Sudden cardiac death (SCD) affects about half of patients who died of Chagas disease followed at a referral center, and may victim up to 20% of patients without any evidence of heart disease [1]. In the vast majority of cases, SCD is caused by either sustained ventricular tachycardia (VT) degenerating into ventricular fibrillation (VF) or direct VF [1]. In patients with non-Chagas disease dilated cardiomyopathy and left ventricular ejection fraction b35%, Implantable CardioverterDefibrillator (ICD) delivered-therapy has been recommended by international guidelines to avert SCD [2]. Conversely, the impact of ICD therapy in patients with Chagas cardiomyopathy and a left ventricular ejection fraction b35% is unknown. Accordingly, we report herein our initial experience with ICD-delivered therapy in patients with severe Chagas cardiomyopathy. From January, 2006 to December, 2009, a total of 32 patients routinely followed at the Cardiomyopathy Outpatient Service of our Institution received ICD therapy for primary prevention of SCD. The study was approved by the Local Ethical Committee (protocol 3626/ 2010). Inclusion criteria were a left ventricular ejection fraction b35% at radionuclide ventriculography, receiving standard therapy for chronic systolic heart failure, and to have a reasonable expectation of 1-year survival after device implantation [2]. All patients were on New York Heart Association class II at the time of ICD implantation and none had syncope before device implantation. Nineteen (59%) patients had a positive serology for Chagas disease and were considered to have Chagas cardiomyopathy, whereas the remaining 13 (41%) patients had non-Chagas disease dilated cardiomyopathy. Device therapy consisted of antitachycardia pacing (ATP) or shock. ATP was the initial therapy for VT. Shock was the initial therapy when heart rate was higher than 200 beats per minute (VF zone), when ATP accelerated VT heart rate, or when ATP failed to abolish VT. Baseline characteristics of the study population are given in Table 1. It is noteworthy that 87% patients were on Beta-Blocker therapy. Sustained VT was detected in 4 (21%) of 19 patients with Chagas cardiomyopathy and in 2 (15%) patients with non-Chagas cardiomyopathy (pN0.05). VF was observed in 4 Chagas cardiomyopathy (21%) patients and in 2 (15%) non-Chagas cardiomyopathy patients (pN0.05). Overall, 10 (31%) patients were shocked. In total, shock occurrence was necessary to terminate 136 out of 172 (79%) episodes of VT/VF. Inappropriate shock occurred in 4 (12%) patients; 2 (10%) in Chagas and 2 (14%) in non-Chagas cardiomyopathy patients (pN0.05). A total of 101 episodes of VFwere observed in 6 patients. Median of episodes per patient was 6.5 (3, 38). Ninety-four episodes of VF were observed in four Chagas and 7 episodes of VF in two non-Chagas cardiomyopathy patients (p=0.03). Median time to first event was 78 (34, 151) days in Chagas and 173 (71, 593) days in non-Chagas cardiomyopathy patients (pb0.005). Median time to VF was 80 (22, 147) days in Chagas cardiomyopathy patients and 151 (43, 804) days in non-Chagas cardiomyopathy patients (pb0.005). Probability of freedom from shock occurrence was similar in Chagas and non-Chagas cardiomyopathy patients (Fig. 1). Median follow upwas 292 (78, 845) days in Chagas and 654 (158, 987) days in non-Chagas cardiomyopathy patients (pb0.005). Three (9%) patients died during the study period; 2 in Chagas and 1 in non-Chagas patients (pN0.05). The frequency of shock occurrence observed in our Chagas cardiomyopathy cohort in this study was similar to that reported in patients with non-Chagas cardiomyopathy [3]. The lower mortality rate observed in our study in comparison to studies of secondary prevention of SCD [4] in Chagas disease patients may partially be explained by the lower frequency of shock occurrence. Also, the frequency of shock episodes observed in this investigation was lower than that seen in Chagas disease patients receiving ICD therapy for secondary prevention of SCD [4–6]. This can account, at least in part, for the similarity of mortality rate observed in our study in comparison to that seen in non-Chagas cardiomyopathy patients [7]. The higher frequency of VF episodes/per patient observed in Chagas cardiomyopathy patients in this study can also reflect the characteristics of the pathogenesis of Chagas disease [8]. In patients with Chagas cardiomyopathy, foci of confluent fibrosis intermingled with mononuclear cell infiltrate disseminated throughout the myocardium are frequently seen. Moreover, myocardial ischemia