On-treatment dNLR and metastatic burden predict survival in advanced non-small-cell lung cancer treated with nivolumab

Aim: Despite the survival benefit achieved with immune checkpoint inhibitors in advanced non–small-cell lung cancer (NSCLC), outcomes remain heterogeneous. While PD-L1 expression is an established biomarker, less is known about the prognostic relevance of on-treatment systemic inflammatory dynamics and the utility of endothelial stress–based indices such as the Endothelial Activation and Stress Index (EASIX) in patients receiving immunotherapy.Material and Methods: We retrospectively evaluated patients with advanced NSCLC treated with nivolumab. PD-L1 expression, metastatic burden, baseline inflammatory parameters, and baseline EASIX were recorded. On-treatment inflammatory markers, including derived neutrophil-to-lymphocyte ratio (dNLR) at 3 months and longitudinal changes in NLR (ΔNLR), were analyzed. Overall survival (OS) and progression-free survival (PFS) were assessed using Kaplan–Meier and Cox regression models. To account for time-dependent bias, 3-month landmark analyses were performed.Results: Eighty-one patients were included. High PD-L1 expression (≥50%) was associated with significantly improved OS but did not retain independent prognostic significance for PFS. A high metastatic burden, particularly involvement of four or more metastatic organs, was associated with inferior OS. In landmark-adjusted analyses, dNLR at 3 months was independently associated with both OS and PFS, while ΔNLR was associated with OS, indicating that favorable on-treatment immune–inflammatory changes correlated with improved survival. Baseline EASIX was not associated with either OS or PFS.Conclusion: In nivolumab-treated advanced NSCLC, survival outcomes are primarily driven by tumor biology, disease burden, and immune–inflammatory dynamics evolving during therapy. On-treatment inflammatory markers, particularly dNLR and ΔNLR, provide more clinically relevant prognostic information than baseline indices, whereas baseline EASIX appears to have limited utility in this setting. These findings support a treatment-specific, time-dependent approach to biomarker-based risk stratification in immunotherapy.