Abstract
Background: The prospective observational Nippon Storm Study aggregated clinical data from Japanese patients receiving implantable cardioverter-defibrillator (ICD) therapy. This study investigated the usefulness of prophylactic ICD therapy in patients with non-ischemic heart failure (NIHF) enrolled in the study.
Methods and Results: We analyzed 540 NIHF patients with systolic dysfunction (left ventricular ejection fraction <50%). Propensity score matching was used to select patient subgroups for comparison; 126 patients were analyzed in each of the primary (PP) and secondary (SP) prophylaxis groups. The incidence of appropriate ICD therapy during follow-up in the PP and SP groups was 21.4% and 31.7%, respectively (P=0.044). The incidence of electrical storm (ES) was higher in SP than PP patients (P=0.024). Cox proportional hazard analysis revealed that increased serum creatinine in SP patients (hazard ratio [HR] 1.18; 95% confidence interval [CI] 1.02–1.33; P=0.013) and anemia in PP patients (HR 0.92; 95% CI 0.86–0.98; P=0.008) increased the likelihood of appropriate ICD therapy, whereas long-lasting atrial fibrillation in PP patients (HR, 0.64 [95% CI, 0.45–0.91], P=0.013) decreased that likelihood.
Conclusions: In propensity score-matched Japanese NIHF patients, the incidence of appropriate ICD therapy and ES was significantly higher in SP than PP patients. Impaired renal function in SP patients and anemia in PP patients increased the likelihood of appropriate ICD therapy, whereas long-lasting atrial fibrillation reduced that likelihood in PP patients.
The implantable cardioverter-defibrillator (ICD) is an established therapy for the prevention of sudden cardiac death (SCD) due to fatal ventricular tachyarrhythmias (VA) in patients with ischemic cardiomyopathy.1,2 However, the role of primary prophylactic ICD therapy in non-ischemic heart failure (NIHF) patients with reduced left ventricular ejection fraction (LVEF) is far less clear. Although approximately 70% of heart failure (HF) patients have non-ischemic etiology according to Japan Cardiac Device Therapy Registry (JCDTR) data in Japan, there is still no large-scale study on the usefulness of prophylactic ICD therapy in Japan.3
Editorial p 101
The Nippon Storm Study was a prospective observational study designed to aggregate clinical data from Japanese patients receiving ICD therapy.4–6 In all, 1,570 patients from 48 ICD centers in Japan were enrolled in the Nippon Storm Study. Of the 813 patients with non-ischemic structural heart disease registered in the Nippon Storm Study, we focused on 540 patients with systolic dysfunction, defined as LVEF <50%. To investigate the usefulness of prophylactic ICD therapy in NIHF patients with systolic dysfunction, we used propensity score matching to compare the incidence of ICD therapies and prognosis between the primary (PP) and secondary (SP) prophylaxis groups. In addition, we used Cox proportional hazard analysis to investigate predictors of ICD therapy in the 540 patients with systolic dysfunction.
Methods
Registration
The details of the overall design of the Nippon Storm Study have been published elsewhere. Briefly, the Nippon Storm Study was organized by the Japanese Heart Rhythm Society and Japanese Society of Electrocardiology. Web site registration of patients was conducted at 48 Japanese ICD centers, and the Japanese Heart Rhythm Society collected data from physicians who input patient data. Data collection began in October 2010 and data accumulation for the registry was terminated in July 2012. According to the guidelines for the implantation of an ICD, the indication for and purpose of implantation were determined by attending cardiologists in each center.7
ICD Programing
Model selection and detailed programming of transvenous ICDs were at the attending physician’s discretion. Some discrimination algorithms, such as PR Logic and Wavelet (Medtronic, Minneapolis, MN, USA), Rhythm ID (Boston Scientific, Marlborough, MA, USA), and Morphology Discrimination plus AV Rate Branch (Abbott, Sylmar, CA, USA), were used. The ventricular fibrillation (VF) zone was >188 to 200 beats/min with at least 1 train of antitachycardia pacing (ATP) before the shock; the ventricular tachycardia (VT) zone was >140 to 160 beats/min with at least 3 trains of ATP before the shock. VF and VT zones could be modified according to an individual patient’s background. Each electrical storm (ES) was managed by physicians according to their preference. If an ES was thought to be triggered by myocardial ischemia, acute decompensated HF, or electrolyte disorder, they were corrected immediately. If needed, an antiarrhythmic drug regimen comprising β-blockers, amiodarone, and lidocaine, sequentially or in combination, was administered. Some patients may have undergone catheter ablation in the acute phase of an ES.
