Background and aims Extra-intestinal manifestations are frequently reported in inflammatory bowel diseases. However, data comparing the effect of vedolizumab and ustekinumab on articular extra-intestinal manifestations are limited. The aim was to evaluate differences in new onset and evolution of pre-existing joint extra-intestinal manifestations during both treatments. Methods An international multicentric retrospective study was performed on inflammatory bowel disease patients who started vedolizumab or ustekinumab between May 2010 and December 2020. Extra-intestinal manifestations were assessed at baseline and joint extraintestinal manifestations were evaluated throughout the 2-year follow-up. Arthropathy was defined by joint inflammation (arthritis/sacroiliitis), diagnosed by a rheumatologist, and arthralgia as articular pain without confirmed inflammation. Additionally, skin, ocular and hepatic extra-intestinal manifestations were assessed at baseline. Uni- and multivariate analyses were performed. Results In total 911 patients (vedolizumab:584; ustekinumab:327) were included. Deterioration of pre-existing arthropathy and rate of new onset arthropathy were not significantly associated with vedolizumab over ustekinumab. Arthropathy was reason to stop treatment in 6 vedolizumab and 2 ustekinumab patients. The odds of developing new arthralgia within 6 months was higher in patients who took vedolizumab compared to ustekinumab (aOR: 2.28 [1.01-5.15], p=0.047). However, this effect was not sustained during the 2-year follow-up (aOR: 1.35 [0.80-2.29], p=0.259). Deterioration of pre-existing arthralgia was comparable between ustekinumab and vedolizumab treated patients. In 2 vedolizumab-treated patients arthralgia was reason to stop treatment. Conclusions Vedolizumab and ustekinumab can be used safely in patients with articular extra-intestinal manifestations. Only a temporary increased risk for developing arthralgia has been observed under vedolizumab.

Environmental hypoxia and hypoxia-induced signalling in the gut influence inflammatory bowel disease pathogenesis, however data is limited to colitis. Hence, we investigated the effect of environmental hypoxia and immune cell-specific deletion of oxygen sensor prolyl hydroxylase (PHD) 1 in a Crohn’s like ileitis mouse model. Therefore, 5-week-old C57/BL6 TNF∆ARE/+ mice and wildtype (WT) littermates were housed in normoxia (21% O2) or hypoxia (8% O2) for 10 weeks. Systemic inflammation was assessed by haematology. Distal ileal hypoxia was evaluated by pimonidazole staining. The ileitis degree was scored on histology, characterized via qPCR and validated in haematopoietic Phd1-deficient TNF∆ARE/+ mice. Our results demonstrated that hypoxia did not impact body weight evolution in WT and TNF∆ARE/+ mice. Hypoxia increased red blood cell count, haemoglobin, haematocrit and increased pimonidazole intensity in the ileum. Interestingly, hypoxia evoked an increase in circulatory monocytes, ileal mononuclear phagocytes and proinflammatory cytokine expression in WT mice. Despite these alterations, no histological or ileal gene expression differences could be identified between TNF∆ARE/+ mice housed in hypoxia versus normoxia nor between haematopoietic Phd1-deficient TNF∆ARE/+ and their WT counterparts. Therefore, we demonstrated for the first time that long-term environmental hypoxia or haematopoietic Phd1-deletion does not impact experimental ileitis development.