Vitamin C and Charcot-Marie-Tooth 1A: Pharmacokinetic considerations.

Charcot-Marie-Tooth 1A disease (CMT1A) is a disease for which no drug treatments are available. In 2004, it was reported that ascorbic acid reduced the severity of neuropathy in transgenic mice overexpressing PMP22, an animal model of human CMT1A, compared with untreated mice. Based on those results, clinical trials were undertaken at different centers worldwide and four of them have been completed, but none of them resulted in significant improvements. Based on the pharmacokinetics of ascorbic acid, we propose that the randomized clinical trial carried out thus far confirmed the tight control of ascorbic acid's absorption and proved its tolerability at one and two years. The pharmacokinetic considerations discussed in this article might largely explain the disappointing results of the recent CMT1A trials.

Charcot–Marie–Tooth 1A disease (CMT1A) is a disease for which no drug treatments are available [1]. A potential breakthrough for human therapy was anticipated in 2004, when Passage et al. reported that ascorbic acid reduced the severity of neuropathy in transgenic mice overexpressing PMP22, an animal model of human CMT1A, compared with untreated mice [2]. These data are corroborated by in vitro observations that ascorbic acid inhibits PMP22 expression (by reducing cAMP levels) [3]. Based on those results, clinical trials were undertaken at different centers worldwide and four of them have been completed [4-8]. None of them resulted in significant improvements. Why did they fail?

These trials have been based on the results of a single paper, which has not been replicated to date and in which a single animal model was used. However, we would like to underscore important issues in ascorbic acid pharmacokinetics, which might have undermined CMT1A trials. The major difference between the animal model of Passage et al. [2] and human trials is that mice are able to synthesize vitamin C. Indeed, Passage et al. were “unable to demonstrate a peak of ascorbic acid in either plasma samples from the saphenous vein or directly from the heart” (page 399 of Passage et al. [2]). In brief, there is no evidence that lower or higher doses in mice would produce different effects. Indeed, Passage et al. tested twice the dose of ascorbic acid and did not record any further improvement. Consequently, the issue of vitamin C dosage in rodents (with the exception of the guinea pig, who is unable to synthesize ascorbic acid) is unresolved and very difficult to transpose onto human settings.

In addition to differences between models, i.e. rodents vs. humans, the question arises as to whether vitamin C was provided in sufficient amounts to exert clinical effects. Before 2000, Recommended Dietary Allowances (RDAs) for vitamin C had a simple goal: prevention of frank vitamin C deficiency (scurvy) with an additional margin of safety. To correctly address this issue, however, we need to trace back and explain how the current RDA for ascorbic acid was established. As with other micronutrients, ethical and practical hurdles strongly limit the correct assessment of the RDA. The first attempt was that of Hodges et al. [9], who induced experimental scurvy in Iowa inmates, who were then replenished until symptoms disappeared. Based on these data, an RDA of 60 mg was established and increased to 90 mg for men and 75 mg for women only in the year 2000. The only complete pharmacokinetic study carried out with vitamin C is that of Levine et al. [10]. Their depletion–repletion study was carried out on hospitalized volunteers and the authors concluded that “Bioavailability was complete for 200 mg of vitamin C as a single dose. No vitamin C was excreted in urine of six of seven volunteers until the 100-mg dose. At single doses of 500 mg and higher, bioavailability declined and the absorbed amount was excreted. Oxalate and urate excretion were elevated at 1000 mg of vitamin C daily compared to lower doses. Based on these data and Institute of Medicine criteria, the current RDA of 60 mg daily should be increased to 200 mg daily, which can be obtained from fruits and vegetables. Safe doses of vitamin C are less than 1000 mg daily, and vitamin C daily doses above 400 mg have no evident value.” In one of its more controversial applications, gram doses of vitamin C were promoted by the two-time Nobel Laureate Linus Pauling as a cancer treatment agent. However, as oral vitamin C produces plasma concentrations that are tightly controlled, only intravenous administration of vitamin C produces high plasma and urine concentrations that might exert supranutritional health effects [11]. This is clearly impractical in randomized clinical trials, even though high dose IV vitamin C is in unexpectedly wide use by complementary/alternative medicine practitioners [12].

