Activation of Bone Remodeling Compartments in BMP-2-Injected Knees Supports a Local Vascular Mechanism for Arthritis-Related Bone Changes.

Background: Synovial membrane-derived factors are implicated in arthritis-related bone changes. The route that synovial factors use to access subchondral bone and the mechanisms responsible for these bone changes remain unclear. A safety study involving intra-articular injection of bone morphogenetic protein-2 (BMP-2)/calcium phosphate matrix (CPM) or CPM addresses these issues.

Methods: Knee joints in 21 monkeys were injected with CPM or 1.5 or 4.5 mg/mL BMP-2/CPM and were evaluated at 1 and 8 weeks. Contralateral joints were injected with saline solution. Knee joints in 4 animals each were injected with 1.5 or 4.5 mg/mL BMP-2/CPM. Contralateral joints were injected with corresponding treatments at 8 weeks. Both joints were evaluated at 16 weeks. Harvested joints were evaluated grossly and with histomorphometry. Knee joints in 3 animals were injected with 125I-labeled BMP-2/CPM and evaluated with scintigraphy and autoradiography at 2 weeks to determine BMP-2 distribution.

Results: All treatments induced transient synovitis and increased capsular vascularization, observed to anastomose with metaphyseal venous sinusoids, but did not damage articular cartilage. Both treatments induced unanticipated activation of vascular-associated trabecular bone remodeling compartments (BRCs) restricted to injected knees. Bone volume increased in BMP-2/CPM-injected knees at 8 and 16 weeks. Scintigraphy demonstrated metaphyseal 125I-labeled BMP-2 localization restricted to injected knees, confirming local rather than systemic BMP-2 release. Autoradiography demonstrated that BMP-2 diffusion through articular cartilage into the metaphysis was blocked by the tidemark. The lack of marrow activation or de novo bone formation, previously reported following metaphyseal BMP-2/CPM administration, confirmed BMP-2 and synovial-derived factors were not free in the marrow. The 125I-labeled BMP-2/CPM, observed within venous sinusoids of injected knees, confirmed the potential for capsular and metaphyseal venous portal communication.

Conclusions: This study identifies a synovitis-induced venous portal circulation between the joint capsule and the metaphysis as an alternative to systemic circulation and local diffusion for synovial membrane-derived factors to reach subchondral bone. This study also identifies vascular-associated BRCs as a mechanism for arthritis-associated subchondral bone changes and provides additional support for their role in physiological trabecular bone remodeling and/or modeling.

Clinical Relevance: Inhibition of synovitis and accompanying abnormal vascularization may limit bone changes associated with arthritis.