Homozygous Cystathionine β-Synthase Deficiency, Combined With Factor V Leiden or Thermolabile Methylenetetrahydrofolate Reductase in the Risk of Venous Thrombosis

HOMOCYSTINURIA, characterized by severe hyperhomocysteinemia, may have its origin in deficiencies in enzymes involved in methionine metabolism.1-3 The concomitant finding of an extremely high risk of premature vascular disease in patients suffering from homocystinuria has led to the “homocysteine theory,”⁴ which hypothesizes that a severely elevated homocysteine concentration, irrespective of its cause, is associated with vascular pathology in arteriosclerosis and thrombosis.

The most frequent cause of homocystinuria is a homozygous deficiency of cystathionine β-synthase (CBS), the first enzyme in the transsulphuration pathway.5 This classical homocystinuria is, next to the life-threatening complications from the vascular system, characterized by ectopia lentis, skeletal abnormalities, and mental retardation.5 In a large international survey among patients with homocystinuria caused by CBS deficiency, Mudd et al6 observed one or more episodes of a thromboembolic event in 158 (25%) of 629 homocystinuria patients. These thromboembolic events can be subdivided in cerebrovascular accidents (31%), venous complications (51%, of which 12% represent pulmonary embolism), myocardial infarctions (4%), and peripheral arterial complications (11%). It could be calculated that each untreated patient had about a 50% chance to suffer from such an event under the age of 30 years.

Resistance to activated protein C (APC resistance) caused by a single R506Q mutation in the factor V gene, the so-called factor V Leiden (FVL) mutation,7 is the most predominant hereditary cause of venous thrombosis identified so far.8 Heterozygous individuals have approximately an eightfold higher risk on thrombosis,9 whereas the risk in homozygotes is increased about 80-fold.10 In the Physicians' Health Study including about 15,000 apparently healthy men, Ridker et al11reported an association also between this mutation and the occurrence of deep venous thrombosis after a mean follow-up of 8.6 years but not with myocardial infarction and stroke.

Only a few studies have investigated a possible interaction between hyperhomocysteinemia and APC resistance caused by FVL in the risk of thrombosis. In a recent study by Den Heijer et al,12,mild hyperhomocysteinemia was associated with an increased risk for a first occurrence of deep-vein thrombosis, but they did not observe an enhanced thrombotic risk in jointly affected patients. These results have been confirmed in a study of D'Angelo et al.13 On the other hand, Ridker et al14 noticed a substantially increased risk of developing future thrombosis in apparently healthy men with coexistent hyperhomocysteinemia and FVL. Insevere hyperhomocysteinemia data are conflicting. In seven highly consanguineous families including 45 patients with homocystinuria of different origins, Mandel et al15observed thrombosis in only those homocystinuria patients with concomitant heterozygous or homozygous FVL. These results were challenged by Quéré et al16 who reported the absence of such an interaction in 15 homocystinuria patients with genetically proven CBS deficiency from 13 unrelated kindreds.

A common homozygous 677C→T mutation in the methylenetetrahydrofolate reductase (MTHFR) gene has been shown to be a major cause of mildly elevated plasma homocysteine concentrations17,18 and has been investigated as a risk factor for venous thrombosis.19,20 The effect of the homozygous 677C→T mutation on the risk of deep venous thrombosis in CBS-deficient patients is unknown.

In the present study, we analyzed 24 homozygous CBS-deficient patients from 18 unrelated kindreds for FVL, and the 677C→T mutation in the MTHFR gene. We investigated whether a possible interaction between FVL, thermolabile MTHFR, and the severe hyperhomocysteinemia caused by CBS deficiency, leads to an enhancement of the excessive thrombotic risk in patients with joint abnormalities.

Twenty-four patients, 14 men and 10 women from 18 unrelated kindreds, with homocystinuria caused by CBS deficiency were studied for FVL and the MTHFR 677C→T polymorphism. Twenty-three patients were pyridoxine responsive. The diagnosis of homocystinuria caused by CBS deficiency in patients was made at a mean age of 24.7 years (range 4 to 54), by establishing severe hyperhomocysteinemia and homocystinuria, hypermethioninemia, and decreased levels of cysteine in plasma. Furthermore, CBS activities measured in extracts of cultured fibroblasts21,22 were less than 2% of the mean in controls, except in one patient, in whom we observed CBS activities in the heterozygous range. However, in this patient we were able to show a defective CBS regulation by S-adenosylmethionine (AdoMet) leading to severe hyperhomocysteinemia.23 Up to now, in 18 of 24 patients homozygous CBS deficiency was confirmed by molecular genetic analysis of the CBS cDNA18,23,24 (Kluijtmans et al, unpublished results).

