Perspectives On Medical Research

Volume 5, 1995

Aping Science

A Critical Analysis of Research at the Yerkes
Regional Primate Research Center

2. Atherosclerosis

Another Yerkes research focus is atherosclerotic vessel pathology, such as coronary artery disease. As described in the Spring 1993 issue of Inside Yerkes,¹ Laurence Harker and colleagues' baboon experiments have aimed to elucidate 1) how atherosclerosis evolves in arteries, 2) how to prevent recurrent atherosclerosis after certain invasive treatments such as balloon angioplasty, and 3) how to avoid clot formation.^2-11 In contrast to other Yerkes publications, Inside Yerkes has not suggested that these studies have directly led, or will lead, to new knowledge of human disease. Instead, the publication states that primate studies are an important last step before applying treatments to patients.¹ However, as discussed in the Introduction, animal experiments cannot reliably test hypotheses about human reactions to stimuli such as drugs.

In defense of such experiments, Harker and colleagues have argued that baboons are needed to model clot-induced arterial occlusion (thrombosis) because

in general, experimental animal models of thrombosis and thromboembolization have had limited usefulness, particularly when the assessment of vascular occlusion has involved experimental animals of questionable relevance to humans and observations at a single point in time following an initiating vascular insult.³

They have stressed the need to use an "appropriate species." In fact, however, there is no a priori reason to expect that baboon experiments will be any more applicable to humans than other species, for, as Lewontin¹² has pointed out, the "closeness to humans" argument is unsound. And, similar experiments with other species have proven to be non-contributory. In the 1980s, 25 compounds were reported to reduce ischemic-stroke damage in nonhuman animals, but none proved effective in humans. David Withers et al. found that each time a treatment for experimental stroke has appeared effective in animal experimentation, it has prompted numerous other costly, wasteful clinical studies. The animal results, perhaps merely artifactual, did not predict human reactions.^13,14

Yerkes officials have argued that Harker and colleagues' baboon model is useful and reliable because it has accorded with human clinical findings that angiotensin converting enzyme (ACE) inhibitors do not prevent re-narrowing of surgically dilated blood vessels, unlike the results of certain rat experiments.¹ However, this anecdotal information hardly proves the point, for the Yerkes researchers' highly artificial model differs considerably from naturally occurring human thrombus formation.

While Inside Yerkes has suggested that this baboon model is relevant to coronary artery disease, in reality the thrombus produced in baboons differs fundamentally from human arterial atherosclerotic plaques, and, therefore, resembles the failed animal models of stroke reported in the 1980s.^13,14 The baboon thrombus was induced by first creating a shunt between a major artery and vein, and then placing into the shunt a device that included a Teflon chamber and a Dacron-lined silicone rubber stent.¹⁵ Importantly, unlike humans, these baboons did not have diseased blood vessels--the vessel wall cellular proliferation and fat accumulation that characterize human atherosclerosis. Following these structural manipulations, the experimental baboon thrombus developed in about an hour and was comprised of blood components, whereas human atherosclerotic plaques feature proliferation of vascular cells and develop over years.¹⁶ Therefore, it was not reasonable to assume that baboons' response to ACE inhibitors would reliably predict humans' response, and Yerkes' retrospective analysis of one drug's effect does not confirm the model's validity.

For every drug whose pharmacological action is found to be similar in nonhuman primates and humans, one can be found with dissimilar action. It cannot be concluded that a particular model is valid because it models correctly the actions of a single drug. Benoxaprofen, a non-steroidal anti-inflammatory drug, showed no toxicity in animal tests, including tests in rhesus monkeys. When used in humans, however, it was noted to produce liver failure and cutaneous photosensitivity. It was withdrawn in 1982 after 3500 documented cases of severe reaction and 61 deaths in the UK alone.¹⁷

Evidently recognizing their model's difficulties, Harker's group now maintains that it is analogous to a "venous-type thronibus."¹⁵ Even with this new way of conceptualizing their methodology, however, there are serious problems. The highly artificial baboon thrombus forms in an arteriovenous shunt, and such shunts resemble arteries more than veins. Also, the more gradual evolution of human thrombi over days or weeks, as contrasted with the more rapidly developing experimental thrombus, may have consequences with respect to symptoms, course, and treatment, and likely result in differences in composition and character.

