Introduction FC is a marker of GI inflammation. Four commercial ELISA-based assays are available, two polyclonal (Calpro [“C”]), Eurospital [“E”]) and two monoclonal (Buhlmann [“B”], Immunodiagnostik [“I”]). “C” is a manual assay, rest are automated. Automation eases testing. Monoclonal assays are reportedly more accurate. Head-to-head comparison of all four assays is unexplored to the best of our knowledge.
Aim Pilot study to compare the four assays to help us select one (preferably automated) that best meets our clinical needs: reliably exclude GI inflammation (new patients) and quantify inflammation (known IBD).
Methods 42 stool samples collected from January to March 2011 were tested. Patients: 18 new (mainly for diarrhoea), 24 follow-up IBD (in remission/chronic active disease/flare). Assay (n): “C” (42), “B” (36), “I” (36), “E” (35). All four assays: 29/42 (sample insufficient in rest to do all 4). Analysis: Blinded to assay details, a single investigator (MS) mapped FC values to inflammation grade (0=nil, 1=mild/possible, 2=severe/definite) based on conventional markers (CRP/imaging/endoscopy/histology) and final diagnosis. Linearity characteristics of each assay was assessed by Excel trendlines. Restricting analysis to the 29 samples tested by all four assays (giving six pairings), inter-assay concordance was determined for each inflammation grade by Kendall co-efficient. p Value <0.02 (Fisher ratio) was deemed significant.
Results All four assays showed linear characteristics with different gradients, minimum and maximum values (Abstract PTU-243 figure 1). “C” had maximum gradient and highest values while “I” had the lowest levels detectable. Assays “B” and “E” had characteristics in between. Inter-assay concordance (Abstract PTU-243 table 1) was statistically significant in absence of inflammation for all pairings. The highest assay concordance across all grades of inflammation was between monoclonal “I” and polyclonal “C”.
Conclusion In this pilot, assays “I” and “C” had the most favourable characteristics/concordance. If this trend is confirmed by larger numbers, we will adopt the monoclonal assay “I” as it is automated.
Competing interests None declared.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.