Three-year Effectiveness and Safety of Vedolizumab Therapy for Inflammatory Bowel Disease

A Prospective Multi-centre Cohort Study

Aurelien Amiot; Melanie Serrero; Laurent Peyrin-Biroulet; Jerome Filippi; Benjamin Pariente; Xavier Roblin; Anthony Buisson; Carmen Stefanescu; Caroline Trang-Poisson; Romain Altwegg; Philippe Marteau; Thibaud Vaysse; Anne Bourrier; Stephane Nancey; David Laharie; Matthieu Allez; Guillaume Savoye; Jacques Moreau; Lucine Vuitton; Stephanie Viennot; Guillaume Bouguen; Vered Abitbol; Mathurin Fumery; Charlotte Gagniere; Yoram Bouhnik; on the behalf of the OBSERV-IBD study group, the GETAID


Aliment Pharmacol Ther. 2019;50(1):40-53. 

In This Article


In the present study, we provided long-term data beyond 1 year, from 294 consecutive patients with IBD treated with vedolizumab in a real-world setting for the first time. We previously published the effectiveness and safety results of the OBSERV-IBD cohort from week 0 to week 54, showing that approximately one-third of UC and CD patients achieved steroid-free clinical remission.[10,12] After 3 years of follow-up, more than one-third of patients with UC still presented with steroid-free clinical remission, indicating a stable effect of vedolizumab between 1 and 3 years of follow-up, whereas the rate of patients with CD who presented with steroid-free clinical remission tended to decrease over time. The safety profile was consistent with previously published data with no new alarm signal, especially for neoplastic diseases.

Vedolizumab has already demonstrated significant differences with anti-TNF agents in its mode of action showing a slower induction of clinical and biological response.[21] In patients treated with an anti-TNF agent, it has been shown that approximately 10%-20% of patients per year lose response to the drug.[22,23] Although long-term data coming from open-label extension studies of the GEMINI trials suggested a lower rate of loss of response than with anti-TNF agents, a recent meta-analysis mostly including studies limited to a 1-year follow-up suggested a higher rate of loss of response.[6,7,24] In the present study, patients with UC disclosed stable results from week 54 to week 162, accounting for one-third of patients in steroid-free clinical remission (40.5% and 36.1%, respectively, P = 0.40). It was not the case for patients with CD in whom the overall rate of steroid-free clinical remission decreased from 27.2% to 19.9% between week 54 and week 162 (P = 0.02). Although such comparison may not be valid dealing with different patient and IBD characteristics, those data are in line with previous data coming from the GEMINI trials suggesting better efficacy of vedolizumab in patients with UC than in patients with CD and those from the ENEIDA registry showing higher discontinuation rate in patients with CD compared to those with UC.[3–5,25] The mode of action of vedolizumab, the transmural nature of CD and heterogeneity in the plasticity of inflammatory processes may explain this difference between UC and CD and should be studied further.[26] Pathogenic and phenotypic differences between UC and CD may also explain differences in the efficacy profile of vedolizumab between CD and UC.

Vedolizumab discontinuation was observed in 80.1% of patients with CD and 61.3% of patients with UC. The majority of vedolizumab discontinuation in our cohort was related to an inadequate response or secondary loss of response. Predictors of vedolizumab discontinuation were related to the severity of the disease in patients with UC (partial Mayo Clinic score >6 and UCEIS >5), whereas it was related to age at vedolizumab induction (>35 years and disease phenotype (history of perianal disease) in patients with CD. The intensity of the systemic inflammatory burden, which was a predictor of short-term effectiveness in our cohort, was no longer significant for long-term treatment. No benefits of concomitant immunomodulators were noted in our cohort. The discontinuation rate of vedolizumab seems to be higher than those with anti-TNF agents in real-world cohorts.[27] However, such comparison may not be valid when comparing biologic-naïve patients with highly refractory patients with previous failure of immunomodulator and anti-TNF agent. It is also conceivable that some patients discontinued vedolizumab overtime due to the advent of other biological agents such as ustekinumab in patients with CD and tofacitinib in patients with UC.

The safety profile of vedolizumab was good and consistent with the results of the GEMINI trials and real-world cohort studies. Almost one half of patients reported an adverse event without any difference between patients with CD and UC. The more frequent adverse events were related to incidental infection in 51.7% of the cases. Infections were mainly related to upper respiratory and nasal mucosal infections and gastrointestinal infections. Adverse events related to IBD relapse occurred in 26.4% of the cases, and even more when considering severe adverse events in 62.5% of cases. It has been suggested that vedolizumab may impair antitumoural immunosurveillance, especially in the intestinal mucosa. In the present study, only one case of incidental cancer was noticed in a 58.3-year-old patient with localised prostate adenocarcinoma in whom vedolizumab therapy was continued concomitant with oncologic management. The occurrence of headache and paraesthesia was less frequent than during the first year of treatment, whereas the occurrence of inflammatory arthralgia/arthritis was more frequent, accounting for 4.7% of patients still treated with vedolizumab after week 54. Headache (n = 16) and paraesthesia (n = 13) were common adverse events. Nine out of 13 patients who experienced paraesthesia after vedolizumab was started underwent brain MRI and electromyography analysis, which did not reveal any abnormalities. Paraesthesia tends to disappear over time in almost all patients. Twelve patients experienced IBD exacerbation.

The OBSERV-IBD cohort is one the largest cohorts of IBD patients who had an inadequate response or intolerance to anti-TNF therapy and who were treated with vedolizumab in a real-life setting.[10,12] All patients treated in France between June 2014 and December 2014 were included and followed up through week 162 with systematic collection of clinical data including HBI and partial Mayo Clinic score and biological indicators allowing measurement of IBD activity. The strength of the OBSERV-IBD cohort relies on the possibility of assessing effectiveness and safety of vedolizumab therapy in patients with IBD independently of exclusion criteria of clinical trials and without excluding early nonresponders in the analysis.[3–5] However, the limitations of our study are the absence of a control group and the absence of a standardised endoscopic assessment. Indeed, the assessment of mucosal healing was performed in 43.6% and 84.5% of patients with CD and UC respectively. Endoscopy was also scheduled at the investigator's discretion, either for patients with inadequate response or for patients with steroid-free clinical remission. Finally, safety data were not collected during the whole study period in patients who discontinued vedolizumab therapy, which could have biased the safety analysis.