NEW YORK (Reuters Health) – Implantable cardioverter-defibrillators (ICDs) may not improve survival for patients in their 80s and 90s, according to a report in the June 2nd online American Journal of Cardiology.

“To me, the question is not whether or not there is an age limit beyond which the ICD can no longer improve longevity in patients with cardiomyopathy,” Dr. Samir Saba from University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania told Reuters Health in an email. “The question is rather to identify the age cutoff at which this happens. For now, the message to the physicians is to have an open discussion with their elderly patients who ‘qualify’ for receiving a defibrillator re the pros and cons of this therapy…and to individualize the decision based on the discussion.”

Dr. Saba and colleagues evaluated the effect of ICDs on all-cause mortality in a retrospective study of 140 octogenarians and 12 nonagenarians, 99 of whom had an ICD.

Patients implanted with an ICD were younger and had fewer comorbidities, better glomerular filtration rates (GFR), and a trend toward lower left ventricular ejection fraction (LVEF), compared with patients not implanted.

Ninety-three patients (61%) died during 2.3 years of follow-up, including 58 (59%) in the ICD group and 35 (66%) in the no-ICD group.

Although 1-year unadjusted survival was better in ICD recipients (72%) than in the no-ICD group (52%), there was no survival difference after adjusting for age, comorbidities, GFR, and LVEF.

In a multivariate Cox model, only age and GFR proved to be independent predictors of survival.

“Of note,” the researchers say, “in none of the ICD recipients followed in our study were there any true instances of documented ventricular fibrillation, which adds to the evidence that ICDs are unlikely to influence survival in this patient population and is consistent with previous reports.”

“Our study on the value of ICDs in the elderly is a retrospective analysis that can be only hypothesis generating,” Dr. Saba explained. “It cannot be used to make recommendations regarding patient management. What is needed is a prospective, randomized study of ICD vs. no ICD therapy in octogenarians with left ventricular dysfunction. Such a study could then change clinical practice if its results support those of our current study.”

“The present study has significant clinical implications because it questions the utility of implanting ICDs in octogenarians and nonagenarians,” the investigators conclude. “Implanting an ICD is an invasive and costly procedure and its utility has not been heretofore studied rigorously.”

“Improving longevity with any therapy depends essentially on the severity of competing causes of death,” Dr. Saba said. “In other words, if patients are so old that they are likely to die within a short period of time (on average), it is unlikely that any therapy that protects against cardiac arrest will make them live longer.”

Am J Cardiol 2 June 2011.