Introduction: Mortality disparities exist in gastrointestinal (GI) cancers among racial/ethnic groups. One potential contributor to this disparity is a gap in who receives surgery. We aim to examine how much of the mortality disparity among racial/ethnic groups for GI cancers is explained by differences in operative rates.

Methods: The National Cancer Database was used to obtain data from patients diagnosed with gastric, pancreatic, and colorectal cancer in 2004-2015. Descriptive statistics were used to compare raw differences for variables among races. Variables included demographics, receipt of surgery, tumor stage and characteristics, and hospital factors. The racial disparity in survival was measured as the hazard ratio (HR) for each minority compared to White patients, controlling for age and year in Cox regression. The contributions of the variables, including surgical resection, to the racial disparities were estimated by measuring how inclusion of each of these variables affected the HRs of minorities compared to White patients. The magnitudes of the contributions to the HRs were estimated using two methods: (1) the addition of each variable to the age- and year-controlled model, and (2) the serial removal of each variable from a multivariate model that included all variables. The main analysis was performed excluding patients with unknown stage or disseminated cancer.

Results: 1.47 million patients with GI cancer were included in the study: 52% colon, 11% gastric, 21% pancreatic, and 16% rectal. Black patients were more likely to be from lower-income areas, from urban areas, and had lower operative rates in all cancers except gastric cancer. On Cox regression of stage 1-3 disease controlling for age and year of diagnosis, the HRs for Black patients compared to White patients were 1.01 (95% CI 0.97- 1.03), 1.11 (1.09-1.13), 1.22 (1.19-1.24), and 1.28 (1.24-1.32) for stomach, pancreas, colon, and rectum tumors, respectively. Based on the multivariate regression, the factors with the greatest influence on the survival disparity were zip income quartile and receipt of surgery. Receipt of surgery independently accounted for 29%, 11%, and 19% of the survival disadvantage observed in Black compared to White patients for pancreas, colon, and rectum cancer. Zip income quartile accounted for 16%, 18%, and 17%, excluding interactions. In contrast, no gap in operative rates or overall survival was observed for stomach cancer. The significance of surgery to outcomes was most pronounced for pancreatic cancer, where adding receipt of surgery to the age- and year-controlled model reduced the HR from 1.11 to 0.99; for colon and rectal cancer the HR changed from 1.22 to 1.15 and from 1.28 to 1.17.

Conclusion: Part of the observed cancer disparities for Black patients may be due to fewer surgeries being performed for Black patients. Correcting the disparities on the receipt of surgery for stage I through III GI cancer would likely have a large impact on mortality disparities.

Citation Format: John Bliton, Michael Parides, John McAuliffe, Peter Muscarella, Haejin In. Racial disparities in receipt of cancer surgery contribute to worse outcomes for patients with gastrointestinal cancers [abstract]. In: Proceedings of the Eleventh AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2018 Nov 2-5; New Orleans, LA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl):Abstract nr A104.