It is often claimed that scientists and doctors are united in their belief in the value of, and necessity for, animal experiments to protect human health. This stream of quotes from scientists and doctors, stretching from as far back as the early 20th Century right up to today, shows that there has been a long tradition of scepticism about this issue. Many of the quotes are from scientists who support or conduct animal research; far from diminishing the impact of their words, this fact ought to give them extra weight.
Note: FDA stands for Food and Drug Administration, the US drug regulatory body.
"It was hoped that the introduction of PD-related mammalian genetic models... would provide a disease-relevant context for testing potential neuroprotective therapies. However, a common feature of the genetically modified rodent models is the paucity or absence of cell death in the substantia nigra pars compacta and the absence of Lewy pathology... In contrast to neurotoxin-based models, which show acute neuronal death, genetically modified rodent models were thought to allow progressive degeneration, in the same way as in human PD, hence it was disappointing to discover that this was not the case."
Wassilios G. Meissner et al, Priorities in Parkinson’s disease research, NATURE REVIEWS | DRUG DISCOVERY VOLUME 10 | MAY 2011 | p385
Until now we have only been able to mimic epilepsy using experimental animal models but this can never give you a true picture of what is actually going on inside the human brain in epilepsy
– Dr Mark Cunningham, Institute of Neuroscience, Newcastle University.
ScienceDaily.com, 1st December 2009
In brain tumour research it is clear that the development of three-dimensional tissue culture models, which utilise both cells and serum of human origin, offer a real alternative to some of the established laboratory animal models which have been found to be wanting when clinical translation to the disease in man is considered. Geoffrey J Pilkington BSc PhD CBiol FIBiol FRCPath, Professor of Cellular & Molecular Neuro-oncology, School of Pharmacy & Biomedical Sciences at University of Portsmouth http://www.politics.co.uk/opinion-formers/press-releases/animal-welfare/adi-meps-buckle-under-vivisection-industry-pressure-$1293593$464772.htm
The animal models are pretty useless, to be honest. Now you can replay the human disease over and over in the dish and ask what are the very early steps that began the process. Dr Clive Svendsen, a cell biologist from the Waisman Center at the University of Wisconsin, Madison, speaking about models for Spinal Muscular Atrophy in Nature, 22nd December
The following quotes are all taken from: Jim Schnabel, Standard Model: Questions raised about the use of ‘ALS mice’ are prompting a broad reappraisal of the way that drugs are tested in animalmodels of neurodegenerative disease. Nature, Vol 454,682-685.
- The results of drug tests in mice have never translated perfectly to tests in humans. But in recent years, and especially for neurodegenerative diseases, mouse model results have seemed nearly useless.
- In the past year, for example, three major Alzheimer’s drug candidates, Alzhemed (3-amino-1-propanesulphonic acid), Flurizan (tarenflurbil) and bapineuzumab, all of which had seemed powerfully effective in mouse models, have performed weakly or not at all in clinical trials involving thousands of human Alzheimer’s patients.
- In the case of ALS [amyotrophic lateral sclerosis, a progressive, usually fatal disease of the nerves], close to a dozen different drugs have been reported to prolong lifespan in the SOD1 mouse, yet have subsequently failed to show benefit in ALS patients. In the most recent and spectacular of these failures, the antibiotic minocycline, which had seemed modestly effective in four separate ALS mouse studies since 2002, was found last year to have worsened symptoms in a clinical trial of more than 400 patients.
- The wrong model?
Perhaps the biggest issue facing the field is whether the mouse models faithfully reproduce the biology of the human disease. Alzheimer’s mouse models typically develop amyloid ‘plaques’ in their brains, but they do not develop an Alzheimer’s-like dementia and anti-amyloid strategies have repeatedly failed to slow the disease in clinical trials.
- Parkinson’s researchers have never had a good mouse model for the full disease process, and even the mouse model for Huntington’s disease — a relatively simple genetic disease— does not fully reproduce the clinical signs seen in humans with the disorder.
- The debate over [the ALS mouse model's] relevance to the majority of human ALS cases feeds a broader worry, namely that it may be unrealistic to think of modelling the ful complexity of ageing-related human brain disorders in mice whose disease course is usually accelerated by a single, crude genetic modification.
- ...Mouse models could therefore end up being not only more difficult and expensive to use with acceptable rigour, but at the same time more narrowly predictive of the human condition.
