Changes in fecal lactoferrin as a predictor of of steroid responsiveness in pediatric patients with ulcerative colitis
Murphy, Sean Thomas
MetadataShow full item record
INTRODUCTION: The management of pediatric patients with ulcerative colitis (UC) is dependent upon the ability to detect meaningful changes in disease status. This is currently done using validated patient-reported clinical disease activity indices, including the Pediatric Ulcerative Colitis Activity Index (PUCAI). While useful for global assessments completed during ambulatory office visits, the sensitivity of this metric may be insufficient to reflect more subtle changes in disease activity or response to medical therapy in hospitalized patients. Intravenous steroids are typically employed in the management of patients admitted for acute exacerbations of UC symptoms. These are typically manifest by worsening bloody diarrhea, abdominal pain, and worsening anemia. There is presently no way of predicting whether a patient admitted for UC will respond to steroid therapy. Current paradigms dictate a five-day trial before considering a transition to more potent medical or definitive surgical approaches to the management of refractory colitis. The development of more sensitive and reliable biomarkers or disease activity metrics could enable clinicians to more expediently identify steroid non-responders. This would minimize patient morbidity, decrease risk of complication, and lower overall cost of care. Previous studies have demonstrated that changes in fecal lactoferrin (FLA) correlate with disease activity in patients with UC. OBJECTIVES: To analyze the predictive value of FLA in the response to steroid treatment of patients admitted for management of UC. METHODS: We recruited pediatric inpatients with UC in the Division of Gastroenterology, Hepatology and Nutrition at Boston Children’s Hospital who were hospitalized for treatment of a flare of their UC symptoms. After obtaining patient consent, we collected a stool sample on days 1 and 3 of their hospital stay. We sent samples to TECHLAB® Inc. (Blacksburg, VA) to be analyzed for levels of FLA. We compared Day 1, Day 3 and ∆FLA (Day 1 – Day 3) in two patient groups: those that responded to conventional steroid therapy and those that required rescue medical or surgical therapy. We reported statistical significance with the Wilcoxon signed-rank test. RESULTS: Of 67 patients consented for the study, 30 provided stool samples on both days 1 and 3 of their inpatient hospitalization. Of the 30 patients, 63.3% responded to steroids while 36.7% required rescue therapy with immunomodulators. ∆FLA for responders, 43.6μg/mL(-239.0, 331.6) (median(interquartile range)), did not differ significantly from non-responders, -74.1μg/mL(-296.7, 221.7), P = 0.3. CONCLUSIONS: Our findings do not demonstrate that measurement of changes in quantitative FLA over three days can be used to assess acute responses to steroid therapy. Increasing the sample size may allow us to better delineate subtle differences between responders and non-responders to steroid therapy.