Implantable cardioverter defibrillators (ICDs) are effective at reducing arrhythmic death in patients with left ventricular dysfunction, but few studies have investigated the outcomes after ICD implantation in patients with chronic kidney disease (CKD). We conducted a 2-center retrospective study of 958 patients who had undergone ICD placement for primary prevention from the 2000 to 2006. The patients were stratified into 5 groups according to the CKD stage (stage 1, glomerular filtration [GFR] 90 to 120 ml/min; stage 2, GFR 60 to 89 ml/min; stage 3, GFR 30 to 59 ml/min; stage 4, GFR 15 to 29 ml/min; and stage 5, GFR 0 to 14 ml/min). The primary end point was death at 1 year. Of the 958 patients included in our analysis, 73 (7.6%) had died at 1 year. The mortality rate at 1 year increased with worsening CKD (1.8%, 5.3%, 9.0%, 22%, and 38% for stage 1 to 5, respectively, p <0.0001 for group). CKD was an independent predictor of mortality; hazard ratio 1.0, 1.075 (95% confidence interval 0.578 to 2.0), 1.372 (95% confidence interval 0.736 to 2.556), 3.092 (95% confidence interval 1.52 to 6.29), and 10.15 (95% confidence interval 4.25 to 24.23) for stage 1 to 5, respectively (p <0.0001 for group). Patients with CKD and left ventricular dysfunction have a poor prognosis despite ICD placement. The 1-year mortality increased as the renal function decreased. In conclusion, physicians should be cognizant of the prognosis when considering whether an ICD should be implanted in patients with CKD.
Heart failure (HF) affects 2% of the population in the United States, with a lifetime incidence of 20%. Despite optimal medical therapy, HF still has an annual mortality rate as great as 10%. Several trials have demonstrated the benefit of implantable cardioverter defibrillators (ICDs) in reducing the incidence of cardiovascular death in patients with left ventricular dysfunction. However, patients with chronic kidney disease (CKD) either have been excluded or were a very low proportion of those included in trials of ICD therapy. However, these patients are at a particularly high risk of cardiovascular death. Data from the Kaiser Permanente Renal Registry shows a greater than threefold increase in cardiovascular events for those with the most advanced CKD compared to those with a glomerular filtration rate (GFR) >60 ml/min. We sought to evaluate the effect of CKD on survival in patients undergoing ICD implantation for primary prevention.
Methods
The present study was a retrospective cohort study performed at 2 centers. The diagnosis codes for cardioverter-defibrillator implantation were reviewed from January 1, 2000 to December 31, 2006 to identify the patients. Social security numbers were used to assess the mortality data from the Social Security Death Index. The institutional review board at both institutions approved the study before we began data collection.
Adults (age >18 years) with left ventricular ejection fraction of ≤0.40 undergoing ICD placement for primary prevention of sudden cardiac death were included. Patients were excluded from the study cohort if the procedure was (1) ICD implantation for secondary prevention after sudden cardiac death or symptomatic ventricular arrhythmias; (2) ICD implantation for primary prevention in patients with arrhythmogenic right ventricular dysplasia, congenital long QT syndrome, Brugada syndrome, or hypertrophic cardiomyopathy with a preserved ejection fraction; (3) ICD generator replacement only; (4) ICD lead revision or extraction; or (5) a biventricular ICD upgrade.
The data were collected by manual review of the electronic medical records at both institutions. The information collected included age, gender, weight, creatinine (most recent within 180 days before implantation), and race. The GFR was calculated using the Modification of Diet in Renal Disease study group formula using the data closest to the date of ICD implantation. The patients were stratified into 5 groups according to the calculated GFR and corresponding CKD stage: stage 1, GFR 90 to 120 ml/min; stage 2, GFR 60 to 89 ml/min; stage 3, GFR 30 to 59 ml/min; stage 4, GFR 15 to 29 ml/min; and stage 5, GFR 0 to 14 ml/min, with latter constituting end-stage kidney disease.
The patients were also assessed for the following; diabetes mellitus, hypertension, coronary artery disease, myocardial infarction occurring within 30 days before ICD implantation, type of cardiomyopathy (ischemic vs nonischemic), left ventricular ejection fraction, reported history of peripheral arterial disease, and device-related complications requiring lead extraction, prolonged antibiotics, or other interventions.
