Another indication for the use of oral vitamin B-12 supplementation

Reports of a randomized controlled supplementation trial that used a standardized nutrient are not common and are always to be applauded, but providing a comparator supplement is even more unusual. The report by Schijns et al. (1) in this issue of the Journal accomplishes both of these aims and provides an answer about the equal value in patients who have undergone a Roux-en-Y gastric bypass (RYGB) for improving low serum cobalamin (vitamin B-12) concentrations, with the use of either oral supplementation or the more standard intramuscular administration of the vitamin. The result is not unexpected based on what is known about cobalamin physiology, but the new information needs to be placed in context by exploring several aspects of the report itself.

It is well known that orally administered cobalamin is as effective in overcoming deficiency states as is intramuscular delivery (2). Moreover, this equivalence has been shown even when the cause of deficiency is the lack of production of gastric intrinsic factor, in patients with pernicious anemia (3). Schijns et al. (1) selected patients who underwent RYGB. In this procedure, a small fundic gastric pouch (∼30 mL) is connected to the jejunum, transected 30–75 cm beyond the ligament of Treitz. The excluded gastroduodenal-jejunal limb is re-anastomosed 75–150 cm distal to the gastrojejunostomy. This rearrangement leads to lack of stimulation of gastric secretion in the bypassed stomach, and because the gastric remnant in continuity does not include parietal cells, the source of intrinsic factor (IF), the production of IF is likely to be at relatively low unstimulated amounts after a meal. However, the concentration of IF in the gastric lumen is far in excess of that of cobalamin after a meal (4), so it is likely that patients with RYGB may still produce adequate IF to support cobalamin absorption by a receptor-mediated mechanism. Whether this is true or not, there remains sufficient small intestine in continuity, including all of the IF-cobalamin receptor ileal region, to support either IF-mediated or IF-free cobalamin absorption. Passive diffusion of cobalamin in the absence of IF accounts for ∼1% of the total ingested dose (4). When oral cobalamin is provided in sufficient amounts, it can account for the absorption of the Dietary Reference Intake for vitamin B-12 of 2.4 µg (0.24% of a 1000-µg dose). Thus, the efficacy of oral cobalamin noted in the current study was a result that confirms what is known about the oral absorption of vitamin B-12.

The patients studied also were sometimes treated with proton pump inhibitors (PPIs) or with metformin, both of which are discussed by the authors as potential confounders for cobalamin absorption. The data on clinical effects of PPIs on cobalamin status have been inconsistent, but the use of PPIs for ≥2 y and doses of >1.5 pills/d have been associated with an increased risk of developing vitamin B-12 deficiency (5). The possible reasons for this are not fully worked out. Lansoprazole at 60 mg/d for 7 d leads to a marked suppression of pepsin output and activity, but also an increase in IF concentration in healthy volunteers, the latter effect possibly related to a marked decrease in secreted volume (6). Although total IF secretion is dependent on gastric acid and fluid secretion from parietal cells, 60 mg omeprazole given for 9 d to healthy volunteers did not affect luminal IF secretion (7). Thus, the risk of developing vitamin B-12 deficiency while taking PPIs may be due to the balance of decreased pepsin activity (due to elevated pH) and maintenance of at least some IF secretion. At any rate, only 11 of 50 patients in the present trial were taking PPIs, the dose and time of treatment not otherwise specified, so the effect, if any, was probably small.

The other medication confounder mentioned by the authors was the use of metformin in 6 of 50 patients. This concern was raised because of the decrease in serum cobalamin and holotranscobalamin (holoTC), with both declines still within the normal range, which was noted in 21 patients taking metformin [reference 18 in (1)], and partial reversal of the holoTC measurement after oral calcium supplementation. However, holoTC is not the gold standard for assessing cobalamin deficiency, unlike the biomarkers methymalonic acid (MMA) and homocysteine (Hcy) used in the present study. Thus, this single report of the effect of metformin on cobalamin status must be considered as preliminary, with the need for further validation of the effect. At any rate, such a small percentage of patients in the current study were taking this drug that its possible effect on cobalamin absorption in the whole group is very small.

Another feature of the study design was the provision of a different form of cobalamin in each arm, hydroxycobalamin for intramuscular use and methylcobalamin orally. Although claims continue to be made for the unique efficacy of methylcobalamin, there is little published evidence that is convincing for its advantage. With the use of whole-body counting after administration of oral cyanocobalamin, coenzyme B-12, methylcobalamin, and hydroxycobalamin at different doses, oral absorption was probably the same for all forms (8). Moreover, after uptake into the cell, cobalamin of any form binds to the MMA and C-type homocysteine protein, which catalyzes removal of the upper axial ligand (9). Thus, there is no theoretical rationale for differing efficacy of various forms of vitamin B-12, and there is yet no convincing clinical evidence for differences in efficacy. The use of different forms of cobalamin used in the current study is probably not a critical variable in the study design.

