Validation of Bioanalytical Methods (essentials) 9783658389123, 9783658389130, 3658389125

Table of contents :
What You Can Find in this Essential
Contents
1: Introduction
2: Method Categories
3: Validation Parameters
3.1 Accuracy
3.2 Precision
3.3 Further Validation Parameters
4: Validation Environment and Validation Planning
4.1 Validation Environment: The Big Difference Between Academic and GMP Laboratories
4.2 Validation Planning
5: Validation of Bioanalytical Methods
5.1 Identity: Validation of a PCR Test
5.2 Determination of Content: Validation of a Virus Titration
5.3 Qualitative Impurities: Validation of a Test for Absence of Microorganisms (Sterility)
5.4 Quantitative Impurity: Validation of an Endotoxin Assay
6: Errors, Problems and Risks Associated with Insufficient Method Validation
7: Summary
What the Reader Can Take Away from this Essential
References

Citation preview

Patric U. B. Vogel

Validation of bioanalytical methods

essentials

essentials provide up-to-date knowledge in concentrated form. The essence of what is important as “state-of-the-art” in the current technical discussion or in practice. essentials provide information quickly, easily and understandably • as an introduction to a current topic in your field • as an introduction to a topic that is still unknown to you • as an insight to be able to have a say on the topic The books in electronic and printed form present the specialist knowledge of Springer authors in a compact manner. They are particularly suitable for use as e-books on tablet PCs, e-book readers and smartphones. essentials are knowledge building blocks from economics, social sciences and humanities, from technology and natural sciences as well as from medicine, psychology and health professions. By renowned authors from all Springer publishing brands.

Patric U. B. Vogel

Validation of bioanalytical methods

Patric U. B. Vogel Vogel Pharmopex24 Cuxhaven, Germany

ISSN 2197-6708     ISSN 2197-6716 (electronic) essentials ISBN 978-3-658-38912-3    ISBN 978-3-658-38913-0 (eBook) https://doi.org/10.1007/978-3-658-38913-0 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer Spektrum imprint is published by the registered company Springer Fachmedien Wiesbaden GmbH, part of Springer Nature. The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany

What You Can Find in this Essential

• • • • •

An introduction to the validation of bioanalytical methods The presentation of the different test categories An overview of method properties that are checked during validation A description of the requirements that must be fulfilled for a method validation Examples of companies neglecting their duty to validate and the consequences thereof

v

Abstract

This book describes the validation of bioanalytical methods. This is an important element in the quality control of drugs, especially biological drugs. During the manufacture of medicinal products, samples are analysed for compliance with quality requirements at many points in the very complex manufacturing process. Depending on the nature of the drug, bioanalytical methods are also used. These must provide trustworthy results so that no false conclusions are drawn when evaluating the results. The trustworthiness is checked by validation.

vii

Contents

1 Introduction ����������������������������������������������������������������������������������������������� 1 2 Method Categories������������������������������������������������������������������������������������� 5 3 Validation Parameters�������������������������������������������������������������������������������13 3.1 Accuracy���������������������������������������������������������������������������������������������13 3.2 Precision���������������������������������������������������������������������������������������������16 3.3 Further Validation Parameters�������������������������������������������������������������18 4 Validation  Environment and Validation Planning�����������������������������������23 4.1 Validation Environment: The Big Difference Between Academic and GMP Laboratories �����������������������������������������������������������������������23 4.2 Validation Planning�����������������������������������������������������������������������������26 5 Validation  of Bioanalytical Methods���������������������������������������������������������29 5.1 Identity: Validation of a PCR Test�������������������������������������������������������29 5.2 Determination of Content: Validation of a Virus Titration�����������������33 5.3 Qualitative Impurities: Validation of a Test for Absence of Microorganisms (Sterility) �����������������������������������������������������������������36 5.4 Quantitative Impurity: Validation of an Endotoxin Assay �����������������37 6 Errors,  Problems and Risks Associated with Insufficient Method Validation ���������������������������������������������������������������������������������������������������39

ix

x

Contents

7 Summary�����������������������������������������������������������������������������������������������������43 What the Reader Can Take Away from this Essential�������������������������������������45 References �����������������������������������������������������������������������������������������������������������47

1

Introduction

Pharmaceuticals represent a heterogeneous group of products that are used, among other things, to treat or prevent diseases. The production of medicinal products takes place in specialised companies under consideration of Good Manufacturing Practice (GMP). The most important European set of rules in this regard, the EU GMP Guide, is a comprehensive work that lays down the “rules of the game” for all quality-relevant processes. This covers everything from the drug manufacturer’s suppliers, the quality of the materials used, the production processes, quality control and other important areas such as how changes are to be made and how activities are to be recorded. A key aspect of drugs is that they must pass extensive laboratory analytical examination before they are released for use and shipped. Many drugs are based on small active substances that consist of a certain number of atoms bonded together and whose composition can be represented relatively easily as a chemical structural formula. These so-called low-­ molecular active substances include, for example, typical antipyretics such as paracetamol or ibuprofen. Many of these drugs are mainly analysed using chemical-­ physical analytical methods. This can be, for example, chromatography, a method for separating and, if necessary, determining the quantity of substances. A comprehensive presentation of bioanalytical methods would require more than 1000 pages, as for example in the excellent textbook Bioanalytik by Lottspeich and Engels, which can be regarded as the bible for students of the life sciences in the field of methodology (Lottspeich and Engels 2012). Nevertheless, we will try to make the essence of bioanalytics understandable in this book. Bioanalytical methods in the narrow sense are used to study biomolecules, i.e. molecules of living organisms. Biomolecules are, for example, proteins, DNA, RNA, carbohydrates or lipids. Bioanalytical in a broader sense means analyzing certain ­properties © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0_1

1

2

1 Introduction

of biological material, i.e. biomolecules, viruses and living cells (see Fig.  1.1). Biological drugs include genetically engineered proteins, DNA molecules, stem cells, antibodies or vaccines based on viruses and bacteria. Depending on the nature of the product, various methods are used for quality control of biological products, such as cell cultures, polymerase chain reaction (PCR), protein biochemical methods, microbiological methods such as nutrient media for the detection of specific microorganisms, etc. However, it would be wrong to think in absolute categories, such as biological medicinal product = bioanalytical methods and chemical medicinal product  =  chemical analytical methods. Depending on the product, classical chemical-physical methods are also used in the quality control of biological medicinal products, e.g. pH-value determinations or determinations of the moisture content, but in most cases bioanalytical methods predominate in the case of biological medicinal products. The entirety of the analytical methods of a medicinal product should provide evidence that the product fulfils all the quality requirements specified in the marketing authorisation and is thus effective and safe.

Proteins

Viruses, cells

DNA

Bioanalytics

RNA

Lipids

Carbonhydrates

Fig. 1.1  Objects of investigation in bioanalysis. (Source: Created by Patric Vogel)

1 Introduction

3

In order to be able to make this statement at all, one must be able to trust the results of these bioanalytical methods. Many trust results or displays intuitively and do not even question whether what they see really corresponds to the truth. There are many situations in everyday life where we trust information. This may be the time on the illuminated sign of a pharmacy, the speedometer display of our car while driving, the receipt of a letter from the doctor with the results of laboratory tests or graphics on the current development of unemployment figures in the news. We trust that the information is accurate. But how do we know? The answer is simple: we don’t. Some of the data can be confirmed by doing our own research or comparing it to other measurements. Other data is not so easy to check. Errors can always occur, with graphics but also with electronic displays, and do not always have to have serious consequences. An incorrect time, e.g. 14.45 instead of 15.00 on the pharmacy clock may have no effect at all when we are on our way to the beach. But the same mistake can have significant consequences if, for example, we are on our way to a job interview and, based on the wrong time, decide to have a leisurely coffee at the bakery next door before presenting at the company and thus unfortunately arrive too late. In the worst case scenario, the job slips through our fingers as we are deemed unreliable or unpunctual. A similar scenario arises with bioanalytical methods used in the GMP area. If we want to make statements on the basis of results from bioanalytical methods as to whether a drug meets the quality requirements, we must first be able to trust the results and for this we need validation. Validation is the process of proving that the method is fit for its intended purpose, i.e. that it produces reliable results. The validation of analytical methods is required in the European area by the Guide to Good Manufacturing Practice of the European Union (Chap. 6 of the EU GMP Guide). This chapter deals with the tasks of quality control, which also includes the performance of laboratory analytical checks (EudraLex 2014). While the GMP guideline only requires validation in one sentence, without formulating how this is to be done, there are other international or national guidelines that describe how method validation has to be done, including which properties are to be checked. These general guidelines are not formulated down to the last detail and allow room for interpretation or different approaches. For certain long-­ established methods, there are even pharmacopoeial monographs that describe in detail how the suitability should be checked in one’s own laboratory. Compliance with the requirements for validations, like any other area of Good Manufacturing Practice, is checked at regular intervals by the competent authorities as part of GMP regulatory inspections. It is not uncommon for deficiencies to be identified which relate to the lack of or inadequate validation of analytical methods (see Chap. 6). This makes it clear how important the validation of methods is for GMP-­ compliant quality control.

