Economic Burden of Hospitalizations for Patients with Newly Diagnosed Acute Myeloid Leukemia in Remission in the United States: Retrospective Analysis of an Administrative Claims Database

INTRODUCTION: Managing patients with acute myeloid leukemia (AML) requires extensive healthcare resource utilization and costs; hospitalization is the largest component of medical costs for AML. In the phase 3 QUAZAR AML-001 trial, oral azacitidine (Oral-AZA [CC-486]) was associated with significant improvements in overall survival and relapse-free survival (RFS) vs. placebo (Wei, 2020). Prolonged RFS with Oral-AZA may translate into substantial economic benefits, with lower hospitalization-related costs due to reduced rates of hospitalization and days in hospital (Oliva, ASH 2020). Real-world economic benefits of prolonged remission remain insufficiently studied.

OBJECTIVE: This study aimed to examine the economic burden of hospitalizations among patients with newly diagnosed AML in remission from a retrospective analysis of real-world data from a US claims database.

METHODS: Using a retrospective cohort design, adult patients were selected from the IBM MarketScan ^TM Commercial and Medicare Supplemental Databases, with ≥2 outpatient claims or 1 inpatient claim with a primary International Classification of Disease 9th/10th Revision (ICD-9/ICD-10) code for AML between July 1, 2012, and September 30, 2019. Patients were required to have 1) at least 6 months of continuous enrollment, with pharmacy benefits, prior to diagnosis, 2) received systemic induction therapy as first-line (1L) therapy for AML on or after the index diagnosis date, and 3) attained remission from 1L systemic therapy.

Patients were followed from remission after systemic induction therapy, with or without consolidation, until the end of the follow-up period. Eligible patients were organized into cohorts based on their duration of remission (DOR): Cohort A included patients with < median DOR and Cohort B had patients with ≥ median DOR. Hospitalization incidence and duration during the follow-up period were calculated using a per-patient per-year (PPPY) metric. Hospitalization-related costs were compared using a generalized linear model (GLM) with gamma distribution and log-link function. Generalized estimating equations (GEE) clustered on the patient were used to compare incidence and duration of hospitalizations. All costs were adjusted for inflation and reported in 2019 US dollars (USD).

RESULTS: In all, 693 patients met all selection criteria and were assessed for DOR. The median DOR for all patients was 167 days; cohorts A and B included 346 and 347 patients, respectively, with median times from remission to end of follow-up of 106.5 and 460 days. Baseline characteristics were comparable between cohorts; mean [SD] age was 55.1 [14.4] years, 49.4% of patients were male, and mean [SD] Charlson comorbidity index [CCI] was 0.8 [1.2]. The PPPY number of hospitalizations was higher in Cohort A than in Cohort B (4.56 vs 2.09, respectively), as was the PPPY total length of hospital stay (51.1 vs 19.7 days). The PPPY hospitalization-related costs were $345,728 in Cohort A and $125,018 in Cohort B; the cumulative hospitalization cost per patient was $148,430 higher in Cohort A at 12 months and $177,500 higher at 18 months (Figure). Multivariate GEE and GLM models with adjustment for patient characteristic covariates (eg, age, sex, CCI) identified prolonged remission was significantly associated with reduced number, duration, and cost of hospitalization (P < 0.001, all comparisons).

CONCLUSIONS: In a real-world setting, prolonged AML remission was associated with significantly lower rates and durations of hospitalization, which were estimated to result in substantial cumulative cost savings. These data underscore the importance of active maintenance therapies that delay relapse for patients with AML. While the costs of long-term therapy face increasing scrutiny, these costs should be weighed relative to the potential economic benefit of prolonged remission and reduced burden of healthcare resource utilization.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal