Animal Models for Human Cancer E-Book
129,99 €
Mehr erfahren.
- Herausgeber: Wiley-VCH
- Kategorie: Wissenschaft und neue Technologien
- Serie: Methods and Principles in Medicinal Chemistry
- Sprache: Englisch
Based on results from the past ten years, this ready reference systematically describes how to prepare, carry out, and evaluate animal studies for cancer therapies, addressing the widely recognized lack of reliable and reproducible results.
Following a short historical introduction and a discussion of the ethics surrounding animal experiments, the book describes correct study design as well as the handling and housing of animals. It then goes on to describe the animal models available for different cancer types, from natural cancer models in mice and dogs to humanized animals. An evaluation of previously unpublished long-term data from the Swiss canine and feline cancer registry is also included. The final part of the book reviews the lessons learned over the last decade on how to interpret data from animal studies for improving human therapy and gives recommendations for future drug development.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 575
Veröffentlichungsjahr: 2016
Ähnliche
Table of Contents
Cover
Previous Volumes of this Series
Title Page
Copyright
List of Contributors
Preface
Reference
A Personal Foreword
References
Chapter 1: Introduction
1.1 Animal Models in Biomedical Research
1.2 Animals in the Drug Development Process: Historic Background
1.3 Problems with Translation of Animal Data to the Clinic
1.4 Animal Studies in Anti-cancer Drug Development
1.5 Toward Relevant Animal Data
1.6 Aim of the Book
References
Chapter 2: Ethical Aspects of the Use of Animals in Translational Research
2.1 Introduction
2.2 Today's R&D Environment
2.3 “Do No Harm”: the Essential Dilemma of Animal Research
2.4 Man and Animals in Philosophy: an Overview of Key Concepts
2.5 Conclusions: Solving the Dilemma
References
Chapter 3: Study Design
3.1 Introduction
3.2 Design Principles
3.3 Experimental Design
3.4 Conclusion
References
Chapter 4: Improving External Validity of Experimental Animal Data
4.1 Introduction
4.2 Variation in the Laboratory
4.3 The Fallacies
4.4 Future Perspectives: an Experimental Strategy Integrating Adaptive Plasticity and Fundamental Methodology
References
Chapter 5: How to End Selective Reporting in Animal Research
5.1 Introduction
5.2 Definition and Different Manifestations of Reporting Bias
5.3 Magnitude of Reporting Biases
5.4 Consequences
5.5 Causes of Reporting Bias
5.6 Solutions
References
Chapter 6: A Comprehensive Overview of Mouse Models in Oncology
6.1 Introduction
6.2 Xenograft Mouse Models
6.3 Genetically Engineered Mouse Models
6.4 Applications for GEMMs in Compound Development
6.5 Humanized Mouse Models: toward a More Predictive Preclinical Mouse Model
6.6 Conclusions: Potentials, Limitations, and Future Directions for Mouse Models in Cancer Drug Development
References
Chapter 7: Mouse Models of Advanced Spontaneous Metastasis for Experimental Therapeutics
7.1 Mouse Tumor Models in Cancer Research
7.2 The Evolution of Metronomic Chemotherapy
7.3 Development of Highly Aggressive and Spontaneously Metastatic Breast Cancer Models
7.4 Is There Any Evidence that Models of Advanced Metastatic Disease Have the Potential to Improve Predicting Future Outcomes of a Given Therapy in Patients?
7.5 Metronomic Chemotherapy Evaluation in Preclinical Metastasis Models
7.6 Experimental Therapeutics Using Metastatic Her-2 Positive Breast Cancer Xenografts Models
7.7 Examples of Recently Developed Orthotopic Models of Human Cancers
7.8 Factors that Can Affect the Usefulness of Preclinical Models in Evaluating New Therapies
7.9 Monitoring Metastatic Disease Progression in Preclinical Models
7.10 Alternative Preclinical Models: PDX and GEMMs
7.11 Recommendations for the Evaluation of Anti-cancer Drugs Using Preclinical Models
7.12 Summary
References
Chapter 8: Spontaneous Animal Tumor Models
8.1 Introduction
8.2 Advantages of Spontaneous Canine/Feline Cancer Registries
8.3 Spontaneous Animal Tumors as Suitable Models for Human Cancers
8.4 The Swiss Canine/Feline Cancer Registry 1955–2008
References
Chapter 9: Dog Models of Naturally Occurring Cancer
9.1 Introduction
9.2 Advantages of Spontaneous Cancer Models in Dogs
9.3 Dog Cancer Models
9.4 Preclinical and Veterinary Translational Investigations in Dogs with Cancer
9.5 Necessary Developments for Realizing the Potential of Canine Models
9.6 Key Challenges and Recommendations for Using Canine Models
9.7 Conclusions
References
Chapter 10: Improving Preclinical Cancer Models: Lessons from Human and Canine Clinical Trials of Metronomic Chemotherapy
10.1 Introduction: Low-dose Metronomic Chemotherapy
10.2 Clinical Trials of Metronomic Chemotherapy
10.3 Veterinary Metronomic Trials in Pet Dogs with Cancer
10.4 Lessons Learned from Clinical Trials: Improving the Predictability of Preclinical Models
10.5 Conclusions
Acknowledgements
References
Index
End User License Agreement
Pages
xi
xii
xiii
xiv
xv
xvi
xvii
xviii
xix
xx
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
Guide
Cover
Table of Contents
Preface
Begin Reading
List of Illustrations
Chapter 3: Study Design
Figure 3.1 The sample size needed per group as a function of the standardized effect size (SES, or signal–noise ratio) for 80% (triangles) or 90% (circles) power assuming a two-sample -test with a two-sided alternative hypothesis and a significance level of 0.05. In the text example the SES was 1.5, so such an experiment would need about eight mice/group for an 80% power or about 10 mice/group for a 90% power.
Chapter 4: Improving External Validity of Experimental Animal Data
Figure 4.1 (a) The concept of standardization. First, by reducing variation in the data, standardization within experiments increases test sensitivity. Because higher test sensitivity allows a reduction of sample size, standardization is also promoted for ethical reasons with a view to reducing animal use. Second, standardization between experiments is believed to reduce between-experiment variation, thereby improving the comparability and reproducibility of results between laboratories. (b) Example of a standardized laboratory environment.
Figure 4.2 Blocking and systematic variation in animal experiments. (a) Example of an experimental design consisting of three blocks (red, blue, green) to compare three treatments (A, B, C). Within each block, experimental subjects are as homogeneous as possible, but between blocks there are differences. (b) Example of a block design for an experiment with mouse genotype (−/−, +/−, +/+) as treatment and three different experimental blocks (red, blue, green) to introduce systematic experimental variation. Blocks (e.g., cages) may differ in, for example, age of animals and/or housing condition (with or without shelter, nesting material or running wheel (RW)). In principle, each block might represent a slightly different experimental setting. Including block as factor in the statistical analysis controls for the environmental variation between the blocks and increases the generalizability of the results without inflating sample size.
