Hi solarsanitizer, thanks for continuing the dialogue! I worked on 2 papers and am now back online.
1. Human clinical trials. Trials are a complicated process but let me simplify it down to 2 different kinds/aspects. In the traditional trials, there are human volunteers who take the drug, really to establish safety on a gross level. This is Phase I. Then the trials go to more humans for efficacy and so forth. There is a reasonably high risk for volunteers in these “Who goes first?” or Phase I trials. For example, speaking of toxicity trials for new drugs in humans, an unnamed clinician quoted in Science stated, “If you were to look in [a big company’s] files for testing small-molecule drugs you’d find hundreds of deaths." (Marshall 2000) This is not comforting.
Then there are vaccine trials and neuroprotectant trials and cancer trials. These operate more or less the same way BUT eventually you get to the patient in acute distress (or who will be) and who needs medical intervention, for example neuroprotectant trials. In those cases, the doctor is in the ER and the patient comes in having a stroke. You have 2 options. Do the usual and get the usual results (sometimes good sometimes not) or do the new trial that your university hospital is participating in. What you really do not want, is to do the new trial drug only to make the patient far worse than he would have been had you done nothing. But that is exactly what has happened with HIV vaccines and neuroprotectants and others. And this was based directly on animal studies.
The coin toss analogy is apropos in that at the end of the testing day you have data (mostly from animals) and you have to make a decision about whether to give this new drug to humans. At the end of the testing day, whether we are talking traditional Phase I trials or actually giving the drug to sick people, you have to make a go/no go decision. All the coin tosses so far really don’t count. The coin tosses so far got you to this point. Now you look back over all the data and make your ultimate decision. If the process works, your decision should be right a vast amount of the time. ALL involved in these decisions agree that the process is not viable. They think this because of things like the fact that only about 8% of drugs that enter Phase I go on to be marketed and that only about 50% of drugs that go through Phase III (there are three phases total) go to market. So by any evaluation, the process, which is largely animal-based, does not work.
Now, lets say we are at the end of that process and we toss a coin to determine go/no go. In light of the fact that ALL HIV vaccines have failed and that some made the patent more susceptible to HIV, and that ALL neuroprotectant drugs have failed and some have made the patient worse, and that ALL spinal cord protection drugs have failed and so on, the coin toss would have eliminated 50% of them hence less harm would have been done. NOTE! I am not advocating we toss a coin to make the go/no go decision. That would be stupid. However, it would probably work better than what we are currently doing ergo what we are currently doing is stupid.
One must also remember that we have lost drugs because of animal tests. Or would have lost drugs had anyone believed the animal tests. Consider the following:
Malcolm Young 2008: ". . . and it is anyone’s guess how many incorrect rejections (misses) there have been in which, for example, a candidate would have worked very beneficially in humans – if only we’d known – but was unfortunately not beneficial to mice." (Young 2008)
Furosemide, commonly called Lasix, is an example of an important medication almost lost secondary to animal studies. It is a diuretic, used to treat high blood pressure and heart disease. Mice, rats and hamsters suffer liver damage from it, but people do not. The drug is metabolized differs from species to species. (Walker and McElligott 1981; Weatherall 1982) Isoniazid an anti tuberculosis drug that has been used for decades causes cancer in lab animals (Clayson 1980; Shubick 1980). Clinical trials of digoxin were delayed secondary to the high blood pressure it caused in animals. (Okita 1967)(Jover et al. 1992)
Anisimov et al.:
By contrast, many of these substances were shown to be carcinogenic in rodents. Widely prescribed human medicines - acetaminophen, chloramphenicol, and metronidazole - are examples. Acetaminophen (paracetamol in the UK), an antipyretic that has been used extensively in developed countries since 1946, is not classifiable by the IARC by its carcinogenic effects on humans. However, animal experiments have shown that it increases the incidence of induced renal adenomas in rodents. These agents are only carcinogenic in high doses, sometimes only at the maximum tolerated dose. Chloramphenicol, an antibiotic, increased the incidence of induced lymphomas in mice, but the drug did not show a carcinogenic effect in humans". Metronidazole, an antibiotic that can destroy Helicobacter pylori and, therefore, probably decrease the risk of stomach cancer in humans, increased the incidence of induced colon cancer in rats. Similarly, some human carcinogens do not affect rodents, For example, the anticonvulsant diphenylhydantoin (phenytoin) is classified as carcinogenic to humans, but showed no carcinogenic effect in experimental mice and rats. These and other data reveal a serious problem in interpreting the results of animal carcinogen experiments in relation to humans. (Anisimov, Ukraintseva, and Yashin 2005)
Lazzarini et al. 2006: "Drugs which were unsuccessful in animal models were not used in clinical osteomyelitis, with few exceptions. Teicoplanin and linezolid were successful in the treatment of osteomyelitis in clinical trials, despite being completely inactive in two animal model studies of staphylococcal osteomyelitis. Therefore, the value of animal models as predictors of failure should also be carefully assessed." (Lazzarini et al. 2006)
Navarro and Senior writing the New England Journal of Medicine in 2006: "Statins have been shown to cause elevations of aminotransferase levels and severe liver injury in animals; in humans such elevations are common but rarely, if ever, lead to clinically significant hepatotoxicity." [They reference (Tolman 2002)] (Navarro and Senior 2006)
FK 506 (Tacrolimus) was almost shelved before proceeding to clinical trials (AMA 1990). Researchers stated: “Animal toxicity was too severe to proceed to clinical trial” (Calne et al. 1989).
