Harms with placebo in trials of biological therapies and small molecules as induction therapy in inflammatory bowel disease: a systematic review and meta-analysis.

Background: Randomised placebo-controlled trials are the gold standard to assess novel drugs in ulcerative colitis and Crohn's disease. However, there might be risks associated with receiving placebo. We aimed to examine the harms associated with receiving placebo in trials of licensed biologics and small molecules for the induction of remission in ulcerative colitis and luminal Crohn's disease in a meta-analysis.

Methods: We performed a systematic review and meta-analysis. We searched MEDLINE, Embase, Embase Classic, and the Cochrane Central Register of Controlled Trials from database inception to May 30, 2024, for randomised placebo-controlled trials of licensed biologics and small molecules for induction of remission in adults (≥18 years) with moderately to severely active ulcerative colitis or luminal Crohn's disease reporting data on adverse events over a minimum treatment period of 4 weeks. There were no prespecified study exclusion criteria. We extracted summary data and pooled data using a random-effects model for any treatment-emergent adverse event, any drug-related adverse event, infection, worsening of inflammatory bowel disease (IBD) activity, withdrawal due to adverse events, serious adverse events, serious infection, serious worsening of IBD activity, or venous thromboembolic events (VTEs), reporting relative risks (RRs) with 95% CIs. The protocol for this meta-analysis was registered with PROSPERO (CRD42024527341).

Findings: The search identified 10 826 citations, of which 47 trials including 20 987 patients (14 267 [68·0%] receiving active drug and 6720 [32·0%] receiving placebo) were eligible. The risk of any treatment-emergent adverse event was no different with active drug than with placebo (7660/14 267 [53·7%] patients on active drug vs 3758/6720 [55·9%] on placebo; RR 0·97, 95% CI 0·94-1·00; I =36%). However, the risks of worsening of IBD activity (563/13 473 [4·2%] vs 530/6252 [8·5%]; 0·48, 0·40-0·59; I =54%), withdrawal due to adverse event (401/13 363 [3·0%] vs 299/6267 [4·8%]; 0·62, 0·48-0·79; I =46%), serious adverse event (682/14 267 [4·8%] vs 483/6720 [7·2%]; 0·69, 0·59-0·80; I =30%), serious infection (140/14 194 [1·0%] vs 91/6647 [1·4%]; 0·67, 0·50-0·89; I =0%), serious worsening of IBD activity (187/11 271 [1·7%] vs 189/5056 [3·7%]; 0·45, 0·34-0·60; I =27%), or VTEs (13/7542 [0·2%] vs 12/2981 [0·4%]; 0·45, 0·21-0·94; I =0%) were all significantly lower with active drug than placebo. Numbers needed to treat with active drug to avoid these potentially serious adverse events ranged from 23 for worsening of IBD activity to 452 for VTEs. 27 randomised controlled trials were judged as low risk of bias across all domains.

Interpretation: Patients with moderately to severely active IBD receiving placebo are more likely to experience significant worsening of IBD activity and some serious adverse events, which might relate to a reduction in risk of these events with active drug. Patients should be counselled about these potential harms, and alternative trial designs to mitigate these harms should be considered.

Funding: None.