A Machine Learning Model Based on Clinical Factors to Predict the Efficacy of First-Line Immunochemotherapy for Patients With Advanced Gastric Cancer - Retrospective Study

Sep 29, 2026

Background: Immunotherapy offers new hope for advanced gastric cancer (AGC), yet first-line immunochemotherapy efficacy varies due to disease heterogeneity, and simple predictive models are lacking. Objective: Identify key clinical factors and build models to predict treatment efficacy using routinely available data, supporting evidence-based guidance and personalized strategies. Methods: Retrospective clinical data were used with training, internal validation, and temporal validation, applying Cox/LASSO modeling, four ML prognostic models, and performance/interpretability analyses. Results: The RSF model showed superior prediction and net clinical benefit; key predictors were age, subtype, immune cell proportions, and liver metastasis, enabling risk stratification by Kaplan-Meier outcomes. Conclusions: The RSF model supports quantifiable individualized decision-making.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages18

Reading Time45 min

AccessFree to download

Abstract

Background and Objective

Methods

Results and Conclusions

Introduction

Gastric cancer burden and unmet need

Rationale for immunochemotherapy and predictive modeling

JMIR MEDICAL INFORMATICS Cheng et al Original Paper A Machine Learning Model Based on Clinical Factors to Predict the Efficacy of First-Line Immunochemotherapy for Patients With Advanced Gastric Cancer: Retrospective Study | Xu Cheng1,2, MSc; Ping Li1,2, PhD; Enqing Meng1,2, MSc; Xinyi Wu 1,2, MBS; Hao Wu 1,2, PhD | | --- | | 1Gastric Cancer Center, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China 2Department of Oncology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China
Corresponding Author:
Hao Wu, PhD
Department of Oncology
The First Affiliated Hospital of Nanjing Medical University
No. 300, Guangzhou Road, Gulou District, Nanjing City, Jiangsu Province Nanjing, 210029
China
Phone: 86 13913855335
[Email:](Email: whdactor@njmu.edu.cn)[ ](Email: whdactor@njmu.edu.cn)[whdactor@njmu.edu.cn](Email: whdactor@njmu.edu.cn)
Abstract | | Background: The development of immunotherapy has provided new hope for patients with advanced gastric cancer (AGC) . However, due to the high heterogeneity of the disease, the efficacy of first-line immunochemotherapy varies among patients. There is still a lack of simple and effective models to predict the efficacy of immunochemotherapy in this setting.
Objective: This study aimed to identify critical factors and develop predictive models to evaluate the efficacy of first-line immunochemotherapy in patients with AGC using clinically available data. The goal was to offer evidence-based guidance for clinical practice and enable personalized treatment strategies.
Methods: To evaluate the effectiveness of first-line immunochemotherapy in AGC, we retrospectively collected clinical data from The First Affiliated Hospital of Nanjing Medical University between January 2018 and October 2023. The data collected were divided into a training set (168/240, 70%) and an internal validation set (72/240, 30%). Additionally, a temporal validation cohort of 76 patients recruited from November 2023 to September 2024 was assembled to further evaluate the predictive performance of the models. We used univariate and multivariate Cox regression analyses, along with the least absolute shrinkage and selection operator (LASSO) regression, and integrated clinical expertise to identify key predictors of treatment efficacy and to construct the LASSO-Cox model. We developed 4 models (LASSO-Cox, random survival forest [RSF], extreme gradient boosting, and survival support vector machine) and evaluated their performance using the C-index, area under the curve (AUC), calibration curves, and decision curve analysis. The optimal model was interpreted using Shapley additive explanations, and its risk scores were used to stratify patients for Kaplan-Meier survival analysis.
Results: Among the 4 prognostic models developed in this study, the RSF model demonstrated superior predictive accuracy and discrimination for progression-free survival, as evidenced by its higher AUC, concordance index, continuous AUC curves, and calibration curves compared with the other 3 models. Additionally, decision curve analysis showed that the RSF model offered greater net clinical benefit. The Shapley additive explanations results identified that age, histological subtype, the proportion of CD19+ B cells, CD16+CD56+ natural killer cells, and the presence of liver metastasis were key prognostic factors influencing patient outcomes. Patients in the low-risk group, as determined by the RSF model’s risk score, exhibited a significantly higher progression-free survival rate than those in the high-risk group, further validating the value ofthe RSF model for risk stratification. Conclusions: This study is the first to use machine learning algorithms to develop a predictive model for the efficacy of first-line immunochemotherapy in AGC, and to identify key predictors of treatment outcome. The results indicate that the RSF model not only enables precise stratification of

FAQs

What clinical problem does this study address?

Efficacy of first-line immunochemotherapy for advanced gastric cancer varies among patients, and there is a lack of simple, effective models to predict that efficacy. The study aims to fill this gap using routinely available clinical data.

How were the predictive models developed and evaluated?

The study retrospectively collected clinical data from a single hospital, split into training and internal validation sets, and added a temporal validation cohort. Four models were built (LASSO-Cox, RSF, XGBoost, and survival SVM) and assessed using C-index, AUC, calibration curves, and decision curve analysis.

Which model performed best and what factors were most important?

The random survival forest (RSF) model showed the best predictive accuracy and discrimination for progression-free survival, and provided greater net clinical benefit. Shapley additive explanations identified age, histological subtype, CD19+ B-cell proportion, CD16+CD56+ NK-cell proportion, and liver metastasis as key prognostic factors.

Sep 29, 2026

Upload by วิน