Follow-up
For precise follow-up, we constructed a new tracking system called “Chaser”, which was intended to minimize the loss of follow-up data. The data of interventions (both appropriate and inappropriate) from the ICD were sent at a maximum interval of 6 months to the office of the Japanese Heart Rhythm Society through the web site. The ICD interventions were classified as ATP, low-energy shocks, and high-energy shocks. An ES was defined as the occurrence of at least 3 separate episodes of VT/VF within a 24-h period.8 Every ES was blindly adjudicated by 2 electrophysiologists (N.A., K.S.) based on the 10 intracardiac electrograms at the time of the event.
Data Analysis
Patient characteristics at baseline were analyzed, including age, sex, underlying heart disease, purpose of indication (primary or secondary), LVEF, severity of HF, underlying heart rhythm, types of implanted defibrillator, medications, and blood test findings, such as hemoglobin, serum creatinine, B-type natriuretic peptide (BNP).
To minimize the effect of treatment selection bias and confounding effects, propensity score matching was performed with NIHF patients for whom complete baseline variables were available (n=540). We used logistic multivariable regression with prophylactic ICD therapy (PP and SP) as the dependent variable and 13 baseline variables (age, male sex, New York Heart Association [NYHA] functional class, LVEF, number of patients with atrial fibrillation [AF], hemoglobin, serum creatinine, BNP, number of cardiac resynchronization therapy with defibrillator (CRT-D) implantations, use of angiotensin-converting enzyme inhibitors [ACEI] or angiotensin II receptor blockers [ARB], β-blockers, mineralocorticoid receptor antagonist [MRA], Class III antiarrhythmic drugs) as independent predictors to calculate the propensity score (Table 1). Harrell’s C-statistic for the propensity score logistic regression model was 0.79. Patients were 1 : 1 greedy matched using the nearest-neighbor method.
Table 1. Baseline Characteristics of Unmatched and Propensity Score-Matched Cohorts of Patients With ICD Therapy as Primary or Secondary Prophylaxis
| |
Unmatched |
Matched |
| PP (n=352) |
SP (n=188) |
P value |
PP (n=126) |
SP (n=126) |
P value |
| Age (years) |
63.6±12.4 |
63.9±12.3 |
0.804 |
63.4±12.9 |
63.3±12.8 |
0.872 |
| Male sex |
255 (72) |
129 (69) |
0.371 |
85 (67) |
85 (67) |
1.000 |
| NYHA FC III/IV |
197 (56) |
66 (35) |
<0.0001 |
50 (40) |
55 (44) |
0.609 |
| LVEF (%) |
26.5±8.3 |
31.5±9.7 |
<0.0001 |
29.8±8.7 |
29.9±9.5 |
0.918 |
| AF |
57 (16) |
34 (18) |
0.630 |
22 (17) |
21 (17) |
1.000 |
| Hb (g/dL) |
13.1±2.0 |
13.0±2.0 |
0.179 |
13.1±1.9 |
13.0±2.0 |
0.655 |
| Creatinine (mg/dL) |
1.4±1.9 |
1.3±1.3 |
0.439 |
1.2±0.9 |
1.3±1.4 |
0.389 |
| BNP (pg/mL) |
647.4±898.4 |
587.1±898.3 |
0.483 |
636.4±1,130.2 |
569.4±715.0 |
0.574 |
| CRT-D |
260 (74) |
83 (44) |
<0.0001 |
71 (56) |
72 (56) |
1.000 |
| ACEI/ARB |
264 (75) |
109 (58) |
<0.0001 |
87 (69) |
80 (63) |
0.424 |
| β-blocker |
257 (72) |
134 (71) |
0.687 |
95 (75) |
91 (72) |
0.668 |
| MRA |
152 (43) |
78 (42) |
0.716 |
62 (49) |
52 (41) |
0.255 |
| Class III antiarrhythmic drugs |
105 (30) |
123 (65) |
<0.0001 |
73 (58) |
71 (57) |
1.000 |
Unless indicated otherwise, values are presented as the mean±SD or n (%). ACEI, angiotensin-converting enzyme inhibitors; AF, atrial fibrillation; ARB, angiotensin II receptor blocker; BNP, B-type natriuretic peptide; CRT-D, cardiac resynchronization therapy with defibrillator; FC, functional class; Hb, hemoglobin; ICD, implantable cardioverter-defibrillator; LVEF, left ventricular ejection fraction; MRA, mineralocorticoid receptor antagonist; NYHA, New York Heart Association; PP, primary prophylaxis; SP, secondary prophylaxis.
