OBJECTIVE: PD-1 inhibitors are increasingly used in advanced cancers, but reliable biomarkers for predicting treatment response remain limited. METHODS: In this retrospective cohort study, we analyzed 335 patients with stage IV cancers treated with anti-PD-1 therapy (The First People's Hospital of Kunshan, 2019-2025). A predictive model integrating peripheral blood markers and clinical factors was developed using logistic and Cox regression. RESULTS: Histologic subtype, dNLR ≤ 3.18, AMC ≤ 0.48, and IL-6 ≤ 12.35 were independent prognostic factors for OS and PFS (multivariate HRs reported). Patients with ≥ 3 favorable factors had significantly better outcomes. The model showed high accuracy for 1-year survival prediction (AUC = 0.83) but limited performance beyond 3 years. CONCLUSION: This study proposes a clinically feasible biomarker-based tool to stratify PD-1 inhibitor responders, though further validation in prospective cohorts is needed.
Feng et al. (Fri,) studied this question.