Biotin, a water-soluble B-complex vitamin involved as a coenzyme in several critical carboxylation reactions, has become a popular over-the-counter supplement for its purported effects in strengthening hair and nails, controlling blood glucose levels, and easing peripheral neuropathy. While the National Academies of Sciences, Engineering, and Medicine’s Food and Nutrition Board recommends daily adequate intake of 30 µg for adults, a typical biotin supplement contains 5,000–10,000 µg. Recently, some studies even have suggested that mega doses (100,000–300,000 µg daily) might be beneficial in multiple sclerosis and other neurodegenerative disorders.
This upsurge in biotin supplementation is having an unintended consequence: spurious immunoassay results in some patients. At least three papers have looked at this phenomenon, and my endocrinology colleagues and I are seeing more instances of it. However, I do not think the medical community at large is at all aware of this issue. Laboratorians have an important role in consulting with physicians who see immunoassay results that do not match their patients’ clinical presentation.
A Very Strong Bond
Two important chemical characteristics of biotin are behind this circumstance. First, biotin has a valeryl side-chain, allowing covalent bonds (biotinylation) to form easily with many small or large molecules, without significantly altering their antigenic or biologic properties. Second, biotin binds to a protein from Streptomyces avidinii—streptavidin—with the exceptionally low dissociation constant of 10-14. The biotin-streptavidin complex is one of the strongest non-covalent bonds observed in nature, about 1,000 times the affinity of the most potent antibody-hapten interaction, making it resistant not only to pH and temperature extremes but also solvents. Consequently, biotin and streptavidin are used in several immunoassays to ensure strong non-specific binding between chemicals of interest.
In sandwich assays, biotin covalently binds to an antibody specific for the analyte being tested; thyroid stimulating hormone (TSH), for example. This antibody is incubated with the serum sample and a second analyte-specific antibody linked to a reporter system, for example, ruthenium. The resulting biotinyl-antibody-analyte-ruthenium-antibody complexes precipitate on the solid phase coated with streptavidin. The ruthenium-generated signal is directly proportional to the amount of hormone present in the serum sample. A large excess of soluble biotin in the serum sample will compete for streptavidin with the biotin-linked antibody and therefore result in incomplete or no solid phase formation, and the reporter system will register a zero signal, interpreted as a falsely low hormone level.
Biotin excess also affects competitive assays, with the opposite effect. In this case, the serum analyte competes with its biotinylated version for binding sites on a specific antibody linked to ruthenium. The amount of biotinylated analyte-antibody-ruthenium complex precipitated on the streptavidin-coated solid phase is inversely proportional to the analyte’s concentration in the sample. A large excess of biotin in the sample will prevent the solid phase formation and yield a low signal, interpreted as a falsely high analyte level.
Of recent reports about these interferences, the most striking involved patients diagnosed erroneously with Graves’ disease. In these cases, mega dose biotin supplementation led to falsely low TSH levels in a sandwich-design assay and to markedly falsely elevated triiodothyronine (T3) and thyroxine (T4) levels, with highly positive TSH receptor antibody, in competitive assays. Remarkably, this combination of falsely high and falsely low test results faithfully recapitulates the laboratory diagnosis of Graves’ disease, raising no red flag. Only the discordance between these results and patients’ asymptomatic clinical presentation have unmasked this potentially disastrous combination.
Other reports have described interference with thyroglobulin, parathyroid hormone, dehydroepiandrosterone sulfate, estradiol, and ferritin assays. The list of immunoassays potentially affected is longer and includes the whole catalog offered by some companies. Commercial platforms from Roche, Beckman, and ISYS all have been shown to be sensitive to the interference. The magnitude and duration of the interference is likely to be variable, and may depend on the patient’s biotin supplement dose. In my practice, I saw from a sandwich assay the same markedly elevated T4 levels but normal sex hormone binding globulin level results. As described in a case report, in vivo experiments showed the interference disappearing 24 hours after a 10 mg dose. However, biotin interference may last longer in patients taking megadoses.
If markedly elevated thyroid hormone levels in the absence of hyperthyroidism symptoms raise suspicion, milder and less obvious cases are likely to go unsuspected. Consider the tests used for cancer screening and monitoring, such as prostate-specific antigen or thyroglobulin, for which there is no immediate clinical alarm bell for falsely elevated or falsely low results.
Are there any solutions to this problem? In my view, changes to assay design, such as the pre-absorption of all sera with streptavidin, are unlikely to be implemented in the near future, as they would be expensive. Awareness is the simple answer. Clinical laboratorians should update busy clinicians, unlikely to read laboratory bulletins, on this problem, while clinicians should build in their practice the habit of specifically questioning patients on biotin use.
Giuseppe Barbesino, MD, is an assistant professor of medicine at Harvard Medical School and an endocrinologist at Massachusetts General Hospital in Boston. +Email: firstname.lastname@example.org