Background/Purpose: Patellofemoral (PF) braces have been shown to increase contract area in this joint, decreasing focal stress  and may also correct patellar malalignment. Bone marrow lesions (BMLs) are lesions in the subchondral bone of osteoarthritis (OA) knees on MRI that are caused in part by focal stress in the joint.  Because BML size correlates with knee pain severity, because they predict structural progression and because recent evidence suggests, unlike cartilage, that their volumes fluctuate substantially in as little as 6 weeks, BML’s may be a good outcome to test short term structure modification in OA trials. We performed a trial testing the efficacy of  the BIOSKIN patellar tracking Q brace (Ossur, inc).  

Methods: Eligible subjects had to have pain daily of >=40/100 for the last 3 months during a PF activity and on examination had tenderness over patellar facet or a positive patellar compression test.  X-rays had to show KL grade 2 or 3 PFOA which was worse than tibiofemoral disease. At baseline,  subjects  were randomly allocated to immediate or 6 weeks provision of the brace; the trial lasted 6 weeks.  Subjects with other forms of arthritis were excluded as were those ineligible for MRIs or with impaired renal function. Gadolinium MRIs were acquired on a 1.5 T MRI at baseline, and at 6  weeks.  Postgadolinium fat suppressed sagittal and PD weighted axial images were used for BML’s which were manually segmented (blinded to sequence and to treatment) at all time points. The primary symptom outcome was the change in pain at 6 weeks (0 – 100 VAS scale) during the PF painful activity in the more symptomatic knee and the primary structural outcome was BML volume in the PF joint of that knee (patella and trochlea). We also examined tibiofemoral (TF) BML volumes as a control untreated region.  Analyses used multiple linear regression with an ITT approach.

Results:          We randomised 126 subjects age 40-70 years (mean age 55.5 years (SD 7.5 years); 72 females (57.1%).  Mean pain score at baseline was 64.6. Ninety of the 125 with MRI scans (72%) had PF BML’s at baseline. Subjects wore the brace for a mean of 7.4 hrs/day. Six subjects withdrew during the RCT (an additional 2 did not attend one session, and therefore had only one MRI). Compared with the control group, the brace group had both a significant reduction in PF-related knee pain and a decrease in the volume of BML’s in the PF  but not in the TF joint (see Table). 

Conclusion: A patellofemoral brace is efficacious for treatment of PF OA reducing both symptoms and BML volumes in the targeted compartment. BML’s may be a responsive outcome measure in OA trials.