Analytical evaluation of three enzymatic assays for measuring total bile acids in plasma using a fully-automated clinical chemistry platform

BACKGROUND: Although the clinical significance of measuring bile acids concentration in plasma or serum has been recognized for long in patients with hepatobiliary disease and/or bile acid malabsorption, the reference separation techniques are expensive and mostly unsuitable for early diagnosis and for measuring large volumes of samples. Therefore, this study was aimed to evaluate the analytical performance of three commercial enzymatic techniques for measuring total bile acids in plasma using a fully-automated clinical chemistry platform. METHODS: Three commercial enzymatic assays (from Diazyme, Randox and Sentinel) were adapted for use on a Cobas Roche c501. We performed imprecision and linearity studies, and we compared results with those obtained using a reference liquid chromatography-mass spectrometry (LC-MS) technique on an identical set of lithium-heparin plasma samples. RESULTS: Total imprecision was optimal, always equal or lower than 3%. All assays had optimal linearity between 3–138 μmol/L. The comparison studies showed good correlation with LC-MS data (Spearman’s correlation coefficients always >0.92), but all plasma samples values were significantly underestimated using the commercial enzymatic assays (-44% for Diazyme, -16% for Randox and -12% for Sentinel). The agreement at the 10 and 40 μmol/L diagnostic thresholds of total bile acids in plasma ranged between 86–92%. This discrepancy was found to be mainly attributable to a heterogeneous composition in terms of bile acids content of the three assay calibrators. CONCLUSIONS: This study suggests that the analytical performance of the three commercial enzymatic assays is excellent, thus confirming that automation of this important test by means of enzymatic assessment may be feasible, practical, reliable and supposedly cheap. Nevertheless, the underestimation of values compared to the reference LC-MS also suggests that the local definition and validation of reference ranges according to the combination between the specific enzymatic assay and the different clinical chemistry platforms may be advisable.