ICDs' Role Remains Uncertain in CKD Patients

LOS ANGELES – Patients with advanced chronic kidney disease who receive an implantable cardioverter defibrillator are not like other patients who receive these devices.

But results from a trio of studies reported at the meeting show how hard it is for researchers to get a clear handle on what makes chronic kidney disease patients different, whether they have different outcomes than defibrillator recipients without CKD, and what is the best approach for judiciously using implantable cardioverter defibrillators (ICDs) in patients with renal dysfunction.

Mitchel L. Zoler/IMNG Medical Media

It’s a particularly relevant question because CKD is linked with a substantially increased risk for sudden cardiac death (patients on dialysis have about a fivefold higher risk for sudden cardiac death, compared with patients with normal or mildly impaired renal function), but patients with CKD were excluded from the major trials that proved the efficacy of ICDs for preventing sudden cardiac death. The lack of data from randomized, prospective trials leaves questions about the efficacy, and perhaps as importantly, the cost efficiency of ICDs in CKD patients.

"Should patients with CKD get ICDs? I don’t know the answer, but I don’t think that, categorically, patients with CKD should be excluded from ICD treatment," commented Dr. Matthew Reynolds, a cardiac electrophysiologist and director of the Economics and Quality of Life Research Center at Harvard Medical School in Boston. But many patients with CKD, especially advanced disease, are not good ICD candidates.

"For a lot of patients with CKD, someone decides not to place an ICD. Those patients are not represented" in studies that focus on ICD recipients, said Dr. Reynolds, who cochaired the session where the three studies were presented.

"For every CKD patients who gets an ICD, an electrophysiologist has made a decision that this was a CKD patient who could benefit. It’s hard to extrapolate from that" to all CKD patients, agreed Dr. Paul Varosy, director of electrophysiology at the Denver Veterans Administration Medical Center. "Patients with CKD who get ICDs are fundamentally different patients."

The aim of one of the studies reported was to assess ICD efficacy in CKD patients by comparing survival between patients with ICDs divided into stage III CKD (an estimated glomerular filtration rate [GFR] of 30-59 mL/min per 1.73 m²), stage IV or V (a GFR of less than mL/min per 1.73 m²), or no CKD. The study examined 3-year follow-up data from 556 patients who received an ICD or similar device during 2006-2010 at the Minneapolis Veterans Administration Medical Center at the University of Minnesota. The series included 301 patients with no CKD, 230 with stage III CKD, and 25 patients with more severe CKD.

The analysis showed a similar incidence of appropriate and inappropriate shocks from the ICDs in all three groups, and roughly similar efficacy of the ICDs for preventing sudden cardiac death, Dr. Selcuk Adabag said at the Annual Scientific Sessions of the American Heart Association. But mortality rates rose substantially higher as renal function worsened, rising from 17% in patients without CKD to 30% in those with stage III disease, and to 56% in those with stage IV or V disease, reported Dr. Adabag, a cardiac electrophysiologist at the University of Minnesota in Minneapolis.

The mortality differences were hardly surprising, and they likely have little direct relationship to the ICDs. "In patients with CKD, the worse the disease, the worse their prognosis. CKD is a complex disease with a whole range of [mortality] risk factors; preventable sudden cardiac death is just one" of the risks these patients face, Dr. Varosy said in an interview.

Selection bias makes the ICD observations questionable, Dr. Reynolds said. "I’m very concerned about concluding that ICDs have similar effectiveness in patients with stage IV or V CKD, compared with no CKD. The study had only 25 patients with stage IV or V disease. I have to think that there was some belief that these were good candidates and that there were a lot of other patients with severe CKD where a physician decided not to implant an ICD. Those patients are not represented in the data," Dr. Reynolds said.

Another study ran a pair of meta-analyses to explore the interactions of ICDs and CKD. In both analyses, CKD was defined as patients on dialysis, those with a creatinine clearance of less than 60 mL/kg per 1.73 m², or patients with a serum creatinine of at least 1.5 mg/dL.

