Background/Purpose: The cause of knee pain in osteoarthritis (OA) is multi-factorial, and there is increasing evidence of the role of inflammation in OA. The goal of this study was to examine the association between development of incident frequent knee pain or improvement of frequent knee pain with changes in semi-quantitative (SQ) MRI imaging biomarkers of inflammation.

Methods: Data from the Osteoarthritis Initiative (OAI) study was used for this analysis. Knees from the OAI that were at risk of developing incident radiographic OA through the 48-month OAI visit were included if they had SQMRI readings and data on frequent knee pain (KP) on at least two successive OAI annual visits. MRIs were read according to the MRI Osteoarthritis Knee Score (MOAKS) system, unblinded to order of acquisition but blinded to clinical data. Frequent knee pain was defined as “pain on most days of the past 30 days”. For each pair of successive annual time points where KP was absent at the first time point, we examined the potentially increased odds of transitioning to KP (no KP to KP+) predicted by whether Hoffa synovitis (HS) had a concurrent change from absent (no HS) to present (HS+) vs other HS patterns. Concurrent effusion synovitis (ES) vs KP transition patterns were analogously compared.  We also compared the HS and ES concurrent transitions when a knee went from KP+ to no KP. State transition modeling was used to estimate the odds of transitioning from no KP to KP+ in successive years predicted by HS  or ES transitions. The absence of Hoffa synovitis or effusion synovitis at both time points was the referent group. Random effects mixed models were used to account for clustering of knees within a person and multiple time points for each knee. Models were adjusted for age, gender and BMI.

Results: We included 345 knees from 313 participants, 66% female and with a mean age of 60 years; 62% of knees were KLG 1 and 38% were KLG 0 at baseline (Table 1). No ES to ES+ and no HS to HS+ were the only changes associated with the transition from no KP to KP+ in multivariate analyses. When stratified by sex, the association of no HS to HS+ with the transition of no KP to KP+ was stronger for men (aOR=7.37, 95% CI: 1.54-35.32) than for women (aOR=3.95, 95% CI: 1.47-10.61), whereas for no ES to ES+, the association with transition of no KP to KP+ was significant only in men (aOR=3.98, 95% CI: 1.39-11.38), but not women (aOR=1.78, 95% CI: 0.84 -3.74). There were no significant associations of these SQMRI biomarkers with the transition from KP+ to no KP.

Conclusion: Incident frequent knee pain was associated with the presence of new Hoffa synovitis or effusion-synovitis, and these associations varied by sex. These SQMRI biomarkers of inflammation are potential targets for therapeutic interventions for pain in knee OA.

Table 1. Association between semi-quantitative MRI biomarkers of inflammation and incident knee pain

*From No KP to KP+ at next time point

**adjusted for age, gender, BMI

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/incident-frequent-knee-pain-is-associated-with-changes-in-semi-quantitative-imaging-biomarkers-of-inflammation/