Classical homocystinuria (HCU; McKusick 236200) was first described in 1962 simultaneously by Carson and colleagues in Northern Ireland1 and Gerritsen and colleagues in Wisconsin, USA.2 Within 2 years of the condition being described, Mudd and colleagues demonstrated a deficiency of hepatic cystathionine β-synthase (CβS; EC 188.8.131.52), a pyridoxine-dependent enzyme in the methionine trans-sulphuration pathway.3 This recessively inherited condition4 is the second most treatable aminoacidopathy after phenylketonuria (McKusick 261600).
Homocysteine lies at a very important metabolic branch point between the trans-sulphuration and remethylation pathways of methionine metabolism. Cofactors involved in the metabolism of methionine include pyridoxine (vitamin B6), vitamin B12 and folate. CβS deficiency is characterized by homocystine in the urine, homocystinuria, and biochemically by severe hyperhomocysteinaemia, hypermethioninaemia and hypocysteinaemia (Figure 1). Approximately 50% of patients worldwide are biochemically responsive to pharmacological doses of pyridoxine.5
Homocystinuria, as we now know, may be caused by CβS deficiency and a variety of very rare other genetic problems that interfere with tetrahydrofolate-dependent or methylcobalamin-dependent conversion of homocysteine to methionine. Therefore, the use of the term ‘homocystinuria’ must be appropriately qualified by the genotypic cause of it. Classical HCU has been historically used to describe CβS deficiency and, for the purpose of this article, HCU is used synonymously with CβS deficiency.
The purpose of this review is to briefly outline the general background to CβS deficiency and the modalities of treatment available for severe hyperhomocysteinaemia in patients with CβS deficiency, to ascertain the biochemical control achieved and to comprehensively review the effects of homocysteine-lowering therapy on the long-term clinical outcome based on published long-term data in an Irish cohort of newborn screened HCU patients.
The worldwide prevalence has been reported at 1 in 344 0006 compared with a much higher incidence in Ireland of 1 in 65 000.7 This prevalence based on newborn screening and clinical ascertainment has been underdetermined due to undetected cases,8 consisting mainly of pyridoxine-non-responsive patients.
Recent reports from several Western countries, based on molecular analysis of CβS mutations, suggest a much higher incidence at about 1 in 6500–20 000.9–12 Studies from the Saudi Arabian countries, particularly in Qatar where there is a high rate of consanguinity (80%),13,14 show that the incidence is highest there, at 1 in 1800, based on molecular and biochemical screening.15
Du Vignead first coined the terms ‘homocysteine’ and ‘homocystine’ in 1932 to represent the sulphydryl (reduced) and the disulphide (oxidized) forms of the next-higher homologues of cysteine and cystine16 (Figure 2).17 Since then, homocysteine nomenclature has been confusing rather than helpful. Mudd and Levy6 pointed out that it was crucial for all workers concerned with homocysteine to be familiar with the long-established nomenclature so as to avoid obscuring or confusing it.
A consensus statement on homocysteine terminology was published by Mudd and colleagues in 2000.18 In healthy subjects, 80–90% of homocysteine exists protein bound. The remaining free or non-protein-bound fraction (10–20%) consists of homocysteine–cysteine mixed disulphide and homocystine (Hcy–Hcy), the disulphide. The total free (non-protein-bound) homocysteine is taken as twice homocystine plus homocysteine–cysteine mixed disulphide. In vivo, only < 2% exists as the reduced form, homocysteine. Together, all these moieties make up what is called total homocysteine (tHcy).18
Genetics of cystathionine β-synthase deficiency
In 1984, the human CβS gene was mapped onto chromosome 21,19 and by in situ hybridization in 1988 it was specifically assigned to chromosome 21q22.3.20 The human complementary deoxyribonucleic acid (cDNA) for CβS was cloned and fully sequenced in 1993, with its transcribed sequence broken up into 18 exons and 17 introns.21
There are currently more than 150 known disease-causing point mutations or short deletions or insertions.22 The two most common mutations responsible for the homocystinuria phenotype include c.919G>A (p.G307S), also known as the ‘Celtic’ mutation, which heralds the more severe pyridoxinenon-responsive phenotype and is the major mutation in patients with Irish ancestry.23 The other mutation, c.833T>C (p.1278T), is associated with pyridoxine responsiveness and is a pan-ethnic mutation detected in the majority of European populations.24,25,26
Among the Qatari tribes M and K with HCU, c.1006C>T (p.R336C) is the only CβS mutation found, thus illustrating a strong founder effect causing a high prevalence of an autosomal recessive disease in a highly consanguineous population.14
Individuals with HCU are clinically normal at birth. If untreated, HCU is accompanied by an abundance and variety of clinical and devastating pathological abnormalities manifesting in four organ systems: eyes, skeleton, vascular and central nervous systems. The clinical picture is extremely heterogeneous, ranging from patients presenting with all of the complications to individuals with no overt clinical involvement.
