Impact of disadvantaged socioeconomic and demographic status on survival in Kaposi sarcoma.

Abstract

e21581 Background: Kaposi sarcoma (KS) is a soft tissue tumor that affects a wide variety of patient populations, though it is typically seen in immunocompromised patients, including those living with organ transplant or HIV/AIDS. Patients from traditionally underserved socioeconomic and demographic populations have historically had poor outcomes across malignancies, however these populations have never been studied on a large national scale in KS. To examine how socioeconomic and demographic status affects survival in KS, we examined the National Cancer Database (NCDB); a large oncology database that includes approximately 75% of newly diagnosed cancers across the United States. Methods: We conducted a retrospective analysis of the NCDB for KS patients diagnosed from 2004 to 2019. Cases missing data for variables of interest were excluded. Overall survival (OS) was examined using Kaplan-Meier and multivariate Cox regression. Results: A total of 5,233 patients with KS were identified, with a mean age of 58 years. Most patients were male (87.1%), White (70.1%), not Spanish/Hispanic (82.7%), and had government insurance (54.3%). Comorbid conditions were described using the Charlson-Deyo Comorbidity Score (CDS), and a substantial number of patients had a score above 3 (28.6%). The mean OS for all patients was 122 months. On adjusted Cox regression, decreased OS was seen in patients over 85 years old (HR 2.14, 95% CI 1.82-2.53, p<0.001) compared to patients aged 40-54 years old. Compared to White patients, Black patients had decreased OS (HR 1.24, 95% CI 1.11-1.38, p<0.001); mean survival in Black patients was 113 months compared to 124 months in White patients (p<0.001). Improved OS was seen in patients of Spanish/Hispanic origin (HR 0.80, 95% CI 0.70-0.91, p<0.001). Any CDS greater than 1 was associated with poor OS (HR 1.46 for CDS 1, 95% CI 1.26-1.70, p<0.001), and treatment at an academic facility had a protective effect (HR 0.75, 95% CI 0.69-0.82, p<0.001), as did private insurance (HR 0.65, 95% CI 0.55-0.77, p<0.001). Biological sex, education, income, and facility location did not independently impact survival. Conclusions: When adjusted for key demographic covariates, advanced age, Black race, and presence of comorbidities were each independently associated with decreased OS in KS patients. Improved OS was seen with Spanish/Hispanic ethnicity, treatment at an academic facility, and private insurance. To our knowledge, our study is the first to describe demographic and socioeconomic features impacting survival for KS patients at this scale. We revealed important disparities in survival and highlight the need for equitable care in KS, especially for patients of Black race.