- But whether preclinical researchers will accept such a radical change remains to be seen.
“I think there’s a sense of desperation that we need a convenient model for bringing drugs to clinical trial,” says Benatar. “And I do sort of hear that concern.” But desperation, he adds, is an inadequate justification for the continued use of a poor model. “It’s a bit like the proverbial drunk who keeps looking for his lost keys under the lamp post, simply because the light’s better there.”
Testing drugs against AD on animals is not easy because animals don’t develop the disease. When using mice, scientists need to artificially induce one or several mutations in the mice and check whether they develop symptoms of the disease that are similar to the human ones. We showed that some of the mice currently bred to develop the disease don’t get better when they receive previously tested drugs. Professor Sascha Weggen, Heinrich-Heine-University, Germany, quoted on the American Society for Biochemistry and Molecular Biology website (16th August), referring to his article published in the Journal of Biological Chemistry, 282: 24504 - 24513.
In my research, animal models don't represent human patients sufficiently well, and that's a problem that extends across pain research as a whole. New and highly sophisticated brain-imaging technology is providing vital insights that animal research has failed to produce. I would like to see far greater uptake of these and other human-relevant approaches to pain research. Professor Qasim Aziz, the Barts and the London School of Medicine and Dentistry, speaking at the Pain workshop organised by Focus on Alternatives, April.
Our study shows that these mouse breeds may not reflect what may really happen in the brains of Alzheimer’s patients if they were treated with such compounds in future clinical studies. These compounds may seem to be ineffective on these mice, while it’s actually the mouse breed that is to blame. Professor Sascha Weggen. Professor Sascha Weggen, Heinrich-Heine-University, Germany, quoted on the American Society for Biochemistry and Molecular Biology website (16th August), referring to his article published in the Journal of Biological Chemistry, 282: 24504 - 24513.
The following quotes are taken from: Andrea Gawrylewski, The Scientist, The Trouble with Animal Models, Volume 21, Issue 7, Page 44.
- "There's no doubt about the absence of an effect [of NYX-059], and that called into question the many other studies in stroke, and how good are the animal models?" says Gilman. “So many agents appeared to be effective in the animal model and failed in human trials.” (Sid Gilman, Director of the Michigan Alzheimer's Disease Research Center in the Department of Neurology at the University of Michigan). Because of these failures, hundreds of millions of dollars, and a potential approach to stroke treatment, have disappeared down the drain. The failure of NXY-059 may have stalled the quest for a neuroprotective agent, at least for some time. "This trial has poisoned stroke studies," says Gilman. "I'm doubtful that investors will want to invest in clinical stroke trial[s] for a while." The fault, it appears, may rest in the slipshod use of animal models.
- To the dismay of clinicians and researchers of acute stroke, the compound showed limited efficacy in neuroprotection versus the placebo. Instead, NXY-059 joined the family of more than a dozen failed neuroprotective agents... All had reached Phase III clinical trials and failed miserably at doing what their animal model tests had suggested they would: stop the cascade of necrosis in the event of stroke, and protect the remaining viable brain cells.
- That NXY-059 had fallen victim to the same fate was particularly disheartening to a stroke roundtable group that had, in 1999, directly addressed the disconnection between animal models for stroke and their counterpart human trials. The group had devised a set of guidelines whose aim was to standardize the path to stroke therapeutics. During its development, NXY-059 had been its poster child. "This drug was being hailed as the first one to follow the standards," says Sean Savitz, assistant professor of neurology at Harvard Medical School. "But it didn't do that."
- The difficulties associated with using animal models for human disease result from the metabolic, anatomic, and cellular differences between humans and other creatures, but the problems go even deeper than that. One of the major criticisms of the NXY-059 testing was the lack of correlation between how the effects of the drug were monitored in animals versus in humans… Indeed, some consider the two phases of testing, from animal to human, completely out of whack, and that only by statistical fluke was SAINT I, the first clinical trial, deemed a success.
- Nonetheless, even genetically manipulated mice have their problems. The current knockout mouse model for amyotrophic lateral sclerosis (ALS) may be completely wrong, according to John Trojanowski at the University of Pennsylvania School of Medicine…Until now, research has focused primarily on SOD-1 knockout mice, with virtually no success in human trials. The new findings relating to the TDP-43 protein suggest that the SOD-1 knockout model for ALS could be wrong. "There was this nagging doubt" about the validity of the current models, Trojanowski says.
