Current guidelines have indicated an implantable cardioverter-defibrillator (ICD) for patients with severe idiopathic dilated cardiomyopathy, for both primary and secondary prevention. Compared to coronary artery disease, the overall benefit has been smaller. A more refined risk assessment, using the left ventricular ejection fraction (LVEF) and prevention mode (primary/secondary), is still needed to guide ICD implantation. Patients included in 2 large ICD registers were analyzed regarding the appropriate therapies and improvement of LVEF, overall and in subgroups of prevention mode and LVEF <20% versus >20%. Overall, 349 patients were included; 70% were men, the mean age was 54 years, and the mean follow-up was 33 months. Cardiac resynchronization therapy (CRT) was used in 57%, and secondary prevention was present in 30%. ICD therapies were delivered to 33% of the patients, in most for ventricular tachycardia. Patients receiving an ICD for secondary prevention and non-CRT were more likely to have arrhythmic events (both p <0.05). The cumulative event rates at 5 years were 53% for secondary and 33% for primary prevention (p <0.001). Depending on the prevention mode and LVEF status (<20% vs >20%), the event rates ranged from 30% to 76%. The mean LVEF improved by 10%, independently of the stimulation mode (CRT 22% to 31%, non-CRT 26% to 35%; p <0.0001). A persistent improvement to >35% was seen in only 25% of CRT patients but in 45% of non-CRT patients (p = 0.004). In conclusion, the results from the present study have demonstrated that in patients with idiopathic dilated cardiomyopathy, the potential for LVEF improvement is considerable and that the rate of ICD interventions strongly depends on the prevention mode and LVEF. These findings could be the basis for additional risk stratification tools.
The current guidelines for placement of an implantable cardioverter-defibrillator (ICD) recommend it as a class 1 indication for primary prevention in patients with idiopathic dilated cardiomyopathy (IDC) and a left ventricular ejection fraction (LVEF) of ≤35% who are in New York Heart Association class II and III. However, the current guidelines have recommended it only as a class II(b) indication if patients are in New York Heart Association class I. For secondary prevention, a clear class 1(a) indication has been given. However, in a real-life setting, the decision making for patients with nonischemic cardiopathy is more complex, because issues such as the etiology, prevention mode (primary or secondary prevention), additional cardiac resynchronization therapy (CRT), co-morbidities, and the potential to improve LVEF over time also should be considered. The present retrospective study had 2 aims: to define the higher and lower risk groups by combining the LVEF and prevention mode and to determine the degree of improvement of LVEF over time.
Methods
Patients with IDC in whom an ICD had been implanted at the University of Basel Hospital or the Erasmus Medical Center in Rotterdam after January 1, 2000 were included in the present analysis, irrespective of the pacing or prevention mode. Because of insufficient baseline or follow-up data, 59 patients were excluded. The date of the last follow-up examination was September 30, 2008. The baseline characteristics included the ejection fraction, prevention and pacing modes, cardiac rhythm, and drug therapy. The results from a baseline transthoracic echocardiogram were available for all patients. For primary prevention, a cutoff rate for ventricular tachycardia (VT) detection of 175 to 180 beats/min with a series of antitachycardia pacing bursts followed by shocks was usually chosen. For secondary prevention, the cutoff rate was usually 20 beats/min less than the clinically observed VT. The detection for ventricular fibrillation was usually set at 210 beats/min with direct shock application. The patients were exclusively followed at the 2 institutions. Follow-up examinations were performed in Basel at 1, 3, and 6 months after implantation and then every 6 months and in Rotterdam at 10 days and 3, 6, 9, and 12 months after implantation and then every 6 months. Every documented arrhythmia was interpreted by 2 experienced physicians (Basel) or by 1 physician and 1 expert (Rotterdam), and the ICD intervention was then declared either appropriate or inappropriate. Only appropriate interventions were considered. Death was categorized as due to congestive heart failure, sudden cardiac death, or noncardiac related.
The baseline and follow-up data have been presented for all patients; however, because CRT has the potential to improve LVEF and thus might also have an effect on the ICD interventions, a separate analysis was performed stratifying patients by CRT status. Another analysis was done by dividing patients into 4 different groups with a different presumed risk: group 1, primary prevention (no arrhythmias or only nonsustained VT during ambulatory electrocardiography, LVEF >20%); group 2, primary prevention (as defined for group 1 but LVEF <20%); group 3, secondary prevention (patients with syncope suggestive of VT, documented VT, or survivors of sudden death from ventricular fibrillation, and LVEF <20%); and group 4, secondary prevention (as defined for group 3 but LVEF >20%). The same assignment to groups was done using a cutoff LVEF of 25%.
