Association of adjuvant somatostatin analogue therapy with disease-free survival in patients with gastroenteropancreatic neuroendocrine tumours: a Chinese multi-institutional propensity score matched and weighted analysis (CASSANET-GEP).

BACKGROUND

Recurrence remains a major concern after curative-intent resection of gastroenteropancreatic neuroendocrine tumours (GEP-NETs), and evidence to guide postoperative adjuvant therapy is limited.

METHODS

We performed a multicentre retrospective cohort study across nine university-affiliated hospitals in China, including patients with grade 1-3 GEP-NETs who underwent curative-intent resection between January 2007 and December 2024; follow-up ended on 1 October 2025. Exposure was postoperative adjuvant SSAs. The primary endpoint was disease-free survival (DFS) and the secondary endpoint was overall survival (OS). DFS and OS were estimated using Kaplan-Meier methods, and associations were evaluated using Cox regression after propensity score matching (PSM) and inverse probability of treatment weighting (IPTW). Time-related bias was addressed using time-dependent Cox and landmark analyses at 3 and 6 months after surgery.

RESULTS

Among 1602 patients, 358 received adjuvant SSAs. After 1:1 PSM, 560 patients were retained with improved baseline balance. Adjuvant SSAs were associated with a lower recurrence risk in the overall cohort in unadjusted (HR 0.466, 95% CI 0.323-0.673; p < 0.001) and IPTW-adjusted analyses (HR 0.419, 95% CI 0.278-0.631; p < 0.001), with similar directionality in G-NETs and P-NETs, whereas estimates in E-NETs were imprecise. In time-dependent Cox analyses treating SSAs as an ever-started time-varying exposure, adjuvant SSAs remained associated with longer DFS in the whole cohort (adjusted HR 0.519, 95% CI 0.353-0.764; p < 0.001), and findings were consistent in landmark analyses at 3 months (adjusted HR 0.523, 95% CI 0.339-0.806; p = 0.003) and 6 months (adjusted HR 0.537, 95% CI 0.356-0.811; p = 0.003). In the overall cohort, 24- and 36-month DFS rates were 92.4% and 89.3% in the SSAs cohort versus 86.9% and 84.2% in controls, respectively; corresponding 24- and 36-month OS rates were 99.1% and 98.7% versus 97.2% and 96.9%, respectively.

CONCLUSIONS

In this large multicentre real-world cohort, adjuvant SSAs were associated with longer DFS after curative-intent resection of GEP-NETs. The overall direction of association remained consistent in propensity-adjusted, time-dependent, and landmark analyses, supporting the robustness of the primary findings. Prospective studies are needed to confirm effectiveness and refine patient selection.