Biologic drugs for induction and maintenance of remission in Crohn's disease: a network meta-analysis.

Abstract / Summary

Crohn's disease is an inflammatory bowel disease (IBD) that can extend to any part of the gut. Many people with Crohn's disease fail to reach and maintain remission using conventional therapies. Biologic therapies with different mechanisms have advanced rapidly over recent decades. Comparative data for efficacy and safety are needed to clarify the relative position of these agents and guide decision-making. To assess the effects of biologic and advanced treatments for induction and maintenance of remission in Crohn's disease. A secondary objective was to rank the included interventions. In June 2025, we searched CENTRAL, MEDLINE, Embase, ClinicalTrials.gov and WHO ICTRP. Randomised controlled trials (RCTs) comparing biologics with any other active comparator, placebo or no treatment for induction or maintenance of remission in adults with Crohn's disease. Our critical outcome was clinical remission/relapse. Important outcomes were clinical response/loss of response, endoscopic remission/relapse, withdrawals due to adverse events, serious adverse events and total adverse events. We assessed the risk of bias using the Cochrane tool (RoB 1). We expressed outcomes as risk ratios (RR) with 95% confidence intervals (CI). We conducted network meta-analyses for a network of induction studies and a network of maintenance studies. We assessed the certainty of the evidence using GRADE. We included 94 RCTs, of which 66 were on induction (n = 20,653) and 22 on maintenance (n = 6823) of remission. Induction For clinical remission, there is high-certainty evidence that ustekinumab is more effective (RR 2.01, 95% CI 1.67 to 2.43, number needed to treat for an additional beneficial outcome (NNTB) = 6, small effect size) and moderate-certainty evidence that adalimumab combined with purine analogues (RR 2.6, 95% CI 1.71 to 3.96, NNTB = 3, moderate effect size), guselkumab (RR 2.49, 95% CI 1.95 to 3.18, NNTB = 4, moderate effect size), adalimumab (RR 2.36, 95% CI 1.75 to 3.18, NNTB = 4, moderate effect size), upadacitinib (RR 1.76, 95% CI 1.34 to 2.3, NNTB = 7, small effect size), vedolizumab (RR 1.57, 95% CI 1.15 to 2.13, NNTB = 10, trivial effect size) and natalizumab (RR 1.37, 95% CI 1.08 to 1.73, NNTB = 15, trivial effect size) are probably more effective than placebo. There is low-certainty evidence that risankizumab (RR 2.26, 95% CI 1.8 to 2.84, moderate effect), BI695501 (adalimumab biosimilar) (RR 2.09, 95% CI 1.31 to 3.33, small effect), a combination of infliximab with purine analogues (RR 2.04, 95% CI 1.14 to 3.64, small effect) and filgotinib (RR 1.55, 95% CI 1.18 to 2.04, trivial effect) may be more effective. The evidence for humicade, a combination of natalizumab with infliximab and purine analogues, CTP13 (infliximab biosimilar) with purine analogues, infliximab, mirikizumab, etrolizumab, certolizumab, tesnatilimab, onercept, andecaliximab and etanercept was of very low certainty and no conclusions can be drawn. There is moderate-certainty evidence that vedolizumab (RR 0.71, 95% CI 0.41 to 1.23), ustekinumab (RR 0.72, 95% CI 0.52 to 0.99) and upadacitinib (RR 0.8, 95% CI 0.46 to 1.39) probably lead to a similar number of withdrawals due to adverse events. There is low-certainty evidence that risankizumab may lead to fewer withdrawals due to adverse events (RR 0.36, 95% CI 0.22 to 0.59, small effect size) and low-certainty evidence that humicade (RR 0.81, 95% CI 0.39 to 1.68), natalizumab (RR 0.79, 95% CI 0.47 to 1.33), apilimod mesylate (RR 0.84, 95% CI 0.36 to 1.95), certolizumab (RR 0.98, 95% CI 0.65 to 1.49), filgotinib (RR 1.06, 95% CI 0.64 to 1.78) and tesnatilimab (RR 1.51, 95% CI 0.82 to 2.8) may lead to a similar number of withdrawals due to adverse events. Maintenance For preventing clinical relapse, there is moderate-certainty evidence that adalimumab (RR 0.69, 95% CI 0.58 to 0.82, NNTB = 2, moderate effect size) is probably more effective than placebo. There is low-certainty evidence that CTP13 (RR 0.55, 95% CI 0.43 to 0.71, large effect size), infliximab (RR 0.68, 95% CI 0.57 to 0.81, moderate effect size) and upadacitinib (RR 0.72, 95% CI 0.56 to 0.94, small effect size) may be more effective than placebo. Vedolizumab (RR 0.78, 95% CI 0.6 to 1.01), ustekinumab (RR 0.83, 95% CI0.67 to 1.03), etrolizumab (RR 0.89, 95% CI 0.67 to 1.17) and tofacitinib (RR 0.9, 95% CI 0.67 to 1.21) may be similar to placebo. The certainty of the evidence ranged between high and very low, with the main limitations being imprecision and risk of bias. There is moderate-certainty evidence that vedolizumab probably leads to a similar number of withdrawals due to adverse events during maintenance (RR 0.84, 95% CI 0.63 to 1.11). There is low-certainty evidence that tofacitinib (RR 0.78, 95% CI 0.47 to 1.31), ustekinumab (RR 0.74, 95% CI 0.4 to 1.38), etrolizumab (RR 1.42, 95% CI 0.73 to 2.77) and upadacitinib (RR 1.51, 95% CI 0.73 to 3.11) may be no different to placebo in withdrawals due to adverse events. The evidence for certolizumab, natalizumab, filgotinib, risankizumab, infliximab combined with purine analogues, and vedolizumab combined with purine analogues was of very low certainty and no conclusions can be drawn. This review has supported evidence generation, but head-to-head comparative trials for key interventional subclasses or specific agents may be needed, as guided by key stakeholders; for example, on the use of biosimilar versions of medications out of patent. The role of concomitant purine analogues is a key confounding factor and future studies may not only want to investigate specifically the role of combinations, but also consider stratifying populations or purposefully excluding participants based on this key class of therapy to avoid such heterogeneity. Given the majority of evidence focusses on the outcomes of clinical remission and relapse, future studies may want to further consider other outcomes of clear interest to clinicians and patients, particularly endoscopic remission. Finally, long-term safety data are limited throughout the networks. Whilst future studies with longer follow-ups could provide increased data, it may be that study designs outside of randomised trials are employed for such outcomes. This project is funded by an NIHR Evidence Synthesis Programme Grant (NIHR132748). A protocol for this review was published in 2017: doi.org/10.1002/14651858.CD012751.

Related Clinical Guidelines