Treatment of Hyperhomocysteinemia in Renal Transplant Recipients

Methods

The institutional review board at Rhode Island Hospital, Providence, Rhode Island, approved the study protocol, and all study participants provided written informed consent. Study participants were 29 clinically stable renal transplant recipients (that is, they had had transplantation 6 months previously, had no clinical evidence of renal graft rejection, and had normal liver aminotransferase values) who were selected from among 35 consecutive renal transplant recipients involved in ongoing investigations [3] of homocysteine metabolism. Six of the 35 patients did not participate in our current study because of deteriorating renal function (n = 2), only part-time residence in Rhode Island (n = 2), and lack of interest (n = 2). Study participants either did not use vitamin supplements or had abstained from taking any supplements containing folic acid, vitamin B₁₂, or vitamin B₆ for at least 6 weeks before the start of the study. On the basis of their initial pretreatment levels of fasting total homocysteine (>20 µmol/L) and post-methionine-loading increases in total homocysteine levels (>30 µmol/L), participants were randomly assigned in blocks to one of four regimens: placebo (n = 8); vitamin B₆, 50 mg/d (n = 7); folic acid, 5 mg/d, and vitamin B₁₂, 0.4 mg/d (n = 7); or vitamin B₆, 50 mg/d, folic acid, 5 mg/d, and vitamin B₁₂, 0.4 mg/d (n = 7). Treatment assignments were made by a pharmacist who was blinded to all other aspects of the study. Laboratory analyses, data entry, and data analyses were performed by code so that treatment assignments remained concealed. Compliance with treatment was assessed by pill counts and determination of the change in plasma vitamin status.

Fasting and 2-hour post-methionine-loading blood samples were collected twice before treatment and twice during the sixth week of treatment, as described elsewhere [3]. Plasma total homocysteine levels were determined by high-performance liquid chromatography with fluorescence detection [7], plasma folate levels were measured by microbiological (Lactobacillus casei) assay [8], plasma pyridoxal 5′-phosphate levels were measured by radioenzymatic (tyrosine decarboxylase) assay [9], and plasma vitamin B₁₂ levels were ascertained by radioassay. Serum creatinine, albumin, and liver aminotransferase levels were measured by using standard automated clinical chemistry laboratory techniques. To eliminate interassay variability, all analytes were batch assayed from aliquots (which had been cryopreserved at −70°C)obtained during each of the four study visits.

Two independent treatment effects were assumed a priori: 1) that vitamin B₆ treatment would reduce the post-methionine-loading increase in plasma total homocysteine levels and 2) that folic acid plus vitamin B (₁₂) treatment would decrease fasting plasma total homocysteine levels. We also assumed that these treatments would not interact. Variance estimates were derived by using natural log-transformed data from a single pretreatment determination of fasting and post-methionine-loading homocysteine levels. On the basis of these data, with 14 patients in each main treatment group, our study had 80% power at a two-tailed α level of 0.05 to demonstrate a 35% reduction in the post-methionine-loading increase in plasma total homocysteine levels with vitamin B₆ treatment and 90% power at a two-tailed α level of 0.05 to show a 35% reduction in fasting plasma total homocysteine levels with folic acid plus vitamin B (₁₂) treatment.

All laboratory analyte values reported are based on averages of two pretreatment and post-treatment values, and all skewed variables were appropriately transformed. Baseline continuous variables were compared by using unpaired t-tests, and categorical variables were compared by using Fisher exact tests. Treatment effects for percentage changes in fasting total homocysteine levels and post-methionine-loading increases in total homocysteine levels were presented as ([average pretreatment level − average post-treatment level] ÷ average pretreatment level) × 100 and were compared by using unpaired t-tests. General linear modeling with analysis of covariance was performed to assess the independent effect of combined folic acid and vitamin B₁₂ treatment or vitamin B₆ treatment on fasting total homocysteine levels and the post-methionine-loading increase in total homocysteine levels; adjustments were made for age; sex; and pretreatment fasting total homocysteine levels, post-methionine-loading increases in total homocysteine levels, pyridoxal 5′-phosphate levels, folate levels, vitamin B₁₂ levels, and creatinine levels. Reported P values were based on two-tailed calculations. All statistical analyses were performed by using SYSTAT software (version 6.0.1, SPS, Chicago, Illinois).

The funding source had no role in the gathering, analysis, or interpretation of the data or in the decision to submit the manuscript for publication.

