Xience V drug-eluting stent Image source: Abbott Vascular

WASHINGTON, D.C.—The Xience V (Abbott) stent consistently had better outcomes regarding target vessel revascularization, target lesion revascularization, MI and stent thrombosis compared to Taxus Liberté (Boston Scientific) at two years, based on the single-center, all-comer COMPARE trial, presented as a late-breaking clinical trial at the 22nd annual Transcatheter Cardiovascular Therapeutics (TCT) conference Sept. 23. Lead author Peter Smits, MD, told Cardiovascular Business News that both the first- and second-generation Taxus stents are “inferior” to Xience.

Smits and colleagues from Maasstad Ziekenhuis in Rotterdam, the Netherlands, sought to determine the outcome of second-generation drug-eluting stents (DES) in a randomized study designed to "reflect everyday clinical practice." The study only used symptom-driven clinical endpoints.

“It is not surprising that the Taxus Liberté is performing similarly to the Taxus Express. Even though it is considered a second-generation stent because the struts are thinner, the polymer, the drug and the drug-release program are still the same. Therefore, the alloy and strut thickness do seem to be influencing outcomes,” Smits said.

Slideshow | COMPARE Trial: 2 year follow up

Peter Smits Maasstad Ziekenhuis Rotterdam

The researchers included all patients eligible for PCI with a life expectancy of more than five years. Patients were excluded if they could not take dual-antiplatelet therapy for 12 months; presented with cardiogenic shock; had planned major surgery within one month; were participating in another trial; or would not provide informed consent.

The primary endpoint was all death, nonfatal MI and target vessel revascularization (TVR) at 12 months of follow-up. The secondary endpoints were cardiac death, non fatal MI, ischemia-driven target lesion revascularization (TLR) rate annually during the five years of follow-up; all death, non fatal MI, TVR at two- to five-year follow-up; and incidence of definite, probable or possible stent thrombosis annually during the five years of follow-up.

Initially, 903 patients received the Taxus Liberté, with 902 at one year of follow-up and 900 at two years of follow-up. Initially, 897 patients received the Xience V, with 895 at one year and 895 at two years—equating to an overall two-year follow-up rate of 99.7 percent.

A summary of the 12-month primary endpoints, which were published in Lancet in January, showed that the lesion-specific composite endpoint, thrombosis and MI all were significant lower in the Xience V stent group.

In the subgroup analysis, a trend was observed toward better outcome with the Xience V stent in almost all subgroups, apart from the diabetic population—an outcome that was similarly observed in the one-year results of the Spirit IV trial, presented at TCT 2009, and later published in New England Journal of Medicine.

At the end of two years, 15.2 percent of the Taxus group and 11.4 percent of the Xience group were on dual-antiplatelet therapy—a statistical significant difference—even though clinical practice guidelines only recommend one year.

For the primary endpoint, the significant difference of 2.9 percent in major adverse coronary event (MACE) rates (all death, nonfatal MI and TVR) at 12 months increased to 4.7 percent at two years follow-up, with a 33 percent relative reduction risk. At two-years follow-up, the MACE rate was 13.7 percent in the Taxus group versus 9 percent in the Xience group, which is “a highly statistically significant difference,” said Smits.

For the secondary endpoint of composite cardiac death, nonfatal MI and ischemia-driven TLR, the difference in event rates between both stent groups gradually increased over time with a significant difference of 3.3 percent at one year, which increased to 4 percent at two years, with a 11.4 percent MACE rate in the Taxus group compared with 7.4 percent in the Xience group, which is “a highly significant difference” with a p value of 0.0038, he reported.

Definite and probable stent thrombosis rates, according to the ARC definitions, which was already “highly significantly different” between both groups, with 2.6 percent in the Taxus group versus 0.7 percent in the Xience group at 12 months, increased to an absolute difference of 3 percent at two years of follow-up. Particularly, Smits noted, that the low stent thrombosis rate of less than 1 percent in the Xience group occurred with only 11 percent of the patients on dual-antiplatelet therapy.

However, all death and cardiac death rates were similar in both groups.

Also, TVR and ischemia-driven TLR rates were significantly higher in the Taxus stent group compared to the Xience V stent group—7.7 versus 3.1 percent and 5.9 versus 2.6 percent, respectively. “Again, these results that showed an early difference gradually increased over time,” Smits said.

In this all-comer trial, “reflecting a real-world patient population, the major secondary endpoints at two years showed superiority of Xience V versus Taxus Liberté [p = 0.0016],” he concluded.

“In our practice, we stopped using Taxus more than a year ago, after we observed the one-year data from COMPARE. Even for diabetics, both stents are essentially equal, so there is no reason to use Taxus,” Smits said.