Figure 4.3 Hypothetical integrative approach combining heterogenization of experimental variables and adaptive plasticity. Example of an experimental design aimed at addressing the effects of a novel compound. We propose combining heterogenization of rearing conditions (e.g., non-challenging vs. challenging neonatal environments) with phenotypic plasticity considerations: in our proposal, subjects reared in a non-challenging environment should be tested under non-challenging adult conditions (home-cage automated testing) and subjects reared in a challenging environment should be tested under challenging adult conditions (traditional testing involving removal from the cage and testing in unfamiliar environments).
Chapter 5: How to End Selective Reporting in Animal Research
Figure 5.1 Forest plot showing that contradiction of two trial results in terms of statistical significance is fully compatible with exact agreement of their estimated effect size. Here the trials by Smith and Jones both measured a preventive treatment effect of 0.778 relative to control. A huge sample size difference explains the seeming contradiction when a rigid significance testing paradigm is applied.
Figure 5.2 Flow chart depicting a system designed to reduce research waste and reporting bias. Note that animalresearch.gov is a non-existing website used to illustrate preregistration of all animal studies. Its name was inspired by clinicaltrials.gov, the US-based website for preregistration of (randomized) clinical trials.
Chapter 7: Mouse Models of Advanced Spontaneous Metastasis for Experimental Therapeutics
Figure 7.1 (a) Schematic of a series of experiment testing metronomic vinblastine chemotherapy on human tumor cell lines growing in immunodeficient mice as either primary tumors of metastases. In some experiments we noted that primary tumors would respond poorly to a given therapy, such as metronomic vinblastine (b), that would effectively cure mice that had the same tumor cell line growing as established metastatic disease [8].
Figure 7.2 Selection of metastatically aggressive subpopulations of tumor cells often requires rounds of
in vivo
selection. For example, a primary tumor of the human melanoma cell line WM239A was established subdermally (orthotopically) in SCID mice, allowed to grow and establish itself, and then was surgically removed. Months later the mice showed evidence of lung metastases—from which the subline 113/6-4L was isolated. In some studies, these cells are re-implanted into a new host for an additional round of selection. A similar approach using orthotopic implantation (into the mammary fat pad) was also used to derive, the highly metastatic 231/LM2-4 variant of the human breast cancer cell line MDA-MB-231, and the H2N.met2.hCG metastatic variant was derived from Her-2 expressing MDA-MB-231 cells. H2N.met2.hCG were then transfected to express and secrete the β-subunit of human chorionic gonadotropin (hCG) protein, which can be used as a surrogate molecular marker of disease burden.
Figure 7.3 Schematic of the selection of H2N.met2.hCG metastatic variants in immunodeficient SCID mice.
Figure 7.4 A comparison of the effects of daily sunitinib (SU) therapy in separate experiments on the growth of orthotopic primary breast cancers (a) versus advanced metastatic disease, after the primary tumor was resected (at Day 20). Therapy started on Day 41 (b). The tumor line used for the study was a variant (called
LM2-4
) of the MDA-MB-231 human breast cancer line, isolated by Munoz
et al.
[12] having aggressive spontaneous metastatic potential from established but resected primary (orthotopic) tumors.
Figure 7.5 A disparity between primary tumors and metastasis in terms of response to therapy was observed with the human breast cancer model 231/LM2-4 treated with the combination of metronomic daily cyclophosphamide and the 5-fluorouracil oral prodrug, UFT (tegafur + uracil). Primary tumor growth curves (a) showed no anti-tumor effect by combining UFT with CTX, and did not foreshadow the remarkable survival results obtained when treating mice with established spontaneous metastases of the same line treated with the identical UFT/CTX doublet metronomic therapy (b).
Figure 7.6 (a,b) Differential response of advanced metastases and primary tumors to therapy. Contrasting outcomes when H2N.met2.hCG tumor-bearing mice are treated with trastuzumab monotherapy, either as localized single primary tumors or established multiple spontaneous visceral metastases. Trastuzumab monotherapy (red) inhibits the growth of tumors compared with controls (black) but had no appreciable impact on the growth of advanced visceral metastases (as assessed by the hCG secreted by the tumor cells, which was subsequently measured in the mouse urine).
Chapter 8: Spontaneous Animal Tumor Models
Figure 8.1 Tumor location and diagnoses.
n
= number of tumors found in a location; % = proportion of the location compared with the total of locations. Figure in and around the slices = relative proportion of tumor diagnoses/location. Tumor diagnoses lower than 1% were added into “Other tumors”; locations lower than 1% and unclassified location were not listed. With the exception of “male sexual organs” the listed locations are not sex specific.
Figure 8.2 The risk (OR) of developing a tumor by sex and castration status subclassified by examination methods.
Figure 8.3 Risk to develop a malignant tumor vs. no tumor – pure breeds compared to crossbreeds. Observations:
n
= 90 085.
Figure 8.4 Geographic distribution of canine patients with tumors and ratio of benign to malignant tumors in Switzerland.
Figure 8.5 Number of samples of the most common tumor diagnoses an their malignancy.
Figure 8.6 Risk (OR) of different breeds to develop a tumor compared with European shorthair cats.
Figure 8.7 The risk (OR) of cats developing a tumor according to their sex or castration status, subclassified by examination methods.
Figure 8.8 Tumor location and diagnoses in feline patients.
n
= number of tumors found in a location; % = proportion of the location compared to the total of locations. Figure in and around the slices = relative proportion of tumor diagnoses/location.
Figure 8.9 Geographic distribution of feline patients with tumors and ratio of benign to malignant tumors in Switzerland.
Chapter 9: Dog Models of Naturally Occurring Cancer
Figure 9.1 Popular breeds and the percent of deaths from cancer. On the Y-axis, breeds are ordered according to percent of deaths due to cancer. (Data from [24].) Next to breed name on X-axis in parenthesis is the popularity of breed rank-ordered for 2013 according to the AKC (https://www.akc.org/reg/dogreg_stats.cfm). All pictures used from (www.akc.org/breeds/).
Figure 9.2 Estimated new cases and deaths for common cancers in the USA. According to Centers for Disease Control (CDC), cancer is narrowly second only to heart disease for the leading causes of death. In 2011, heart disease claimed 596 577 lives, while cancer took only slightly less at 576 691 lives (http://www.cdc.gov/nchs/fastats/leading-causes-of-death.htm). Coupled with loss of life is also the financial burden. In 2010, cancer prevelance cost have been estimated between 124.5 and 216.6 billion USD (http://www.cancer.org/cancer/cancerbasics/economic-impact-of-cancer, [26]). This suggests that landmark revolutions are needed in treatment and care of cancer patients. No animal model completely recapitulates the humans cancers perfectly, but the naturally occurring cancers in dogs more closely resemble the disease then do other animal models. Data for Figure taken from “
How Common is Cancer,
” http://seer.cancer.gov/statfacts/html/all.html.
