Background: Cardiac resynchronization therapy (CRT) reduces morbidity and mortality, but 1/3 of patients (pts) doesn't respond and the rate of adverse events (AE) is high.
Aim: To determine the predictive value of autonomous nervous system assessment with 123I-metaiodobenzylguanidine (MIBG) cardiac scintigraphy regarding AE and its relationship to CRT response.
Methods: Prospective cohort study of heart failure (HF) pts with reduced ejection fraction submitted to CRT. NYHA class, MIBG scintigraphy, cardiopulmonary exercise testing and echocardiography were performed before (M1) and 3–6 months after CRT (M2). Predictors of AE (all-cause mortality, hospitalization for HF or arrhythmia and cardiac transplantation) 2 years after M2 were determined using multivariate regression analysis. Responders had ≥1 of the following: NYHA decrease ≥1, left ventricle ejection fraction (LVEF) absolute increase >5, LVEF relative increase >15%, peak O2 consumption (VO2) increase >10% and left ventricle end-diastolic or end-systolic volume decrease >15%.
Results: Of the 121 pts analyzed (70±14 years, 69.2% men, 29.1% ischemic, 73.1% baseline NYHA III-IV, baseline LVEF 27±11), 35.5% had ≥1 AE (time to first AE 7.1 months). Overall 70 AE occurred (27 deaths, 15 cardiac deaths, 30 hospitalizations for HF, 12 hospitalizations for arrhythmia and 1 cardiac transplantation) with 18.2% (22) of pts having >1 AE. Pts with AE had lower late heart-to-mediastinum ratio (lHMR) at M1 (1.37±0.25 vs. 1.41 0.24 p 0.043) and M2 (1.28±0.18 vs. 1.46±0.20 p 0.001) and lower early heart-to-mediastinum ratio (eHMR) at M2 (1.39±0.20 vs. 1.51±0.20 p 0.001). These parameters predicted AE in the univariate analysis (OR 0.083 95% CI 0.008–0.870 p 0.038 for lHMR at M1, OR 0.001 95% CI 0.0001–0.109 p 0.004 for lHMR at M2 and OR <0.001 95% CI 0.0001–0.071 p 0.004 for eHMR at M2). In the multivariate model against commonly known predictors of outcome (NYHA class, brain natriuretic peptide (BNP), LVEF and peak VO2), lHMR at M1 (OR 0.036 95% CI 0.002–0.797 p 0.036) and eHMR at M2 (OR <0.001 95% CI 0.0001–0.006 p 0.004) remained independent predictors of AE. Hospitalization for HF was the only component of the composite endpoint independently predicted by MIBG scintigraphy (OR 0.004 95% CI 0.0001–0.790 p 0.041 for lHMR at M2). eHMR at M2 was related to response by LVEF criteria (OR 205.784 95% CI 1.662–25485–822 p 0.030). eHMR at M1 LVEF (OR 15.260 95% CI 1.058–220.164 p 0.045). In responders by LVEF, peak VO2 and LVESV, MIBG scintigraphy parameters remained independent predictors of AE (OR 0.002 92% CI 0.0001–0.301 p 0.015 for lHMR at M1, OR 0.002 95% CI 0.0001–0.301 p 0.015 for lHMR at M2, OR 1.061 95% CI 1.005–1.119 p 0.031 for washout at M1).
Conclusion: MIBG scintigraphy predicted AE with added value to NYHA class, BNP, peak VO2 and LVEF. It was capable of identifying pts at increased risk of AE in spite of response to CRT and could be used to recognize pts in need of new therapies.