Studies of the synovial membrane in chronic rheumatic joint disease

Abstract

Background: The synovial membrane (SM) outlines the inner cavity of synovial joints except for cartilage surfaces. The SM is the target organ of immune-mediated responses in chronic arthritis, including rheumatoid arthritis (RA). RA is a common joint disease (prevalence of 0.5-1.0%) which is characterized by recurrent joint inflammation, with increasing joint damage and disability. The study of the SM in vivo provides important information about disease pathogenesis and response to treatment.

The aim of this thesis was to study a method for biopsy retrieval, arthroscopy, SM variation by gene and protein expression and SM response to anti-rheumatic therapy.

Results:

In these studies we have added to existing evidence that arthroscopy is a safe and reliable method for synovial tissue retrieval.

We have constructed and validated an easy to use macroscopic score for arthroscopy. This score needs validation against other methodologies.

We have studied variability of synovial gene expression, where we could see that samples close to one another were more related than samples further apart, especially from other patients. Arthroscopic biopsies had less intrinsic variation than had orthopaedic material.

We have shown that injection of intra articular glucocorticosteroids (GCs) reduces protein expression of synovial proinflammatory molecules (TNF, IL-1β, extra-nuclear HMGB-1, ICAM-1 and VEGF) and T cells. mRNA expression was reduced for IL-1α and IL-1β, but not for TNF or HMGB-1. Unexpectedly, no changes were evident in macrophage infiltration and the vascular compartment. Vascular proinflammatory cytokine expression persisted.

We have shown that mPGES-1 (an enzyme in the prostaglandin E2 synthesis pathway) is strongly expressed in RA lining cells, and also in sublining macrophages, synovial fibroblasts and endothelial cells. We also saw that mPGES-1 and COX-2 co-localised in cells from synovial fluid and tissue. Finally mPGES-1 and COX-2 were down-regulated by local GC treatment, but not by systemic TNF antagonists.

We have shown that markers of destruction were positively affected by both GCs and TNF antagonists, although differently. RANKL was down-regulated by GCs and OPG was up-regulated by TNF-blockade, leading to a favourable reduction of RANKL/OPG ratio by both treatments.

We have provided evidence that TNF-antagonists induce apoptosis in RA SM macrophages, but not lymphocytes.

Conclusion: Arthroscopy is a well tolerated, safe and reliable tool that will continue to be important for synovial tissue sampling. Synovial variation has to be considered when sampling the SM. By local GC treatment signs of inflammation persist, perhaps contributing to later disease relapse. We provided evidence for two inducible enzymes in the prostaglandin synthesis pathway, and showed that inhibition depended on therapy. mPGES-1 inhibitors would be an interesting future therapy possibly without the adverse effects of the cyclooxygenase inhibitors. Interestingly both GC and anti-TNF therapy reduced the RANKL/OPG ratio, an indirect sign of reduced osteoclast activity, although by different path ways. These findings provide support for a biological background for the slowing of joint destructions by these treatments. The increase in apoptosis of macrophages indicate that TNF antagonists make the local environment more normal, as reduced apoptosis is a classical feature in active RA SM.

Two important questions remain: Is it possible to predict treatment response, and thereby minimise patient suffering? Is it possible to understand disease pathogenesis, and develop a cure? These are questions that need answering, and using SM as a source for research will possibly provide with important leads. Studying old and especially the new targeted therapies represent important opportunities to these questions.