As the main purpose of this study, the incidence of ICD therapy was compared between the PP and SP groups of NIHF patients. For time-to-event outcomes, the Kaplan-Meier method was used to estimate the cumulative incidence of outcomes for each group, with the significance of differences in outcomes between groups analyzed using the log-rank test. Finally, prognosis was compared between the PP and SP groups of NIHF patients. Predictors of appropriate ICD therapy were analyzed using Cox proportional hazard analysis on the 540 patients at baseline.
Statistical Analysis
Continuous baseline variables are presented as the mean±SD and categorical baseline variables are presented as n (%). When any 2 groups were compared, the χ2
test was used for categorical variables and Student’s t-test was used for continuous variables. For time-to-event outcomes, survival curves were created using the Kaplan-Meier method, and log rank tests were used for statistical hypothesis testing. Hazard ratios (HR) and 95% confidence intervals (CIs) were calculated using the Cox proportional hazards model. Statistical analyses were performed using JMP version 16.0 (SAS Institute, Cary, NC, USA). Two-sided P<0.05 was considered statistically significant.
This study was conducted in accordance with the Declaration of Helsinki and approved by the institutional review board of each institution. All patients provided written informed consent.
Results
Study Cohort
We focused on 540 patients with NIHF with systolic dysfunction, defined as LVEF <50%. After propensity score matching, the analysis was restricted to 252 patients: 126 (50%) with ICD/CRT-D for PP and 126 (50%) with ICD/CRT-D for SP of SCD. A flow diagram of patient classification is shown in Figure 1.
Baseline Characteristics of Study Patients
The baseline characteristics of the overall cohort of patients with NIHF are presented in Table 2. The mean age was 64±12 years and 71% of patients were male. Mean LVEF was 28±9%. Most patients (83%) had moderate to severe HF symptoms (NYHA Class II/III) and the mean BNP was 623±898 pg/mL. CRT-D was implanted in 64% of patients. Most patients (65%) underwent ICD/CRT-D implantation for PP of SCD. The underlying non-ischemic heart disease was dilated cardiomyopathy in 333 patients (62%), cardiac sarcoidosis in 52 patients (10%), hypertrophic cardiomyopathy in 50 patients (9%), valvular heart disease in 33 patients (6%), hypertensive heart disease in 15 patients (3%), congenital heart disease in 13 patients (2%), non-ischemic cardiomyopathy with unknown cause in 9 patients (1%), arrhythmogenic right ventricular cardiomyopathy in 8 patients (1%), myocarditis in 5 patients, amyloidosis in 3 patients, and “other” in 19 patients (4%).
Table 2. Baseline Characteristics of All Patients With Non-Ischemic Heart Failure in the Present Study
| Age (years) |
64±12 |
| Male sex |
384 (71) |
| LVEF (%) |
28±9 |
| NYHA class |
| Class I |
61 (11) |
| Class II |
217 (40) |
| Class III |
230 (43) |
| Class IV |
32 (6) |
| AF |
91 (17) |
| QRS duration (ms) |
144±37 |
| ICD only |
122±32 |
| CRT-D |
158±33 |
| BNP (pg/mL) |
627±898 |
| Hb (g/dL) |
13±2 |
| Serum creatinine (mg/dL) |
1.4±1.7 |
| ICD |
197 (36) |
| CRT-D |
343 (64) |
| PP |
352 (65) |
| SP |
188 (35) |
| Medications |
| β-blockers |
391 (72) |
| ACEI/ARB |
373 (69) |
| MRA |
230 (43) |
| Class III antiarrhythmic drugs |
44 (8) |
Values are presented as the mean±SD or n (%). Abbreviations as in Table 1.
Many aspects of the baseline patient characteristics were differently distributed between the PP and SP groups (Table 1). Patients in the PP group had more severe HF and most patients underwent CRT-D implantation. Conversely, more patients in the SP than PP group were treated with Class III antiarrhythmic drugs. After propensity score matching, the baseline characteristics considered for propensity score calculation were almost equally distributed between the 2 groups.
Incidence of ICD Therapies in Propensity Score-Matched Cohorts
In propensity score-matched cohorts, the incidence of appropriate ICD therapy over a mean follow-up period of 775±291 days was 21.4% (n=27) and 31.7% (n=40) in the PP and SP groups, respectively (log-rank test, P=0.044; Figure 2). Conversely, in the mode-specific analysis of appropriate ICD therapy, the cumulative incidence of ATP therapy did not differ significantly between the PP and SP groups (16.7% patients vs. 21.4%, respectively; log-rank test, P=0.215); there was also no significant difference in shock therapy between the 2 groups (4.8% vs. 9.5%, respectively; log-rank test, P=0.105; Figure 3).
To clarify the impact of LVEF on the incidence of ICD therapy, patients were divided into 2 groups (LVEF ≤35% and LVEF 36–50%), and the cumulative incidence of appropriate ICD therapy was compared between the PP and SP groups. In patients with LVEF ≤35%, the incidence of appropriate ICD therapy was significantly higher in the SP than PP group (34.5% vs. 21.0%, respectively; log-rank test, P=0.026). Conversely, there was no significant difference between the PP and SP groups for patients with LVEF 36–50% (22.6% vs. 23.5%, respectively; log-rank test P=0.729; Figure 4).
In addition, there was no significant difference in the incidence of inappropriate ICD therapy between the PP and SP groups (8.7% vs. 10.3%, respectively; log-rank test, P=0.630).
Incidence of ES and Its Impact on Patient Prognosis
The cumulative incidence of ES was 3.2% (n=4) and 10.3% (n=13) in the PP and SP groups, respectively. The incidence of ES was significantly higher in the SP than PP group (log-rank test, P=0.024; Figure 5). During the follow-up period, the mortality rate in the PP and SP groups did not differ significantly (12.7% [n=16] and 13.5% [n=17], respectively; log-rank test, P=0.797). All-cause mortality included cardiovascular death in 16 patients (48%) and non-cardiovascular death in 17 patients (52%). The most common cause of cardiovascular death was worsening HF (n=12; 75%); arrhythmic death occurred in 3 patients. There was no association between the occurrence of ES and cardiovascular death.
Predictors for Appropriate ICD Therapy in Patients With NIHF
Multivariate analysis was performed on 540 patients at baseline to clarify the predictors for appropriate ICD therapy in each of the PP and SP groups. In the SP group, Cox proportional hazard analysis revealed that increased serum creatinine concentration (HR 1.18; 95% CI 1.02–1.33; P=0.013) significantly increased the likelihood of appropriate ICD therapy (Table 3A). In the PP group, anemia increased (HR 0.92; 95% CI 0.86–0.98; P=0.008), whereas long-lasting AF decreased (HR 0.64; 95% CI 0.45–0.91; P=0.013) the likelihood of appropriate ICD therapy (Table 3B).
Table 3. Results of Cox Proportional Hazard Analysis of Predictors of Appropriate ICD Therapy at Baseline in the SP Group (A) and the PP Group (B)
| |
Univariate |
Multivariate |
| HR |
95% CI |
P value |
HR |
95% CI |
P value |
| (A) SP group |
| Age |
0.99 |
0.97–1.01 |
0.255 |
|
|
|
| Male sex |
0.93 |
0.59–1.47 |
0.339 |
|
|
|
| NYHA FC III/IV |
1.17 |
0.76–1.80 |
0.471 |
|
|
|
| LVEF |
1.00 |
0.98–1.03 |
0.829 |
|
|
|
| AF |
0.78 |
0.43–1.41 |
0.403 |
|
|
|
| Hb |
0.95 |
0.85–1.05 |
0.302 |
|
|
|
| Creatinine |
1.19 |
1.02–1.34 |
0.012 |
1.18 |
1.02–1.33 |
0.013 |
| BNP |
1.00 |
0.99–1.00 |
0.694 |
|
|
|
| CRT |
0.69 |
0.45–1.07 |
0.102 |
0.70 |
0.45–1.08 |
0.109 |
| β-blocker |
0.96 |
0.60–1.53 |
0.863 |
|
|
|
| Class III antiarrhythmic drugs |
1.02 |
0.66–1.56 |
0.939 |
|
|
|
| (B) PP group |
| Age |
1.00 |
0.99–1.01 |
0.864 |
|
|
|
| Male sex |
0.88 |
0.67–1.15 |
0.339 |
|
|
|
| NYHA FC III/IV |
0.98 |
0.76–1.25 |
0.864 |
|
|
|
| LVEF |
0.99 |
0.98–1.01 |
0.398 |
|
|
|
| AF |
0.65 |
0.46–0.92 |
0.016 |
0.64 |
0.45–0.91 |
0.013 |
| Hb |
0.92 |
0.86–0.98 |
0.007 |
0.92 |
0.86–0.98 |
0.008 |
| Creatinine |
1.04 |
0.97–1.11 |
0.250 |
|
|
|
| BNP |
1.00 |
0.99–1.00 |
0.132 |
|
|
|
| CRT |
1.06 |
0.80–1.41 |
0.665 |
|
|
|
| β-blocker |
0.99 |
0.75–1.30 |
0.917 |
|
|
|
| Class III antiarrhythmic drugs |
1.24 |
0.95–1.61 |
0.108 |
|
|
|
CI, confidence interval; HR, hazard ratio. Other abbreviations as in Table 1.
