Background The immunopathology of spondyloarthritis (SpA) is determined by inflammation and structural damage, in particular osteoproliferation, of axial and peripheral joints. The failure of TNF blockers to prevent ongoing osteoproliferation raised the concept that inflammation and osteoproliferation are uncoupled processes in SpA. However, inflammation and osteoproliferation are linked in HLA-B27 tg rats, high CRP is associated with radiographic progression in axial SpA, and NSAID treatment can retard osteoproliferation.

Objectives We propose the alternative hypothesis that inflammatory mediators distinct from soluble TNF drive pathologic osteoproliferation in SpA. Based on our observations on soluble versus transmembrane TNF (tmTNF) expression in SpA synovitis, we explored if and how tmTNF drives experimental arthritis and spondylitis.

Methods tmTNF mice (TgA86)¹, provided by Dr Kollias (Athens), were studied clinically over time for arthritis and spondylitis development. At 3, 6 and 8 months joints were collected and analyzed for inflammation and osteoproliferation. To assess the contribution of stromal versus hematopoietic tmTNF expression, tmTNF tg mice and CD45.1 congenic WT mice were lethally irradiated at 4 weeks of age and rescued by 10⁷ bone marrow cells of either WT or tmTNF tg mice. Bone marrow reconstitution, measured in blood 8 weeks after bone marrow transplantation (BMT), was >95% in all cases. Mice were evaluated for 16 weeks, thereafter joints were collected for histologic and radiographic analysis.

Results tmTNF mice (100%; n>50) spontaneously developed arthritis, visualized by deformed joints and loss of grip strength, and spondylitis as evidenced by crinkled tails and hunchback formation, starting at 4 weeks of age and progressing over time. Analysis of 3 months old mice revealed that arthritis was characterized by inflammation of synovium and enthesis. Hypertrophic chondrocytes, as marker for osteoproliferation, were observed outside the bone in the connective tissue next to the inflammation. In spondylitis, inflammation was found in connective tissue located at the junction of the annulus fibrosus with the vertebral bone. Hypertrophic chondrocytes were observed at the edge of the vertebral body, in conjunction with the ongoing inflammation. X-ray images from 8 months old mice also revealed bridging of the tail vertebra. These typical SpA-like features were not observed in any of the non-transgenic littermates.

In the functional experiments, irradiated tmTNF tg mice receiving tmTNF tg BM developed arthritis and spondylitis with 100% incidence 3 weeks after BMT, albeit the arthritis was less severe than in non-irradiated tmTNF tg mice. Interestingly, tmTNF tg mice receiving WT BM also developed both arthritis and spondylitis with the same incidence, onset and severity as the control group. In sharp contrast, WT mice that received tmTNF tg BM did not develop any arthritis, and spondylitis occurred less frequently (66%) and later (10 weeks after BMT) than in the control group.

Conclusions tmTNF overexpression induces experimental arthritis and spondylitis with radiographic and histologic proven new bone formation, indicating that inflammatory mediators can indeed drive osteoproliferation. The data point out the relevance of the transmembrane form of TNF and the role of the stromal compartment in the pathophysiology of SpA.

References

1. Alexopoulou L, et al. Eur J Immunol 1997; 27(10):2588-92.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.2816