Non-immunological biomarkers for assessment of villous abnormalities in patients with celiac disease.

ABSTRACT

BACKGROUND AND AIM: Demonstration of villous abnormalities is an essential component of diagnosis of celiac disease (CeD) that requires duodenal biopsies. There is a need for non-invasive biomarker(s) that can predict the presence of villous abnormalities. METHODS: Levels of plasma citrulline, plasma intestinal fatty acid binding protein (I-FABP), and serum regenerating gene 1α (Reg1α) were estimated in treatment naïve patients with CeD and controls. The levels of these biomarkers and their cyclical pattern were validated in a predicted model of enteropathy. Optimum diagnostic cut-off values were derived, and the results were further validated in a prospective validation cohort. RESULTS: While level of plasma citrulline was significantly lower, the levels of plasma I-FABP and serum Reg1α were significantly higher in patients with CeD (n = 131) in comparison with healthy (n = 216) and disease controls (n = 133), and their levels reversed after a gluten-free diet (GFD). In the model of predicted enteropathy (n = 70), a sequential decrease and then increase in the level of plasma citrulline was observed; such a sequential change was not observed with I-FABP and Reg1α. The diagnostic accuracy for prediction of presence of villous abnormality was 89% and 78% if citrulline level was  ≤ 30 µM/L and I-FABP levels were ≥ 1100 pg/mL, respectively. The results were validated in a prospective validation cohort (n = 104) with a sensitivity and specificity of 79.5% and 83.1%, respectively, for predicting villous abnormalities of modified Marsh grade > 2 at calculated cut-off values of citrulline and I-FABP. CONCLUSIONS: Plasma citrulline  ≤ 30 µM/L is the most consistent, highly reproducible non-invasive biomarker that can predict the presence of villous abnormality and has the potential for avoiding duodenal biopsies in 78% patients suspected to have CeD.