pleural effusion (Mehlhaff et al. 1984; Hahn, McEntee et al. 1997). Coughing and other signs referable to the respiratory tract are a less consistent finding in the cat, noted only in about onethird of cases (Mehlhaff, Monney et al. 1985). Other nonspecific and frequently detected signs in animals with lung tumors include lethargy, malaise, inappetence, weight loss, exercise intolerance, and pyrexia. Such nonspecific signs are most commonly seen in the cat (Mehlhaff, Monney et al. 1985). Occasionally, dogs and cats are presented for lameness (Mehlhaff, Monney et al. 1985; Hann, McEntee et al. 1997). This can be secondary to hypertrophic osteopathy (Mehlhaff, Monney et al. 1985; McNiel et al. 1997) whereby the distal limbs appear diffusely swollen and painful. Lameness can also result from distant metastatic spread of the primary lung neoplasm to bone or skeletal muscle (Hahn, McEntee 1997; Langlais et al. 2006), with a focal soft tissue swelling and pain on palpation of the affected bone (Mehlhaff, Monney et al. 1985). Neurologic signs secondary to tumor involvement of neurologic tissue has also been reported (Ferreira et al. 2005). In cats, the “digit lung syndrome” is characterized by metastatic spread of lung tumors to digits, with lameness, evidence of digit ulceration, and phalangeal bone lysis described (Figure 8.50) (Jacobs and Tomlinson 1997; Gottfried et al. 2000).

Paraneoplastic syndromes are not commonly associated with lung tumors in either dogs or cats, but can include hypertrophic osteopathy (Figure 8.51), hypercalcemia, fever, and primary lung tumor-induced secretion of adrenocorticotropic hormone (Ogilvie et al. 1989; Hahn, McEntee et al. 1997). Surgery is the most effective and commonly recommended treatment for lung tumors in cats and dogs (Melhlaff et al. 1984; Mehlhaff, Monney et al. 1985; Miles et al. 1988; Ogilvie et al. 1989; O’Brien et al. 1993; McNiel et al. 1997; Hahn and McEntee 1998; Kuntz 1998; Liptak et al. 2004a). Adjuvant therapy Limited data are available on the efficacy of chemotherapy in the treatment of lung tumors. Patients that may benefit from chemotherapy include those with unresectable lesions that require palliation, those with recurrent disease, or cases with negative prognostic factors such as high-grade, poor differentiation, and regional lymph node involvement. Vindesine (with or without cisplatin) has been shown to be of some benefit in dogs (Mehlhaff et al. 1984). Vinorelbine has been used in a few dogs with lung tumors, with a partial response (Poirier et al. 2004). Vincristine, cyclophosphamide, and methotrexate have been used in combination in dogs (Mehlhaff et al. 1984). The administration of mitoxantrone has been described in a cat with well-differentiated

Respiratory Tract and Thorax  313

adenocarcinoma (Clements et al. 2004) and in dogs combined with doxorubicin (McNiel et al. 1997). Because of the small number of animals treated with adjuvant chemotherapy, it is not possible to assess the impact of chemotherapy on prognosis. Intrapleural administration of cisplatin has been shown to be of some benefit in animals with malignant pleural effusion (Moore et al. 1991). The delivery of cytotoxic chemotherapy and cytokines by aerosol for the treatment of primary or metastatic lung cancers has also been described with encouraging results (Hershey et al. 1999; Khanna, Vail et al. 2003). The only lung mass that shows a good and rapid response to chemotherapy is lymphomatoid granulomatosis (Berry et al. 1990). Radiation therapy for lung tumors is largely untried in veterinary oncology because of the potential for serious and life-threatening effects on lungs, such as pneumonitis and subsequent fibrosis (LaRue et al. 1995; McNiel et al. 1997). It is likely that adjuvant radiation therapy would increase survival when combined with surgery in incompletely excised lung tumors (Moore, Ogilvie et al. 2006). More sophisticated methods of irradiation for lung tumors, such as intensity-modulated radiation therapy, have been recently introduced in veterinary oncology with some promising results (Ballegeer et al. 2006). Photodynamic therapy may have some efficacy as an adjuvant treatment for pulmonary metastases from sarcoma (Anderson et al. 2003). Prognosis The most significant prognostic factor in dogs with isolated lung lesions is lymph node involvement (Ogilvie et al. 1989; McNiel et al. 1997; Polton et al. 2008). Dogs with metastatic spread to regional nodes have diseasefree intervals and median survival times shorter than those of dogs without lymph node involvement (6 days versus 351 days and 26 days versus 452 days, respectively) (McNiel et al. 1997). Histologic grade also has prognostic value. Dogs with well-differentiated tumors have significantly longer survival times and disease-free intervals (median, 790 and 493 days, respectively) than do dogs with moderately (median, 251 and 191 days, respectively) or poorly (median, 5 and 0 days, respectively) differentiated tumors (McNiel et al. 1997). Other factors of prognostic value include the presence of clinical symptoms. Survival time is longer (median survival time 545 days) in dogs with lung tumors that are an incidental finding than in dogs that are symptomatic (median survival time 240 days) (McNiel et al. 1997). Primary tumor size has shown some influence on prognosis, with larger tumors having a poorer prognosis

