Primary tumor size as a prognostic factor in duodenal neuroendocrine neoplasms ≤20 mm: a population-based SEER analysis

Objective: The prognostic value of primary tumor size in duodenal neuroendocrine neoplasms (dNENs) ≤20 mm remains uncertain. Although size is used to guide management, its independent effect on survival is unclear. Design and methods: We performed a population-based study using the US SEER database (1975–2021), including adults with histologically confirmed dNENs ≤20 mm. Disease-specific survival (DSS) was the primary endpoint. Associations between tumor size, stage, lymph-node involvement, and DSS were evaluated using chi-squared tests, logistic regression, Kaplan–Meier curves, and Cox models. Receiver operating characteristic (ROC) analysis identified the optimal size cutoff. Mediation analysis assessed whether stage or grade mediated the size–DSS relationship. Results: Among 3,515 eligible cases, median age was 65 years, 51% were male, and 89% had well-differentiated NETs. Most tumors were localized (85.2%) and ≤10 mm (74.4%). Median follow-up was 56 months, with 160 disease-specific deaths (4.6%). In univariable analysis, size >10 mm was associated with worse DSS, but the association disappeared after adjustment for grade, stage, and surgery. Grade G3 and advanced stage independently predicted poorer survival, while radical surgery was protective. Larger tumors were more frequently high grade and advanced stage (P < 0.001), and each 1 mm increase raised the odds of nodal metastasis by 12% (OR 1.12). ROC analysis identified a size threshold of 10.5 mm (AUC = 0.74). Stage mediated only ∼2% of the size–DSS association. Conclusions: In duodenal NENs ≤ 20 mm, tumor size >10 mm correlates with disease extent but does not independently affect DSS, supporting its role as a stratification marker rather than a prognostic factor.