Reed Miller

April 04, 2011

April 4, 2011 (New Orleans, Louisiana) — The investigators of the Surgical Treatment for Ischemic Heart Failure (STICH) trial believe five-year results of the study presented at ACC 2011 provide support for coronary bypass surgery on heart-failure patients with coronary disease, even though CABG did not beat medical therapy alone in the primary analysis of the trial. [1].

Dr Eric Velazquez

This morning at the American College of Cardiology 2011 Scientific Sessions, Dr Eric Velazquez (Duke Clinical Research Institute, Durham, NC) presented long-awaited results of "hypothesis one" from the National Institutes of Health–sponsored STICH, a randomized comparison of coronary artery bypass graft (CABG) surgery and medical therapy alone in 1212 patients with coronary artery disease amenable to CABG with a left ventricular ejection fraction under 35%. The primary end point was all-cause death. A substudy of STICH on myocardium viability testing was presented by Dr Robert O Bonow (Northwestern University, Chicago, IL). Both studies are published online simultaneously in the New England Journal of Medicine.

The "hypothesis-one" results are the final piece of a puzzle for a trial that battled sluggish enrollment for almost a decade and answers a question for which many doctors already assumed they had the answer.

Also commenting on the trial after it was presented, Dr Bernard Gersh (Mayo Clinic, Rochester, MN) praised it as a landmark, even though it missed its primary end point. "We've known for decades that in patients with LV dysfunction, ischemia is the major cause of mortality, and that in patients with ischemia, LV dysfunction is the major cause of mortality. [Velazquez et al] have now proven the concept. This is an incredible trial. A stunning achievement." ACC 2011 cochair Dr Edward McNulty (University of California, San Francisco) agreed, calling the STICH study "precisely the kind of comparative-effectiveness study that is vitally needed."

As expected, in STICH CABG was associated with an early risk of death as a result of the surgical intervention itself, but this disadvantage for surgery disappeared after two years postprocedure.

Over a median follow-up of 56 months, 41% of the medical-therapy group and 36% of the CABG group died (p=0.12), but the difference was statistically significant following adjustment for baseline characteristics (p=0.039). There was also a slight advantage for CABG in cardiovascular-specific causes of death: 33% of the medical-therapy group and 28% of the CABG group died of an adjudicated cardiovascular cause (p=0.05). The study also found a slight advantage for CABG in some of the composite secondary end points. In the medical-therapy group, 68% of the patients died from any cause or were hospitalized for cardiovascular causes, compared with 58% of the CABG group (p<0.0001).

Velazquez also pointed out that there was also significant crossover between the study groups that may have diminished the benefits of CABG in the primary intention-to-treat analysis. The study was designed to allow up to 20% of the medical-therapy group to eventually undergo CABG; 17% of the medical-therapy group eventually underwent CABG, and 9% of the patients assigned to CABG never had bypass surgery. The median time until surgery in the medical-therapy patients who crossed over was 142 days. The most common reasons for crossover to CABG were progressive symptoms, acute decompensation, or family or physician decision.

An as-treated analysis of 592 patients treated with medical therapy for the first year after randomization and 620 patients who had CABG within the first year showed a mortality benefit for CABG (hazard ratio 0.70, p<0.001). Also, a per-protocol comparison of the 537 patients randomly assigned to medical therapy who never crossed over and the 555 patients in the CABG group who actually had surgery also showed a mortality benefit for CABG (HR 0.76, p=0.005).

Velazquez said the patients in this study will be followed for 10 years and there will be additional analyses forthcoming, including an investigation into the completeness of revascularization in the CABG patients.

Investigator, Observers Focus on Secondary Analyses

In an editorial accompanying the paper in the journal [2], Dr James Fang (Case Medical Center, Cleveland, OH) commends the trial as extremely well conducted. He concludes, "With the results of the STICH trial, we should be comfortable with the notion that, in general, surgery is not superior to optimal medical therapy for ischemic left ventricular dysfunction."

But in the press conference, Velazquez said, "We really don't know what impact STICH will have in terms of the large population who are at risk and have coronary disease, because prior to STICH we didn't have, necessarily, an indication to pursue an evaluation of coronary disease to reduce an outcome in patients who were primarily symptomatic of heart failure. The studies since the 1970s have excluded these patients, and then, in patients with modest LV dysfunction, the studies suggested that if they had predominantly heart failure instead of angina, they derive no benefit.

"The take-home message for me is that the STICH trial supports bypass surgery on top of best medical therapy vs medical therapy alone to reduce cardiovascular morbidity and mortality and that many patients who are now treated for heart failure without ever being assessed for the potential of having angiographic coronary disease should be evaluated for that, because [coronary disease] does not present the same way in every patient," he said. "Heart failure without angina shouldn't exclude patients from an angiographic evaluation."

In the published paper, Velazquez et al stress that great care was taken by the trial investigators to ensure that both therapies were optimized. All of the surgeons in the trial had previous operative mortality under 5%, and medication-therapy compliance was very high in both the CABG and medical-therapy–only arm; insofar as the medical compliance was better in the trial than would be expected in the "real world," the study results may have been biased against CABG, Velazquez said. "Our utilization of medical therapy can always improve, and there might be underutilization of medical therapies in these populations referred to bypass surgery. Medical therapy has to be optimized for all patients." On the other hand, the CABG results in STICH should be generalizable to any surgeon who is comfortable with bypass surgery and are comparable to those of previous studies in both Europe and the US.

Following the presentation of the data, surgeon Dr Steven Bolling (University of Michigan Cardiovascular Center, Ann Arbor) said, "The analysis of intention-to-treat vs actual treatment is interesting, but the biological effect that our patients feel is what treatment they receive, and under that analysis, as a surgeon, you must conclude that patients with left ventricular dysfunction should receive coronary bypass."

Gersh predicted that forthcoming additional analysis of the STICH data may reveal more about the mechanism of benefit. Surgery might benefit these patients by improving systolic function or perhaps improving diastolic function, which had been inhibited by myocardial stiffness caused by ischemia, he said. "I suspect that much of the improvement will be from a benefit on sudden cardiac death, which is what we saw in early trials of bypass surgery. Or could it be that the benefit is on recurrent infarction, which these patients obviously cannot tolerate," Gersh speculated.

Velazquez reports consulting fees/honoraria from Novartis and research grants from Johnson & Johnson and Boehringer Ingelheim. Fang reports consulting arrangements with Accorda, Novartis, Lantheus, fellowship training support from Medtronic, and lecture payments from Boston Scientific.Gersh reports consulting fees from Boston Scientific, Merck, Ortho-McNeil Janssen, Scientific Affairs, St Jude, Medispec, GE, and Abbott.