An intriguing link exists between gut and joint inflammation in spondyloarthritis (SpA). About 50% of patients has subclinical (eg, microscopic) gut inflammation, which represents a risk factor for development of Crohn’s disease, sacroiliac inflammation and evolution in to ankylosing spondylitis. However, the underlying mechanisms are still relatively poorly understood. Our goal was to examine the relationship between TNF, microscopic gut inflammation and axial inflammation.

Therefore, we examined in situ expression of TNF, TNFR1 and TNFR2 using triple in situ hybridisation in gut biopsies of human SpA patients and found marked upregulation of TNF in inflamed versus non-inflamed biopsies. We also noted a predominant upregulation of TNFR1 on intestinal epithelium and TNFR2 in lamina propria respectively. Of interest, IL-17 and IL-23 were also markedly increased while IL-22 was most abundant in chronically inflamed samples. In line with this, we found that patients with gut inflammation had a higher need for anti-TNF therapy and their degree of clinical response after anti-TNF was also markedly higher.

We speculated that TNF in the gut represents an important risk factor for disease severity and progression in SpA. To investigate this further we generated intestinal specific human TNF transgenic mice, in which hTNF is under control of a rat iFABP (fatty acid binding protein) promoter, generating a mouse-model over-expressing human TNF in the ileum. These mice, together with wild type littermates, were evaluated for the development of arthritis up until the age of 13 weeks after which they were euthanized and ankle and sacroiliac joints as well as ileum were processed for histology. Transgenic mice exhibit a runt phenotype and hallmarks of inflammatory bowel disease, including increased intestinal permeability and inflammation compared to their wild-type littermates. While in peripheral joints no clear signs of arthritis were observed, the sacroiliac joints in transgenic mice, by contrast, showed marked signs of inflammation as well as bone erosion and destruction. These data propose a new paradigm that gut-derived TNF is sufficient to trigger sacroiliitis and provide an alternate explanation on the relationship between gut inflammation, evolution to inflammatory bowel disease and axial inflammation in SpA.