Deep venous thrombosis was diagnosed by means of flebography and pulmonary embolism by pulmonary perfusion-ventilation scintigraphy.

DNA was isolated from peripheral blood lymphocytes by a standard method.25 FVL was investigated by allele-specific polymerase chain reaction amplification and capillary electrophoresis with on-line ultraviolet detection, as described by Van der Locht et al.26 Screening for the 677C→T polymorphism in the MTHFR gene was performed essentially according to Frosst et al.17 

Determination of homocystine, homocysteine-cysteine mixed disulfide, and methionine in serum of the homocystinuria patients has been performed as described earlier by us.27 The total amount of nonprotein-bound homocysteine was calculated as twice the concentration of homocystine plus the concentration of the homocysteine-cysteine mixed disulfide.

The differences in MTHFR genotype distributions and allele frequencies among homocystinuria patients versus Dutch controls28 have been assessed by (Yates corrected) χ² analyses. Differences in serum homocysteine and methionine concentrations in carriers versus noncarriers of the homozygous 677C→T mutation in the MTHFR gene have been assessed by nonparametric Wilcoxon-Mann-Whitney-U tests. All P values reported are two-tailed, and P < .05 was considered statistically significant.

Of 24 homocystinuria patients, 3 individuals, all belonging to the same kindred, were carriers of FVL; no homozygotes for FVL were observed. In the study group, 6 individuals, mean age, 23 years (range, 9 to 40), suffered from a thrombotic complication; 4 patients (3 men and 1 woman) had deep venous thrombosis and 2 patients (both women) had pulmonary embolism, whereas 18 homocystinuria patients remained free of venous thrombotic disease.

Venous thrombosis occurred in these 6 patients before the start of homocysteine-lowering treatment, which had been prescribed immediately after the diagnosis of homocystinuria had been established. All 6 patients with thrombosis proved to be vitamin B₆responders. Only one thrombotic event occurred in a homocystinuria patient with concomitant FVL. Five patients suffered from thrombosis without the presence of FVL (Table 1).

We also screened for the 677C→T polymorphism in the MTHFR gene in these patients. The overall allele frequency of the T-allele was 0.44, which is significantly different from the allele frequency (0.30) observed in Dutch controls (χ² = 3.88; DF = 1; P = .049). In our study group, we observed 5 (21%) homozygotes for the T-allele, 11 (46%) heterozygotes, and 8 (33%) wild-type (CC) individuals, which statistically tends to be significantly different from the genotype distribution in Dutch controls (χ² = 5.43; Degrees of Freedom (DF) = 2;P = .07). The relatively high frequency of the homozygous TT genotype among these homocystinuria patients was not caused by a high prevalence of TT genotype among siblings (data not shown). Three of six patients with thrombosis had the homozygous TT genotype (Table 1), one was heterozygous, whereas two patients had the CC genotype.

We assessed the serum nonprotein-bound homocysteine and methionine concentrations and the homocysteine/methionine ratio in the homocystinuria patients (n = 21); 5 patients with the concomitant homozygous TT genotype were compared with 16 heterozygous (CT) and wild-type (CC) individuals (Table 2). Mean serum homocysteine concentration in concomitant carriers of the TT genotype was elevated compared with those observed in individuals with CT and CC genotypes (P = .6). Methionine concentrations were less elevated in TT carriers, whereas the homocysteine/methionine ratio was slightly increased compared with noncarriers (P = .4 andP = .7, respectively). However, all these differences did not reach statistical significance because of the small number of patients.

In homocystinuria caused by CBS deficiency, venous thrombotic events are a common manifestation; about 13% of a large cohort of homocystinuria patients studied retrospectively suffered from one or more episodes of venous thrombosis.6 This observation, and the concomitant finding of a high risk for arteriosclerosis in homocystinuria, led to the conclusion that severely elevated plasma homocysteine concentrations constitute a risk factor for both arteriosclerosis and thrombosis.

