Nomogram based on CT imaging and clinical data to predict the efficacy of PD-1 inhibitors combined with chemotherapy in advanced gastric cancer.

Background: PD-1 inhibitors, in combination with chemotherapy, have become the first-line treatment option for patients with advanced metastatic gastric cancer. However, some patients still do not benefit from this treatment, highlighting an urgent need for simple and reliable markers to predict the efficacy of immunotherapy. Methods: Immunotherapy efficacy was evaluated using RECIST 1.1 and categorized into complete remission (CR), partial remission (PR), stable disease (SD), and disease progression (PD). Patients with CR, PR, and SD were classified as non-PD responders, while PD patients were categorized as PD responders. Clinical characteristics and CT imaging features of gastric cancer patients from two centers, before receiving PD-1 inhibitor combination chemotherapy, were retrospectively analyzed. A univariate logistic regression analysis was performed for each variable, and separate models for clinical and imaging characteristics, as well as a nomogram, were developed. Area under the curve (AUC), accuracy, sensitivity, specificity, and decision curve analysis (DCA) were used to evaluate all models. Results: Data from 272 patients (non-PD responders = 206, PD responders = 66) from Center 1 were collected for this study. Data from 76 patients (non-PD responders = 54, PD responders = 22) from Center 2 were used as an external validation cohort to verify the robustness of the models. We developed a clinical model, an imaging features model, and a nomogram. The nomogram, combining clinical and imaging features, demonstrated superior performance with an AUC of 0.904 (95% CI: 0.862-0.947) in the training set and an AUC of 0.801 (95% CI: 0.683-0.918) in the validation set, with sensitivity, specificity, and accuracy of 0.889, 0.682, and 0.829, respectively. Calibration curves further confirmed the agreement between actual results and predictions. Conclusions: A nomogram combining clinical features and CT imaging features before treatment was developed, which can effectively and simply predict the efficacy response of advanced gastric cancer patients treated with PD-1 inhibitors combined with chemotherapy. This tool can aid in optimizing treatment strategies in clinical practice.