4

1 Introduction

This book describes the validation of bioanalytical methods. In addition to general (quality of reagents, training of staff, equipment qualification) and formal aspects (work instructions, validation protocols), validation is illustrated for each test category using an example. Two important terms are the method category and so-called validation parameters. There are a large number of different bioanalytical methods with which properties, or quality attributes, are determined. In most cases, the methods can be assigned to one of a few method categories. For example, if the amount of insulin is determined, this method falls into the category of content. In addition, it could also be analyzed whether certain undesirable substances are contained in the insulin preparation. Such methods fall into the category of impurities. In Chap. 2 the method categories are explained with examples. The method categories are based on the properties of the analysed sample. In addition, validation parameters are properties of the method, e.g. how accurately it can measure, how much individual measurements vary, what minimum amount of the analysed biomolecule can still be measured, etc. In Chap. 3 we will get to know all relevant method properties. The examples chosen in this book include: • • • •

Content: Validation of a virus titration of a live vaccine Identity: Validation of a polymerase chain reaction Contamination: Validation of a test for the absence of microorganisms (sterility) Contamination: Validation of an endotoxin test

2

Method Categories

Validation means that a certain sequence of experiments or laboratory tests verifies whether results of the method used are reliable. These tests are different and serve to examine certain aspects. It also depends on the purpose of a method which properties, so-called validation parameters, are to be checked. The important thing is that there are only a handful of method categories and validation parameters. These are listed in the ICH Q2 (R1) guideline along with recommendations on how to conduct the tests (number of measurements, etc.) (ICH 2005). There are four designated method categories: Content, Identity, Impurities qualitative and Impurities quantitative. This represents a rough classification that can be applied to many analytical methods. If, for example, 12 different analytical methods are used on a biological medicinal product for release, most of the methods used can be classified in one of the four broad categories. However, there are also individual methods that cannot be classified in one of these categories (see Fig. 2.1). The content is the quantity determination of the active substance in the drug, that is, the component that produces the actual effect. The active substance can be almost anything, from small chemical substances such as paracetamol, to proteins or DNA, to living organisms. While chemical methods are used for paracetamol, as mentioned before, the others would fall into the field of bioanalytics. For example, insulin is a protein (growth hormone) whose amount can be determined using a chromatographic method called high-performance liquid chromatography (HPLC) (Moses et al. 2019). This HPLC is a complex system consisting of several components (see Fig. 2.2A). In this example, the analysis involves adding the insulin fluid to a closed system. The insulin is transported with a liquid buffer through the tubing or metal lines until it hits a column. This column contains a surface to which insulin can attach, i.e. bind. However, this binding is changeable © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0_2

5

6

2  Method Categories

Content

Other methods

Identity

Category

Contamination (quantitative)

Contamination (qualitative)

Fig. 2.1  Method categories. (Source: Created by Patric Vogel)

(reversible) and depends on environmental factors. Other molecules present in the sample migrate through the column. By changing the properties (e.g. concentration, pH) of the liquid constantly pumped through the system, one achieves that the insulin at a certain point passes back into the liquid, i.e. detaches from the column surface. As this happens quite abruptly, all the insulin is released within a short time and is carried along with the liquid flow. At the end of the system, the fluid passes through a detector that constantly measures what is passing through the line. The more insulin that was in the sample, the stronger the signal. This signal is plotted as a curve (the amount passing the detector increases until it reaches a maximum and then decreases again). Since different substances have different binding strengths to the column surface, they dissolve at different times (= substance separation) and can be represented as different curves, so-called peaks (see Fig. 2.2B). The results of a test run are also called a chromatogram (see Fig. 2.2). Depending on the drug, additional signals can be generated, e.g. by other components (also impurities) contained in the test sample, which, however, can be distinguished from the active substance if the method is well adjusted (see Fig. 2.2C). The signal is converted into the actual unit, e.g. amount/millilitre, by means of a so-called

2  Method Categories

7

Fig. 2.2  Chromatography system with theoretical representation of substance separation and an example chromatogram for a substance separation. (Source: Image A (left): Adobe Stock, file no.: 5773410; Image B (top right): Adobe Stock, file no.: 181689009; Image C (bottom right): Adobe Stock, file no.: 82895709, modified afterwards; Images licensed by Patric Vogel)

standard curve (different amounts of a known insulin preparation). In principle, there are usually several methodical possibilities to determine the content of ­biomolecules or cells. In the case of insulin, for example, the HPLC described is even a substitute that has only become established in recent years. Before (and still today), an in vivo assay in live rabbits was/is the standard test to measure the biological activity of insulin for batch release (Hamza 2018). In recent years, as regulatory agencies have increasingly pushed to replace these ethically questionable live animal drug tests with in vitro laboratory methods, an HPLC method was developed in this example (Hack et al. 2016). DNA is a biomolecule whose quantity can be determined, for example, by means of absorption measurement. For this purpose, a solution containing DNA is placed in a so-called photometer. This device sends light of a certain wavelength through the liquid. DNA has the property to absorb this light, i.e. to intercept it. Behind the liquid there is a detector that catches the light that passes through. The more DNA there is in the solution, the more light is absorbed and the less light

8

2  Method Categories

reaches the detector. The photometer then calculates for itself (using a stored standard curve) how much DNA is present in the solution. This method is very simple, but is not always, at least not alone, suitable as a true content determination, as it does not capture the state form of the DNA, which is important for the activity of many biological DNA-based products. Whole organisms or living cells are perhaps the most complex form of an active substance. These are used, for example, in vaccines (e.g. the vaccine against tuberculosis called Bacille-Calmette-Guérin (BCG)). The content of this live vaccine, which is based on attenuated live tuberculosis bacteria, can be determined by microbiological methods (Milstien and Gibson 1990). This involves diluting the vaccine suspension and placing the liquid on special culture media (round plastic dishes with a high rim containing a fairly solid jelly-like substance containing all the nutrients that bacteria need to grow). Since bacteria multiply continuously by dividing into two, a bacterial colony visible to the naked eye forms at each site where a bacteria was present after a few weeks (other bacteria form a visible colony after only one day) (Fig. 2.3). The colonies are counted and, taking into account the dilution, it is calculated back how many bacteria there are in the vaccine solution, e.g. per ml. However, biological drugs are now more versatile. For example, stem cells are whole, viable cells that are the active substance. But there are also various therapeutic products under development, such as attenuated bacteria that have been genetically modified and are to be used to fight certain tumours. There are different modes of action, ranging from bacteria that damage tumour cells themselves, to bacteria that produce tumour-associated antigens, to bacteria that carry only DNA molecules. In the latter case, the genetic information is translated into proteins by the immune cells themselves after they have “eaten” the bacteria (Xiong et  al. 2010; Sedighi et  al. 2019). These products are thought to damage the tumor or stimulate a strong immune response that targets the tumor. These examples show how the methodology used is highly dependent on the nature of the test sample being analyzed. In some cases, a simple count of these bacteria as a content determination would no longer be sufficient to measure the content or biological activity, as the bacteria are in some cases only a vehicle to smuggle the active substance (recombinant DNA molecule) into body cells. Another category is identity. Here it is examined whether it is really the drug that is supposed to be contained according to the package insert and label. Almost the entire range of bioanalytical methods can be used as identity tests, depending on the properties of the drug. For example, a drug containing a protein as active

2  Method Categories

9

Fig. 2.3  Principle of detection and quantity determination of viable bacteria. (Source: Image A (top): Adobe Stock, file no.: 329919276; Image B (bottom): Adobe Stock, file no.: 264200390; images licensed by Patric Vogel)

substance can be tested for identity by means of the so-called Western blot. In this bioanalytical method, proteins are separated according to their size in a gel that is exposed to an electric field. Small proteins migrate quickly through the gel, while larger proteins need more time to migrate through the pore-containing gel. This separates the proteins according to their size. The proteins are then transferred to a solid membrane and detected on this surface using antibodies. The antibodies bind specifically only to these proteins. The final result is a coloured area on the membrane.

10

2  Method Categories

Polymerase chain reaction (PCR) may also be used as an identity test if the medicinal product contains nucleic acids (e.g. viral vaccines, live cells or nucleic acid-based vaccines). In this process, a small region from the gene sequence of the product is amplified and, after separation (using a similar principle to the s­ eparation of proteins), made visible under UV light by electrophoresis using fluorescent substances that bind to DNA. The other two major method categories are qualitative (or limit) and quantitative impurities. This is about everything that should not be contained in the drug or only in small amounts of foreign substances, as the patients could be harmed by this. For example, additives that were necessary during the manufacture of the drug but should not be contained in the final product. On the other hand, all forms of harmful pathogens, such as viruses, bacteria or fungi. The difference between qualitative and quantitative is whether the quantity is important. For example, viruses are not supposed to be in the final product, so specifying quantity would make little sense. Either they are there (rejection and destruction of the batch) or they are not (quality requirements regarding contamination by viruses met). Other contaminants are not harmful in every concentration. In many cases, authorities (or a pharmacopoeial monograph) define a limit that must not be exceeded. Here it is important to be able to precisely determine the amount (ultimately like the content described above). In the case of a lower limit value (upper limit values or value ranges are also possible), the medicinal product complies with the requirements if the measured value is below the limit value. If the measured value is above the limit value, the batch of the medicinal product does not meet the quality requirements. An example of an impurity quantified by bioanalytical methods is endotoxin. This impurity originates from certain bacteria and can cause fever or worse reactions in higher amounts. Especially if the drug was produced by means of propagation in bacteria, these may be present in the product despite purification. While most methods fall into one of these four broad method categories, there are other methods, often chemical-physical in nature, that are somehow different. For example, the pH of a liquid drug might be very important in obtaining activity. The pH does not measure the amount of active substance or the identity. However, the pH also does not logically pass as an impurity. Nevertheless, it may be a property of the drug that is very important in maintaining the activity of the protein whose activity may be damaged at extreme pH. Methods that do not fit into the method categories are often standard methods whose performance is described in a pharmacopoeial monograph. These are methods that have been shown to be successful over years or decades in various laboratories or reference laboratories.

