Clinical implications of assay specific differences in f-calprotectin when monitoring inflammatory bowel disease activity over time. (4th March 2017)

• Record Type: Journal Article
• Authors: Amcoff, Karin; Stridsberg, Mats; Lampinen, Maria; Magnuson, Anders; Carlson, Marie; Halfvarson, Jonas
• Is Part Of: Scandinavian journal of gastroenterology. Volume 52:Number 3(2017)
• Page Start: 344
• Page End: 350
• Publication Date: 2017-03-04
• Subjects: Biomarker -- Crohn's disease -- faecal calprotectin -- inflammatory bowel disease -- ulcerative colitis; Gastroenterology -- Periodicals; Digestive organs -- Diseases -- Periodicals; 616.33
• DOI: 10.1080/00365521.2016.1256424
• Languages: English
• ISSNs: 0036-5521

Abstract

Abstract: Objective: With several faecal calprotectin (FC) assays on the market, it has been difficult to define a uniform threshold for discriminating between remission and active disease in patients with inflammatory bowel disease (IBD). We aimed to compare the results of different FC-assays in IBD patients, followed over time. Material and methods: IBD patients provided faecal samples and reported clinical activity every third month prospectively over a two year period. FC was measured with two ELISA – (Bühlmann and Immunodiagnostik) and one automated fluoroimmunoassay (Phadia). Results: In total, 13 patients provided 91 faecal samples. The median (IQR) concentration of FC was higher at active disease than at remission for all assays: Bühlmann 845 (1061–226) μg/g versus 62 (224–39) μg/g, Phadia 369 (975–122) μg/g versus 11 (52–11) μg/g, and Immundiagnostik 135 (302–69) μg/g versus 8 (56–4) μg/g. The Bühlmann assay produced the largest absolute difference but the corresponding relative difference seemed to be more pronounced when analysed by the Phadia – (ratio of means 8.5; 95% CI 3.3–21.9) or the Immundiagnostik assay (ratio of means 7.4; 95% CI 3.1–17.6) than by the Bühlmann assay (ratio of means 5.3; 95% CI 2.7–10.6). Consequently, the specificity for discriminating active disease from remission varied between assays (34–75%) when the cut-off 50 μg/g was used, whereas the differences in sensitivity were less pronounced. Conclusions: Cross-comparisons revealed overall poor agreement between the assays as well as differences in the dynamics of FC. These findings suggest that standardisation of the method is needed to implement FC as a disease monitoring tool at large-scale.