The Stent Controversy Continues
Five new studies add to the evidence base comparing the benefits and risks of drug-eluting and bare-metal stents, but they dont resolve the controversy.
Recent reports comparing drug-eluting stents (DES) with bare-metal stents (BMS) have suggested potentially greater risks for stent thrombosis (Journal Watch Cardiology Dec 20 2006) and for death and MI (Eur Heart J 2006; 27:2784) with DES. However, very low event rates with both types of devices and variable definitions of stent thrombosis complicate comparisons within individual trials. Five new studies enhance the evidence base on both sirolimus-eluting stents (SES) and paclitaxel-eluting stents (PES).
SES and PES Safety at 4 Years: Pooled Comparisons with BMS
Two research groups pooled randomized-trial data to study stent safety at 4 years: One analyzed data from 1748 patients in four trials of SES versus BMS; the other analyzed those data plus findings from 3513 patients in five trials of PES versus BMS.
The mortality rate was nonsignificantly higher with SES than with BMS overall (6.7% vs. 5.4%, P=0.28) but significantly higher within the diabetes subgroup (12.2% vs. 4.4%, P=0.008). MI rates did not differ significantly between SES and BMS recipients. Compared with BMS, SES were associated with more stent thromboses, as defined by the angiographic study protocol (10 vs. 5 thromboses in all, P=0.20; 5 vs. 0 after 1 year, P=0.025) and as defined by broader Academic Research Consortium (ARC) criteria (30 vs. 28 in all; 23 vs. 14 after 1 year).
PES and BMS recipients had similar rates of death or MI. Nonsignificantly more protocol-defined stent thromboses occurred with PES than with BMS (20 vs. 14 in all). After 1 year, stent thromboses were significantly more common with PES than with BMS (9 vs. 2 thromboses; P=0.028). Target-vessel revascularization (TVR) was significantly less common with both SES and PES (hazard ratios, 0.38 and 0.62, respectively) than with BMS.
Rates of death, MI, and late stent thrombosis did not differ significantly between PES and SES recipients. Some late stent thromboses even occurred in DES recipients on dual antiplatelet therapy.
SES vs. BMS in 14 Randomized Trials
A third group pooled data from 4958 patients in 14 SESBMS randomized trials that had mean follow-up of at least 1 year (maximum, 5 years). SES and BMS recipients had similar risks for death and for death or MI combined. The two groups had similar overall rates of protocol-defined stent thrombosis; rates were slightly lower with SES than with BMS before 1 year but significantly higher after 1 year. Among diabetes patients, mortality was nonsignificantly higher with SES than with BMS.
Protocol-Defined vs. ARC-Defined Stent Thrombosis
A fourth group focused specifically on stent thrombosis, according to both ARC and study-protocol criteria, in eight randomized DESBMS trials (4 SES, 4 PES). ARC criteria categorize the thromboses (as definite, probable, or possible) and count blindly adjudicated clinical events without angiographic proof. The number of thromboses during 4 years of follow-up was lowest by protocol definitions, next lowest by definite or probable ARC designations, and highest by any ARC designation. According to all three definitions, overall stent-thrombosis rates were similar with DES and BMS. However, after 1 year, rates tended to be higher with SES than with BMS by all definitions, and they tended to be higher with PES than with BMS by protocol definitions and by definite or probable ARC designations. These authors calculated the percentage of deaths attributable to stent thrombosis at 7% to 8% with DES and 6% to 11% with BMS.
DES vs. BMS in a Swedish Registry
In a nonrandomized study, researchers compared all 6033 DES recipients and all 13,738 BMS recipients in Sweden during 2003 and 2004. The DES recipients tended to be younger and more often female, and had a higher prevalence of diabetes and multivessel disease. Multivariable-adjusted mortality risk was significantly higher with DES than with BMS starting at 6 months: 18% higher from baseline to 3 years, 32% higher from 6 months to 3 years.
Together, the four studies of 19 randomized DESBMS trials (with up to 5 years of follow-up) yielded these major findings:
- A slightly lower risk for stent thrombosis with DES than with BMS before 1 year but a significantly higher risk after 1 year, representing an absolute disadvantage with DES of about 0.5% to 1.0%
- No difference in stent thrombosis between SES and PES
- Documentation that even some DES recipients on long-term dual antiplatelet therapy experienced late stent thrombosis
- No statistically significant difference in death or MI between BMS and either SES or PES, nor between SES and PES
- A marked reduction in TVR with DES, particularly SES, with no evidence that a need for TVR surfaces later with DES than with BMS
- A tendency toward higher mortality with SES than with BMS in diabetes patients
Two accompanying perspective articles, one from experts at the FDA, note that these randomized trials tended to enroll patients with the simplest lesions. The lack of mortality difference between DES and BMS recipients (despite increased risk for late stent thrombosis with DES) is potentially attributable to two phenomena in BMS recipients: (1) cardiac events associated with restenosis and (2) censoring of stent thrombosis data after treatment for restenosis.
Current usage patterns suggest that as many as 60% of stent procedures are performed for off-label indications such as saphenous-vein graft lesions, bifurcations, acute MI, chronic occlusions, and multivessel disease situations in which the risk for stent thrombosis may be particularly high. Perhaps this fact (and not merely randomized trials lack of power to assess mortality differences) helps to explain the Swedish observational finding of significantly greater risk for death with DES than with BMS after 1 year.
For now, DES should be reserved for patients at greatest risk for restenosis who can receive long-term dual antiplatelet therapy. Such therapy is no guarantee against late stent thrombosis, but stopping it prematurely is a risk factor (see recent AHA advisory: Journal Watch Cardiology Jan 31 2007). Although patients without approved DES indications might have the most to gain from restenosis prevention, off-label DES use poses potentially high risk, so physicians must exercise excellent judgment before using DES off label. We need larger, longer-term randomized trials that enroll patients with more-complex lesions to understand how best to use this important technology.
Howard C. Herrmann, MD
Published in Journal Watch Cardiology February 12, 2007
Spaulding C et al. A pooled analysis of data comparing sirolimus-eluting stents with bare-metal stents. N Engl J Med 2007 Mar 8; 356:989-97.
Stone GW et al. Safety and efficacy of sirolimus and paclitaxel-eluting coronary stents. N Engl J Med 2007 Mar 8; 356:998-1008.
Kastrati A et al. Analysis of 14 trials comparing sirolimus-eluting stents with bare-metal stents. N Engl J Med 2007 Mar 8; 356:1030-9.
Mauri L et al. Stent thrombosis in randomized clinical trials of drug-eluting stents. N Engl J Med 2007 Mar 8; 356:1020-9.
Lagerqvist B et al. Long-term outcomes with drug-eluting stents versus bare-metal stents in Sweden. N Engl J Med 2007 Mar 8; 356:1009-19.
Maisel WH. Unanswered questions Drug-eluting stents and the risk of late thrombosis. N Engl J Med 2007 Mar 8; 356:981-4.
Farb A and Boam AB. Stent thrombosis redux The FDA perspective. N Engl J Med 2007 Mar 8; 356:984-7.
Curfman GD et al. Drug-eluting coronary stents Promise and uncertainty. N Engl J Med 2007 Mar 8; 356:1059-60.
To ensure that your Reader Remark is not formatted as one long paragraph, precede new paragraphs with either a blank line or an indentation.