Figure 9.3 Number of publications related to dogs and cancer. We performed a search using the PubMed database (http://www.ncbi.nlm.nih.gov/pubmed) for publications related to dogs and cancer. We used the following search terms: “
dog OR dogs OR canine OR dogs AND cancer
.” Years were grouped and average publications for years calculated. From this, we calculated the annual growth rate and total percent change (percent growth rate = percent change/number of years; http://www.miniwebtool.com/percent-growth-ratecalculator/?present_value=391&future_value=773&num=12).
Figure 9.4 Germinal center of DLBCL. Antigen-activated B-cells differentiate into centroblasts that undergo clonal expansion in the dark zone of the germinal center. During proliferation, the process of somatic hypermutation introduces base-pair changes that can lead to changes in the amino acid sequence. Centroblasts then differentiate into centrocytes and move to the light zone, where the modified antigen receptor, with help from other immune cells, is selected for improved binding to the immunizing antigen. Newly generated centrocytes that produce an unfavorable antibody are removed. Cycling of centroblasts and centrocytes between dark and light zones appears to be mediated by a chemokine gradient. Antigen-selected centrocytes eventually differentiate into memory B cells or plasma cells. Centroblasts with genetic alterations that do not undergo apoptosis as expected can become GBC DLBCL. Likewise, plasma cells can become ABC DLBCL. Not shown: Thymic cell that leads to PMBL. Listed are several known malignant transformations. Adapted from [163, 207, 208].
Figure 9.5 Integration of pet dogs with cancer into translational drug development studies. Canine cancer models compliment the use of both conventional preclinical models (mouse, research-bred dog, and non-human primate) and human clinical trials and their inclusion in preclinical and translational studies will facilitate the rapid intermediate evaluation of agents prior to or after early human trials. Translational drug development studies in dogs may answer important questions about a new drug candidate such as toxicity, biological activity, and establish pharmacokinetic pharmacodynamic relationships for an agent before it enters human studies. Importantly, the comparative approach may answer questions that emerge in early phase human trials such as optimized dosing schedules, combination therapies, and the establishment of surrogate biomarkers or molecular imaging endpoints that will inform the evaluation of these agents as they move into later stages of development. Importantly, the totality of information generated from this comparative and integrative approach will likely reduce the late attrition rate of new cancer therapeutics and contribute to the identification of agents most likely to succeed in human clinical trials. Reprinted from [117] with permission from Macmillan Publishers Ltd: Nature Rev Cancer, copyright 2008.
Figure 9.6 Comparative imaging study in dogs with sinonasal tumors investigating the spatiotemporal stability of Copper(II)-diacetyl-bis(N
4
-methylthiosemicarbazone) (Cu-ATSM) and 3′-deoxy-3′-
18
F-fluorothymidine (FLT) positron emission tomography distributions in during radiation therapy. (a) Sagittal positron emission tomography/computed tomography (PET/CT) slices are shown from a dog with sinonasal carcinoma pretreatment (pre) and mid-treatment (mid) with intensity modulated radiation therapy. Cu-ATSM (middle) and FLT (bottom) scans demonstrate stable spatial distributions of both radiotracers during therapy. (b) Voxel-based scatter plots comparing pretreatment (pre) and mid-treatment (mid) Cu-ATSM and FLT standardized uptake value (SUV) distributions and Spearman rank correlation coefficients (upper left) for dogs with nasal carcinoma or sarcoma. Spatial distributions and uptake of dose painting targets Cu-ATSM and FLT remain stable through mid-treatment, regardless of histology. Reprinted from [305] with permission © 2014 Elsevier.
Chapter 10: Improving Preclinical Cancer Models: Lessons from Human and Canine Clinical Trials of Metronomic Chemotherapy
Figure 10.1 (a) Study characteristics, (b) treatment settings, (c) types of cancer studied, and (d) ranges of performance status of eligible patients of 80 metronomic chemotherapy trials in adults recently reviewed by Lien
et al.
[10].
Figure 10.2 Possible different approaches to the discovery of new biomarkers of metronomic chemotherapy. IL-8, interleukin-8; VEGF, vascular endothelial growth factor; and SNPs, single nucleotide polymorphisms.
List of Tables
Chapter 3: Study Design
Table 3.1 The use of a spreadsheet to assign treatments to experimental subjects at random.
Table 3.2 Block randomization.
Chapter 5: How to End Selective Reporting in Animal Research
Table 5.1 Additional administrative steps for editors in a system in which the initial submission contains no results
Chapter 8: Spontaneous Animal Tumor Models
Table 8.1 Canine tumor entities used as models for human tumors
Table 8.2 Feline tumor entities as models for human tumors
Table 8.3 Relative and absolute breed distribution of feline patients with and without tumors
Table 8.4 Comparative numerical ranking of tumors diagnosed in different organs/organ systems of humans, dogs, and cats in Switzerland
Chapter 9: Dog Models of Naturally Occurring Cancer
Table 9.1 Proportional mortality due to tumors/neoplasm over years by breed
Table 9.2 Breed cancer-specific mortality for OS and mammary tumors
Table 9.3 Over-representation of specific cancers in specific breeds
Table 9.4 Comparative aspects of spontaneous cancers in humans and dogs
Chapter 10: Improving Preclinical Cancer Models: Lessons from Human and Canine Clinical Trials of Metronomic Chemotherapy
Table 10.1 Clinical trials of metronomic chemotherapy in pet dogs
Methods and Principles in Medicinal Chemistry
Edited by R. Mannhold, H. Kubinyi, G. Folkers Editorial Board:
H. Buschmann, H. Timmerman, H. van de Waterbeemd, John Bondo Hansen
Previous Volumes of this Series:
Holenz, Jörg (Ed.)
Lead Generation
Methods and Strategies
2016
ISBN: 978-3-527-33329-5Vol.68
Erlanson, Daniel A. / Jahnke, Wolfgang (Eds.)
Fragment-based Drug Discovery
Lessons and Outlook
2015
ISBN: 978-3-527-33775-0Vol. 67
Urbán, László / Patel, Vinod F. / Vaz, Roy J. (Eds.)
Antitargets and Drug Safety
2015
ISBN: 978-3-527-33511-4Vol. 66
Keserü, György M. / Swinney, David C. (Eds.)
Kinetics and Thermodynamics of Drug Binding
2015
ISBN: 978-3-527-33582-4Vol. 65
Pfannkuch, Friedlieb / Suter-Dick, Laura (Eds.)
