P611 Baseline faecal MMP-9 level correlates to endoscopic sub-score and post-treatment level correlates to mucosal healing in moderate-to-severe ulcerative colitis patients
B. Feagan1, B. Levesque1, J. Woo*2, G. Chen3, H. Wang3, Y. Xiao3, E. Huntzicker2, D. French2, V. Smith4, S. D. Patterson2, M. Subramanian5, B. Kanwar5, W. Sandborn6, B.R. Bhandari7
1Robarts Clinical Trials, London, Ontario, Canada, 2Gilead Sciences, Biomarker, Foster City, California, United States, 3Gilead Sciences, Biostatistics, Foster City, California, United States, 4Gilead Sciences, Biologics and Target Biology, Foster City, California, United States, 5Gilead Sciences, Liver Disease Therapeutic Area, Foster City, California, United States, 6University of California San Diego, Gastroenterology, San Diego, California, United States, 7Gastroenterology and Nutritional Medical Services, Bastrop, Louisiana, United States
Ulcerative colitis (UC) is a chronic disease driven by colonic accumulation of unregulated inflammatory lymphoid and myeloid cells. Infiltrating granulocytes possess cytoplasmic matrix-metalloprotease-9 (MMP-9). Upon activation by local proteases, released pro-MMP-9 is converted to active MMP-9, leading to degradation of local collagen, mucosal tissue injury, and leakage of cells and inflammatory products. Faecal MMP-9 has been identified as a new, noninvasive marker for UC (Annahazi A et al. 2013). In this study, we re-examined the association between dry weight faecal MMP-9 (ng/g) to colon MMP-9 expression, faecal inflammatory markers, Mayo score, and endoscopic sub-score.
Fifty patients with moderately to severely active UC (Mayo score ≥ 6, and endoscopic sub-score ≥ 2) were enrolled into a Phase I study of GS-5745 (a humanised anti-MMP9 mAb). Baseline faecal MMP-9 and faecal inflammatory markers (NGAL, lactoferrin, and calprotectin) were measured by ELISA, and colon MMP-9 expression were measured by immunohistochemistry. Correlations between biomarkers were assessed using the Spearman’s rank correlation coefficients, and bootstrap confidence intervals for the correlation coefficients were computed using 1000 bootstrap samples. Difference of baseline faecal MMP-9 level between patient groups was tested using the Wilcoxon rank-sum test with no multiple testing correction.
Baseline faecal MMP-9 levels showed high correlation with faecal calprotectin (r = 0.626; 95% CI 0.394–0.777) and moderate correlation with colon MMP-9 tissue expression (r = 0.433; 95% CI 0.149–0.645). When stratified by Mayo score, patients with a baseline Mayo score 9–11 had higher median faecal MMP-9 levels compared with patients with Mayo score of 6–8 (1 805 ng/g, n = 14 and 607 ng/g, n = 36 respectively; p = 0.06). Patients with a baseline endoscopic sub-score of 3 had significantly higher level than patients with a sub-score of 2 (median value at 1 506 ng/g, n = 17 and 681 ng/g, n = 33, respectively; p = 0.04). When faecal MMP-9 levels were examined in patients treated with GS-5745, those who achieved mucosal healing at day 36 (endoscopic sub-score of 0–1) had significantly lower levels of median faecal MMP-9 compared with patients without mucosal healing (endoscopic score of 2–3; 258 ng/g and 842 ng/g, respectively, p = 0.01; Table 1).
Results from this study demonstrate that faecal MMP-9 levels normalised to dry weight are associated with endoscopic sub-scores, and faecal MMP-9 has potential to be used as a surrogate marker for mucosal healing.
Table 1 Mean and median faecal biomarker values of UC subjects treated with GS-5745