Age and Effectiveness of Implantable Cardioverter-Defibrillators

TO THE EDITOR: Although some studies suggest that patients aged 75 years or older derive benefit from primary prophylactic implantable cardioverter-defibrillator (ICD) therapy similar to that of their younger counterparts (1), consideration of age and noncardiac comorbid conditions is critical in the decision to implant a device. Santangeli and colleagues’ meta-analysis (2) adds an intriguing observation to the ongoing debate on ICD therapy in elderly persons. On the basis of their primary analysis of 3 major randomized trials of prophylactic ICD therapy in patients with severe left ventricular dysfunction (DEFINITE [Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation], MADIT-II [Multicenter Automatic Defibrillator Implantation Trial II], and SCD-HeFT [Sudden Cardiac Death in Heart Failure Trial]), Santangeli and colleagues conclude that ICDs may not improve survival in patients aged 60 to 65 years or older (hazard ratio, 0.81 [95% CI, 0.62 to 1.05]; P 0.11). This finding has potentially important clinical implications and causes great concern, as the reported age cut-off is within the range of the average age of patients reported in ICD trials and below the mean age of patients with ICDs implanted in routine clinical practice (3). However, careful reading of the article raises the suspicion that the survival effect of prophylactic ICD therapy in elderly patients has been critically underestimated. According to Table 3 and Figure 1, the meta-analysis seems to only include MADIT-II patients aged 60 to 69 years (n 426), whereas the large group of patients aged 70 years or older (n 436) was excluded for unclear reasons. Of note, patients aged 70 years or older had a greater survival benefit from the ICD than those aged 60 to 69 years (4). Therefore, an analysis exclusive of MADIT-II patients aged 70 years or older would contribute to an underestimation of the effects of ICDs in elderly persons. Also, it is not entirely clear how the authors weighted results from SCD-HeFT (5). Although Figure 1 shows the hazard ratio for the comparison of patients who received ICD therapy (n 829) versus placebo (n 847), the meta-analysis seems to account for all 2521 SCD-HeFT patients (see the Methods section and Table 3), including the 845 patients assigned to receive amiodarone who were not expected to be part of the present study. In addition, the number of elderly SCD-HeFT patients in Table 3 is largely exaggerated. The weight attributed to results from SCD-HeFT has probably been mistakenly enhanced, which would further underestimate the effects of ICDs on survival in elderly patients. We concur with Santangeli and colleagues’ call for controlled trials of implantable device therapies in elderly patients. However, we are concerned that the present conclusions misjudge the actual benefits achieved by ICD implantation and may discourage physicians from recommending a potentially life-saving treatment in appropriately selected patients. In fact, a recalculation of the effects of ICDs on survival among elderly patients in DEFINITE, MADIT-II, and SCD-HeFT may yield a hazard ratio less than 0.76 that is likely to be statistically significant. Frieder Braunschweig, MD, PhD Karolinska University Hospital SE-17 176 Stockholm, Sweden