- A recent study at the Massachusetts Institute of Technology shows distinct differences between gene regulation in humans and mouse liver… they found that transcription factor binding sites differed between the species in 41% to 89% of the cases.
- Many of the underlying limitations associated with mice models involve the inherent nature of animal testing. The laboratory environment can have a significant effect on test results, as stress is a common factor in caged life.
- Richard N. Sifers, Associate Professor, Departments of Pathology and Molecular and Cellular Biology, University of Oklahoma Health Sciences Center, Baylor College of Medicine commented on the above article:
“We, as scientists, must admit that models are simply models! Although statistical robustness is certainly needed in animal studies, it must be accepted that models do not, and often cannot, recapitulate sophisticated human physiology. Similarities exist between apples and oranges (both are round and contain seeds), but one had better focus on an apple tree if interested in understanding the intimate details of its fruit. In a similar manner, only in very limited ways will any non-primate model recapitulate human physiology. However, these vast differences are not identified until one examines systems at a biochemical level, and this has become a very rare event. For too long, we have studied evolution in terms of investigating similarities between different species. These examples gave us clues as to the existence of evolution. However, the evolutionary process, by definition, actually refers to the vast differences that exist between species, and even between cells within a given species. Although many of the genes are shared, the regulation of their products can differ considerably!
I suspect that even if all the animal models faithfully mimicked the actual primary defects found in human diseases that they would still fall short of mimicking the human situation. Finally, model systems are certainly appropriate for some endeavors, but they will likely fall short (more times than not) when trying to identify drugs that will correct human diseases. For this reason, I sometimes wonder to what extent science actually advanced (in terms of understanding human disease) during the genomics era?”
All the following quotes are from: Faden & colleagues, Archives of Neurology. 2007;64:794-800.
- The ability of pharmacological agents to limit secondary biochemical damage and cell death has been well established in numerous animal models of stroke, head injury, and spinal cord injury, yet the results of such neuroprotective treatment strategies in human injury have been disappointing.
- However, after numerous failed clinical trials of drugs showing preclinical promise, questions have been raised about the relevance of current experimental models for human brain or spinal cord injury.
- Critical preclinical issues include the clinical predictiveness/ relevance of animal models, the adequacy of pharmacological methodology, and outcome measures used.
- Numerous methodological problems have been raised with regard to development of animal models that have meaningful clinical relevance. One area of concern relates to the choice of species, strain, or sex of the animal. For example, how well do models of brain trauma or ischemia in rodents reflect injury in higher species? Moreover, even within a given species, the same model can produce vastly different injury levels and outcomes across various strains. This issue is critical both for evaluating drug targets and regarding the choice of outcomes.
- Lastly, most animal models are highly constrained. To increase consistency, genetically identical animals are used under highly controlled experimental conditions.
- Such studies highlight the important caveat that even strong experimental data in lower species may not predict therapeutic effectiveness in human injury.
The field of Parkinson's Disease (PD) research was greatly stimulated by the therapeutic attempts of neurosurgeons using dopaminergic brain transplantation in animals and humans which came to the fore in the 1980s and have since largely receded. There have been too many false positives to record here. Many possible false negatives may also have been ignored as part of widely documented publication bias. The most common non-human primate model of PD results from monkeys being poisoned with the neurotoxin MPTP. It is widely acknowledged that profound disparities (anatomical, physiological, neurochemical, pathological, and temporal) exist between the MPTP non-human primate model and humans with idiopathic PD. Despite these paramount concerns of human reproducibility, hundreds of studies involving thousands of animals have followed with conflicting and non-predictive results. There is no evidence to suggest that their overall predictive concordance with human PD treatment, if subjected to the meticulous quantitative analysis of Perel and co-workers (BMJ 2007;334:197-200), would exceed the best case 50:50 coin toss probability established. Dr Marius Maxwell, Oxford, Cambridge & Harvard-trained neurosurgeon.
We have to remember that what works in mice does not unfortunately always work in men. Dr Lee Dunster, head of research and information at the Multiple Sclerosis Society, quoted in http://news.bbc.co.uk/1/hi/health/5245048.stm 6th August 2006.
Experimental allergic encephalomyelitis is not a suitable animal model for testing treatments for multiple sclerosis and it is time to explore alternative experimental and therapeutic approaches. Abhijit Chaudhuri, consultant neurologist, British Medical Journal, 2006;332:416-419.
