|
|
|
|
|
|
|
agent like amiodarone that had beta blocking and some sodium channel blocking properties as well. Why industry went off on almost a quest for pure, specific potassium channel blockers is an interesting example of herd psychology. This is an area where animal models could have been developed to test the hypothesis of which drugs should be sought for development and which would be optimum for clinical testing. I believe this was not done. This is not an indictment of the pharmaceutical industry, but one of the entire academic community. The development of preclinical in vivo models is under emphasized. Could we not develop a CAST arrhythmia model in rats to look at post-MI mortality? Could not proarrhythmia be assessed in an atrial fibrillation animal model: It seems that counting bodies in rats, mice, or guinea pigs, for that matter, or other species is far less expensive and far less disturbing than assessing the impact of these agents on mortality and clinical trials. Clearly, the model needs to be validated, but this is certainly possible when the clinical outcome has been assessed in at least one large-scale clinical study. Another example is the ACE inhibitors and their use with post-MI and CHF patients. Indeed some of pioneering work was done by investigators accomplished in the animal laboratory. Besides clinical studies, an animal model does exist. But when the A-II receptor antagonists come to the fore for development, they are not evaluated in terms of mortality in the existing animal model. It would be useful to evaluate the A-II antagonists in this model since there are differences between the ACE inhibitors and the A-II antagonists. One difference is that the ACE inhibitor increases bradykinin while the A-II receptor blocker antagonizes the effects of angiotensin without changing the amounts of bradykinin in the system. This difference may be immaterial or bradykinin can even be deleterious and the A-II antagonists could show greater beneficial effect. Testing in an animal model though would go a long way toward testing this hypothesis and helping us design perhaps clinical studies to evaluate the A-II effects. Undoubtedly, animal studies employing models need to be validated, but they also need to be employed frequently to aid in determining an optimum development strategy. In models varying dose, varying the combination of drugs is feasible. In a definitive clinical mortality trial one dose of the therapy is probably all that is possible. |
|
|
|
|
|
|
|
|
Another area in which modeling could be so very useful is the use of surrogate endpoints. Whether it be blood-pressure reduction, lipid lowering, VPC suppression, or thrombus dissolution, the endpoint often used clinically and the one that is so readily obtainable clinically is a surrogate endpoint. Surrogate endpoints are an important basis for drug development, but one that is quite precarious. Targeting PVCs for developing an antiarrhythmic in the 90s post-CAST is of no value and, in fact, would be the subject of ridicule. All surrogates are suspect and need to be validated. The possibility of validation, even if only partial, with animal studies is an area needing further evaluation. |
|
|
|
|
|