Background: Several randomized controlled trials (RCTs) have demonstrated that dexrazoxane reduces anthracycline cardiotoxicity in adults, but use in children has been hindered by lack of direct evidence of cardioprotection and concerns regarding second malignant neoplasms (SMNs). This study aimed to systematically review the evidence regarding dexrazoxane in children.
Methods: We searched Medline, Embase, the Cochrane Library, and abstracts for RCTs and nonrandomized studies (NRSs) that compared dexrazoxane to no cardioprotection among children. We combined findings using random-effects models. All statistical tests were two-sided.
Results: Eleven eligible publications reported results from five RCTs (1254 patients), and 15 publications reported results from 12 NRSs (3385 patients). Dexrazoxane did not impact clinical cardiotoxicity in RCTs because of a low cardiotoxic event rate (three events among all patients) but was associated with a reduction in subclinical cardiotoxicity. Among NRSs, dexrazoxane was associated with a reduction in clinical cardiotoxicity (relative risk (RR) = 0.29, P = .001) and clinical+subclinical cardiotoxicity (RR = 0.43, P < .001). Among RCTs, 17 of 635 (2.7%) patients treated with dexrazoxane developed an SMN compared with seven of 619 (1.1%) who did not receive dexrazoxane (RR = 2.37, P = .06). Two RCTs that used concurrent etoposide reported an increased risk of acute myeloid leukemia, while one that used cranial radiation reported an increased risk of brain tumors. Event-free survival did not differ (P = .91).
Conclusion: Dexrazoxane is associated with a statistically significant risk reduction for most cardiotoxic outcomes. Dexrazoxane is associated with a statistically borderline increase in SMNs, possibly because of an interaction with concurrent cancer therapies. The decision to use dexrazoxane in children should balance the risks of cardiotoxicity and SMNs specific to each treatment protocol.
© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: [email protected].