Researchers have long sought a neuroprotectant drug that could salvage some of the brain tissues starved of oxygen after a stroke. But of more than 100 drugs that have entered major clinical trials over the past couple of decades after strong results in animals, none has proven effective in patients. That's probably because the artery-clogging clots in the human brain are caused by a diverse collection of underlying diseases that are not reflected in animal models. Pearson, Nature 444, 532 - 533.
Compounds to treat cancer and central nervous system disorders, including stroke, have particularly high failure rates - partly because animal models are such a poor mimic of the way these diseases affect humans. Pearson, Nature 444, 532 - 533.
Translation of positive results obtained in the laboratory into the clinic has been exceptionally elusive, and the stroke [research] community needs to think long and hard about whether these animal models are financially and ethically viable. The Lancet, 368:1548.
For stroke, the [animal] models and the clinical condition are extremely different, which is reflected in the failure of molecules in the human condition. Carney, Drug Discovery Today, 10, 1025–1029.
Already then (1980) I had a few reservations about such experimental models...I pointed out my doubts and asked whether the treatment would have an equal efficacy in humans as it had in rats. There was no good answer. Twenty-four years later and having been a principal investigator and a steering committee member in many acute stroke trials, I still have my doubts. Kaste, Use of animal models has not contributed to development of acute stroke therapies: Pro, Stroke 36, 2323–2324.
Since the early days of neuroprotecting agents in treatment of acute stroke, more than 700 drugs have been studied and more than 4000 papers describing their neuroprotective efficacy have been published, and yet none of those drugs has been accepted by regulatory authorities to be used for treatment of patients with acute stroke in the United States or the European Union. Kaste, Use of animal models has not contributed to development of acute stroke therapies: Pro, Stroke 36, 2323–2324.
Animal models have helped us better understand the pathophysiology of ischemic brain damage, but have they otherwise contributed much to clinical practice so far? I cannot say that they have, whereas randomized, clinical trials (RCTs) have had a major impact. The need of discipline, an essential part of any RCT, has influenced ordinary patient care in many positive ways. I do not expect either that more developed animal models could contribute to emergency stroke care so that a neuroprotective agent would be able to reduce the volume of an infarct in patients with stroke by 50% as they do in rats, at least if the therapy is not combined with thrombolysis or other neuroprotective therapies. Kaste, Use of animal models has not contributed to development of acute stroke therapies: Pro, Stroke 36, 2323–2324.
It must be recognized that no animal stroke model will precisely mimic human acute ischemic stroke, a condition that is quite heterogeneous. Recognizing the inherent limitations of animal stroke modeling should provide important lessons for both basic and clinical stroke researchers. Fisher and Tatlisumak, Use of animal models has not contributed to development of acute stroke therapies: Con, Stroke 36, 2324–2325.
Many stroke clinicians have been perplexed by the failure of other compounds trialed over the past 2 decades, despite strong evidence for efficacy in animal models. Donnan and Davis, Stroke 36, 2326.
One striking exception to the conventional pathway of drug development has been the positive results using recombinant factor VIIa to attenuate hematoma growth in patients with primary intracerebral hemorrhage. The biologic plausibility of this approach was based on clinical studies of the dynamics of hematoma growth documented by repeated computed tomography scans rather than animal models. The compound was already in clinical use as a hemostatic agent for another indication. This illustrates our view that although the majority of candidate stroke compounds need to be evaluated in preclinical animal models, there is always a place for astute clinicians to recognize the potential of compounds already in use for another clinical indication. Donnan and Davis, Stroke 36, 2326.
The following quotes are all taken from: Stephen H. Curry, Trials and Tribulations in the Search for Stroke Drugs, Preclinica, Vol. 2, No. 6, November/December, p378-381.
- Side effects in patients are often very difficult to predict from animal safety studies or even from healthy human volunteers.
- Animal studies can be and often are conducted with a variety of treatment delays, in order to determine the window of opportunity in the species concerned, but there is little basis for extrapolation of that information to humans.
- Central to this entire problem is the extreme difficulty of extrapolation from animals to humans.
- Some remarkable efficacies have been demonstrated in laboratory animals. Logical attempts have been made to relate the doses used and the drug concentrations in the blood supply to the brain in the various species, so that the human exposure is similar to that shown to be efficacious in animals. Yet, with remarkable and alarming repeatability, the promise of the animal results has not been borne out in the clinic, raising important questions concerning whether the entire process is flawed. It may be that experimental strokes in animals differ fundamentally from spontaneous stroke in humans or that the models are not being used properly.
