Validation and comparison between current prognostication systems for pancreatic neuroendocrine neoplasms: A single-institution experience with 176 patients
This article aims to validate and compare the performance of 6 prognostication systems—the World Health Organization 2010 grading criteria, the European Neuroendocrine Tumour Society and the American Joint Committee for Cancer staging systems, the Memorial Sloan Kettering Cancer Center staging and grading systems, as well as the Bilimoria criteria in a cohort of patients with pancreatic neuroendocrine neoplasms at a single institution.Methods.
A retrospective review of 176 patients with histologically proven pancreatic neuroendocrine neoplasm was performed. The prognostic ability of the various prognostication systems for pancreatic neuroendocrine neoplasm was assessed by analyzing the homogeneity, discriminatory ability, monotonicity of gradient, and Akaike information criteria.Results.
The 5-year overall survival for the 176 patients was 69% and 5-year recurrence-free survival in 119 patients who underwent curative resection was 78%. Comparison between the 6 prognostication systems demonstrated that the World Health Organization 2010 system had the lowest Akaike information criteria score and was hence the best prognostication system in predicting overall survival and recurrence-free survival rates in our cohort of patients. The European Neuroendocrine Tumour Society was superior to the American Joint Committee for Cancer in prognosticating overall survival rates for pancreatic neuroendocrine neoplasms, as there was a statistically significant difference in overall survival across the different stages when stratified by the European Neuroendocrine Tumour Society, while the use of the American Joint Committee for Cancer was limited to distinguishing between patients in stages I and II versus stages III and IV only.Conclusion.
All 6 prognostication systems were useful in the prognostication of pancreatic neuroendocrine neoplasm. The World Health Organization 2010 grading system was the best prognostication system in predicting both overall survival in our entire cohort of patients and recurrence-free survival in the subset of patients who underwent curative resection.