The CMT trials that have been published to date appear to fit pharmacokinetics predictions. As an example, the first trial [4] used twice daily dosing corresponding to less than 2 g/d for the oldest children subjects. Plasma concentrations of in the range of approximately 100 μM are predictable and were, indeed, attained. Another study used either 1 g/d or 3 g/d [5], given in a single administration before breakfast. The resulting plasma concentrations depend on when samples were drawn in comparison to dosing (which was not specified in the paper). If – as likely – they were fasting specimens, the reported value of 106 μM is slightly lower than expected, but still fitting pharmacokinetic data [10].

Pareyson et al. chose to administer 1.5 g/d, divided into two doses of 1 g in the morning and 500 mg in the evening [7]. This dosage is approximately 2.5 fold that used in the Passage et al. study in the mouse (1.12 mg/week). According to a conversion table published by Reagan-Shaw et al. [13], 1.5 g/d for a human being would correspond to 0.43 mg/day (hence 3.01 mg/week) for a mouse. This dose is undoubtedly efficacious in achieving maximal and steady plasma levels with low likelihood of producing side effects. The authors reported significant increases of plasma ascorbate concentrations, though the absolute values are difficult to compare with other studies because of the different methodologies employed to measure vitamin C (spectrophotometry vs. HPLC with electrochemical detection).

Two final considerations are warranted. The first one is that data obtained from in vitro experiments should be interpreted with caution, as culture media are devoid of vitamin C; consequently, cultured cells are in scorbutic states and exogenous vitamin C first replenishes their stores and then exerts biological effects [14,15]. The other reflection is that upon vitamin C transporters, which are saturable vitamin C can enter cells both in its reduced and oxidized form, ascorbic acid (AA) and dehydroascorbate (DHA), utilizing respectively sodium-dependent transporters (SVCT, also indispensable for the uptake of vitamin C by Schwann cells [16]) or glucose transporters (GLUT) [17]. However, vitamin C in excess of their Km will neither be uptaken nor transported inside the cell. To circumvent it, dosing must be quite frequent, and higher than used in these four studies. Even if such dosing were used, it would be difficult to get plasma concentrations above 160 μM unless participants took gram doses of ascorbate continually every 4–6 h, which is impractical in long-term.

Of note, a review of over 200 articles on vitamin C and health concluded that a daily intake of 100 mg of vitamin C is associated with lower incidence of heart disease, stroke, and cancer [18,19]. It should be underscored that these amounts (1) are consistent with those reported by Levine et al. to saturate cells [10] and (2) are attainable through a balanced diet. Whether CMT1A patients have impaired vitamin C transport system is, at present, unknown.

In addition to discussing the optimal dose, one should also consider potential side effects of high, possibly excessive amounts of vitamin C. Indeed, Toth tested a high, i.e. 5 g/d dose of ascorbic acid on CMT patients and reported poor tolerability, mainly in terms of gastrointestinal discomfort [20]. A number of potential problems associated with very large doses of ascorbic acid have been suggested, mainly based on in vitro experiments or isolated case reports, including genetic mutations, birth defects, cancer, atherosclerosis, kidney stones, “rebound scurvy,” increased oxidative stress, excess iron absorption, vitamin B12 deficiency, and erosion of dental enamel [21]. However, none of these alleged adverse health effects have been confirmed, and there is no reliable scientific evidence that large amounts of vitamin C (up to 10 g/d in adults) are toxic or detrimental to health. Along with the latest RDA published, as mentioned, in 2000, a tolerable upper intake level (UL) for vitamin C was set for the first time. A UL of 2 g/d was recommended in order to prevent most adults from experiencing diarrhea and gastrointestinal disturbances [22]. Such symptoms are not generally serious and often they resolve with temporary discontinuation or reduction of high-dose vitamin C supplementation. However, they increase the possibility of unmasking if they occur during double-blind trials. Finally, a review by Hathcock et al. [23] concluded that “Intakes of vitamin C well in excess of 2 g/d have sometimes been associated with gastrointestinal upset or skin rashes, but other evidence suggests that intakes up to 4000 mg/d are well tolerated in the general population”.

In conclusion, based on current knowledge, (1) intakes of vitamin C in excess of 400 mg/d are not expected to provide additional benefits as cells are already saturated and (2) the Institute of Medicine set an upper limit of 2 g/d which, albeit restrictive, would prevent adverse effect and, consequently, increased attrition. The randomized clinical trial carried out thus far confirmed the tight control of ascorbic acid's pharmacokinetics and proved its tolerability at one and two years.

The pharmacokinetic considerations discussed above might largely explain the disappointing results of the recent CMT1A trials.