Nonischemic cardiomyopathy and implantable cardioverter-defibrillators
Nonischemic cardiomyopathy (NICM) represents a heterogeneous group of conditions that can cause left ventricular dysfunction.1 Studies have suggested that patients with NICM who have received implantable cardioverter-defibrillator (ICD) therapy in addition to standard medical therapy have better survival.2,3 There are few data on whether the duration of NICM is related to the degree of benefit from ICD insertion, however. Current reimbursement policies and qualification for ICD insertion are based on the duration of the NICM diagnosis. The Centers for Medicare and Medicaid Services has approved ICD insertion for patients with NICM of at least 9 months’ duration.4,5 If patients with NICM of 3 to 9 months’ duration are enrolled in a registry, they may undergo ICD insertion. Patients who have been diagnosed with NICM within a 3-month period do not qualify for ICD insertion at this time. We conducted a study to assess whether the duration of NICM is associated with the degree of benefit obtained from ICD insertion using a post-hoc analysis of data from the Defibrillators in Nonischemic Cardiomyopathy Treatment Evaluation (DEFINITE) trial.
Subjects and methods
Details of the DEFINITE trial have been published previously.2 Subjects enrolled in the trial (n = 458) were randomly assigned to 1 of 2 groups: (1) standard medical therapy for heart failure, including β blockers and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers; or (2) standard medical therapy plus a single-chamber ICD. To be included in the study, subjects had to have a left ventricular ejection fraction of < 36%, the presence of premature ventricular complexes or nonsustained ventricular tachycardia, a history of symptomatic heart failure, and the presence of NICM. Criteria for the minimum length of time from diagnosis of NICM to randomization were not specified in the trial protocol. If the treating physician did not believe that a reversible process was the cause of left ventricular dysfunction, the subject was assigned to a treatment group with the duration of cardiomyopathy reported to the nearest month. In this post-hoc analysis, all subjects were classified into one of the following categories according to the duration of NICM diagnosis at the time of randomization: ≤ 3 months, > 3 months, ≤ 9 months, or > 9 months. The methodology of data analysis has been reported previously.6 Baseline characteristics, Kaplan-Meier survival curves, the effect of treatment assignment, and differences in all-cause mortality were compared in subjects with shorter and longer durations of NICM diagnosis. The analyses were adjusted for those variables that were different between treatment arms for subjects with varying durations of NICM diagnoses: QRS duration, diabetes, the subject’s New York Heart Association (NYHA) functional class, and white race.
The period from diagnosis of NICM to randomization was a mean of 2.9 ± 4 years, with a median of 1 year and a range of from < 1 month to 38.5 years. The interquartile range was 2 months to 4.6 years. The median duration of diagnosis was 20 months in the standard medical therapy group compared with 8 months in the ICD group (P = .032 based on the Wilcoxon rank-sum test). The demographic characteristics of the subjects are shown in Table 1. Subjects with longer durations of NICM had a higher prevalence of diabetes, more severe NYHA functional class, and a longer QRS duration.
Survival was not significantly different among subjects with varying durations of cardiomyopathy. For subjects with ≤ 3 months’ duration of NICM diagnosis, 10.2% died after 2.5 years compared with 16.0% of subjects with NICM for > 3 months (hazard ratio [HR] = 1.41; 95% confidence interval [CI], 0.82-2.41; P = .22). In subjects with an NICM diagnosis of ≤ 9 months’ duration, 11.2% died at 2.5 years compared with 16.8% of those with cardiomyopathy for > 9 months (HR = 1.38; 95% CI, 0.85-2.24; P = .194). After controlling for assignment to ICD versus standard therapy, survival overall between the recently and remotely diagnosed groups was similar (HR = 1.36; 95% CI, 0.79-2.34 and HR = 1.32; 95% CI, 0.81-2.15 at 3- and 9-month cut points; P = .26 at both cut points). Survival overall between the recently and remotely diagnosed subjects treated with standard therapy alone was also similar (HR = 1.00; 95% CI, 0.51-1.96; P = 1.00 and HR = 1.04; 95% CI, 0.55-1.95; P = .91 at 3- and 9-month cut points).
