Washington, DC - A randomized trial that has already made a mark on cardiology had a few more surprises left when investigators once again looked into the data [1].

In the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT), which helped establish the implantable cardioverter defibrillator (ICD) for primary prevention of sudden death in low-LVEF heart failure, syncope was a major predictor of death, yet device therapy offered no protection against either syncope or the increased mortality risk, the researchers found. At the same time, they say, among those who received devices, syncope appeared to increase the likelihood of appropriate discharges.

Prior, mostly observational studies have indicated that many patients with cardiomyopathy and heart failure have episodes of syncope and that such patients also have elevated risks of ventricular arrhythmias and death, explained the lead author of the current analysis, Dr Brian Olshansky (University of Iowa, Iowa City), to heartwire. It had been largely assumed, he said, that ventricular arrhythmias promoted both the syncope and increased mortality and therefore that such patients would likely benefit from an ICD.

They had more arrhythmias, they got a lot more shocks than the patients who didn't have syncope, but in fact the defibrillator didn't protect them against dying.

Its randomization of such patients to an ICD, amiodarone, or placebo in addition to top-notch medical therapy allowed SCD-HeFT to show whether the assumption was correct, according to Olshansky. As it turns out, the current analysis, published in the April 1, 2008 issue of the Journal of the American College of Cardiology, suggests that ICDs didn't help the trial's patients with syncope.

"Indeed," Olshansky said, "they had more arrhythmias, they got a lot more shocks than the patients who didn't have syncope, but in fact the defibrillator didn't protect them against dying."

The SCD-HeFT analysis has caveats even beyond its post hoc nature. For example, the trial's ICDs were "single chamber, shock only, with backup ventricular pacing," less versatile than the typical device in use today. And it wasn't specifically designed to look at syncope, nor was syncope evaluated in any consistent way, Olshansky et al acknowledge.

Still, the implications are potentially far-reaching. The analysis suggests that ICDs are "useless" in cardiomyopathy patients with syncope who qualify for primary-prevention devices according to current LVEF-based criteria, according to Olshansky. It's probably premature to use syncope to exclude otherwise-eligible patients from getting ICDs, he said, "but these are certainly the first data that suggest that should be the case."

It's most likely some kind of neurocardiogenic response, we don't know for sure.

As previously reported by heartwire, SCD-HeFT randomized 2521 patients with NYHA class 2-3 heart failure and an LVEF <35% to receive the "shock-only" ICDs or either amiodarone or a matching placebo. The significant 23% drop in five-year mortality risk in the ICD group was independent of heart-failure etiology; amiodarone had no significant effect on survival.

In the current analysis covering an average 45.5 months of follow-up, syncope occurred before, after, and both before and after randomization in 6%, 14%, and 2%, respectively, of the SCD-HeFT population.

The mortality hazard ratio (95% CI) for syncope before randomization was a nonsignificant 0.98 (0.73-1.33). Syncope after randomization, however, posed risks of 1.41 (1.13-1.76; p=0.002) for all-cause mortality and 1.55 (1.19-2.02; p=0.001) for cardiovascular death.

"Importantly," Olshansky et al write, "syncope's association with mortality risk was independent of randomization arm." All-cause mortality was elevated among patients with syncope, compared with those without syncope, by 54% in the ICD group, 33% among those receiving amiodarone, and 39% in the control group; the risk elevation achieved greater significance among ICD recipients, but it didn't vary significantly by randomization group.

Among ICD recipients, syncope developing either before or after randomization predicted appropriate device discharges during follow-up; the HRs were 1.75 (1.10-2.80; p=0.019) and 2.91 (1.89-4.47; p=0.001), respectively.

Maybe they need further evaluation of hemodynamic response to various maneuvers, like a tilt-table test or exercise test.

The analysis doesn't indicate what caused syncope in the trial, observed Olshansky. "It's most likely some kind of neurocardiogenic response, we don't know for sure." Getting a fix on the cause, he said, is tough if the patient isn't being monitored at the time of the episode.

"I suspect what's happening is that the syncope itself is an indicator of poor hemodynamic response in heart failure. The people with syncope are probably sicker, even if you can't measure it by ejection fraction, functional class, or other methods." Because they're sicker, Olshansky speculated, they have more ventricular arrhythmias and therefore more device discharges.

It's also possible that in sicker hearts, catecholamine levels rise "in an attempt by the heart to increase contractility, and that leads to this serious neurocardiogenic reflex—not the typical kind you'd see in a healthy person," he observed.

The findings may also point to a need for more extensive workups of candidates for primary-prevention devices, he said. "Maybe they need further evaluation of hemodynamic response to various maneuvers, like a tilt-table test or exercise test. . . . I think maybe we should look into what's going on with these patients more than we have, instead of just blindly accepting ventricular arrhythmia as the cause for the [syncope] event."