(Ogilvie et al. 1989; McNiel et al. 1997). The tumor stage (T) is also prognostic for median survival time (T1 tumors, 790 days; T2 tumors, 196 days; T3 tumor, 81 days) (McNiel et al. 1997). Primary tumor stage T1 and histologic type (papillary tumor type) were found to be statistically significant favorable prognostic indicators (Polton et al. 2008). Dogs with adenocarcinoma (mean survival time [MST] 19 months) have a much better prognosis than dogs with squamous cell carcinoma (MST 8 months) (Mehlhaff et al. 1984). Dogs with neoplasms located in the periphery of the lung lobe have a better prognosis than those with tumors located at the hilus, as the former tumors are more likely to be completely resected (Withrow 2007). Dogs with small (diameter less than 5 cm), isolated, well-differentiated adenocarcinomas, without evidence of spread to regional lymph node and without pleural effusion, have the best prognosis; 1 year survival can be expected in more than 50% of these animals (Mehlhaff et al. 1984; Ogilvie et al. 1989). The prognosis in cats with primary lung tumors is considered less favorable than for dogs because of the advanced stage of the disease at the time of diagnosis and the aggressive metastatic behavior of the tumor. Over 75% of cats are not candidates for surgical excision at the time of presentation because of the presence of metastatic disease or due to the local extension of the tumor (Hahn and McEntee 1997; Hahn and McEntee

Table 8.2.  Clinical stages of primary lung tumors. T: Primary lung tumor (based on clinical and surgical evaluation)

T0: No evidence of tumor T1: Solitary tumor surrounded by lung or visceral pleura T2: Multiple tumors of any size T3: Tumor invading neighboring tissues

N: Regional lymph nodes (based on surgical and histologic evaluation)

N0: No evidence of lymph node involvement N1: Bronchial lymph node involved N2: Distant lymph node involved

M: Distant metastasis (based on surgical and histologic evaluation)

M0: No evidence of distant metastasis M1: Distant metastasis detected

Clinical stages (TNM) of primary lung tumors. (Modified from Owen LN. 1980. Geneva: WHO; adapted from McNiel et al 1997.)

314  Veterinary Surgical Oncology

1998). Grade of differentiation and lymph node involvement are of prognostic value in cats. Cats with moderately differentiated tumors (median survival time 698 days) and those without enlarged nodes (median survival time 421 days) have significantly longer survival times than cats with poorly differentiated lung tumors (median survival time 75 days) or enlarged nodes (median survival time 73 days) (Hahn and McEntee 1998).

Metastasectomy for Sarcomas Sarcomas are very common cancers in dogs and cats. Even with permanent local control and adjuvant chemotherapy the metastatic rate varies by tumor type, stage, site, species, and histologic grade. Common examples include canine appendicular osteosarcoma with an ultimate metastatic rate over 90% versus soft tissue sarcomas with a metastatic rate of less than 25%. Metastasis used to be considered a universal harbinger of imminent cancer-related death. Wide variation in metastatic rates, sites, symptoms, paraneoplastic syndromes, progression, and outcomes exist, however, and speaks to a biological heterogeneity of metastatic disease progression that may occasionally allow a meaningful intervention with durable survivals. Canine osteosarcoma is a template for the following biological, temporal, imaging, and procedural principles that can be applied generally to all tumor types, assuming that permanent local control has been achieved with surgery or radiation and that the patient has often received adjuvant chemotherapy (O’Brien et al. 1993). The biology of metastasis is complex but can be thought of in the seed (tumor cell) and soil (site of metastasis) paradigm. Simplistically, most sarcomas spread hematogenously to the lungs, followed by the bones, lymph nodes (especially synovial cell sarcoma), and other soft tissue sites (Diemel et al. 2009). With this in mind, routine follow-up after primary treatment usually includes a physical exam and chest radiographs every 3 months for the first year, with decreasing intervals for year 2 and beyond. Once metastasis is identified, full staging may be in order (involving CT of the chest, an abdominal ultrasound, and a nuclear bone scan), or it can be delayed until a decision on possible treatment is made. Pulmonary metastasis of osteosarcoma is the most common setting where data exist to support metastasectomy in animals and will be used as an example (O’Brien et al. 1993). General considerations for metastasectomy include the following: 1. Time to detection.  The longer the interval from primary tumor control to detection of metastasis,

