Prophylactic ICD Placement

More than 80 percent of patients who experience a life-threatening ventricular tachyarrhythmia do not survive to benefit from an ICD. Thus, the concept of the ICD for primary prevention of SCD has received considerable attention. Several large trials have been conducted to address primary prevention.14-19 The Sudden Cardiac in Heart Failure Trial (SCD-HeFT)15 tested the hypothesis that amiodarone or an ICD would improve survival compared with placebo in patients with HF. The study enrolled 2521 patients with NYHA II or III HF and an LVEF <35% of either ischemic or nonischemic etiology. Patients were randomly allocated to treatment with an ICD, amiodarone, or placebo. Background therapy was strong: 87% were on angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs) and 78% on beta blockers at last follow-up. ICD therapy decreased mortality by 23% compared with control, with an absolute reduction in mortality of 7% over a mean follow-up of 45 months. The ICD benefit was consistent across ischemic and nonischemic etiologies of HF. Subgroup analysis found benefit in NYHA Class II, but failed to demonstrate benefit from ICDs in patients with NYHA Class III, although such subgroup findings must be interpreted cautiously. A similar subgroup differentiation between NYHA class II and III has not been observed in any other ICD trial.1,20 There was no benefit for amiodarone.

Trials in patients with ischemic cardiomyopathy, including patients with and without HF symptoms, but excluding Class IV HF, also support benefit of an ICD in reducing total mortality. The Multicenter Automatic Defibrillator Implantation Trial (MADIT)-I enrolled patients with a mean LVEF of 26%, a prior MI, unsustained VT, and inducible VT at EPS. Those randomized to receive an ICD had a 54% reduction in total mortality at 27 months over those who had been given "conventional" antiarrhythmic therapy.14 The Multicenter Unsustained Tachycardia Trial (MUSTT) randomized patients with an LVEF <=40% and an inducible sustained VT into a group receiving beta blockers and ACE inhibitors or a group receiving these agents plus antiarrhythmics or, if they did not respond, an ICD. Over a 5-year follow-up, those receiving the ICD experienced a 27% lower risk of arrhythmic death or cardiac arrest. Total mortality was reduced by 20%.18 MADIT-II showed a significant survival benefit in patients with prior MI and LVEF <=30%, but also a trend toward increased HF episodes in patients receiving an ICD.1 In contrast, a study of patients undergoing coronary artery bypass (CABG) surgery who have LV dysfunction and abnormal signal-averaged electrocardiograms, but no symptomatic ventricular arrhythmias, found no benefit for prophylactic ICD implantation.16

A trial of ICDs for patients with nonischemic dilated cardiomyopathy and nonsustained VT reported a favorable trend for mortality reduction that did not reach statistical significance. A meta-analysis of trials in patients with nonischemic cardiomyopathy and the SCD-Heft trial support a survival benefit for ICDs in patients with nonischemic cardiomyopathy.15,21

In evaluating these trial results, it must be remembered that numerous factors can impact ICD effectiveness, including the population tested, mode of implantation, concurrent risk of death (eg, concomitant CABG surgery and its inherent risk), background medical therapy, and the influence of ICDs on nonarrhythmic deaths. For example, only 5% of the control group in MADIT was on beta-antagonist therapy, compared with 27% in the treatment arm. It is likely that this contributed to the decrease in cardiac nonarrhythmic deaths in the ICD group.14 These trials all exclude patients with severe NYHA functional class IV HF, although the Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) trial, examining biventricular pacing with or without concomitant ICD, included some NYHA class IV patients.17 In such patients, ICD implantation is under consideration as a "bridge" to transplant.22

Unresolved Issues

Timing of ICD Placement. Time on background therapy is a significant issue. All trials of ICDs have enrolled patients receiving stable doses of appropriate medical therapy, including beta-blockers and ACE inhibitors or ARBs. Furthermore, it is well known that treatment with these agents, particularly beta blockers, for a period of several months frequently results in reduction in LV volumes and improvement or normalization of LVEF. For these reasons, it is appropriate to delay consideration of ICD implant, particularly in the setting of newly diagnosed nonischemic dilated cardiomyopathy, until after several months of appropriate medical therapy. Although there is no agreement regarding the ideal treatment duration prior to reevaluation of LVEF for consideration of ICD placement, data regarding the time course of reverse remodeling following institution of beta-blocker therapy support a period of 3 to 6 months. Although physicians may choose to place a prophylactic ICD in selected patients prior to such a course of therapy, studies to support such a practice are not available.

Background Therapy. Device effectiveness generally is studied in patients already on background medical therapy. Studies establishing the additional benefit of new agents change the definition of standard or optimal background medical therapy. For example, a recent study of an aldosterone antagonist showed a 15% reduction in total mortality and a 17% reduction in cardiovascular mortality on top of standard medical therapy, mainly because of a 21% reduction in SCD.23

Significance of Arrhythmias Detected by ICDs. Trials of prophylactic ICDs have shown that the occurrence of an episode of VT or ventricular fibrillation (VF) is a marker for increased mortality and HF hospitalizations despite effective termination of VT/VF by the ICD.24-27 VT may be a marker for deterioration of ventricular function. When a patient experiences sustained VT detected by the ICD a careful evaluation and reassessment of HF status and therapy is warranted. The impact of therapies to reduce VT episodes is not clear, although ICD shocks reduce quality of life, and prevention is warranted.

Family History of Sudden Death. There are no systematic data to guide therapy when there is a family history of SCD. This includes patients with familial cardiomyopathy with a history of SCD in 1 or more sibling or parent. In such patients routine implantation of an ICD cannot be routinely recommended, but implantation may be considered on an individual basis after careful consideration of the circumstances.