P775: CLINICAL OUTCOMES AND SOCIOECONOMIC FACTORS IN CHRONIC MYELOMONOCYTIC LEUKEMIA IN THE UNITED STATES USING THE NATIONAL CANCER DATABASE: 2004-2017

Abstract

Background: Chronic myelomonocytic leukemia (CMML) is a rare myeloid malignancy characterized by proliferation of dysplastic monocytic cells and carries a risk for leukemic transformation. Previously, we described the incidence and outcomes of CMML at the population level (2004-2015) using the Surveillance, Epidemiology, and End Results Program (SEER) and National Cancer Database (NCDB) (Ruan et al., ASH 2020). We studied population-based outcomes in CMML using the NCDB from 2004-2017. Aims: Our aim was to understand the impact of social determinants of health on outcomes in CMML. Secondly, we studied if the outcomes of CMML have improved. Methods: The NCDB is a nationwide oncology outcomes database for >1500 cancer programs in the US and Puerto Rico, capturing 70% of all newly diagnosed cases of cancer. We queried the NCDB using the ICD-O-3 diagnosis code 9945/3 from 2004-2017. Only patients (pts) with diagnostic confirmation via histopathologic examination were included. Percent without high school degree (HSD) reflects the education level within the pt’s zip code, which was stratified into 4 quartiles. Median household income was estimated by matching the pt’s zip code at time of diagnosis against US Census 2000 data, which was stratified into 4 quartiles. Hazard ratios (HR) with confidence intervals (CI) were calculated using a Cox proportional hazards model. Overall survival (OS) was estimated using the Kaplan-Meier method. Variables significant in univariable analysis were included in a multivariable analysis. Results: We identified 7143 CMML pts in the NCDB. The median age at diagnosis was 72 years (interquartile range [IQR] 65-81) and 4386 (61%) were males. With a median follow up of 6.1 years (IQR 3.6-9.4), the median OS was 1.4 years (IQR 0.4-3.8). 1600 (23%) had private insurance, 4862 (70%) had Medicare, 298 (4%) had Medicaid, and 129 (2%) were uninsured. In univariate analysis, private insurance had an improved OS compared to patients with Medicaid (HR 1.2, 95% CI 1.06-1.4; p value 0.007), Medicare (HR 1.5, 95% CI 1.4-1.6; p value < 0.001), and other government insurance (HR 1.5, 95% CI 1.1-2.0; p value 0.004). Pts living in a zip code with percent of adults without an HSD of < 14% had an improved OS compared to zip codes with > 29% (HR 1.2, 95% CI 1.1-1.3; p value < 0.001). Median household income by zip code of ≥ $46,000 had an improved OS compared to < $30,000 (HR 1.2, 95% CI 1.1-1.4; p value < 0.001). Receiving treatment at an academic program had an improved OS compared to community cancer programs (HR 1.2, 95% CI 1.1-1.3; p value < 0.001), comprehensive community cancer programs (HR 1.3, 95% CI 1.3-1.4; p value < 0.001), and integrated network cancer programs (HR 1.3, 95% CI 1.2-1.4; p value < 0.001). Finally, pts diagnosed in 2012-2017 had an improved OS compared to pts diagnosed in 2004-2011 (HR 0.8, 95% CI 0.8-0.9; p value < 0.001). Variables significant in univariate analysis were included in multivariate analysis (Table 1). When education level and median household income per zip code were included in multivariate analysis, education level remained statistically significant and income did not. Image:Summary/Conclusion: CMML continues to have a poor prognosis, but OS has improved in recent years. Socioeconomic factors including insurance status, level of education and median household income per zip code, and facility type appear to be associated with outcomes. On multivariable analysis, age < 65, private insurance (vs government insurance), living in a zip code with increased percent HSD, and treatment at an academic program had improved outcomes.