This is the first large-scale,  randomized head-to-head trial of Endeavor vs. Cypher; 2,333 patients were included.The primary safety endpoints of this all-comers study were all-cause mortality, cardiac mortality, MI, and definite stent thrombosis (using the ARC definition) at nine months. Efficacy endpoints were clinically-driven TLR and clinically significant restenosis for the same period.

      At follow-up, patients in the Endeavor group had experienced significantly higher rates of restenosis (HR 6.59, 95% CI, 2.57-16.9; P<.0001) and TLR (see Figure 1) as well as MI (HR 3.47, 95% CI, 1.14-10.5; P=.03). Definite stent thrombosis also was more frequent in the Endeavor arm (see Figure 2). There was no significant difference in cardiac mortality (HR 2.17, 95% CI, 0.75-6.24; P=.14) between the two groups.

      Follow-up did not rely on control angiography or study-related patient contact but instead on patient-driven clinical events available through the comprehensive network of Danish registries and databases.

      Endeavor patients were more likely to have had prior PCI. Lesion characteristics in both  arms were comparable. About half of the patients in each group had stable angina; approximately 38% in each group had NSTEMI or unstable angina.

Results may represent a true biologic finding
David J. Cohen, MD, of Saint Luke’s Mid America Heart Institute in Kansas City, Mo., said the trial’s major strength is the lack of routine angiographic follow-up, eliminating potential bias from the so-called occulostenotic reflex. "SORT-OUT III is the first [study] to demonstrate the association between higher late loss and more frequent clinical restenosis in a real-world population," he said.

      The difference in outcomes between Endeavor and Cypher could be attributable to unreasonably low rates of stent thrombosis in the Cypher group, Cohen said, adding that the two types of stents could also have been applied differently because of a perceived safety advantage with Endeavor.

      On the other hand, one-year data from Endeavor IV showed similar stent thrombosis results, suggesting that the current finding may be real. This may well be a true biologic finding, Cohen said. The purported safety advantage of the Endeavor stent has been based on the virtual absence of late stent thrombosis after one year, well beyond the time frame of this study. This result implies that the biology of subacute and late thrombosis may differ from that of very late stent thrombosis.