Screening main beneficiaries of PD-1inhibitors are vital for desired outcomes in advanced gastric cancer. The present study aimed to construct an individualized radiomics model to predict the treatment benefits of Programmed cell death 1 (PD-1) inhibitors in advanced gastric cancer.
Clinical data of 58 patients with advanced gastric cancer receiving PD-1 inhibitors were retrospective analyzed to identify major clinicopathological factors related to treatment outcomes. The pretreatment-enhanced computed tomography (CT) imaging data were extracted, and an individual radiomics nomogram was constructed based on the imaging features and clinicopathological risk factors. The performance and clinical utility of radiomics model to predict anti-PD-1efficacy was evaluated through receiver operator characteristic (ROC), and further survival prediction was explored.
Serum carcinoembryonic antigen and tumor metastasis site showed a correlation with PD-1 inhibitors response. The radiomics nomogram based on these two clinical factors and imaging features effectively predicted the treatment response of PD-1 inhibitors in both training (the area under curve [AUC] 0.813, 95% confidence interval [CI]: 0.803–0.822) and validation cohorts (AUC, 0.750, 95% CI: 0.717–0.782), obviously prior to the above clinical factors (P < 0.01). Patients with a low-risk probability of disease progress discriminated by the nomogram showed a significantly longer median progression-free survival time than that those with a high risk (9.7 vs. 5.8 months, HR 0.278, 95% CI: 0.084-0.918, P = 0.036).
The CT-based radiomics nomogram could predict the treatment benefits of PD-1 inhibitors in advanced gastric cancer, thereby guiding the decision-making of clinical treatment.
Clinical trial identification
This clinical trial is a retrospective observational study that was approved by the Ethics Committee of Huazhong University of Science and Technology, (number UHCT-IEC-SOP-016-02-03).
Legal entity responsible for the study
Wuhan Union Hospital.
National Key R&D Program of China (2018YFC1313300), the Natural Science Foundation of Hubei Province of China (Grant No. 2019CFB720) and the Chinese Society of Clinical Oncology (CSCO) (Grant No. Y-MX2016-051).
All authors have declared no conflicts of interest.