Iron cardiomyopathy is a leading cause of death in transfusion-dependent thalassemia major (TM) patients and MRI (T2*) can recognize preclinical cardiac iron overload, but, is unavailable to many centers. We evaluated the ability of 12-lead electrocardiography to predict cardiac iron loading in TM. 12-lead electrocardiogram and cardiac T2* measurements were performed prospectively, with a detectable cardiac iron cutoff of T2*less than 20 ms. Patients with and without cardiac iron were compared using two-sample statistics and against population norms using age and gender-matched Z-scores. 45/78 patients had detectable cardiac iron. Patients having cardiac iron were older and more likely female but had comparable liver iron burdens and serum ferritin. Increased heart rate (HR) and prolonged corrected QT interval (QT(c)) were present, regardless of cardiac iron status. Repolarization abnormalities were the strongest predictors of cardiac iron, including QT/QT(c) prolongation, left shift of T-wave axis, and interpretation of ST/T-wave morphology. Recursive partitioning of the data for females using T-axis and HR and for males using QT, HR, and T-axis produced algorithms with AUROC's of 88.3 and 87.1, respectively. Bradycardia and repolarization abnormalities on 12-lead electrocardiography were the most specific markers for cardiac iron in thalassemia major. Changes in these variables may be helpful to stratify cardiac risk when cardiac MRI is unavailable. However, diagnostic algorithms need to be vetted on larger and more diverse patient populations and longitudinal studies are necessary to determine reversibility of the observed abnormalities.
Copyright © 2011 Wiley-Liss, Inc.