Impact of social determinants of health on access to molecular targeted therapy in AML and MDS: A real-world analysis

Abstract Introduction Molecularly targeted therapies (MTT) have redefined the treatment paradigm for acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS). While molecular diagnostics and targeted agents are broadly available, their adoption in clinical practice remains variable. In other cancers, social determinants of health (SDOH) have been shown to influence access to precision care, yet their impact in myeloid malignancies is largely unknown. We hypothesized that individual-level SDOH may impact whether patients with molecular targetable mutations (MTM) receive corresponding MTT. To evaluate this, we conducted a real-world evidence (RWE) analysis to assess associations between SDOH and the receipt of MTT in patients with AML or MDS with MTM. Methods We analyzed de-identified patient-level data from the Symphony claims database (Jul 2018–Dec 2024), linked with genomic testing results from a national genomic provider (Dec 2018–Feb 2024) and with a consumer dataset to determine the social determinants of health (SDOH). The study population included individuals aged >45 years, diagnosed with AML or MDS and with MTM, with SDOH data, and with a 6 month clinical history at baseline. Two cohorts were defined: (1) patients with MTM who received MTT (target group) and (2) patients with MTM who only received non-MTT (comparator group). MTM was defined by the presence of IDH1, IDH2, or FLT3 mutations among AML patients and IDH1, IDH2, or SF3B1 mutations among MDS patients. MTT was defined as receiving ivosidenib (IDH1), enasidenib (IDH2), or midostaurin/gilteritinib (FLT3) for both cohorts. We evaluated eleven SDOH domains: unemployment, financial strain, food insecurity, housing instability, interpersonal safety concerns, low educational attainment, mental health disorders, physical inactivity, poor access to care, substance use, and transportation barriers. Associations between individual SDOH domains and receipt of MTT were assessed using chi-square tests, with effect size estimated by Cramér's V. Results SDOH data were available for 36% (n=6,553) of AML and 40% (n=7,074) of MDS patients. AML Cohort Of the 1,534 AML patients with MTM and with SDOH data, 1,026 patients (67%) received MTT, 265 (17%) received non-MTT and 243 (16%) did not receive any treatment. Most were aged ≥65 (target: 80.2%; comparator: 90.2%), of Western European ancestry (target: 68.9%; comparator: 73.6%), and male (target: 51.7%; comparator: 54%). Ethnicity and gender did not differ significantly (p>0.05). SDOH profiles were similar, except financial strain (target: 16.2%; comparator: 28.3%; p<0.01; Cramér's V=0.122). No differences in employment, housing, food security, transportation, or education (all p>0.05). Among the patients who received MTT from oncologists, 68% were treated at NCI or academic centers and 29% of those did so in the same Zip3 location as their treatment location. MDS Cohort Of the 2,574 MDS patients with MTM and with SDOH data, 1,863 patients (72.4%) received MTT, 265 (10.3%) received non-MTT and 446(17.3%) did not receive any treatment Most were aged ≥65 (target: >95%; comparator: >95%). Western European ancestry was most common (target: 69.9%; comparator: 75.1%), followed by Hispanic (target: 2.8%; comparator: 5.5%) and African American (target: 3.2%; comparator: 4.7%). No significant group differences in ethnicity (p=0.08) or gender (target: 52.3% male; comparator: 59.2% male; p=0.3). Financial strain (target: 20.5%; comparator: 22.7%; p=0.7) and other SDOH domains were similar. Only 44% of patients receiving MTT from oncologists were treated at NCI/academic centers and 38% of those did so in the same Zip3 location as their treatment location. Conclusions In this cohort of patients who had MTM, individual-level SDOH, including financial strain, education, housing, and transportation, did not consistently predict access to MTT. The modest association between financial strain and therapy receipt in AML warrants further investigation but was not observed in MDS. These findings suggest that system-level barriers, such as institutional protocols, clinician practices, and payer dynamics, may play a more substantial role in driving disparities in precision oncology. Future research should prioritize identifying and addressing these structural impediments to promote broad and consistent access to molecular testing and targeted therapies for individuals with myeloid malignancies.