Immune-inflammatory-fibrinogen score as a novel prognostic biomarker in patients with gastric cancer undergoing radical gastrectomy: a multicenter study
摘要
Gastric cancer (GC) progression involves changes in immune responses, inflammation, and coagulation. The prognostic value of related biomarkers remains unclear. This study aimed to develop a novel immune-inflammation-fibrinogen score (FSL score) to predict outcomes in GC patients after radical gastrectomy.
MethodsClinicopathological data from 401 GC patients enrolled in a randomized controlled trial (2015–2016; ClinicalTrials.gov: NCT02327481) were retrospectively analyzed as the training cohort, with 173 patients included for external validation. Cox regression was used to construct the FSL score based on preoperative hematological markers and to evaluate its association with overall survival (OS) and recurrence-free survival (RFS). A nomogram incorporating the FSL score and clinicopathological factors was developed and evaluated using the C-index, time-dependent AUC, calibration curves, AIC, BIC, and decision curve analysis (DCA), and compared with the AJCC 8th TNM staging system.
ResultsPatients with high FSL scores had significantly better 3-year OS (training cohort: 87.4% vs. 71.3%, p = 0.001; validation cohort: 68.4% vs. 42.6%, p = 0.009) and RFS (training cohort: 81.0% vs. 67.6%, p = 0.001; validation cohort: 72.4% vs. 54.9%, p = 0.025). The FSL score independently predicted OS and RFS (all p < 0.05). A nomogram integrating the FSL score, preoperative CEA level, pT stage, pN stage, and postoperative chemotherapy outperformed the TNM system for OS (C-index: 0.806 vs. 0.757; AIC: 1378.25 vs. 1395.73; BIC: 1390.49 vs. 1394.10) and RFS (C-index: 0.803 vs. 0.763; AIC: 1310.10 vs. 1319.08; BIC: 1320.11 vs. 1322.26), with good discrimination, calibration, and clinical utility confirmed by DCA and external validation.
ConclusionsThe FSL score is a promising prognostic biomarker for GC patients undergoing radical gastrectomy. The proposed nomogram enables accurate and individualized survival prediction.