Background In inflammatory bowel disease (IBD), a new biological therapy has recently been approved. Vedolizumab is a humanised IgG1 monoclonal antibody to α4β7 integrin that modulates gut lymphocyte trafficking. Although an exclusively local effect of vedolizumab could be expected based on the restricted presence of the α4β7–mucosal vascular addressin cell adhesion molecule 1 complex in the gut, past combined success with anti-tumour necrosis factor, and previous demonstration of α4β7 integrin in the joint, led to the expectation of a therapeutic efficacy in spondyloarthritis. Nonetheless, the effect of vedolizumab on extraintestinal manifestations—and especially the joint—has not been reported so far.
Case report A series of five patients with IBD who were treated with vedolizumab and promptly developed new onset or exacerbation of sacroiliitis or arthritis are reported.
Conclusions Vedolizumab therapy does not seem to show any efficacy in and might even induce arthritis and/or sacroiliitis. However, larger cohort studies are needed to provide information on the prevalence, the evolution and underlying mechanism.
- Magnetic Resonance Imaging
Statistics from Altmetric.com
Spondyloarthritis (SpA) and inflammatory bowel disease (IBD) have clinical, imaging and genetic evidence supporting overlap in their pathogenesis, which is best reflected in the shared efficacy of anti-tumour necrosis factor (TNF). In IBD, a new humanised IgG1 monoclonal antibody to α4β7 has been approved, which modulates gut lymphocyte trafficking. In clinical trials, vedolizumab induced a clinical response in Crohn's disease (CD) and ulcerative colitis (UC) in respectively one-third and up to half of treated patients.1 ,2 Although an exclusively local effect could be expected based on the restricted presence of the α4β7–mucosal vascular addressin cell adhesion molecule 1 (MadCAM) complex in the gut,3 ,4 combined success with anti-TNF, and previous demonstration of α4β7 in the joint,5–9 led to an anticipated efficacy in SpA. Nonetheless, the effect on extraintestinal manifestations—and especially the joint—has not been reported.
Materials and methods
Information on the recruitment of patients and requirements for inclusion can be found in the online supplementary file.
A 50-year-old woman had been diagnosed with CD and psoriasis at the age of 39. Two months after initiation of vedolizumab, she reported progressive inflammatory back pain with 30 min morning stiffness, which responded well to non-steroidal anti-inflammatory drugs (NSAIDs). MRI of the sacroiliac (SI) joints showed severe bilateral sacroiliitis. At follow-up after 6 months, the patient showed decrease of Bone Marrow Edema (BME) with etoricoxib 90 mg daily (figure 1).
A 28-year-old woman diagnosed with UC at the age of 16 was referred for intermittent pain in the lower limbs after treatment with vedolizumab. She had not been diagnosed with SpA, nor did she present with other SpA features. The patient presented with a painful left shoulder and arthritis of the right wrist. On ultrasound, intercarpal effusion, synovial hyperproliferation and power Doppler (PD) of 1–2 was present, in addition to severe PD signal in the tendon sheath of the flexor pollicis longus and flexor carpi radialis muscle.
A 30-year-old man had been diagnosed with ankylosing spondylitis (AS) and CD at the age of 25. He had been treated with golimumab and infliximab until august 2015 and was in clinical remission. Due to secondary inefficacy at the level of the gut, the patient was switched to vedolizumab. Four weeks later, he presented with arthralgia, back pain, pain at night and 90 min morning stiffness with elevated C reactive protein. Furthermore, MRI unveiled acute inflammatory lesions of the axial skeleton (figure 2).
A 47-year-old woman, diagnosed with CD at the age of 42, consulted for intense low back pain following treatment with vedolizumab. The pain was situated at the right side, worse in the early morning and resulted in awakening at night. She displayed no other features of SpA, nor had she been diagnosed with SpA in the past. MRI of the SI joints displayed unilateral sacroiliitis of the right SI joint.
A 26-year-old woman was diagnosed with UC at the age of 18. She presented with polyarticular joint pain after initiation of vedolizumab, predominantly in the lower limbs with moderate response to low-dose NSAIDs. At clinical examination, a painful, warm left wrist and right elbow, and synovitis of several metatarsophalangeal joints of the right foot were detected, in combination with tenderness at the Achilles tendon enthesis of the left heel. Blood analysis confirmed an inflammatory aetiology. Treatment strategies and response to treatment for all cases are shown in table 1.
