Improving survival prognostication of gastroenteropancreatic neuroendocrine neoplasms: Revised staging criteria

Abstract Purpose Current staging criteria for gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs), while useful, have limitations. In this study, we used a population-based registry to evaluate the prognostic utility of the current staging systems and assess whether evidence-based modifications can improve survival predictions. Methods We identified patients with confirmed GEP-NENs from the Surveillance, Epidemiology and End Results registry. We assigned tumour-node-metastasis status according to American Joint Committee on Cancer and European Neuroendocrine Tumor Society criteria. We derived a revised staging classification using Kaplan–Meier methods and Cox regression to assess disease-specific survival and compared the accuracy of potential models based on the Akaike Information Criterion (AIC) and Harrell's C-index. The revised classification was validated in an independent set. Results We identified 10,268 patients with GEP-NENs. We found that multiple stages, as determined by current criteria, misclassified patients' prognosis. In particular, stage IIIB (T1-4N1) had overlapping survival with stage IIIA (T4N0). A revised system which reclassifies N1 disease into different stages based on T status (T1-2N1, T3N1 and T4N1) had an improved AIC (difference = 38) and C-index (0.86) compared to current staging. These revisions improved predictions in patients with both low and high-grade tumours from all primary sites. Results also were confirmed across all primary sites in the validation set. Conclusion Current staging guidelines misclassify the prognosis of N1 patients. Our results suggest that a revised system could lead to better prognostication for GEP-NEN patients. Further validation followed by implementation of these revisions may improve treatment selection and design of clinical trials.