Network Meta-Analysis: Histologic and Histo-Endoscopic Improvement and Remission With Advanced Therapy in Ulcerative Colitis.

Abstract

Background: Histology has prognostic value in ulcerative colitis (UC). However, direct comparisons of histological endpoints are lacking.

Aim: To perform a network meta-analysis (NMA) to compare histological endpoints with biologics and small molecules.

Methods: We searched four databases up until July 2024 for randomised controlled trials (RCTs) on advanced therapies for moderate-to-severe UC reporting histological endpoints. Outcomes included histological improvement or remission, and histo-endoscopic improvement after induction or during maintenance. We used a random-effects frequentist model and have reported outcomes as relative risk and 95% confidence interval. We estimated relative drug efficacy with the P-score. We conducted subgroup analysis by trial phase and evaluated risk of bias and evidence certainty.

Results: We included 24 RCTs (15 therapies, 8874 patients). Nineteen provided data on induction and 10 on maintenance; outcome definitions were similar. Etrasimod 2 mg/day ranked highest in achieving histologic improvement (P-score 0.98) and remission (P-score 0.90) following induction. Globally, guselkumab 200-400 mg ranked first for histo-endoscopic improvement, while etrasimod 2 mg/day and upadacitinib 45 mg/day were superior in the subgroup analysis. During maintenance, upadacitinib 30 mg/day was superior in achieving histologic improvement and remission (P-score 0.88 for both) and histo-endoscopic improvement (P-score 0.94). Etrasimod 2 mg/day ranked second for histologic remission (P-score 0.70) and histo-endoscopic improvement (P-score 0.73), while mirikizumab 200 mg/month ranked second for histologic improvement.

Conclusion: These results support the ability of small molecules to achieve stringent endpoints in moderate-to-severe UC. Histological outcome data for biologics was sparser, particularly during maintenance. Head-to-head RCTs are imperative to better inform clinical practice.