Boston, MA- Implantable cardioverter defibrillators (ICDs) can deliver shocks to prevent sudden death in patients with heart failure, but, statistically, such patients' all-cause mortality risk climbs whenever those shocks are delivered, according to a secondary, prospectively defined analysis from the SCD-HeFT trial [1]. ICD discharges, the study further found, seemed to raise the risk of death whether or not they were triggered by the kind of arrhythmias that cause sudden death.

The analysis doesn't show that ICD discharges are killing people; rather, it highlights a recognized but perhaps underappreciated phenomenon—that shocks flag patients who are particularly sick. "Ventricular arrhythmias are a high-risk marker for a more vulnerable substrate," observed lead author Dr Jeanne E Poole (University of Washington Seattle) for heartwire.

When patients report to their doctor that they've experienced a shock, "that shouldn't be the end of the discussion," she said. "They've survived sudden death, but now that group of patients is at very high risk of dying of predominantly heart failure in the next months to years. . . . Our main message is that this is a high-risk group by virtue of having had these arrhythmias; they deserve a real thoughtful and careful evaluation, and they should see their physicians for follow-up."

In another message to come out of the trial, from a report published simultaneously with the analysis from Poole et al, ICD therapy—compared with medical management only—seemed to have no adverse effect on quality of life, measured using several standard instruments, over at least a year [2].

"Although ICD therapy for primary prevention of sudden death was not expected to improve the quality of life, the possibility of harm from either psychological or physical complications of the therapy was of concern," write the authors, led by Dr Daniel B Mark (Duke University Medical Center, Durham, NC). "In our overall comparisons, we found no statistically or clinically significant evidence of either psychological or physical harm."

The trial had randomized 2521 patients with NYHA class 2-3 heart failure and LVEF <35% but no history of sustained VT or cardiac arrest to receive standard contemporary HF medical therapy alone or with amiodarone or ICD implantation. Patients getting devices received single-lead Medtronic ICDs programmed to deliver only shock therapy with a maximum output of 30 J; the devices' antitachycardia pacing function wasn't turned on.

The primary findings of a 23% reduction in mortality (p=0.007) over a median of 45.5 months with device therapy, regardless of HF etiology, and no significant effect in the amiodarone group were publicized in March 2004, published the following year [3], and covered both times by heartwire.

The two current reports based on the trial, which helped launch the era of primary-prevention ICDs, appear in the September 4, 2008 New England Journal of Medicine.

"These two reports show that modern ICD therapy is prolonging survival in patients with heart failure, with relatively little compromise in the quality of life," write Drs Jeff Healey and Stuart Connolly (McMaster University, Hamilton, ON), in an accompanying editorial [4]. "It is somewhat disturbing to realize that actually receiving a shock is such an important predictor of death (commonly in association with progressive heart failure). However, it should not be surprising that many patients in whom sudden death from arrhythmia is averted by an ICD ultimately die from heart failure. . . . Several randomized trials of ICD therapy have shown that a reduction in deaths from arrhythmia is offset either partially or completely by an increase in deaths from other causes. Nevertheless, SCD-HeFT has shown us that a prophylactic ICD does buy some time and that this time is worthwhile to patients."

Mark and his colleagues assessed quality-of-life measures in virtually all SCD-HeFT patients at baseline and after three, 12, and 30 months. Cardiac-specific physical functioning, as measured by the Duke Activity Status Index (DASI), didn't differ between the ICD, amiodarone, and control groups at any of those intervals. Scores of psychological well-being on the Mental Health Inventory 5 (MHI-5) were significantly improved in ICD recipients compared with the control group at three months (p=0.01) and 12 months (p=0.003), but the advantage had disappeared by 30 months. A similar pattern was seen for measures of pain, emotional health, social functioning, and other domains on the Medical Outcomes Study 36-item Short Form (SF-36).

Patients in the ICD group, compared with controls, showed significant improvements on the Minnesota Living with Heart Failure Questionnaire (MLHFQ) at three months (p=0.006) and 30 months (p=0.05), but the difference fell short of significance at 12 months (p=0.07).

Quality-of-life scores plunged (p<0.001 to p=0.02) for measures of general health, physical and emotional function, and other domains for the 49 patients in the ICD group who had received a shock within one month of a scheduled evaluation, compared with others in the ICD group, according to Mark et al. But the differences between the two subgroups grew weaker as time after the shock increased.

According to editorialists Healey and Connolly, "The quality-of-life analysis in SCD-HeFT suggests that, in general, the side effects associated with a prophylactic ICD are not onerous, and the quality of life among ICD recipients is sufficiently good that the cost per quality-adjusted life-year that is saved with ICD therapy [as was previously reported] makes the therapy economically attractive."

However, they add, "This analysis may have underestimated some of the negative effects of ICD therapy. With an average follow-up of less than four years, the time frame is too short to capture all potentially relevant complications associated with ICDs, including late lead failure, the need to replace the pulse generator, and safety advisories from the manufacturer."

Over the follow-up period, 33.2% of patients with ICDs received at least one shock, defined as therapy delivered for an episode of ventricular arrhythmia regardless of the number of actual device discharges, according to the authors. Only appropriate shocks, defined as discharges following onset of VT or VF, occurred in 128 patients; 87 patients received only inappropriate shocks (such as for supraventricular tachycardia or at P- or T- wave oversensing, or those delivered after the spontaneous termination of nonsustained VT); and 54 experienced both kinds.

In multivariate analyses adjusted for baseline features that could influence outcomes, appropriate ICD shocks were associated with a fivefold increase in all-cause mortality risk; inappropriate shocks doubled the risk. The risk went up exponentially for patients who experienced multiple shocks of both kinds, compared with patients who didn't experience any shock.

Of the 77 deaths among patients who experienced any kind of ICD shock, 33 were from progressive heart failure, 16 were sudden and caused by VT or VF, nine were from other cardiac causes, and 19 were noncardiac or of unknown cause.

"What is the mechanism underlying these associations?" ask Healey and Connolly. "Although it is plausible that shocks somehow have an adverse effect on myocardial function, this is unlikely to be a major factor. What is much more likely is that the occurrence of a ventricular arrhythmia that causes a shock is signaling a meaningful change in the patient's clinical status. The important message is that the first occurrence of shocks is not a random event in an otherwise-stable clinical course but a sign of clinical deterioration in the underlying disease process."