Primary Prevention ICD Efficacy Again Questioned in Heart Failure With CKD

February 08, 2018

CHICAGO, IL — Treatment with an implantable cardioverter defibrillator (ICD) did not improve survival in patients with reduced-LVEF heart failure and chronic kidney disease (CKD) compared with similar ICD-eligible patients in the same community-based cohort who didn't get a device.[1]

But those who received an ICD did show greater 3-year risks for hospitalization from any cause and from heart failure, in the propensity-adjusted analyses based on 5877 patients seen from 2005 to 2013.

The findings concur with some studies and are at odds with others in questioning whether primary prevention ICDs prolong life in such patients, who might be too sick to benefit because of competing risks from comorbidities or device-related complications, according to a report published February 5 in JAMA Internal Medicine.

One of the study's limitations is that it wasn't possible to determine how much of the observed mortality was arrhythmic, lead author, Dr Nisha Bansal (University of Washington, Seattle), told | Medscape Cardiology.

"So we don't know, was it a lack of a survival benefit due to a defibrillator, which doesn't seem to be effective with arrhythmias in patients with CKD, or is it that there's no difference when you take into account all the other causes of death?"

Bansal proposed that possible complications related to the device itself, such as infections, may have contributed to the increased hospitalizations in the especially vulnerable population with CKD.

Other studies have concluded against a survival benefit from primary prevention ICDs, sometimes seeing a significant mortality risk, in patients with CKD, she and her colleagues note. But even before ICDs were established therapy, when such a trial would have been more feasible, few compared device therapy vs no device therapy in patients with CKD.  

The current comparative study may be the largest of its kind in the literature, and the data are recent enough to account for changes in heart failure prescribing, Bansal said.

"Whether or not this is sufficient to convince the community that CKD should be considered in the guidelines, I think that needs more discussion," she said.

"I think some multidisciplinary discussion between nephrologists and cardiologists is warranted to figure out whether this could be implemented into clinical practice or whether the community needs more data."

Not Perfectly in Line With Existing Evidence

In an interview, Dr Sana Al-Khatib (Duke University Medical Center, Durham, NC) said the study's findings "don't align perfectly well with the existing evidence that we have today." It does, however, once again prompt the question, "Do ICDs benefit patients with chronic kidney disease? And until we have a randomized clinical trial, which probably will never be done, I don't think we can answer that question definitively."

The clinical trial evidence so far, said Al-Khatib, who was not involved in the current study, "does not seem to support any survival benefit from ICDs in patients with end-stage renal disease," even though their risk for arrhythmic death is high. But patients with milder forms of CKD were included in the major primary prevention ICD trials, and in subgroup analyses "the signal seemed to be in favor of using an ICD."

She said the current cohort seems to be between the two extremes of kidney dysfunction. Most patients had moderate CKD, which was to some extent represented in the major ICD trials.

"In mild-to-moderate chronic kidney disease, the existing evidence actually supports a role for ICDs," Al-Khatib said. "I think it's in patients with more severe CKD that the existing evidence tells us that maybe ICDs are not beneficial."

A recent society-sponsored guideline update on prevention of sudden cardiac death, for which Al-Khatib chaired the writing committee, singled out patients with kidney disease as a group not well represented in ICD trials and deserving of further study.

It drew a similar conclusion about the very elderly. Al-Khatib pointed out that the current cohort's mean age was about 73 years, with more than 40% of the patients older than 75 years, whereas patients in the primary prevention ICD trials tended to be in their 50s and 60s. Competing mortality risks in the study's older patients might be expected to attenuate the ICD relative survival benefit.

"I wonder if maybe the age of the patient population is playing a role. These are not only patients with chronic kidney disease, they're older patients with chronic kidney disease," she said. "This is basically hypothesis-generating—it reminds people that we need more data on these patients."

Propensity-Matched and Confounder-Adjusted

The analysis included adults with heart failure, an LVEF of 40% or less, and CKD, based on estimated glomerular filtration rate, who were eligible for ICD therapy and were seen at centers in four Kaiser Permanente regions in the Western United States.

The 1556 who received ICDs were propensity-matched to 4321 who did not get a device; 28.6% and 32%, respectively, were women.

Mortality over a mean of 3.1 years was 43.2% overall; 69.6% of patients were hospitalized for any cause and 32.7% were hospitalized for heart failure.

The mortality risk didn't significantly differ between the ICD and non-ICD groups, but both hospitalization endpoints were significantly increased for ICD recipients in adjusted analyses.

Table. Hazard Ratio for Outcomes of ICD Use in Reduced Ejection Fraction Heart Failure With CKDa

Endpoints Hazard Ratio (95% CI)
All-cause mortality 0.96 (0.87–1.06)
Hospitalization from any cause 1.25 (1.20–1.30)
Heart failure hospitalization 1.49 (1.33–1.60)
aHazard ratio adjusted for age, sex, race, smoking, prevalent heart failure, acute MI, unstable angina, CABG, PCI, cerebrovascular events, other arterial thromboembolic events, AF or atrial flutter, VT or VF, valvular disease, peripheral artery disease, rheumatic heart disease, pacemaker, dyslipidemia, hypertension, diabetes, dementia, depression, chronic lung disease, chronic liver disease, systemic cancer, LVEF, hemoglobin, systolic blood pressure, renal function, year of study entry, and cardiovascular medications (ACE inhibitors, angiotensin receptor blockers, β-blockers, and diuretics).


One of the study's likely limitations, Bansal acknowledged, is the predominance of patients with nonischemic cardiomyopathy. Nonischemic heart failure is more heterogeneous in its etiology than is ischemic heart failure, she observed, and treatments are more well defined for the latter. Only 32% of the cohort were known to have coronary artery disease, so the findings might not be generalizable to the broader population of patients with heart failure and reduced ejection fraction.

Bansal and the other authors had no disclosures. Al-Khatib said she has no conflicts or financial relationships with industry.

Follow Steve Stiles on Twitter: @SteveStiles2. For more from | Medscape Cardiology, follow us on Twitter and Facebook.


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: