Impact of Crohn's Disease Therapies on Histology in Randomized Controlled Trials: Systematic Review With Meta-analysis.

Abstract

The effectiveness of Crohn's disease treatments for inducing histological outcomes has not been addressed systematically. We performed a systematic review and meta-analysis of randomized controlled trials in Crohn's disease to assess the impact of therapies on mucosal histopathology. Databases (MEDLINE, CENTRAL, Web of Science, EMBASE) were searched for randomized controlled trials including adult patients and evaluating histological outcomes. Risk of bias was evaluated using the Critical Appraisal Skills Programme. Histological outcomes, pooled frequencies, pooled odds ratios, and standard mean differences of the histological scores were compared between the intervention and placebo groups using a random-effects model. Out of 2070 records, 10 studies were included. The quality of the studies ranged from moderate to high, but they were clinically and methodologically diverse. All interventions were superior to placebo. Histological response was achieved by 68% of patients, and 38% achieved remission. Pooled odds ratio for histological remission in patients receiving intervention vs placebo was 4.14 (95% CI, 2.28-7.50; I2 0%; P < .01). Heterogeneity in histological response estimates was significant, and subgroup analysis of the odds ratio results was limited by the low number of studies per group. The standard mean difference of histological scores was higher for patients receiving intervention in both induction and maintenance studies (-2.95; 95% CI, -4.17 to -1.74; I2 83% P < .00; and -2.58; 95% CI, -3.89 to -1.27; I2 56% P < .00). Crohn's disease therapies are effective for achieving histological outcomes. Adherence to recently published consensus on histopathology harmonization assessment in Crohn's disease would facilitate adequate comparison between studies in the future.