Assessing vitamin B-12 absorption and bioavailability: read the label

See corresponding article on page 1504.

Clinical vitamin B-12 (B12) deficiency is characterized by hematological (macrocytic anemia or other cytopenias) and neurological (degeneration of the brain and spinal cord, peripheral neuropathy, dementia) manifestations (1). While low dietary intake of B12 due to low or absent intake of animal source foods (i.e., vegetarian and vegan diets) can cause significantly diminished B12 status, most cases of severe, clinically evident B12 deficiency are caused by intestinal malabsorption. The classical cause of B12 malabsorption is pernicious anemia in which autoimmune destruction of the gastric parietal cells leads to loss of gastric intrinsic factor (IF) required for binding of B12 in the small intestine and subsequent absorption in the ileum by receptor-mediated endocytosis. Other causes of B12 malabsorption that lead to varying severities of B12 deficiency include atrophic gastritis, pancreatic insufficiency, bacterial overgrowth in the small intestine, gastrectomy, ileal resection, inflammatory bowel disease (e.g., Crohn disease), and long-term treatment of gastric reflux with proton pump inhibitors or H₂-receptor antagonists (1).

The gold standard test for assessing B12 absorptive capacity was described by Schilling in 1953 (2). The “Schilling test” consisted of providing a physiological oral dose of radioactively labeled B12 (usually 1 µCi of cobalt-57– or cobalt-58–labeled cyanocobalamin), followed ∼2 h later by a large intramuscular dose of unlabeled B12 (1 mg). The intramuscular dose saturates plasma B12 transport proteins, so the absorbed, unbound labeled B12 is flushed into the urine. A low amount of labeled B12 in the urine indicated malabsorption and required repeating the test with coadministration of IF, after a course of antibiotics, or with pancreatic enzymes to pinpoint the cause of malabsorption as either pernicious anemia, intestinal disease, bacterial overgrowth, or pancreatic insufficiency, respectively. The egg-yolk B12 absorption test, a modified version of the Schilling test in which ⁵⁷Co-B12 is mixed with egg prior to oral administration, was used to identify malabsorption from food-bound B12 due to conditions such as atrophic gastritis (3).

In recent years, the Schilling test has become obsolete due to reduced availability of the test components (i.e., ⁵⁷Co- or Co-B12, IF), and radiation exposure concerns. Consequently, differentiating pernicious anemia from other malabsorption disorders is now performed indirectly on the basis of clinical features and limited blood analysis (for example, autoantibodies against IF or gastric parietal cells) (1). Notwithstanding, there remains a need for a direct diagnostic test of B12 absorptive capacity (4).

Within the last decade and a half, 2 potential alternatives for assessing B12 absorption have been put forward. The first alternative is the “CobaSorb” test, which consists of oral consumption of 3 separate doses of unlabeled B12 at 6-h intervals, followed by measurement of the change in the amount of B12 bound to the serum transport protein transcobalamin (holotranscobalamin) (5). A small or absent increase in holotranscobalamin is indicative of malabsorption. The second alternative to the Schilling test is carbon-14–labeled B12 (¹⁴C-B12), which is synthesized by supplying ¹⁴C-labeled substrate to a genetically modified strain of Salmonella enterica, which then synthesizes B12 with a single ¹⁴C label located specifically in the dimethylbenzimidazole moiety of the molecule (6). After oral consumption of ¹⁴C-B12, enrichment of carbon-14 in blood, urine, and stool samples can be measured using accelerator MS. Measuring serum ¹⁴C-B12 has been used to assess absorption and bioavailability of B12 from aqueous solution (6), endogenously enriched chicken eggs (7), and fortified bread (8). Neither the CobaSorb test nor ¹⁴C-B12 has been thoroughly validated for use in diagnosing B12 malabsorption.

In this issue of the Journal, a promising new option for assessing B12 absorption and bioavailability is described. Devi et al. (9), employing the same strain of S. enterica used in the ¹⁴C-B12 studies described above, have succeeded in labeling the molecule with the stable isotope, carbon-13. In contrast to ¹⁴C-B12, the ¹³C label is located in multiple locations within the corrin ring of the molecule. These authors have also assessed the pharmacokinetics and bioavailability of high (18.3 µg) and low (2.3 µg) oral doses of ¹³C-B12 (in the form of cyanocobalamin) in human subjects. (For reference, the recommended dietary allowance for vitamin B12 for adults is 2.4 µg/d.) Enrichment of plasma samples at various time intervals after oral dosing was assessed by ultra-HPLC-MS. In general, the ¹³C-B12 behaved as expected: it was absorbed into the blood with peak plasma concentration occurring at ∼7 h, consistent with what has been shown using ⁵⁷Co-, ⁵⁸Co-, or ¹⁴C-B12 (6, 10). In addition, mean bioavailability for the low dose of ¹³C-B12 was 46%, very close to what has previously been reported (8, 10), whereas the mean bioavailability of the high dose was 7.6%, consistent with the known phenomenon of diminishing percentage absorption of B12 with increasing doses because of saturation of physiological IF-mediated B12 absorption in the ileum (10). Devi et al. (9) also demonstrated that essentially all of the ¹³C-B12 in the plasma was in the form of methylcobalamin, consistent with the expected metabolism of cyanocobalamin to methylcobalamin within the ileal cells (10, 11).

While the focus of the Devi et al. (9) paper is on proof of principle that a stable isotope-labeled molecule can be synthesized and used to estimate the pharmacokinetics and bioavailability of B12, the existence now of ¹³C-B12, as well as ¹⁴C-B12, breathes new life into the prospects of replacing the obsolete Schilling test. Specifically, these labeled forms of the vitamin may allow for a single measurement of isotopically enriched B12 in plasma within 6–8 h after oral dosing as a quantitative indicator of absorption. This would eliminate the need for both the intramuscular flushing dose of unlabeled B12 and the measurement of labeled-B12 in a 24-h urine sample that were required in the Schilling test. However, we are not yet ready to replace the Schilling test, as neither ¹³C-B12 nor ¹⁴C-B12 has been validated for diagnostic use in patients with pernicious anemia and other conditions of B12 malabsorption.

One final note of interest: although the label in ¹⁴C-B12 is located in the dimethylbenzimidazole moiety of the molecule (6), the stable isotope form of the vitamin has the ¹³C label introduced into multiple locations within the corrin ring of the molecule (9). This makes possible dual-isotope studies of the biochemical and physiological fate of B12, which in addition to being absorbed intact, metabolized, and utilized as a cofactor for B12-dependent reactions, is likely also metabolized and converted to analogue forms of B12 by gut microbes (6, 12), the significance of which is not well understood. We therefore anticipate not only a new day of clinical diagnosis of B12 malabsorption, but also advances in our fundamental understanding of its biological fate in humans and microbes alike. It simply comes down to reading the label.