P139. Different scoring methods affect the accuracy of clinical outcomes in ulcerative colitis. Post-hoc analysis in a phase 2a study with interferon-β-1a
P. Mannon1, W. Reinisch2, P. Miner3, A. McAllister4, Y. Xi4, F. Cataldi4
1UAB, Birmingham, AL, United States; 2Vienna University, Vienna, Austria; 3OFDR, Oklahoma City, OK, United States; 4Biogen, Cambridge, MA, United States
Aim: Placebo response rates in ulcerative colitis trials often make it difficult to demonstrate efficacy of new drugs. Contributing factors includes duration of the trial and geographical location. Furthermore, trials in UC have used a variety of non-validated clinical scores. Based on encouraging results from an NIH open-label study of interferon-β-1a (IFN) we conducted a placebo-controlled study in North America and East Europe. We collected clinical data using both the Mayo score and the Simple Clinical Colitis Activity Index (SCCAI) (used in the NIH study) in order to benchmark results to the pilot study data as well as to an emerging standard in UC trials. We conducted a post-hoc analysis comparing the common domains of the two scores: stool frequency (SF) and rectal bleeding (RB).
Methods:123 subjects with moderate to severe UC were randomized to receive IFN 30 μg IM or placebo (P) twice at WK, 62 IFN and 61 P, for 12 weeks. The primary endpoint was clinical response at Week 8, defined as a decrease from baseline in the total Mayo score. The secondary endpoint was the % of subjects with a decrease in the SCCAI score at Week 8. The total Mayo score collects: SF, RB, Physician Global Assessment and Endoscopy; the SCCAI collects: SF for day and night; urgency; RB; general well being and extra intestinal manifestations. The data collection methodology is also different between scores: the Mayo score counts numbers of stools “more than normal” whereas the SCCAI measures absolute numbers of daily stools. Lastly, the Mayo averages the score of the subject self-assessment collected over a period of 7 days vs. the prior 24-hour only of the SCCAI.
Results:At week 8, 53% of patients on IFN vs. 44% on P (P = 0.35) were in clinical response based on the total Mayo score and 64% on IFN vs. 46% on P (P = 0.05) based on the SCCAI. There was virtually no discordance between the RB responses measured by either score. However, there was a clear discordance for the self-reported SF between scales at baseline and furthermore in the efficacy outcome at week 8 (Figure). In fact, there was a 7 point difference favoring P over IFN when SF efficacy was measured by the Mayo score. The opposite was observed when SF was measured by the SSCAI with 31 points difference favoring IFN over P.
Conclusions: IFN showed significant efficacy on the secondary endpoint measured by the SCCAI in the induction of clinical response at week 8. However, the primary endpoint measured by the Mayo score was not significant. There was discordance in SF between the SCCAI and the Mayo scores that likely contributed to the difference in results of efficacy estimates.