Impact of hospital volume and type on survival in hepatocellular carcinoma: Results from the National Cancer Database.

Abstract

417 Background: To assess the impact of hospital volume and type on survival in patients with hepatocellular carcinoma (HCC). Methods: Patients with histopathological or imaging-based diagnosis of HCC were identified from the 2003-2015 National Cancer Database (NCDB). First-line treatment was stratified as liver transplant, surgical resection, interventional oncology (IO) and chemotherapy. Hospital volume was stratified as high (ranking among top 10% in case numbers) and low volume, separately for each treatment modality. Hospital type was categorized as academic and non-academic. Overall survival was assessed using multivariable Cox proportional hazards models. Results: A total of 63,877 patients were included (transplant n = 10,596, surgical resection n = 11,132, IO n = 12,286, chemotherapy n = 29,863). Of 1,261 hospitals systems which treated HCC, 226 (17.9%) were academic centers and 1,035 (82.1 %) were non-academic centers. Mean number of cases treated annually was higher in academic centers (55.2; 34.6; 40.7; 79.9) versus non-academic centers (10.7; 6.25; 6.6; 11.9 for transplant; surgical resection; IO and chemotherapy; p < 0.001, respectively). Young African American patients and those with private insurance, high income and education were more likely to receive treatment at academic centers. Geographical difference were evident among US regions, with highest proportion of HCC treated at academic centers in New England states (83.6%) and lowest in South Atlantic states (48.6%). Overall survival was superior for academic versus non-academic centers (HR = 0.89, 95% CI: 0.87-0.91, p < 0.001) and high versus low volume centers (HR = 0.79, 95% CI: 0.77-0.81, p < 0.001), after multivariable adjustment for potential confounders. These effects were evident among all HCC treatment modalities. Conclusions: HCC treatment in academic centers shows distinct patterns according to patient demographics and US geography. Among all treatment modalities, both academic setting and hospital volume independently affected HCC outcomes, with highest patient survival observed in high-volume academic centers.