Authors: Mia Church1, Sophie Coyle1, Colin Willoughby2, James Church3, Colm O’Brien4, Sarah-Louise McDaid-McCloskey1, Fionn O’Leary4, Pirro Hysi5, Ajda Pristovec5, Steven Watterson1, Sarah Atkinson2, Meredith Gregory-Ksander5, Victoria McGilligan1

Affiliations: 1 Ulster University, Personalised Medicine Centre, School of Medicine, Derry/Londonderry, Northern Ireland, 2 Ulster University, Centre for Genomic Medicine, School of Biomedical Science, Coleraine, Northern Ireland, 3 Altnagelvin Area Hospital, Department of Ophthalmology, Derry/Londonderry, Northern Ireland, 4 Mater Misericordiae University Hospital, Department of Ophthalmology, Dublin, Ireland, 5 Harvard Medical School, Schepens Eye Research Institute of Massachusetts Eye and Ear, Department of Ophthalmology, Boston, United States

Background/ Introduction: Glaucoma is a leading cause of irreversible blindness globally, yet circulating biomarkers reflecting disease risk and pathogenesis remain poorly defined. This study aimed to identify and replicate plasma protein signatures associated with glaucoma across independent prevalent, incident and clinical cohort analyses.

Material & Methods: Plasma proteomic data from the UK Biobank Pharma Proteomics Project (UKB-PPP), comprising 2,917 proteins measured using the Olink Explore 3072 platform, were analysed in two complementary analyses. A logistic regression model identified proteins associated with prevalent glaucoma (780 cases), adjusting for age, age², sex, BMI, lifestyle factors, medication use, spherical equivalent and 20 genetic principal components. A Cox proportional hazards model identified proteins associated with incident glaucoma (~900 cases) over follow-up. Findings were cross-referenced with an independent Northern Ireland clinical cohort (39 glaucoma cases) and a UK Twins cohort (170 prevalent glaucoma cases).

Results: The prevalent analysis identified 290 significant proteins and the incident analysis identified 136 significant proteins (p<0.05). Three proteins: GLYR1, SATB1 and B3GNT7, were shared between prevalent and incident UKB analyses. Seven proteins: IL-6, CCL23, CXCL10, HGF, OSM, IL-24 and IL-4, overlapped between the incident UKB analysis and the Northern Ireland clinical cohort. IL-18 and DNER were identified in both the local cohort and twins cohort analyses. These overlapping proteins are predominantly involved in inflammatory cytokine signalling, implicating immune dysregulation in glaucoma pathogenesis.

Conclusion: Cross-cohort proteomics analysis identified replicated inflammatory protein signatures associated with glaucoma across prevalent, incident and clinical populations, highlighting IL-6, CXCL10, HGF, IL-18 and DNER as priority candidates for further investigation as glaucoma biomarkers.