Unité de rythmologie, Institut de cardiologie, groupe hospitalier Pitié-Salpêtrière, 47 boulevard de l’Hôpital, 75013 Paris
Although non-ischemic dilated cardiomyopathies (NIDCM) account for a minority of patients enrolled in implantable-cardiovertor-defibrillator (ICD) trials, the relative benefit yield by such devices seems to be the same whatever the etiology of the cardiomyopathy is. This risk reduction is estimated between 20 and 35%, in primary or secondary prevention, compared with an optimal pharmacological therapy, sometimes associated with antiarrhythmic drugs. However, ventricular arrhythmias and overall mortality are less frequent in patients with NIDCM, leading to a less significant absolute benefit of ICD. In primary prevention, patient’s selection is only based on NYHA classification and ejection fraction because others criteria failed to stratify the risk of sudden death. Among patients in NYHA class II, the impact on prognosis brought by the ICD seems more important than in patients in class III, who are essentially exposed to death by terminal heart failure. In the same way, prophylactic ICD implantation becomes only cost-effective as a long-term therapy especially with extended longevity devices, except for patients with an end stage heart failure. Results of recent ICD trials have led to extensive indications, according to national and international guidelines published in 2006, for any patient in NYHA class II or III with an ejection fraction lower than 35%. Nevertheless, ICD implantation must be in every case a decision requiring the patient’s agreement, based on rhythmic and hemodynamic prognosis and including the risk of adverse effects brought by the ICD.