Novel recurrence risk stratification of resected pancreatic neuroendocrine tumor. (1st January 2018)

• Record Type: Journal Article
• Authors: Gao, Heli; Liu, Liang; Wang, Wenquan; Xu, Huaxiang; Jin, Kaizhou; Wu, Chuntao; Qi, Zihao; Zhang, Shirong; Liu, Chen; Xu, Jinzhi; Ni, Quanxing; Yu, Xianjun
• Is Part Of: Cancer letters. Volume 412(2018)
• Page Start: 188
• Page End: 193
• Publication Date: 2018-01-01
• Subjects: Pancreatic neuroendocrine tumor -- Recurrence -- Risk stratification -- Ki-67; Cancer -- Periodicals; Neoplasms -- Periodicals; Cancer -- Périodiques; Electronic journals; 616.994
• DOI: 10.1016/j.canlet.2017.10.036
• Languages: English
• ISSNs: 0304-3835

Abstract

Abstract: Radical surgical resection represents the only hope of cure for pancreatic neuroendocrine tumor (PanNET). Adjuvant therapy is rarely used because there is no evidence to distinguish patients with high recurrence risk. Here we investigated the recurrence feature of resected PanNET and established a novel risk stratification to predict its recurrence. We analyzed 505 PanNET patients who underwent R0 resection at our institute from January 2004 through May 2015. The median follow-up was 71months (range: 12months–143months), 129 patients (25.5%) experienced recurrence with median disease-free survival (mDFS) of 19months. Restricted cubic spline (RCS) functions revealed a positive, linear relationship between Ki-67 index and recurrence. Multivariate analysis showed T stage, N stage, insulinoma and Ki-67 index were independent predictors of recurrence (P < 0.05). Based on scores of these independent factors, we generated a recurrent-risk stage system with HCI of 0.806, superior to TNM stage (HCI 0.704) and grading system (HCI 0.706). Resected PanNET were classified into low risk (65.3%, mDFS not reached), intermediate risk (16.6%, mDFS 48months, 95%CI 26.5–73.4), high risk (13.3%, mDFS 24months, 95%CI 19.4–50.5) and very high risk (4.8%, mDFS 10months, 95%CI 6.9–13.0) (Hazard ratio: 2.650, 95%CI: 2.233–3.145, P < 0.001). This novel risk stratification thus identified PanNET patients of different recurrent-risk. Patients with very high recurrence risk may be suitable for post-operative clinical trials investigating adjuvant treatment. Highlights: Recurrence risk stratification based on 505 postoperative PanNET (HCI = 0.806) was built with independent predictors. Insulinoma showed preferred prognosis despite grade and size. This model identified pNET patients with high and very high risk of recurrence. RCS revealed a positive, linear relationship between Ki-67 and recurrence.