Primary Resection and Metastasectomy for Well-Differentiated Grade 3 Gastro-Entero-Pancreatic Neuroendocrine Tumors: A National Cancer Database Analysis.

Abstract

In patients with synchronous liver-metastatic low-grade (WHO grade 1/2) gastro-entero-pancreatic neuroendocrine tumors (GEPNETs), resection of primary and liver metastases is associated with improved overall survival (OS). The benefit of surgical resection in metastatic well-differentiated grade 3 GEPNETs, a recently defined tumor grade, remains unclear. We aimed to define survival benefit from primary resection and liver metastasectomy in patients with well-differentiated WHO grade (G) 3 GEPNETs. Patients diagnosed with well-differentiated WHO grade 1-3 synchronous liver-metastatic GEPNETs between 2018-2021 were identified from the National Cancer Database (NCDB). Kaplan-Meier and multivariable Cox proportional hazards modeling were utilized to evaluate the association of primary tumor and liver resection with OS. Among 3,198 patients (G1: n=1434; G2: n=1113; G3: n=651), surgical intervention decreased with higher grade (G1: 67.3%; G2: 62.4%; G3: 31.3%, P<0.001). Ninety-day mortality was higher in G3 tumors (21.1%) compared with G1 (1.7%) and G2 (3.2%) (P<0.001), due to suspected early disease-related mortality. After a median follow-up of 35.5 months, median OS among G3 tumors was 8.1 months without surgery, 11.8 months with primary-only resection, and 24.4 months with liver and primary resection (combined liver/primary resection vs. no surgery: log-rank P=0.007), however survival curves converged over time with similar 3-year OS (25.7% vs. 19.3%). In contrast, G1-2 patients undergoing combined resection had improved 3-year OS (88.2% vs. 67.1%, P<0.001). Primary and liver resection are associated with improved OS in G1-2 liver-metastatic GEPNETs. In G3 disease, surgical resection provides limited long-term survival benefit and should be considered selectively in a multidisciplinary setting.