- It may be impossible to produce a model of human stroke in the laboratory.
- In the most exhaustive analysis of the relevance of animal models to date, Green et al. first documented the most recent failures (in the 3 years up to and including 2003), bringing the score card to greater than 37 potential neuroprotective agents in more than 114 clinical trials, with no clinically efficacious agent identified or in use in the Western world.
- A further challenge with NMDA antagonists is a relative lack of subcortical protection, and attention has also been drawn to the fact that rat brains contain relatively little white matter, which is a major area of damage in humans.
- Humans tend to have longer pharmacokinetic half-life values than do rats, but not in proportion to body weight.
- Thus, in the human, the rise in glutamate after initiation of a stroke is slower than in the rat, the maximum concentration reached is lower, but the glutamate half-life is many times that of the rat.
Any suggestion about which is the ‘best’ rodent model to test putative neuroprotective agents is likely to prove contentious. Most investigators have personal variants of the major published models and all claim that their model has particular relevance, even though the predictive value cannot be demonstrated in the absence of a clinically efficacious drug. Green & colleagues, Animal models of stroke: do they have value for discovering neuroprotective agents? Trends in Pharmacological Sciences, Vol.24 No.8, p402-408.
It was proposed recently that animal models should be modified to reflect more accurately the complexity of human stroke. This is a near impossible goal, particularly for the purposes of drug screening. Green & colleagues, Animal models of stroke: do they have value for discovering neuroprotective agents? Trends in Pharmacological Sciences, Vol.24 No.8, p402-408.
Although it was suggested .10 years ago that animal models were flawed and would not produce a reliable prediction of clinical efficacy, they have continued to play a key role in the evaluation of putative neuroprotectants. Green & colleagues, Animal models of stroke: do they have value for discovering neuroprotective agents? Trends in Pharmacological Sciences, Vol.24 No.8, p402-408.
Alzheimer's, Parkinson's and other neurodegenerative diseases occur in humans and it is in human tissue that we will find the answers to these diseases. Dr. John Xuereb, Director of the Cambridge Brain Bank and Wolfson Brain Imaging Centre, BBC Radio Cambridge 7th February.
Discovery of drugs that act on the human central nervous system, are best studied in human-cell based systems. Drs Palfreyman, Charles and Blander, The importance of using human-based models in gene and drug discovery, Drug Discovery World, Fall 2002, p33-40.
The following quotes are from: Behan PO, Chaudhuri A, Roep BO. The pathogenesis of multiple sclerosis revisited. Journal Royal College Physicians Edinburgh, 32:244-265.
- Experimental allergic encephalomyelitis (EAE: the animal “model” used for MS) is totally different clinically, immunologically and histologically from MS although it does have similarities with other human demyelinating diseases.
- Experimental allergic encephalomyelitis has been used as a model for MS and it is the best studied organ-specific autoimmune disease: it is not acceptable as a model for MS for many reasons, some of which have already been alluded to...Treatment protocols in EAE are many but those used in humans, based on the findings in animal EAE, have singularly failed in alleviating the symptoms and signs of MS. Our contention is that it is inaccurate to extrapolate the findings in this putative animal model to the pathogenesis of human MS.
- Suffice it to say that the immunological reactions involved in EAE, whilst of clear importance in their own right in elucidating pathogenesis, have not been shown to occur in humans. Thus therapeutic strategies based on findings in EAE have failed to succeed in humans.
- A large number of potential MS therapies have been envisaged from experiments in animals with EAE...The list is far from complete but it should be stressed that the application of these potential therapeutic techniques to humans has been consumately associated with failure.
- Attempts at therapy based on experiments in rodents with EAE have so far failed.
- It is difficult, if not impossible, to induce EAE in neonatal animals;108 immunisation is only successful after a certain age. This is very important since the age at which animals become susceptible to developing EAE corresponds in humans to about the age of two years...If sensitised lymphocytes invading the brain are thought to be involved in the pathogenesis of MS, one might therefore not expect to find MS occurring below the age of two. However, there have been several reports of MS occurring in infants.
- The acceptance of EAE as a model for MS is an unfortunate error that has its basis on faith rather than science. Whilst EAE is a good example of an experimental organ-specific autoimmune disorder in animals, it cannot be accepted as a model for MS for a wide variety of reasons. This is particularly important in relation to the development of MS pharmacotherapy.
