Gender differences in long-term dexrazoxane cardioprotection in doxorubicin-treated children with acute lymphoblastic leukemia

Abstract

10005 Background: Doxorubicin (DOX) causes progressive cardiac dysfunction, particularly in females. Adding dexrazoxane (DZR) to DOX treatment resulted in reduced myocardial injury in children with acute lymphoblastic leukemia (ALL) during Dana-Farber Cancer Institute Protocol 95–01. Methods: We centrally remeasured echocardiograms from childhood high-risk ALL survivors in their first continuous remission who were randomly assigned to treatment with DOX only (n = 66; 30 mg/m2/dose for 10 doses) or DOX plus DZR 30 minutes prior (n = 68; 300 mg/m2/dose). Results: Demographics and median follow-up (DOX 5.3 vs. DZR/DOX 5.5 y) were similar in both arms. Mean left ventricle (LV) end systolic dimension (ESD) z-score was significantly larger than predicted for body-surface area for DOX (mean = 0.46, P-value [deviation of mean from normal] = 0.01) but not so for DZR/DOX (mean = 0.06, P = 0.74); DOX LV fractional shortening (FS; -0.78, P = .001; DZR/DOX = -0.38, P = 0.11) and thickness to dimension ratio (-0.96, P < .001; DZR/DOX = -0.32, P = 0.08) were also abnormal. LV end diastolic posterior wall thickness (EDPWT) was reduced in both groups, though more so for DOX (-1.19, P < .001) than DZR/DOX (-0.74, P < .001). By gender, 5 years post treatment LVESD z-score was significantly larger than normal for DOX males (mean = 0.48, P = 0.04) but not DZR/DOX males (0.19, P = 0.41) or females (DOX female = 0.38, P = 0.22; DZR/DOX female = -0.17, P = 0.56). LVFS z-score was significantly different from normal in DOX females (mean = -1.29, P < .001), but not DZR/DOX females (-0.22, P = 0.54) or males (DOX = -0.45, P = 0.15; DZR/DOX = -0.52, P = 0.09), as was LV thickness to dimension ratio (DOX female = -1.03, P < .001; DZR/DOX female = 0.02, P = 0.93). DZR/DOX females were the only group with normal LVEDPWT z-score (mean = -0.43, P = 0.07; DOX/female = -1.43, P < .001; DOX male = -1.05, P < .001; DZR/DOX male = -0.94, P < .001). Conclusions: While its impact is seen in all groups, primarily females drive the long-term DZR cardioprotective effect. DZR/DOX females exhibit more normal LV dimensions and more appropriate wall thickness for LV dimension, both of which are consistent with less LV remodeling. DZR/DOX females also have more normal LV function than females who received DOX only or males of either group. [Table: see text]