Predictive Toxicology
From Vision to Reality
2014
ISBN: 978-3-527-33608-1Vol. 64
Kirchmair, Johannes (Ed.)
Drug Metabolism Prediction
2014
ISBN: 978-3-527-33566-4Vol. 63
Vela, José Miguel / Maldonado, Rafael / Hamon, Michel (Eds.)
In vivo Models for Drug Discovery
2014
ISBN: 978-3-527-33328-8Vol. 62
Liras, Spiros / Bell, Andrew S. (Eds.)
Phosphodiesterases and Their Inhibitors
2014
ISBN: 978-3-527-33219-9Vol. 61
Hanessian, Stephen (Ed.)
Natural Products in Medicinal Chemistry
2014
ISBN: 978-3-527-33218-2Vol. 60
Lackey, Karen / Roth, Bruce (Eds.)
Medicinal Chemistry Approaches to Personalized Medicine
2013
ISBN: 978-3-527-33394-3Vol. 59
Edited by Marianne I. Martic-Kehl and P. August Schubiger
Animal Models for Human Cancer
Discovery and Development of Novel Therapeutics
Series Editors
Prof. Dr. Raimund Mannhold
Rosenweg 7
40489 Düsseldorf
Germany
Prof. Dr. Hugo Kubinyi
Donnersbergstr. 9
67256 Weisenheim am Sand
Germany
Prof. Dr. Gerd Folkers
Collegium Helveticum
STW/ETH-Zentrum
Schmelzbergstr. 25
8092 Zürich
Switzerland
Volume Editors
Dr. Marianne I. Martic-Kehl
Collegium Helveticum
STW/ETH-Zentrum
Schmelzbergstr. 25
8092 Zürich
Switzerland
Prof. Dr. P. August Schubiger
Collegium Helveticum
STW/ETH-Zentrum
Schmelzbergstr. 25
8092 Zürich
Switzerland
All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.
Library of Congress Card No.: applied for
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library.
Bibliographic information published by the Deutsche Nationalbibliothek
The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at <http://dnb.d-nb.de>.
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA,
Boschstr. 12, 69469 Weinheim, Germany
All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form–by photoprinting, microfilm, or any other means–nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law.
Print ISBN: 978-3-527-33997-6
ePDF ISBN: 978-3-527-69591-1
ePub ISBN: 978-3-527-69589-8
Mobi ISBN: 978-3-527-69590-4
oBook ISBN: 978-3-527-69588-1
Cover Design Grafik-Design Schulz
List of Contributors
Carlos E. Alvarez
The Ohio State University
Medicine and Veterinary Medicine
700 Children's Drive, W431
Columbus, OH 43205
USA
Karin Blumer
Novartis International AG
Fabrikstr. 6
4002 Basel
Switzerland
Guido Bocci
University of Pisa
Clinical and Experimental Medicine
Via Roma 55
56126 Pisa
Italy
Gianluca Boo
Collegium Helveticum
ETH Zürich
Schmelzbergstr. 25
8092 Zürich
Switzerland
Lex M. Bouter
VU University Medical Center
Department of Epidemiology and Biostatistics
De Boelelaan 1117
1081 HV Amsterdam
The Netherlands
Urban Emmenegger
University of Toronto
Sunnybrook Health Sciences Centre
2075 Bayview Avenue
Toronto, ON M4N3M5
Canada
Joelle M. Fenger
Department of Veterinary Clinical Sciences
The Ohio State University
Veterinary Medical Center
601 Vernon Tharp Street
Columbus, OH 43210
USA
Michael F.W. Festing
University of Leicester
MRC Toxicology Unit
Lancaster Road
Leicester LEI 9HN
UK
Giulio Francia
The University of Texas at El Paso
Department of Biological Sciences,
Border Biomedical Research Center
El Paso, TX 79902
USA
Ramona Graf
Collegium Helveticum
ETH Zürich
Schmelzbergstr. 25
8092 Zürich
Switzerland
Katrin Grüntzig
Collegium Helveticum
ETH Zürich
Schmelzbergstr. 25
8092 Zürich
Switzerland
Robert S. Kerbel
University of Toronto
Biological Sciences Platform
Sunnybrook Research Institute
S-217, 2075 Bayview Avenue
Toronto, ON M4N 3M5
Canada
William C. Kisseberth
The Ohio State University
Department of Veterinary Clinical Sciences
448 VMAB
1900 Coffey Rd.
Columbus, OH 43210
USA
Esther K. Lee
University of Toronto
Sunnybrook Health Sciences Centre
2075 Bayview Avenue
Toronto, ON M4N3M5
Canada
Cheryl A. London
Department of Veterinary Biosciences
The Ohio State University
454 VMAB
1900 Coffey Rd.
Columbus, OH 43210
USA
Simone Macri
Istituto Superiore di Sanità
Department of Cell Biology and Neuroscience
Viale Regina Elena 299
00161 Roma
Italy
Marianne I. Martic-Kehl
Collegium Helveticum
STW/ETH-Zentrum
Schmelzbergstr. 25
8092 Zürich
Switzerland
Irving Miramontes
The University of Texas at El Paso
Department of Biological Sciences
Border Biomedical Research Center
500 W. University Avenue
El Paso, TX 79902
USA
Anthony J. Mutsaers
Department of Clinical Studies
Department of Biomedical Sciences
Ontario Veterinary College
University of Guelph
50 Stone Road
Guelph, ON N1G2W1
Canada
Karla Parra
The University of Texas at El Paso
Department of Biological Sciences
Border Biomedical Research Center
500 W. University Avenue
El Paso, TX 79902
USA
Andreas Pospischil
Collegium Helveticum
ETH Zürich
Schmelzbergstr. 25
8092 Zürich
Switzerland
Gerben ter Riet
University of Amsterdam
Academic Medical Center, J2-116
Meibergdreef 9
1105 AZ Amsterdam
The Netherlands
S. Helene Richter
University of Münster
Department of Behavioural Biology
Badestraße 13
48149 Münster
Germany
Jennie L. Rowell
Center of Excellence in Critical and Complex Care
College of Nursing
The Ohio State University
390 Newton Hall
1585 Neil Ave.
Columbus, OH 43210
USA
Chiara Spinello
Istituto Superiore di Sanità
Department of Cell Biology and Neuroscience
Viale Regina Elena 299
00161 Roma
Italy
P. August Schubiger
Collegium Helveticum
ETH Zürich
Schmelzbergstr. 25
8092 Zürich
Switzerland
Divya Vats
ETH Zürich
Institute for Biomedical Engineering
Wolfgang-Pauli-Str. 27
8093 Zürich
Switzerland
Isain Zapata
Center for Molecular and Human Genetics
The Research Institute at Nationwide Children's Hospital
700 Children's Drive
Columbus, OH 43205
USA
Preface
The second demand of The Three R guiding principles for more ethical use of animals in testing (already established in 1959 [1]) reads:
Reduction: use of methods that enable researchers to obtain comparable levels of information from fewer animals, or to obtain more information from the same number of animals.
