The data stop here: Primary-prevention ICDs for the very elderly

August 26, 2008

Boston , MA - An 85-year-old presents with symptomatic bradycardia and is considered a likely candidate for a pacemaker. At further evaluation, the patient is found to have an LV ejection fraction of 32%. With the patient meeting the guidelines' LVEF criteria for an implantable cardioverter-defibrillator (ICD), the decision is made on the spot to give the patient, not a single- or dual-chamber pacemaker, but an ICD programmed to manage the sinus-node disease (SND).

Such a scenario isn't unusual in actual clinical practice, at least in the US, according to Dr Michael O Sweeney (Brigham and Women's Hospital, Boston, MA). But there is no solid support for it in the clinical trials on which the guidelines for primary-prevention ICD therapy are based, he observed for heart wire . Patients with SND, he said, are typically much older than those entered into the trials, whose mean ages were around 60 years—and more likely to have potentially life-limiting comorbidities. They are far less likely to die from ventricular arrhythmias and more likely to die from noncardiac causes, suggest data Sweeney and his colleagues recently published in the Journal of Cardiovascular Electrophysiology[1].

Those broad findings aren't exactly novel, but they substantiate the group's cautionary message that a great many ICDs are implanted in the US based on scant clinical-trial evidence in elderly patients who are especially unlikely to benefit from them. The group calls for studies to define the best ways to treat the very elderly with SND who qualify for a device.

"It is almost a runaway situation," he said. "We don't exclude octogenarians from ICD therapy nor does Medicare exclude coverage, but it remains to be proven at the highest level of clinical evidence that the very elderly have a meaningful extension of life with defibrillators."

What happens is that our colleagues in the community who refer patients for ICDs tend not to refer those who are very elderly.

In an editorial accompanying Sweeney's report[2], Dr David Schwartzman (University of Pittsburgh, PA) observed that "the establishment of primary-prevention indications has produced a rapid rise in ICD implantations within the elderly community, and there is every indication that cardiac resynchronization (CRT) and AF-ablation therapies will follow this path." The public, "which has tended conceptually toward unlimited care for the most people," he writes, doesn't adequately grasp the implications of an impending convergence of a growing elder population and tightening resources.

Sweeney acknowledged any controversy about the issue is likely limited to the US. "The primary-prevention ICD is almost a US genre of one. This is a discussion that wouldn't even take place in most other Western countries," Sweeney said. "In many parts of the world, an 80-year-old who needs a pacemaker not only won't be offered an ICD, they'll probably get a single-chamber pacemaker."

In Canada, observed Dr Paul Dorian (St Michael's Hospital and University of Toronto, ON), "there is a limit on the number of ICDs we can implant. So what happens is that our colleagues in the community who refer patients for ICDs tend not to refer those who are very elderly. We just don't see many of those patients."

Dorian said he might well recommend an ICD for someone who is 80 or 85 years old "who is extremely healthy otherwise" but who needs a pacemaker and has an LVEF <30%, per Canadian guidelines. "But I suspect it happens a lot less frequently than it does south of the border."

Scope of the issue in the US

Such patients, Sweeney said, "create a unique point-of-service dilemma for the physician, who has made a commitment to hardware, at least in the form of bradycardia support, but the patient additionally seems to qualify on paper for an ICD." Suddenly, an elderly patient who might be well served by a fairly inexpensive pacemaker that will last up to a decade is getting a far costlier device that might last only half as long, he observed. "I don't have a statistic for how often that's happening, except we all know it's happening."

We're in an era when we're in great need of some thoughtful discussion about how best to approach these patients.

One can infer from the ICD Registry, maintained by the Heart Rhythm Society and American College of Cardiology to accommodate federal reimbursement requirements, that at least one-sixth of patients who receive primary-prevention ICDs are 80 or older.

"That, to me, is remarkable," Sweeney said. "There's been an understandable but unsupported extrapolation of observations of benefit made in younger patients to patients on average 20 years older, for which we really don't have any strong scientific support."

