An integrated prognosis prediction model based on real-word clinical characteristics for immunotherapy in advanced...

Immune checkpoint inhibitors (ICIs) benefit only a subset of patients in advanced esophageal squamous cell carcinoma (ESCC). Our study aims to develop and validate a clinically accessible model to better identify those who may respond to ICIs.

This study enrolled advanced ESCC patients treated with ICIs at Peking University Cancer Hospital from January 14, 2016, to January 26, 2024 for the training cohort and at Harbin Medical University Cancer Hospital between January 10, 2019, and July 6, 2022 for the validation cohort. Combined positive score (CPS) was recorded to assess the predictive value of programmed cell death ligand-1 (PD-L1). Baseline clinical and laboratory characteristics were identified as predictors through Akaike information criterion (AIC) and Cox proportional hazards regression. The prediction model underwent internal validation through bootstrapping and was externally validated in the validation cohort.

The training cohort consisted of 430 patients, while the validation cohort included 184 patients. PD-L1 expression failed to discriminate survival outcomes. The prediction model incorporates 10 variables: stage, bone metastasis, line of therapy, treatment, lactate dehydrogenase, carcinoembryonic antigen, carbohydrate antigen 199, systemic immune-inflammation index, lymphocyte count and prognostic nutritional index. The model achieved a C-index of 0.725 in the training cohort, 0.722 following bootstrapping, and 0.691 in the external validation cohort. An interactive online prediction tool ( https://escc-survival.shinyapps.io/shiny_app/ ) was subsequently developed.

This is the first large-scale, real-world model for individualized survival prediction for advanced ESCC patients treated with ICIs, offering a practical tool for optimizing clinical decisions.