Effect of centralization on long-term survival after resection of pancreatic ductal adenocarcinoma

Centralization of pancreatic surgery has resulted in improved short-term outcomes in a number of healthcare systems. The aim of this study was to see whether hospital volume influenced long-term prognosis, use of adjuvant therapy or histopathological evaluation of patients undergoing surgical resection for pancreatic ductal adenocarcinoma (PDAC).

Patients undergoing surgical resection of PDAC in Finland between 2002 and 2008 were identified from national registers. Demographic, histopathological, operative and oncological data were recorded, and the histopathological slides of patients who survived for more than 4 years were reviewed. Operative volume was defined according to the annual rate of pancreatoduodenectomy as: high-volume centres (HVCs; 20 or more resections per year), medium-volume centres (MVCs; 6-19 resection annually) and low-volume centres (LVCs; 5 or fewer resections annually).

Some 467 patients who had undergone resectional surgery for PDAC at 22 centres were included. Patient demographics and resection types did not differ between centres. Thirty- and 90-day mortality rates were significantly lower in HVCs compared with LVCs: 0 versus 5·5 per cent (P = 0·001) and 2·5 versus 11·0 per cent (P = 0·003) respectively. Tumours in HVCs were generally at a more advanced stage than those in LVCs (stage IIB: 65·7 versus 40·6 per cent respectively; P < 0·001), but with no greater use of adjuvant therapy. Significantly more patients survived for 2 years (43·3 versus 29·7 per cent; P = 0·034) and 3 years (25·4 versus 14·1 per cent; P = 0·045) after surgery in HVCs than in LVCs. More information was missing in the histopathological reports from LVCs and MVCs than in those from HVCs (P ≤ 0·002).

Both short- and long-term survival was significantly better for patients operated on in HVCs. Histopathological analysis appears to be more comprehensive in HVCs.