The major ICD trials have typically excluded patients with indications for bradycardia pacing, observes Schwartzman's editorial, so it's unknown how SND might influence the value of an ICD. "It is possible that sinus-node dysfunction betrays a more susceptible patient substrate, one that might derive greater benefit from ICD therapy; alternatively, or additively, it is possible that sinus-node dysfunction implies greater frailty, possibly attributable to unmeasured comorbidity, which could render ICD therapy less valuable," he writes.

"There's no question that the underlying risk of these patients is usually higher, their risk of complications is usually higher, and their preferences are much more varied than in the younger population," observed Dr Harlan Krumholz (Yale University School of Medicine, New Haven, CT) for heartwire . "Even if we assume that they get the same benefit [from ICDs] that was seen in the trials, their entire scenario is different. . . . We're in an era when we're in great need of some thoughtful discussion about how best to approach these patients."

The new insights

Sweeney and his colleagues looked retrospectively at >1100 patients with SND who were randomized to receive either single-chamber or dual-chamber pacing in the Mode Selection Trial (MOST) and had documented baseline ejection fractions.

About 62% of the group had NYHA class 2-4 heart failure; about 44% were >75 years old, and 23% were >80. "On average, patients enrolled in MOST were nearly 15 years older than patients in the Sudden Cardiac Death- Heart Failure Trial (SCD-HeFT)," Sweeney et al note, citing a trial that was instrumental in establishing current guidelines for primary-prevention ICDs.

Hazard ratio (95% CI)* for outcomes by LVEF (relative to LVEF >50%) in MOST

Outcome LVEF <35% LVEF 36 % - 49%
All-cause mortality 1.66 (1.20-2.29) 1.29 (0.92-1.82)
Heart-failure mortality 10.17 (3.35-30.82) 7.75 (2.50-24.01)
Sudden death 3.68 (1.72-7.89) 2.73 (1.27-5.87)        
*Adjusted for prior diagnosis of HF, NYHA class, prior coronary revascularization, history of atrial fibrillation or atrioventricular block, discharge antiarrhythmic therapy, QRS duration, and randomized pacing mode

"Patients who presented with SND had [ejection fractions] EF <35%, and were treated with permanent pacemakers had a 10-fold increased risk of HF death and a fourfold increased risk of SCD [sudden cardiac death] compared with patients who had a normal EF," write Sweeney et al. Their absolute rate of SCD was 22% over the average 33-month follow-up. Annualized over four years, their SCD rate was 3.9%, "only slightly higher than the 2.6% annualized SCD rate in the non-ICD treatment arm of SCD-HeFT," the group writes. "The annualized all-cause mortality rate (13.1%), however, was nearly double that of non-ICD patients in SCD-HeFT (7.0%)."

Much of the data from which guidelines are generated don't apply to the particular patient in front of us.

With their mean age of 74 years substantially higher than that in SCD-HeFT and other influential trials, MOST patients had a similar SCD risk compared with those trials but steeply increased mortality from non-SCD causes. Given that the cause of death in MOST was unknown in only a few percent of patients, the numbers suggest that >85% of patients died of causes that an ICD could not have prevented.

"Whether the use of ICDs (or cardiac resynchronization therapy) to provide appropriate bradycardia support would prolong life among very elderly pacemaker patients with EF <35% should remain an open line of scientific inquiry," write Sweeney et al.

The patient in front of us

Age, by itself, isn't a good basis for denying an otherwise-qualified patient from getting an ICD, according to Dr William T Abraham (Ohio State University, Columbus). "There are older patients who may have a good expectation for survival with an appropriately indicated defibrillator, and there are younger patients who because of comorbidities and other issues may actually have a poorer expectation for survival," he said. Age can be part of the equation, but "one shouldn't make decisions to limit the utilization of ICDs on the basis of age alone without a broader consideration of those other things."

Not all patients want to live forever. In fact many of them, if they're going to die of something, would prefer to die quickly and painlessly than to die slowly in the hospital.