Results

As shown in Table 1, block randomization was successful with respect to both main-effect treatment assignments. The one exception was age in the vitamin B₆ treatment group. All patients completed the entire study protocol. Average compliance by pill count was 97.2%, a finding confirmed by marked increases (≥ 680%; P < 0.001) in the geometric-mean levels of both folate and pyridoxal 5′-phosphate. Independence of the hypothesized main a priori treatment effects is evident in the individual group (7 patients per group) percentage changes in fasting and post-methionine-loading increase in plasma total homocysteine levels (Table 2). Any vitamin B₆ treatment (14 patients) resulted in a 22.1% reduction in the post-methionine-loading increase in plasma total homocysteine levels (P = 0.049), whereas any combined folic acid and vitamin B₁₂ treatment (14 patients) caused a 26.2% reduction in fasting plasma total homocysteine levels (P = 0.004). Analysis of covariance revealed that both of these main treatment effects persisted (P = 0.042 for vitamin B₆ treatment; P = 0.027 for folic acid plus vitamin B₁₂ treatment) after adjustment for age; sex; and pretreatment levels of post-methionine-loading or fasting plasma total homocysteine, plasma B vitamins, and serum creatinine. Finally, when we used 90th-percentile cut points previously reported [3] from matched controls without renal disease for fasting plasma total homocysteine levels and the post-methionine-loading increase in plasma total homocysteine levels, 6 of 9 patients in the active-folic acid plus vitamin B₁₂ group compared with 1 of 5 patients in the placebo-folic acid plus vitamin B₁₂ group had baseline fasting total homocysteine levels reduced to less than 14 µmol/L; 3 of 3 patients in the active-vitamin B₆ group compared with 0 of 4 patients in the placebo-vitamin B₆ group had their baseline post-methionine-loading increase in total homocysteine levels reduced to less than 26 µmol/L.

Discussion

These data are the first placebo-controlled evidence in any patient population that vitamin B₆ treatment independently reduces the post-methionine-loading increase in plasma total homocysteine levels. Our findings are consistent with those of three open-label, uncontrolled investigations [10-12] that examined the effect of vitamin B₆ treatment on post-methionine-loading homocysteine levels in persons without renal disease. We also provide placebo-controlled confirmation of an earlier uncontrolled study [6] showing that combined folic acid and vitamin B₁₂ treatment decreases fasting total homocysteine levels in renal transplant recipients. We evaluated only 29 patients, however, and the uncertainty associated with this modest sample size remains an important limitation. Investigation of more renal transplant recipients undergoing longer treatment is required to confirm the external validity of our findings.

Selhub and Miller [13] hypothesized that two forms of hyperhomocysteinemia result from derangements in distinct metabolic pathways. Impairment of the remethylation pathway, due primarily to inadequate folate or vitamin B₁₂ status, results in hyperhomocysteinemia under fasting conditions. Conversely, impairment of the transsulfuration pathway is associated with normal or only mildly elevated homocysteine levels under fasting conditions but with substantial elevations after methionine loading. This hypothesis is supported by data from a study in folate-deficient animals [14] and findings from adult humans with severe vitamin B₁₂ deficiency that revealed marked fasting hyperhomocysteinemia but no abnormal post-methionine-loading increase in homocysteine levels above those seen during fasting [10]. A study of vitamin B₆-deficient animals [15], in which fasting homocysteine levels were normal but methionine loading resulted in a dramatic increase in homocysteine above fasting levels, has also recently provided confirmatory data. Our cross-sectional data from renal transplant recipients [3] and the effects of the placebo-controlled B-vitamin homocysteine-decreasing treatments in our present study are consistent with these earlier observations [10, 13-15]. In particular, the unique finding that vitamin B₆ treatment independently reduced the post-methionine-loading increase in total homocysteine levels provides strong evidence that improvement of suboptimal vitamin B₆ status can result in enhanced transsulfuration of homocysteine.

A large, multicenter, European case–control study recently showed that post-methionine-loading hyperhomocysteinemia confers a risk for vascular disease that is equivalent in magnitude to and is independent of fasting hyperhomocysteinemia [16]. The simultaneous occurrence of both fasting and post-methionine-loading hyperhomocysteinemia was associated with a level of risk for vascular disease that was greater than the additive individual risk for fasting or post-methionine-loading hyperhomocysteinemia occurring in isolation [16]. Combined fasting and post-methionine-loading hyperhomocysteinemia is a common finding (prevalence, approximately 30%) in stable renal transplant recipients [3] that may contribute to the excess risk for arteriosclerotic outcomes in these patients.

We conclude that vitamin B₆ should be added to the combination of folic acid and vitamin B₁₂ for effective reduction of both fasting and post-methionine-loading homocysteine levels in renal transplant recipients and, possibly, other patient groups. Vitamin B₆, as well as folic acid and vitamin B₁₂, should be included in the treatment regimens of clinical trials designed to test the hypothesis that decreasing homocysteine levels reduces the rate of arteriosclerotic outcomes in renal transplant recipients.

Drs. Gohh and Beaulieu: Division of Renal Diseases, Rhode Island Hospital, 593 Eddy Street, Providence, RI 02903.

Ms. Nadeau and Drs. Jacques, Selhub, and Rosenberg: Vitamin Bioavailability Laboratory, The Jean Mayer USDA Human Nutrition Research Center, 711 Washington Street, Boston, MA 02111.

Dr. Hume: Department of Family Medicine, Memorial Hospital of Rhode Island, 111 Brewster Street, Pawtucket, RI 02860.