Background: Overall five year survival following pancreaticoduodenectomy for ductal adenocarcinoma is poor with typical reported rates in the literature of 8-27%. The aim of this study was to identify the histological variables best able to predict long-term survival in these patients.
Methods: A prospective database of patients undergoing pancreaticoduodenectomy between April 2002 and June 2009 was analysed to identify patients with histologically proven pancreatic ductal adenocarcinoma. Patients with ampullary tumours, cholangiocarcinoma, duodenal adenocarcinoma and neuroendocrine tumours were excluded. The histology reports for these patients were reviewed. Uni-variate and multi-variate survival analysis was performed to identify variables useful in predicting long-term outcome.
Results: 134 patients underwent pancreaticoduodenectomy for pancreatic ductal adenocarcinoma during this period. 5 year survival in this series was 18.6%. Uni-variate analysis identified nodal status and the metastatic to resected lymph node ratio as predictors of survival. Using multi-variate Cox Regression analysis a metastatic to lymph node ratio of >15% (p < 0.01) and the presence of perineural invasion (p < 0.05) were identified as independent predictors of patient survival. Metastatic to resected lymph node ratio is better able to stratify prognosis than nodal status alone with 5 year survival of those with N0 disease being 55.6% and 12.9% for N1 disease. However for those with <15% of resected nodes positive, 5 year survival was 21.7% and in those with >15% nodes positive it was 5.2% (p = 0.0017).
Conclusion: The metastatic to resected lymph node ratio can provide significant prognostic information in those patients with node positive disease after pancreaticoduodenectomy for pancreatic ductal adenocarcinoma.
© 2011 Elsevier Ltd. All rights reserved.