Deformation Analysis of Myocardial Layers Detects Early Cardiac Dysfunction after Chemotherapy in Bone Marrow Transplantation Patients: A Continuous and Additive Cardiotoxicity Process

Abstract

Chemotherapy-induced cardiotoxicity has not been extensively validated in bone marrow transplantation (BMT) patients. Speckle-tracking echocardiography is a sensitive method for the detection of subclinical cardiac dysfunction. Cardiac function was prospectively assessed in 80 patients (44 men; mean age, 45±11years) after BMT for non-Hodgkin's lymphoma and acute or chronic myeloid leukemia by means of various echocardiographic techniques. Before chemotherapy for BMT, 89% of the patients had previously been treated with anthracyclines. Patients had normal left ventricular ejection fraction (LVEF). Left ventricular (LV) global longitudinal strain (GLS), subendocardial and subepicardial longitudinal strain, circumferential strain, LV twist, and right ventricular GLS were measured by speckle-tracking, and (2) three-dimensionally derived LVEF and right ventricular ejection fraction were also assessed. Abnormal LVEF was defined as <53%. Studies were performed before (baseline) and 1, 3, 6, and 12months after chemotherapy conditioning followed by BMT. Impaired LV GLS values were observed at 1month after chemotherapy and at 3, 6, and 12months compared with baseline (-20 ±2.2% at baseline, -18.4 ±2.1% at 1month, -17.3 ±2.2% at 3months, -17.1 ±2.1% at 6months, and -17.1 ±2.2% at 12months; P=.001). Early LV GLS changes were driven mostly by changes in subendocardial longitudinal strain (-22.5 ±2.4% at baseline, -20.5 ±2.3% at 1month, -19.2 ±2.3% at 3months, -19.2 ±2.4% at 6months, and -19.1±2.4 at 12months; P=.001), whereas significant subepicardial strain changes were observed at 3months after BMT. Compared with baseline, right ventricular GLS was also impaired early after chemotherapy. Compared with baseline, LVEF was slightly reduced (P=.02) at the end of the follow-up. Among echocardiographic markers, LV GLS at 1month had the strongest predictive value for abnormal LVEF (<53%) at 12months (area under the curve 0.86; 95% CI, 0.76-0.96). A cutoff LV GLS value of -18.4% had sensitivity of 84.6% and specificity of 71.9% for the identification of abnormal LVEF at the end of follow-up. In BMT patients, myocardial deformation analysis detected early and progressive subclinical cardiac dysfunction. Impaired LV GLS had predictive value for the detection of abnormal LVEF at 12-month follow-up. Thus, myocardial deformation study should be applied early after BMT to prevent irreversible cardiac dysfunction by appropriate treatment.