Tendon disease is often the first presentation of rheumatoid arthritis (RA). Tenosynovial proliferation results in scarring and adhesion formation. Synovial invasion into tendons occurs in 50% of cases and is associated with multiple tendon ruptures and a poorer prognosis. Recent work on diseased joints suggests that hypoxia may play a key role in synovial invasion. We hypothesised that hypoxia promotes and maintains the inflammatory response in RA tenosynovium via neovascularisation.

To measure oxygen tension (pO₂) within the synovium of rheumatoid tendons and joints in vivo. To study markers of hypoxia and neovascularisation in invasive and non-invasive tendon synovium by immunohistochemistry. To investigate the effects of hypoxia on in vitro cultures of invasive and non-invasive tenosynovium.

Patients undergoing elective hand surgery for RA were recruited into the study. In vivo oxygen tension measurements were taken intra-operatively using an established microelectrode technique. Readings were taken from tendon and joint synovium. Diseased tissue was harvested from areas of oxygen sampling for immunohistochemical analysis of CD31 expression.

Tissue was also harvested from invasive and non-invasive tenosynovium. Joint tissue was used as a control. Serial sections of tissue were stained for the pro-angiogenic factor vascular endothelial growth factor (VEGF) and the hypoxia-regulated transcription factor hypoxia inducible factor (HIF)-2α.

Separate tissue from these three areas of interest was enzymatically digested and cultured in hypoxic (1% oxygen) and normoxic (21% oxygen) conditions. Supernatants were harvested at 24 hours and analysed for expression of key inflammatory cytokines by ELISA.

We observed profound hypoxia in the synovium of RA tendons and joints in vivo despite immunohistochemical evidence of markedly increased vascularity, measured as expression of CD31. Immunohistochemical analysis revealed twofold greater VEGF expression in tenosynovium (P < 0.05), compared with joint synovium from the same patient. Levels of HIF-2α were found to be similar in tendon and joint synovium.

One of the consequences of synovial hypoxia could be the modulation of inflammatory and angiogenic cytokines. We found that hypoxia upregulated in vitro expression of pro-angiogenic cytokine VEGF in both joint and tenosynovial cultures (P < 0.05) by an average of 128%, and in parallel increased levels of the pro-inflammatory cytokine tumour necrosis factor alpha by an average of 105%. In contrast, hypoxia downregulated expression of monocyte chemoattractant protein-1 (P < 0.05) by an average of 44%.

Hypoxia is a feature of rheumatoid tenosynovitis. Key cytokines are regulated by hypoxia in this disease process. HIF-2α may represent the link between hypoxia and VEGF-driven angiogenesis in rheumatoid tendon disease.