Readjudication Finds Higher Event Rates for BMS, DES
Changes in standard thrombosis definitions, timing yield event rate changes.
Readjudication of pooled data from studies carried out on the Endeavor and Cypher stents has resulted in higher event rates for bare-metal and drug-eluting stents.
Donald E. Cutlip, MD, an interventional cardiologist at Beth Israel Deaconess Medical Center, presented readjudicated data on stent thrombosis after drug-eluting stenting in a session yesterday afternoon. The reclassification was carried out by researchers from the clinical research institutes at Harvard and Duke and the Cardiovascular Research Foundation, as well as groups in Europe, the FDA and members of industry.
Redefinitions
The researchers proposed three standard definitions for thrombosis: definite/confirmed (acute coronary syndrome, and either angiographic confirmation of thrombus/occlusion or pathologic confirmation of acute thrombosis), probable (unexplained death within 30 days and target vessel MI without angiographic confirmation of thrombosis or other identified culprit lesion), and possible (unexplained death after 30 days).
Event timing was recategorized as early (including acute events within 24 hours and subacute events from one day to 30 days), late (30 days to one year) and very late (more than one year).
Clinical trial data were pooled from the Endeavor and Cypher studies, including Endeavor I, II and III and the Continuing Access Registry, and RAVEL, SIRIUS, E-SIRIUS and C-SIRIUS.
Pooled Endeavor data
After reclassification and in terms of cumulative incidence out to three years, proportionally more events were added in the bare metal stent groups than in the Endeavor DES groups; the difference in event rates was significant (1.0% vs 3.3%; P = 0.01).
Most of the DES events added were defined as possible, including five added events after 30 days (from 1,318 patients). Six events in the bare metal stent arm were added (from 594 patients).
Only one additional TLR thrombosis in each arm was added after reclassification.
“The overall increase [in events] is driven mostly by increased late and very late ‘possible’ events, with definite or probable events similar to prior reports using protocol definitions and trending lower for the DES arm,” Cutlip said.
Pooled Cypher data
After reclassification at the end of four years, event rates remained similar, but events were added to both groups and there was no significant difference between the two (sirolimus-eluting, 3.5% vs. bare metal, 3.3%).
“There was also an apparent early hazard with the bare metal stent that persists out to two years, where the rates were higher than with the drug-eluting stent, with the increased proportionate number of events with the drug-eluting stent beginning just after one year leading to the equal rates after four years,” Cutlip said.
The increase in events was driven by the addition of ‘possible’ stent thromboses; 16 of 25 new events in the drug-eluting stent arm and 13 of 25 events in the bare metal stent arm.
Reclassification to include TLR resulted in 10 newly added stent thrombosis events in the bare metal stent patients after prior TLR. “Our statisticians are busy at work with more specific analyses to define specifically how prior TLR should be treated in these patients,” Cutlip said.
“The cumulative incidence of any stent thrombosis or definite plus probable stent thrombosis out to four years was not different between Cypher and the Bx Velocity bare metal stent,” Cutlip concluded.