Despite the inherent difficulties associated with this model, Harker and colleagues have also used such arteriovenous shunt systems^2,18 to test other potential anti-clotting agents, such as hirudin,⁷ anti-platelet antibodies,¹⁰ fatty acids found in fish oil,¹¹ and protein C.⁴ These arteriovenous shunts, they claim, were chosen for study "because of their clinical relevance, defined flow geometry, reproducibility of thrombus formation with predictable occlusion, and their resistance to aspirin therapy."⁷

Such a rationale does not validate the model’s applicability to humans. First, there is no clear evidence for this model’s "clinical relevance," Second, while many of the model’s supposed attributes relate to its controllability, there is no reason to expect that this highly artificial baboon condition faithfully reproduces human thrombus formation. Trading biological accuracy for greater "controllability" does not enhance applicability. Third, this model system may falsely favor new anticoagulant regimens over older, clinically proven therapies. Since, as the researchers have noted, their model system is resistant to both aspirin and heparin treatment,⁷ there is increased likelihood that they will "demonstrate" superior efficacy for other anti-coagulation therapies. Fourth, their drug protocol has often differed markedly from typical human exposures. For example, in the fish oil protocol, the investigators fed baboons 1 gram/day of a combination of fish oils, a dose which greatly exceeded that used in previous clinical trials.¹¹ Whether such massive dosing would be safe for humans, much less helpful in preventing clot formation, is unclear. Similar to the choice of model system, such a dosing policy increases the likelihood of demonstrating a "new and interesting" finding, but decreases the likelihood that the results would apply to humans. In conclusion, differences between the model system and normal human clotting, compounded by marked differences in experimental drug dosing compared to the clinical setting, make the research results virtually uninterpretable for humans.


1. Yerkes Regional Primate Research Center. Inside Yerkes. Atlanta, Spring 1993.

2. Bass A, Krupski WC, Hanson SR, et al. Exteriorized chronic aorto-caval arteriovenous access shunts in the baboon (Papio cynocephalus). Journal of Medical Primatology 1993;22:331-339.

3. Harker LA, Kelly AB, Hanson SR. Experimental arterial thrombosis in nonhuman primates. Circulation 1991;83(suppl IV):IV-41-IV-55.

4. Hanson SR, Griffin HJ, Harker LA, Kelly AB, Esmon CT, Gruber A. Antithrombotic effects of thrombin-induced activation of endogenous protein C in primates. Journal of Clinical Investigation l993;92:2003-2012.

5. Harker LA. New antithrombotic strategies for resistant thrombotic processes. Journal of Clinical Pharmacology 1994;34:3-16.

6. Gruber A, Harker LA, Hanson SR, Kelly AB, Griffin JH. Antithrombotic effects of combining activated protein C and urokinase in nonhuman primates. Circulation 1991;84:2454-2462.

7. Kelly AB, Marzec UM, Krupski W, et al. Hirudin interruption of heparin-resistant arterial thrombus formation in baboons. Blood 1991;77:1006-1012.

8. Cadroy Y, Maraganore JM, Hanson SR, Harker LA. Selective inhibition by a synthetic hirudin peptide of fibrin-dependent thrombosis in baboons. Proceedings of the National Academy of Sciences of the United States. 1991;88:1177-1181.

9. Krupski WC, Bass A, Kelly AB, Marzec UM, Hanson SR, Harker LA. Heparin-resistant thrombus formation by endovascular stents in baboons. Circulation 1990;81:570-577.

10. Cadroy Y, Hanson SR, Kelly AB, et al. Relative antithrombotic effects of monoclonal antibodies targeting different platelet glycoprotein-adhesive molecule interactions in nonhuman primates. Blood 1994;83:3218-3224.

11. Harker LA, Kelly AB, Hanson SR, et a!. Interruption of vascular thrombus formation and vascular lesion formation by dietary n-3 fatty acids in fish oil in nonhuman primates. Circulation 1993;87:1017-1029.

12. Lewontin RC. Primate models of human traits.

13. Wiebers DO, Adams HP, Whisnant JP. Animal models of stroke: Are they relevant to human disease? Stroke 1990;21:l-3.

14. Wiebers DO, Adams HP, Whisnant JP. Relevance of animal models to stroke [letter]. Stroke 1990;21:1091-1092.

15. Hanson SR, Harker LA, Kelly AB, Lumsden A. Evaluation of small vessel prostheses: Results of inhibition of angiotensin converting enzyme. Annual Progress Report of the Yerkes Regional Primate Research Center to NCRR/NIH. Atlanta, 1992, pp 179-180.

16. Benditt EP, Benditt JM. Evidence for a monoclonal origin of human atherosclerotic plaques. Proceedings of the National Academy of Sciences of the United States 1973;70:1753-1756.

17. US General Accounting Office. FDA Drug Review: Post approval risks 1976-1985. Washington DC, US Government Printing Office, 1990.

18. Hanson SR, Harker LA. Baboon models of acute arterial thrombosis. Thrombosis and Haemostasis 1987;58:801.