New sights of spleen-preserving versus splenectomy in distal pancreatectomy for pancreatic neuroendocrine tumors: a systematic review and meta-analysis

Background: For resectable pancreatic neuroendocrine tumors (pNETs) located in the body or tail of pancreas that require surgery, distal pancreatectomy with splenectomy (DPS) is standard. While splenic-preserving distal pancreatectomy (SPDP) may reduce complications and provide additional benefits, its feasibility in pNETs remains uncertain. This study compares the perioperative outcomes of SPDP versus DPS for pNET patients. Methods: A comprehensive literature search was conducted in PubMed, Embase, and Web of Science included studies published before June 1, 2025. The analysis focused on primary endpoints of intraoperative blood loss (ml) and lymph nodes harvested, as well as secondary endpoints including operative time (min), transfusion, R0 resection, postoperative major complications (PMCs), postoperative pancreatic fistula (POPF), postoperative hemorrhage (PPH), hospital stay (day), lymph node metastasis (LNM), and reintervention. The pooled analysis is presented as odds ratios (OR) or mean differences (MD) with 95% confidence interval (CI). The protocol is registered on PROSPERO (CRD420251079167). Results: 4 retrospective studies involving 457 patients (226 with SPDP and 231 with DPS) were analyzed. 401 patients have well-differentiated G1/G2 tumors, and the majority of SPDP patients have small tumors. Compared to DPS, SPDP had less intraoperative blood loss (SMD, -0.50, 95% CI [-0.90 to -0.11], P = 0.01), fewer lymph nodes examined (MD, -3.30, 95% CI [-5.35 to -1.24], P = 0.002), shorter operative time (MD, -31.78 min, 95% CI [-57.98 to -5.58], P = 0.02), fewer PMCs (OR, 0.57, 95% CI [0.34 to 0.95], P = 0.03) and lower transfusion rates (OR, 0.25, 95% CI [0.07 to 0.83], P = 0.02). In terms of length of hospital stay, SPDP demonstrated more favorable outcomes (MD, -1.13 days, 95% CI [-2.02 to -0.24], P = 0.01). No significant differences were observed regarding R0 resection (OR, 1.40, 95% CI [0.43 to 4.58], P = 0.58), LNM (OR, 0.95, 95% CI [0.49 to 1.85], P = 0.88), or other perioperative outcomes. Conclusion: This study proposes that SPDP may represent a feasible option for selected patients with small, well-differentiated G1/G2 pNETs, suggesting a potential role in reducing surgical risks. These findings should be interpreted as hypothesis-generating, highlighting the need for further investigation. Systematic review registration: https://www.crd.york.ac.uk/PROSPERO/recorddashboard, identifier CRD420251079167.