Background: Previous studies have found racial disparity in pancreatectomies for resectable pancreatic adenocarcinoma. The aim of this study was to investigate if racial disparities were worse in the performance of pancreaticoduodenectomy for borderline resectable pancreatic adenocarcinoma.
Methods: This study used the National Cancer Database (2004-2016) and included patients with non-metastatic and head of the pancreas borderline resectable pancreatic adenocarcinoma. Multivariable, Poisson regression models with robust standard errors evaluated the relative risk (RR) of undergoing a pancreaticoduodenectomy among non-White patients (Black, Asian, and non-White Hispanic) compared with White patients. A Poisson regression model with hospital fixed effects was performed to evaluate if findings were due to within-hospital or between-hospital variation. Interaction between race and neoadjuvant therapy was also evaluated.
Results: There were 15,482 patients (median age 68 years, interquartile range 60-76 years; 48.6% male) with borderline resectable pancreatic adenocarcinoma who were predominantly White (84.3%, n = 13,058; non-White, 15.7%, n = 2424). Overall, 18.4% (n = 2853) had a pancreatic resection. Non-White patients had a significantly lower likelihood of undergoing a pancreatic resection for borderline resectable pancreatic adenocarcinoma when compared with White patients (RR 0.75, 95% confidence interval 0.68-0.83; p < 0.001). These findings persisted in the hospital fixed-effects model. In the interaction analysis, there were no significant differences in the likelihood of pancreatic resection if patients received neoadjuvant therapy.
Conclusions: Non-White patients were 25% less likely to undergo a pancreatic resection for borderline resectable pancreatic adenocarcinoma compared with White patients. This racial disparity was due to variation in care within-hospitals and disappeared if non-White patients were treated with neoadjuvant therapy.