Biotin Measurement by Targeted LC-MS/MS: Method Development and Application to the Evaluation of Immunoassay Interference in Thyroid Function Tests

Abstract

Abstract Introduction While biotin is widely used as a health and beauty supplement, patients rarely report the use of biotin supplements to clinicians. This can negatively impact patient care since biotin can falsely elevate or lower results in immunoassays that use biotinylated components. On the Beckman DxI platform, total T3 (T3), free T3 (FT3), and free T4 (FT4) assays are falsely elevated by excess biotin, while TSH and total T4 (T4) are unaffected. The goals of this study were to develop a method to measure serum biotin, to determine biotin concentrations in samples submitted for thyroid function testing, and to correlate biotin concentrations with immunoassay result patterns suggestive of biotin interference. Methods We developed and validated a targeted LC-MS/MS method to measure serum biotin (AMR: 5.4-500 mcg/L). Two collections of patient samples were tested using automated Beckman DxI assays for TSH (RR: 0.4-5.0 μIU/mL), T3 (RR: 73-178 ng/dL), FT3 (RR: 2.4-4.1 pg/mL), T4 (RR: 4.8-10.8 mcg/dL), and FT4 (RR: 0.6-1.2 ng/dL). The first collection included consecutive samples submitted for TSH and FT4 testing. To increase the probability of identifying samples with biotin interference, a middleware preselection algorithm was used to identify samples with a results pattern suggestive of biotin interference (normal to high TSH with elevated T3, FT3, and/or FT4). Results Two of the 94 samples consecutively submitted for thyroid function tests demonstrated biotin levels >AMR. TSH and FT4 were within normal ranges for one, while the other had normal TSH and elevated FT4 (biotin: 26 mcg/L). The preselection algorithm targeting normal to high TSH with elevated T3, FT3, and/or FT4 results flagged 83 samples, 21.7% of which had measurable biotin (5.71-194.3 mcg/L). Immunoassay results for TSH, T3, FT3, T4, and FT4 were obtained where there was sufficient volume for testing. Of these 38 samples, all samples with biotin >24 mcg/L had elevated results for 2(+) affected assays. Conclusions In this study, a targeted LC-MS/MS method to measure biotin in serum samples was developed and used to estimate the distribution of biotin concentrations in samples submitted for thyroid function testing. Comparison of immunoassay result patterns with biotin concentrations suggested an interference threshold around 24 mcg/L. The evaluation of consecutively submitted samples showed that biotin in excess of this interference threshold in patients receiving thyroid function tests on the Beckman DxI platform is rare. However, combining the preselection algorithm with quantitative biotin measurement by LC-MS/MS demonstrated that biotin interference does occur in these patient samples and should be evaluated when there are confounding results or result patterns suggestive of interference. In addition, the biotin LC-MS/MS method described here has additional clinical utility in the evaluation of suspected interference in any immunoassay that utilizes biotinylated components.