Introduction
The therapeutic armamentarium for inflammatory bowel disease (IBD) is rapidly increasing recent years with newer therapies such as vedolizumab (VDZ). Their effectiveness has already been demonstrated against placebo, but head-to-head trials compared to tumour necrosis factor alpha inhibitors (anti-TNF) are scarce.
Aims & Methods
Aim: To compare the effectiveness of VDZ versus anti-TNF in bio-naïve IBD patients in a real world setting.
Methods: A retrospective monocentric study was performed in IBD patients treated with VDZ and anti-TNF at a Belgian University Hospital, from January 2018 to June 2021 with follow-up until December 2021. Primary endpoints considering efficacy after induction and after 1 year of therapy were: clinical response (CRp) and remission (CRm), biochemical response (BRp) and remission (BRm) and endoscopic response (ERp) and remission (ERm). Secondary endpoint was treatment survival during follow-up. For assessment of risk factors influencing efficacy and treatment survival, a multivariate logistic regression and Cox regression model were applied. Propensity scores for VDZ vs. anti-TNF were weighted for age at start biological, IBD type, perianal disease activity, disease duration and previous IBD-related intestinal resection.
Results
In total, 42 patients received VDZ and 57 anti-TNF. At baseline, the VDZ group had an older median age (38 vs. 31 years, P=0.024), longer disease duration (P=0.017) and more moderate to severe comorbidities (21.4% (9/42) vs. 0% (0/57), P<0.001). In the anti-TNF group, percentage of CD patients was higher (77.2% (44/57) vs. 57.1% (24/42),P=0.034), patients had more upper GI, penetrating or stricturing disease (P=0.013 for location and P=0.004 for behavior), perianal disease (P=0.034) and used more immunomodulators (IMM) during treatment (52.6% (30/57) vs. 16.7% (7/42), P<0.001). After treatment induction, CRp (P=0.311) and CRm (P=0.216) rate was not significantly different. The anti-TNF group had a significantly higher BRp rate compared to the VDZ group (87.8% (43/49) vs. (23/34), P=0.026). After adjusting for propensity score, age, IBD type, systemic corticosteroid (CS) or IMM use, the higher rate of BRp remained (aOR 3.83 [1.01-14.60], P=0.049). Similar results were found for BRm (87.8% (43/49) vs. 67.6% (23/34), P=0.026). After adjustment for the same confounders, the higher rate of BRm in the anti-TNF group remained (aOR 3.83 [1.01-14.60], P=0.049). No significant difference was seen for ERp (P=1.000) and ERm (P=0.338). After 1 year, univariate analysis showed no significant difference for CRp (P=0.736) and CRm (P=0.404), nor for BRp (P=0.907) and BRm (0.890) or for ERp (P=0.163) and ERm (P=0.533). Treatment survival using Cox regression model with adjustment for propensity and IMM or CS use, showed no significant longer treatment duration for anti-TNF compared to VDZ (aHR 0.947 [0.36-2.49], P=0.913).
Conclusion
In this tertiary real-world cohort of bio-naïve patients with IBD, anti-TNF and VDZ showed a similar efficacy, with an advantage of anti-TNF regarding onset of action.