the better, with 300 days suggested as the break point from early versus late onset. In theory, those metastases with late detection are less aggressive yet somehow have evaded complete and durable chemotherapy cytotoxicity. In rare cases, single new lung nodules may represent primary lung carcinoma. 2. Number of lesions.  One or two metastases are biologically much less than three or more. The typical scenario is the asymptomatic detection of lung nodule(s) on plain radiographs. Three-view films in a conscious (fully aerated) patient are recommended (Prather et al. 2005). CT scans may help define lesions not clearly seen on plain films but can also reveal small nonneoplastic lung nodules of indeterminate origin (granulomas, osteoid osteomas, etc.) (Nemanic et al. 2006). Positron emission tomography/CT (PET/CT) should help distinguish metabolically active tumor tissue from nontumor tissue (Ballegeer et al. 2006). CT may help determine the exact site of the tumor within the lung (surface tumors vs. tumors deep in the parenchyma), which may aid in the decision to do an open thoracotomy rather than thoracoscopy. The number of pulmonary lesions is only a relative criteria for metastasectomy in humans where as many as 50 lesions are occasionally removed. 3. Doubling times.  Once one or two nodules are noted in thoracic radiographs, it is usually recommended that repeat radiographs be taken in 30 days to determine a rate of growth and possible clinical detection of new lesions that would make surgical intervention more problematic. It is generally felt that patients most amenable to metastasectomy will still be good candidates in 30 days and that slow-growing tumors are better candidates than fast-growing tumors. The use of chemotherapy during this waiting time is not clear. 4. Technique.  Open intercostal thoracotomy is the usual surgical technique, although a sternal split can be considered for bilateral lesions or the rare larger lesions. Open approaches allow for more thorough palpation of lung lobes for unsuspected lesions (Quiros and Scott 2008). Selective lung deflation may aid in digital palpation of nodules. Thoracoscopy, or video-assisted thoracic surgery (VATS), is becoming increasingly popular in the hands of experienced surgeons when lesions are definable to accessible (visible) sites and not adjacent to the hilus (Lansdowne et al. 2005). Most lesions are peripheral, discrete, subpleural, and amenable to lobectomy, wedge resection with staples, or “cherry-picking” of small and subpleural lesions (this entails a suture ligature below the base of the lesion).

Respiratory Tract and Thorax  315

(a)

(b)

Figure 8.52.  (A) A ventrodorsal projection radiograph of a dog with a rib chondrosarcoma. (B) A lateral projection radiograph of a dog with a rib chondrosarcoma. (Image courtesy of Dr. Julius Liptak)

5. Outcomes after metastasectomy.  In carefully selected patients (late onset, one or two lesions, slow growth, and resectable), long-term survival can occur after metastasectomy. Even with variable inclusion criteria, the 1–year survival rate for canine osteosarcoma has been reported to be 30% for dogs with osteosarcoma and subsequent pulmonary metastasectomy (O’Brien et al. 1993). The role of postmetastasectomy chemotherapy is undefined, but considerations can be made for new adjuvant drugs that have not been previously used, clinical trials of new agents, or even metronomic low-dose chemotherapy regimes (Anderson et al. 2008). Postmetastasectomy surveillance begins again at 3-month intervals and repeat metastasectomy can be considered (Bielack et al. 2009). Metastasis of cancer is generally a poor prognostic sign, especially with carcinomas. Carefully selected patients may benefit from surgical resection of late onset, slow-growing, solitary metastasis. The paraneoplastic disease of hypertrophic osteopathy may also resolve after pulmonary metastasectomy (Liptak et al. 2004). Nonpulmonary sites of metastasis are managed with the same general principles, for example, radiation or resection for boney metastasis, lymphadenectomy for lymph node metastasis, local resection for stump recurrences (especially with synovial cell sarcoma) (Pfannschmidt et al. 2009).

Thoracic Wall Resection Biopsy procedures An incisional biopsy is indicated to determine the tumor type. This can be done as a wedge incisional biopsy or with a Trucut biopsy (needle-core biopsy) technique. The advantage of the wedge biopsy is that a larger biopsy can be obtained. The size of the biopsy has been shown to be an important factor in achieving a correct diagnosis (Montgomery et al. 1993). The advantage of the Trucut biopsy technique is that it is a faster procedure that may not require general anesthesia and has less disruption of the tissue planes. Regardless of technique, the biopsy should be planned so that the entire biopsy tract can be easily removed at the time of definitive surgery. The biopsy should be taken centrally over the tumor with only one incision. There should be minimal disruption of tissue planes deep and lateral to the biopsy incision. Imaging tests Three-view thoracic radiographs are a good starting point to evaluate a mass over the thoracic wall. This will serve to determine the intrathoracic extent of the lesion and to evaluate for pulmonary metastasis. With chondrosarcomas in particular, the palpable mass on the exterior of the thorax may be a small percentage of the total mass (Figure 8.52). The point of