2  Method Categories

11

By being included in the pharmacopoeia, the respective method acquires an almost elite reference status (= pharmacopoeia method) and is thus generally considered reliable. However, it would be careless to generalize this to every laboratory. For this reason, these methods should also be tested for suitability in one’s own laboratory, since insufficient laboratory practice alone can lead to inaccurate or false ­results. For example, the pH value is measured with so-called pH meters. A poorly calibrated pH meter or sloppy work when cleaning the pH electrode can also falsify a standard method.

3

Validation Parameters

There are several properties of a method that are important. These properties include trueness, specificity, precision, linearity, working range, detection limit and limit (ICH 2005). These different properties are also called validation parameters. The question of which of these properties should be investigated experimentally by laboratory tests depends on the method category (see Table  3.1). Guidelines for method validation also exist in other geographical regions such as the USA (FDA 2015) and even specifically for bioanalytical methods (FDA 2018).

3.1 Accuracy Accuracy is the ability of the method to determine the nominally correct value of a sample. If we remember the examples in Chap. 1, we very often trust indications or results in everyday life without questioning their correctness. Let us take as a simple example the measurement of body size at the doctor’s office. Often there is a rail on the wall with markings, similar to a tape measure. You stand with your back against the rail and a movable horizontal stop is pushed onto your head so that you can then read off your height. Normally we assume that the measurement gives a correct result. But what if at some point, unobserved, a child pulled on the rail and it shifted a little, say 3 cm? Such a comparably small error might not be noticed for a long time. The result, however, would be that every body measurement subsequently taken would be erroneous. This is called systematic measurement error. Similar effects can also be present with bioanalytical methods. Depending on the setting and calibration of the instruments used and the complexity of the method itself, deviations from the correct value may occur. Nobody tears at it as in the © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0_3

13

14

3  Validation Parameters

Table 3.1  Validation parameters to be determined for different method categories Parameter Accuracy Specificity Precision (repeatability) Precision (intermediate precision) Linearity Workspace Detection limit Quantification limit

Content x x x x

Identity – x – –

x x – –

– – – –

Impurities qualitative (limit) – x –

Impurities quantitative x x x x

– – x –

x x – x

e­ xample above, but there may be other reasons that lead to a systematic measurement deviation. The reasons can be varied, from interference from other components in the drug affecting the measurement to incorrectly adjusted measuring instruments. In the context of validation for quantitative tests, i.e. tests designed to determine the amount of a substance (active substance or impurity), accuracy must be determined. Let’s take insulin as an example. Insulin is actually expressed in the international units IU (derived from the English term international unit). To simplify matters, we measure insulin here in weight/volume, e.g. microgram/ml (microgram is one thousandth of a gram). We use the chromatography method mentioned in Chap. 3. Our test sample has a quantity (concentration) of 100 μg/ml. But how does the chromatography system actually know what 100 μg/ml is? And how do we know if our test sample really has 100 μg/ml? Reference standards are used for this purpose, if available. These are preparations that are sold by certified bodies. These preparations are well analysed, i.e. one knows that the indicated concentration is correct. This value is, apart from a few exceptions, where the reference substances themselves have quality defects, e.g. due to improper storage, the recognised correct value, which is beyond any doubt. Now the software of the device is “taught” which measuring signal measured by the detector corresponds to which concentration. For this purpose a so-called standard curve is created, i.e. different concentrations (e.g. 200  μg/ml, 150  μg/ml, 100 μg/ml, 50 μg/ml and 10 μg/ml) are measured with the chromatography system and beforehand it is entered which test sample has which concentration. From this, the software of the chromatography system calculates which measurement signal corresponds to which concentration and calculates a standard curve. Subsequently,

3.1 Accuracy

15

unknown test samples that lie within the concentration range can be measured. The system measures the signal strength of the test sample, compares it with the standard curve and determines the result. If we now want to check the correctness of our new method, we measure this reference standard. The result tells us whether the method measures accurately or has a deviation. Many measurements are not absolutely accurate, so depending on the sample and the method, a tolerance range is given, say 99.0–101.0 μg/ml. The tolerance range, also called the acceptance range, in some cases comes from the pharmacopoeia. In other cases, this range is set by the manufacturer, but must also be accepted by regulatory authorities. The decision as to whether the method measures correctly is then evaluated based on compliance with the acceptance range. We measure the reference standard with the method. If the result is e.g. 99.6 μg/ml, the correctness is confirmed, i.e. we obtain a result with our method that corresponds to the expected value (correct value) within an acceptable tolerance range. If the measurement result would be e.g. 101.7 μg/ml, our requirements would not be fulfilled, i.e. the validation parameter accuracy would not be successfully proven. In this case, the method would have to be revised to bring the measurement result closer to the correct value. In practice, trueness is often not determined with one concentration. The ICH guidelines recommend an evaluation over a range of concentrations. In addition, the result will often not be exactly the same if we repeat the measurement a few times (see Sect. 3.2 Precision), this phenomenon concerns the precision of the method. Many bioanalytical methods have an important limitation, especially for the analysis of novel biological drugs. Here, commercially available reference standards are often not available. A company that has been intensively researching a new stem cell therapy under the strictest secrecy for the last few years, and in the course of this has been isolating and multiplying new stem cells, will not easily find a suitable reference standard in a catalogue of official authorities or private certified manufacturers. These are based on decades of experience with certain long-­ term drugs, such as insulin. But even in such cases, there are various possibilities. For example, the manufacturer could produce its own reference standards, e.g. by storing a batch for which efficacy has been proven under constant conditions for a long time. In-house production is required by the Food and Drug Administration (FDA) in the Americas, for example (FDA 2018). In this way, the properties of the material are preserved. The performance of bioanalytical methods could then be such that each manufactured batch is tested in direct comparison with this standard preparation. To meet quality requirements, each manufactured batch must perform as well or better.

16

3  Validation Parameters

3.2 Precision Precision is another property (validation parameter) that must be investigated for quantitative methods. Precision describes how close together the results of successive measurements are. Let us return to the example of body size measurement. Let us assume that the measurement is correct, i.e. has no systematic error. A person with a height of exactly 183 cm is measured 3 times at intervals of 10 min. The results are 182.5 cm, 183 cm and 184.1 cm. Although we have measured the same person, we do not get exactly the same result. During the measurements, the person may have tilted their head slightly differently or may have had a different body tension. This leads to the fact that on average we arrive relatively accurately at the correct height, but the individual measurements show variations. The difference but also the correlation between accuracy and precision is illustrated in Fig. 3.1. If the results of a bioanalytical method give the correct value on average, but show certain differences in the individual measurements, there is a high accuracy despite fluctuations (case A in Fig. 3.1). If the individual measurements are very close together but far from the correct value, the method is highly precise but measures inaccurately (case B in Fig. 3.1). The ideal case is that the method measures both accurately and precisely (case C in Fig. 3.1). The worst case is that the method measures incorrectly on average and also shows large differences in the individual values (case D in Fig. 3.1). Some analysts, especially those with a chemistry background or experience in the analysis of classical drugs, complain about the imprecision and lack of precision of many bioanalytical methods, and yes, there is a grain of truth in this. Many biomolecules, especially when they are part of a more complex structure, cannot be analyzed with the same razor-sharp precision as chemically simple active substances. For example, viruses are very small and simple compared to bacteria, but their complexity, consisting of viral genome, envelope proteins and membranes

High accuracy

High precision

High precision + accuracy

Poor precision + accuracy

Fig. 3.1  Relationship between correctness and precision. (Source: Adobe Stock, File No.: 308832180; Licensed by Patric Vogel)

3.2 Precision

17

and multiplication properties, is a thousand times higher than that of chemically defined substances such as paracetamol. For this reason, the acceptance ranges of many bioanalytical methods are somewhat broader than those of chemical-physical measurement methods. Different levels are distinguished for the parameter precision (see Fig. 3.2). The simplest case is the repeated measurement of the same test sample several times in direct succession in a coherent series of measurements in one test run. This level is called repeatability. All samples have very similar conditions. The scatter that occurs, i.e. the difference between the individual results, represents the lowest possible variability of the method. The measured differences can result, for example, from inaccuracy in pipetting (transferring liquid) of the sample or reagents, but also from slight measurement inaccuracies of the laboratory equipment used or individual factors such as differences in mixing intensity in a manual mixing procedure. If the same test is repeated at a different time, somewhat larger deviations usually occur, since various factors have an influence on the deviations, such as room temperature, humidity, adjustment, but also individual differences in the work between different laboratory analysts. Often individual factors are varied specifically, such as carrying out the test on a different day, by a different person or by using a different laboratory instrument with the same design. The next level is not always

Routine

Reproducibility

Intermediate precision

Repeatability

Fig. 3.2  Levels of precision. (Source: Created by Patric Vogel)

18

3  Validation Parameters

relevant as it relates to test performance in different laboratories. There may be even greater differences in precision here, for example due to the use of laboratory equipment from other manufacturers, etc. A comparison of methods between laboratories in terms of precision is called reproducibility. The last level is routine, i.e. the case where the bioanalytical method is frequently used for analyses and is thus subject to natural variations. It is important that the method has the required level of precision not only under very controlled conditions (all measurements directly one after the other), but also under routine conditions.