This is the first report on new onset or exacerbation of arthritis/sacroiliitis in vedolizumab-treated patients. Despite the proven efficacy in IBD, both in anti-TNF-naive1 ,2 and in anti-TNF-exposed patients,10 α4β7 blockade seems to facilitate synovitis with similar distribution to SpA in some patients, irrespective of the response to treatment at the level of the gut. According to the European Medicines Agency, arthralgia, back pain and pain in the extremities were reported common (≥1/100) to very common (≥1/10) in clinical trials. Yet, no mention of proven arthritis or sacroiliitis was made. One can speculate that these may have been present, but were insufficiently investigated. Surprisingly, in trials with natalizumab, a non-selective inhibitor of α4, solely arthralgia was reported. Meanwhile, a large safety study could not demonstrate higher rates of arthralgia in patients treated with vedolizumab compared with placebo.11 Similarly, no mention was made of increased musculoskeletal symptoms with natalizumab in the ENCORE trial, nor in multiple sclerosis.12 ,13
Lymphocyte trafficking is facilitated by adhesion molecules, divided into integrins, selectins and the immunoglobulin superfamily. The integrins consist of an α-subunit and β-subunit, of which α4 forms a heterodimer with β1 or β7. The α4β1 serves as a ligand for vascular cell adhesion molecule-1 (VCAM-1) and is expressed on leucocytes and endothelial cells, whereas α4β7 integrin serves as a ligand for both MadCAM-1 and VCAM-1, and is expressed on a subset of lymphocytes such as CD4+ and CD8+ T cells. Indirect evidence in humans has shown that T cells with high α4β7 expression have a predilection for the gut mucosa. Similarly, MadCAM-1 is selectively expressed on the mucosal lymphoid organ high endothelial venules (HEVs) and on the gut lamina propria venules.4 ,6 ,14 In particular, vedolizumab has been shown to inhibit the interaction between α4β7 and MadCAM-1, but not VCAM-1.15
One of the many hypotheses is that integrins and adhesion molecules play a role in the interception of recirculating activated lymphocytes between the gut and the synovial membrane due to the inhibition of the α4β7 integrin homing at the level of the gut. As the preferred interaction could not take place, these activated cells could easily drift across tissues in search of a landing as the overall survival of cells was not affected.16 Even though both axial and peripheral disease manifested in our patients, irrefutable data on the presence of such adhesion molecules at the level of the spine are lacking. Nevertheless, MadCAM-1 was found to be upregulated in the HEVs of bone marrow in a small sample of patients with active axial SpA.14 Although MadCAM-1 generally appears to be restricted to the gut, α4β7 has been demonstrated in the inflamed joint with increased expression in various inflammatory conditions compared with healthy controls as a consequence of exposure to inflammatory cytokines.5–9 ,17 Alternatively, in the presence of vedolizumab, cellular recruitment may be mediated by yet to be determined adhesion molecules. This recirculation theory might explain the short mean interval of 64 days between vedolizumab initiation and the expression of symptoms. The remarkable discrepancy in joint symptoms by natalizumab compared with vedolizumab, and the human leukocyte antigen (HLA)-B27 negativity of these cases, may point towards a predominantly innate immune mechanism of joint disease. In IBD, the prevalence of HLA-B27 is comparable to healthy controls. However, in association with SpA, the prevalence is much higher.18 Notwithstanding, these interactions may be volatile and/or display low affinity, or be subject to a distinct pattern of interactions, which might explain why these features do not manifest in every patient.
Nevertheless, we cannot rule out that vedolizumab simply does not have any efficacy in SpA, which can be explained by vedolizumab solely intervening in the interaction of α4β7 with gut-specific MadCAM-1.6 The overlap between SpA and IBD manifests in up to 30% of patients with IBD, and in patients with concomitant disease, gut flares have been associated with joint flares.19 Admittedly, these joint flares may have occurred before any efficacy by vedolizumab at the level of the gut had taken place. However, the effect of vedolizumab on the gut did not seem to be related to the joint symptoms as a good response at the level of the gut did not necessarily result in a better outcome of the SpA feature over time, which demonstrates at least some level of disconnect. Although vedolizumab therapy was not discontinued in these patients due to a lack of alternative therapeutic options regarding their IBD, more than half of these patients necessitated additional chronic therapy in order to reach acceptable SpA disease activity.
Moreover, due to the prior exposure to anti-TNF, underlying SpA might have been suppressed in the past. This inefficacy theory is supported by the flare of sacroiliitis in the patient with pre-existing SpA in remission under anti-TNF therapy. Still, in the remaining four patients, the time frame in which symptoms occur, the dissociation of gut and joint response, and the absence of HLA-B27, family history or other SpA features render this lack of efficacy theory less likely as opposed to the induction of SpA features.
In conclusion, vedolizumab does not seem to show any efficacy in and might even induce arthritis and/or sacroiliitis. Larger cohort studies are needed to provide information on the prevalence, evolution and underlying mechanism.
The investigators would like to thank their patients for giving consent.
Handling editor Tore K Kvien
Contributors GV, MDV, PH, FVdB and DE: study concept and design. GV, KT, GDB, LVP, FC, HC, JDK, PC, MDV, PH, JA, IV, PS, BC, MA, FVdB and DE: data acquisition. GV, PC, FVdB and DE: analysis and interpretation of data. GV, MDV, MA, FVdB and DE: manuscript preparation. GV, KT, GDB, LVP, MDV, PH, MA, FVdB and DE: manuscript revision.
Funding This study was supported by a grant of Ghent University to GV. DE is supported by a fund of Scientific Research–Flanders (FWO) and the Research Council of Ghent University. DE is also a member of a multidisciplinary research platform (MRP) of Ghent University and is supported by Interuniversity Attraction Pole (IUAP) grant Devrepair from the Belspo Agency (project P7/07).
Competing interests DE has received grants or speakers fees from Boehringer Ingelheim, Pfizer, UCB, Merck, Novartis, Janssen and Abbvie. FVdB received consultancy and/or speaker fees from Abbvie, Celgene, Janssen, Merck, Novartis, Pfizer and UCB.
Patient consent Obtained.
Ethics approval The study protocol was approved by the Ethical Committee of Ghent University Hospital.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.