Abstract PS13-38: Cardiotoxicity among patients with breast cancer treated with doxorubicin: A real-world database study

Abstract

Abstract Background While novel targeted agents are increasingly used to care for patients with breast cancer, doxorubicin (DOX) continues to play a role in management of patients, particularly those with aggressive disease. Dose-dependent cardiomyopathy is a challenge in its use. Strategies have been proposed to mitigate this, including administration by continuous intravenous (CIV) infusion as an alternative to bolus (BOL) administration. This study used real world data to explore the impact of DOX administration mode on cardiotoxicity, duration of DOX and time to treatment failure (TTF). Methods IBM MarketScan claims were used to identify patients age ≥ 18 who received at least 2 DOX administrations (excluding liposomal DOX) after cancer diagnosis. Patients with history of cardiac events were excluded. Cardiac events based on a range of International Classification of Disease (ICD) codes were compared for BOL versus CIV overall, by tumor site and by regimen during three follow-up periods, early (within 1 year), middle (>1 to 5 years) and late (>5 years), from DOX initiation using Fisher’s exact test. Duration of DOX and TTF, defined as time from initiation of DOX to subsequent systemic therapy, hospice or death, were evaluated using Kaplan-Meier method and unadjusted Cox proportional hazards models. Results: A total of 38,924 patients with breast cancer met eligibility criteria (13,186 with confirmed metastatic disease). The most common regimen used was DOX plus cyclophosphamide (n=31,815, 81.7%). Most patients had codes for both modes on the same claim date and could not be definitely assigned to BOL or CIV infusion groups; however, 917 and 5,433 patients had exclusive BOL and CIV codes, respectively. Among patients receiving DOX monotherapy (n=687), 361 and 100 had exclusive BOL and CIV codes, respectively. For patients with exclusive infusion type codes, the mean duration of DOX treatment was not significantly different for BOL vs CIV (58.9 vs 56.2 days, p=0.33 overall; 74.4 vs 74.8 days for monotherapy, p=0.97). Overall, cardiac events for BOL vs CIV were 5.1% vs 4.7% (p=0.55) during the early period, 3.1% vs 5.0%, (p=0.01) during the middle period, and 0.4% vs 1.0% (p=0.10) in the late period. There were no differences in cardiac events for BOL vs CIV among those treated with DOX monotherapy (p=0.90, 0.56 and 0.52 for the early, middle, and late period, respectively). TTF was shorter for BOL vs CIV (262.3 vs 366.0 days, p<0.001). However, when evaluating TTF, there was a significant relationship between cardiotoxicities and longer TTF (hazard ratio, HR=0.85, 95% confidence interval, CI: 0.81-0.88, p<0.001). This relationship was statistically significant for the early, middle and late periods, respectively (all p<0.001). Conclusions: These data suggest that cardiac events may occur at a similar rate for BOL and CIV. This study is limited by the retrospective nature of this study and the ability to determine causality; the use of strict coding rules to correctly assign patients to BOL vs CIV groups maintained scientific integrity and a large sample size but did result in the loss of eligible patients. Future research, including adjusted analyses, are needed to further investigate the relationship between mode of infusion and clinical outcomes. Citation Format: Poorni Manohar, Hannah M Linden, Lisa M Hess, Tomoko Sugihara, Yajun E Zhu, Howard G Muntz, Lee D Cranmer. Cardiotoxicity among patients with breast cancer treated with doxorubicin: A real-world database study [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS13-38.