The first analysis looked at the impact of ICDs on all-cause death among CKD patients at high risk for sudden cardiac death. The literature search found five reports that addressed this issue, in a total of 17,460 patients, which showed that ICD placement linked with a statistically significant 35% cut in total mortality, compared with similar, propensity-score matched patients who did not get ICDs, Dr. Nader Makki said at the meeting.

The second analysis involved 15 reports with 5,333 patients, and used multivariate adjustment to find that patients who had received an ICD and also had CKD had a statistically significant, 2.9-fold higher mortality rate than did ICD recipients without CKD, highlighting the high risk for nonarrhythmic death that CKD patients face, said Dr. Makki, a researcher at the University of Iowa in Iowa City. "Despite a paucity of randomized trials in the CKD population, these data support use of ICDs for the prevention of sudden cardiac death in patients at risk," he said. "We believe that ICDs are underused in this population," patients with CKD.

But these analyses are compromised by the data they used, said Dr. Reynolds. The data were "heavily weighted with a couple of large studies that used claims data. My concern is the potential for confounding by indication. Patients who get ICDs are somehow not as sick as those who don’t."

The third study looked at the impact of CKD on the aftermath of ICD replacement among the 1,744 patients enrolled in the REPLACE (Implantable Cardiac Pulse Generator Replacement Registry) trial, which tracked the incidence of complications following replacement of ICD generators and leads. Among the enrolled patients, researchers had renal-function data for 1,662 patients. About 80% had either mild or moderate renal dysfunction, while 6% had an estimated GFR less than 30 mL/min per 1.73 m², and the remained had normal renal function.

The incidence of complications was 15% overall, and the analysis showed roughly this rate across all strata of CKD severity. Even among patients with severe CKD, the complication rate was about 20%, and not significantly different from patients with even normal renal function, Dr. Suneet Mittal reported at the meeting. The split between major and minor complications also varied little by renal function, and the incidence of infections was 1%-2% across all five levels of renal function examined.

But CKD severity played a big role in 6-month mortality. A multivariate analysis showed that CKD stage played a significant role in survival; for each stage of worsened renal function the risk of death rose by 50%. (Other significant predictors were recent heart-failure hospitalization, severe heart failure, treatment with an antiarrhythmic drug, and history of cerebrovascular disease.) In actual numbers, the 6-month mortality rate was 2% among patients with either none or mild renal dysfunction that jumped to a 9% rate in patients with an estimated GFR of 15-29 mL/min per 1.73 m², and to 16% in those on dialysis, said Dr. Mittal, a cardiac electrophysiologist at Columbia University in New York. The analyses also showed that the increased mortality risk linked to severe CKD became apparent by a month after ICD replacement, and it was not driven by procedure-related complications.

"Why do these patients [with severe CKD] do so much worse? We thought that if we looked at all their complications we’d pick up the reason, but clearly there are things that we have not yet recognized," Dr. Mittal said. "There are a lot of things that we didn’t capture in the study."

Although the analysis shed little light on what was behind the high mortality risk in severe CKD patients, it effectively highlighted how risky ICD replacements are in any patient.

"The incidence of complications, 15%-23%, was striking. It’s substantially greater than what’s been seen in any registry-based data or administrative-based data," said Dr. Varosy. He cited the unique ability of this registry to capture all the minor complications following ICD replacement.

"These are frightening numbers [because] we tell patients that there is a very small risk" from ICD replacement procedures, commented Dr. Kalyanam Shivkumar, professor and director of the University of California Los Angeles Cardiac Arrhythmia Center.

Dr. Reynolds said that he has been a consultant to Medtronic and Biosense Webster and has received research grants from Edwards Lifesciences. Dr. Varosy had no disclosures. Dr. Adabag said that he had received research grants from Medtronic and Boston Scientific. Dr. Makki had no disclosures. The REPLACE registry was sponsored by Biotronik and Dr. Mittal said that has been a consultant to Biotronik. Dr. Shivkumar had no disclosures.

