Implantable cardioverter-defibrillators in the elderly and potentially terminally ill patients

During the past three decades, the implantable cardioverter-defibrillator (ICD) has evolved from a last-ditch effort for the treatment of patients with malignant ventricular arrhythmias to a widely used treatment option for patients who are at risk of sudden cardiac death.

Since ICD treatment requires surgery and implantation of a complex and exquisitely engineered device, it is sometimes difficult to decide whether ICD implantation is appropriate in an individual patient. Randomized clinical trials are designed to identify a relatively homogeneous population at risk of the treatment endpoint of interest to evaluate the effects of a specific therapy.

"While clinical guideline statements developed by professional societies are extremely important for synthesizing scientific data and expert consensus and thus provide general guidance to the clinician for best practices, they cannot account for all of the variables in clinical medicine," says Fred Kusumoto, M.D., an electrophysiologist at Mayo Clinic in Florida.


Two seminal trials established the ICD as an effective therapy that reduces mortality. In the Multicenter Automatic Defibrillator Implantation Trial II (MADIT-II), patients with an ejection fraction < 30 percent due to prior myocardial infarctions were randomized to best medical therapy or best medical therapy with an ICD. After two-year follow-up, patients randomized to receive an ICD had a 20 percent reduction in risk of death.

Using a slightly different patient population, the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) randomized patients with an ejection fraction < 35 percent due to either ischemic or nonischemic cardiomyopathy, and NYHA Class II or Class III heart failure, to receive an ICD, amiodarone or placebo. After 46-month follow-up, patients who received an ICD had a 23 percent reduction in mortality compared with the placebo and amiodarone groups.

When applying the results of clinical trials to management of individual patients, it is important to clearly understand the inclusion and exclusion criteria for the trial and carefully examine the characteristics of the patients who were actually enrolled. In MADIT-II and SCD-HeFT, there were no upper age limits, but the average age of patients enrolled in both studies was 60 to 64 years, and both included mostly men (84 percent in MADIT-II and 76 percent in SCD-HeFT).

Between 2000 and 2010 in the United States, the population older than 65 years grew at a faster rate (15 percent) than the total population (10 percent), and there are now nearly 6 million people in the United States who are older than 85 years. In a recent analysis of the National Cardiovascular Data Registry for ICDs, 47 percent of ICD implants were in patients > 70 years of age, and 16 percent were in patients > 80 years of age.

In addition to under-representing the elderly and women, clinical trials have not enrolled patients with potentially terminal medical illnesses. This represents a significant gap in our understanding of the clinical application of ICD therapy. For example, in patients with stage I to III breast cancer, five-year survival rates can range from 45 to 95 percent. In a patient with breast cancer who also has cardiomyopathy and is at risk of sudden cardiac death, is an ICD an appropriate treatment option?

Careful examination of MADIT-II and SCD-HeFT can provide some insight into these important questions. The survival benefit from ICD therapy appears to increase over time, with the survival curves beginning to diverge at one year in MADIT-II and two years in SCD-HeFT, so that by three years there is an absolute 6 to 9 percent reduction in mortality associated with ICD implantation.

To put this data into perspective, chronic beta blocker therapy after myocardial infarction was associated with an absolute 2.5 to 3.5 percent reduction in mortality in the Carvedilol Post-Infarct Survival Control in Left Ventricular Dysfunction Study (CAPRICORN) and the Beta-Blocker Heart Attack Trial (BHAT).

There is some information available on the use of ICDs in the elderly. In a prespecified subanalysis of SCD-HeFT, ICD therapy was not beneficial in patients 65 years or older (HR: 0.86; 97.5 percent CI, 0.62-1.118; p = ns). In contrast, in a post hoc analysis of MADIT-II, mortality reductions with ICD therapy were similar in patients < 75 years and e 75 years.

A recent meta-analysis studied the effects of age by combining data from MADIT-II and SCD- HeFT with three smaller primary prevention trials, the Multicenter Unsustained Tachycardia Trial (MUSTT), the Defibrillators in Nonischemic Cardiomyopathy Treatment Evaluation (DEFINITE) trial and the first Multicenter Automatic Defibrillator Implantation Trial (MADIT-I). When these studies are combined, 597 patients > 75 years old were identified (12 percent of the total number of patients enrolled), and the ICD remained beneficial for reducing all-cause mortality (HR 0.73; 95 percent CI, 0.51-0.97; p = 0.03).


During the last five years, several studies have documented and studied the relationship between unnecessary shocks and clinical outcomes. In the recently published Multicenter Automatic Defibrillator Implantation Trial-Reduce Inappropriate Therapy (MADIT-RIT), careful programming of the ICD to reduce inappropriate therapy was compared with standard programming. Programming that required higher tachycardia rates or longer periods of time in tachycardia before a shock was delivered were associated with a 45 to 55 percent reduction in mortality.

In MADIT-RIT, approximately 50 percent received an ICD that could also provide cardiac resynchronization therapy, and the two-year mortality of the conventionally programmed group (10 percent) was lower than the two-year mortality of the ICD group in MADIT-II (16 percent). "The findings of MADIT-RIT do bring up the intriguing notion that the large original ICD trials actually underestimated the possible benefits of ICD therapy," says Dr. Kusumoto.

It is obvious that the decision whether to implant an ICD in an older patient or a patient with a potentially terminal condition is highly nuanced and requires a sophisticated and balanced approach. The decision is best served by close collaboration among a group of physicians who have expertise in different medical specialties and taking care of different patient groups.

Once implanted, the ICD must be programmed carefully and based on individual patient characteristics to maximize the benefit of therapy. While clinical guidelines and algorithms are extremely useful for providing broad recommendations, the decision to implant an ICD in the patient "sitting in front of you," says Dr. Kusumoto, requires thoughtfulness and consideration of the clinical situation and the patient's wishes.