Abdom Radiol (NY). 2026 Apr 12. doi: 10.1007/s00261-026-05425-0. Online ahead of print.
ABSTRACT
OBJECTIVE: To explore the value of tumor volume reduction rate (VRR) measured by enhanced CT, compared with RECIST 1.1 criteria, in evaluating the efficacy of neoadjuvant chemotherapy combined with immunotherapy(NACI) for patients with locally advanced gastric cancer (LAGC), and its impact on prognosis.
METHODS: This retrospective study included 107 gastric adenocarcinoma patients. VRR and RECIST 1.1 response were calculated from pre-treatment and post-treatment CT. Interobserver agreement was evaluated using the intraclass correlation coefficient(ICC), while correlation with pathological tumor regression grade (TRG) and diagnostic performance (ROC analysis) were assessed. The optimal VRR cut-off was determined. Prognostic value was evaluated via survival analysis.
RESULTS: The ICC for the tumor VRR was 0.966 (95% CI: 0.951-0.977), which was significantly higher than the ICC for the tumor longest diameter reduction rate (0.701, 95% CI: 0.590-0.785). VRR showed a moderate negative correlation with TRG (r= -0.389, P < 0.001), which was significantly better than the weak correlation between RECIST 1.1 and TRG (r = 0.196, P = 0.043). The area under the curve (AUC) for VRR in predicting major pathological response was significantly higher (AUC = 0.7953) than that for RECIST 1.1 (AUC = 0.6127, P = 0.0017). Survival analysis showed that grouping based on VRR (2-year PFS: 69.4% vs. 37.8%, P = 0.018) and TRG (2-year PFS: 82.8% vs. 46.2%, P = 0.0029) significantly distinguished patient prognosis, whereas RECIST 1.1 criteria did not (P = 0.74). The combined model integrating VRR and TRG failed to provide more refined prognostic stratification. Among patients with TRG 2-3, whether VRR was greater than 47.4% did not significantly stratify prognosis (P = 0.7935).
CONCLUSIONS: (1) The tumor VRR demonstrates significantly superior observer agreement compared to the RECIST 1.1 criteria. (2) The correlation with TRG and diagnostic efficacy of VRR before and after neoadjuvant chemotherapy combined with immunotherapy are significantly superior to RECIST 1.1 criteria. (3) Regarding 2-year PFS, both VRR and TRG are effective indicators for prognostic evaluation after this treatment regimen, whereas RECIST 1.1 criteria lack statistical significance for 2-year PFS. (4) The combined model of VRR and TRG did not enable more precise prognostic stratification, indicating a need for further research.
PMID:41966643 | DOI:10.1007/s00261-026-05425-0