LOS ANGELES – Patients with advanced chronic kidney disease who receive an implantable cardioverter defibrillator are not like other patients who receive these devices.

But results from a trio of studies reported at the meeting show how hard it is for researchers to get a clear handle on what makes chronic kidney disease patients different, whether they have different outcomes than defibrillator recipients without CKD, and what is the best approach for judiciously using implantable cardioverter defibrillators (ICDs) in patients with renal dysfunction.

Mitchel L. Zoler/IMNG Medical Media

It’s a particularly relevant question because CKD is linked with a substantially increased risk for sudden cardiac death (patients on dialysis have about a fivefold higher risk for sudden cardiac death, compared with patients with normal or mildly impaired renal function), but patients with CKD were excluded from the major trials that proved the efficacy of ICDs for preventing sudden cardiac death. The lack of data from randomized, prospective trials leaves questions about the efficacy, and perhaps as importantly, the cost efficiency of ICDs in CKD patients.

"Should patients with CKD get ICDs? I don’t know the answer, but I don’t think that, categorically, patients with CKD should be excluded from ICD treatment," commented Dr. Matthew Reynolds, a cardiac electrophysiologist and director of the Economics and Quality of Life Research Center at Harvard Medical School in Boston. But many patients with CKD, especially advanced disease, are not good ICD candidates.

"For a lot of patients with CKD, someone decides not to place an ICD. Those patients are not represented" in studies that focus on ICD recipients, said Dr. Reynolds, who cochaired the session where the three studies were presented.

"For every CKD patients who gets an ICD, an electrophysiologist has made a decision that this was a CKD patient who could benefit. It’s hard to extrapolate from that" to all CKD patients, agreed Dr. Paul Varosy, director of electrophysiology at the Denver Veterans Administration Medical Center. "Patients with CKD who get ICDs are fundamentally different patients."

The aim of one of the studies reported was to assess ICD efficacy in CKD patients by comparing survival between patients with ICDs divided into stage III CKD (an estimated glomerular filtration rate [GFR] of 30-59 mL/min per 1.73 m²), stage IV or V (a GFR of less than mL/min per 1.73 m²), or no CKD. The study examined 3-year follow-up data from 556 patients who received an ICD or similar device during 2006-2010 at the Minneapolis Veterans Administration Medical Center at the University of Minnesota. The series included 301 patients with no CKD, 230 with stage III CKD, and 25 patients with more severe CKD.

The analysis showed a similar incidence of appropriate and inappropriate shocks from the ICDs in all three groups, and roughly similar efficacy of the ICDs for preventing sudden cardiac death, Dr. Selcuk Adabag said at the Annual Scientific Sessions of the American Heart Association. But mortality rates rose substantially higher as renal function worsened, rising from 17% in patients without CKD to 30% in those with stage III disease, and to 56% in those with stage IV or V disease, reported Dr. Adabag, a cardiac electrophysiologist at the University of Minnesota in Minneapolis.

The mortality differences were hardly surprising, and they likely have little direct relationship to the ICDs. "In patients with CKD, the worse the disease, the worse their prognosis. CKD is a complex disease with a whole range of [mortality] risk factors; preventable sudden cardiac death is just one" of the risks these patients face, Dr. Varosy said in an interview.

Selection bias makes the ICD observations questionable, Dr. Reynolds said. "I’m very concerned about concluding that ICDs have similar effectiveness in patients with stage IV or V CKD, compared with no CKD. The study had only 25 patients with stage IV or V disease. I have to think that there was some belief that these were good candidates and that there were a lot of other patients with severe CKD where a physician decided not to implant an ICD. Those patients are not represented in the data," Dr. Reynolds said.

Another study ran a pair of meta-analyses to explore the interactions of ICDs and CKD. In both analyses, CKD was defined as patients on dialysis, those with a creatinine clearance of less than 60 mL/kg per 1.73 m², or patients with a serum creatinine of at least 1.5 mg/dL.

