Understanding Improved Survival Among Black Individuals with Myelodysplastic Syndromes

Introduction:

Recent data have demonstrated worse survival among White patients with myelodysplastic syndromes (MDS) compared to Black patients with the same diagnoses. Whether these differences in survival are due to disease-related factors or comorbidities is unknown. To better understand the relationship between race and survival in MDS, we performed an observational study comparing the demographics, disease severity, and rate of antecedent comorbidity in individuals with MDS from different racial groups.

Methods:

Using the Surveillance, Epidemiology, and End Results registry linked to Medicare claims data (SEER-Medicare), we identified a cohort of adults age ≥ 65 years diagnosed with MDS between January 1, 2002 and December 31, 2019. Patients with MDS diagnosed before age 65 years, with HMO coverage, or without consistent Medicare coverage during months -1 to -13 were excluded from analysis. MDS disease severity was estimated using the previously defined SEER- Medicare myelodysplastic risk score (SMMRS) and pre-existing comorbidities were identified using ICD-9 and ICD-10 codes. Charlson Comorbidity Index (CCI) scores were determined for each study participant. A distinct cohort of non-cancer controls was utilized to better understand usage trends for MDS-diagnostic tools such as bone marrow biopsies. Descriptive statistics were used to compare patient characteristics and bone marrow biopsy rate among racial groups. A Cox proportional-hazards model was conducted to analyze racial differences in overall survival after adjusting for sex, SMMR, and CCI.

Results:

Our cohort included 47,147 individuals with MDS. Of these, 41,537 (88%) were White, 2,578 (5%) were Black, 1,146 (2.5%) were Asian, and 1,886 (4%) were other or unknown. While overall male predominance was noted in MDS (56% male and 44% female), female predominance was observed among Black individuals with MDS (46.4% male and 53.6% female). Sex based distribution for White, Asian and Other race groups matched the overall population. Additionally, Black individuals with MDS had a higher proportion of pre-existing comorbidity with 51.1% of the population having a CCI ≥ 3 compared to only 39% and 44% of White and Asian individuals with MDS, respectively. Interestingly, while Black and White individuals with MDS had similar distribution of SMMR scores, overall survival was better for the Black population [Adjusted Cox model: hazard ratio = 0.85 (0.81, 0.89), p<0.001]. This survival advantage was largely driven by Black females of whom 23.9% were alive at the end of the observation period compared to only 17.7% of Black men, 18.6% of White women, and 15.8% of White men. Lastly, we determined the rate of pre-diagnostic bone marrow biopsies among individuals with MDS and in a distinct cohort of 60,679 non-cancer controls. A smaller proportion of Black MDS patients had pre-diagnostic bone marrow biopsies compared to White MDS patients (75.6% vs 81.1%, p < 0.001). Similarly the percentage of non-cancer controls with cytopenia who had bone marrow biopsies was lower for Black individuals compared to White individuals (13.6% vs 27.8%, p <0.001).

Conclusion:

Black individuals with MDS in the SEER-Medicare database have an improved survival compared to individuals of other races despite higher prevalence of pre-existing co-morbidities and similar MDS disease severity. Improved survival appears to be driven by better survival among Black females who make up a majority of the Black MDS population. Importantly, while SEER-Medicare data are often utilized as a representative US cancer population, a widespread reduced rate of utilization of bone marrow biopsy for Black Medicare recipients points to a potential under-diagnosis of MDS in this population. Analysis of clinical data from two institutions is ongoing to confirm these findings and determine if there are molecular correlates of race-specific survival trends.