Background Extensive evidence has correlated epigenetic alterations in articular tissues with both the presence and progression of human osteoarthritis, but few analyses of blood cell epigenetic patterns have been done in OA.

Objectives We examined the DNA methylation aging rate in peripheral blood mononuclear cells (PBMCs) at baseline from knee OA patients with rapid radiographic progression compared to well-matched nonprogressors enrolled in the Osteoarthritis Initiative (OAI).

Methods PBMC DNA was obtained from baseline blood draws of 64 OA patients enrolled in the OAI longitudinal study. All patients had baseline symptomatic and radiographic OA. 32 rapidly-progressive OA patients, defined as ≥1.0mm radiographic joint space loss within 24 month follow-up were compared to 32 non-progressive patients. There were no differences in age, sex, race, BMI, baseline K/L grade, or calculated PBMC subset composition between rapid- and non-progressors. DNA methylation was quantified with Illumina HumanMethylation 450k arrays. Preprocessing was performed in GenomeStudio and normalized to internal controls. Epigenetic age was estimated with the algorithm described by Horvath et. al., using 353 age-associated CpG sites. This epigenetic age was compared to chronological age to calculate epigenetic-chronological age discordance (ΔAge) and group differences compared with a Student t-test. ΔAge was correlated with individual CpG methylation sites of rapid progressors, and Pearson values calculated. Correlation was considered significant if Pearson's r values were ≤-0.55 or ≥0.55 (p≤0.001). Pathway analysis of correlated genes was performed with the Ingenuity Pathway Analysis (IPA) system.

Results The baseline DNA methylation aging rate in rapidly progressive (RP) knee OA patients was decelerated compared to nonprogressors (NP) and to chronological age (ΔAge-RP: -4.9±1.4 vs. ΔAge-NP: -0.071±1.3 mean±SEM years less than chronological age, p=0.015). 1165 CpG sites were correlated with ΔAge in rapid progressors, corresponding to 755 genes. Ontologic analysis of highly correlated genes showed association of the STAT3 pathway (p=6E-4), Notch signaling (p=1E-3), axonal guidance signaling (p=7E-3), CREB signaling (p=2E-2), NFAT signaling (p=2E-2), and autophagy (p=4E-2) among others. Associated upstream regulators included FGF2 (p=3E-5), SMAD4 (p=9E-4), SMAD5 (p=1E-3, TNF (p=4E-3), and TGFB1 (p=4E-3), among others

Conclusions Our data reveal that a decelerated peripheral blood differential DNA methylation age epigenotype is present at baseline in rapidly progressive knee OA patients, but not in nonprogressive knee OA patients. The genes correlated with this methylation age deceleration cluster in pathways previously associated with OA in articular tissues, suggesting that these pathways may systemically epigenetically dysregulated. Our data reinforce the notion that OA is a heterogeneous disease composed of distinct subgroups, and suggests that future epigenetic investigation of immune cell subsets may be beneficial in unraveling OA pathogenesis.

Acknowledgements We gratefully acknowledge the assistance of Dr. Michael Nevitt, PhD, MPH, and Dr. John Lynch, PhD, of the University of California San Francisco and the OAI for their assistance in sample selection.

Disclosure of Interest None declared