ICDs in Chagas heart disease: the standard treatment for secondary prevention of sudden death

We appreciate the critical reading of our article by Anis Rassi, Jr and Anis Rassi,1 but we disagree with them. Secondary prevention studies have shown striking survival benefit with implantable cardioverter-defibrillator (ICD), irrespective of the aetiology of the cardiac disease,2–6 with 50% reduction in sudden cardiac death and 28% reduction in mortality from all causes (absolute reduction of 3.5% per year). This benefit, observed in randomized clinical trials (RCTs), was confirmed and extended in observational studies, including populations different from original trials.6,7 Our results suggest secondary prevention with ICD can be more useful in patients with Chagas disease (ChD) than other cardiopathies, due to the high arrhythmogenic profile of the former.8 Indeed, we found that ChD increased by 2.2 times the chance of patients receiving an ICD appropriate therapy and that ChD was the only independent predictor of combined outcome death or appropriate therapy in this sample. Moreover, we have previously demonstrated that Chagas cardiomyopathy was associated with poorer survival compared with idiopathic disease,9 further suggesting that the benefit of ICD implantation could be higher in ChD than in other cardiopathies. About the three specific comments by Rassi Jr and Rassi, two of them are directly related to the composition of our study population. We studied consecutive patients referred to university hospital mainly from a state-wide emergency care system, most of them resuscitated from cardiac arrest. Moreover, we followed the current guidelines for implantation of ICD developed by Brazilian Health Ministry, which restricts the indication for well-selected cases for secondary prevention. However, the cohort profile of our study sample was not significantly different from that included in both RCTs and observational studies for secondary prevention. The reported all-cause mortality rate in our study (12.3%, median follow-up: 8.7 months) is slightly higher than that observed in the AVID trial10 (10.7% in 1 year), but lower than the mortality rate of patients documented in other observational studies in Latin America.11–13 The mean left ventricular ejection fraction (LVEF) of our sample (37%) was also quite similar to the value reported from the meta-analysis of the three major clinical trials that demonstrated the benefit of ICD for secondary prevention (34%).6 Although, as claimed by Rassi Jr and Rassi, this meta-analyses suggested that the benefit of the secondary prevention was restricted to patients with lower LVEF, it must be recalled that this conclusion arose from a post hoc subgroup analysis of these studies, thus considerably reducing the strength of this statement. Finally, the suggestion that a ‘RCT comparing ICD with amiodarone is urgently needed’ for secondary prevention of sudden death in ChD cardiomyopathy is not supported by current evidence.14 Studies conducted in patients with dilated cardiomyopathy showed no reduction in overall mortality with the use of amiodarone, in spite of a modest reduction in the incidence of sudden and cardiovascular death, but revealed a two- and five-fold increase risk of pulmonary and thyroid toxicity.15 The lack of effect on global mortality by amiodarone was evident in both ischaemic and non-ischaemic cardiomyopathies.15 Thus, an RCT comparing ICD, an effective treatment that reduces overall mortality in a wide range of situations of high risk of sudden death, with amiodarone, which had not proved to reduce global mortality, could lead to unjustified loss of lives, is dubious and, probably, unethical. Conflict of interest: none declared.