Evaluation of cardiac failure and recovery in acute myelogenous leukemia.

Abstract

7038 Background: Cardiac failure causes significant comorbidity and death in acute myeloid leukemia (AML) patients, with incidence estimated at 10-15% (Neuendorff N Blood Adv. 2020). We sought to describe the post-treatment cardiac experience in AML survivors including factors associated with onset, persistence, and outcome of cardiac dysfunction during and following treatment. Methods: 86 patients with AML or other high-grade myeloid neoplasms treated with induction chemotherapy with or without allogeneic hematopoietic cell transplant were identified at the University of Washington/Fred Hutchinson Cancer Center from 1/2000-1/2022. All patients sustained cardiac dysfunction, defined as left ventricular ejection function (LVEF) decline of 10% with absolute LVEF below 50% during chemotherapy. A multivariable analysis was performed to compare those who recovered (LVEF recovery) and those who did not (LVEF non-recovery), examining baseline health factors, disease risk (ELN 2017), time to dysfunction and recovery, most probable cause of cardiac dysfunction based on cardiologist evaluation, and goal-directed medical therapy (GDMT) use and dosage. Results: 48% (41 out of 86) failed to recover following LVEF decline. This cohort tended to be older, male, and had more ASCVD risk factors ( p < 0.05). Neither the choice of chemotherapy regimen ( p= 0.43) nor the cumulative anthracycline dose ( p= 0.44) was significantly different between the LVEF recovery and non-recovery groups. Onset of cardiac dysfunction tended to occur later in the LVEF recovery cohort (median 72 vs. 48.5 days, p= 0.660). In subjects who recovered, the median time to recovery was 87 days (range 2-268 days). When chemotherapy was identified as the probable cause, it was associated with greater likelihood of eventual recovery ( p= 0.014), whereas ischemic coronary etiology was associated with non-recovery ( p= 0.005). Failing to recover from cardiac dysfunction correlated with shorter survival (median 229 vs 650 days, p = 0.005). Multivariable analysis is shown in the Table below. Importantly, the use of GDMT with higher doses of renin-angiotensin inhibitor medications (RAASi) were associated with higher probability of recovery (not shown, p < 0.05). Conclusions: The present study demonstrates recovery occurring in over half of AML patients who experienced cardiac dysfunction during their chemotherapy. Recovery was more common in patients whose cardiac dysfunction is precipitated by chemotherapy and in patients who received GDMT with aggressive use of RAASi. Since cardiac recovery was associated with improved survival, next steps include establishing a cardiac monitoring algorithm to initiate early and aggressive GDMT. [Table: see text]