316  Veterinary Surgical Oncology

Figure 8.53.  Postcontrast transverse CT of an osteosarcoma involving the fifth rib. (Image courtesy of Dr. Simon Kudnig)

origin of the mass may be evident with radiography alone. However, this test is not as sensitive as 3D imaging, and it may not be possible to correctly determine the origin or extent of disease based on radiography alone. If the tumor is of rib origin, radiographs may indicate the degree of bone lysis or production. This has not been found to help distinguish between osteosarcoma and chondrosarcoma (Liptak, Kamstock et al. 2008). CT is the imaging modality of choice to evaluate chest wall tumors (Figure 8.53). It allows the accurate evaluation of the origin of the tumor and the other structures that are involved in the thorax. The length of rib involvement and extent of soft tissue involvement can also be accurately assessed for surgical planning. The lungs can also be evaluated for evidence of metastatic disease with CT. Nuclear scintigraphy should be considered in cases of osteosarcoma to determine if there is evidence of metastatic disease or if the rib mass is a metastatic site. If nuclear scintigraphy is not available, long-bone survey radiography is an alternative method to evaluate for other bone lesions. Liptak, Kamstock and colleagues (2008) found a 16% rate of metastasis to bone with primary rib osteosarcoma in a recent study. Abdominal ultrasound is recommended for staging in certain tumor types. This test should be performed in cases of known hemangiosarcoma or when a pleural effusion accompanies the thoracic wall mass. Description of surgical procedures Rib tumors Preoperative surgical planning is very important for the surgical treatment of rib tumors. Rib tumors are usually sarcomas, and aggressive resection is critical to success-

Figure 8.54.  Resection of a subcutaneous hemangiosarcoma. In this case, skin was included in the chest wall resection.

ful treatment. Completeness of surgical excision has been shown to have a significant effect on survival and disease-free interval in dogs with chest wall tumors (Pirkey-Ehrhart et al. 1995; Liptak, Kamstock et al. 2008). The planned margins of resection should include one rib cranial and one rib caudal to the lesion (Liptak, Kamstock et al. 2008; Baines et al. 2002). Dorsal and ventral margins should be 3 cm along the ribs (Liptak, Kamstock et al. 2008; Baines et al. 2002). One school of thought is that the entire rib should be removed by disarticulating the rib with the vertebra and sternum. However, this does not appear to be necessary. The reported maximum number of ribs that can be resected is six (Liptak, Dernell et al. 2008; Pirkey-Ehrhart et al. 1995; Orton 2003) Anecdotally, seven ribs have been successfully removed (NJ Ehrhart NJ and SJ Withrow, personal communication). However, the removal of seven or eight ribs increases the risk of causing severe respiratory compromise and dysfunction. The location of the tumor may determine the ability of the surgeon to remove more than six ribs. This is better tolerated in the caudal thorax, where diaphragmatic advancement is possible. In the cranial thorax, the creation of a flail chest with the removal of more than six ribs may cause ventilatory failure (NJ Ehrhart, personal communication). In general, skin is not resected en bloc with tumors arising from the rib. This is contrary to a recent paper that used en bloc resection and a myocutaneous flap for reconstruction of primary rib chondrosarcoma in five dogs (Halfacree et al. 2007). There may be exceptions to this if there is extensive invasion in the soft tissues lateral to the rib seen on CT (Figure 8.54). The common rib tumors such as osteosarcoma and chondrosarcoma tend to remain somewhat encapsulated within the periosteum. This concept is similar to

Respiratory Tract and Thorax  317

Figure 8.55.  The previous biopsy tract is removed with the resection; however, mobile skin is spared over the tumor. The skin involving the biopsy tract is sutured to the underlying fascia. This is the same case as Figure 8.53. (Image courtesy of Dr. Simon Kudnig)

the resection of distal radial osteosarcomas that spare surrounding soft tissues in a limb-sparing procedure. The overlying muscle also serves as another plane that is a barrier to extension of the tumor into the subcutaneous tissue and skin. The 3D imaging should serve as a guide for whether or not skin resection is necessary. Another rule of thumb is to assess the mobility of the skin over the mass. If the skin is mobile over the mass, the need for resection is less likely. If a biopsy is performed, the biopsy tract will need to be resected with the tumor (Figure 8.55). The latissimus dorsi muscle can also be spared if appropriate based on imaging and its mobility. The latissimus dorsi muscle is an important source of autogenous tissue for reconstruction of the defect. The patient is placed in lateral recumbency with the appropriate side up. The entire exposed hemithorax and a large portion of the abdominal skin should be clipped and prepared for surgery. The patient is positioned with the uppermost limb extended forward. The skin incision is made over the mass. The incision may be curvilinear, vertical, or T-shaped, depending on the size and location of the mass. An ellipse of skin is created around the biopsy tract. The biopsy tract should be sutured to the underlying tissues to prevent motion (Figure 8.55). The latissimus dorsi muscle is undermined and retracted dorsally if it is to be spared. An intercostal approach is made at either the cranial or caudal margin of the resection. This should be the margin that is most obvious based on preoperative imaging and palpation (Figure 8.56). The length of the intercostal approach is dictated

Figure 8.56.  Intraoperative photograph of a dog with a rib chondrosarcoma. The lateral thorax has been entered at the proposed intercostal space of excision. (Image courtesy of Dr. Julius Liptak)

Figure 8.57.  Intraoperative photograph of a dog with a rib chondrosarcoma after thoracic wall resection. (Image courtesy of Dr. Julius Liptak)

by the size of the tumor. Rib cutters are used to cut the ribs at the dorsal and ventral margins. The intervening intercostal tissue is cut with Mayo scissors or electroscalpel. The intercostal vessels caudal to the ribs should be located and ligated dorsally and ventrally. Care must be taken to locate and ligate the internal thoracic artery as unaddressed disruption of this vessel has been reported to cause fatal hemorrhage (Liptak, Dernell et al. 2008). The final cranial or caudal intercostal incision is made to remove the chest wall en bloc (Figure 8.57). Generally, rib tumors are located in the area of the costochondral junction (Montgomery et al. 1993). The thoracic mass

318  Veterinary Surgical Oncology

may invade or be adherent to intrathoracic structures such as the lungs and pericardium. Adhesions should not be broken down as this may result in contamination of the field with tumor cells. The involved tissues should be removed with the chest wall. This may involve a pericardiectomy or a partial lung lobectomy using a TA stapler (Mattieson et al. 1992; Liptak, Dernell 2008; Pirkey-Ehrhart et al. 1995). For cranial rib resections, care must be taken to avoid the brachial plexus. Most importantly, the origin of the nerve roots that make up the radial nerve can be found coursing around the first rib. Resection including the first rib has been reported to be successful with no impairment of limb function (Liptak, Kamstock et al. 2008). Invasive soft tissue masses of tissues lateral to the ribs In general, invasive soft tissue masses lateral to the ribs are soft tissue sarcomas. They should be removed using the same principles as with sarcoma removal anywhere on the body. As with rib tumors, these tumors require 3D imaging for appropriate presurgical planning. On 3D imaging, these masses would involve the soft tissues from the subcutaneous tissue and skin medially to the chest wall. If there is not a fascial plane between the tumor and the chest wall, these tumors require a true en bloc resection from skin through the chest wall (Figure 8.54). The patient will be prepared similarly for surgery. A more extensive clip may be needed to allow for skin reconstruction. Using a sterile marking pen, the mass should be drawn on the patient. The margins of excision should then be measured and drawn. The margins of excision should be 3 cm around the mass (Ehrhart et al. 2005). The skin incision follows the planned margins. This incision is continued medially through the underlying muscle to the level of the chest wall. An electroscalpel may be helpful once the skin incision has been made to aid in hemostasis. The skin over the tumor can be sutured to the underlying tissues to prevent the tissue planes from sliding on each other. It is important that the incision in the underlying tissues continues perpendicular to the original skin incision to ensure adequate margins. There is a tendency to make the subsequent incisions in deeper tissues toward the mass, causing a coning down of the excision. This can lead to inadequate surgical margins. Once the ribs and intercostal muscles are reached, intercostal incisions are made in the most appropriate locations based on the planned en bloc resections. Always err on the side of a larger resection when planning the cranial and caudal intercostal incisions. Bone cutters are used along the ribs and Mayo scissors or electroscalpel is used between the ribs to con-

tinue the en bloc resection. The entire mass and soft tissue margins are removed. Although it is rarely reported in veterinary medicine (Munday et al. 2006; Ferreira et al. 2005), lung or chest wall tumors may penetrate the chest wall at the level of the brachial plexus. Such tumors may require lung and chest wall resection combined with forequarter amputation. The same principles of en block resection are followed. Reconstruction after thoracic wall resection There are several options for reconstruction of the thoracic wall defect after rib resection or true en bloc resection. These options include the use of prosthetic mesh implants such as Marlex, Gor-Tex, and Vicryl. Marlex is the most commonly reported implant used in veterinary and human thoracic wall reconstruction techniques. Local tissue flaps have also been reported either alone or in combination with prosthetic mesh. The most commonly used local tissue is a latissimus dorsi flap, either alone or as a myocutaneous flap (Liptak, Dernell et al. 2008; Halfacree et al. 2007; Raffoul et al. 2001; Mansour et al. 2002). Marlex mesh was first described for reconstruction of thoracic wall defects in dogs by Bright in 1981. Its use has been largely adopted from widespread use in human thoracic reconstruction. The major advantages of Marlex mesh are that it is readily available, strong, easy to sterilize and it has been shown to develop rapid ingrowth of fibrous tissue (Bright 1981). In dogs, it is recommended to cut a piece of mesh that is slightly larger than the defect. The mesh is folded over by 1 cm around its periphery to allow a more solid area to hold sutures Figure 8.58). The mesh is sutured in place by laying the

Figure 8.58.  Intraoperative photograph of mesh that is used to reconstruct a defect after chest wall resection.

