Predicting response to vedolizumab in Crohn’s disease – a single centre experience

Claire Kane, Ross Porter, John Thomson, Malcolm Smith, Umesh Basavaraju, Gillian H Bain, Mairi H McLean

Research output: Contribution to journalMeeting abstractpeer-review


Introduction: There is a clinical need to develop personalised clinical biomarker scoring systems to direct biologic treatment for Crohn’s disease (CD). Vedolizumab is a humanised anti-α4β7 integrin biologic, inhibiting migration of lymphocytes to the gastrointestinal mucosa. Dulai and colleagues [Gastroenterology, 2018; 155(3)] developed a new clinical prediction tool to determine probability of response to Vedolizumab. Our aim was to retrospectively determine whether this published clinical prediction tool could identify responders to Vedolizumab in our single tertiary IBD centre.

Methods: Patients receiving Vedolizumab for CD between 2015–2018 were identified from the IBD database. The Dulai clinical score (>19 predicting response and ≤13 predicting non-response) was retrospectively calculated from baseline pre-treatment characteristics including absence of previous anti-TNF therapy (+3), no previous gastrointestinal surgery (+2), no fistulising disease (+2) along with pre-treatment CRP (mg/L) (variable) and albumin levels (g/L)(variable) as previously published. Clinical response at 26 weeks was determined based on documented improvement in clinical symptoms and no requirement for steroids. Radiological or endoscopic improvement in disease activity and reduction in faecal calprotectin were considered if data was available. Analysis was performed with SPSS V25 with Mann-Whitney U test, Fisher’s exact test and ROC analysis, with significance p<0.05.

Results: Forty-nine patients with CD received Vedolizumab. 10 were excluded from analysis as they did not reach the 26-week endpoint. Following 26 weeks of Vedolizumab treatment, 13/39 (33.3%) CD patients had evidence of response. Responders versus non-responders had a median pre-treatment predictive score of 16.3 (8.2–20.8) and 14.1 (6.6–13.5), respectively (p=0.222). ROC analysis reported area under the curve of 0.621 (95% CI 0.44–0.81) (p=0.222). A score >19 had poor sensitivity [23.08% (95% CI 50.38,53.81)] but good specificity [84.62% (95% CI 65.13, 95.64)] for predicting response to therapy. A score ≤13 had poor sensitivity [42.31% (95% CI 22.35, 63.08)] but good specificity [84.62% (95% CI 54.55, 98.08)] for predicting non-response. There was no association between predictive score >19 or ≤13 and patients who did (p=0.666) or did not (p=0.151) respond to Vedolizumab therapy, respectfully.

Conclusions: In this retrospective single centre tertiary IBD centre analysis, the previously published Dulai clinical score did not predict response to Vedolizumab in CD. Future prospective analysis and consideration of a modified score including additional parameters is warranted.
Original languageEnglish
Issue numberS2
Publication statusPublished - 1 Jun 2019
Externally publishedYes


  • Inflammatory Bowel Disease
  • Crohn's Disease
  • Biomarker

Fingerprint Dive into the research topics of 'Predicting response to vedolizumab in Crohn’s disease – a single centre experience'. Together they form a unique fingerprint.

Cite this