AUTHORS
Parisa Abedi, MD, Luca Lanino, MD2, Rohan K Achar, MD, MSCR3, Henry Dumke, MD4, Razan Mansour, MD5, Benjamin Rolles, MD6, Abdulrahman Alhajahjeh, MD7, Jan Philipp Bewersdorf, MD8, Luis E. Aguirre, MD, MPH9, Nikolai Podoltsev, MD, PhD10, Lourdes Mendez, MD11, Lohith Gowda, MD12, Stuart Seropian, MD13, Alain Mina, MD14, Amer M. Zeidan, MD15, Maximilian Stahl, MD16
AFFILIATIONS
1 Yale School of Medicine, New Haven, Connecticut, United States
2 Yale School of Medicine, New Haven, Connecticut, United States
3 Yale School of Medicine, New Haven, Connecticut, United States
4 Yale School of Medicine, New Haven, Connecticut, United States
5 Yale School of Medicine, New Haven, Connecticut, United States
6 Yale School of Medicine, New Haven, Connecticut, United States
7 Yale School of Medicine, New Haven, Connecticut, United States
8 Yale School of Medicine, New Haven, Connecticut, United States
9 Yale School of Medicine, New Haven, Connecticut, United States
10 Yale School of Medicine, New Haven, Connecticut, United States
11 Yale School of Medicine, New Haven, Connecticut, United States
12 Yale School of Medicine, New Haven, Connecticut, United States
13 Yale School of Medicine, New Haven, Connecticut, United States
14 Yale School of Medicine, New Haven, Connecticut, United States
15 Yale School of Medicine, New Haven, Connecticut, United States
16 Yale School of Medicine, New Haven, Connecticut, United States
ABSTRACT
Background: Allogeneic hematopoietic cell transplantation (HSCT) remains the only curative therapy for many patients (pts) with acute myeloid leukemia (AML), yet access is highly variable. We quantified the impact of socioeconomic status (SES) on HSCT access and assessed how much of this gap is explained by insurance type, practice setting, or treatment timing versus other unmeasured barriers associated with low SES.
Methods: We conducted a retrospective cohort study of pts with newly diagnosed AML using the Flatiron Health Research Database (2014–2024), stratified by first-line therapy (intensive chemotherapy [IC] vs. venetoclax plus hypomethylating agents [HV]). Area-level SES was assigned using Flatiron Health's ACS-derived Yost Index (2015–2019 vintage, census block group), coded as a quintile (Q1 = lowest SES) and dichotomized as Low (Q1–Q2) vs High (Q3–Q5; reference) for primary comparisons. HSCT was ascertained from structured treatment records. Analyses included logistic regression and time-to-event models (cause-specific Cox; Fine-Gray with death as a competing event). We adjusted for potential confounders including age, sex, ethnicity (race), ECOG performance status, diagnosis year, molecular risk per ELN 2022 risk classification for subgroup and mediation analyses. We performed mediation decomposition using a difference-method approach with 1,000 bootstrap resamples to derive 95% percentile confidence intervals. Modifiable mediators included insurance type, practice setting, and time-to-first treatment. Additionally, we evaluated SES × ELN risk, SES × ethnicity and SES × insurance type interactions using likelihood-ratio tests.
Results: Among 5,978 pts, HSCT rates were 39.5% in IC-treated pts and 8.7% in HV-treated pts. In 120-days landmark-adjusted analysis, HSCT was associated with lower mortality (IC: adjusted hazard ratio [aHR] 0.69, 95% CI 0.59–0.82, p<0.001; HV: aHR 0.50, 95% CI 0.30–0.84, p=0.009), with 2-year OS gains of 16.8% in IC (69.2% vs 52.4%) and 27.7% in HV treated pts (56.4% vs 28.7%).
Low- vs high-SES pts had lower HSCT rates in IC (30.5% vs 44.1%, p<0.001) and directionally in HV (7.1% vs 9.4%, p=0.12), with a monotonic gradient across quintiles in (IC: 28.0% Q1 to 45.9% Q5, ptrend<0.001; HV: 6.7% Q1 to 13.0% Q5, ptrend=0.002). SES remained an independent predictor of lower HSCT utilization across both cause-specific Cox and Fine-Gray models. Specifically, pts with low vs. high SES experienced a 35% to 48% relative reduction in adjusted HSCT receipt: In the IC cohort, low SES was associated with a significant reduction in HSCT receipt in both cause-specific Cox (aHR 0.60, 95% CI: 0.53–0.67, p < 0.001) and Fine-Gray models (sun distribution hazard ratio [sHR] 0.59, 95% CI: 0.52–0.67, p < 0.001), with a similar pattern observed in the HV cohort for both Cox (aHR 0.55, 95% CI: 0.37–0.82, p = 0.003) and Fine-Gray analysis (sHR 0.52, 95% CI: 0.33–0.82, p = 0.004).
The SES disparity persisted across all ELN 2022 risk strata (Favorable: aOR 0.56; Intermediate: aOR 0.53; Adverse: aOR 0.38; all p < 0.001), with no significant interaction between SES and ELN risk (p = 0.52). Ethnicity also did not modify the SES effect (p = 0.16 for IC; p = 0.23 for HV). Lower SES independently reduced HSCT access across all racial groups (IC White: aOR 0.48; IC Black: aOR 0.46; HV White: aOR 0.58; HV Black: aOR 0.11 [Firth-penalized]; all p < 0.03). Finally, SES effects were homogeneous across insurance types in IC (p = 0.53) but heterogeneous in HV treated pts (p = 0.015). In the HV cohort, the SES disparity was confined to commercially insured pts (9.3% vs. 12.9%; aOR 0.48, p = 0.009) and absent among Medicare-insured pts (5.3% vs. 5.7%; aOR 0.74, p = 0.49).
Together, potentially modifiable mediators including insurance type, practice setting, and treatment timing explained only a minority of the disparity in HSCT access based on SES (IC: 17.9%, 95% CI: 11.4–26.8%; HV: 25.5%, 95% CI: 4.9–81.2%).
Conclusion: Among adults with AML, lower SES is associated with a 35–48% reduction in HSCT receipt across first-line regimens. This disparity demonstrates a dose-response across deprivation quintiles and consistent across age, molecular risk, and ethnicity, and widens under full confounder adjustment. Because equalizing practice setting, insurance, and treatment timing addresses only a minority of this gap, achieving equity will require interventions beyond traditional insurance and practice policies.
DISCLOSURES
Parisa Abedi, MD: No relevant conflicts of interest to declare.