Discussion
Major Findings
In propensity score-matched Japanese NIHF patients, the incidence of appropriate ICD therapy and ES were significantly higher in the SP than PP group. There was no significant difference in the incidence of inappropriate ICD therapy between the 2 groups. The most common reason for death was worsening HF, and the occurrence of ES was not associated with mortality. In the analysis of predictors of ICD therapy at baseline in all 540 patients, Cox proportional hazard analysis revealed that impaired renal function significantly increased the likelihood of appropriate ICD therapy in the SP group. In the PP group, anemia increased, whereas long-lasting AF decreased, the likelihood of appropriate ICD therapy.
Significance of PP by ICD for SCD in Japanese NIHF Patients
Advances in HF treatment over the past 20 years, represented by guideline-directed medical therapy (GDMT), have reduced SCD in patients with HF by 44%.9 ICD has also been shown to reduce SCD in HF with reduced LVEF (HFrEF) patients,10 but the usefulness of primary prophylactic ICD therapy in patients with non-ischemic cardiomyopathy is far less clear than in patients with ischemic cardiomyopathy. Recent meta-analyses have demonstrated the effectiveness of ICD therapy in reducing the risk of all-cause mortality,11 and in reducing the risk of all-cause mortality and arrhythmic death.12 However, most of the trials included in those meta-analyses were conducted over 15 years ago and may not have fully used the current optimal medical therapy for HFrEF as recommended by the latest guidelines.
Conversely, the Danish ICD Study in Patients With Dilated Cardiomyopathy (DANISH),13 in which 1,116 non-ischemic HFrEF patients undergoing GDMT were randomized to primary prophylactic ICD therapy (n=556) or to GDMT only (n=560), raised a new issue. Although DANISH demonstrated a significant risk reduction of SCD in the ICD compared with GDMT group (HR 0.50; 95% CI 0.31–0.82; P=0.005), there was no beneficial effect of ICD on all-cause mortality over 5 years of follow-up.13 DANISH allowed the use of cardiac resynchronization therapy (CRT) in both arms of the study. Most study patients (58%) had already been treated with CRT, and this may be one of the factors that obscured the survival benefit of ICD. Further, the disadvantages due to ICD-related complications, such as device infection, were demonstrated in a subanalysis. Nevertheless, a prespecified secondary analysis of DANISH demonstrated a significant reduction in mortality with an ICD in patients aged ≤70 years.14 A recently reported meta-analysis on primary prophylaxis of ICD in patients with NIHF, reflecting the results of DANISH, has shown a 25% relative risk reduction in mortality with ICD.15 Furthermore, a recent prospective propensity score-matched analysis from the Swedish Heart Failure Registry showed a clear benefit of primary prophylaxis ICD therapy on overall cardiovascular survival in patients with HFrEF.16 That Swedish study enrolled patients with indications for primary prophylaxis with ICD meeting the European Society of Cardiology criteria, and an ICD was implanted in only 10% of enrolled patients.16 The use of ICD as primary prophylaxis was associated with 27% and 12% reductions in 1- and 5-year all-cause mortality, respectively, as well as a 29% reduction in the 1-year (but not 5-year) risk of cardiovascular death. Notably, this survival benefit of primary prophylaxis ICD therapy was consistent in patients with vs. without ischemic heart disease, men vs. women, patients aged <75 ≥75 years, and those with vs. without CRT.16
In contrast with Western countries, there is still no large-scale study on the usefulness of prophylactic ICD therapy in Asia. Recent reports have shown that approximately 70% of patients have a non-ischemic etiology, and there was an apparent increase in the percentage of non-ischemic heart disease patients aged ≥75 years receiving ICD (from 10.9% in 2006 to 16.4% in 2016; P=0.0008) and CRT-D (from 17.1% in 2006 to 27.8% in 2016; P=0.0001).3 To clarify the significance of ICD as primary prophylaxis in Japanese NIHF patients, we used propensity score matching to minimize the impact of diversity in patients’ clinical characteristics and compared the usefulness of ICD between patients in the PP and SP groups. The results showed that the incidence of appropriate ICD therapy was significantly lower in PP than SP group, but appropriate ICD therapy was observed in 27 of 126 patients (21%) in the propensity score-matched PP group. The incidence of appropriate ICD therapy in the PP groups was close to that in Western studies. Furthermore, in the mode-specific analysis of ICD therapies, the ratio of ATP therapy to overall ICD therapy tended to be higher in the PP than SP group (78% vs. 68%, respectively). These results suggest that differences in the settings for tachycardia detection (e.g., detection rate or the length of the detection interval) affect the results, but also show the considerable risk of developing fatal VA in primary prophylaxis patients. However, in the present study, unlike in DANISH and the Swedish Heart Failure Registry, the mortality benefit of ICD use in the PP group could not be clarified. This may be explained by the short (>2 years) observation period in the present study and the fact that 52% of deaths were due to non-cardiovascular causes.
