It was hypothesized many years ago by Radin that osteoarthritis was a disease of subchondral bone. Advances in imaging notably MRI and DXA scanning have greatly enhanced our understanding of this process. There are a number of facets of bone involved in cartilage damage and/or loss in the knee. These include bone marrow lesions (BMLs), tibial plateau area, subchondral bone density (sBMD) and osteophytes. In addition, there is evidence that BMLs, bone attrition and osteophytes are also associated with symptoms although the latter is controversial. BMLs and tibial bone area also predict knee replacement independent of pain and radiographic changes suggesting they may be a marker of fast progression. Most of these changes tend to worsen over time. BMLs can worsen, stay static or improve over time and this fluctuation is associated with change in symptoms. There is also evidence that BMLs are linked to pain in the hip and tenderness in the hands. Taking these observations into account there are some key potential targets. We don't know how to alter bone size or bone attrition. sBMD could be altered by antiresorptive agents but this would potentially be deleterious as it is higher sBMD that precedes cartilage damage and causing enhanced BMD loss would have the additional effect of increasing the risk of osteoporosis and fractures. Osteophyte progression in the spine can be slowed by alendronate based on a post hoc analysis of the original osteoporosis trial. Risedronate appeared to have no effect on the knee but this may reflect measurement issues. Observational data suggests BMLs are much less common in those taking bisphosphonates (but not other antiresorptives) while risedronate may slow progression of BMLs. Recently, a proof of principle trial of the potent IV agent zoledronic acid was performed where all subjects had BMLs. Chondroitin may also have an effect on BMLs. Strontium ranelate has also been shown to help pain and slow down radiographic change in knee osteoarthritis but it remains uncertain if this is mediated by its bone effects or a cartilage effect. Back pain is also less common with zoledronic acid. These data imply that we should be choosing lesion specific therapy for osteoarthritis especially bone targeted therapy for those with BMLs (and possibly osteophytes?).
Disclosure of Interest None declared