Mode of death and factors associated with cardiac and non-cardiac death in the MADIT-RIT trial

Abstract

Purpose: To describe mode of death and identify factors associated with cardiac and non-cardiac death in the MADIT-RIT trial. Methods: The trial randomized 1500 patients with a prophylactic indication for ICD or CRT-D to one of three ICD programming arms: (A) conventional programming (VT zone ≥ 170 bpm); (B) high-rate cut-off (VT zone ≥ 200 bpm); and (C) prolonged 60 second delay before therapy. Multivariate Cox model with best subset regression was used to identify predictors of cardiac- and non-cardiac death. Results: During a mean follow-up of 1.4±0.6 years, 71 (5%) of 1500 patients died, including 40 cardiac (56.3%), 23 non-cardiac (32.4%) and 8 unknown (11.3%). Of the 40 cardiac deaths, 14 died of arrhythmic causes, 21 of causes related to heart failure and 5 of unspecified cardiac causes. Of the 23 non-cardiac deaths, 5 were vascular, 8 were cancerous, 10 were for non-cardiac diseases. Randomization to conventional ICD programming (A) was identified as an independent predictor of cardiac and non-cardiac death compared to arm B, but not when compared to arm C. Predictors of cardiac death included lower diastolic blood pressure, lower ejection fraction, ischemic cardiomyopathy, appropriate shock and inappropriate ATP only (Table). Predictors of non-cardiac death included hypertension, NYHA class III, ICD:CRT-D, increased age and inappropriate shock (Table). View this table: Table 1. Cardiac and non-cardiac mortality Conclusion: Cardiac and non-cardiac mortality was associated with different clinical characteristics and different types of ICD therapy.