Higher mortality, no long-term antirestenosis benefit with SES in SVGs: DELAYED RRISC

Antwerp, Belgium - Authors of a new study that found a higher mortality and no reduced restenosis over the long term in patients with saphenous vein graft (SVG) disease treated with sirolimus-eluting stents (SES) say they have started using mainly bare-metal stents for SVG stenosis in their practice^[1]. But Dr Paul Vermeersch (Antwerp Cardiovascular Institute Middelheim, Belgium) and colleagues also emphasize that their findings are from a secondary, post hoc analysis of a randomized trial and require further study.

Co-primary investigator for the study, Dr Pierfrancesco Agostoni (Antwerp Cardiovascular Institute Middelheim), told heartwire that he and his colleagues are not recommending that other cath labs also stop using drug-eluting stents (DES) in SVGs; "however, we were so impressed by the results of our study that we just stopped putting sirolimus-eluting stents [SES] in SVGs. We surely welcome—and are willing to participate in—additional larger trials to solve this issue."

Vermeersch, Agostoni, et al's study was a late reexamination of patients who originally participated in the Reduction of Restenosis In Saphenous Vein Grafts with Cypher Stent (RRISC) study, now dubbed DELAYED RRISC (Death and Events at Long-term Follow-up Analysis: Extended Duration of RRISC). In the original RRISC study, 75 patients with 96 SVG lesions had been randomized to either a SES or bare-metal stent and followed for six months. For DELAYED RRISC, the authors looked at all-cause mortality, MI, and target vessel revascularization (TVR) out to three years.

At a median follow-up of 32 months, 11 deaths, of which seven were cardiac, had occurred in the SES group, but none had occurred in the bare-metal-stent group. One of the SES deaths was a confirmed stent thrombosis, and three others occurred suddenly and "possibly" due to stent thrombosis. Rates of MI were no different between the two treatment groups, but more important, neither were rates of TVR.

"The late catch-up phenomenon noted in SVGs is somewhat new, but probably the sample size is too small to be definitive," Agostoni commented.

Outcomes at 32 months

More surprising, however, are the mortality findings, Agostoni noted. "Even if we think that the four noncardiac deaths can occur by chance, still, seven deaths vs zero is quite striking in a cohort of 75 patients."

Agostoni pointed to the 95% confidence intervals for the mortality findings: 17% to 45% for SES and 0% to 9% for the bare-metal-stent group. "If we think of the worst-case scenario for bare-metal stents, ie, 9% mortality, and the best-case scenario for SES, ie, 17% mortality, mortality in SES is still double that of bare-metal stents." That said, he continued, bare-metal-stent performance in this study was "exceptional," such that "the play of chance cannot be completely ruled out."

In an accompanying editorial, Dr Stephen G Ellis (Cleveland Clinic, OH) sounds a note of caution^[2]: "How should the cardiology community respond when informed of an apparent considerable increase in mortality from a procedure, when data are derived from the only randomized trial assessing the question, but that trial is small, the analysis post hoc, and the result somewhat counterintuitive?" Ellis writes. "That is the question raised by the report by Vermeersch et al."

Ellis points out that previous studies of SES for SVG lesions, while predominantly from uncontrolled registries, have not produced signals of increased mortality in this group. Moreover, Vermeersch et al's results come from an analysis that was never planned when the study was first launched, he points out. That said, Ellis also insists more rigorous studies are needed and "appropriate."

"It is quite possible, if not likely, that there is no excess risk" of SES in SVGs, he writes. "In the meantime, however, the cautious physician might well wish to refrain from using SES in SVG whose closure might be expected to lead to a large infarction."

Ellis concludes, "One must not overreact to the outcome of a study that it was not specifically designed to address. Such outcomes should be considered hypothesis generating. At the same time, the paucity of appropriate trial data tends to exaggerate the importance of whatever data we might have and, importantly, physicians must treat their patients on the basis of the best available data, however flawed it might be. "

For their part, Vermeersch, Agostoni, et al point out that an adverse effect of DES in SVGs is at least biologically plausible. Compared with native coronary artery lesions, lesions that develop in SVGs are typically softer, more lipid-rich, and more prone to rupture; atherosclerosis tends to progress more aggressively and may enhance inflammation and thrombotic reactions after stenting, they suggest.

"Plaque composition and plaque behavior in SVGs are different from native arteries; thus these factors can influence the outcomes of SES in SVGs," Agostoni told heartwire . "We are currently analyzing some pathological samples taken from SES in SVGs in patients who subsequently died, and we hope to get additional information from a detailed pathological analysis."

Heartwire from Medscape © 2007 

Cite this: Higher mortality, no long-term antirestenosis benefit with SES in SVGs: DELAYED RRISC - Medscape - Jul 09, 2007.