The first analysis looked at the impact of ICDs on all-cause death among CKD patients at high risk for sudden cardiac death. The literature search found five reports that addressed this issue, in a total of 17,460 patients, which showed that ICD placement linked with a statistically significant 35% cut in total mortality, compared with similar, propensity-score matched patients who did not get ICDs, Dr. Nader Makki said at the meeting.

The second analysis involved 15 reports with 5,333 patients, and used multivariate adjustment to find that patients who had received an ICD and also had CKD had a statistically significant, 2.9-fold higher mortality rate than did ICD recipients without CKD, highlighting the high risk for nonarrhythmic death that CKD patients face, said Dr. Makki, a researcher at the University of Iowa in Iowa City. "Despite a paucity of randomized trials in the CKD population, these data support use of ICDs for the prevention of sudden cardiac death in patients at risk," he said. "We believe that ICDs are underused in this population," patients with CKD.

But these analyses are compromised by the data they used, said Dr. Reynolds. The data were "heavily weighted with a couple of large studies that used claims data. My concern is the potential for confounding by indication. Patients who get ICDs are somehow not as sick as those who don’t."

The third study looked at the impact of CKD on the aftermath of ICD replacement among the 1,744 patients enrolled in the REPLACE (Implantable Cardiac Pulse Generator Replacement Registry) trial, which tracked the incidence of complications following replacement of ICD generators and leads. Among the enrolled patients, researchers had renal-function data for 1,662 patients. About 80% had either mild or moderate renal dysfunction, while 6% had an estimated GFR less than 30 mL/min per 1.73 m², and the remained had normal renal function.

The incidence of complications was 15% overall, and the analysis showed roughly this rate across all strata of CKD severity. Even among patients with severe CKD, the complication rate was about 20%, and not significantly different from patients with even normal renal function, Dr. Suneet Mittal reported at the meeting. The split between major and minor complications also varied little by renal function, and the incidence of infections was 1%-2% across all five levels of renal function examined.

But CKD severity played a big role in 6-month mortality. A multivariate analysis showed that CKD stage played a significant role in survival; for each stage of worsened renal function the risk of death rose by 50%. (Other significant predictors were recent heart-failure hospitalization, severe heart failure, treatment with an antiarrhythmic drug, and history of cerebrovascular disease.) In actual numbers, the 6-month mortality rate was 2% among patients with either none or mild renal dysfunction that jumped to a 9% rate in patients with an estimated GFR of 15-29 mL/min per 1.73 m², and to 16% in those on dialysis, said Dr. Mittal, a cardiac electrophysiologist at Columbia University in New York. The analyses also showed that the increased mortality risk linked to severe CKD became apparent by a month after ICD replacement, and it was not driven by procedure-related complications.

"Why do these patients [with severe CKD] do so much worse? We thought that if we looked at all their complications we’d pick up the reason, but clearly there are things that we have not yet recognized," Dr. Mittal said. "There are a lot of things that we didn’t capture in the study."

Although the analysis shed little light on what was behind the high mortality risk in severe CKD patients, it effectively highlighted how risky ICD replacements are in any patient.

"The incidence of complications, 15%-23%, was striking. It’s substantially greater than what’s been seen in any registry-based data or administrative-based data," said Dr. Varosy. He cited the unique ability of this registry to capture all the minor complications following ICD replacement.

"These are frightening numbers [because] we tell patients that there is a very small risk" from ICD replacement procedures, commented Dr. Kalyanam Shivkumar, professor and director of the University of California Los Angeles Cardiac Arrhythmia Center.

Dr. Reynolds said that he has been a consultant to Medtronic and Biosense Webster and has received research grants from Edwards Lifesciences. Dr. Varosy had no disclosures. Dr. Adabag said that he had received research grants from Medtronic and Boston Scientific. Dr. Makki had no disclosures. The REPLACE registry was sponsored by Biotronik and Dr. Mittal said that has been a consultant to Biotronik. Dr. Shivkumar had no disclosures.

