A New Scoring System to Predict Recurrent Disease in Grade 1 and 2 Nonfunctional Pancreatic Neuroendocrine Tumors

The aim of this study was to predict recurrence in patients with grade 1 or 2 nonfunctioning pancreatic neuroendocrine tumors (NF-pNET) after curative resection. Surgical resection is the preferred treatment for NF-pNET; however, recurrence occurs frequently after curative surgery, worsening prognosis of patients. Retrospectively, patients with NF-pNET of 3 institutions were included. Patients with distant metastases, hereditary syndromes, or grade 3 tumors were excluded. Local or distant tumor recurrence was scored. Independent predictors for survival and recurrence were identified using Cox-regression analysis. The recurrence score was developed to predict recurrence within 5 years after curative resection of grade 1 to 2 NF-pNET. With a median follow-up of 51 months, 211 patients with grade 1 to 2 NF-pNET were included. Thirty-five patients (17%) developed recurrence. The 5- and 10-year disease-specific/overall survival was 98%/91% and 84%/68%, respectively. Predictors for recurrence were tumor grade 2, lymph node metastasis, and perineural invasion. On the basis of these predictors, the recurrence score was made. Discrimination [c-statistic 0.81, 95% confidence interval (95% CI) 0.75-0.87] and calibration (Hosmer Lemeshow Chi-square 11.25, P = 0.258) indicated that the ability of the recurrence score to identify patients at risk for recurrence is good. This new scoring system could predict recurrence after curative resection of grade 1 and 2 NF-pNET. With the use of the recurrence score, less extensive follow-up could be proposed for patients with low recurrence risk. For high-risk patients, clinical trials should be initiated to investigate whether adjuvant therapy might be beneficial. External validation is ongoing due to limited availability of adequate cohorts