December 21, 1998: Garret A. FitzGerald, University of Pennsylvania
Response to AHRQ on CERTs
We welcome this initiative and wish to suggest topics and approaches to implementation as solicited in the Federal Register of November 3, 1998.
We believe that this mechanism affords a unique opportunity to integrate the efforts of basic and clinical pharmacology with health services research and drug information services to refine the quality of health care while reducing its cost , ultimately by refining the costs of drug development.
Existing structures at the University of Pennsylvania including the Center for Experimental Therapeutics, the Leonard Davis Institute, the Departments of Pharmacology and Epidemiology and Biostatistics, the General Clinical Research Center and the Drug Information Service will collaborate in this effort. A newly developed Center for Pharmacogenetics will address issues relating to genetic—environmental interactions relevant to interindividual differences in drug response.
We propose a CERTs focused upon a specific disease entity—Cardiovascular Disease—the nation's leading cause of death . We shall integrate the expertise of the aforementioned structures within this disease modality to achieve the objectives of the CERTs program. This can best be illustrated by two examples:
The effectiveness of statins in reducing cardiovascular morbidity and mortality has been remarkable. However, given that little information is available as to the comparative efficacy of the individual statins and the disparity in their cost, this has major pharmacoeconomic implications. Manufacturers of statins are currently placing emphasis on identifying relevant mechanisms of action distinct from lipid lowering which may afford cardiovascular benefit.
We have recently developed novel methodology, based on GCMS analysis of isoprostanes, for the study of oxidant stress in vivo. We have shown that isoprostane excretion is enhanced in both murine and human hypercholesterolemia and that tissue levels are elevated in the atherosclerotic vasculature. Using vitamin E at doses titrated to suppress isoprostane formation, we demonstrated the functional importance of lipid peroxidation in murine atherogenesis despite sustained hypercholesterolemia.
These observations raise several questions:
- Do statins exert an antioxidant effect in vivo and if so, is this a class effect related to their potency in reducing cholesterol?
- Will addition of a rationally selected dose of an antioxidant, such as vitamin E, add to the antiatherosclerotic effects of a statin?
These questions may now be addressed experimentally in humans using isoprostane analysis and novel imaging methodology, such as ultrafast CT of human coronaries in studies based on the GCRC, as well as in further studies in mouse models of atherosclerosis. Meantime, studies emanating from the LDI can assess relative clinical efficacy and effectiveness of individual statins using metaanalyses and case control studies. Furthermore, these instruments can also be used to address the vitamin—statin interaction. Basic and clinical studies can provide information relevant to modeling. Outcomes of such studies and our clinical and basic research may also formulate hypotheses that can be addressed in prospective clinical trials. Processing of this information and its dissemination by the Drug Information Service will be expanded by the development of a website, a newsletter and the development of an integrated Experimental Therapeutics/Health Services Research Seminar Series—the CERTs Series—which will include lectures targeted at the general public, as well as practitioners. Such a series will also be videotaped and distributed on the Web.
Cyclooxygenase (COX)-2 inhibitors
These drugs are likely to be approved in the new year for the treatment of arthritides. We have recently found that they suppress vascular production of prostacyclin (PGI2) without concomitant platelet inhibition. PGI2 is a potent inhibitor of platelet aggregation, vasoconstriction and vascular proliferation in vivo and the implications of this observation for treatment of target age groups at risk of cardiovascular disease is unknown. Our studies indicate that this is a class effect; the mechanism appears to be that physiological rates of shear stress lead to "constituitive" upregulation of COX-2 in endothelial cells.
The trials leading to approval of these drugs in arthritis are roughly an order of magnitude too small to detect a reasonable cardiovascular risk, so this is likely to emerge only from prospective trials designed to address this issue, metaanalyses of arthritis trials and postmarketing surveillance. These instruments may be applied within the context of a CERTs by investigators in the Leonard Davis Institute and the Center for Epidemiology and Biostatistics, which can make assessments of (i) clinical effectiveness and relative cost effectiveness as well as (ii) postmarketing surveillance respectively.
The CERTs can integrate these efforts with basic studies in murine models of COX deficiency and atherosclerosis and GCRC based studies of the effects of COX-2 inhibitors on vascular function in patients at risk. Again, information from this integrated approach can be disseminated as described above. Aside from the value of assessing cardiovascular risk, this model may also be relevant to other aspects of COX-2 inhibitor usage. Thus, although these drugs may be less likely to cause GI ulceration de novo, COX-2 is upregulated in healing ulcers and the implications of its inhibition in patients who have existing ulcers when they start taking these drugs in unknown. Similarly, COX-2 is widely expressed in kidney and brain. However, the potential risks of COX-2 inhibition on renal and central function in patients with preexisting disorders will only become evident after their approval for arthritis. A CERTs would provide an excellent structure for integrating basic, clinical and health services research in an effort to address these issues.
While these two examples reflect the potential of a CERTs to harness the variegate expertise necessary to address issues of drug efficacy and effectiveness, particularly as an entity transitions from the context of clinical trials to use in the real world, they also highlight the iterative process of continued drug development that a CERTs can foster, by integrating pharmacological and health services research. Additional to the structures mentioned above which render the University of Pennsylvania a unique environment in which this can be achieved are an integrated Health System covering more than 1 million lives; a comprehensive international clinical trials network (the EUPENN Network) for mechanism based phase 1 and 2 studies of cardiovascular drugs and the track record of coordinated effort by the Centers for Experimental Therapeutics and Clinical Epidemiology and Statistics, the GCRC and the Leonard Davis Institute in our recent submission of an application for a K30 award.
We hope that these suggestion are helpful as you deliberate on the final CERTs structure and look forward to applying to AHCPR for funding for a CERTs at this institution.
Garret A. FitzGerald, M.D.
Department of Pharmacology
University of Pennsylvania
153 Johnson Pavillion
3620 Hamilton Walk
Philadelphia, PA 19104-6084
Telephone: (215) 898-1185
Fax: (215) 573-9135