Skip to main content

Confusion Regarding Predictive Value?

The June 9 edition of The Express (UK-based newspaper) featured an article by Danny Buckland titled: The extraordinary secret of how mice are curing cancer. The article examines the use of avatar mice to cure cancer. I have discussed avatar mice previously (see here and here). The mice have some DNA from a patient’s cancer and are used as a screening device for cancer treatments in addition to then being treated with the same drugs the patient is being treated with. The premise is that how the cancer reacts in mice will correspond to the how it reacts to drugs in the patient. This premise is false and ignores millions of years of evolution, modern genetics, and complexity science. But it does make money for some people.

The mice are being called an example of personalized medicine—tailoring treatments to match a patient’s genome. They are not. This is just an example of how a new hot topic in medical science—personalized medicine—is used to justify old nonsense—mouse research. The puff piece is complete with a testimonial from a patient in the UK being treated based on the mice: “It is a very intelligent approach.” And: “They will not be wasting my time or my life with Drug A or B that may not work. If they have given that to the mice, they will know and it will be easier to decide and I will benefit in the future.”

You could read virtually the same from any company selling another miracle cure.

There are other testimonials: “Professor Justin Stebbing, a consultant oncologist at Imperial College, London, who has used the treatment, says: ‘As oncologists we aim to improve quantity and quality of life but often there is a lack of data for therapeutic options and, in other situations, well recognised treatments are too toxic. This is the ultimate in personal medicine, delivering the right treatment to the right patient to the right tumour at the right time.’ ”

At least the article states: “The treatment is expensive and only available privately.”

The above clearly claims that animal models have predictive value for human response to drugs and disease. As does this, from April Peake, writing for who quotes David Willetts, Science Minister of the UK, as saying: “The Government is committed to working to reduce the use of animals in scientific research, but we do recognise that there remains a strong scientific case for the careful regulated use of animals in scientific research and that this does play a role in ensuring new medicines are safe and effective.”

As does the below from St. Jude Children's Research Hospital.

As a research institution, St. Jude has a unique mission to generate the knowledge that will save the lives of children stricken with cancer and other catastrophic diseases. In the course of our research to find cures for these deadly diseases, we do use laboratory animal models, mostly rats and mice bred specifically for that purpose. There is no substitute for animal testing when evaluating the effects of diseases and proposed treatments to fight those diseases. Most biological systems do not behave in a predictable manner and cannot be replicated by computer simulations. St. Jude is usually legally required, and always ethically obligated, to test treatments on laboratory models to ensure safety and efficacy before those treatments are studied in children. Without this research, St. Jude would not be able to provide hope for cures to our patients and their families.

But there are also statements from the scientific community like the following. Donald A. Prater, DVM, Deputy Director, in the European Office of the United States Food and Drug Administration stated during an October 10, 2012 lecture at the conference, “Advancing Safety Science and Health Research Under Horizon 2020 with Innovative, Non-Animal Tools" hosted by the European Parliament at Brussels: “Product development is increasingly costly, success rates remain low, many uncertainties exist, including, as a major component, failures in predicting toxicity despite extensive animal testing.”

Zhang et al state:

 [The publication of the report Toxicity Testing in the 21st Century: A Vision and a Strategy by the National Research Council of the National Academies of Science (NAS)] is a long-due response to the call by many for alternatives to the currently standard, whole-animal-based methodologies, which are inefficient, costly, and have had only limited success in making informative connections to human health risk associated with environmental chemical exposures.[1]

Elias Zerhouni, former director of NIH and current head of R&D at Sanofi was quoted in the June 25, 2012 issue of Forbes as saying: “R&D in pharma has been isolating itself for 20 years, thinking that animal models would be enough and highly predictive, and I think I want to just bring back the discipline of outstanding translational science, which means understand the disease in humans before I even touch a patient.”

In the inaugural issue of Disruptive Science And Technology, a publication by Mary Ann Liebert, Inc. Drake et al describe a new in vitro-based testing system for vaccines and the immunotherapeutics:

While it can be fairly argued that animal models have prompted significant gains in our understanding of the mammalian immune system, it has also been shown in numerous examples that laboratory species do not always mimic human diseases or recapitulate/predict human immunity.[2, 3] Likewise, current-day culture techniques generally fail to provide predictive data on human immunity. For example, few, if any, existing assay protocols are available to researchers to investigate human antibody responses against primary/naïve antigens.[4]

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” [5].

Barnes and Hayes state:

The need for new medicines for the treatment of neurological and psychiatric diseases is as great as it was twenty years ago…Moreover the limited predictive value of animal models of CNS [Central Nervous System] disease is also a challenge. Often disease phenotypes cannot be directly mimicked in animals (e.g. hallucinations) and even where there are correlates (e.g. sleep, pain, movement), the triggers used to mimic disease are often based on poorly understood mechanisms or existing pharmacology. [6]

Such statements could be easily multiplied (see here for more examples).

Notice that people with a vested interest in animal models want society to believe that animal models have predictive value for humans whereas scientists addressing the problems responsible for the lack of cures frankly admit that animal models are largely to blame due their lack of predictive value. If the science supporting the predictive value of animal models is so straightforward, why is there so much in the literature that says animal models are simply not predictive? I am not asking anyone to take my word or even the word of scientists like the above for the lack of predictive value of animal models. My advice has always been to study the issue for yourself. (I suggest you start by reading FAQs About the Use of Animals in Science: A handbook for the scientifically perplexed.) But the fact that such conflict can be found in the scientific literature should raise concern regarding the status quo as well as the honesty of science institutions and scientists with the vested interest in a product or practice.


1.         Zhang Q, Bhattacharya S, Andersen ME, Conolly RB: Computational systems biology and dose-response modeling in relation to new directions in toxicity testing. J Toxicol Environ Health B Crit Rev 2010, 13(2-4):253-276.

2.         Lewis AD, Johnson PR: Developing animal models for AIDS research--progress and problems. Trends in Biotechnology 1995, 13(4):142-150.

3.         Watkins DI, Burton DR, Kallas EG, Moore JP, Koff WC: Nonhuman primate models and the failure of the Merck HIV-1 vaccine in humans. Nat Med 2008, 14(6):617-621.

4.         Drake III DR, Singh I, Nguyen MN, Kachurin A, Wittman V, Parkhill R, Kachurina O, Moser JM, Burdin N, Moreau M et al: In Vitro Biomimetic Model of the Human Immune System for Predictive Vaccine Assessments. Disruptive Science and Technology 2012, 1(1):28-40.

5.         Gura T: Cancer Models: Systems for identifying new drugs are often faulty. Science 1997, 278(5340):1041-1042.

6.         Barnes JC, Hayes AG: CNS drug discovery: Realising the dream. Drug Discovery World 2002(Fall):54-57.


Popular Video