DALLAS — Paclitaxel-eluting coronary stents posed a substantially increased risk of causing late stent thrombosis, compared with bare-metal stents, in a metaanalysis of five studies with more than 3,500 patients.
The results were less clear for sirolimus-eluting coronary stents, with no statistically significant differences in late thrombosis rates, compared with bare-metal stents, said Dr. Anthony A. Bavry, who reported on the metaanalysis at the annual scientific sessions of the American Heart Association.
The clearly increased risk linked with paclitaxel-eluting (Taxus) stents contrasted with the results of a similar metaanalysis that was reported by Dr. Bavry and his associates at the Cleveland Clinic a year ago. In that metaanalysis, the researchers failed to document a clear-cut difference in risk.
The new analysis included follow-up of as much as 3–4 years, which allowed for the collection of very late events. This was in contrast to the earlier report, which had a follow-up of only 1 year.
The researchers used the results from a total of 18 randomized, controlled comparisons between drug-eluting stents and bare-metal stents, as well as the results from two registries.
The trials included four studies that examined nonpolymeric paclitaxel-eluting stents in a total of 1,423 patients, nine studies that examined sirolimus-eluting (Cypher) stents in a total of 3,162 patients, and five studies with polymeric paclitaxel-eluting stents in 3,513 patients.
The two registries included a total of 2,171 patients.
The incidence of stent thrombosis more than 30 days after stent placement among those patients who received polymeric paclitaxel-eluting stents was 6.3/1,000 patients, compared with 1.1 events/1,000 patients in the bare-metal stent arms of these trials—a statistically significant relative risk for paclitaxel-eluting stents of 3.6, Dr. Bavry reported.
The stent thrombosis incidence for the same time period in patients who received sirolimus-eluting stents was 3.0/1,000 patients, compared with 4.5/1,000 patients in the bare-metal stent arms, a nonsignificant difference.
When thrombotic events were limited to those that occurred more than 6 months after stent placement, the rate for paclitaxel-eluting polymeric stents was 5.0 per 1,000 patients, compared with 0 among patients who received bare-metal stents, a statistically significant relative risk of 7.1. In the studies in sirolimus-eluting stents, the rate for the drug-eluting stents was 3.0 per 1,000, compared with 1.2/1,000 among the bare-metal patients, a nonsignificant difference.
Thrombotic events more than 1 year after stent placement occurred at a rate of 5.7 per 1,000 patients with the paclitaxel-eluting stent, compared with 0 in the bare-metal stent arm, a statistically significant relative risk of 5.7. The difference between groups in the sirolimus stent studies, 3.5/1,000 vs. 0 for bare-metal stents, did not reach statistical significance.
Overall, no thrombotic events were seen more than 1 year following placement of a bare-metal stent in these studies.
In contrast, however, the median time to a thrombotic event was about 16 months in patients who received a sirolimus-eluting stent and about 18 months in those who got a paclitaxel-eluting stent.
“The polymer seems to be the culprit, leading to delayed arterial healing and an increased risk of late thrombosis,” added Dr. Bavry.
Because late thrombosis episodes are relatively rare, metaanalysis might be the best way to determine their incidence, he said.