The following quotes are from: Gladstone & colleagues. Toward wisdom from failure: lessons from neuroprotective stroke trials and new therapeutic directions. Stroke 33:2123-2136.
- Neuroprotective drugs for acute stroke have appeared to work in animals, only to fail when tested in humans.
- Of >49 neuroprotective agents studied in >114 stroke trials, none has proven successful clinically. Similarly, neuroprotective therapy has been unsuccessful in clinical trials of head trauma. With the failure of so many trials, some clinicians may ask, "Is neuroprotective stroke therapy just a fantasy invented by basic scientists?" Will it ever play a clinical role? The answer is unclear. Gladstone & colleagues. Toward wisdom from failure: lessons from neuroprotective stroke trials and new therapeutic directions.
- Reasons for the failure of so many neuroprotective agents in clinical trials, despite their apparent benefit in animal (mostly rodent) models, have been the subject of intense discussion recently. In addition, despite significant similarities between the rodent and human genomes, the differences that do exist are sufficient to remind us that conclusions reached regarding genomic and proteomic characteristics in rodent studies may not apply to human stroke.
- Neuroanatomical, pathophysiological, pharmacokinetic, and genetic differences between rodents and humans notwithstanding, there has been a fundamental "disconnect" in the way that the efficacy of putative neuroprotective agents has been assessed in animal studies compared with clinical trials.
- Experimental stroke models are homogeneous, whereas human stroke is heterogeneous.
The following quotes are from: Recommendations for Standards Regarding Preclinical Neuroprotective and Restorative Drug Development, Stroke, 30:2752.
- Despite much animal research concerning the pathophysiology of focal ischemic brain injury, little of this work has translated into effective treatment modalities for stroke in humans.
- What remains curious is that although many of these agents appear quite effective in preclinical studies with small-animal models of ischemia (rats, mice, or gerbils), none of these have been conclusively effective in humans.
- A drug dose effective in the mouse or rat may or may not be effective in large animals &/ humans. It may not be sufficient or correct to merely scale up the dose of a drug in milligrams per kilogram from the mouse to larger animals & humans. Pharmacokinetics & pharmacodynamics may vary considerably among species.
- Another aspect to consider is that neuroprotective drug dose ranges & toxicities in animals may not overlap with those tolerated in humans.
- Gerbil models should be avoided because many pharmacological agents act as protectants in gerbils but not in other species.
- No animal model can exactly mimic stroke in humans.
- It is uncertain if benefit in young, healthy animals can be extrapolated to elderly, sick humans.
- There are a number of dissimilarities between the rodent brain & that of humans & nonhuman primates that may lead to differences in response to an identical ischemic insult. Therefore it is difficult to know how to scale up dosing regimens from rodents to humans... As noted above, recovery in rodent models occurs rapidly over the 1st few weeks after stroke. Recovery in humans with stroke may occur over a longer time, up to several months after stroke. Thus timing & duration of drug administration for humans is not easily extrapolated from rat models. A similar concern applies to drug dosage.
For cortical regions, such as the language areas, we cannot use the macaque brain even as a rough guide as it probably lacks comparable regions." Crick and Jones, Nature, 361: 109-110.
Ultimately, the answers to many of our questions regarding the underlying pathophysiology and treatment of stroke do not lie with continued attempts to model the human situation more perfectly in animals, but rather with the development of techniques to enable the study of more basic metabolism, pathophysiology and anatomical imaging detail in living humans. Professor David Wiebers, Mayo Clinic, Stroke, 21: 1-3.
Over-reliance upon such animal models may impede rather than advance scientific progress in the treatment of this disease. Professor David Wiebers, Mayo Clinic, Stroke, 21: 1-3.
The repeated failures of laboratory proven stroke therapies in humans can be due only to the inapplicability of animal models to human cerebral
vascular disease." Neff, Stroke, 20:699-700.
It is clearly not easy to establish a reliable indicator of the relative toxicities of the various anti-depressants, although some form of comparison is necessary. The 'natural experiment' of cases of self-poisoning has to be taken as the starting point as the results of experiments in animals cannot reliably be extrapolated to man and the induction of even mild toxicity in man, besides being unethical, may in any case not correspond to the drug's lethal toxicity. Drs Cassidy and Henry, British Medical Journal, vol 295, p 1021-1024.