Now put that in light of the below.
Park 2011: "A comparison of ~100 Merck drug candidates indicated that there was no correlation between incidence of liver toxicity observed in vivo in preclinical safety studies and level of CB (measured as pmol bound per mg protein, either in vitro or in vivo; TABLE 1)." (Park et al. 2011)
In an editorial introduction to one article by Ellis and Fidler and another by Van Dyke (Van Dyke 2010), Nature Medicine stated: "The complexity of human metastatic cancer is difficult to mimic in mouse models. As a consequence, seemingly successful studies in murine models do not translate into success in late phases of clinical trials, pouring money, time and people’s hope down the drain." (Ellis and Fidler 2010)
Ellis and Fidler: “Preclinical models, unfortunately, seldom reflect the disease state within humans (Fig. 1).” (Ellis and Fidler 2010)
Rothwell: "Prediction of the clinical consequences of interventions based on their effects in model systems has also often proved difficult. For example, the contrast between the countless reports of apparently beneficial effects of oestrogens on vascular biology in the laboratory and the harm later seen in large randomised trials of hormone replacement therapy is by no means atypical." (Rothwell 2006)
Alan Oliff, former executive director for cancer research at Merck Research Laboratories in West Point, Pennsylvania stated in 1997: "The fundamental problem in drug discovery for cancer is that the [animal] model systems are not predictive at all." (Gura 1997)
Chabner and Roberts: "Fewer than 10% of new drugs entering clinical trials in the period from 1970 to 1990 achieved FDA approval for marketing, and animal models seemed unreliable in predicting clinical success . . ." (Chabner and Roberts 2005)
2. “So far, the necessity of animal research has not been disproved. Challenged on ethical grounds yes, but not disproved.” I say in the books, and elsewhere, that animal use in science can be categorized into 9 different uses. Each use has to be judged on its own merit. For example, the use of animals for spare parts, say aortic valves, has been tested by time and vast numbers and has passed with flying colors. If someone wishes to put forth the hypothesis that pig valves don’t work in people, the burden of proof is on him.
On the other hand, experience has also shown major organ xenotransplants have not been successful to date. As of today, surgeons are not putting baboon hearts into humans. If someone wants to take the position that, as of today, major organ xenotransplants are not viable, there is no burden on him to show that. The burden of proof would be on the person claiming the opposite. Furthermore, the null hypothesis basically says there is no correlation until it has been proven. This supports the above both for pig valves and xeno. So you cannot paint with a broad brush in these matters.
Another example. The use of animals to demonstrate basic anatomical principles is viable. It has been done for centuries and the benefits are obvious. Animals can also be used in basic research to generate new ideas. I have stated that often. That has also been done for centuries. On the other hand, show me to proof that animal models can be used to predict human response to drugs and disease. You cannot say the words “animal research” and include all aspects of the enterprise (scientific enterprise not business enterprise). As I said previously, and many times elsewhere, the use of animals in science works for some specific uses. Most, in fact. In my “9 ways animals are used in science” classification, they work in 7 out of the 9 ways. But you cannot use historical successes of animal use in one area to prove it works in another. There is a name for that fallacy but it escapes me at the moment. Aging! It sucks, but it beats the alternative.
