Related Experiment Video
Updated: Oct 24, 2026
DIPLOMA Approach for Standardized Pathology Assessment of Distal Pancreatectomy Specimens
Published on: February 1, 2020
Prognostic Performance of AJCC Version 9 versus Other Staging Systems for Pancreatic Neuroendocrine Tumors
Jun-Xi Xiang1, Yi-Fan Wei1, Jin He2
1Department of Hepatobiliary Surgery and Institute of Advanced Surgical Technology and Engineering, the First Affiliated Hospital of Xi'an Jiaotong University, Xi'an, China.
Background:
Several anatomical staging systems are used for pancreatic neuroendocrine tumors (pNETs), but their comparative prognostic performance in a common surgical cohort remains uncertain. We compared AJCC Version 9, European Neuroendocrine Tumor Society (ENETS), modified ENETS (mENETS), and modified AJCC (mAJCC) staging.
Methods:
This retrospective multicenter cohort included 891 patients who underwent pNET resection between 2000 and 2023 and had assignable stage under all four systems. Performance was assessed using homogeneity, monotonicity, Harrell concordance index (C-index), Akaike information criterion (AIC), and integrated time-dependent area under the curve (iAUC). C-index uncertainty and optimism were evaluated with 1,000 bootstrap resamples. Overall survival (OS) was the primary outcome.
Results:
All four systems stratified OS (all log-rank p < 0.001). In the final common comparable cohort, the estimated 3-, 5-, and 10-year OS rates were 91.8%, 87.7%, and 67.5%, respectively. ENETS had the highest likelihood-ratio chi-square (26.90), mENETS the highest linear-trend chi-square (22.10), and mAJCC the highest apparent C-index (0.681; bootstrap 95% CI, 0.616-0.742). AJCC Version 9 had the lowest AIC (1126.65) and highest iAUC (0.652). Bootstrap-corrected C-indices were 0.666 for AJCC Version 9, 0.654 for ENETS, 0.647 for mENETS, and 0.661 for mAJCC; confidence intervals overlapped. The equal-weight composite ranked AJCC Version 9 first as a descriptive summary.
Conclusions:
All four systems provided prognostic stratification, and no system dominated every evaluation domain. AJCC Version 9 had the most balanced descriptive profile, but the small differences do not establish clinical superiority or independently direct treatment or surveillance decisions.
