Conduct of Clinical Trials in UC

Purpose: The impact of interobserver variation in endoscopy on trial outcome has never been explored. A clinical study of a peptide vs mesalazine provides evidence of the influence of the variability of endoscopic scoring on clinical outcome. Methods: To investigate the efficacy of a peptide vs mesalazine, 335 patients with moderately active UC were randomised in 2 independent cohorts to receive peptide 200mg or 600mg daily + mesalazine 2.4g/day, or mesalazine 4.8g/day for 8 weeks. Moderate UC was defined as a Mayo score 7–11, with a sigmoidoscopy score (SS)≥2. Videoendoscopy, histology and clinical data were collected at baseline and 8 weeks. To minimize variability, investigators and an independent blinded observer were trained to score videosigmoidoscopy to prespecified standards. Three definitions of remission were used: clinical (stool frequency (SF) subscore 0 & rectal bleeding subscore (RB) 0); complete (SF 0, RB 0 & SS≤1) and registration (RB 0 & SS≤1). Results: The blinded observer disagreed with investigator scoring for endoscopic disease severity at baseline by 12–23% across both cohorts. The impact of the “observer” on the treatment effect of peptide vs 5ASA control for clinical, complete and registration remission rates within each cohort was a median difference of 19% (range 10 to 22.4%). Results were more variable for registration than clinical or complete remission.Table: Remission (registration)-Central Reader vs Investigator.Conclusions: Observer assessment of endoscopic activity for UC needs to be validated and techniques standardized. It is possible to conduct clinical trials with an independent observer. Remission rates defined by regulatory authorities for registration trials are highly dependent on the observer. The quality and robustness of UC trials will benefit from the inclusion of an independent observer and internationally agreed definitions of activity and remission. Research was funded by P&G Pharmaceuticals, Mason, OH.