Inflammatory-Erosive Arthritis, Arthritic Flare, and Bone Marrow Edema
Contrast Enhanced MRI scans of the knee of a patient with inflammatory psoriatic arthritis before (left) and after (right) anti-TNF therapy shows the drug effects on bone marrow edema. 3D reconstructed images of the edema signals differentiate the synovial fluid (blue) and the pathologic bone marrow edema (red) as described in (Anandarajah et al., 2008).
Based on the success of biologic therapy, it is clear that rheumatoid arthritis (RA) is mediated by increased TNF levels, and an emerging role for B cells has been identified. However, the functional relationship between TNF and B cells in RA pathogenesis is unknown. Additionally, how arthritic
flare, as defined by the initiation of synovitis and focal erosion, occurs in select joints during a chronic-systemic autoimmune disease remains an enigma. Thus, there is a need to:
- elucidate the cellular and molecular mechanisms of RA flare,
- validate novel therapeutic targets, and
- develop novel outcome measures to assess active RA, response to therapy, and flare potential.
To this end, we developed high resolution (105µm) in vivo contrast enhanced (CE) MRI of murine arthritis with in vivo CT, which permits longitudinal quantitation of synovial volume, erosion volume and popliteal lymph node (PLN) volume.
In our natural history studies in TNF-Tg mice, we made several novel findings that serve as the basis for this program. First, PLN appears to be the most sensitive biomarker of knee arthritis, as quantitation of its size (inflammation) and CE following i.v. Gd-DTPA (function) demonstrated significant changes that precede synovitis and focal erosion. Moreover, we discovered two distinct PLN phenotypes that exist in the same animal, which could explain the role of B cells in arthritic flare. Initially,
expanding PLN have striking CE and paracortical sinuses that are mostly void of cells, which appear to protect the adjacent knee from arthritis. In contrast, arthritis appears to follow PLN
collapse, which is characterized by a decreased volume and CE. These non-CE regions correspond to
plugged sinuses that are completely filled with B-cells, and closely resembles LN shutdown during an immune response. As the CXCL13-CXCR5 chemokine axis has been identified to critically regulate B cell migration, these findings lead us to propose a unifying hypothesis to explain the roles of TNF and B cell activation in RA flare which posits that dysregulated TNF production from RA joints is initially counteracted by the draining LN, which serves to remove excess fluid, cells and inflammatory/catabolic factors from the joint to prevent synovitis and focal erosion. This compensatory mechanism ultimately breaks down due to B-cell accumulation and
clogging of the sinuses from either:
- functional changes of B cells in the draining LN, and/or
- secondary to synovial lymphoid neogenesis, promoting synovitis and joint destruction from excess lymph.
Currently, the lab is testing the hypothesize that:
- Lymphatic flow into collapsed PLN of TNF-Tg mice is significantly reduced vs. that of expanded PLN in the contralateral leg, which correlates with inflammatory-erosive arthritis.
- B-cell depletion in a collapsed PLN restores its function and ameliorates arthritis.
- Disruption of CXCL13 signaling prevents the collapse of expanding PLN and arthritis.
- And, these expanding and collapsed PLN also exist in patients with RA and can be studied via CE-MRI and Doppler ultrasound.
3D reconstructed image of the ankle bone of a mouse with erosive arthritis as described in (Mensah et al., 2010).
Focal erosions in RA, which leads to irreversible structural damage, deformities and loss of function, are known to be mediated by inflammation-induced osteoclasts, which are multinucleated cells that are responsible for bone resorption. Interestingly, patients with both RA-like disease and systemic lupus erythematosus (SLE), have a non-erosive form of inflammatory arthritis called Jaccoud's arthritis (JA). Recently, we have shown that the mechanism for this non-erosive JA is medicated through lupus-induced interferon-alpha (IFNa), which inhibits osteoclast formation at the cell fusion step. Since the dendritic cell specific transmembrane protein (DC-STAMP) is known to mediate osteoclast precursor fusion during osteoclast formation, and we have recently shown that DC-STAMP is regulated by this process, we are currently investing its signal transduction mechanism and its potential as a biomarker of erosive disease.
One of the greatest mysteries of arthritis, and the greatest clinical concern in treating the disease, is the source of joint pain. The greatest obstacle towards understanding joint pain is the absence of an objective-quantitative outcome measure that can be used in animals and people. Bone marrow edema signals observed on CE-MRI have been studies in RA, osteoarthritis (OA) and various spondyloarthropathies (SpA). To explore this further in murine models of RA, OA and degenerative disk disease (DDD), we have developed custom surface coils that interface with a clinical 3T MRI that allows us to perform natural history and therapeutic intervention studies.