Heterogeneity in Definitions of Endpoints for Clinical Trials of Ulcerative Colitis

A Systematic Review for Development of a Core Outcome Set

Christopher Ma; Remo Panaccione; Richard N. Fedorak; Claire E. Parker; Tran M. Nguyen; Reena Khanna; Corey A. Siegel; Laurent Peyrin-Biroulet; Geert D'Haens; William J. Sandborn; Brian G. Feagan; Vipul Jairath


Clin Gastroenterol Hepatol. 2018;16(5):637-647. 

In This Article

Abstract and Introduction


Background & Aims Advances in development of therapeutic agents for ulcerative colitis (UC) have been paralleled by innovations in trial design. It would be useful to identify a core outcome set, to standardize outcome definitions for efficacy and safety in clinical trials. We performed a systematic review of efficacy and safety outcomes reported in placebo-controlled randomized controlled trials of patients with UC.

Methods We searched MEDLINE, EMBASE, and the Cochrane Library from inception through March 1, 2017, for placebo-controlled randomized controlled trials of adult patients with UC treated with aminosalicylates, immunosuppressants, corticosteroids, biologics, and oral small molecules. We collected information on efficacy and safety outcomes, definitions, and measurement tools, stratified by decade of publication.

Results We analyzed data from 83 randomized controlled trials (68 induction and 15 maintenance) comprising 17,737 patients. Clinical or composite-clinical efficacy outcomes were reported in all trials; the UC Disease Activity Index and Mayo Clinic Score were frequently used to determine clinical response or remission. We found substantial variation in definitions of clinical or composite-clinical endpoints, with more than 50 definitions of response or remission. Endoscopic factors, histologic features, and fecal or serum biomarkers were used to determine outcomes in 83.1% (69 of 83), 24.1% (20 of 83), and 24.1% (20 of 83) of trials, respectively. A greater proportion of trials published after 2007 reported objective outcomes (96.5% endoscopic, 26.3% histologic, and 36.8% biomarker outcomes), but no standardized definitions of histologic or biomarker endpoints were found. Patient-reported efficacy and quality-of-life outcomes were described in 25 trials (30.1%) and safety outcomes were reported in 77 trials (92.8%).

Conclusion In a systematic review, we found that despite recent advances in clinical trials methods, there is a great deal of variation in definitions of endpoints, including response and remission, in randomized controlled trials of patients with UC. Researchers should identify a core set of outcomes to standardize efficacy and safety reporting in UC clinical trials.


Approval of new therapies for ulcerative colitis (UC) has been informed by data from randomized controlled trials (RCTs). Since the first placebo-controlled RCT in UC over 60 years ago,[1] clinical trials have become larger and more sophisticated in their design. In recent years, treatment targets in UC have evolved, from symptom-based scoring assessments to normalization of objective measures of inflammation such as endoscopic appearance, biomarkers, and histology.[2] However, the U.S. Food and Drug Administration has continued to require inclusion of patient-reported outcomes (PROs) as clinical trial endpoints to accurately capture the patient's experience.[3] Assessment of safety endpoints has also evolved. As novel therapies are designed to target specific components of the immune response, treatment-specific short- and long-term adverse events (AEs) have been recognized, including the risk of opportunistic infections, malignancies, infusion or injection reactions, and the formation of antidrug antibodies.

Given the progress that has occurred in this field, it is surprising that no formalized consensus has been established regarding what to measure, how to measure, or when to measure efficacy and safety endpoints in UC trials. In response to this unmet need, we have proposed the collaborative development of a core outcome set (COS) for use in UC RCTs.[4] A COS is a consensus-derived minimum set of outcomes that should be measured and reported in all clinical trials of a given disease.[5] COS implementation reduces heterogeneity in outcome reporting and enhances quality of evidence synthesis in systematic reviews and meta-analyses.

Selection of outcomes for development of a COS is a multistep process, starting with a comprehensive review of existing endpoint definitions and measurement tools.[4,5] Accordingly, we systematically review the efficacy and safety outcomes reported in placebo-controlled UC trials. Specifically, we aim to explore the evolution of endpoint reporting in UC trials over time and characterize the key outcome domains of a conceptual framework for COS development.