As shown in Table 2, Figure 1, and Figure 2, ICD insertion correlated with a reduced risk of death in subjects with a shorter duration of NICM diagnosis. In subjects with diagnosed cardiomyopathy for ≤ 3 months, there was a significantly reduced risk of death (HR = 0.37; 95% CI, 0.140-0.998; P = .049). In subjects with diagnosed cardiomyopathy for ≤ 9 months, there was a borderline significant difference in risk of death (HR = 0.48; 95% CI, 0.230-1.025; P = .058). In subjects diagnosed with NICM for longer periods, ICD insertion was not significantly related to risk of death (HR = 0.82; 95% CI, 0.47-1.43; P = .48 and HR = 0.86; 95% CI, 0.46-1.94; P = .64 at 3- and 9-month cut points).
Although subjects with a shorter duration of NICM diagnosis in the ICD therapy group had a significantly reduced risk of death, the nonsignificant interaction term between the time to diagnosis and ICD benefit indicates that there was no significant difference in occurrences of death in subjects with varying durations of NICM diagnosis who were assigned to ICD therapy (P = .17 and P = .25 at 3- and 9-month cut points, respectively). In other words, ICD therapy significantly decreased the risk of death in subjects with a shorter duration of NICM diagnosis; however, the data do not show a significant difference in the degree of effect on risk of death among subjects with varying durations of NICM diagnosis. These results did not measurably change after adjusting for covariates. The important difference to note from the adjusted data is that there was a significant benefit from ICD therapy in subjects with a diagnosis of NICM for ≤ 9 months (HR = 0.43; 95% CI, 0.200-0.093; P = .032), whereas with the unadjusted data, there was only a trend toward significance.
Our study results show that there was a benefit from ICD insertion in subjects who were recently diagnosed with NICM. There was a significant improvement in survival among subjects with recently diagnosed NICM whether a recent diagnosis was considered < 3 months’ or < 9 months’ duration after controlling for covariates. As shown in Figures 1 and 2, this benefit was not caused by death soon after randomization, because the survival curves continued to diverge over time. Although ICD benefit was significant in subjects with recently diagnosed NICM and not in subjects with remotely diagnosed NICM, the relative difference in ICD benefit between these groups was not significant. Therefore, we can state that ICD insertion is beneficial for patients with recently diagnosed NICM, but we cannot state that there is a different degree of benefit from ICD therapy between those with recently versus remotely diagnosed NICM.
Our subject groups had different baseline characteristics. There was a lower percentage of NYHA functional class 1 heart failure and longer QRS duration in subjects with a longer duration of NICM diagnosis, which may very well be expected with a longer duration of cardiomyopathy. Differences that are likely to be the result of chance are that subjects with a longer duration of cardiomyopathy were more likely to be diabetic and to be nonwhite. Adjusting for differences between the 2 groups did not considerably affect the HR for ICD benefit in either the recently or the remotely diagnosed groups.
Our study was limited by the fact that the DEFINITE trial was not powered to assess the time from diagnosis to randomization, and the results of post-hoc analyses should be interpreted with caution. Nonetheless, our finding that there is benefit from ICD insertion in subjects who have recently been diagnosed with NICM is strengthened by the fact that the benefit was not significantly changed after adjusting for covariates. The duration of cardiomyopathy differed between the standard therapy and ICD groups, and the numbers of subjects in each treatment assignment in recently and remotely diagnosed groups were different (Table 1). Our results should be valid, however, given that use of the Cox regression analysis does not necessitate that sample sizes be equivalent.
There is risk of death in subjects with a recent diagnosis of NICM, and ICD implantation seems to reduce this risk. The DEFINITE trial suggested that there was a 35% decrease in mortality with ICD therapy in subjects with NICM, but the results did not reach significance.2 A recent meta-analysis showed that ICD insertion decreases mortality in patients with NICM.7 Our results indicate that patients with recently diagnosed NICM may benefit from ICD therapy.
The results of this study and those of other recent trials indicate that implantation of a cardioverter-defibrillator may be suitable for patients with NICM and may improve survival. Insertion of an ICD may be appropriate in these patients as soon as they are identified and as soon as reversible causes of left ventricular dysfunction have been excluded.