Treatment-associated healthcare utilization in older adults with AML: A retrospective single-center study

Abstract

Abstract Background: Treating AML in older adults (OA) presents challenges, including elevated disease risk, increased therapy toxicities, and discrepant patient preferences. As therapies evolve, understanding the healthcare utilization (HCU), namely duration of index hospitalization, rehospitalization rates, emergency department (ED) usage, intensive care unit (ICU) admissions, and hospice enrollment, in the current treatment landscape is relevant for counseling patients and healthcare systems. We aim to evaluate HCU patterns in OA treated for AML and compare HCU patterns between hypomethylating agent-based therapy (HMA) and induction chemotherapy (IC), which may improve clinical decision-making and patient expectations as well as serve as a baseline for improvement. Methods: We conducted a retrospective, single-center analysis of AML patients aged 65 and older who received AML-directed treatment at UMass Memorial Medical Center between 2016 and 2023, excluding those treated only with hydroxyurea. AML classification followed 2016 WHO criteria due to the era of patient inclusion. HCU data were extracted from electronic health records (EHRs). Descriptive statistics were used for patient characteristics. Wilcoxon rank-sum test was used to compare medians between patients who received HMA and IC. Kaplan-Meier and log-rank test were used to evaluate and compare survival outcomes. Results: A total of 85 patients were included with median age 75 (range 65-92; 76 HMA; 72 IC) and median Charlson comorbidity index (CCMI) of 6 (range 4-13; 7 HMA; 5 IC). Most patients were male (61%) and non-Hispanic White (93%). The most frequent WHO 2016 subgroups were myelodysplasia-related changes (38%), not otherwise specified (28%), NPM1-mutated (18%), and therapy-related (9%). On cytogenetic and molecular analysis 26% had complex karyotype, 22% had TP53 mutation, and 13% had del(17p). TP53 aberrations were more frequent in the HMA group (28%) compared to the IC group (5%). For initial treatment, 35 patients (41%) received HMA and venetoclax (median duration 16 days; range 1-32), 30 patients (35%) received HMA monotherapy; 20 patients (24%) received induction chemotherapy (IC), the majority of which consisted of mitoxantrone and cytarabine with or without FLT3 inhibitor, and 6 patients (7%) received HMA with a FLT3 inhibitor. The minority of the cohort (14%) received their first cycle outpatient, and those treated inpatient had a median length of stay (LOS) from treatment initiation of 18 days (range 2-70), with 18% experiencing a 30-day readmission. For those hospitalized, index LOS was 12 days in the HMA group and 30 days in IC (p<0.01). Incidence of febrile neutropenia and documented infection during C1 was 37% and 43%, respectively, in HMA and 84% and 63% in IC. ICU stays were similar in HMA group (15%; median duration 4 days) and IC group (16%; median duration 3 days). Nine patients (14%) in the HMA group and 2 patients (11%) in the IC group died during cycle 1 (C1). Median overall survival (OS) in the HMA group was 3.9 months versus 19.3 months in the IC group (p=0.02). Twelve patients (14%) underwent allogeneic stem cell transplant (6 HMA; 6 IC). In those who survived C1, the median number of rehospitalizations was 3.3 per year (3.7 in HMA; 3.0 in IC; p=0.81). The median number of days spent in the hospital per year was 33 (35 in HMA; 26 in IC; p=0.70). The median number of ED visits was 2.5 (2.5 in HMA; 2.2 in IC; p=0.68). Subsequent admissions to the ICU occurred in 44% of patients, and 10% had >1 ICU admission. Of 75 patients who died, 40% were enrolled on hospice at the time of death with median duration of 5 days (range 1-37). Conclusion: We describe HCU in a contemporary OA AML cohort with HMA and venetoclax as the most frequent treatment received. IC was associated with longer survival, likely reflecting differences in patient characteristics (lower CCMI in IC; higher disease risk in HMA). Index hospitalization was shorter in the HMA-group. Subsequent healthcare utilization was similar between HMA and IC groups, with ~1 month of patients' limited survival spent hospitalized. Hospice remains underutilized. As survival improves for OA with AML, minimizing HCU from this described baseline is a key secondary outcome for both our healthcare system and, most importantly, our patients.