All available reports of echocardiographic examinations performed ≥6 months after implantation were collected. Subanalyses were performed of those patients whose LVEF improved to >35%, because this cutoff value has been used as the indication for primary preventive ICD implantation. The end points in the present study were appropriate ICD therapy (either shock or antitachycardia pacing), all-cause mortality, and improvement of LVEF to >35%.
Continuous data are presented as the mean ± SD. Chi-square tests to compare categorical variables and a 2-sided Student’s t test to compare continuous variables were used, as appropriate. Kaplan-Meier curves were drawn to show the mortality according to the prevention mode and intervention and event-free survival in the 4 patient groups. For both analyses, the patients were censored at the first adequate ICD intervention and/or at death, if they had not had previous ICD interventions. The Statistical Package for Social Sciences, version 15 (SPSS, Chicago, Illinois) was used. A p value of <0.05 was considered significant.
Results
The study cohort consisted of 349 patients (125 from Basel and 224 from Rotterdam). The mean follow-up was 33 months. A CRT-ICD was used in 198 patients (57%), a single-chamber ICD in 97 patients (28%), and a dual-chamber ICD in 54 patients (15%). For the 105 secondary prevention patients (30%), the index arrhythmia was ventricular fibrillation in 28, sustained VT in 62, and syncope associated with poor LVEF in 15. Detailed baseline characteristics are listed in Table 1 .
| Variable | All Patients (n = 349) | CRT Patients (n = 198) | Non-CRT Patients (n = 151) | p Value |
|---|---|---|---|---|
| Men | 245 (70.2%) | 111 (73.3%) | 134 (67.7%) | 0.28 |
| Secondary prevention mode | 105 (30.1%) | 37 (18.7%) | 68 (45.1%) | <0.0001 |
| Sinus rhythm | 309 (88.5%) | 177 (89.4%) | 132 (87.4%) | 0.62 |
| Age (years) | 53.9 ± 14.3 | 56.5 ± 12.4 | 50.5 ± 15.9 | <0.0001 |
| New York Heart Association class | ||||
| I-II | 149 (42.7%) | 56 (28.3%) | 93 (61.6%) | NA |
| III-IV | 200 (57.3%) | 142 (71.7%) | 58 (38.4%) | |
| QRS width (ms) | 147 ± 39 | 167 ± 29 | 120 ± 32 | <0.0001 |
| Left ventricular ejection fraction | 24.1 ± 7.6% | 22.3 ± 6.4% | 26.5 ± 8.3 | <0.0001 |
| β Blocker | 260 (74.5) | 155 (78.3%) | 105 (70.0%) | 0.08 |
| Angiotensin-converting enzyme inhibitor/angiotensin receptor blocker | 328 (94.0%) | 186 (94.0%) | 142 (94.0%) | 1 |
| Diuretics | 272 (77.9%) | 164 (82.8%) | 108 (71.5%) | 0.01 |
| Amiodarone | 90 (25.8%) | 50 (25.3%) | 40 (26.5%) | 0.8 |
| Statins | 92 (26.4%) | 59 (29.8%) | 33 (21.9%) | 0.11 |
| Digoxin | 86 (24.6%) | 56 (28.3%) | 30 (19.9%) | 0.08 |
| Follow-up (mo) | 32.9 ± 24.9 | 29.6 ± 21.2 | 37.2 ± 28.6 | 0.004 |
| Mortality | 43 (12.3%) | 23 (11.6%) | 20 (13.2%) | 0.74 |
| Transplantation | 28 (8.0%) | 13 (6.6%) | 15 (10.0%) | 0.32 |
| Delivered defibrillator therapy | 115 (32.9%) | 56 (28.3%) | 59 (39.1%) | 0.04 |
During follow-up, 115 patients (33%) experienced appropriate ICD interventions. In 44 patients (38%), the first arrhythmia was treated within the ventricular fibrillation zone of the device. Interventions were more often seen in the secondary prevention (49% vs 26%, p = 0.0001) and non-CRT (39% vs 28%, p = 0.04) patients. Both CRT and non-CRT patients experienced more ICD interventions if a secondary prevention indication was present (CRT 51% vs 23%, p = 0.001; non-CRT 47% vs 32%, p = 0.09). The baseline LVEF was not associated with the ICD interventions, although patients with a poorer LVEF (<20%) tended to receive more ICD interventions (38% vs 30%, p = 0.12). Figure 1 shows the actuarial event rates for a first appropriate ICD intervention stratified by LVEF and prevention mode. Group 1 included 156 patients, group 2 included 87, group 3 included 40, and group 4 included 65 patients. With the LVEF cutoff of 25%, no differences were observed among the 4 groups. The cumulative event rate at 5 years was 53% for secondary and 33% for primary prevention and 30% for group 1, 46% for group 2, 53% for group 3, and 76% for group 4.
