Comparative efficacy of neoadjuvant imatinib therapy in gastric versus rectal gastrointestinal stromal tumors: a two-center retrospective cohort study.

OBJECTIVE

This study aimed to evaluate and compare tumor response, survival outcomes, and prognostic factors in gastric versus rectal GIST patients treated with neoadjuvant IM.

METHODS

We conducted a two-center retrospective cohort study of 134 patients with locally advanced gastric (n = 83) and rectal (n = 51) GISTs who received neoadjuvant IM followed by surgery between 2012 and 2024. Radiologic and pathologic responses were assessed, and survival outcomes were analyzed using Kaplan-Meier and Cox regression models.

RESULTS

Preoperative IM (Pre-IM) duration was stratified by the median of 9 months (IQR: 6–12 months). Before IM therapy, the mean tumor size was significantly larger in the gastric group than in the rectal group (12.07 ± 5.42 cm vs. 6.42 ± 2.96 cm, P < 0.001). Both groups demonstrated significant tumor shrinkage after IM (post-treatment size: 8.35 ± 4.33 cm vs. 4.25 ± 2.11 cm, P < 0.001), with comparable percentage reduction in size (29.7%±19.3% vs. 32.5%±17.7%, P = 0.402). Radiologic and pathologic response rates were also similar between the groups. However, rectal GISTs exhibited significantly better overall survival (5-year OS: 98.0% vs. 78.5%, P = 0.003) and a trend toward improved relapse-free survival (5-year RFS: 87.2% vs. 70.1%, P = 0.057). These survival benefits persisted across subgroups stratified by tumor size and Pre-IM duration, with a significant OS advantage observed in rectal GIST patients treated for ≥ 9 months (P = 0.025). Multivariate Cox regression confirmed rectal tumor location as an independent predictor of improved RFS (HR = 0.242, P = 0.022) and OS (HR = 0.044, P = 0.003).

CONCLUSION

Gastric and rectal GISTs showed similar antitumor responses to neoadjuvant IM, while survival appeared more favorable for rectal tumors, particularly with extended Pre-IM. However, this result may be influenced by confounding factors and requires further validation in larger prospective studies.