In a new meta-analysis Dr. Gregg Stone and members of the Drug-Eluting Stent in Primary Angioplasty (DESERT) Cooperation concluded that reduction in target-vessel revascularization (TVR) associated with drug-eluting stents (DES) in primary PCI provided a powerful reason for continued use of DES in primary PCI. An accompanying editorial by James Brophy focused on several potential DES weaknesses in the study. (Click here for our CardioExchange news story.) In this post Dr. Stone responds to the editorial.
The editorialist makes valid points about the precision that can be achieved from any meta-analysis, including ours. He ignores, however, that the reduction in TVR is robust, and the trend toward a reduction in mortality (with no evidence of a late risk for death) favors the interpretation that DES should be favored in STEMI. Many MIs are small and not of prognostic importance, and very late stent thromboses may not carry the prognostic import of earlier such events. Mortality thus trumps all. Moreover, these data were generated with first generation DES, and the network meta-analysis we recently published in the Lancet suggest that newer DES, especially fluoropolymer-based everolimus-eluting stents, may have lower rates of definite stent thrombosis than not only first generation DES, but also BMS. However, all meta-analyses are hypothesis-generating, and require testing in adequately powered randomized trials before being accepted as definitive. Clinical equipoise continues to be present in the clinical community regarding the utility of DES vs BMS in STEMI. Thus, later this year we will be initiating the HORIZONS-AMI II trial, in which 7,000 – 10,000 patients with STEMI will be randomized to PROMUS Element everolimus-eluting stents vs BMS, with this study powered to demonstrate not only improved clinical outcomes but a reduction in stent thrombosis with DES compared to BMS. If positive, this demonstration will truly be paradigm shifting.
You can follow any responses to this entry through the RSS 2.0 feed.
Both comments and pings are currently closed.
NEJM — Recent Cardiology Articles
I am literally stunned, that in the wake of COURAGE, SYNTAX, etc, and meta-analysis that show no benefit from DES in reducing cardiovascular events, that Dr. Stone still thinks a surrogate endpoint like TVR has any relevance whatsoever. Surrogate endpoints are far too often inventions of CROs and industry to get drugs and devices approved for commercial profit. They are not tools to answer genuine hypotheses
Roger Chou, MD: How the AHRQ creates and grades a meta-analysis
http://currentmedicine.tv/2012/specialties/biostatistics/roger-chou-md-how-the-ahrq-creates-and-grades-a-meta-analysis/
I am equally stunned that TVR is considered a surrogate endpoint by some. TVR is not a change in a laboratory variable but an invasive and an expensive procedure with signficant impact on quality of life.
As meta-analysis go, nothing beats patient level pooling of study data as was done in this study.
Further, the newer generation DES appear to have even lower risk of stent thrombosis compared with BMS. While I await the results of the HORIZONS II study, I think there is enough evidence to support the use of DES in primary PCI in carefully selected patients.
Count me in as one who believes we need hard mortality and serious morbidity endpoints, not TVR, which is often a discretionary event.