Absolute quantification of aqueous humor proteins for prognostic stratification in uveal melanoma

Background: Uveal melanoma (UM) is the most common primary intraocular malignancy in adults and carries a high risk of metastatic progression, particularly in tumors with high-risk gene expression profiles or advanced clinical stage. Direct tumor biopsy provides valuable prognostic information but carries procedural risk, highlighting the need for minimally invasive approaches to capture tumor-derived biomarkers. Aqueous humor (AH) contains detectable molecular material released by intraocular tumors, yet absolute quantification of protein biomarkers has not previously been performed. This study aims to establish the first calibrated AH proteome map for UM and to identify protein biomarkers associated with molecular classification and disease stage. Methods: AH samples from 70 UM eyes were collected before plaque brachytherapy. Fine needle tumor aspiration provided gene expression profiling (GEP1, GEP2) and PRAME status. All samples underwent NGS based high-plex proximity extension assay (PEA), and 27 samples with remaining volume underwent qPCR-based PEA targeting 92 proteins. Regression models derived from 66 overlapping proteins enabled extrapolation of absolute pg/mL concentrations across the full cohort. Proteins with median concentrations >5 pg/mL were assessed for differences by GEP class, PRAME status, and AJCC stage. Pathway enrichment and upstream regulators were identified using Ingenuity Pathway Analysis. Results: Twenty-three proteins were present at clinically quantifiable concentrations. The lowest protein levels were observed in GEP1/PRAME- tumors, with progressive increases seen across advanced AJCC stages. Significant elevations were noted in Stage IV disease for CXCL8 (p=0.00024), IL10 (p=0.00049), and PDCD1 (p=0.00094). Pathway analysis demonstrated enrichment in immune and tumor-associated signaling pathways. VEGFA (z=3.1) and CCL2 (z=1.7) were identified as predicted upstream regulators and were elevated in advanced-stage tumors. Conclusions: This study provides the first absolute quantitative map of AH proteins in UM and identifies a panel of 23 clinically detectable biomarkers associated with tumor biology, molecular classification, and stage. These findings position AH proteomics as a minimally invasive liquid biopsy platform capable of capturing stage-linked immune activity in UM and supporting the development of standardized biomarkers for prognostic stratification.