P213 A systematic review and meta-analysis of non-invasive biomarkers for assessing disease activity in Inflammatory Bowel Disease
M.H. Mosli*1, 2, G. Zou1, 3, S.K. Garg4, S.G. Feagan1, J.K. MacDonald1, W.J. Sandborn1, 5, N. Chande6, B.G. Feagan1, 3, 7
1Robarts Clinical Trials Inc., Robarts Research Institute, University of Western Ontario, London, Ontario, Canada, 2King Abdulaziz University, Department of Medicine, Jeddah, Saudi Arabia, 3University of Western Ontario, Department of Epidemiology and Biostatistics, London Ontario, Canada, 4University of Minnesota, Department of Medicine, Minneapolis, United States, 5University of California San Diego, Division of Gastroenterology, La Jolla, United States, 6London Health Sciences Centre, Victoria Hospital, London, Canada, 7University of Western Ontario, Department of Medicine, London , Canada
Endoscopic disease activity in inflammatory bowel disease (IBD) is associated with poor outcomes. Endoscopic evaluation is the gold standard for the assessment of disease activity, but is invasive, expensive and potentially time consuming. Identification of non-invasive biomarkers of disease activity in IBD is a research priority.
The primary objective was to evaluate the diagnostic accuracy of 3 non-invasive biomarkers (fecal calprotectin [FC], stool lactoferrin [SL] and C-reactive protein [CRP]) used for the evaluation of disease activity in IBD. MEDLINE, EMBASE, the Cochrane Library, the ISI Web of Knowledge and conference abstracts were searched from inception to November 2014 for relevant studies. Grey literature databases (e.g. SIGLE) were also searched to identify studies not indexed in traditional databases. All cohort and case-control studies that evaluated the diagnostic accuracy of FC, SL or CRP for assessment of disease activity in symptomatic patients with previously diagnosed IBD (ulcerative colitis and Crohn's disease) were included. True positive, true negative, false positive and false negative rates were extracted for each biomarker and used to construct 2X2 tables for each cutoff. Sensitivity, specificity and area under the curve (AUC) estimates for FC, SL and CRP were calculated for each study based on different cut-offs and pooled together into single estimates for each test. Receiver operator characteristics (ROC) curves were then used to identify the cut-off values for each biomarker that best predicted endoscopic disease activity.
Nineteen studies (2456 participants) met our inclusion criteria. Sensitivity, specificity, and AUC values for the 3 biomarkers are summarized in Table 1.
“Table 1. Diagnostic accuracy of fecal calprotectin, stool lactoferrin, and C-reactive protein”
The best cut-off values to detect endoscopically active disease in IBD determined by ROC analysis were 50 μg/g, 7.25 μg/mL and 10 mg/dL for FC, SL and CRP, respectively.
FC and SL are highly accurate biomarkers that can be used to screen symptomatic IBD patients for endoscopic disease activity prior to colonoscopy.