Estimating survival probabilities of advanced gastric cancer patients in the second-line setting: The gastric life nomogram

Objective: We built and externally validated a nomogram for predicting the overall survival (OS) probability of advanced gastric cancer patients receiving second-line treatment. Methods: The nomogram was developed on a set of 320 Italian patients and validated on two independent sets (295 Italian and 172 Korean patients). Putative prognostic variables were selected using a random forest model and included in the multivariable Cox model. The nomogram's performance was evaluated by calibration plot and C index. Results: ECOG performance status, neutrophils to lymphocytes ratio, and peritoneal involvement were selected and included into the multivariable model. The C index was 0.72 (95% CI 0.68-0.75) in the development set, 0.69 (95% CI 0.65-0.73) in the Italian validation set, but only 0.57 (95% CI 0.52-0.62) in the Korean set. While Italian calibrations were quite good, the Korean one was poor. Regarding 6-month OS predictions, calibration was best in both Caucasian cohorts and worst the in Asian one. Conclusions: Our nomogram may be a useful tool to predict 3-or 6-month OS in Caucasian gastric cancer patients eligible for second-line therapy. Based on three easy-to-collect variables, the Gastric Life nomogram may help clinicians improve patient selection for second-line treatments and assist in clinical trial enrollment. (C) 2018 S. Karger AG, Basel