Neighborhood Disadvantage Impacts Outcomes in Patients with Myelodysplastic Syndromes across Different Racial Backgrounds

INTRODUCTION The impact of racial differences on the outcomes of patients (pts) with myelodysplastic syndromes (MDS) remains unclear. Even though several groups have reported population-based studies in MDS, the results have been mixed (Ma et al. Cancer, 2007; Goksu et al. Blood, 2020; Tinsley-Vance et al. Blood, 2021; Laadem et al. Leukemia Research, 2023). The comparison of socioeconomic (SE) factors, an independent prognostic factor by itself, was limited across all studies. There is no perfect tool to assess all the SE factors. However, the area deprivation index (ADI) includes 17 SE factors and is considered one of the most scientifically validated tools for assessing health equity (Powell et al. Forefront, 2023). We report the outcomes of pts with MDS treated at MDACC stratified by area deprivation index (ADI). METHODS All pts with MDS treated at MDACC from 3/2013 to 3/2023 with self-reported racial background and available ADI data were included. ADI data was downloaded from https://www.neighborhoodatlas.medicine.wisc.edu/September 28, 2023. Propensity score matching (PSM) was used to adjust for covariates between races. The Nearest Neighbor Matching (NNM) algorithm and the Matchit package were used to perform the matching. We subset the dataset to pair two races, such as Non-Hispanic Black (NHB) vs. Non-Hispanic White (NHW), Hispanic (H) vs. NHW, and NHB vs. H. To balance the covariate across race, we used the following covariates: gender, age (binary), IPSS-R risk, prior treatment, ADI State Rank, and ADI National Rank. We adjusted the matching options (including caliper to ensure quality matches) to keep the Standardized Mean Deviation less than 2, a common standard. RESULTS 1540 pts (NHW-1329, H-88, NHB-85, Asian-37, Native American-1) were identified with available racial and ADI data. Asian and Native American pts were excluded due to low numbers. After PSM, 61 NHB and 366 NHW pts were included for comparison. There was no significant difference in the IPSS-R cytogenetics risk groups or IPSS-M mutations regardless of hypomethylating agent (HMA) exposure. Among pts with no prior HMA, treatment on clinical trial was more common among NHW than NHB (246/298, 83% vs. 30/48, 67%, p=0.004). In multivariable analysis (MVA), age ≥60 years (y, HR-1.31, p=0.03), higher ADI National Rank (HR-1.01, p=0.01), high/very high IPSS-R (HR-2.63/5.55, p=0.001/<0.001), and TP53 mutations (HR-2.68, p<0.001) led to shorter overall survival (OS). Interestingly, race did not impact OS, irrespective of prior HMA exposure. Similarly, 73 H and 360 NHW pts were included for comparison after PSM. Among HMA naïve pts, H pts were younger (median, 65 vs. 68 Y, p=0.01), and a relatively lower number of pts were treated on clinical trial (37/55, 67% vs. 231/280, 83%, p=0.02). Mutations in IDH2 were more frequent in H pts (12% vs. 4%, p=0.02). MVA showed higher ADI National Rank (HR-1.01, p<0.001), intermediate/high/very high IPSS-R (HR-3.23/4.77/6.24, p=0.01/0.001/<0.001), TP53 mutation (HR-2.26, p<0.001) led to shorter OS. Allogeneic stem cell transplant was associated with longer OS (HR-0.41, p<0.001). Race did not affect the OS in both the HMA-exposed and HMA naïve groups. CONCLUSION Our data shows that race did not impact the OS in MDS pts after matching the 17 different SE factors using ADI. This is by far the most extensive single-center data that has incorporated the advanced SE tool currently available for measuring outcomes in pts with MDS. There is still a need for a comprehensive tool to estimate the SE status of pts. Focusing on the health equity of pts living in disadvantaged neighborhood (higher ADI) regardless of racial background, would help improve treatment outcomes in pts with MDS.