Dr Lynne Warner Stevenson (Brigham and Women's Hospital, Boston, MA) agrees that SCD-HeFT and the other relevant trials entered much younger patients than those in MOST and the broader heart-failure population, so their conclusions can't directly apply to those groups. There's no good trial evidence for using primary-prevention ICDs in patients >80 years, "but that is generally true for elderly patients with almost any therapy. They are underrepresented in trials because they frequently have comorbidities that would limit their inclusion in trials. But they are usually not excluded directly because of age," she said. "Much of the data from which guidelines are generated don't apply to the particular patient in front of us."

Pacemaker therapy for symptomatic SND, Stevenson said, "is entirely justifiable at any age. But the idea of addressing a theoretical risk of a sudden death sometime in the future is a very murky benefit, and it gets murkier as the patient gets older."

In the US, "we are perhaps too generous with therapies that are not likely to provide benefit, as a general course. This may be particularly true in older patients who have other conditions that we're not taking into account. . . . When we look at any therapy that affects life expectancy but does not affect quality of life, we have to be very careful that we use that therapy in patients who are expected to have a good quality of life," she said.

"As we look at the healthcare problem in the US, we're going to have to make some exceedingly difficult decisions. We're going to have to be either rational or be severely rationed. And a relatively easy target is to reduce the number of expensive procedures that do not provide anticipated benefit."

For some patients, an ICD to reduce a hypothetical risk of sudden death while it merely delivers bradycardia support might not be worth the morbidity that comes with it.

All it might take to decide an ICD isn't appropriate, she said, is two questions. "Has the patient been hospitalized more than once for heart failure in the last year? Does the patient have an estimated glomerular filtration rate of less than 60 [mL/min]? If the answers to those questions are yes, the chances are small the patient will benefit from an ICD."

And the patient simply may not want one. Heart failure, regardless of pacing needs, can be debilitating, Dorian observed. "Not all patients want to live forever. In fact many of them, if they're going to die of something, would prefer to die quickly and painlessly than to die slowly in the hospital. That's an important consideration that some physicians don't necessarily discuss with their patients in enough detail."

Also, ICDs have risks, some of which don't exist with conventional pacemakers, observed Dorian. They can include infections, inappropriate shocks, and lead fractures that prompt more surgery. "Pacing leads are much more reliable than defibrillator leads, historically," he observed. "The odds of these things happening are small, but if the benefit is small, then all it takes is a small extra risk to even things out. The point is that defibrillators are not innocuous devices."

Sweeney notes that "for some patients, an ICD [to reduce] a hypothetical risk of sudden death while it merely delivers bradycardia support might not be worth the morbidity that comes with it."

A need for further research?

In their report, Sweeney and his colleagues make a clear call for trials to explore primary-prevention ICD use in the very elderly. Medicare "has not placed age limits on ICD therapies, and many clinicians have apparently not considered whether ICD therapy is less likely to be lifesaving in the very elderly," they write. The practice will become increasingly common "until future clinical trials can specifically address the indications for ICD therapy in elderly pacemaker patients."

To heartwire , Sweeney said, "They would not necessarily be easy trials to do, but I think, frankly, they are clinical trials we're going to have to face in order to rationalize these resources. I think we need to know whether the average 85-year-old lives longer with a defibrillator or not on a primary-prevention basis."

Abraham, on the other hand, said SCD-HeFT, the second Multicenter Automatic Defibrillator Implantation Trial (MADIT-2), and the other studies on which the guidelines are based included "reasonable proportions" of older patients and already accounted for those who had a poor expectation for survival. "I don't believe it's necessary to do another clinical trial. Rather, we should use the data that we already have in hand and apply some common sense in patient selection."

Stevenson observed that with about 16% of ICDs now going into patients who are octogenarian or older, "it won't take long to see how long their lives are lengthened in a meaningful way. We won't need a randomized trial to do that." There's no need for even a prospective registry study, she said. "The responsible thing for us to do is to track the results of the therapies we're already giving people."

MOST was supported by the US National Heart, Lung,and BloodInstitute, Medtronic, Guidant, and St Jude Medical; disclosures for the authors are in the report.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: