Purpose: To explore associations between MRI-defined structural abnormalities and clinical features related to knee osteoarthritis (OA).
Methods: Structural and clinical knee OA features were assessed in 87 women (45 with knee OA symptoms). Structural features were quantified by the Kellgren and Lawrence grade on radiography and by the Boston-Leeds Osteoarthritis Knee Score on MRI. Clinical features were assessed using the Knee Injury and Osteoarthritis Outcome Score, functional tests and muscle strength measurements. Associations were examined using regression analyses.
Results: Limited significant associations between structural and clinical features were found. An increased meniscal signal was associated with more pain/symptoms (P < 0.027). An anterior cruciate ligament tear was associated with poorer stair climbing test performance (P = 0.045). In a stepwise linear regression model, patellofemoral cartilage integrity and pain explained 28 % of the isometric quadriceps strength variability. The amount of cartilage lesions, loose bodies and pain explained 38 % of the isokinetic quadriceps strength variability. Synovitis/effusion and patellofemoral cartilage integrity combined with pain explained 34 % of the isometric hamstring strength variability.
Conclusion: Although MRI-detected cartilage lesions, synovitis/effusion and loose bodies did explain part of the muscle strength variability, results suggest that MRI does not improve the link between joint degeneration and the clinical expression of knee OA. MRI contributes less than expected to the understanding of pain and function in knee OA and possibly offers little opportunity to develop structure-modifying treatments in knee OA that could influence the patient's pain and function.
Level of evidence: Case series with no comparison groups, Level IV.