Procedure-specific morbidity and pathologic assessment after resection of small (≤20 mm) pancreatic neuroendocrine tumors

Pancreatic neuroendocrine tumors (PanNETs) are increasingly detected, and the optimal extent of resection remains uncertain because perioperative burden and oncologic assessment may differ by procedure. Enucleation (EN) and central pancreatectomy (CP) are often grouped together as parenchyma-sparing resections (PSRs) despite substantial differences in technical complexity, postoperative risk, and oncologic evaluation.We conducted a multicenter retrospective study of consecutive patients who underwent resection for pathologically confirmed PanNETs <= 20 mm. Procedures were categorized as EN, CP, pancreatoduodenectomy (PD), total pancreatectomy (TP), or distal pancreatectomy (DP). Primary outcomes were clinically relevant postoperative pancreatic fistula (CR-POPF) and major morbidity, defined as Clavien-Dindo grade >= IIIa. A prespecified pairwise comparison focused on EN and CP, and pathologic assessment patterns were analyzed descriptively across procedures. Of 460 patients, 164 (35.7%) underwent EN, 40 (8.7%) CP, 84 (18.2%) PD/TP, and 172 (37.4%) DP. CR-POPF occurred in 115 patients (25.0%), major morbidity in 68 (14.8%), and 90-day mortality in 4 (0.9%). Among PSRs, CP had higher major morbidity than EN (25.0% vs 10.4%, p =.020), higher CR-POPF (50.0% vs. 24.4%, p =.003), and a longer median hospital stay (19.5 vs. 10.0 days, p <.001). Lymph node (LN) assessment was less frequent after EN than CP (40.4% vs. 87.5%, p <.001). Among assessed cases, LN metastases were identified in 44/320 patients (13.8%). EN and CP are not interchangeable as parenchyma-sparing options for small PanNETs. Among patients selected for resection, procedure selection should consider tumor location, function, technical feasibility, and oncologic requirements rather than size alone.