Background/Purpose

We sought to investigate if cartilage thickness change over 24 months predicts clinically relevant progression (radiographic and/or symptomatic) in knee OA over a 48 month period. 

Methods

The OA Biomarkers Consortium undertook a nested case-control study of progressive knee OA within the Osteoarthritis Initiative (OAI). Main inclusion criteria were KLG 1, 2 or 3 at baseline and availability of knee radiograph and magnetic resonance imaging (MRI) at baseline and 24 months.  The primary case group (n=194) was defined by the combination of knee radiographic progression (medial tibiofemoral joint space loss (mTF JSL) >= 0.7mm) AND pain progression (persistent worsening in WOMAC pain score, reaching a MCID threshold of 9 points on a 0-100 scale), each achieved for the first time at the 24, 36 or 48 month follow-up compared to baseline; referred to as “JSL and pain progressor”.  We defined two additional case subsets in order to distinguish structural from pain progression: 103 subjects with JSL but no pain increase comprised “JSL only progressors”: and 103 with a persistent increase in pain but no JSL comprised “Pain only progressors”.  “Nonprogressors” (n=200) were participants with a knee eligible for the study that did not meet either progression definition (pain or mTF JSL). Manual segmentation of the femorotibial cartilages was performed by trained readers to generate cartilage thickness.  The measures used in this analysis were (i) change in cMFTC (central medial femorotibial compartment), and (ii) change in ccMF (central medial femur) and cMT (central medial tibia). Association between cartilage thickness measures and progressors vs nonprogressors was assessed using a logistic regression model adjusting for baseline age, sex, BMI, KLG and baseline pain level. Additional analyses were conducted to detect marginal effects of cartilage thickness on changes in pain and mTF JSL.

Results

Participants had mean age of 61.5 years, 59% female and the majority were obese.  Changes in cartilage thickness from BL to 24 months in the cMFTC, ccMF and cMT were greater for all three progressor groups combined compared to nonprogressors and significantly associated with increased odds of being a progressor (Table).  The ORs for cartilage thickness changes ranged from 1.6 to 2.8, with the largest ORs associated with cartilage thickness changes in the ccMF OR 2.8 (95%CI 2.1 to 3.7). Further analysis suggested that these associations were with structural progression and not pain progression.

Table. Change in cartilage thickness at 24 months and prediction of case control status OR (95% CI) p value. ORs represent the change in odds of being a progressor per 1SD increase of normalized changes in cartilage thickness.

Conclusion

Changes in cartilage thickness markers over 24 months clearly differentiate progressors from nonprogressors. These associations are largely attributable to mTF JSL and not to pain increases.