Respiratory Tract and Thorax  319

Figure 8.59.  Intraoperative photograph of a dog with a rib chondrosarcoma. The defect in the chest wall has been repaired using mesh. (Image courtesy of Dr. Julius Liptak)

mesh along the pleural surface of the defect and suturing it in place with circumcostal and mattress sutures through the chest wall (Figure 8.59) (Bright 1981). The mesh is stretched when it is sutured so that it is taut and affords some rigidity to the chest wall. When possible, soft tissues are closed over the mesh prior to skin closure. When a large number of ribs are removed, chest wall stability has also been restored by using spinal plates at the sites of rib resection. The plates are affixed in place using hemicerclage wires through the remaining ribs (Ellison et al. 1981). There are currently no guidelines to determine when this method should be employed. Rigid reconstruction may not be necessary and is not currently commonly used for chest wall reconstruction. The latissimus dorsi flap was first reported as a method for chest wall reconstruction in humans in the late 1800s (Mansour et al. 2002). It is used commonly in people, often in combination with Marlex mesh (Mansour et al. 2002). It has been reported in dogs as a muscle flap and a musculocutaneous flap when skin is resected (Liptak, Dernell et al. 2008; Halfacree et al. 2007). The latissimus dorsi muscle originates from the superficial leaf of the lumbodorsal fascia associated with the dorsal spinous processes of the thoracolumbar vertebrae and the last two to three ribs. Its insertion is also wide. It courses toward the shoulder and holds the dorsal scapular border against the chest. The latissimus dorsi muscle inserts on the teres tuberosity of the humerus (Pavletic 2003; Dyce et al. 1987b). It has been classified as a type V muscle, with the thoracodorsal

artery being the main blood supply. The intercostal arteries also supply segmental branches to the muscle (Pavletic 2003; Purinton et al. 1992). To harvest the latissimus dorsi flap, the skin incision made for tumor resection may need to be extended. The latissimus muscle is freed up ventrally with blunt dissection. The origin of the muscle is excised dorsally. The dissection is continued along the dorsal border of the muscle, parallel to the muscle fibers, allowing the muscle to be rotated into the defect. The dominant vascular pedicle containing the thoracodorsal artery is preserved (Pavletic 2003). Alternately, the muscle flap can be harvested from its insertion at the level of the humerus and used to repair defects more caudally (Seguin, personal communication). For the latissimus dorsi myocutaneous flap, the flap is planned using a sterile marking pen. The dorsal border is drawn from a point that is ventral to the acromion and caudal to the border of the triceps muscle. The line is drawn to the head of the 13th rib. The ventral border of the flap is drawn from a point at the forelimb skin fold, caudal to the triceps muscle. The line is drawn parallel to the dorsal border to the 13th rib. The caudal border is drawn by connecting the dorsal and ventral borders along the 13th rib. The ventral border is incised first, and the ventral border of the latissimus dorsi muscle is located. The flap is developed by continuing the incisions from the skin to the ventral aspect of the latissimus dorsi. The myocutaneous flap can be rotated into the defect, and the donor site skin can be closed primarily (Figure 8.60) (Pavletic 2003; Halfacree et al. 2007). The omentum can also be used to provide an autogenous, airtight seal if there is not enough local tissue to close the defect. This can be achieved by making a flank incision and tunneling the omentum in the subcutaneous tissue and into the defect. Skin is then closed over the omentum (Figure 8.61) (Orton 2003). For tumors of the caudal chest wall, diaphragmatic advancement can be used to decrease the size of the thoracic defect to be closed or completely close the defect. This is performed by detaching the diaphragm from its attachments laterally and ventrally to allow mobilization of the diaphragm. The diaphragm is then advanced and sutured to the remaining chest wall to allow for an airtight, rigid fixation (Figure 8.62). A partial or complete caudal lung lobectomy may have to be performed concurrently to prevent atelectasis of the affected hemithorax and subsequent ventilation/ perfusion mismatch (Orton 2003). This is best performed with a TA stapling device. The amount of the lung lobe resected depends on the degree of diaphragmatic advancement.

