Balancing the oncologic effectiveness versus the cardiotoxicity of anthracycline chemotherapy in childhood cancer

One of the most complex issues in cancer treatment is the unavoidable conflict between administering cytotoxic agents with variable tumor selectivity and the resulting dose-dependent short- and long-term damage to normal tissues. Further, there is great uncertainty as to whether late outcomes from prior treatment protocols are relevant to the anticipated late outcomes from current protocols. Virtually all recipients of anthracycline therapy should be considered to have some degree of cardiotoxicity. However, the severity of cardiotoxicity, not its presence, should determine what actions are appropriate. Currently, changes in ejection fraction and other imaging or serologic biomarkers (singly or in combination) during therapy have weak predictive value for chronic cardiomyopathy after the end of therapy, and their clinical utility requires further verification. Cardiotoxicity justifying individual dose modification during therapy requires evidence that it improves survival. The low prior probability of congestive heart failure during anthracycline therapy with the current monitoring protocols means that the ejection fraction has an unacceptably low predictive value. As a result, continued reliance on published recommendations for withholding chemotherapy based on asymptomatic changes in ejection fraction increases the risk of treatment failure more than it decreases the likelihood of irreversible cardiac injury. However, abnormalities in ventricular size and function after the end of therapy do predict chronic, progressive cardiomyopathy and justify longitudinal monitoring. Here, we discuss the cardiotoxicity of some of these chemotherapeutic agents and provide a framework for deciding when the evidence of cardiotoxicity is strong enough to justify a change in management.