Biomarkers to predict anthracycline-induced cardiotoxicity in adult hematologic malignancies: A systematic review.

Abstract

e24012 Background: Anthracyclines are a mainstay of treatment in hematologic malignancies, particularly in lymphomas and acute leukemias. However, they can also cause dose- and duration-dependent cancer therapy-related cardiac dysfunction (CTRCD), including left ventricular dysfunction. Guidelines regarding the use of cardiac biomarkers in monitoring for CTRCD are equivocal, with the European Society of Cardiology 2022 recommending routine monitoring of troponin and natriuretic peptides, while the American Society of Clinical Oncology 2017 guidelines only provide a moderate recommendation to obtain serum biomarkers. Given that cumulative anthracycline dosages tend to be higher in the management of hematologic malignancies, this systematic review seeks to characterize the relationship between cardiac biomarkers and the detection of anthracycline-induced cardiotoxicity (AIC) in hematologic cancer patients. Methods: A comprehensive search was conducted on PubMed, SCOPUS, and Dissertations and Theses. The search encompassed a broad range of terms related to hematologic malignancies, anthracyclines, and biomarkers. Inclusion criteria included articles that obtained biomarker levels and documented cardiotoxic events. Excluded articles included data from non-hematologic malignancies in which the data from hematologic cancers could not be separated. Results: 1764 articles were screened, including 78 full-text screens. 21 articles were included in the review. The biomarkers investigated were troponin-I in 9 (42.9%) articles, pro-BNP in 8 (38.1%), troponin-T in 7 (33.3%), BNP in 6 (28.6%), ANP in 3 (14.3%), and pro-ANP in 1 (4.8%). A correlation between cardiotoxicity and biomarkers was identified in 12 (57.1%) of the studies. 6 (28.6%) studies utilized multivariate analyses, and all 6 found biomarkers to be predictive of CTRCD when part of a multivariate analysis or included in an algorithm. 3 of these 6 studies incorporated echocardiography. Conclusions: As outlined in this review, changes in biomarker concentrations have been associated with cardiotoxic events, but the utility of individual markers as predictors of CTRCD appears to be limited. Investigators have yet to identify a consistent threshold or monitoring strategy to aid clinicians in risk management for anthracycline therapy. However, our review provides evidence that incorporating multiple biomarkers into multivariate algorithms may hold more promise than the use of any one marker alone. The authors recommend further validation of biomarker-based multivariate prediction models, especially for resource-limited settings in which laboratory monitoring may be more accessible than imaging.