3.3 Further Validation Parameters Specificity is the ability of the method to specifically detect only the desired biomolecule without other components that are part of the product and those that could potentially be present contributing to the signal, producing the same signal or falsifying it. Take as an example a genetically engineered protein that is analyzed by a chromatographic separation method such as HPLC. These proteins are often produced by propagation in bacteria in bioreactors called fermentors. After fermentation is complete, the bacteria and their components are removed. Here, bacterial components, e.g. cell wall components, bacterial proteins, RNA, DNA and lipids are found in the biomass. These are largely removed by subsequent purification steps, but usually not completely. When using HPLC, specificity means that the protein can be clearly distinguished from other components, i.e. the signal is not distorted by these interfering factors. The detection limit is the lowest amount of substance that can be detected against a background of method noise. Method noise depends on certain factors of the method. For example, a fluorescence signal is measured during real-time PCR. The more DNA that is formed during real-time PCR, the higher the fluorescent signal. However, even in the absence of DNA, there is a small amount of background fluorescence that is referred to as the background noise of the method. Signals from samples containing DNA can only be detected as true signals if they stand out clearly from this background noise. This background noise is part of every methodology. The detection limit is only determined for methods that are qualitative in nature, i.e. answer the presence of a substance with yes or no (present/absent). Certain impurities (e.g. contaminating microorganisms) must not be present in the product, the question of how much is irrelevant here. The quantification limit is determined for impurities in quantitative methods. This is the smallest amount, of impurity, that can just be correctly and accurately determined. This in turn can be compared to the actual speed indicator on cars,

3.3 Further Validation Parameters

19

even if the speed is not an impurity. In many cars, the speed is shown not only on the speedometer needle but also on the display of the on-board computer. However, there are cars where this electronic display is only shown from a speed of 30 km/h, for example. In this case, the quantification limit of the speed of the electronic display is 30 km/h, although the speedometer needle can still display low speeds (hopefully correctly) and thus has a low quantification limit. It would be the same with the measurement of height. Provided the rail with the measurement scale doesn’t start until 60 cm. Then the quantification limit of the height measurement would be 60 cm. Children who are shorter can stand under it, but we cannot measure how tall they are. The quantification limit can be the same as the detection limit or higher than the detection limit, but it can never be lower. The linearity describes the range in which the measurement signal is proportional to the sample concentration. No matter what the product, the content values of the test samples fluctuate because often not every batch manufactured has the exact same content. This means that we have to measure samples with different concentrations and naturally want the results to show us these differences. Impurities can show even greater variations in quantity. However, it should not be automatically assumed that bioanalytical methods always measure linearly (i.e. the more in, the higher the signal). There are cases where at lower and higher concentrations the signal does not increase, although the sample concentration differs (curved curve in Fig. 3.3). A good example to illustrate this is a test called ELISA (abbreviation of Enzyme-­linked immunosorbent assay). This test is also frequently used, for example, when we have our blood taken by a doctor to test for certain infectious 16 14 12

Signal

10 8

Increase sample quantity, but constant it signal

6 4 2 0

0

10

20

30

40

50

Sample quantity

Fig. 3.3  Linear and non-linear relationships between concentration and signal. (Source: Created by Patric Vogel)

20

3  Validation Parameters

diseases. We take as an example a so-called sandwich ELISA, i.e. the test starts with antibodies that specifically bind our product. The antibodies are on a solid phase. Next we add our product, e.g. a protein, which specifically binds to the antibodies. The measurement is completed later when we add another antibody, which in turn docks to the other side of our protein and is bound to an enzyme that causes a color reaction. The more protein in the sample, the stronger the colour development, the intensity of which is read out with a laboratory instrument. In our example we have 5 binding sites for protein molecules. If our test sample contains 2 protein molecules, they bind and cause a signal. If our test sample has 4 protein molecules, the resulting signal is twice as high as proportional to the quantity. However, if there are 5 or more protein molecules, all binding sites are occupied. When measuring a test sample with 10 protein molecules, the signal would be as high as with 5 molecules. The measurement is so-called saturated and the signal is no longer proportional to the sample quantity (Fig.  3.4). This means that the

Addition of the sample

Binding of the sample

Scenario A: Measurement signal proportional to sample quantity

Scenario B: Measurement signal not proportional to sample nquantity

Fig. 3.4  Example of saturation effects in ELISA. (Source: Created by Patric Vogel)

3.3 Further Validation Parameters

21

measurement in these ranges is no longer reliable and would actually give us misleading results. For this reason, the linear measuring range must be determined experimentally for quantitative methods. The working range results from the other parameters. It lies in the linear range, is limited upwards by the highest sample concentration that can be measured accurately and precisely. Downwards, the working range is limited by the quantification limit. In the example of body height, if the bar goes up to 210 cm, the working range would be 60–210  cm (measurement of persons below 60  cm and above 210 cm is not possible in our example).

4

Validation Environment and Validation Planning

4.1 Validation Environment: The Big Difference Between Academic and GMP Laboratories If you are interested in scientific articles in journals, you will often find titles like “Validation of …”. Here scientists have developed a method for the analysis of certain samples, checked it for suitability and publish their data. Basically, anyone, whether a university researcher or a GMP specialist, can check the properties of the method mentioned in Chap. 3. Then surely validation should have the same importance? Far from it! A GMP-compliant validation goes far beyond the performance of a limited number of tests. The decisive difference lies in the environment. It is this environment that ensures that the results obtained during validation are sustainable and, above all, long-lasting. This environment includes certain factors that are indispensable in a quality management system: • • • • • • • • • •

Adequate premises Training system Supplier qualification Review and approval of reagents for analyses Equipment qualification Risk analyses Document management system Deviation system Change control Establishment of acceptance criteria

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0_4

23

24

4  Validation Environment and Validation Planning

The quality of the laboratory reagents used must be checked. Here, the verification does not necessarily have to be carried out via many laboratory tests, provided that the reagents are purchased from trustworthy suppliers who have also established a quality management system and carry out the analyses themselves, which are required, for example, in the pharmacopoeia. This is documented on so-called Certificates of Analysis (CoA), which are supplied with the product. Laboratory equipment is needed to perform analytical methods. There is a wide range of different laboratory equipment and supplies that are used for quality control experiments, e.g. pipettes (to be able to transfer exact quantities of liquids), shakers (to shake solutions or temper them at a certain temperature), centrifuges (devices that rotate substances at high speed to achieve separation between components), incubators (for e.g. cell cultures or culture media for growing and detecting microorganisms), but also, for example, chromatographic equipment (for the separation and detection of types of molecules) or machines for the polymerase chain reaction (PCR). There may be dozens or hundreds of devices, depending on the size of the quality control laboratory. Laboratory equipment may not be used for analytical tests without prior verification, called qualification. Qualification is a multi-stage process in which the suitability of the equipment is demonstrated and documented. A variety of documents are created during this review, called qualification documents. Depending on the device, the review can include various aspects. In addition to safety aspects for the personnel, the functions of the laboratory equipment are checked here, among other things. A qualification usually consists of the following stages (Elroy 2018): • Design qualification (DQ) → Definition of requirements for the device • Installation qualification (IQ)  →  Installation and check for compliance with requirements • Functional qualification (OQ) → Testing of device function (e.g. temperature control, light measurement) for compliance with requirements • Performance qualification (PQ)  →  Test whether device meets requirements even under routine conditions (e.g. high measuring frequency, real samples) After successful qualification, laboratory devices are released for use (e.g. with a label), so that e.g. laboratory personnel can immediately recognize that this device may be used.

4.1 Validation Environment: The Big Difference Between Academic and GMP…

25

Clearly understandable Standard Operating Procedures (SOPs) for performing the method must be available. SOPs should clearly describe the objective, have a complete material list and details on the preparation of e.g. buffers, reagents and the test procedure. Bioanalytical methods are usually very complex, so that in some cases the test performance can take up to several weeks and consist of hundreds of steps. The description must be complete enough so that everyone can follow the individual steps until the final result is obtained. Example: “Stock solution XY is diluted five times 1:10 with buffer”. The experienced laboratory technician will know what is meant by this and how to do it. However, this description leaves questions unanswered. Which buffer, in which volumes (1 ml and 9 ml or 10 ml in 90 ml) and in which vessels? The description should minimize sources of error and differences in execution by different people as much as possible. In addition, further quality systems are established in the pharmaceutical company which, together with the aspects already mentioned, represent the environmental corset. This includes the handling of changes and unplanned deviations. If something is changed or runs differently than planned due to errors, the influence on the validation must be evaluated. In the case of complex bioanalytical methods, the mere exchange of a laboratory reagent can turn the results upside down. The systems mentioned help to avoid stumbling carelessly into problems that can influence the assessment of product quality or patient safety. Once all the prerequisites of the environment have been created, a plan is the be-all and end-all. Only those who have a plan can define criteria on the basis of which the success of the validation can be evaluated. If only the results are looked at, one is significantly influenced in the evaluation. If you do not define acceptance criteria, you will not think about whether the deviation from the target value found is okay until the results are available. In such a case, one naturally tends to intuitively push the boundaries. Of course, the validation plan must be trained just like all other procedures. The personnel must be trained in the use of the premises, equipment and methods in a documented manner. The validation itself consists of a sequence of specific tests. It is important that the validation activities are adequately documented, i.e. it is possible to trace what was done, when, where, how and by whom. The minimum documentation requirements are also described in the GMP Guide (EudraLex 2011). A general principle in the GMP environment is that if something was not written down, it did not happen (Patel and Chotai 2011). An important aspect that is involved here is not only quality control personnel but also, for example, quality assurance personnel. The latter is primarily concerned with ensuring that everything is “above board”, i.e. that the specifications are ad-

26

4  Validation Environment and Validation Planning

hered to (Vogel 2020b). The methods used must also be regularly kept up to date with the current “state of the art”, a requirement that does not come from the GMP guidelines but, in Germany for example, directly from the law in the form of the Arzneimittel- und Wirkstoffherstellungsverordnung (AMWHV) (Blasius 2014).

