Prognostic relevance of the novel TNM-classification system for neuroendocrine tumors of foregut origin

Abstract

15147 Background: Until recently no TNM-classification existed for neuroendocrine tumors (NET) of the gastroenteropancreatic system. In 2006 Rindi et al. proposed a new TNM-classification system for foregut-NET of gastric, duodenal and pancreatic origin. The present study validates prognostic significance of the new TNM-classification in a cohort from a German referral center. Methods: The case files of 193 patients with histologically proven foregut NET were analyzed retrospectively. Patients were classified according to Capella (1995) and the WHO (2000) and to the new TNM-classification. Uni- and multivariate analysis was performed using log-rank and Cox regression methods. Results: Primary tumors localization was gastric (n=48), duodenal (23) and pancreatic (122); 35 patients (18%) died during the observation period. Overall 5- and 10-year survival rates (YSR) were 76% and 65% resp., while 5- and 10-YSR according to NET-related deaths (74% of all deaths, 26/35) were 83% and 74% resp.. The classifications by Capella and the WHO significantly discriminated between low and high grade malignant NET (p<0,001 and 0,002 resp.), but did not allow further prognostic differentiation. In contrast, the newly proposed TNM-classification system was able to significantly differentiate stage IV NET from all other stages (I vs. IV p=0,001; II vs. IV p=0,001; III vs. IV p=0,008) as well as stage III from stage I NET (p=0,038). Furthermore, the new classification system which includes a grading system according to the mitotic index (Ki67-index; G1 <3%, G2 3–20%, G3 >20%) was also able to significantly distinguish all three groups from each other (G1 vs. G3 p<0,001; G2 vs. G3 p<0,001). On multivariate analysis stage I and II as well as stage III-NET had a decreased hazard ratio (HR) when compared to stage IV-NET (I/II: HR 0,094, p=0,003; III: HR 0,38, p=0,096). In the same model G1 and G2-NET were associated with an decreased HR as compared to G3-NET (G1: HR 0,05, p=0,01; G2: HR 0,20, p=0,002). Conclusions: Here we demonstrate prognostic relevance of the newly proposed TNM-classification system for foregut-NET with statistical significance which therefore provides a new and powerful tool for future classification of NET. No significant financial relationships to disclose.