Response to Placebo, Measured by Endoscopic Evaluation of Crohn’s Disease Activity, in a Pooled Analysis of Data From 5 Randomized Controlled Induction Trials

Background & Aims Endoscopy is used to measure activity of Crohn’s disease (CD) and determine eligibility and outcomes of participants in randomized controlled trials of therapeutic agents. We aimed to estimate the rate of response to placebo in trials, based on endoscopic evaluation of CD activity, and identify factors that affect this response. Methods We collected patient-level data from randomized, double-blind, placebo-controlled trials of therapeutic agents for CD that included centrally-read endoscopic assessments with validated scoring indices. We analyzed data from induction trials of eldelumab, filgotinib, risankizumab, and ustekinumab (from 188 patients given placebo). The primary outcome was the rate of response to placebo, based on endoscopic assessment of CD activity (>50% reduction in the simple endoscopic score for CD). Rate of remission, based on endoscopic score, was a secondary outcome. Overall rates of response to placebo were calculated using the inverse variance-weighted average method and presented with 95% CIs. We performed a multi-variable meta-regression analysis to identify determinants of response to placebo, assessed endoscopically, using patient-level data from the filgotinib and ustekinumab trials. Results The pooled rate of response among patients given placebo was 16.2% (95% CI, 10.5%–22.0%) and the rate of remission in this group was 5.2% (95% CI, 1.7%–8.8%). Prior exposure to tumor necrosis factor antagonists (odds ratio, 0.31; 95% CI, 0.10–0.93; P = .036) and increased concentration of C-reactive protein at baseline (odds ratio, 0.93; 95% CI, 0.87–0.98; P = .014 per 10 mg/L increase) were independently associated with lower rates of response to placebo. Conclusions Rates of response and remission to placebo, determined by centrally-read endoscopy, in induction trials of therapies for CD are low. These estimates are important for sample size calculations for randomized placebo-controlled trials that use the Simple Endoscopic Score for CD as an endpoint. They also provide a benchmark to interpret findings from non-placebo controlled, prospective, randomized, unblinded trials. Endoscopy is used to measure activity of Crohn’s disease (CD) and determine eligibility and outcomes of participants in randomized controlled trials of therapeutic agents. We aimed to estimate the rate of response to placebo in trials, based on endoscopic evaluation of CD activity, and identify factors that affect this response. We collected patient-level data from randomized, double-blind, placebo-controlled trials of therapeutic agents for CD that included centrally-read endoscopic assessments with validated scoring indices. We analyzed data from induction trials of eldelumab, filgotinib, risankizumab, and ustekinumab (from 188 patients given placebo). The primary outcome was the rate of response to placebo, based on endoscopic assessment of CD activity (>50% reduction in the simple endoscopic score for CD). Rate of remission, based on endoscopic score, was a secondary outcome. Overall rates of response to placebo were calculated using the inverse variance-weighted average method and presented with 95% CIs. We performed a multi-variable meta-regression analysis to identify determinants of response to placebo, assessed endoscopically, using patient-level data from the filgotinib and ustekinumab trials. The pooled rate of response among patients given placebo was 16.2% (95% CI, 10.5%–22.0%) and the rate of remission in this group was 5.2% (95% CI, 1.7%–8.8%). Prior exposure to tumor necrosis factor antagonists (odds ratio, 0.31; 95% CI, 0.10–0.93; P = .036) and increased concentration of C-reactive protein at baseline (odds ratio, 0.93; 95% CI, 0.87–0.98; P = .014 per 10 mg/L increase) were independently associated with lower rates of response to placebo. Rates of response and remission to placebo, determined by centrally-read endoscopy, in induction trials of therapies for CD are low. These estimates are important for sample size calculations for randomized placebo-controlled trials that use the Simple Endoscopic Score for CD as an endpoint. They also provide a benchmark to interpret findings from non-placebo controlled, prospective, randomized, unblinded trials.