4.2 Validation Planning The introduction of a new bioanalytical method always begins with a development phase (see Fig. 4.1). The first step is to look for a suitable methodology that fulfils the desired purpose. If, for example, a company is working on a vaccine against the novel disease COVID-19, it must inevitably also establish the methodology for quality control. In the case of content determination, for example, a suitable technology naturally depends on the type of vaccine being developed. However, other factors also play a role, such as the laboratory equipment already available and expertise in various bioanalytical areas. In areas where many similar products already exist, there is expertise which methodology is generally well suited. For novel products without a long history, there are numerous options from which to choose a suitable method. For example, for mRNA vaccines, there are quite a few different analytical methods for each property (content, identity, etc.), from which the one that seems most appropriate is selected (Poveda et al. 2019). Once a methodology is selected, a final method is determined during method development and pre-tested for various properties to be relatively certain that the method can successfully pass method validation. The introduction of a method also includes risk analyses. As part of these risk analyses, important factors are analyzed and the risk is assessed. These serve to identify critical factors in order to minimise the risk that at a later stage an undetected source of error will negatively influence the validity of the method. This includes, for example, evaluating the influence of various reagents, but may include other aspects in anticipation. For example, the decision may ultimately be against the most appropriate methodology if the risk analysis reveals that the reagents are from a source that does not comply with Good Manufacturing Practice. Another reason could be that the supplier of laboratory reagents has, for example, a monopoly for certain important reagents that cannot be obtained elsewhere and that the supplier is also unable to promise to offer the reagents in the long term. What does one want to do with a bioanalytical method that can no longer be used for batch analysis after the product has been approved because the supplier has ceased production of its laboratory reagents? However, risk analysis is not a one-time activity when the method is introduced. The authorities expect risk analyses to be

4.2 Validation Planning

27

Development phase Choice of methodology Establishment of the method in the laboratory Tests for preliminary assessment of whether method fit for the purpose intended

Risk analysis

Validation plan Determination of who does what, when, with what and how Definition of success criteria

Execution and documentation

Validation report Presentation of the results Assessment of the conformity of the results with the performance criteria

Fig. 4.1  Flow chart of method development and validation. (Source: Created by Patric Vogel)

carried out over the entire life cycle of a method (from development, through the entire period of use, to phase-out). This does not have to be every month, but, for example, when an important aspect changes. This could be the change of the method to a new laboratory device, a change of the supplier of the reagents or technological advances because of which the own method should be replaced by a better method. In addition to risk analyses, this naturally also requires other procedures such as change control, a procedure designed to prevent a change from having a negative impact on product quality (Vogel 2020a).

28

4  Validation Environment and Validation Planning

The validation phase itself begins with a validation plan, a document that contains all the essential points. It defines who does what, when, where and with what. This means that in addition to the general information, such as the responsibilities of personnel and the qualification status of the devices, the exact definition of the test material to be used is also included. In addition, all laboratory analytical tests for determining the relevant validation parameters (e.g. accuracy, precision, linearity, etc.) are described. An important aspect is the definition of acceptance criteria for the individual validation parameters. This is another important difference between a GMP laboratory and the non-regulated academic laboratory. Only those who record the acceptance criteria in advance in a valid document will later have an objective decision-making basis for evaluating the results. Unless a valid plan exists, some tend to adjust the success criteria based on the results obtained, as mentioned earlier. Before starting the analysis, a so-called standard operating procedure (SOP) is of course required, which describes very precisely the sampling, sample preparation, the analysis itself and the evaluation of the results. The next step is the execution of the experiments by trained personnel as well as the accurate documentation of all performed experiments as well as the results. Based on these records, a validation report (in some cases also combined as a protocol and report) is prepared, which summarizes the results as well as an assessment of whether all success criteria have been met. The validation report is prepared, reviewed and approved in the same way as the validation plan. After successful completion of a method validation, the freshly validated method may then be used for routine testing of the relevant medicinal product, provided, of course, that it has also been accepted by the competent regulatory authorities at the time of marketing authorisation and that marketing authorisation has been granted. The information on the method and the validation must already be included in the marketing authorisation dossier.

5

Validation of Bioanalytical Methods

We now come to examples of validation of bioanalytical methods. We will look at one example from each of the test categories. There are many other possibilities besides the presented methods, it depends, as already mentioned, on the nature of the drug, the advantages of methodologies and the expertise of the manufacturer.

5.1 Identity: Validation of a PCR Test A popular identity test for biological drugs containing nucleic acid (gene therapy products, vaccines based on e.g. viruses, DNA vaccines) is the polymerase chain reaction (PCR). The basic principle is always the same, namely a product that differs from others in its gene sequence. For example, a dangerous influenza virus could be genetically engineered to be attenuated so that it can safely be used as a vaccine against influenza. In this hypothetical example, we cut out a specific region from the genome of the dangerous virus. The missing gene sequence causes the virus to be unable to replicate as quickly. After this vaccine receives approval, each batch produced (= number of vials produced in one processing run) must be tested for identity in addition to the other properties (content, impurities, etc.). For this purpose we use a PCR.  Strictly speaking, it is a subtype of PCR, namely RT-­ PCR. The addition RT stands for “reverse transcription”. The nucleic acid of influenza viruses consists of RNA, which cannot be amplified in normal PCR. For this purpose, reverse transcription is used to first transcribe the RNA into DNA and then to amplify a specific part of the gene sequence in the PCR. The nature of our hypothetical vaccine and the principle of our identity test is shown in Fig. 5.1. A small region was excised from the viral genome using genetic © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0_5

29

30

5  Validation of Bioanalytical Methods Region removed from gene sequence for production of attenuated vaccine

P

P

Gene sequence of original virus P

P

Section of the gene sequence from the original virus

Gene sequence vaccine P

P

PCR products 400 bp

200 bp

Fig. 5.1  Schematic representation of the genetic manipulation of an influenza vaccine and the results in an RT-PCR (P: primer binding sites). (Source: Created by Patric Vogel)

engineering methods. The gene sequence, which we now amplify in RT-PCR, is located around the excised region. The so-called primer binding sites define which region from the viral genome is amplified. In this case, the primer binding sites are upstream and downstream of the excised region. This results in a small PCR fragment when our vaccine is analysed by RT-PCR, whereas RT-PCR results in a larger fragment in the dangerous wild type (= initial strain). The size of PCR fragments is given in base pairs (bp) and represents the number of individual nucleotides (building blocks of DNA) in that DNA strand. In our example, the PCR fragment would be 200 bp when analyzing our vaccine and 400 bp when analyzing the original virus (Fig. 5.1). The PCR products are then visualized, for example, with the aid of a classical separation technique, electrophoresis. Here, the PCR products are applied to a thin gel that is itself immersed in a liquid in a tank (Fig. 5.2A). An electric field is then applied. Since DNA is always negatively charged, the DNA migrates through the gel to the + pole during electrophoresis (Fig. 5.2B). Since the gel consists of a pore structure of many small passages, the PCR fragments have resistance. Small fragments move rapidly through the gel, while larger fragments have difficulty winding through the gel. This achieves separation between small and large fragments and completes the electrophoresis. The gel is placed on a UV stage (Fig. 5.2C). Since the gel has been previously incubated with a fluorescent substance that binds to

5.1 Identity: Validation of a PCR Test

31

Fig. 5.2  Exemplary representation of the steps of an electrophoresis for the separation of DNA (A: Loading of the gel; B: Electrophoresis; C: Placing the gel on the UV table; D: Visualization of DNA by excitation with UV light). (Source: Image A (top left): Adobe Stock, file no.: 275562527; Image B (top right): Adobe Stock, file no.: 208256696; Image C (bottom left): Adobe Stock, file no.: 213648287; Image D (bottom right): Adobe Stock, file no.: 208256769; Images licensed by Patric Vogel)

DNA and fluoresces when excited with UV light, the areas containing PCR ­fragments can be seen after the UV light is turned on (Fig. 5.2D). The size of the fragments can be estimated using a DNA standard with different bands, which is also applied to the gel. According to Table 3.1, only one validation parameter is important for an identity test, the specificity. To prove that our identity test is specific, we need to measure a number of samples in the validation. If we want to be sure that a positive RT-PCR signal (the PCR fragment) is generated only when our vaccine is present, we need to analyze different test samples where no PCR fragment is expected, such as water, buffers, or the final formulation solution of the vaccine that contains certain components, such as substances to improve the stability of the vaccine. Besides, the dangerous virus and other strains should also be checked to show that

32

5  Validation of Bioanalytical Methods

Fig. 5.3  Example of the result of a PCR validation (A and B: Two different DNA size standards; C and D: PCR fragment from the original virus; E and F: PCR fragment from the vaccine; G and H: Negative sample without nucleic acid). (Source: Adobe Stock, File no.: 342360041, modified afterwards; Licensed by Patric Vogel)

our vaccine can be distinguished from this strain by the different size of PCR fragment. As long as the desired fragment shows up only in our sample but not in the others, the specificity is successfully demonstrated (Fig. 5.3). If we consider the specifications from Table 3.1, one could now get the impression that the validation can be completed by a single laboratory test in one day. This is not completely wrong, but the validation is not the first experimental test. During the development of this method, additional tests are made that allow an assessment of the method performance. For example, the ICH Q 2 (R1) guideline already requires that during development the robustness of the method should be analyzed (ICH 2005). Robustness is about showing that the results are not affected by small variations in the test conditions (e.g. temperature, concentration of reagents).