Quoted from Wikipedia July 2015
The present volume on Animal Models for Human Cancer: Discovery and Development of Novel Therapeutics by Marianne Martic-Kehl and P. August Schubiger focuses in essence on the design and numerical evaluation of animal tests in cancer drug development. This field of research needs special attention because of various reasons. The most challenging one being the quality and validity of human tumor model in rodents, the most frequent used animal in cancer drug development. Since mice do not develop spontaneous human cancer, genetically modified organisms are the standard. Higher animals, like cats and dogs, would be the choice then, implicating an ethical conflict that is also at a “higher” level, besides the cost.
The volume editors have been able to invite distinguished experts from leading institutions and research organizations to reason about the current situation, to analyze pros and cons, and to come up with new suggestions to improve the situation. While this is definitely difficult to do for the animal experiment itself, much neglect can be detected in its statistical evaluation. How can it be that good experimental practice is violated in so many papers that have all passed peer review. Randomization, multiple use of the animals, clear and validated endpoints, and sufficient numbers are very often omitted or not commented on, as a meta-study of several hundred of recent publications in the field has revealed. Here, the second demand of the 3Rs is affected to a considerable extent, and at the same time could easily be followed by a more rigorous peer review system.
Marianne Martic-Kehl and P. August Schubiger deserve deep respect for tackling problems, which have been of concern for several decennia, but have always been and still are a kind of taboo. You don't make friends in the scientific community by asking those nasty questions. Hence, the scientific community should be grateful for the researchers having the “guts” to point to neglect and suggest alternatives and improvements.
In addition, we are very much indebted to Frank Weinreich and Waltraud Wüst, both at Wiley-VCH. Their support and ongoing engagement, not only for this book but for the whole series Methods and Principles in Medicinal Chemistry adds to the success of this excellent collection of monographs on various topics, all related to drug research.
Düsseldorf
Weisenheim am Sand
Zürich
Raimund MannholdHugo KubinyiGerd Folkers
January 2016
Reference
1. Russell, W.M.S. and Burch, R.L. (1959).
The Principles of Humane Experimental Technique
, London: Methuen.
A Personal Foreword
As animal well being is a prerequisite for reliable experimental results, it is of utmost importance to seek for methods and procedures that can reduce suffering of the animals and improve their welfare.
This sentence, closing Vera Baumans' conclusion in a 2004 paper on ethical dilemmas in animal research [1], unfolds the dilemma of what suffering means for animals. What is the perspective to be taken, where to position the borderline between objectivity and subjectivity? Are anthropocentric views the good or the flipside of the coin?
While it is uncontested that animals feel pain, the question remains of which kind it is. The McGill pain questionnaire will not apply to rodents. This question has been around in the literature for almost 4000 years. Animals as “tools” for research are ascribed to the times of Hippocrates, which is still under dispute. The famous Roman physician Galenus, however, became known as the Father of Vivisection. The debate probably flared up for the first time in the seventeenth century. One of the founding fathers of the enlightenment, the Dutch philosopher Baruch de Spinoza, “admitted that animals suffer, but we are within our moral rights to use them, as we please, treating them in the way which best suits us; for their nature is not like ours, and their emotions are naturally different from human emotions” [2]. This view seems to be enforced by what is widely known as the “Cartesian Gap,” assuming that Descartes categorized animals as “meaty machines” or as automata. However, when it comes to emotions I get the impression that Descartes – at least for the human being – is somewhat conciliating: Ainsi que souuent vne mesme action, qui nous est agreable lors que nous sommes en bonne humeur, nous peut déplaire lors que nous sommes tristes & chagrins.1 It has been more than 100 years later that Jeremy Bentham fiercely and provokingly opposed the Cartesian perspective:
The French have already discovered that the blackness of the skin is no reason why a human being should be abandoned without redress to the caprice of a tormentor. It may one day come to be recognized that the number of the legs, the villosity of the skin, or the termination of the os sacrum [tailbone),are reasons equally insufficient for abandoning a sensitive being to the same fate. What else is it that should trace the insuperable line? Is it the faculty of reason, or perhaps the faculty of discourse? But a full grown horse or dog is beyond comparison a more rational as well as more conversable animal, than an infant of a day, or even, a month old. But suppose they were otherwise, what would it avail? The question is not, Can they reason? nor Can they talk? but Can they suffer? [3]
Bentham is regarded among the first to foster animal rights, the ability to suffer being the benchmark, the insuperable line, instead of the ability to reason. Again approximately after a century Darwin established the biological similarities between man and animal. Not surprisingly, however, his seminal scientific findings led to an increase in animal experimentation [1], since they paved the ground for a rationale to use animals as a model for human physiology and biological function.
It was Darwin's contemporary, the great physiologist Claude Bernard, who established this similarity between man and animal as a scientific method and became the founding father of modern experimental medicine [4]. His key message precisely describes the contemporary paradigm of biomedical research:
Le médecin qui est jaloux de mériter ce nom dans le sens scientifique doit, en sortant de l'hôpital, aller dans son laboratoire, et c'est là qu'il cherchera par des expériences sur les animaux à se rendre compte de ce qu'il a observé chez ses malades, soit relativement au mécanisme des maladies, soit relativement à l'action des médicaments, soit relativement à l'origine des lésions morbides des organes ou des tissus. C'est là, en un mot, qu'il fera la vraie science médicale.2
Claude Bernard incorporated the principles of “hard science,” in particular physics and chemistry, into the realm of medical research and made them the cornerstones of his scientific method [4]. Since (physical or chemical) experiments in humans are clearly beyond any moral or legal acceptance, animal “deputies” became the scientific object to serve as the mere substance in modeling a human disease.
With the advent of genetic modification techniques in the 1980s, transgenic animals and in particular rodents—mice being the working horses of modern biomedical animal experimentation—opened a new era in disease modeling. Single gene function, genetic components, and regulatory networks could be correlated with diseased conditions in humans. Still, we are facing the problem of “bridging the gap” in between mouse genomics and the disease phenomenon in the individual human being. Hence, the front edge research in biomedical is focusing on non-human primates (NHPs).
