No drug-eluting stent stands out from the pack and very few conclusions can be drawn regarding their comparative effectiveness, according to Stephen Ellis, MD, head of invasive and interventional cardiology at the Cleveland Clinic, who spoke at the Cardiovascular Revascularization Therapies (CRT) 2008 meeting in Washington, D.C., last week.
Ellis reviewed the lifecycle of innovative technologies, in which he tracked its cycle from unbridled enthusiasm to rapid adoption to untoward complications to second wave effect, where he included drug-eluting stents (DES), which finally leads to realistic applications (where data outweighs perception).
In a meta-analysis of all reported results from six different trials, which appeared in 2007 in the Annals of Internal Medicine, Ellis evaluated five-year survival rates in diabetics. Out of the 233 patients who underwent PCI, 80.7% survived, and out of 266 patients who underwent CABG, 82.7% survived, which is not statistically significant.
He also referenced the COURAGE trial, which evaluated PCI for survival angina in 2,287 class I-ll patients. The PCI success rate was 89%, and survival for MI in the long-term produced no difference.
A separate German meta-analysis found that 17 studies with 7,513 patients comparing PCI to medical therapy produced 20% relative reduction for PCI’s impact on mortality. These findings contradicted the finding of the COURAGE trial.
An important lesson Ellis learned is to be aware of placing too much importance on one trial.
In the 2007 Lancet collaborative network meta-analysis, the investigators examined 38 trials of 18,023 patients, comparing sirolimus-eluting stents to paclitaxel-eluting stents to bare-metal stents, and each to the other. In the cumulative incidence of target lesion revascularization, BMS had 22%, while PES had 9% and SES 7%.
More importantly, Stettler and colleagues found that the rates of mortality, cardiac death and MI were no different after 4 years in any of the stents. In the same meta-analysis, the cumulative incidence of ARC definite stent thrombosis at four years indicates a stent thrombosis rate of 2.5% for PES, whereas the BMS and SES rates were nearly identical and slightly lower.
Based on the conflicting evidence from so many different trials, Ellis came up with four additional lessons:
- More data isn’t always good, if you can’t understand it;
- Beware of tenuous underlying assumptions, such as BMS are all the same;
- Beware of looking at point estimates, (look at and understand confidence intervals); and
- Sometimes you just have to use the best available data…and sometimes you just have to wait until we have more available data.
Ellis also highlighted several questions and vulnerabilities of DES:
- Real world results will be the same as randomized controlled trials with simple lesions (Endeavor?)
- Xience/Promus long term results will mimic Cypher’s
- ARC definitions are the best (are all MIs in stented vessels stent related?)
- Just because we haven’t seen stent thrombosis in Cypher/TAXUS trials after year four, does that mean that is the end of the risk period?
- How do we best manage patients with DES undergoing surgery?
“DES are neither as good as we thought nor as bad as the naysayers and media made them out to be,” Ellis concluded.