Abstract 16326: Replication of Previously Reported Anthracycline Cardiotoxicity Risk Variants in a Large Electronic Health Record Based Cohort
Introduction: Anthracyclines have well described cardiotoxic effects that limit their clinical use. Prior studies have identified putative risk alleles for anthracycline cardiotoxicity (ACT), however these studies vary widely in definitions of cardiotoxicity and number, age, and characteristics of included subjects, making it difficult to interpret their relevance for patient care. Moreover, many observed associations have not been replicated.
Hypothesis: The purpose of this study was to identify risk variants with the most consistent association with ACT. We hypothesized that testing published variants for association with a clinically accepted definition of ACT in a large clinical cohort would identify variants most robustly associated with ACT.
Methods: We performed a comprehensive literature search to identify variants associated with ACT. Genotype and clinical data were available for 404 individuals from Vanderbilt University’s BioVU biobank with a history of anthracycline exposure and measures of pre- and post- chemotherapy left ventricular ejection fraction (LVEF). ACT was defined as either a reduction in LVEF >10% to a nadir of <50% or a reduction in LVEF >20%. Early- and late-onset ACT were defined as within or after 1 year of chemotherapy completion, respectively.
Results: A total of 58 putative risk variants were identified from 9 separate reports, 7 of which were from pediatric cohorts. The study cohort was 59% women, 90% white, and mean age was 53±16 years. There were 58 cases of ACT of which 36 were late-onset. After adjustment for age, gender, race, and anthracycline dose, four variants were associated with late-onset ACT, including rs8187710 (MDR2), rs1695 (GSTP1), rs1800562 (HFE), and rs1883112 (NCF4). A single variant, rs4148808 in ABCB4, was associated with early-onset ACT after similar adjustments. Replicated variants tended to arise from studies with ACT definitions that included specific LVEF and time course components.
Conclusions: We replicated associations between several published variants and a widely accepted definition of ACT in a large clinical cohort. These findings support the biological importance of these variants for ACT and prioritize them for future study of strategies for pre-exposure risk stratification.
Author Disclosures: E.P. Held: None. E. Farber-Eger: None. M. Blair: None. Q.S. Wells: None.
- © 2016 by American Heart Association, Inc.