Background: The diagnosis and management of primary Sjögren’s syndrome (pSS) remains a challenge, and treatments preventing progression of the disease are yet to be discovered. Proximity extension assay (PEA) allows for highly sensitive and specific detection of aberrances in the serum and tissue proteome. Variabilities of proteomic profiles will allow for stratification of patient subsets and may aid identification of patients at risk for extraglandular manifestations, monitoring of the disease course, as well as evaluation of immunomodulatory effects of novel treatments.

Objectives: To obtain an in-depth understanding of the salivary gland and serum proteome in patients with pSS and to relate proteomic variations to clinical, serological, and genetic parameters.

Methods: Paired lysates of salivary gland biopsies and serum from patients with pSS (n=80) and sicca controls (n=19), as well as serum samples from patients with pSS and healthy controls from a Swedish (n=456 and n=141) and a Norwegian (n=233 and n=137) cohort were analyzed by Olink PEA technology. Patients were stratified based on occurrence of extraglandular manifestations and the presence or absence of ANA, anti-Ro/SSA, and anti-La/SSB autoantibodies. Risk-associated HLA alleles were imputed. Proteins detected in >80% of samples were analyzed, and all analyses were adjusted for age and sex. Cut-off for significance was set at 5% FDR and ± 0.5 log2 FC.

Results: Initial screening of salivary gland lysates and serum using six different PEA panels encompassing 512 unique protein markers determined the Immuno-Oncology panel to be the most relevant for further use in the larger cohorts. In the Swedish discovery cohort, 28 serum proteins significantly associated with pSS compared to healthy controls, out of which 16 were successfully validated in the Norwegian replication cohort. The most significant proteins comparing patients and healthy controls in the discovery cohort which could also be validated were galectin-9, CCL19, TNF, and soluble PD-1. Pulmonary involvement was present in 23 patients and associated with a striking up-regulation of 22 inflammatory proteins. Differences in distinct protein levels were also noted for cutaneous, lymphadenopathy, renal, and biological manifestations with CXCL10 being the most frequently detected marker. Notably, levels of the 16 validated protein biomarkers were found to gradually increase according to autoantibody profiles, with the lowest levels in ANA negative patients and highest levels in patients positive for both anti-Ro/SSA and anti-La/SSB. Statistical analysis of trend among these patient subgroups revealed that the protein level gradients were significant for all 16 proteins (p<0.0001). Correspondingly, higher inflammatory protein levels as well as a higher protein type I interferon (pIFN) score associated with frequency of the risk-associated HLA-DRB1*03 and DRB1*15.

Conclusion: These data increase our understanding of the dysregulated proteome in patients with pSS and indicate how such aberrances relate to the presence of autoantibody specificities, risk-associated HLA as well as extraglandular manifestations.

REFERENCES: NIL.

Acknowledgements: This study was partly funded by the NECESSITY grant of EU Innovative Medicines Initiative 2.

Disclosure of Interests: Albin Björk: None declared, Johannes Mofors: None declared, Lauro Meneghel: None declared, Marika Kvarnström: None declared, Roland Jonsson: None declared, Roald Omdal: None declared, Helena Forsblad-d’Elia: None declared, Sara Magnusson Bucher: None declared, Per Eriksson: None declared, Thomas Mandl Working as medical lead at UCB., Peter Olsson: None declared, Juliana Imgenberg-Kreuz: None declared, Gunnel Nordmark: None declared, Marie Wahren-Herlenius: None declared.