Limitations of Post-treatment TNM Staging After Neoadjuvant Targeted Therapy and Surgery for Gastrointestinal Stromal Tumors: A SEER-Based Study with Two-Center External Validation.

内容摘要

The American Joint Committee on Cancer (AJCC) TNM system is widely used for gastrointestinal stromal tumor (GIST) staging, but its value after neoadjuvant targeted therapy is uncertain. We evaluated TNM performance after targeted therapy and surgery using the Surveillance, Epidemiology, and End Results (SEER) database and a two-center validation cohort. SEER patients with surgically treated GIST diagnosed from 2000 to 2022 were classified as no systemic therapy (NS), systemic therapy after surgery (SA), or systemic therapy before surgery (SB). Cancer-specific survival (CSS) and Harrell C-index were analyzed overall and by site. External validation included 171 locally advanced patients with GIST treated with neoadjuvant imatinib followed by surgery. Reconstructed post-treatment ypTNM was evaluated for recurrence-free survival (RFS) and overall survival (OS). The SEER cohort included 5561 patients: 3304 NS, 1870 SA, and 387 SB. TNM discrimination was highest in NS (C-index 0.755), intermediate in SA (0.683), and lowest in SB (0.676). In SEER, discrimination decreased in gastric GISTs after preoperative systemic therapy (NS 0.739; SB 0.616) but was preserved in non-gastric GISTs (NS 0.763; SB 0.769). In validation, ypTNM predicted RFS (C-index 0.700) and OS (0.702), with weaker discrimination in gastric than non-gastric GISTs for both RFS (0.655 vs. 0.753) and OS (0.663 vs. 0.807). TNM/ypTNM staging shows reduced, site-dependent performance after preoperative targeted therapy for GIST. In both SEER and external validation analyses, prognostic discrimination was weaker in gastric GISTs and better preserved in non-gastric GISTs. Risk stratification after neoadjuvant imatinib should integrate ypTNM with pretreatment burden, primary site, genotype, response, rupture status, and perioperative imatinib exposure.