In addition, in the present study, GDMT for HFrEF was clearly suboptimal compared with recent Western trials, which contributed to the high incidence of appropriate ICD therapy in the PP group. Neurohormonal antagonists significantly reduce the risk of SCD, and current guidelines require that HFrEF patients receive GDMT before consideration of ICD implantation for primary prophylaxis. However, in clinical practice in Japan, the rate of use of such medications and the doses achieved are lower than in Western trials. In the Change the Management of Patients with Heart Failure (CHAMP-HF) Registry, which analyzed 2,588 patients with chronic HFrEF, renin-angiotensin system (RAS) inhibitors and β-blockers were prescribed to 73% and 67% of patients, respectively, and MRAs were prescribed to 33% of patients.17 In addition, only 17% of patients receiving RAS inhibitors and 28% of patients receiving β-blockers achieved the target dose recommended by the guidelines, and in only 1% of these patients were GDMTs simultaneously administered with the target dose. Therefore, it is considered that the results of the present study are in line with the actual situation of daily clinical practice, and may also imply the potential role of ICD in the primary prophylaxis of SCD in NIHF patients.
Predictors of ICD Therapy in NIHF Patients
Many Western studies of NIHF patients treated with primary prophylaxis ICD have investigated predictors of ICD therapy, but the predictors have not been clarified. In the present study, we demonstrated that anemia increased, whereas long-lasting AF decreased, the likelihood of appropriate ICD therapy in the PP group. To date, many studies have shown that AF is an independent predictor of inappropriate ICD therapy. Conversely, the effect of AF on the prognosis of patients with ICD/CRT-D is controversial.18,19 In the present study, the rate of severe HF (NYHA Class III/IV) was significantly higher among patients with than without AF in the PP group (68.4% vs. 53.3%, respectively; P=0.039); the proportion of patients with CRT-D was also higher among patients with than without AF (81% vs. 70%, respectively). There was also a tendency for a higher percentage of AF than non-AF patients to be prescribed β-blockers (79% vs. 73%) and Class III antiarrhythmic drugs (33% vs. 30%). These factors are likely to have influenced the results.
One limitation of prophylactic ICD treatment may be patient selection based on LVEF. Recent studies demonstrated that LVEF is less sensitive and specific a predictor of SCD,20 and that an alternative index is required. Late gadolinium enhancement (LGE) detected by cardiac magnetic resonance imaging (CMR) has been used as an index for more accurate risk stratification of VA. It is well known that LGE on CMR in patients with ischemic HF is an indicator of myocardial fibrosis and is useful as a risk assessment of VA. For patients with NIHF, LGE has also been shown to provide useful information for risk assessment of all-cause mortality, cardiovascular events, and VA in numerous studies. The prospective CMR substudy of the randomized DANISH trial demonstrated that myocardial fibrosis detected as LGE on CMR predicts all-cause mortality in patients with NIHF.21 However, in that study, ICD implantation did not affect all-cause mortality for either patients with (HR 1.18; 95% CI 0.59–2.38; P=0.63] or without (HR 1.00; 95% CI 0.39–2.53; P=0.99) LGE.21 In the present study, the underlying non-ischemic heart diseases were diverse because the study patients were selected solely based on LVEF, and we did not analyze each non-ischemic heart disease individually. Furthermore, there was a lack of sufficient data on CMR. The establishment of highly predictive alternatives to LVEF will enable the selection of suitable patients for primary prophylaxis with ICD.