A more significant concern involves the interpretation of the data in the present study. Although all of the patients were selected with the use of a cutoff of ≤200 pmol/L for serum cobalamin, this value is derived from the mean ± 2 SDs of population samples, whereas the more stringent cutoff of ≤148 pmol/L represents the lower value for the mean ± 3 SDs (10). The lower cutoff will miss 3–5% of cobalamin deficiency defined by using serum MMA and Hcy values, whereas the higher cutoff will produce more false positives. This latter concern is highlighted by the finding that most patients in the current study (∼60%) had normal MMA values and ∼82% had normal Hcy values. Thus, this population may represent patients with subclinical cobalamin deficiency (11). This syndrome describes patients with mild biochemical abnormalities but no symptoms or diagnostic markers of cobalamin deficiency. Although a roundtable consensus panel of NHANES data did not endorse subclinical cobalamin deficiency as a distinct clinical entity, it also noted that its definition depended on the cutoff value of serum cobalamin used, on the unknown natural history of the laboratory abnormalities (whether static, progressive, or resolving), and by the unknown need for therapy (10, 11). In the current study, there was a comparison between groups, but not from baseline to the end of study in either group for MMA and Hcy. Although the serum vitamin B-12 concentrations increased equally in both groups, there was no apparent change in MMA or Hcy concentrations in either group, as expected by the lack of abnormal baseline values in the majority of patients, and consistent with the lack of a diagnosis of cobalamin deficiency in most of the patients as diagnosed by MMA concentrations.

There is a further potential confounder in determining which patients might eventually develop clinical cobalamin deficiency, which relates to unrecognized deficiency of riboflavin. When riboflavin status is examined with the use of the functional marker, erythrocyte glutathione reductase activation coefficient (EGRAC), many young women have been identified as deficient, and the lower the riboflavin status, the greater the effect on hemoglobin concentrations with riboflavin supplementation (12). In addition, riboflavin status affects serum Hcy values in patients with the methylenetetrahydrofolate reductase polymorphism 677C→T, and the methionine synthase polymorphism 66A→G, and the cobalamin status affects Hcy concentrations in the latter polymorphism as well (13). Thus, it is possible that the development of clinical cobalamin deficiency might depend in part on concomitant riboflavin deficiency, a possibility not adequately examined in patients with RYGB or in other cobalamin-deficiency states.

In the absence of changes in the gold-standard measurements of cobalamin deficiency, it is not clear from the current study which patients with RYGB need to be treated with which formulation to prevent clinical cobalamin deficiency. What the study does show, however, is the equivalence in 6 mo of elevating serum cobalamin with either preparation. Because body stores of cobalamin can be adequate for ≥2 y in the absence of obvious malabsorption (4), there will be a need for longer follow-up of this cohort of patients to see if there are any differences in the incidence of clinical deficiency. Given the fact that supplemental cobalamin is provided as the isolated vitamin and is not dependent on intragastric release of protein-bound cobalamin, it seems likely that long-term use of oral cobalamin should be adequate to maintain serum cobalamin concentrations for years. Providing supplemental cobalamin in some form does seem to be indicated in many patients with RYGB, because the prevalence of low serum cobalamin concentrations in these patients is considerable, possibly due to poorly digested, protein-bound dietary cobalamin, and to poor mixing of pancreatic enzymes and dietary protein in the small intestine. Moreover, as dietary intake is decreased, foods rich in cobalamin may also decline. An alternate strategy to treating all RYGB patients would be to test regularly and to initiate therapy only when the serum cobalamin concentration falls below normal, with consideration that there is no gold-standard cutoff value. For supplementation once begun, the current study provides an alternative to the regular use of intramuscular cobalamin. This point has been made before for patients with pernicious anemia (3). Perhaps the time has arrived for the wider uptake of daily oral cobalamin supplementation in patients at risk of cobalamin malabsorption due to abnormalities in IF production but with normal IF-cobalamin ileal receptors in place, including the ever-increasing population of RYGB patients.

Notes

Supported in part by Nutrition and Obesity Research Center (NORC) grant DK056341 from the NIH (S Klein, Principal Investigator).

Abbreviations used: Hcy, homocysteine; holoTC, holotranscobalamin; IF, intrinsic factor; MMA, methylmalonic acid; PPI, proton pump inhibitor; RYGB, Roux-en-Y gastric bypass.