5.2 Determination of Content: Validation of a Virus Titration

33

5.2 Determination of Content: Validation of a Virus Titration The potency of a drug is an important property because it allows statements to be made about its efficacy. The efficacy has been proven in clinical studies and is often determined after approval in the form of an in  vitro content determination (called potency assay in English). There are also other drugs in which this test is performed in vivo in living animals (see the example of insulin in Chap. 2). As an example, let us take a viral titration of a live vaccine. A virus titration is the determination of the number of infectious virus particles. In our example, the product is a rotavirus vaccine. This vaccine protects against a common intestinal infection, especially in children. There are different methods to determine the virus titre, e.g. in chicken eggs or in cell cultures. Our example is about a PFU titration, which is mentioned by the World Health Organization (WHO) as a possible virus titration for the quality control of rotavirus vaccines (WHO 2007). The abbreviation PFU comes from the English and means plaque-forming units. Here, the virus is detected by forming a circular “plaque” (area in which cells are destroyed). This method can be used whenever the virus causes visible cell damage (Lambert et al. 2008), which is the case with rotaviruses. Brief description of the method: The vaccine is diluted and a portion is added to cell cultures. These cell cultures are shortly thereafter overlaid with a semi-solid medium, stained after a few days and checked to see how many circular virus foci are present. We further assume that we have analyzed various influencing factors, i.e. robustness, in the development and are satisfied. We further assume that we have previously discussed the acceptance criteria shown in Table 5.1 with the competent authority. For validation, we now perform different experiments. The accuracy is tested with 3 different concentrations (105, 106,5, 108 PFU/ml) each with 3 replicates (measurements), for linearity 5 different concentrations are required. So we do the concentrations 106 and 107 PFU/ml in one experiment (precision + linearity). The precision is determined by 2 extra experiments with 6 repetitions each (6 measurements), which are done on different days. As soon as the virus titration has been carried out and the results have been documented on the protocols, we start the evaluation. For the evaluation of validations of quantitative methods, we unfortunately cannot avoid a little laboratory statistics. There are 3 types of calculations that are necessary here. The accuracy is calculated and given as a percentage by comparing the measured value and the true value. For a result of 9 with a correct value of 10,

34

5  Validation of Bioanalytical Methods

Table 5.1  Results of validation and evaluation Parameter Accuracy

Acceptance criteria (target) Content 70–130% of the correct value

Specificity Repeatability Intermediate precision Linearity

– ≤30% ≤30%

Range

Coefficient of determination R2 ≥ 0.90 All concentrations that are accurate, precise and linear

Result (actual) 105 PFU = 91% 106,5 PFU = 97% 108 PFU = 88% – 15% 21%

Evaluation Passed

R2 = 0.96

Passed

105–108 PFU/ml accurate, precise and linear

Passed

– Passed Passed

this would be 9/10*100 = 90%. This is also often described as recovery rate (how much of what should come out I find in the analysis). The variability of the results is expressed by the standard deviation (SD). The standard deviation is the average distance of all results from the mean. If the results are 9, 10 and 11, the standard deviation is 1. As far as the dispersion of the individual values is higher, e.g. 7, 10 and 13, the SD is also larger, in this case 3. The relative standard deviation, also called coefficient of variation (CV), is often used. This value expresses what percentage the SD is of the mean. In the first case (8, 9, and 10, mean = 10 and SD = 1), the CV would be 10% (the SD is one-tenth of the mean). In the second case (7, 10, 14; mean = 10 and SD = 3), the CV would be 30% (the SD is 30% of the mean). For precision tests, the ICH guideline also recommends the calculation of confidence intervals, but we omit this for ease of presentation. Another statistical method we need is linear regression. We need this to be able to evaluate the linearity. We measured 5 sample concentrations (virus titer 105, 106, 106,5, 107 and 108). The expected values (correct value) are plotted on the x-axis against the result (measured viral titer) on the y-axis. The figure includes the log-­ transformed viral titer for ease of visualization (106 = 6 log10, 105 = 5 log10, etc.) The log-transform is a commonly used means of evaluating viral titer (Lock et al. 2010; McFarland et al. 2018). We recognize that the points do not all lie exactly on a straight line. This is because our measurements do not match 100% with the correct value. This is not a big deal, as we have allowed ourselves a tolerance range in which the results are allowed to fluctuate. To assess linearity, we use linear regression to plot a compensation line. This is based on the individual data points and represents the best fit for the true relationship between the correct value and the

5.2 Determination of Content: Validation of a Virus Titration

35

measurement result. Linear regression also allows the relationship to be quantified. This is expressed by the coefficient of determination R2 which can take values between 0 and 1. If all data points are close to the line, R2 takes a high value (close to 1), if the measurement results vary more, R2 is lower. In our case R2  =  0.96 (Fig. 5.4). The results of the accuracy experiment all meet the success criterion of 70– 130%. The precision (repeatability and intermediate precision) also meet the criteria. The linearity is also successfully passed. This results in a working range (range in which the method measures accurately, precisely and linearly) of 105–108 PFU/ ml (Table 5.1). If you look closely, you will see that the specificity was not tested. Unfortunately, this is not possible with PFU titrations. There are many viruses that can form these circular plaques, i.e. we cannot tell from the plaques that they were caused by the rotavirus vaccine. The ICH guideline has a special clause for these cases. If the specificity cannot be determined, it is possible to compensate this parameter by another method. In addition to the content, we also have to do an identity test. If we have a specific identity test that detects the viruses in the cell culture with e.g. antibodies, it can be concluded during quality control that the result is due to the

8,5 R2 = 0,9587

Result virus titre (log PFU/m)

8 7,5 7 6,5 6 5,5 5 4,5

4,5

5

5,5

6 6.5 7 7,5 Expected virus titer (log PFU/ml)

8

8,5

Fig. 5.4  Results of PFU titration to evaluate trueness and linearity. (Source: Created by Patric Vogel)

36

5  Validation of Bioanalytical Methods

vaccine. I.e. we have compensated the weakness of the PFU titration by considering the identity test. Finally, the results are summarized and evaluated in a report. If all validation parameters have been successfully demonstrated, the method validation has been passed.

5.3 Qualitative Impurities: Validation of a Test for Absence of Microorganisms (Sterility) Depending on the manufacturing technology, a wide variety of impurities can be present in pharmaceuticals. These also include bacteria and fungi. Not all medicinal products have to be absolutely free of these microorganisms, but all medicinal products that are injected into the body (so-called parenterally administered). Because so many products must be tested for sterility (absence of bacteria and fungi), a highly standardized methodology has developed over time that has its own pharmacopoeial monograph. This monograph 2.6.1 specifies in minute detail how the testing must be done, including composition of nutrient solutions, volume, temperatures for incubation, duration of incubation, and amount of units from the batch that must be tested. So unlike many other monographs, there is no room for leeway here. In summary, the test sample is placed in two different nutrient solutions and incubated for 14 days at two different temperatures. One temperature tends to promote the growth of bacteria, the other the growth of fungi. At the end (and in between) the solutions are visually inspected. If there are microorganisms in the test sample, a turbidity (caused by the microorganisms) of the nutrient solution develops. If the test sample is sterile, the nutrient solutions remain clear. If turbidity is present, an aliquot is taken, subcultured (spread on solid culture media and incubated again) and then finally determined, e.g. by biochemical methods, which microorganism species is present. The monograph also has passages describing how the method must be checked for suitability for each product. Even though the monograph does not use the term validation, this suitability test corresponds to a GMP-compliant assurance of reliability. Defined reference germs (bacterial and fungal species) must be used in this test. As can be seen from Table 3.1 (see Chap. 3), two validation parameters are important for qualitative impurities, the specificity and the detection limit. These are investigated by adding the prescribed reference bacteria to the product (so-­ called spiked) in small quantities prior to analysis. A direct comparison of the

5.4 Quantitative Impurity: Validation of an Endotoxin Assay

37

p­ roduct and the product spiked with the reference germs is then carried out. The product must be negative. The spiked samples should be positive. Subsequently, each of the reference germs is identified by biochemical methods. In this way, both validation parameters are detected. Small amounts of the reference germs are used (detection limit) and proof is provided that all reference germs can be detected in the specific product matrix. This verification is important because drugs have very different compositions. Some have certain additives such as oils, fats or other substances that interfere with the detection of the microorganisms. Therefore, this demonstrates that all reference germs can be specifically detected in the specific product matrix. If this detection is successful, the so-called product-specific validation has been achieved and the method can subsequently be used to detect the absence of microorganisms in batches of the product for quality control purposes.

