Predicting Risk Versus Predicting Potential Survival Benefit Using 123I-mIBG Imaging in Patients With Systolic Dysfunction Eligible for Implantable Cardiac Defibrillator Implantation: Analysis of Data From the Prospective ADMIRE-HF Study.

Abstract

Cardiac (123)I-metaiodobenzylguanidine ((123)I-mIBG) imaging improves prognostication in patients with left ventricular (LV) dysfunction. Whether (123)I-mIBG can identify optimal candidates for implantable cardiac defibrillator (ICD) placement is unclear. We examined whether (123)I-mIBG enhances risk assessment and identifies patients with enhanced survival with ICD in a patient cohort with reduced LV function who were candidates for ICD implantation. We identified 777 patients (66 sites, 12 countries) without ICD at the time of enrollment in Adreview Myocardial Imaging for Risk Evaluation in Heart Failure (ADMIRE-HF) and index (123)I-mIBG study. Patients completed prescribed study protocol and follow-up. Heart-to-mediastinum (H/M) ratio was determined from (123)I-mIBG results. Survival modeling used a Cox proportional hazards mixed-effects model, including a propensity score, to adjust for nonrandomized ICD implantation after (123)I-mIBG. All-cause death occurred in 75 patients (9.6%), and 196 (25%) patients had ICD implantation on follow-up. After adjusting for multiple factors, although the H/M ratio added incremental prognostic value and enhanced reclassification, neither H/M results, BNP levels, nor left ventricular ejection fraction interacted with ICD use in the survival model, indicating that these variables did not identify patients with enhanced survival with ICD implantation. Nonetheless, H/M results did identify the number of lives saved by ICD use per 100 treated. We found that although (123)I-mIBG imaging enhances the risk stratification of patients with left ventricular dysfunction who are ICD candidates, it does not identify which patients may have improved survival with ICD placement. However, (123)I-mIBG identifies the absolute benefit gained with ICD use, thus may play a role in optimizing the cost-effectiveness of this intervention. URL: http://www.clinicaltrials.gov. Unique identifiers: NCT00126425 and NCT00126438.