LOS ANGELES – Patients with advanced chronic kidney disease who receive an implantable cardioverter defibrillator are not like other patients who receive these devices.

But results from a trio of studies reported at the meeting show how hard it is for researchers to get a clear handle on what makes chronic kidney disease patients different, whether they have different outcomes than defibrillator recipients without CKD, and what is the best approach for judiciously using implantable cardioverter defibrillators (ICDs) in patients with renal dysfunction.

Mitchel L. Zoler/IMNG Medical Media

It’s a particularly relevant question because CKD is linked with a substantially increased risk for sudden cardiac death (patients on dialysis have about a fivefold higher risk for sudden cardiac death, compared with patients with normal or mildly impaired renal function), but patients with CKD were excluded from the major trials that proved the efficacy of ICDs for preventing sudden cardiac death. The lack of data from randomized, prospective trials leaves questions about the efficacy, and perhaps as importantly, the cost efficiency of ICDs in CKD patients.

"Should patients with CKD get ICDs? I don’t know the answer, but I don’t think that, categorically, patients with CKD should be excluded from ICD treatment," commented Dr. Matthew Reynolds, a cardiac electrophysiologist and director of the Economics and Quality of Life Research Center at Harvard Medical School in Boston. But many patients with CKD, especially advanced disease, are not good ICD candidates.

"For a lot of patients with CKD, someone decides not to place an ICD. Those patients are not represented" in studies that focus on ICD recipients, said Dr. Reynolds, who cochaired the session where the three studies were presented.

"For every CKD patients who gets an ICD, an electrophysiologist has made a decision that this was a CKD patient who could benefit. It’s hard to extrapolate from that" to all CKD patients, agreed Dr. Paul Varosy, director of electrophysiology at the Denver Veterans Administration Medical Center. "Patients with CKD who get ICDs are fundamentally different patients."

The aim of one of the studies reported was to assess ICD efficacy in CKD patients by comparing survival between patients with ICDs divided into stage III CKD (an estimated glomerular filtration rate [GFR] of 30-59 mL/min per 1.73 m²), stage IV or V (a GFR of less than mL/min per 1.73 m²), or no CKD. The study examined 3-year follow-up data from 556 patients who received an ICD or similar device during 2006-2010 at the Minneapolis Veterans Administration Medical Center at the University of Minnesota. The series included 301 patients with no CKD, 230 with stage III CKD, and 25 patients with more severe CKD.

The analysis showed a similar incidence of appropriate and inappropriate shocks from the ICDs in all three groups, and roughly similar efficacy of the ICDs for preventing sudden cardiac death, Dr. Selcuk Adabag said at the Annual Scientific Sessions of the American Heart Association. But mortality rates rose substantially higher as renal function worsened, rising from 17% in patients without CKD to 30% in those with stage III disease, and to 56% in those with stage IV or V disease, reported Dr. Adabag, a cardiac electrophysiologist at the University of Minnesota in Minneapolis.

The mortality differences were hardly surprising, and they likely have little direct relationship to the ICDs. "In patients with CKD, the worse the disease, the worse their prognosis. CKD is a complex disease with a whole range of [mortality] risk factors; preventable sudden cardiac death is just one" of the risks these patients face, Dr. Varosy said in an interview.

Selection bias makes the ICD observations questionable, Dr. Reynolds said. "I’m very concerned about concluding that ICDs have similar effectiveness in patients with stage IV or V CKD, compared with no CKD. The study had only 25 patients with stage IV or V disease. I have to think that there was some belief that these were good candidates and that there were a lot of other patients with severe CKD where a physician decided not to implant an ICD. Those patients are not represented in the data," Dr. Reynolds said.

Another study ran a pair of meta-analyses to explore the interactions of ICDs and CKD. In both analyses, CKD was defined as patients on dialysis, those with a creatinine clearance of less than 60 mL/kg per 1.73 m², or patients with a serum creatinine of at least 1.5 mg/dL.

