Area deprivation index does not influence survival in patients with lower-risk myelodysplastic syndromes at a major cancer center

Abstract

Abstract Introduction: The Area Deprivation Index (ADI) is a well-validated tool for assessing health equity (Powell et al., Forefront, 2023). ADI measures neighborhood disadvantage using 17 indicators of poverty, education, housing, and employment. It is an independent prognostic factor in solid tumors (e.g., breast, ovarian, prostate) and acute leukemias. However, its impact on myelodysplastic syndromes (MDS) is unclear. We evaluated outcomes in patients with lower-risk (LR) MDS, defined by an IPSS-R score ≤3.5, treated at MD Anderson Cancer Center (MDACC). Methods: In this retrospective analysis, previously untreated adult patients with LR-MDS managed at MDACC from 2000 to 2023 were included. Patient and disease characteristics and addresses at the index visit to the MDACC leukemia clinic were abstracted from electronic medical records. Patient zip codes were linked to the 2020 ADI national percentile ranks from the Neighborhood Atlas. ADI percentiles were quartiled (Q) with Q1 (0-25) representing the least disadvantaged neighborhoods and Q4 (76-100) comprised of neighborhoods with the greatest deprivation; state ranks followed the same structure, with Q1 representing the least disadvantaged neighborhoods. Kruskal-Wallis and Fisher's Exact Test were used to compare covariates across ADI quartiles. Stem cell transplantation (SCT) was included as a time-dependent covariate. Backward elimination was applied in multivariate analysis (MVA), retaining variables with p ≤ 0.1. Results: Of 1,481 patients identified, 1,447 patient zip codes were able to be linked to ADI national and state percentile ranks. Patients were stratified by ADI national quartiles: Q1, 232 (16%); Q2, 441 (31%); Q3, 535 (37%); Q4, 239 (17%). When stratified by ADI state quartiles: Q1, 389 (27%); Q2, 593 (41%); Q3, 329 (23%); Q4, 136 (9%) he median age of the cohort was 69 years (19–93). No significant differences were observed in median age, therapy-related MDS (t-MDS), absolute neutrophil count, hemoglobin, platelet count, bone marrow blasts, complex cytogenetics, TP53 mutations (monoallelic or biallelic), type of first-line treatment, clinical trial enrollment for first-line treatment, SCT, or transformation to acute myeloid leukemia across ADI national quartiles. In MVA, only the first-line treatment in clinical trials was associated with longer overall survival (OS; HR 0.73, 95% CI 0.57–0.94, p = 0.014). Older age (HR 1.02, 95% CI 1.00–1.03, p = 0.01), t-MDS (HR 1.78, 95% CI 1.36–2.31, p <0.001), intermediate risk by IPSS-R (HR 2.58, 95% CI 1.67–4.00, p <0.001), mutations in CBL (HR 2.49, 95% CI 1.40–4.42, p = 0.002), DNMT3A (HR 1.42, 95% CI 1.03–2.00, p = 0.032), PTPN11 (HR 4.77, 95% CI 1.28–17.85, p = 0.020), SRSF2 (HR 1.42, 95% CI 1.00–2.00, p = 0.045), STAG2 (HR 1.71, 95% CI 1.07–2.74, p = 0.024), monoallelic TP53 (HR 2.3, 95% CI 1.60–3.31, p <0.001), and biallelic TP53 (HR 2.60, 95% CI 1.41–4.80, p = 0.002) were associated with increased risk of death. Neither the ADI state nor the national quartiles affected OS in univariate or MVA. Median OS for LR-MDS patients, stratified by ADI national quartiles, was: Q1, 54.3 months (95% CI 45–69.8); Q2, 56.1 months (95% CI 51.3–65.3); Q3, 52.6 months (95% CI 48.1–62.3); Q4, 44.2 months (95% CI 35.9–53.7); p = 0.67. Conclusion: Our findings indicate that the ADI does not influence overall survival in patients with LR-MDS treated at a major cancer center, likely due to access to clinical trials, which independently predicted improved survival. Future studies in other healthcare settings would be valuable in understanding if this lack of association between neighborhood deprivation and survival in LR-MDS is representative of the patient population more broadly.