The ICD interventions were similarly distributed across the 4 New York Heart Association classes (39% in class I, 39% in class II, 28% in class III, and 33% in class IV; p = 0.23). However, comparing less symptomatic patients in class I and II to the more symptomatic patients in class III and IV, a strong trend was seen toward more ICD interventions in the less symptomatic group (58 [39%] of 149 patients and 57 [28%] of 200, respectively, p = 0.052).
No differences were seen between the patients in Basel and Rotterdam regarding the baseline LVEF, use of CRT or non-CRT devices, ICD interventions, or mortality. The patients treated in Basel more often had echocardiographic follow-up data available (74% vs 58%, respectively, p = 0.002), and more of the patients treated in Rotterdam had undergone heart transplantation (1% vs 12%, respectively, p <0.0001).
Follow-up echocardiographic data were available for 221 patients (63% of the whole population). Echocardiography had been performed a mean of 26 ± 20 months after implantation ( Table 2 ). An improvement in LVEF of 5% to 9% was observed in 33 patients (15%), 10% to 19% in 62 patients (28%), and >20% in 40 patients (18%). Overall, “normalization” of the LVEF (i.e., to ≥50%) had occurred in 27 patients (12%).
| Variable | All Patients | CRT Patients | Non-CRT Patients | p Value |
|---|---|---|---|---|
| Follow-up echocardiograms ⁎ | 221 (63.3%) | 132 (66.7%) | 89 (58.9%) | 0.15 |
| Left ventricular ejection fraction | ||||
| At baseline | 23.6 ± 7.6% | 21.6 ± 6.4% | 26.4 ± 8.4% | <0.0001 |
| During follow-up | 32.7 ± 11.1% | 31.0 ± 9.8% | 35.2 ± 12.4% | <0.0001 |
| Improvement | 9.1% | 9.4% | 8.8% | NA |
We also divided the patients into 2 age groups according to the median age of 55 years. A greater proportion of the older patients had received a CRT device (49% vs 65%, p = 0.0025). The prevention mode, ICD interventions, combination of mortality and heart transplantation, mean improvement in LVEF, and percentage of follow-up echocardiograms were similar (all p = NS), as was the percentage of patients improving their LVEF to >35%. However, in patients <55 years old, the LVEF improved to >35% in more non-CRT patients (76% vs 35%, p = 0.001) with the potential of CRT improving the LVEF to >35% less pronounced than in older patients (13% vs 38%, respectively, p = 0.001). Table 3 shows the rate of defibrillator interventions in different patient groups.
| Left Ventricular Ejection Fraction | CRT Patients | Non-CRT Patients | ||
|---|---|---|---|---|
| Primary Prevention (n = 107) | Secondary Prevention (n = 26) | Primary Prevention (n = 50) | Secondary Prevention (n = 39) | |
| Improvement | 22–32% ⁎ | 22–28% † | 25–32% ⁎ | 28–35% ‡ |
| Defibrillator interventions | 26 (24%) ⁎ | 18 (69%) ⁎ | 18 (36%) ‡ | 24 (19%) ‡ |
| Improved to >35% | 29 (27%) | 5 (19%) | 24 (48%) | 16 (41%) |
| Defibrillator interventions | 3 (10%) ‡ | 4 (80%) ‡ | 7 (29%) | 7 (43%) |
| Not improved to >35% | 78 (73%) | 21 (81%) | 26 (52%) | 23 (59%) |
| Defibrillator interventions | 23 (29%) ‡ | 14 (67%) † | 11 (42%) ‡ | 17 (67%) † |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