The United States Social Security Death Index was used to determine the date of death, if it had occurred. The survival time was calculated according to the last date of contact (if the patient was still alive) or the date of death after ICD implantation. All patients included in the present study had undergone ICD implantation ≥1 year before data collection to allow the assessment of 1-year mortality. The primary outcome of the study was the 1-year mortality, with a secondary outcome of the mean survival time. The cause of death was analyzed stratified by CKD stage. Death was categorized as noncardiac, progressive HF, arrhythmic, or unknown.
To adjust for study center variability, the homogeneity between the 2 centers was checked using the 2-sample t test for continuous variables and the chi-square test for categorical variables. The data from the 2 centers would be pooled if not significantly different. Otherwise, a variable indicating each center for each patient would be included in the analysis. The demographic and baseline characteristics were summarized by center using descriptive statistics: the mean ± SD or median (range) for continuous variables and the frequency (percentage) for categorical variables. The Cox proportional hazard model was used to assess CKD as an independent predictor of mortality. Logistic regression analysis was used to compare the survival differences among patients with advanced CKD to those with less-severe disease (e.g., stage 1 vs stage 5). Kaplan-Meier curves were used to assess 1-year survival and the median survival time. Fisher’s exact test was used to analyze the cause of death among the CKD stages. p values <0.05 were used to indicate statistical significance. Statistical Analysis Systems, version 9.2 (SAS Institute, Cary, North Carolina) was used for data management and analysis.
Results
All patients undergoing ICD implantation within the study period (n = 1,162) were screened for inclusion. Of these, 204 patients were excluded because the ICD was implanted for secondary prevention or unstable ventricular dysrhythmias. The remaining 958 patients met the inclusion criteria, and the data were extracted from their records. Data collection was complete for these 958 patients included in the final analysis. No differences were found in any of the variables among the 2 centers; thus, the results were pooled for analysis.
The baseline characteristics of the patients are listed in Table 1 . Age, the presence of diabetes mellitus, and a history of peripheral arterial disease were significantly more likely with worsening CKD stage ( Table 2 ), with the exception that patients with stage 5 CKD were slightly younger. Of the 958 patients included in the analysis, 73 (7.6%) had died within the first year after ICD implantation. A stepwise increase in 1-year mortality was seen for patients with worse CKD (p <0.0001). The median survival time for the patients with stage 1 to 5 CKD was 78, 90, 80, 42, and 21 months, respectively (p <0.0001). Figure 1 illustrates the progressive increase in mortality with time out to 36 months, after which the survival rate for those with Stage 5 CKD was 0%. The likelihood of death at 1 year was significantly greater for patients with stage 4 or 5 CKD than for those with stage 1 CKD; however, this effect was not seen for patients with stage 2 or 3 CKD ( Table 3 ). The risk of death within 1 year for patients with Stage 5 CKD was significantly greater than that for those with stage 1 CKD (odds ratio 35, 95% confidence interval 5.85 to 209.6, p <0.0001). Of the 73 patients who reached the primary end point of 1-year mortality, 22 (30%) had an undetermined cause of death. Of the remaining 51 patients for whom the cause of death was known, no significant differences were seen among the CKD stages (p = 0.27).
| Variable | Baseline (n = 958) |
|---|---|
| Median age (years) | 67 |
| White | 88% |
| Coronary artery disease ⁎ | 81% |
| Diabetes mellitus | 36% |
| Hypertension | 72% |
| Peripheral arterial disease | 19% |
| Ischemic cardiomyopathy | 75% |
| Myocardial infarction (within 30 days of implant) | 5% |
⁎ Defined as any degree of stenosis on coronary angiogram or documentation in medical record.
| Variable | CKD Stage | p Value | ||||
|---|---|---|---|---|---|---|
| 1 (n = 114) | 2 (n = 432) | 3 (n = 345) | 4 (n = 54) | 5 (n = 13) | ||
| Mean age (years) | 59 ± 13 | 66 ± 12 | 71 ± 10 | 72 ± 9.4 | 65 ± 7.5 | <0.0001 |
| Diabetes mellitus | 29 (25) | 143 (33) | 128 (37) | 29 (53) | 11 (84) | <0.0001 |
| Ejection fraction | 29.2 ± 8.3 | 27.8 ± 8 | 26.7 ± 7.9 | 27.2 ± 8.1 | 27.9 ± 6.5 | 0.0409 |
| Ischemic cardiomyopathy | 80 (70) | 324 (75) | 255 (74) | 45 (83) | 11 (84) | 0.3951 |
| Peripheral arterial disease | 13 (11) | 65 (15) | 83 (24) | 17 (31) | 6 (46) | <0.0001 |
