Soluble PD-1/PD-L1 biomarkers in NSCLC: prognostic and diagnostic value

BACKGROUND: Tissue-based immunohistochemistry (IHC) for programmed death-ligand 1 (PD-L1) remains the clinical standard for guiding immunotherapy selection in non-small cell lung cancer (NSCLC) but is constrained by intratumoral spatial heterogeneity and procedural sampling risks. Circulating soluble PD-1 (sPD-1) and soluble PD-L1 (sPD-L1) generated either through alternative mRNA splicing or by ADAM10/17-mediated proteolytic shedding offer minimally invasive liquid-biopsy alternatives that capture systemic immune dynamics in real time. METHODS: This review synthesizes clinical and translational data from prospective validation cohorts, surgical trials, and multi-center meta-analyses to evaluate the stage-stratified diagnostic utility, prognostic reliability, and treatment-response predictive value of sPD-1 and sPD-L1 across diverse therapeutic contexts. RESULTS: Cross-sectional baseline assessments show modest standalone diagnostic sensitivity due to significant absolute concentration overlaps between oncological cohorts and healthy controls. However, elevated pre-treatment sPD-L1 functions as a robust independent negative prognostic indicator of overall survival (OS) and a predictor of primary resistance to immune checkpoint inhibitors (ICIs) across advanced stages, while demonstrating no prognostic relevance under conventional cytotoxic chemotherapy. Longitudinally, an acute postoperative sPD-L1 increase of ≥20% at four months marks micro-metastatic persistence and predicts disease recurrence in early-stage surgical cohorts (Odds Ratio = 10.29). Conversely, sPD-1 demonstrates distinct stage-specific volatility; pre-treatment profiles lack stable prognostic indicators, but an early on-treatment plasma surge during anti-PD-1 monotherapy independently maps to prolonged progression-free survival and superior OS (Hazard Ratio = 0.24). Finally, composite liquid biosignatures (e.g., sCombo or joint bsPD-L1/MMP screening) significantly optimize predictive resolution over single-analyte measurements. CONCLUSIONS: Circulating sPD-L1 serves as a reliable negative indicator of therapeutic efficacy and surgical durability, whereas on-treatment tracking of sPD-1 captures protective host T-cell clonal reactivation. Cross-platform assay standardization and predefined cutoff validation remain essential thresholds before these liquid biomarkers can be implemented into standard clinical practice and thoracic surgery. Integrating these circulating checkpoints into high-throughput multi-omics platforms and AI-based analytics is a promising next step toward reproducible predictive signatures that could guide personalized treatment in thoracic oncology.