320  Veterinary Surgical Oncology

(a)

(b)

Figure 8.60.  (A) Intraoperative photograph showing elevation of the latissimus dorsi muscle flap. (B) Intraoperative photograph showing rotation of the latissimus dorsi flap cranially. The flap will be brought ventrally to fill the defect in the chest wall. (Image courtesy of Dr. Julius Liptak)

Liptak, Dernell, and colleagues (2008) recently compared autogenous, prosthetic, and composite methods of chest wall reconstruction retrospectively. They found that the complication rate was higher for prosthetic and composite techniques compared with autogenous reconstruction, with complications being 12.8 times more likely to occur with prosthetic techniques and 3.0 times more likely to occur with composite techniques compared with autogenous tissue reconstruction. The current recommendation for chest wall reconstruction is to use a latissimus dorsi flap when possible and to add Marlex mesh if necessary to augment the repair. In the human literature, there is a lot of importance placed on maintaining absolute rigidity of the chest wall after resection (Mansour et al. 2002; Weyant et al. 2006). This is usually achieved using a Marlex meshmethyl methacrylate sandwich (Weyant et al. 2006). However, the use of autogenous tissue alone has been successfully reported in human cases where the use of prostheses was precluded by infection (Raffoul et al. 2001). Maintaining rigidity of the thoracic wall after reconstruction is also stressed in the veterinary literature (Ellison et al. 1981; Bright 1981). Care should be taken to suture the muscle flap or mesh so that it is taut. However, absolute rigid reconstruction of the thoracic

Figure 8.61.  The omentum has been placed over the thoracic defect to provide an additional autogenous tissue to close the defect. (Image courtesy of Dr. Simon Kudnig)

wall has not been shown to be necessary in dogs (Liptak, Dernell et al. 2008). Sternectomy Masses of the sternum can be approached in much the same way as masses in other areas of the thoracic wall.

(a)

(b)

(d)

(c)

(f)

(e)

Figure 8.62.  (A) Intraoperative picture of a dog with a mass over the caudolateral thorax. Resection of the caudal thoracic and cranial abdominal portions of the lateral abdominal wall. The surgeon’s hand is in the thorax, pushing the diaphragm caudally. (B) Resection of the caudal thoracic and cranial abdominal portions of the lateral abdominal wall. The diaphragm is being excised from the lateral body wall. (C) Resection of the caudal thoracic and cranial abdominal portions of the lateral abdominal wall. The diaphragm is being excised from the lateral body wall. (D) Resection of the caudal thoracic and cranial abdominal portions of the lateral abdominal wall. The diaphragm is advanced cranially and sutured to the caudal edge of the remaining thoracic wall. (E) Resection of the caudal thoracic and cranial abdominal portions of the lateral abdominal wall. The diaphragm is advanced cranially and sutured to the caudal edge of the remaining thoracic wall. (F) Resection of the caudal thoracic and cranial abdominal portions of the lateral abdominal wall. The diaphragm is advanced cranially and sutured to the caudal edge of the remaining thoracic wall. (Images courtesy of Dr. Julius Liptak)

321

322  Veterinary Surgical Oncology

Figure 8.63.  CT scan of a cat with a sternal chondrosarcoma. (Image courtesy of Dr. G. Romenelli)

Incisional biopsy for tissue diagnosis and 3D imaging will guide preoperative planning (Figure 8.63). The decision to include skin in the resection will be based on the same logic as for rib tumors. The overlying pectoral muscle may be resected with the tumor. The amount of sternum resected will depend on the amount of bone involved on CT imaging. Threecentimeter margins should be taken cranial and caudal to the tumor. The ribs are cut 3 cm from the sternal mass. The sternum is cut using an oscillating bone saw, and the ribs are cut using bone cutters. The segment of the chest wall is removed en bloc (Figure 8.64). The defect is reconstructed using Marlex mesh (Figure 8.64), a Marlex mesh–poly(methyl methacrylate) sandwich, heterogenous bone and mesh, spinal plates (Figure 8.65), or an autogenous muscle flap. The pectoral muscle can be used in the reconstruction if it has not been removed with the en bloc resection. A deep pectoral muscle has recently been reported as a method to repair the chest wall after sternectomy in dogs (Liptak, Dernell et al. 2008). This can be used alone or in combination with a latissimus dorsi flap and/or Marlex mesh. The deep pectoral muscle is a type V muscle. The dominant vascular pedicle is the lateral thoracic artery that enters the deep face of the muscle cranially and supplies the craniodorsal portion of the muscle. The segmental branches of the internal thoracic artery enter along the sternal attachment and supply the

