Predicting recurrence after curative resection for gastric cancer: External validation of the Italian Research Group for Gastric Cancer (GIRCG) prognostic scoring system

Background Most nomograms for Gastric Cancer (GC) were developed to predict overall survival (OS) after curative resection. The Italian Research Group for Gastric Cancer (GIRCG) prognostic scoring system (PSS) was designed to predict the recurrence risk after curative treatment based on pathologic tumor stage and treatment performed (D1-D2/D3 lymphadenectomy). This study was carried out to externally validate the GIRCG's PSS. Patients and methods Adopting the same criteria used by GIRCG to build the PSS, 185 patients with GC operated with curative intention were selected. The median follow-up period was 77.8 months (1.93-150.8) for all patients and 102.5 months (60.9-150.8) for patients free of disease. The NRI (net reclassification improvement) was calculated to estimate the overall improvement in the reclassification of patients using the PSS in place of the TNM stage system. Results GC recurrence occurred in 70 (37.8%) patients. The mean time to recurrence was 22.2 (range 1.9-98.1) months. For patients with recurrence, the gain in the proportion of reclassification was 0.257 (p < 0.001), indicating an improvement of 26%. For patients without recurrence, the gain in the proportion of reclassification was -0.122 (p < 0.001), indicating a worsening of 12%. The NRI calculated was 0.135 (p = 0.0527). Conclusion The GIRCG's PSS, which predicts the likelihood of recurrence after radical surgical treatment for GC, is more accurate than TNM system to predict recurrence mainly for high-risk patients. Yet, the PSS does not have the same effectiveness for low-risk patients, overestimating the chance of recurrence occurs even for disease-free patients.