Study Limitations
This study has several limitations. First, because it was a prospective observational study of a multicenter registry, there was a lack of randomization and thus the potential for hidden bias. To minimize the effect of treatment selection bias and confounding effects, we performed propensity score matching with NIHF patients for whom complete baseline variables were available. Although caution should be used when applying our results to other geographic settings, especially patients in Western countries, our cohort reflects the real-world clinical setting of Asian patients with ICD. Second, the model selection and detailed programming of ICD were done at the attending physician’s discretion. Therefore, differences in the setting of tachycardia detection rates and the length of the detection interval affected the incidence of ICD therapy, and it is possible that overtreatment was given to self-terminating VA. In addition, differences in tachycardia discrimination algorithms between manufacturers may have affected the incidence of inappropriate ICD therapy. Third, GDMT for NIHF was clearly suboptimal compared with recent Western trials, which also contributed to the incidence of ICD therapy. Fourth, the underlying heart disease of NIHF was determined at the discretion of each institution and no CMR data were collected in this study. This may have obscured the predictors of primary prophylaxis. Fifth, because the patients enrolled in this study were enrolled between 2010 and 2012, the indication criteria for ICD and CRT-D were based on the Guideline for Non-Pharmacotherapy of Cardiac Arrhythmias (JCS 2011). Therefore, the indication criteria for CRT-D in particular differ slightly from the current guideline,22 in which the presence or absence of left bundle branch block and QRS width are stated as indication criteria for CRT-D. These differences may have affected the incidence of ICD therapies and the prognosis of CRT-D-implanted patients. Finally, in this study, 13 baseline variables were propensity score matched and analyzed. As a result, more than half the patients were excluded, decreasing the number of patients enrolled. Other confounders, such as QRS width and left bundle branch block, may have influenced the incidence of ICD therapy.
Conclusions
In propensity score-matched Japanese NIHF patients, the incidence of appropriate ICD therapy and ES were significantly higher in the SP than PP group. Severe HF and impaired renal function increased, whereas CRT decreased, the likelihood of ICD therapy in the SP group.
Acknowledgments
The authors gratefully acknowledge all 48 Japanese implantable cardiac shock device centers involved in this study. The authors also acknowledge the office of the Japanese Heart Rhythm Society, especially Ms. Yoko Sato, as well as Ms. Shoko Narumi and Mr. Jin Ono for data collection.
Sources of Funding
No financial support was received for this study from any specific company, except the Japan Arrhythmia Device Industry Association. This study was supported, in part, by JSPS KAKENHI Grant no. JP085700004. The authors declare no other relationships with industry and no specific unapproved use of any compound or product.
Disclosures
The authors have no conflicts to disclose.
Authorship
All authors attest they meet the current ICMJE criteria for authorship.
Patient Consent
Each patient completed a written informed consent form.
IRB Information
This study was approved by the Institutional Review Board’s Ethics Committee at Kindai University (Reference no. 10-04) and was conducted according to the principles of the Declaration of Helsinki, as revised in 2013. All patients provided written informed consent.
Data Availability
The deidentified participant data will not be shared.
References
- 1.
Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, et al. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 2002; 346: 877–883.
- 2.
Bardy GH, Lee KL, Mark DB, Poole JE, Packer DL, Boineau R, et al. Amiodarone or an implantable cardioverter-defibrillator for congestive heart failure. N Engl J Med 2005; 352: 225–237.
- 3.
Yokoshiki H, Shimizu A, Mitsuhashi T, Ishibashi K, Kabutoya T, Yoshiga Y, et al. Trends in the use of implantable cardioverter-defibrillator and cardiac resynchronization therapy device in advancing age: Analysis of the Japan Cardiac Device Treatment Registry database. J Arrhythm 2020; 36: 737–745.
- 4.
Kurita T, Noda T, Nitta T, Furushima H, Shimizu A, Ohe T, et al. Nippon Storm Study design. J Arrhythm 2012; 28: 277–279.
- 5.
Noda T, Kurita T, Nitta T, Abe H, Watanabe S, Furushima H, et al. Appropriate duration of driving restrictions after inappropriate therapy from implantable cardiac shock devices-interim analysis of the Nippon Storm Study. Circ J 2014; 78: 1989–1991.
- 6.
Noda T, Kurita T, Nitta T, Chiba Y, Furushima H, Matsumoto N, et al. Significant impact of electrical storm on mortality in patients with structural heart disease and an implantable cardiac defibrillator. Int J Cardiol 2018; 255: 85–91.
