Average treatment effect of facility hepatopancreatobiliary cancer volume on survival of non-resected pancreatic adenocarcinoma

Abstract

To examine the average treatment effect of hepato-pancreato-biliary (HPB) cancer volume on survival outcomes of patients with non-resected pancreatic adenocarcinoma (PDAC). We queried the National Cancer Database (2004-2018) for patients with HPB malignancies (PDAC, pancreatic neuroendocrine neoplasms, hepatocellular carcinoma, biliary tract cancers). We determined the 25th, 50th, and 75th percentiles based on the total annual HPB volume. We then identified patients with non-resected PDAC. We utilized inverse probability (IP)-weighted Cox regression to estimate the effect of facility volume on overall survival (OS). We identified 710,988 patients with HPB malignancies. The 25th, 50th, and 75th percentiles of total annual HPB volume were 32, 71, and 177 cases/year, respectively. We included a total of 196,150 patients with non-resected PDAC. Patients treated at ≥25th, ≥50th, and ≥75th percentile facilities had improved median OS compared to those treated at facilities below these thresholds (5.8 vs. 4.2months, 6.5 vs. 4.5months, 7.5 vs. 4.8months, respectively; p<0.001 for all). Treatment at facilities ≥25th, ≥50th, and ≥75th percentile resulted in lower hazards of death than treatment at lower-percentile facilities (HR: 0.87, 95% CI: 0.84-0.90; HR: 0.87, 95% CI: 0.83-0.91; HR: 0.85, 95% CI: 0.79-0.91, respectively). Our data suggest that consolidation of care of patients with PDAC to high-volume centers may be beneficial even in the nonoperative setting.