P382 A panel of serum markers for early detection of endoscopic healing with infliximab in patients with ulcerative colitis

Abstract

Background: The need for surrogate markers to detect endoscopic healing in inflammatory bowel disease (IBD) is imminent. Previously, neutrophil gelatinase B-associated lipocalin and matrix metalloproteinase-9 (NGAL-MMP-9) complex was found to be superior to CRP for detection of endoscopic healing with infliximab (IFX) in patients with ulcerative colitis (UC)[1]. Both cathelicidin LL-37 and chitinase 3-like 1 (CHI3L1) are secreted by neutrophils and were previously associated with IBD [2,3]. The aim of this study was to investigate whether LL-37 and/or CHI3L1 could improve detection of endoscopic healing with IFX in UC patients. Methods: Serum samples were obtained from 145 UC patients (41% female, median [IQR] age 41.3 [30.8–51.9] years) who underwent endoscopy following IFX initiation and from 75 controls (56% female, median [IQR] age 33.6 [29.2–51.8] years). Endoscopic healing with IFX was defined as a Mayo endoscopic subscore of 0 or 1. CRP, NGAL-MMP-9 and neutrophils were previously determined [1], and LL-37 and CHI3L1 were measured with ELISA (Hycult Biotech and R&D systems, respectively). For all markers, optimal cut-offs were determined with ROC analysis and binary variables were entered in a logistic regression model to generate the Ulcerative Colitis Response Index (UCRI). Non-parametric tests were performed and p-values <0.05 were considered significant. Results: Median (IQR) time to serum sampling after IFX was 8.2 (6.0–14.0) weeks. 83 patients (57%) had endoscopic healing, whereas 62 patients (43%) did not. Median [IQR] LL-37 levels (ng/ml) were significantly lower in healers (24.3 [16.1–41.4]) compared to non-healers (37.3 [24.0–53.8], p=0.002), but remained elevated compared to controls (16.7 [10.2–27.1]; p<0.001). Median [IQR] CHI3L1 levels (ng/ml) were significantly lower in healers (20.9 [14.3–34.4]) compared to both non-healers (30.0 [22.7–53.9], p<0.001) and controls (31.9 [19.6–48.6], p=0.003). UCRI consisted of CRP (Odds ratio [95% CI] 3.3 [1.4–7.5]), CHI3L1 (3.1 [1.3–7.7]), neutrophils (4.9 [2.1–11.2]) and LL-37 (2.5 [1.0–6.4]). The AUC of UCRI was 0.83 and Q1 (0.0–2.6) was able to discriminate healing with 54% sensitivity, 92% specificity, 60% NPV and 90% PPV, whereas Q4 (7.2–9.8) was able to discriminate non-healing with 37% sensitivity, 95% specificity, 67% NPV and 85% PPV. Finally, UCRI could detect endoscopic healing as early as 3 weeks after IFX initiation (Hazard ratio [95% CI] 4.1 [2.6–6.5]). Conclusions: In the search for surrogate markers of endoscopic healing, UCRI was shown to accurately identify UC patients who fail to achieve healing with IFX and may help in early decision making to therapy switch.