Genetic Predictors of Anthracycline Mediated Cardiomyopathy with Induction Chemotherapy in Acute Myeloid Leukemia (AML)

Abstract

Introduction: Anthracycline induced cardiomyopathy (CMP) is a concern in AML patients (pts) induced with "7+3” (7 days of cytarabine and 3 days of an anthracycline) or liposomal cytarabine/daunorubicin (CPX-351, Vyxeos). However, it is not well understood (i) how common anthracycline CMP with the availability of routine echocardiography for baseline CMP screening and (ii) how often post induction CMP is the reason for not pursuing a potentially curative allogeneic stem cell transplant (allo-SCT). In addition, while it is known that the cumulative anthracycline dose and preexisting left ventricular dysfunction are risk factors for anthracycline mediated CMP, it is not known whether mutations recurrently mutated in myeloid neoplasms predict risk for anthracycline mediated CMP. Methods: AML pts treated at Dana-Farber Cancer Institute (2014-2022) with 7+3 or Vyxeos induction were included if they met the following criteria: (i) pre and post induction chemotherapy echocardiogram to assess the left ventricular ejection fraction (LVEF) and (ii) comprehensive next generation sequencing was obtained prior to treatment. Post induction CMP was defined as LVEF < 50%. We recorded cardiovascular comorbidities including hypertension (HTN), diabetes mellitus (DM), hyperlipidemia (HLD), coronary artery disease (CAD), severe valvular disease and smoking. Continuous variables were summarized as median (range, R), while categorical variables were reported as frequency (percentage). Univariable and multivariable regression analysis was performed to identify predictors of post induction CMP. Results: A total of 422 pts were included; 37 patients (8.8%) developed post induction CMP (Table 1). There were no statistical differences between age, AML subtype, prior cancer therapy and cumulative anthracycline dose (for prior cancer therapy and induction chemotherapy) between pts with and without post induction CMP. Only 3 pts in this cohort (0.7%) had an abnormal LVEF (<50%) pre induction arguing for a highly preselected pt population thought to be fit enough to receive anthracycline containing induction chemotherapy. The post induction LVEF was 43% (R 20-48%) and 60% (R 50-75%) for patients with and without post induction CMP (p<0.001); this corresponded to an absolute decrease in LVEF post induction compared to pre-induction of -18% and -3%, respectively (p<0.001). While 73% of pts without post induction CMP received an allo-SCT, only 46% of pts with post induction CMP received an allo-SCT (p<0.001). However, in only 4/20 pts with post induction CMP who did not receive an allo-SCT (20%) was the emergent CMP the sole reason for not pursuing an allo-SCT. In the other 16/20 pts, other medical reasons were the reason for not pursuing an allo-SCT including sepsis/fungal infection (7 pts), active disease (3 pts), allo-SCT not indicated given favorable genetics (2 pts) and other reasons (4 pts). In univariable and multivariable analysis, age, sex, cardiovascular risk factors and cumulative anthracycline dose did not predict an increased risk for anthracycline mediated CMP (Table 2). While myeloid mutations commonly seen in patients with clonal hematopoiesis (DNMT3A, TET2, ASXL1) did not predict an increased risk for CMP, presence of a JAK2 V617F mutation predicted a significantly increased risk of CMP in univariable (hazard rate [HR] 5.57, 95% CI: 1.14-22.1, p=0.019) and multivariable analysis (HR 5.91, 95% CI: 1.18-24.1, p=0.017). Interestingly, JAK2 V617F transgenic mice have been shown to develop cardiac hypertrophy and collagen fibrosis (Shi et al. Journal of Hematology & Oncology 2014) arguing for an underlying biological effect of JAK2 mutations on cardiac function. Conclusions: In a selected pt population with a normal pre-induction LVEF (99.3% of pts), 8% of pts developed post-induction CMP, however, in only 4/123 pts (3%) this was the sole reason for not receiving an allo-SCT. While cardiovascular risk factors and cumulative anthracycline dose were not predictive of post induction CMP, presence of a JAK2 V617V mutation (but not other myeloid mutations) was associated with an increased risk for anthracycline CMP. This association is intriguing as it could potentially help with patient selection at higher risk for anthracycline CMP (e.g. consideration of anthracycline free induction in JAK2 mutated pts). However, this will require validation in a larger patient cohort. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal