Diagnostic performance of somatostatin receptor positron emission tomography in preoperative evaluation of pancreatic neuroendocrine tumor lymph node metastases

Hallbera Gudmundsdottir; Fabricio J Hernandez-Delima; Bryan C McDowell; Patrick Starlinger; Cornelius A Thiels; Susanne G Warner; Thorvardur R Halfdanarson; Patrick J Navin; Annie T Packard; Sean P Cleary

doi:10.1016/j.surg.2026.110136

Diagnostic performance of somatostatin receptor positron emission tomography in preoperative evaluation of pancreatic neuroendocrine tumor lymph node metastases

Surgery. 2026 Mar 17:194:110136. doi: 10.1016/j.surg.2026.110136. Online ahead of print.

Affiliations

¹ Department of Surgery, Mayo Clinic, Rochester, MN.
² Department of Radiology, Mayo Clinic, Rochester, MN.
³ Department of Medical Oncology, Mayo Clinic, Rochester, MN.
⁴ Department of Surgery, Mayo Clinic, Rochester, MN; Division of General Surgery, University of Toronto, Toronto, ON, Canada. Electronic address: Sean.Cleary@uhn.ca.

PMID: 41849846
DOI: 10.1016/j.surg.2026.110136

Abstract

Background: Regional lymph node involvement impacts prognosis for patients with pancreatic neuroendocrine tumors and may influence management decisions for small tumors. Conventional cross-sectional imaging modalities have low sensitivity in detecting regional lymph node metastases, but the diagnostic performance of somatostatin receptor positron emission tomography for this is unknown.

Methods: Patients with pancreatic neuroendocrine tumors who underwent preoperative gallium 68 DOTATATE positron emission tomography followed by resection with lymphadenectomy from 2017 to 2022 were reviewed. Preoperative gallium 68 DOTATATE positron emission tomography was retrospectively reviewed by radiologists. The presence of suspicious regional lymph nodes was correlated with pathologic reports to determine diagnostic accuracy.

Results: Among 130 patients who met inclusion criteria, DOTATATE positron emission tomography detected suspicious lymph nodes in 24%, whereas pathologic evaluation demonstrated lymph node involvement in 42%. Overall, sensitivity was 46% and specificity was 92%, with a positive predictive value of 81% and negative predictive value of 29%. Among patients with lymph node involvement confirmed on pathologic review, 92% of patients with true-positive results had a Krenning score of 4 compared with 59% of patients with false-negative results (P = .013), and 58% of patients with true-positive results had lymph nodes >1 cm on preoperative cross-sectional imaging compared with 8.7% of patients with false-negative results (P < .001).

Conclusion: Gallium 68 DOTATATE positron emission tomography has limited sensitivity but high specificity for preoperative diagnosis of regional pancreatic neuroendocrine tumor lymph node metastases. Detection rate improves with higher Krenning score and larger lymph node size. This study has important implications for preoperative planning, particularly for patients with small tumors who may be candidates for either observation or resection.