Due to physiologic differences between rodents and higher primates, such as life span, brain size and complexity and motor repertoire, as well as the availability of cognitive behavioral testing, NHPs are considered one of the best animal models; especially for complex disorders that correlate with aging, cognitive behavioral function, mental development, and psychiatric dysfunctions. In addition to neural psychiatric related disorders, metabolic function, reproductive physiology, and immunology are other areas of research where the NHP model has been widely used. [5]
Given the complexity of a disease or illness, the causality of which is, as we increasingly understand, far from a simple “one gene, one disease” situation that can be reduced to a single biochemical step in the cell only in a few cases. Many more parameters in animal experimentation have to be considered than just measurement of the chemistry and physics, often termed “surrogate parameters.” Referring back to the opening quotation, environmental conditions play a crucial role in obtaining reliable scientific results from the models. Overcoming structuralistic views and granting animals a body–mind relation too, probably very similar to humans, does not facilitate animal experimentation and its interpretation. The human–animal boundary is closer.
But besides ethical and moral concerns in general, there are good scientific and economical reasons to scrutinize and carefully optimize laboratory experimentation with animals. Those experiments are costly, need special infrastructure, lots of paperwork and hence quite a number of laboratory staff; each outcome of the experiments should contribute to our knowledge. The battle between hypothesis-driven or explorative research can already be found in the musings of Claude Bernard.2 The demand for both is to extract the maximum of information. Scanning current scientific papers seems, provocatively, to be rather the exception than the rule as the editors and their distinguished invited authors of the present volume show in their contributions. Good laboratory practice as randomization, clear endpoints, sound statistics, selective reporting, and publication bias are at stake and are often ruthlessly abandoned. There seems to be much room for improvement.
Emerging from a fellowship at the Collegium Helveticum, Marianne Martic-Kehl and P. August Schubiger, with their background of active researchers in life sciences, focused for several years on getting hard data about the practice of animal experiments, mostly in rodents, with importance placed on cancer research. Over the years they have continuously confronted their colleagues with their findings and elicited fierce debates in the interdisciplinary environment of the Collegium. The project culminated in a final symposium, the results of which yielded the contents of this book. The author is extremely grateful to both of the editors to have picked up a kind of taboo topic in biomedical research and sometimes stubbornly to follow its traces in the vast universe of biomedical publications.
The book deserves a wide readership the scientific community and beyond.
Gerd Folkers
January 2016
References
1. Baumans, V. (2004) Use of animals in experimental research: an ethical dilemma?
Gene Ther
.,
11
, 64–66.
2. Soccio, D.J. (2009)
Archetypes of Wisdom
. 9th edn, Boston, MA: Cengage Learning, p. 55f.
3. Bentham, J. (1828) Of the limits of the penal branch of jurisprudence. In:
An Introduction to the Principles of Morals and Legislation
. A new edition, corrected by the author. Footnote 122: Interest of the inferior animals improperly neglected in legislation, London: Dover Classics, pp. 235, 236.
4. LaFollette H. and Shanks, N. 1994) Animal experimentation: the legacy of Claude Bernard.
Int. Stud. Philos. Sci.
,
8
, 195–210.
5. Chan, A.W.S. (2013) Progress and prospects for genetic modification of nonhuman primate models in biomedical research.
ILAR J.
,
54
, 211–223.
1
Correspondence, 1630, A Monsieur *** (Isaac Beeckman), 17 october 1630.
2
The doctor who is jealous to deserve that name in the scientific sense must, coming out of the hospital, go to his laboratory, and there in the laboratory by experiments on animals, he will seek to account for what he has observed in his patients, whether about the action of drugs or about the origin of morbid lesions in organs or tissues. This is, in a word, where he will do the true medical science. Claude Bernard,
Introduction À L'étude De La Médecine Expérimentale
(1865). Ebook Project Gutenberg (This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away, or re-use it under the terms of the Project Gutenberg License included with this eBook or online at
www.gutenberg.net
).
Chapter 1Introduction
Marianne Isabelle Martic-Kehl, Michael F.W. Festing, Carlos Alvarez and P. August Schubiger
1.1 Animal Models in Biomedical Research
Modern biomedical research relies heavily on the use of laboratory animals, particularly mice, rats, and fish, which, according to UK data for 2013, accounted for 94% of animals used in research. The research included fundamental studies aimed at understanding biological processes, the preclinical testing of potential new drugs and therapies, the development of diagnostic reagents and, in the case of monoclonal antibodies (mAbs), the production of therapeutic agents themselves.
In all developed countries the use of animals should (and probably is) strictly regulated in order to minimize pain and distress. All research workers should be familiar with the “Three Rs,” Replacement, Refinement, and Reduction described in the book The Principles Of Humane Experimental Technique [1]. Thus, where possible, non-sentient alternatives to the use of animals should be used as a “Replacement,” but if animals must be used, then “Refinements,” such as anesthesia and analgesia as well as enriched housing conditions, should be used to minimize pain, distress, or lasting harm, and, finally, the number of animals used should be “Reduced” to the minimum necessary to meet the objectives of the study.
There are continued, successful, efforts to develop alternatives to the use of animals. For example, large numbers of animals were once used for assaying many biological reagents such as hormones and vaccines. These have now largely been replaced by in vitro methods such as direct immunological or chemical assays. Fundamental research uses large numbers of mice but also makes extensive use of cell cultures and tissues from animals, which have been humanely euthanized. An important “Refinement” has been the development of disease-free or so-called “specific pathogen free” mice, rats, guinea pigs, rabbits, cats, and a few other species. These are free of clinical and sub-clinical infections that can cause problems if the animals are stressed by an experimental treatment. “Reduction” is achieved by good experimental design in which neither too many animals are used, which would be wasteful, nor too few, which might mean that important reactions are missed.
In vivoresearch plays an important role in life science, particularly in preclinical drug development. The standard drug development process nowadays consists of several sub-phases (research phase, preclinical development, clinical phases I–III), which take several years and give rise to costs between US$ 50 million and 2 billion [2–4].
This development process is highly prone to attrition. One critical step is the translation of preclinical animal research results to the clinic. In the last decade, it has been frequently revealed in many research fields that translation rates are minimal, non-existent, or generally shrinking [5–7]. A literature investigation by Thomson-Reuters revealed that the success rates of development projects in phase II clinical trials fell from 28% to 18% between 2009 and 2010 [7]. In more than half of the cases the reason for attrition identified was insufficient efficacy.
1.2 Animals in the Drug Development Process: Historic Background
Animals were already used as surrogate organisms for humans in the nineteenth century for the purpose of understanding chemistry-based drug effects on physiological function. The first Pure Food and Drugs Act in the USA (1906) described official standards for drugs and proper labeling and prohibited the interstate commerce of unsafe drugs. In 1938 the Food, Drug, and Cosmetic Act further required proof of safety and authorized inspections as a consequence of the sulfanilamide-disaster in 1937 that killed over a hundred people in the USA. The reason for the toxicity was that the cough syrup containing the antibiotic sulfanilamide also contained the toxic solvent diethylene glycol, which made the syrup very popular with children because of its sweet taste.