cranioventral portion of the muscle (Purinton et al. 1992). The origin of the deep pectoral muscle is the ventral sternum and the fibrous raphe along midline. The insertion is to the greater tubercle of the humerus and the medial brachial fascia (Evans et al. 2000). The flap is harvested by incising its midline sternal attachments and undermining the muscle. The cranial attachment should be preserved to maintain the branches of the internal thoracic artery, if possible. The flap is rotated across ventral midline into the contralateral defect. Alternately, the flap can be rotated cranially and dorsally on the lateral thoracic pedicle (Liptak, Dernell et al. 2008). The amount of rigidity required depends on the size and location of the defect. If many ribs are removed and/ or if the manubrium is resected, an effort should be made to restore rigidity to the reconstruction. This can be done with a Marlex mesh–poly(methyl methacrylate) sandwich and autogenous tissue. There was an increased incidence of early complications seen in sternal resections and reconstruction in dogs in a recent retrospective study (Liptak, Dernell et al. 2008). This may indicate that this location is less forgiving in terms of reconstruction techniques. Aftercare Patients require intensive postoperative care for pain management and to monitor for respiratory dysfunction. Pain should be managed using opioid analgesics in combination with NSAIDs (if not contraindicated). Ketamine and lidocaine continuous rate infusionss can also be considered. A chest tube should be placed intraoperatively to monitor for and manage pleural effusion, should it occur. Nasal oxygen will be helpful in some cases because chest wall resection will create some degree of hypoventilation after surgery due to pain and due to the change in the conformation of the chest wall. Blood gas analysis should be part of the postoperative monitoring plan to evaluate for potential respiratory complications such as hypoventilation, ventilation/perfusion mismatch, and aspiration pneumonia. An indwelling urinary catheter should be considered to allow strict rest of the patient for 12–24 hours and to allow the clinician to monitor urine output. All patients should be on intravenous fluids until they are eating and drinking. In a large thoracic resection, fluid loss from hemorrhage and evaporative losses intraoperatively can be substantial. Careful monitoring of the patient’s hydration status in the postoperative period is important. Cosmetic and functional outcome The reported maximum for removal of ribs is six. This is generally well tolerated. There is no effect on the

(a)

(b)

(c)

(d)

Figure 8.64.  (A) Preoperative picture of a cat with a sternal chondrosarcoma (B) Intraoperative picture of a cat with a sternal chondrosarcoma. (C) Intraoperative picture of a cat with a sternal chondrosarcoma after sternectomy. (D) Intraoperative picture of a cat with a sternal chondrosarcoma after sternectomy and reconstruction with mesh. (Images courtesy of Dr. G. Romanelli)

(a)

(b)

Figure 8.65.  (A) Intraoperative photograph of a sternectomy for soft tissue sarcoma in a Jack Russell terrier. The mass has been resected with 2.5 cm margins en bloc. (B) Intraoperative photograph of a sternectomy for soft tissue sarcoma in a Jack Russell terrier. The mass has been resected with 2.5 cm margins en bloc. Lubra plates are used to reconstruct the sternal defect.

323

324  Veterinary Surgical Oncology

functional outcome once the patient has recovered from surgery. Potential complications Reported complications of chest wall resection include seroma, pleural effusion, and peripheral and limb edema (Liptak, Dernell et al. 2008; Pirkey-Ehrhart et al. 1995). Lameness, infection, and dehiscence are also possible complications. A rare but severe complication associated with chest wall resection is respiratory failure leading to death. This can be due to a combination of factors, including changes in the conformation of the chest wall due to a large resection, aspiration pneumonia, sepsis and systemic inflammatory response or acute respiratory distress syndrome. Owners should be counseled before surgery that postoperative ventilation may become necessary and that an inability to wean the patient off of the ventilator may necessitate euthanasia. Common tumor types The most common reported primary tumors of the chest wall are osteosarcoma and chondrosarcoma. Other reported tumor types include fibrosarcoma, hemangiosarcoma, soft tissue sarcoma, and leiomyosarcoma (Baines et al. 2002; Montgomery et al. 1993; Liptak, Dernell et al. 2008; Mattieson et al. 1992; Pirkey-Ehrhart et al. 1995). Adjunctive therapy Rib osteosarcoma has been shown to have a similar biological behavior to appendicular osteosarcoma. PirkeyEhrhart et al. (1995) showed that dogs with rib osteosarcoma treated with surgery alone had a MST of 90 days, whereas dogs with rib osteosarcoma treated with surgery and chemotherapy had a MST of 240 days. Liptak, Kamstock et al. (2008) showed a MST of 290 days for rib osteosarcoma, with 19 or 23 of these dogs being treated with chemotherapy. Chondrosarcoma treated with surgery alone has a much more favorable prognosis, with reported median survival times of 1,080 days (Pirkey-Ehrhart et al. 1995) and greater than 3,820 days. Pulmonary metastasis has been reported in cases of chondrosarcoma (Liptak, Kamstock et al. 2008), but chemotherapy