- 7.
JCS Joint Working Group. Guidelines for non-pharmacotherapy of cardiac arrhythmias (JCS 2011). Circ J 2013; 77: 249–274.
- 8.
Exner DV, Pinski SL, Wyse DG, Renfroe EG, Follmann D, Gold M, et al. Antiarrhythmics versus implantable defibrillators. Electrical storm presages nonsudden death: The Antiarrhythmics Versus Implantable Defibrillators (AVID) trial. Circulation 2001; 103: 2066–2071.
- 9.
Shen L, Jhund PS, Petrie MC, Claggett BL, Barlera S, Cleland JGF, et al. Declining risk of sudden death in heart failure. N Engl J Med 2017; 377: 41–51.
- 10.
Theuns DA, Smith T, Hunink MG, Bardy GH, Jordaens L. Effectiveness of prophylactic implantation of cardioverter-defibrillators without cardiac resynchronization therapy in patients with ischaemic or non-ischaemic heart disease: A systematic review and meta-analysis. Europace 2010; 12: 1564–1570.
- 11.
Romero J, Díaz JC, Grushko M, Quispe R, Briceno D, Avendano R, et al. Clinical impact of implantable cardioverter-defibrillator in primary prevention of total mortality in non-ischaemic cardiomyopathy: Results from a meta-analysis of prospective randomized clinical trials. Europace 2018; 20: f211–f216.
- 12.
Stavrakis S, Asad Z, Reynolds D. Implantable cardioverter defibrillators for primary prevention of mortality in patients with nonischemic cardiomyopathy: A meta-analysis of randomized controlled trials. J Cardiovasc Electrophysiol 2017; 28: 659–665.
- 13.
Køber L, Thune JJ, Nielsen JC, Haarbo J, Videbæk L, Korup E, et al. Defibrillator implantation in patients with nonischemic systolic heart failure. N Engl J Med 2016; 375: 1221–1230.
- 14.
Elming MB, Nielsen JC, Haarbo J, Videbæk L, Korup E, Signorovitch J, et al. Age and outcomes of primary prevention implantable cardioverter-defibrillators in patients with nonischemic systolic heart failure. Circulation 2017; 136: 1772–1780.
- 15.
Al-Khatib SM, Fonarow GC, Joglar JA, Inoue LYT, Mark DB, Lee KL, et al. Primary prevention implantable cardioverter defibrillators in patients with nonischemic cardiomyopathy: A meta-analysis. JAMA Cardiol 2017; 2: 685–688.
- 16.
Schrage B, Uijl A, Benson L, Westermann D, Ståhlberg M, Stolfo D, et al. Association between use of primary-prevention implantable cardioverter-defibrillators and mortality in patients with heart failure: A prospective propensity score-matched analysis from the Swedish Heart Failure Registry. Circulation 2019; 140: 1530–1539.
- 17.
Greene SJ, Butler J, Albert NM, DeVore AD, Sharma PP, Duffy CI, et al. Medical therapy for heart failure with reduced ejection fraction: The CHAMP-HF Registry. J Am Coll Cardiol 2018; 72: 351–366.
- 18.
van Gelder IC, Phan HM, Wilkoff BL, Brown ML, Rogers T, Peterson BJ. Prognostic significance of atrial arrhythmias in a primary prevention ICD population. Pacing Clin Electrophysiol 2011; 34: 1070–1079.
- 19.
Bunch TJ, Day JD, Olshansky B, Stolen KQ, Mullin CM. Newly detected atrial fibrillation in patients with an implantable cardioverter-defibrillator is a strong risk marker of increased mortality. Heart Rhythm 2009; 6: 2–8.
- 20.
Wellens HJ, Schwartz PJ, Lindemans FW, Buxton AE, Goldberger JJ, Hohnloser SH, et al. Risk stratification for sudden cardiac death: Current status and challenges for the future. Eur Heart J 2014; 35: 1642–1651.
- 21.
Elming MB, Hammer-Hansen S, Voges I, Nyktari E, Raja AA, Svendsen JH, et al. Myocardial fibrosis and the effect of primary prophylactic defibrillator implantation in patients with non-ischemic systolic heart failure: DANISH-MRI. Am Heart J 2020; 221: 165–176.
- 22.
Nogami A, Kurita T, Abe H, Ando K, Ishikawa T, Imai K, et al. JCS/JHRS 2019 guideline on non-pharmacotherapy of cardiac arrhythmias. J Arrhythm 2021; 37: 709–870.
