In my December 16, 2011 blog, I discussed the Safety of Medicines Bill and the comments by British MP David Amess. The British government via the Department of Health responded to the Safer Medicines’ campaign on March 30, 2012. The response can be accessed here. The response is typical and illustrates why asking for alternatives to current testing regimes using animals is poor science as well as futile.
I will not reproduce the response here but suffice it to say that the government stated once again that 1) animal testing works and 2) that nothing is available to replace it. They also said that if such tests become available they would be amenable to requiring them. However, for the time being, lives would be lost if animal testing were not required. There are many errors of fact on both sides in this campaign but I will once again point out three.
1. There are no tests, be they animal or nonanimal, that predict what any given drug will do in you when you take it. By predict, I mean that the test has a high enough positive and negative predictive values that any human can be assured that an adverse drug reaction is unlikely to the same degree that a lethal allergic reaction is unlikely. The only way this state of affairs will ever be realized is when
i) your genome is routinely mapped and filed,
ii) the functions of genes are known,
iii) the effects that drugs have on genes are known, and
iv) your physician can compare your genome to the drug profile.
Currently, there are drugs that have been connected to specific effects via specific genes. This is great news and is saving lives but it is only the tip of the iceberg. If society wants more of this genome-based, or personalized, medicine then more funding should be directed to the relevant research instead of animal models of disease X, Y, and Z. Finding genes in animals is not productive for personalized medicine as the functions of genes vary among species and human data is available, or easily obtainable, and can be so used.
Moreover, the tests currently in use, be they in vitro, in silico, or animal-based, are not predictive for humans in general in a vast majority of cases. The basic chemistry of a drug can be determined by the usual chemistry tests as can some other properties, but this is not what the above is referring to. When the government speaks about safety they mean toxicity testing. In reality, toxicity is influenced by absorption, distribution, metabolism, and elimination, as is efficacy. And all these parameters need to be known in order to make a knowledgeable decision regarding risk. Animals are used in an attempt to predict all of these and they fail. I have addressed this many times. The positive predictive value (PPV) and negative predictive value (NPV) for animal models is simply too low to be used as predictive tests for human response. So when the British government states that current animals tests are adequate and are therefore life saving, they are factually incorrect on the science.
2. When the animal protection community states that alternatives are available that are predictive they are also incorrect. Granted, nonanimal tests are available for specific tests but to the best of my knowledge most of these tests also give PPVs and NPVs far below what is considered predictive in medical science. Some of these nonanimal tests may give PPVs and NPVs that are higher than the animal tests but this is scientifically immaterial as, if a test fails to be predictive to the accepted standard, it fails to be predictive. It is irrelevant whether the test fails by 10 points or 40 points. A PPV of 0.3 and a PPV of 0.60 are both insufficient. The solution to the problem is to stop requiring tests that fail to accomplish the goal, not to implement other tests that also fail. Pharmaceutical science needs PPVs and NPVs around 0.99 and no toxicity test that is relevant to drug development provides this to the best of my knowledge. If anyone can refute this, please do. But do not send me studies that do not have a PPV and NPV and do not send me studies that measure something indirectly and then make ten assumptions and conclude their test is predictive. That is not science; that is wishful thinking.
Furthermore, there are hundreds of tests that use animals or human cells and there is a difference between what the regulatory agencies like the FDA require and what they accept. Considering the sheer numbers of in vivo and in vitro tests, there must be some test somewhere that uses intact animals or animal parts or human cells, that the FDA requires or accepts, and that gives a result that is predictive for some aspect of the drug development process. I do not know what such a test is (Caco-2 cells for drug absorption perhaps) or the PPV and NPV of the test, but I would not be surprised if there were such tests. A list of these with predictive values would be nice. But as a rule of thumb Pharma is lacking predictive tests and they and patients are paying a huge price for this. Regardless, if the government or the animal protection community is going to claim that animal tests are predictive or that nonanimal tests are predictive for drug development, then the burden of proof is on them to show PPVs in the 0.95-1.0 range. I am positive the government cannot do this for toxicity testing and I have never seen nonanimal tests approach this either.
3. Lives are being saved because of animal testing but not in the way such is portrayed. If you want to ensure that a drug does not harm anyone, simply do not develop or approve it. Problem solved. That is how animal tests are saving lives, by not allowing drugs to come to market. In reality every drug is a potential help to one person and potential problem to another, penicillin and aspirin being but two examples. Reactions to drugs vary even among family members and they vary greatly among humans in general. The variation is even greater among species. This is one reason personalized medicine is needed, to sort out who will benefit from a drug and who will be harmed by it or just not helped by it.
But this is not what vivisection activists and the British government mean when they say animal testing is saving lives. They mean that animal tests can predict human response and therefore the animal tests are keeping drugs that would damage all or most people off the market. This is sheer nonsense. Animal models are not predictive modalities for this purpose. The British government can make such claims in order to placate their citizens but such nonsense would not have prevented the thalidomide disaster nor is it preventing deaths and disasters today. (For more on thalidomide see The History and Implications of Testing Thalidomide on Animals.) Maybe the citizens of England do not care. Maybe they would rather ensure that GSK makes money than have safe drugs. Regardless, it is a sad state of affairs when a society allows itself to be pacified by nonsense.
If animal lovers that feel they are scientists want to help animals first they should make sure they are competent in the field under consideration. Most are not. Incompetence not withstanding, if they still feel compelled to take action in matters beyond their comprehension they should then put forth some effort in making certain that the solution to the problem they are addressing really is a solution. It would be nice if, in addition to the above, they would also make sure the problem actually exists as they are presenting it. The problem with the British government’s position is that it is not factual. That is where the position should be attacked. Not along lines that are themselves fact-challenged.
For more on all this, see Animal Models in Light of Evolution and FAQs About the Use of Animals in Science: A handbook for the scientifically perplexed.