ORLANDO — The largest study of coronary stenting ever conducted has documented significantly better clinical outcomes with drug-eluting than with bare-metal stents.
The observational study involving 262,700 Medicare patients with 30-month follow-up found a 25% decrease in deaths and a 23% reduction in myocardial infarctions in drug-eluting stent (DES) recipients after adjustment for 102 potential confounding variables.
Moreover, there was no hint of a late increase in stent thrombosis in the 83% of patients who received DES, contrary to the findings in a number of earlier, far smaller randomized trials.
“The concerns about the safety and effectiveness of these devices, I think, have been laid to rest,” Dr. Pamela S. Douglas said in presenting the study findings at the annual meeting of the American College of Cardiology.
“We feel that our sample size of over a quarter-million patients makes our data much more robust than what's been out there, and we can now say with some confidence that drug-eluting stents are not killing people,” added Dr. Douglas, professor of research in cardiovascular medicine at Duke University, Durham, N.C., and director of the cardiovascular imaging program at the Duke Clinical Research Institute.
The most remarkable feature about this study, according to many observers, was its unique design. The study was carried out by linking data from the ACC-National Cardiovascular Data Registry to Medicare claims data without using patient names or other individual identifiers. This was accomplished through sophisticated statistical techniques involving a weighted probability matching process.
Adjusted rates of revascularization and bleeding were 9% lower in patients who received drug-eluting stents than in those who received bare metal stents. Rates of late stent thrombosis could not be looked at directly, so the investigators analyzed the incidence of ST-elevation myocardial infarction as a marker. They found no late excess in DES recipients.
Small randomized trials have shown no significant reduction in rates of death or MI but a large decrease in coronary revascularization with DES—just the opposite of the findings in this huge real-world study. Because of this discrepancy, Dr. Douglas and coworkers did a subanalysis limited to the less than 20% of patients in their study population who fit the eligibility criteria for the major stenting randomized trials. The results were “exactly the same” as in the full study, according to the cardiologist.
Although the huge patient numbers involved in the new study were dazzling, not everyone found the results compelling.
“I didn't see anything terribly illuminating here. In every registry the people who get the drug-eluting stents do better. But why do they do better? That's the real question,” Dr. Spencer B. King III said in an interview.
The answer, he added, almost certainly lies in unmeasured confounding variables that simply cannot be controlled for in an observational study, no matter how big.
For example, the study database did not contain information on long-term use of dual antiplatelet therapy including clopidogrel (Plavix), but the use of such therapy was surely greater in DES recipients, which could explain their lower rates of mortality and MI.
And selection bias was likely at work, too; some patients are probably selected for bare metal stents because they're sicker, unable to afford costlier DES, or deemed unlikely to comply with dual antiplatelet therapy, according to Dr. King, executive director for academic affairs at Saint Joseph's Health System in Atlanta.
Discussant Donald E. Cutlip of Beth Israel Deaconess Medical Center, Boston, also cited confounding issues as a study limitation, but said that the study nonetheless carries an important message: “It's not that drug-eluting stents should now be used in all patients because of their improved safety, but that with appropriate device selection, using bare metal stents in the right patients, we can still treat a large proportion of patients with drug-eluting stents safely, including possibly with reductions in death and MI,” Dr. Cutlip said.
The study was published simultaneously with Dr. Douglas's presentation (J. Am. Coll. Cardiol. 2009;53:1629–41).
The study was sponsored by the Agency for Healthcare Quality and Research. Dr. Douglas reported having no financial conflicts of interest.
'Our sample size of over a quarter-million patients makes our data much more robust than what's been out there.' DR. DOUGLAS
ELSEVIER GLOBAL MEDICAL NEWS