5.4 Quantitative Impurity: Validation of an Endotoxin Assay The last method presented here, the test for endotoxins, is an example of a quantitative test for impurities. Certain drugs are produced using bacteria, e.g. certain therapeutic proteins or DNA vaccines. Some bacteria produce so-called endotoxins, which are combined biomolecules that can cause, for example, fever or even anaphylactic shock. During production, after the product has been multiplied, various purification steps take place to remove these and other bacterial components. In these cases (and others), a test for endotoxins must be performed during quality control. There is also a pharmacopoeial monograph for this, Ph Eur 2.6.14, but it is much less restrictive than the sterility test described earlier. The endotoxin monograph allows a variety of different methodological approaches and measurement principles to detect endotoxins. If an own test is established or a normal kit (kits are compilations of reagents necessary for the test performance) is used, all validation parameters according to Table 3.1, i.e. accuracy, precision, specificity, linearity, limit of quantification and working range must be examined. The validation is then very similar to the description of the content determination in Sect. 5.2. However, there is also a highly standardized test kit, called Endosafe®, which is available in two variants (for single samples and up to 5 samples) and which is certified in the American region for use with pharmaceutical products. This means that the method of production and the test kit manufacturer’s own quality control have been tested and certified by the competent authority, the American Food and

38

5  Validation of Bioanalytical Methods

Drug Administration (FDA). In this test, the test sample is filled into 4 wells of a test cassette, which is inserted into an analyser. The liquid is sucked in. At the end of two channels, the test sample is measured for endotoxin (double determination). In the other two channels, the test sample in the channel is mixed with an endotoxin reference standard. All reactions are measured in the detector at the end, producing a color if endotoxin is present. Again, the measurement signal is proportional to the amount of endotoxin. The high confidence in this test results from the fact that the manufacturer demonstrates specific validation parameters for each individual kit lot. This includes linearity, trueness, precision, quantification limit and working range. The type and number of tests differ from the general specifications. For example, linearity is demonstrated by 3 concentrations with 10 replicates each instead of 5 concentrations. Besides, precision is tested by duplicate determinations, instead of six-fold determinations. This shows that the approach mentioned in guideline ICH Q2 (R1) is not the only possible one to demonstrate the validity of a method. Nevertheless, the manufacturer recommends performing a product-specific validation. This is not surprising, since when comparing the mentioned validation parameters with the parameters mentioned in Table 3.1 for quantitative impurities, it is noticeable that specificity is missing. This is provided in one’s own laboratory by measuring the product, since the test cassettes directly contain a reference standard with which the influence of one’s own product matrix on the measurement result can be checked.

6

Errors, Problems and Risks Associated with Insufficient Method Validation

The performance and validation of analytical, including bioanalytical, methods is an obligation if you want to be a drug manufacturer. It is just one small of many building blocks used to ensure the quality of drugs. But it is an important building block that can cause a lot of trouble if the obligation is not fulfilled. This is often seen during so-called GMP inspections, where official representatives visit the pharmaceutical company and check for compliance with Good Manufacturing Practice. Since these inspections only last a few days, not all processes in the company can be illuminated. But again, it’s only a matter of time until the next time you see them, and GMP inspectors know from your records what they checked last time and may look at new procedures next time. Insufficient or missing method validation is not infrequently part of the complaints. In the US, deficiencies are dealt with very openly. The American GMP authority, the Food and Drug Administration (FDA), regularly publishes so-called “warning letters”. These are publicly accessible letters to the manufacturer in question in which serious deficiencies are described. In the worst case, i.e. if the manufacturer does not react appropriately and, among other things, remediate the deficiencies, these can lead to withdrawal of approval. The deficiencies range from cases where the method validation is incomplete (i.e. only available once, but in need of improvement), to cases where no validation is available, to cases where disagreeable test results are repeated as often as desired or no analytical test is performed at all. For example, during an inspection of a Canadian manufacturer, it was found that a product had not been tested for impurities. Impurities are an important aspect of medicinal products. Depending on the medicinal product, this may refer to certain biomolecules that are an indispensable part of the manufacturing process © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0_6

39

40

6  Errors, Problems and Risks Associated with Insufficient Method Validation

and cannot be completely separated, but must not exceed a certain limit. It may also refer to living organisms, such as bacteria or fungi. In the case of the Canadian company, the lack of one of these tests was particularly critical. The product could theoretically (due to the manufacturing process) contain a certain carcinogenic, or cancer-causing, impurity. The absence of this test is serious because it does not allow an assessment to be made as to whether recipients of the drug may be harmed by taking it. For example, in this case, the FDA required that retain samples of all circulating batches be retrospectively tested for this impurity (ECA 2020). Another example is a French company which, as a contract laboratory, carries out GMP-relevant analyses for the products of other manufacturers. During the inspection by the inspectors, it was found that some analytical methods used were not validated. In addition, some test results that did not meet expectations were invalidated, i.e. invalidated and repeated without sufficient justification (ECA 2018). The unsubstantiated repetition of test results is a dangerous game. For example, a drug might be specified to have a level between 90 and 110 μg/ml (a microgram is one thousandth of a gram). Insofar as the first analysis shows 113 μg/ ml, this result does not meet the requirements. If this were to be confirmed, the batch would have to be thrown in the bin, i.e. destroyed. If the analysis is invalidated because it is assumed to be an analytical error for no real reason and the repeat analysis measures 114 μg/ml, the batch would again not meet the quality requirements. If this result is again invalidated and repeated and then a value of 109 μg/ml is obtained and the batch is released based on this result, this is referred to as testing-into-compliance (Vogel 2020c). This term means an unpleasant result is repeated until the result happens to meet the quality requirements. Economic interests are more in the foreground here, as the destruction of a manufactured batch is very costly. Basically, it is patient safety that suffers, which is why inspectors take sharp action against such offences. In 2013, a batch of an insulin preparation in syringe form was recalled. The background to the recall was that, after release, a strongly fluctuating content of 50–150% of the target dosage was found in some insulin prefabricated pens. This could have led to under- and overdoses during use. Incorrect operation of a valve during filling was cited as the cause (Apotheke ADHOC 2013). The question is why this quality defect was not noticed during quality control if a validated method was used? Quality control is never complete. A certain amount of the batch is tested for certain properties, such as content. It is important that the samples taken are representative of the batch, i.e. using samples, for example, only from the beginning of a filling is dangerous as errors can also happen during a filling process, such as contamination of a filling needle with bacteria. This would then potentially only affect units after this incident. That is why the sampling should be

6  Errors, Problems and Risks Associated with Insufficient Method Validation

41

representative. Nevertheless, something can be overlooked during quality control if only very few samples of the batch are affected. Overall, recent analyses show that inadequate analytics or weaknesses related to method validation were among the top 5 deficiencies FDA inspectors listed in “warning letters” during the period October 2018-September 2019 (GMP Navigator 2019). In some cases, problems with the validation of bioanalytical methods can even be one of the reasons for the insolvency of smaller companies, as was the case with a German biotech company in 2008. The company wanted to launch a cancer therapy on the market (ÄrzteZeitung 2008), but failed in the validation of a tumour test because a parameter required by the regulatory authority could not be verified. As a result, the company filed for insolvency due to insolvency (DGAP 2008).

7

Summary

The validation of bioanalytical methods is an important element in pharmaceutical operations that serves to verify product quality and thus patient safety. Method validation ensures that assessments and decisions are made on the basis of reliable results. Analytical methods cannot make products better, but reliable results help to separate the wheat from the chaff, i.e. to identify those batches with real quality deficiencies in the large flood of good quality batches. However, validation does not mean that analytical results are “beyond all doubt”. Undetected errors can creep in at any time, especially with the sometimes very complex bioanalytical methods. Therefore, in the case of unexpected results of bioanalytical methods, it is necessary to get to the bottom of the cause of the error and understand whether it was really a product deficiency (correct result) or a laboratory error. In doing so, economic reasons must not be in the foreground, but rather the scientific question of what caused this result. In addition, many things can change in the life cycle of a bioanalytical. These include, in part, highly complex substances that are required for the performance. Slight changes in composition, purity or other factors can have a significant impact on the results. For this reason, method validation must be embedded in a tight corset of other quality systems, including ensuring the proper functioning of laboratory equipment, change control (no one is allowed to change substances or work instructions just like that), checking the quality of materials used, but also the competence of employees, which is ensured by regular training. In addition, a continuous or interval-­ based review of method performance is important, for example in the form of regular revalidations. Thus, proof is repeatedly provided that the quality systems function well and that the bioanalytical methods are valid, i.e. reliable.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0_7