Recently, Bertina et al7 described a point mutation in the factor V gene, an R506Q substitution, which renders factor V resistant to cleavage by activated protein C. Several studies have investigated either APC resistance,11,29-31 mild hyperhomocysteinemia,32-36 or thermolabile MTHFR19,20,37 as risk factors for venous thrombosis. Only a few reports have investigated a possible interaction between hyperhomocysteinemia or thermolabile MTHFR and FVL in the risk of thrombosis.12-16,20,37 

However, both in mild and severe hyperhomocysteinemia, data are confusingly conflicting. Very recently, a synergistic interaction between mild hyperhomocysteinemia and FVL was reported from the Physicians' Health Study, in which a 20-fold higher risk on thrombosis was calculated in jointly affected individuals compared with nonaffected subjects.14 These results were in contradiction with those reported by Den Heijer et al12 and D'Angelo et al.13 Cattaneo et al20 reported a synergistic interaction between thermolabile MTHFR and FVL in Italian thrombosis patients. However, the numbers of jointly affected individuals in all these studies were relatively small.

Mandel et al15 reported on an interaction between severe hyperhomocysteinemia and FVL. In their study they observed thrombotic events only in patients with both inherited anomalies, which led them to conclude that FVL is a prerequisite for the occurrence of thrombosis in homocystinuria patients. However, the results in their study may have been biased by the high degree of consanguinity in their families, which is reflected in the presentation of other rare inborn errors of metabolism, like phenylketonuria, lysinuric protein intolerance, and Immerslund-Gräsbeck syndrome in some of their pedigrees.15 

Our present study contradicts the hypothesis of Mandel et al.15 In our study group, five of six homocystinuria patients, who suffered from one or more thrombotic events, were not carriers of an FVL allele. The same results were obtained by Quéré et al,16 who also concluded that FVL is not an absolute requirement for the development of thrombosis in homozygous CBS-deficient patients. In their study, 5 of 15 homocystinuria patients had presented with venous thrombosis, of which only 2 individuals had FVL. Thus, 3 patients without the FVL mutation had had venous thrombosis.16 

A remarkable finding in the present study was the relatively high frequency of the MTHFR TT-genotype among our six homocystinuria patients who suffered from a thrombotic event. In this subgroup, three (50%) of six patients had the MTHFR TT-genotype, suggesting that, although the number of individuals is small, thermolabile MTHFR might be an important additional factor in the risk of thrombosis in homozygous CBS-deficient patients. In the 18 homocystinurics who did not present with thrombotic complications, the frequency of the MTHFR TT-genotype was approximately 10%, virtually identical to that observed in the Dutch population.28 This major contributing effect of the 677C→T mutation may be explained by the coordinate regulation of homocysteine metabolism by AdoMet.38,39 In homocystinuria caused by CBS deficiency, AdoMet concentrations are elevated,40 which results in an inhibition of MTHFR. In case of a combination with the homozygous 677C→T mutation, this joint effect will be an even more substantial reduction of homocysteine remethylation, theoretically leading to a further accumulation of serum homocysteine and an attenuation in the increase of serum methionine. The measurement of serum nonprotein-bound homocysteine and methionine in concomitant TT carriers and noncarriers supports this hypothesis: serum homocysteine was slightly higher, whereas serum methionine was less elevated in these patients with concomitant thermolabile MTHFR. However, probably because of the small sample sizes, these differences did not reach statistical significance.

In our study group, the three patients with FVL were siblings, and only one of them presented with thrombosis. Interestingly, the patient with thrombosis was a carrier of the MTHFR TT-genotype as well, in contrast to his two siblings. Institution of anticoagulation and homocysteine lowering treatment after diagnosis could have prevented the development of thrombosis in these two subjects. However, both individuals were untreated for a long period, 21 and 26 years, respectively, at which ages in five of six homocystinurics deep venous thrombosis had developed already. In this pedigree, thermolabile (TT) MTHFR might have been a critical contributing factor in the risk of thrombosis in the homocystinuric sibling with joint anomalies.

The authors greatly acknowledge Mia Willemsen and Kitty Verbeek for expert technical assistance.