Comparative Analysis of Characteristics and Outcomes of Patients with Early-Onset Versus Average-Onset Small Bowel Neuroendocrine Tumors.
OBJECTIVES The incidence of early-onset gastrointestinal malignancies is increasing globally; however, early-onset small bowel neuroendocrine tumors (EO-SBNETs) remain poorly characterized. We evaluated clinical characteristics, treatment patterns, incidence trends, and survival outcomes of EO-SBNETs compared with average-onset SBNETs (AO-SBNETs). METHODS We conducted a retrospective population-based cohort study using the Surveillance, Epidemiology, and End Results (SEER) database (2004 to 2021). Patients with SBNETs were identified using ICD-O-3 histology codes and categorized as early-onset (30 to 49 y) or average-onset (≥50 y). Baseline characteristics were compared using χ2 tests. Overall survival (OS) was estimated using the Kaplan-Meier method and evaluated with multivariable Cox regression models that adjusted for demographic, socioeconomic, clinical, and treatment-related factors. Temporal incidence trends were assessed using linear regression. RESULTS Among 3713 patients with SBNETs, 925 (25.0%) had EO-SBNETs. EO-SBNETs were more frequently observed in females and racial and ethnic minority populations, including non-Hispanic Black and Hispanic patients (all P<0.05). Stage at diagnosis and treatment patterns, including rates of surgery, lymphadenectomy, chemotherapy, and radiation therapy, were comparable between the groups. Incidence increased significantly over time in both EO-SBNETs and AO-SBNETs (annual increase 14.9% vs. 14.8%, P<0.001), without differences in temporal trends. The median OS was not reached for AO-SBNETs and was 188 months for EO-SBNETs. On multivariable analysis, early-onset disease was independently associated with improved survival (HR=0.50, 95% CI: 0.41-0.60, P<0.001). CONCLUSIONS EO-SBNETs represent a rising and clinically distinct subgroup characterized by similar presentation and treatment but relatively favorable long-term outcomes compared with AO-SBNETs, suggesting potential biological differences warranting further investigation.