The first analysis looked at the impact of ICDs on all-cause death among CKD patients at high risk for sudden cardiac death. The literature search found five reports that addressed this issue, in a total of 17,460 patients, which showed that ICD placement linked with a statistically significant 35% cut in total mortality, compared with similar, propensity-score matched patients who did not get ICDs, Dr. Nader Makki said at the meeting.

The second analysis involved 15 reports with 5,333 patients, and used multivariate adjustment to find that patients who had received an ICD and also had CKD had a statistically significant, 2.9-fold higher mortality rate than did ICD recipients without CKD, highlighting the high risk for nonarrhythmic death that CKD patients face, said Dr. Makki, a researcher at the University of Iowa in Iowa City. "Despite a paucity of randomized trials in the CKD population, these data support use of ICDs for the prevention of sudden cardiac death in patients at risk," he said. "We believe that ICDs are underused in this population," patients with CKD.

But these analyses are compromised by the data they used, said Dr. Reynolds. The data were "heavily weighted with a couple of large studies that used claims data. My concern is the potential for confounding by indication. Patients who get ICDs are somehow not as sick as those who don’t."

The third study looked at the impact of CKD on the aftermath of ICD replacement among the 1,744 patients enrolled in the REPLACE (Implantable Cardiac Pulse Generator Replacement Registry) trial, which tracked the incidence of complications following replacement of ICD generators and leads. Among the enrolled patients, researchers had renal-function data for 1,662 patients. About 80% had either mild or moderate renal dysfunction, while 6% had an estimated GFR less than 30 mL/min per 1.73 m², and the remained had normal renal function.

The incidence of complications was 15% overall, and the analysis showed roughly this rate across all strata of CKD severity. Even among patients with severe CKD, the complication rate was about 20%, and not significantly different from patients with even normal renal function, Dr. Suneet Mittal reported at the meeting. The split between major and minor complications also varied little by renal function, and the incidence of infections was 1%-2% across all five levels of renal function examined.

But CKD severity played a big role in 6-month mortality. A multivariate analysis showed that CKD stage played a significant role in survival; for each stage of worsened renal function the risk of death rose by 50%. (Other significant predictors were recent heart-failure hospitalization, severe heart failure, treatment with an antiarrhythmic drug, and history of cerebrovascular disease.) In actual numbers, the 6-month mortality rate was 2% among patients with either none or mild renal dysfunction that jumped to a 9% rate in patients with an estimated GFR of 15-29 mL/min per 1.73 m², and to 16% in those on dialysis, said Dr. Mittal, a cardiac electrophysiologist at Columbia University in New York. The analyses also showed that the increased mortality risk linked to severe CKD became apparent by a month after ICD replacement, and it was not driven by procedure-related complications.

"Why do these patients [with severe CKD] do so much worse? We thought that if we looked at all their complications we’d pick up the reason, but clearly there are things that we have not yet recognized," Dr. Mittal said. "There are a lot of things that we didn’t capture in the study."

Although the analysis shed little light on what was behind the high mortality risk in severe CKD patients, it effectively highlighted how risky ICD replacements are in any patient.

"The incidence of complications, 15%-23%, was striking. It’s substantially greater than what’s been seen in any registry-based data or administrative-based data," said Dr. Varosy. He cited the unique ability of this registry to capture all the minor complications following ICD replacement.

"These are frightening numbers [because] we tell patients that there is a very small risk" from ICD replacement procedures, commented Dr. Kalyanam Shivkumar, professor and director of the University of California Los Angeles Cardiac Arrhythmia Center.

Dr. Reynolds said that he has been a consultant to Medtronic and Biosense Webster and has received research grants from Edwards Lifesciences. Dr. Varosy had no disclosures. Dr. Adabag said that he had received research grants from Medtronic and Boston Scientific. Dr. Makki had no disclosures. The REPLACE registry was sponsored by Biotronik and Dr. Mittal said that has been a consultant to Biotronik. Dr. Shivkumar had no disclosures.