Systematic review and meta-analysis of definitions and reported incidences of anthracycline cardiotoxicity.

Abstract

e23081 Background: Anthracycline-induced cardiotoxicity (AIC) is a well-known phenomenon, but the true incidence is poorly defined, and definitions of AIC in the literature are variable. We performed a systematic review and meta-analysis to elucidate AIC definitions and the incidence of clinical heart failure (HF) and subclinical AIC reported in adult breast cancer and lymphoma patients after anthracycline (AC) therapy. Methods: 3,428 abstract and titles were identified via PubMed search. Inclusion criteria for full text review included minimum of 100 adult patients receiving ACs; median/mean follow up of at least 12 months; and specified evaluation of left ventricular ejection fraction. Outcomes were divided into clinical HF and subclinical AIC. Meta-analysis was performed with MedCalc software with random effects model. Results: 366 full text articles were reviewed. 24 studies, including a total of 11,065 patients, reported clinical HF findings; 21 studies, including a total of 8,098 patients, reported subclinical AIC findings. Meta-analysis showed an overall incidence of clinical HF of 3.1% (range 0 to 29%, CI 1.9 to 4.6%, I2= 93.9%); an incidence of 2.0% (CI 0.9 to 3.6%, I2 = 91.6%) in breast cancer patients receiving AC; and 4.8% in lymphoma patients receiving AC (CI 2.1 to 8.6%, I2 = 93.6%). Subclinical AIC was seen in 13.8% overall patients (range 2 to 45%, CI 10.4 to 17.7%, I2= 95.3%); 10.3% of breast cancer patients (CI 7.0 to 14.3%, I2= 95.0%); and 19.8% of lymphoma patients (CI 12.3 to 28.6%, I2= 94.4%). Incidence of HF correlated with increasing age (r = 0.51, p = 0.06) and cumulative dose (r = 0.51 , p = 0.19). There was notable heterogeneity of AIC definitions between studies (Table). Conclusions: Cardiotoxicity is not uncommon among patients who receive AC, particularly at doses given for the treatment of lymphoma. However, there was considerable heterogeneity in HF incidences. More uniform criteria and precise estimates are needed to inform clinical practice and trials. [Table: see text]