Thanks for reading and commenting, solarsanitizer!
(Now the usually disclaimers. 1. Unless otherwise noted, I am addressing the use of animals to predict human response to drugs and disease. 2. I have addressed most of this elsewhere and occasionally get sloppy and make partial arguments because a) in my arrogance and narcissism I assume everyone has read everything I have previously written and b) the actual arguments are really, really long. So if an argument appears incomplete that is probably because it is. 3. Material presented in blogs is by definition incomplete. Don’t confuse blogs with scientific papers or books. 4. If anyone wants to really understand all this she needs to read the books: Animal Models in Light of Evolution for the scientifically literate and, if not real comfortable when talking about complexity science and evo devo, then FAQs About the Use of Animals in Science: A handbook for the scientifically perplexed. Knowledge does not consist of facts. Facts are necessary but not sufficient. Knowledge is connecting all those facts together. Think of facts as nodes in a network and all the connections between the nodes as lines. The more lines the more knowledge.)
AMA. 1990. Animals in research: the American Medical Association's position. JAMA 263 (13):1766.
Anisimov, V. N., S. V. Ukraintseva, and A. I. Yashin. 2005. Cancer in rodents: does it tell us about cancer in humans? Nat Rev Cancer 5 (10):807-19.
Calne, R. Y., D. S. Collier, S. Lim, S. G. Pollard, A. Samaan, D. J. White, and S. Thiru. 1989. Rapamycin for immunosuppression in organ allografting. Lancet 2 (8656):227.
Chabner, B. A., and T. G. Roberts, Jr. 2005. Timeline: Chemotherapy and the war on cancer. Nat Rev Cancer 5 (1):65-72.
Clayson, DB. 1980. The carcinogenic action of drugs in man and animals. In Human Epidemiology and Animal Laboratory Correlations in Chemical Carcinogenesis edited by F. Coulston and P. Shubick: Ablex Pub.
Ellis, L. M., and I. J. Fidler. 2010. Finding the tumor copycat. Therapy fails, patients don't. Nat Med 16 (9):974-5.
Gura, T. 1997. Cancer Models: Systems for identifying new drugs are often faulty. Science 278 (5340):1041-2.
Jover, R., X. Ponsoda, J. V. Castell, and M. J. Gomez-Lechon. 1992. Evaluation of the cytotoxicity of ten chemicals on human cultured hepatocytes: Predictability of human toxicity and comparison with rodent cell culture systems. Toxicology in vitro : an international journal published in association with BIBRA 6 (1):47-52.
Lazzarini, L., K. A. Overgaard, E. Conti, and M. E. Shirtliff. 2006. Experimental osteomyelitis: what have we learned from animal studies about the systemic treatment of osteomyelitis? J Chemother 18 (5):451-60.
Marshall, E. 2000. Gene therapy on trial. Science 288 (5468):951-7.
Navarro, V. J., and J. R. Senior. 2006. Drug-related hepatotoxicity. N Engl J Med 354 (7):731-9.
Okita, G. T. 1967. Species difference in duration of action of cardiac glycosides. Federation proceedings 26 (4):1125-30.
Park, B. Kevin, et al. 2011. Managing the challenge of chemically reactive metabolites in drug development. Nat Rev Drug Discov 10 (4):292-306.
Rothwell, P. M. 2006. Funding for practice-oriented clinical research. Lancet 368 (9532):262-6.
Shubick, P. 1980. Statement of the Problem. In Human Epidemiology and Animal Laboratory Correlations in Chemical Carcinogenesis, edited by F. Coulston and P. Shubick: Ablex Pub. .
Tolman, K. G. 2002. The liver and lovastatin. Am J Cardiol 89 (12):1374-80.
Van Dyke, T. 2010. Finding the tumor copycat: approximating a human cancer. Nat Med 16 (9):976-7.
Walker, R. M., and T. F. McElligott. 1981. Furosemide induced hepatotoxicity. J Pathol 135 (4):301-14.
Weatherall, M. 1982. An end to the search for new drugs? Nature 296:387-90.
Young, Malcolm. 2008. Prediction v Attrition Drug Discovery World (Fall):9-12.