Background/Purpose:  

Lateral compartment knee tibio-femoral radiographic osteoarthritis (ROA) is associated with pain and disability, and the epidemiology of this form of knee OA is not well known. We examined MRI abnormalities and their association with lateral compartment ROA.

Methods:  

We analyzed knees from participants in the Osteoarthritis Initiative (OAI), an NIH-funded study of persons aged 45-79 at baseline with, or at high risk of knee ROA.  Baseline radiographs with central readings and MRIs were used to assess knee structure lesions.  We defined cases as knees with Kellgren/Lawrence (K/L) grade >=2 and with joint space narrowing (JSN) >0 in the lateral compartment but JSN = 0 in the medial compartment; controls were defined as having K/L=0 and JSN=0 in all compartments.  Cases and controls were matched by sex and age in 10-year bands. MRI readings for cases and controls were obtained from two ancillary studies in the OAI and further MRI readings were generated for this analysis.  MRIs were scored at each subregion for cartilage lesions, bone marrow lesions, and meniscal tear or extrusion using the MOAKS scoring system. The score for each lesion was dichotomized into “any” or “none”.  Lesions in medial and lateral compartments were defined separately for whether any of the subregions in each compartment exhibited that lesion. We compared cases and controls for each structural lesion separately in both lateral and medial compartments using conditional logistic regression, adjusting for age, race, body mass index (BMI), and clinic site.

Results:  

Included in this analyses were 200 cases (mean age: 63.8, 34.5% men, mean BMI 29.5, and 76.5% white) and 200 controls (mean age: 63.1, 34.5% men, mean BMI 27.1, and 87.5% white). Isolated lateral compartment ROA was strongly associated with lateral compartment cartilage lesions (odds ratio (OR)=89.9, 95%CI 29.7-271.9), BMLs (OR=40.0, 95% CI: 21.1-75.7) and meniscal tear or extrusion (OR= 93.8, 95% CI 43.5-202.3). Medial compartment cartilage lesions, BMLs and meniscal abnormalities were also associated with isolated lateral compartment ROA; the magnitude of association, however, was much smaller. (See table)

Conclusion:  

Knees with isolated radiographic OA in the lateral compartment have more MRI abnormalities of cartilage, meniscus and BMLs in the lateral compartment than knees without radiographic OA. Isolated lateral compartment ROA is also associated with medial compartment MRI lesions, but less strongly.