First-line immunochemotherapy modulates peripheral immune landscape in advanced gastric cancer and gastroesophageal junction cancer (GC/GEJC): Prediction of response to combination therapy.

e14540 Background: The use of immunochemotherapy has brought about a change in the treatment of advanced or metastatic GC/GEJC, but only a subset of patients could benefit substantially. Therefore, seeking efficacious biomarkers to identify favorable populations for the regimen is imperative. This study aims to explore predictive markers and potential mechanism of combination therapy for GC/GEJC in real-world settings based on blood proteomic biopsy. Methods: This is a prospective, single-center trial in which all patients are adults with advanced or metastatic GC/GEJC confirmed by pathology or imaging. 30 eligible patients will receive PD-1 mab (sintilimab/nivolumab (NIVO) Q3W) and chemotherapy (SOX/XELOX Q3W) regardless of programmed death ligand-1 status. Blood samples were collected before treatment and after every two treatment cycles for Olink Target 96 IO analysis. CT scans were performed after every two cycles and will be categorized as responding (complete response plus primary response) or non-responding (progress disease plus stable disease) according to RECIST 1.1. Results: As of January 17, 2024, a total of 21 patients were enrolled. Eight of these patients received 16 samples before and after treatment, with 4 each of the 8 patients responding and non-responding. By analyzing the baseline samples comparing the responding and non-responding groups, it was found that patients in the responding group had lower expression of serum PD-L1 (sPD-L1) ( p=0.015). In the post-treatment samples, angiopoietin 1 was found to be poorly expressed in the responding patients ( p<0.001). In contrast, high expression of angiopoietin 2 in post-treatment blood samples was more pronounced in non-responders compared to pre-treatment ( p=0.020). Additionally, we observed that immunotherapy can modify the immune microenvironment by up-regulating the T cell receptor signaling and NK cell-mediated toxic effects pathways. These pathways may be potential targets for treatment. Conclusions: In summary, low sPD-L1 level before treatment may be a predictive biomarker for better efficacy of immuno-combination chemotherapy in advanced GC/GEJC; whereas elevation of angiopoietin after treatment predicts poor efficacy and suggests that antiangiogenic therapy may subsequently be added in non-responding patients. We are further recruiting more patients to obtain more accurate conclusions. Clinical trial information: ChiCTR2300078000 .