The first privately financed, nationally supervised, and evaluated drug was streptomycin, which was approved in 1945. However, the use of regulatory authorities and the movement toward supervised drug development was still limited to the Anglo-Saxon world. At that time in continental Europe the view persisted that diseases were non-comparable processes, not suitable for statistical evaluation “since the treatment never concerns populations but only individual patients” [8]. In 1954, the German company Grünenthal patented the sedative thalidomide and in 1957 launched it as Contergan® in West Germany. Since the company considered the drug particularly safe, they marketed it as a sleeping pill for pregnant women, as well as a morning sickness preventative in early pregnancy. However 2 years later the first reports of nerve damage related to Contergan® appeared, and in 1961 Grünenthal had to withdraw the drug from the market when thousands of babies were born with extremity abnormalities. This tragedy, known as the thalidomide-disaster, led to the first drug law in Germany (1961). The USA revised its existing drug law a year later to require proof of efficacy and sufficient pharmacological and toxicological results from animal trials before granting a license for market authorization. This provided the basis for the drug development process, as it is known today in developed countries. To date, it has not been possible to fully replace live animals as human surrogates, and therefore it is of great concern that such experiments should be performed in the most ethical way possible [9–12].
In various official pronouncements, for example, of the Royal Society in the UK, the UK Department of Health or the US Department of Public Health, it is stated that “Virtually every medical achievement of the last century has depended directly or indirectly on research with animals.” Whether or not this claim can be backed by any proven evidence was unclear until 2008, when Robert Matthews investigated it. In his article, he came to the conclusion that even though the statement does not generally hold true, “animal models can and have provided many crucial insights that have led to major advances in medicine and surgery” [13].
Indeed, research using laboratory and domestic animals has underpinned many major advances in human medicine. Perhaps Louis Pasteur in the nineteenth century should be credited with the first use of scientific methods to develop new treatments for infectious disease. He used dogs and rabbits to develop methods for immunizing dogs and humans against rabies, a viral disease (although viruses were not known at that time), and sheep to immunize sheep against anthrax, a bacterial disease. His methods laid the groundwork for the development of vaccines used today to control diseases such as polio, measles, mumps, and rubella. Two infectious viral diseases, smallpox and rinderpest, a serious disease of cattle, have even been entirely eliminated from the wild and polio has nearly been eliminated. The first oncogenic retrovirus was discovered in 1911 by Peyton Rous, who found that cancer can be induced in chickens by injecting them with a cell-free extract from a chicken tumor. Further studies of the biology of murine retroviruses, such as the Bittner mammary tumor virus and murine leukemia virus, meant that when the human immunodeficiency virus (HIV) appeared, at least the biology of retroviruses was largely understood. Although infectious diseases have now largely been controlled in developed countries, new zoonotic diseases such as that caused by the Ebola virus, which is maintained in wild animals in West Africa, and various strains of the influenza virus present in wild and domestic birds remain a constant threat, especially in view of the rapidity with which diseases can be transmitted throughout the world. Moreover, antibiotic-resistant bacteria also remain a constant threat.
Transplantation of kidneys, hearts, and other organs has saved many lives. This was made possible by the discovery of immunological tolerance by Peter Medawar in the 1950s. At that time it was known that skin grafts between two individuals would be rejected, but it was assumed that this was a physiological problem. Medawar showed that reciprocal skin grafts between two different strains of mice are rejected. However, if lymphocytes of a donor strain were injected into baby mice of a recipient strain, treated adult mice of the recipient strain would then permanently tolerate grafts from the donor strain. This showed for the first time that graft rejection is an immunological rather than a physiological phenomenon, and that it can be controlled by immunological methods. The development of drugs such as cephalosporin to dampen the immune system, again using laboratory animals, has made organ transplantation possible.
The first chemotherapy was developed by Paul Ehrlich, who, in 1909 screened 606 chemicals for activity against the spirochete causing syphilis, using rabbits infected with the organism, and found one, salvarsan, that was effective. For a time it was the most widely prescribed drug in the world. The development of new drugs now depends on an understanding of the biology of the disease, the identification of possible drug targets and the screening of large numbers of chemicals likely to interact with the target, using in vitro and in vivo methods involving research animals. Any potential new drugs will be tested in animals for safety and efficacy, usually in mice, rats, and dogs before proceeding to clinical trials.
The discovery of insulin in the early 1920s has saved many millions of human lives. Banting and Best ligated the pancreatic duct of dogs and found that cells associated with the production of digestive enzymes degenerated, leaving islands of cells. These secreted the hormone later designated as insulin, and they showed that it could be used to maintain diabetic dogs. The biochemist Collip developed methods of purifying it from porcine and bovine pancreases, using several thousand rabbits to assay it. Fortunately, although these insulins are different from human insulin, they are sufficiently similar to be effective in humans. Before that time, type I diabetes was usually fatal. For many years, batches of porcine or bovine insulin had to be assayed using mice or rabbits. Frederick Sanger sequenced the insulin protein in 1955 and genetically modified human insulin is now produced in bacterial cultures and assayed chemically.
Antibiotics have probably saved more lives than any other medical intervention. Penicillin was discovered by Alexander Fleming in 1928, but he was unable to isolate it and verify that it was effective. This was done by Ernst Chain and Howard Florey, who were able to show that it was both effective and non-toxic in mice. They went on to develop a method for producing it on a large scale. Many other antibiotics have been discovered since then. For example, Selman Waksman discovered streptomycin in research involving mice, guinea pigs, and chickens.
Nutritional deficiency diseases are fortunately now rare in developed countries, but are still a problem in some underdeveloped ones. Frederick Gowland Hopkins showed that young rats given diets of purified protein, carbohydrate, minerals, and fat stopped growing, but when they were given a small amount of milk they grew. He postulated the existence of substances required in the diet in minute amounts, which were later called vitamins. The vitamin that he discovered was designated vitamin A. His work coincided with that of Christiaan Eijkman, who was attempting to find the cause of beriberi, a disease characterized by loss of feeling in the feet and difficulties in breathing. He injected the blood of soldiers hospitalized with beriberi into chickens, but also noticed that the chickens fed on scraps of the same polished rice diet as the soldiers also got sick, whereas those receiving unpolished rice remained healthy. The disease was caused by a deficiency of what we now call vitamin B1 (thiamine). Hopkins and Eijkman shared the 1929 Nobel Prize for their work.
