30 ELEVATION OF PLASMA ONCOSTATIN M IS ASSOCIATED WITH PRIMARY BIOCHEMICAL NONRESPONSE AND SECONDARY CLINICAL NONRESPONSE TO INFLIXIMAB IN PEDIATRIC CROHN’S DISEASE PATIENTS

Abstract

Oncostatin M (OSM) and OSM receptor (OSMR) were found to be highly expressed in inflamed intestinal tissue in patients with Crohn’s disease (CD) and ulcerative colitis (UC).1 West et al. also found elevations of mucosal OSM was strongly associated with treatment nonresponse to tumor necrosis factor-α inhibitors (anti-TNF).1 We hypothesized that increased abundance of plasma OSM would also be associated with anti-TNF nonresponse with our primary aim to identify an OSM cut-point for early anti-TNF nonresponse. We enrolled anti-TNF naïve children with CD starting infliximab into this prospective, observational study. The primary outcome was biochemical response and defined by a fecal calprotectin <400 µg/ml prior to the first maintenance dose (infusion 4). This outcome was restricted to patients with both a pre-infliximab fecal calprotectin >400 µg/ml and a stool sample collected at infusion 4. Clinical remission at 12 months was defined as a weighted pediatric CD activity index <12.5, off steroids, and on infliximab without surgery in the first year. The ideal OSM cut-point was identified using the Youden index from the receiver operating characteristic curve analysis of biochemical responders and nonresponders. OSM was measured using ELISA (ThermoFisher, Frederick, MD) with an intra-assay CV<10%. Fecal calprotectin was measured with ELISA (Buhlmann, Switzerland). Infliximab concentrations were measured with ELISA (Immundiagnostik, Germany). We measured pre-infliximab OSM in 40 consecutive CD patients. The cohort was 35% female, 90% white race with a mean age of 13 (SD 4) years at the start of infliximab. One patient remained on methotrexate during induction and 50% of the cohort received prednisone prior to the first induction dose. 33/40 met criteria for evaluation for the primary aim with 16/33 (49%) in biochemical remission at infusion 4. The median (IQR) OSM in biochemical nonresponders was 96 (69-198) pg/ml compared to 166 (84-1766) pg/ml in nonresponders (p=0.035, Figure1). Using the Youden index, we identified plasma OSM >143 pg/ml was 71% sensitive and 75% specific for biochemical nonresponse (Figure 2). The median infusion 4 fecal calprotectin in the OSMhigh (>143 pg/ml) group was 985 (382-2501) µg/g compared to 317 (141-676) µg/g in OSMlow group (p=0.04). We found only 25% of the OSMhigh patients achieved biochemical response compared to 71% with OSMlow (p=0.015). Thirty-five percent of OSMhigh patients achieved steroid-free clinical remission at month 12 (odd’s ratio of 0.21, 95% CI 0.05-0.77, p=0.023). We found no differences in infliximab concentration at infusions 2, 3 and 4 between OSMhigh and OSMlow patients. In a small pediatric CD cohort, we identified a pre-infliximab cut-point of >143 pg/ml was predictive of early biochemical nonresponse and secondary clinical nonresponse at year 1. Figure 1. Oncostatin M (OSM) was measured with ELISA prior to the first infliximab infusion. Fecal calprotectin was measured prior to infusion 1 and 4. Groups compared by Mann-Whitney test. Figure 2. Oncostatin M (OSM) was measured prior to the first infliximab infusion. Area under the curve (AUC) was determined by the receiver operating characteristic curve for biochemical nonresponse at infusion 4 (defined by a fecal calprotectin ≥400 µg/g). The ideal cut-point for OSM was determined by the Youden index.