43

44

7 Summary

The field of method validation ultimately also represents a professional field. In smaller companies, there are certainly all-rounders who have to devote themselves to many different tasks and for whom method validation is only one of several tasks. However, depending on the size of the pharmaceutical company, there are also people, called validation specialists, who exclusively deal with the validation of methods and do nothing else all day long. The validation of bioanalytical methods is by no means a boring field. With the large number of different methods, the question of how to provide proof of reliability always arises. Not all methods can be validated according to the same pattern. That is why a little “brain power” always has to be put into planning. This field also needs to keep up with other biomedical advances. For example, ATMPs (advanced therapeutic medicinal products) are a rapidly growing branch of medicines, to which the GMP Guide has dedicated a single section (EudraLex 2017). The more complex the product, the more difficult the necessary bioanalysis becomes in some cases. But here too, there are tricks to approach validation in a reasonable way (Viganò et al. 2018).

 hat the Reader Can Take Away from this W Essential

• For the analysis of the properties of biological drugs, bioanalytical methods are often necessary, with which biomolecules, but also cells and viruses can be investigated • Validation under GMP conditions is very different from validation in the academic field and requires a highly regulated environment. • In the validation of bioanalytical methods, a series of experimental trials is used to demonstrate that the results are reliable • Successful validation is a basic prerequisite for the use of the method for testing medicinal products • Deficiencies regarding the validation status are not uncommon, but must be remedied if the deficiency is identified during regular GMP inspections.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0

45

References

Apotheke ADHOC (2013) Novo Nordisk: Fehler beim Abfüllen. https://m.apotheke-­ adhoc.de/nc/nachrichten/detail/pharmazie/insulin-­rueckruf-­novo-­nordisk-­fehler-­beim-­ abfuellen/. Accessed on: 30.08.2020. ÄrzteZeitung (2008) Biotech-Firma LipoNova stellt Insolvenzantrag. https://www.aerztezeitung.de/Wirtschaft/Biotech-­Firma-­LipoNova-­stellt-­Insolvenzantrag-­356298.html. Accessed on: 30.08.2020. DGAP (2008) LipoNova AG: Vorläufiger Insolvenzverwalter bestellt. Keine Einigung mit Aufsichtsbehörde über Tumorassay für Reniale. https://www.dgap.de/dgap/News/ corporate/liponova-­vorlaeufiger-­i nsolvenzverwalter-­b estellt-­k eine-­e inigung-­m it-­ aufsichtsbehoerde-­ueber-­tumorassay-­fuer-­reniale/?newsID=271040. Accessed on: 29.08.2020. Blasius H (2014) Arzneimittelherstellung Mehr als Produktion und Qualitätskontrolle. DAZ. online https://www.deutsche-­apotheker-­zeitung.de/daz-­az/2014/daz-­30-­2014/arzneimittelherstellung. Accessed on: 10.08.2020. ECA (2018) Warning letter due to non-validated analytical methods. https://www.gmp-­ compliance.org/gmp-­news/warning-­letter-­due-­to-­non-­validated-­analytical-­methods. Accessed on: 30.06.2022 ECA (2020) FDA warning letter for Canadian drugmaker  – HPLC and data integrity in the focus. https://www.gmp-­compliance.org/gmp-­news/fda-­warning-­letter-­for-­canadian-­ drugmaker-­hplc-­and-­data-­integrity-­in-­the-­focus. Accessed on: 18.08.2020. Elroy J (2018) An introduction to analytical instrument qualification & validation – Meeting FDA expectations. https://www.pharmaceuticalonline.com/doc/an-­introduction-­to-­ analytical-­instrument-­qualification-­validation-­meeting-­fda-­expectations-­0001. Accessed on: 10.08.2020. EudraLex (2011) Volume 4 – Good Manufacturing Practice (GMP) guidelines Part I, chapter 4: Documentation. https://ec.europa.eu/health/sites/health/files/files/eudralex/vol-­ 4/2014-­11_vol4_chapter_6.pdf. Accessed on: 12.08.2020.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 P. U. B. Vogel, Validation of bioanalytical methods, essentials, https://doi.org/10.1007/978-3-658-38913-0

47

48

References

EudraLex (2014) Volume 4 – Good Manufacturing Practice (GMP) guidelines Part I, chapter 6: Quality control. https://ec.europa.eu/health/sites/health/files/files/eudralex/vol-­ 4/2014-­11_vol4_chapter_6.pdf. Accessed on: 15.08.2020. EudraLex (2017) Volume 4  – Good Manufacturing Practice (GMP) guidelines Part IV. https://ec.europa.eu/health/sites/health/files/files/eudralex/vol-­4/2017_11_22_guidelines_gmp_for_atmps.pdf. Accessed on: 25.08.2020. FDA (2015) Analytical procedures and methods validation for drugs and biologics guidance for industry. https://www.fda.gov/files/drugs/published/Analytical-­Procedures-­and-­ Methods-­Validation-­for-­Drugs-­and-­Biologics.pdf. Accessed on: 18.08.2020. FDA (2018) Bioanalytical Method Validation Guidance for industry. https://www.fda.gov/ files/drugs/published/Bioanalytical-­Method-­Validation-­Guidance-­for-­Industry.pdf. Accessed on: 29.08.2020. GMP Navigator (2019) Chargenfreigabe ohne Gehaltsbestimmung und andere GMP-­ Verstöße  – ein Blick auf die FDA warning letters der letzten Monate. https://www. gmp-­navigator.com/gmp-­news/chargenfreigabe-­ohne-­gehaltsbestimmung-­und-­andere-­ gmp-­verstoesse-­ein-­blick-­auf-­die-­fda-­warning-­letters-­der-­letzten-­monate. Accessed on: 18.08.2020. Hack R, Rüggeberg S, Schneider L (2016) The replacement of the rabbit blood sugar bioidentity assay by an in vitro test for batch release of insulin glargine drug substance. https://ec.europa.eu/environment/chemicals/lab_animals/3r/pdf/rudiger_hack.pdf. Accessed on: 17.08.2020. Hamza AA (2018) Biological assay of insulin: an old problem re-discovered. Journal of Advanced Technologies in Endocrinology Research. https://chembiopublishers.com/ JATER/JATER180002.pdf. Accessed on: 25.08.2020. ICH (2005) Validation of analytical procedures: text and methodology. Q2(R1). https:// database.ich.org/sites/default/files/Q2%28R1%29%20Guideline.pdf. Accessed on: 05.08.2020. Lock M, McGorray S, Auricchio A et al. (2010) Characterization of a reocombinant adeno-­ associated virus type 2 reference standard material. Hum Gene Ther 21:1273–1285; https://doi.org/10.1089/hum.2009.223. McFarland EJ, Karron RA, Muresan P et  al. (2018) Live-attenuated respiratory syncytial virus vaccine candidate with deletion of RNA synthesis regulatory protein M2-2 is highly immunogenic in children. J Infect Dis 217:1347–1355. https://doi.org/10.1093/infdis/ jiy040 Milstien JB, Gibson JJ (1990) Quality control of BCG vaccine by WHO: a review of factors that may influence vaccine effectiveness and safety. Bull World Health Organ 68:93–108. Moses A, Bierrum J, Hach M et al. (2019) Concentrations of intact insulin concurs with FDA and EMA standards when measured by HPLC in different parts of the distribution cold chain. J Diabetes Sci Technol 13:55–59. https://doi.org/10.1177/1932296818783783 Lambert F, Jacomy H, Marceau G, Talbot PJ (2008) Titration of human coronaviruses, HCoV-229E and HCoV-OC43, by an indirect immunoperioxidase assay. Methods Mol Biol 454:93–102. https://doi.org/10.1007/978-­1-­59745-­181-­9_8 Lottspeich F, Engels JW (2012) Bioanalytik. Wiesbaden: Springer VS. Patel KT, Chotai NP (2011) Documentation and records: Harmonized GMP requirements. J Young Pharm 3:138–150. https://doi.org/10.4103/0975-­1483.80303

References

49

Poveda C, Biter AB, Bottazzi ME et  al. (2019) Establishing preferred product characterization for the evaluation of RNA vaccine antigens. Vaccines (Basel) 7:131. https://doi. org/10.3390/vaccines7040131 Sedighi M, Zahedi Bialvaei A, Hamblin MR et al. (2019) Therapeutic bacteria to combat cancer; current advances, challenges, and opportunities. Cancer Med 8:3167–3181. https://doi.org/10.1002/cam4.2148 Viganò M, Budelli S, Lavazza C et al. (2018) Tips and tricks for validation of quality control analytical methods in Good Manufacturing Practice mesenchymal stromal cell production. Stem Cell Int 2018:3038565. https://doi.org/10.1155/2018/3038565 Vogel PUB (2020a) COVID-19: Suche nach einem Impfstoff. Wiesbaden: Springer VS. Vogel PUB (2020b) Qualitätskontrolle von Impfstoffen. Wiesbaden: Springer VS. Vogel PUB (2020c) GMP-documentation – From blank page to GMP-compliance. 1st edition ebook Amazon kindle. ASIN: B088557MFW. WHO (2007) Annex 3 guidelines to assure the quality, safety and efficacy of live attenuated rotavirus vaccines (oral). WHO Technical Report Series No 941. https://www.who.int/ biologicals/publications/trs/areas/vaccines/rotavirus/Annex%203%20rotavirus%20vaccines.pdf?ua=1. Accessed on: 09.08.2020. Xiong G, Husseiny MI, Song L et al. (2010) Novel cancer vaccine based on genes of Salmonella pathogenicity island 2. Int J Cancer 126:2622–2634. https://doi.org/10.1002/ ijc.24957