ILAR Journal Online, Volume 32(2) 1990: Animal Models in Biomedical Research

Online Issues

<< All Back-issues

<< This Issue's Table of Contents

ILAR Journal V32(2) 1990
Animal Models in Biomedical Research

Modeling in Biomedical Research: An Assessment of Current and Potential Approaches
Applications to Studies in Cardiovascular/Pulmonary Function and Diabetes
A National Institutes of Health Conference, May 1-3, 1989

Dear Readers: In May 1989, the NIH Division of Research Resources, Division of Research Services, and Office of Medical Applications of Research held a conference on the topic of modeling in biomedical research. After hearing presentations on the subject by scientific experts, an independent panel issued a report containing conclusions and recommendations. The panel's report has been printed, and the NIH has requested that the executive summary be printed in ILAR News. Copies of the full statement from this conference can be obtained from the Office of Medical Applications of Research, NIH, Building 1, Room 260, Bethesda, MD 20892 (301/496-1143).
-- Editor

Executive Summary

Models are indispensable for biomedical research. There is no branch of life science or medicine in which the current knowledge base is not determined in some way by the results of research with models. The status reports presented at this conference, representing two of the most active subdisciplines--cardiovascular/pulmonary physiology and pathology and the attack on diabetes--are eloquent testimony to that assertion. These examples of outstanding research highlight another important point: Progress in the war against these and other diseases depends not only on a steady flow of insights from research employing models but also on research based on a variety and often a combination of models.

The two groups of diseases singled out for special consideration in this conference illustrate the case. Progress has been made in reducing the toll taken by cardiovascular disease, in part because of insights gained through mathematical analysis and computer simulation of the cardiac cycle. New studies in the comparative anatomy of myocardium and advances in the biophysics and molecular biology of channels and receptors have contributed to this progress. These successes have been based in small part on study of simple physical analogs of the heart. And always and without fail, progress has resulted from submission of modeling such results to the test of validity in the intact mammal--the last stopping point before application of new knowledge and therapies to the situation in man.

The same is precisely true for diabetes mellitus. This group of diseases, in which the fundamental mechanisms remain elusive to this day despite decades of intense study, is nevertheless better understood today than ever before, with the possibility of prevention now apparently realistic. Such understanding has resulted from the closer collaboration of clinicians, basic scientists, and theorists, whose computational work has been either the goal of new and incisive observations on the disease in animal models and human victims or the explanation of hitherto enigmatic phenomena associated with the disease state.

It follows, therefore, because cardiovascular disease and diabetes mellitus are not likely to be fundamentally different from other categories of human pathology, that models and the ideas derived from them are inextricably woven into the fabric of knowledge and practice in the biomedical sciences. The future of biomedical research depends on an even denser intertwining.

It is no longer practical to design drugs for human and veterinary use without the aid of sophisticated computer modeling, including the most advanced graphics. The physiological compartment models on which so much of our understanding of complex control mechanisms is based cannot be imagined or tested without mathematics and computing. The molecular analysis of signals and gates controlling flows into and out of compartments depends on experiments with lower animals or molecules derived from them. The setting of treatment protocols with drugs depends on prior knowledge gained from animal screening and testing. In the end, the validity of every proposal about the nature and mitigation of human disease must be verified by appropriate testing in an appropriate mammalian model system.

This last is a critical point. Some advances in modeling of the past decade, driven by explosive growth of computing power and molecular biology, have allowed reduction in the number of vertebrate animals required in certain systems for the development of drugs. The more there is of design and the less of trial and error, the more directly the results of research can be applied to man. The manifold costs of higher animal testing can be reduced. But those same triumphs of modeling are simultaneously creating brilliant opportunities for new kinds of research and therapeutic intervention. These opportunities themselves call for validation in the appropriate mammalian models and eventually by means of appropriate clinical trials in man. The evidence of this resides in nearly every case of a medical "breakthrough" since the 1960s.

Therefore, it is not possible to predict the consequences for the number of mammals used in research of current advances in theory building and analysis. The writing of computer programs, isolation and cloning of genes, proliferation of cultured cell types that carry out differentiated functions in vitro, prediction by equation of complex control outcomes in whole animals--all of these will become, in the decade ahead, the tools of most biomedical research groups. But it is extremely unlikely that these remarkable new tools will substitute to any significant extent for experimental vertebrate animals. The tools will unquestionably help to reduce the toll of human suffering. Continued improvement of the techniques by which experimental animals are cared for and employed in research will unquestionably improve their lot. But we cannot now predict that the numbers of animals needed will decline. It is much more likely, in fact, that the numbers required will remain unchanged so long as the manpower engaged in biomedical research and the intensity of effort devoted thereto remain unchanged.

Overall and quantitatively speaking, simple model systems from the physical analog or the differential equation to the particularly suited invertebrate animal will not provide "alternatives" to mammalian experimentation. They will not reduce the quantity of research on higher animals. What they are providing and will provide in even greater abundance during the decades to come are new insights and opportunities undreamed of earlier for the alleviation of human suffering caused by disease.

It is, therefore, our first recommendation that the National Institutes of Health (within its intramural and extramural programs) and other agencies charged with the support of biomedical research seek new means and create new programs to encourage theoretical biology, to support new collaborations and new models, and to catalyze their application to the attack on disease that is the hallmark of developed societies and their obligation to the developing world. We urge interagency collaboration across the federal government to accomplish this objective.

Our second recommendation is as much to our colleagues--scientists, physicians, administrators--as to the agencies of government. It is that we join, finally, in responding with the truth about animals in research to the misinformation and disinformation that has been so widely distributed and has been given some currency in the media. We hold the truth to be:

Research animals are indispensable for the progress of human and veterinary medicine and the maintenance of human and animal health.
Although the numbers of such animals needed in the near term for urgent research may not rise, they are not likely to fall significantly.
The advance of modeling and model systems enhances science; it will not substitute for animal research and testing.