The few examples cited above demonstrate how, historically, animal research has contributed to the development of many areas of medicine. The development of mAbs is a relatively recent advance that has resulted in a limitless supply of highly specific diagnostic reagents as well as many promising new therapeutic agents. B-cell multiple myelomas have been recognized in humans for many years, and it was also known that they produced mAbs, known as Bence–Jones proteins. In the late 1960s it was found [14] that the BALB/c inbred strain of mice produced myelomas when injected i.p. with mineral oil. These myelomas were immortalized and could be maintained as permanent cell cultures. In 1975 Kohler and Millstein fused these myeloma cells with spleen cells from mice that had been immunized to sheep red blood cells and found that the “hybridomas” secreted antibodies to sheep red blood cells. They were able to select out individual hybridoma cells, each of which produced a monoclonal antibody. Subsequently, the immunoglobulin genes of the mice were replaced by the equivalent human genes by means of genetic engineering, so that human rather than murine mAbs could be produced. This avoids any possible problems associated with adverse reactions to mouse proteins. mAbs are now used to treat several diseases such as some forms of cancer and as anti-inflammatory agents to treat diseases like rheumatoid arthritis and Crohn's disease. Many more are being tested in clinical trials. Because of their high specificity they are also widely used in the diagnosis of disease.
Other examples where animals have made important contributions include blood transfusion, joint replacements, reproduction, and in vitro fertilization (allowing many otherwise infertile couples to have children), heart valve replacement, cancer, and stroke. Moreover, veterinary medicine and human medicine are converging. Dogs and humans get many of the same diseases, such as cancer, obesity, and type II diabetes. Dogs also get a number of hereditary diseases, in some cases as a result of many generations of selective breeding, which are inappropriate to or incompatible with good health.
1.3 Problems with Translation of Animal Data to the Clinic
Despite the impressive examples described above, many articles in scientific and non-scientific journals have criticized the quality and reporting of animal research in drug development in the last decades. For some diseases, animal models were found to have no predictive value for clinical applications. A mouse model developed to investigate cystic fibrosis, for example, turned out to show symptoms different from human patients, even though the same genetic modification was introduced [15]. Another example is the search for HIV vaccinations using non-human primates as a surrogate organism. Chimpanzees and macaques infected with the simian immunodeficiency virus (SIV), a virus similar to HIV and from which HIV is assumed to have developed, turned out to be responsive to various vaccination candidates, whereas none has been translated successfully to human patients so far [16–19].
Certainly, humans are not 70 kg mice and it is probably utopic to assume that the efficacy of any drug candidate can be predicted 100% reliably by the investigation of an animal surrogate; nevertheless, literature analyses in various fields of research have revealed a variety of potential causes apart from pure genetics for the low predictive value of animal research data. Poor experimental planning, inappropriate statistical analysis, and insufficient reporting are keywords frequently summarized in the literature [6, 13, 16, 19–27] and it is highly conceivable that the predictive value of animal research can be increased substantially by eliminating such methodological shortcomings.
A disease field where considerable work has been done to detect potential reasons for translation failures is acute stroke. Literature analysis revealed that almost 500 intervention candidates have shown satisfactory efficacy in animal models, whereas only three interventions have been proven to be effective in patients suffering from acute stroke [5, 28]. For various interventions with positive outcome in animal models, meta-analyses were performed to investigate potential reasons for this high failure rate. Judging from a checklist with 10 quality criteria, researchers found that low-quality studies tended to overestimate effect sizes [5].
In analyzing the quality shortcomings of the studies, the investigators identified two main groups, which can be summarized as general, stroke-independent or stroke-specific shortcomings. The latter included the use of animal models (mostly mice) that did not reflect the general health state of an average stroke patient. Human patients are often elderly, suffering from additional health problems such as hypertension or diabetes [5], whereas mice are young and healthy apart from the artificially introduced lesion to trigger stroke symptoms. Furthermore, other researchers identified discrepancies in the administration schedule of a particular drug candidate. Treatment onset occurred much sooner in animals (median 10 min) than in patients (median 5 h) [29].
The other, more general, group of quality shortcomings concerns the frequent neglect of study design and performance concepts in animal research, which are standard for clinical trials. These include random allocation of animals to test and control groups, blinded performance and assessment of study outcome, and sample size calculation before study performance in order to guarantee a certain study power (which should by convention minimally be 80–90%) [5, 29, 30].
Similar issues with study quality were observed in amyotrophic lateral sclerosis research [31].
1.4 Animal Studies in Anti-cancer Drug Development
Failure rates of drug effects in the clinical test phase after successful animal experiments were reported to be highest in the field of oncology [32]. In 2011, the licensing success rate for anti-cancer drugs reached 5%, in contrast to 20% for that of cardiovascular diseases [32].
In a systematic review of 232 publications in this field, only 41% reported randomization, and only 2% reported blinded assessment of outcome. None reported allocation concealment and only one reported sample-size calculations [33]. Even though many articles have been published in the last decade emphasizing the importance of such study design features, there has been no increase in their reporting between the late 1990s and 2011 [33]. The only exception was the increased incorporation of conflict of interest statements in more recent articles than in older ones [33]. This phenomenon can easily be explained by more extensive author's guidelines of numerous scientific journals, which require a conflict of interest statement.
It seems that external enforcement, for example, by journal editors, is necessary to achieve an improvement in reporting quality—and, presumably, performance quality—of animal studies. There was also a tendency for higher quality studies to report more small or non-existent effects compared with low-quality studies [33].
Anti-angiogenic cancer drugs represent a striking example both of how clinically irrelevant animal models can mislead decision-making, and also how well clinically relevant animal models can provide important information not only on efficacy, but also on potential harmful side effects of drug candidates. After marketing of the drugs, evidence was found that certain anti-angiogenic drugs could trigger metastatic evasion of cancer cells in patients [34]. Retrospectively, it was found that this phenomenon could have been foreseen by investigating metastatic cancer models (highly clinically relevant) in mice. Primary cancer models did not show similar results [34].
Tumor location within the animal model can also play a crucial role in predictive value. The easiest and cheapest way of inoculating tumors into an animal is to use a subcutaneous injection into the shoulder or flank. It is then easy to follow tumor growth. However, tumor cells then grow in an area different from their naturally occurring stromal conditions, which might crucially influence their growth and reaction to cancer drugs. For preclinical drug testing, it would therefore make sense to use orthotopic tumor models where tumors are inoculated into the organ of tumor cell origin. It is also possible that the animal species has to be carefully considered; a spontaneous dog or cat tumor may be genetically and behaviorally closer to human tumors than a human tumor induced in mice, which would not occur naturally.
1.5 Toward Relevant Animal Data
Problems and shortcomings in animal research as described above for various fields of disease, and for cancer research in particular, need to be resolved in order to achieve maximal relevancy of animal data.
Choosing the most representative animal and disease model depends on the field of research, but methodological shortcomings of study design, evaluation, and reporting are found almost universally. Aspects of study design such as randomization, blinded assessment of experimental outcome, and sample size calculations have been mentioned in the previous sections, but there are further aspects to be added.
A clear
