The Analysis Of Risk Factors Affecting The Disease-free Survival Of G1/2 Pancreatic Neuroendocrine Tumor Patients After Curable Surgery And The Establishment Of A Novel Predictive System

Objective Disease progression of G1/2 pancreatic neuroendocrine tumors is relatively slow and the life expectancy seems to be better compared with other pancreatic malignant tumors.Clinical data of patients who underwent curable surgery of G1/2 pancreatic neuroendocrine tumors from December 2002 to May 2018 in our institute were retrospectively analyzed in this study to explore the prognostic factors affecting the diseasefree survival(DFS)after surgery.Based on these independent risk factors,we built a prediction model for DFS and stratified patients with recurrence risk according to the prediction model,aiming at providing clinical recommendation of comprehensive postoperative treatment.Method 245 patients were ultimately included in this study.In the part of risk factor analysis,univariate and multivariate Cox proportional hazards models were performed on potential prognostic factors to establish the prognostic index(PI)equation,and Kaplan-Meier curves depicting DFS were computed with the log-rank test applied to verify the significance in differences.The prediction model for DFS was built using multivariate Cox regression analysis,while the nomogram and calibration plot were draw using R software.We also calculated the C-index and area under curve(AUC)of receiver working curve(ROC)of the proposed model and ENETS stage system.We presented a novel scoring system for recurrence risk stratification based on the prediction model and draw Kaplan-Meier curves by different risk layers.Result Univariate and multivariate analysis suggested T stage,N stage and WHO grade were independent risk factors for DFS,the PI equation could be expressed as: PI = ey,y = 1.102 *(T stage-1.706)+ 1.576 *(N stage-0.151)+ 2.681 *(WHO grade-1.294).The proposed prediction model compromised tumor size,N stage and WHO grade.The calibration plot displayed good ability of consistency.The AUC and C-index,which represented the ability of discrimination,of the proposed model were superior to those of ENETS stage system.In addition,we established a novel scoring system for recurrence risk stratification to sort patients into three groups: score<15.4 for low risk(79.0%,3-and 5-year recurrence risk 0.8% and 4.3%,median DFS not reached),score>15.4,<24.5 for intermediate risk(20.8%,3-and 5-year recurrence risk 11.6% and 21.4%,median DFS 70 months,)and score>24.5 for high risk(10.2%,3-and 5-year recurrence risk 37.3% and 68.7%,median DFS 49 months).Kaplan-Meier analysis reminding DFS revealed significant differences among three groups(P<0.001).Conclusion(1)T stage,N stage and WHO grade were independent risk factors for DFS after curable surgery of G1/2 pancreatic neuroendocrine tumors.(2)We constructed a prediction model for DFS compromising tumor size,N stage and WHO grade.The prediction model displayed good consistency and discrimination.(3)We proposed a novel scoring system for recurrence risk stratification based on the prediction model to identify patients with high risk of recurrence after curable surgery and provide evidence for adjuvant therapy.