Pan-Cancer Proteomic and Transcriptomic Meta-Analysis of PD-1/PD-L1 Signaling Reveals Predictive Biomarkers for Anti-PD-1 Therapy in Lung Cancer.

Programmed death-ligand 1 (PD-L1) is an immune checkpoint molecule that enables tumor cells to escape immune surveillance, and its blockade by immune checkpoint inhibitors has become an effective therapeutic strategy in various cancers. Previous studies have primarily focused on genomic and transcriptomic features within specific cancer types, while proteomic analyses remain relatively limited. Here, we conducted a comprehensive pan-cancer meta-analysis encompassing 12 human cancer types to systematically characterize PD-1/PD-L1 signaling pathways at both the proteomic and transcriptomic levels. We observed clear cancer-type-specific patterns of PD-1/PD-L1 expression. Pathway-crosstalk analyses further revealed multiple pathways and phosphorylation events influencing PD-1/PD-L1 activity. Immune-infiltration profiling identified MMP9 neutrophils as key immune subsets associated with PD-1/PD-L1 pathway activity. Using proteomic data, we constructed a PD-L1-centered protein-protein interaction network and identified TAP proteins as potential predictive biomarkers in small-cell lung cancer. We also established a biomarker signature capable of predicting clinical response to anti-PD-1 therapy in non-small-cell lung cancer and validated this gene set in two independent previously published cohorts. Finally, we developed an interactive web application (https://yuyingsuo-simm.shinyapps.io/PD-L1_Profiling/) to facilitate visualization of PD-L1 features and exploration of its associations with genes and pathways of interest.