Skip to main content
. 2026 Aug 12;17:1907116. doi: 10.3389/fimmu.2026.1907116

Figure 1.

Flowchart outlining a study of 738 patients with locally advanced gastric or EGJ adenocarcinoma treated with neoadjuvant PD-1 inhibitor and chemotherapy, divided into training, internal validation, and external validation cohorts. A pretreatment peripheral blood panel assesses endocrine, nutrition, and inflammation markers, followed by machine learning model comparison using a five-marker ENI panel, leading to ENI-Score calculation and benefit stratification.

Study design and ENI-score workflow, from multicenter cohort allocation through pretreatment marker collection, machine-learning model comparison, and tertile-based clinical stratification.