In 1985, Mudd et al.5 carried out a landmark study documenting the natural history of 629 HCU patients due to CβS deficiency. The data obtained resulted in time-to-event graphs for each of the major recognized complications, namely ectopia lentis, radiographical osteoporosis, clinically detected thromboembolic events and mental retardation. Of the untreated pyridoxine non-responsive patients, these time-to-event graphs predicted that:
82% will have ectopia lentis by the age of 10 years
27% will have had a clinically detected thromboembolic event by the age of 15 years
64% will have radiological evidence of spinal osteoporosis by the age of 15 years and
23% will not survive to the age of 30 years.
These well-recognized time-to-event graphs for the first time provided the clinician with a prognostic probability for the development of each stated complications when a patient remains untreated. More importantly, the result of this landmark study also establishes baselines for future evaluation of the effects of treatment in this disease.
The cardinal biochemical features of HCU are:
markedly increased concentrations of plasma homocysteine (> 100 μmol/l) and methionine with a low cysteine concentration and
increased concentration of urine homocystine.
The diagnosis can be confirmed by measuring CβS enzyme activity in cultured skin fibroblasts.27 Residual CβS activity in cultured fibroblasts of affected individuals has ranged from none detected up to 10% of mean control activity. There is no correlation between the amount of residual enzyme activity and clinical pyridoxine responsiveness. In addition, in vitro response to pyridoxine of the isolated enzyme does not always correlate with an in vivo response.26 Therefore, it is recommended that all patients diagnosed with HCU must have a pyridoxine trial to ascertain clinical pyridoxine responsiveness before the actual commencement of homocysteine-lowering therapy.
Molecular CβS gene testing is now readily available. The sensitivity of the methodology used should be known and it is important to differentiate between mutation scanning, mutation screening and direct sequencing of the CβS gene when interpreting molecular reports.
Clinical aims of treatment
The clinical aims of treatment in HCU depend on the age at diagnosis. If CβS deficiency is diagnosed in the newborn infant through newborn screening, as ideally it should be, the aim must be to prevent the development of ocular, skeletal, intravascular and thromboembolic complications and to ensure the development of normal intelligence. On the other hand, if the diagnosis is made late when some recognized complications have already occurred, then the clinician's goal must be to prevent life-endangering thromboembolic events and to prevent further escalation of the complications already suffered.28
To achieve these clinical aims in treatment, one must try to control, and if possible normalize, the severe hyperhomocysteinaemia that is characteristic of this condition.28
Modalities of treatment
There are currently three recognized modalities of treatment for CβS deficiency:
pyridoxine (vitamin B6), in combination with folic acid and/or vitamin B12
methionine-restricted, cysteine-supplemented diet and
For those who are vitamin responsive, pharmacological doses of pyridoxine in combination with folic acid and/or vitamin B12 is the treatment of choice.31,32 For vitamin non-responders, the treatment is with a methionine-restricted, cysteine-supplemented diet.31–34 Pyridoxine, folic acid and vitamin B12 have been continued in pyridoxine non-responders as cofactors of methionine metabolism. More recently, betaine, a methyl donor that remethylates homocysteine to methionine, has been used as an adjunct to treatment.35
Pyridoxine trial – ascertainment for clinical pyridoxine responsiveness
When the diagnosis is confirmed with the findings of hyperhomocysteinaemia, hypermethioninaemia and hypocystinaemia, the newborn is admitted for treatment and stabilization. Pyridoxine 50 mg t.i.d. is prescribed to assess vitamin responsiveness with deproteinized amino acids monitoring every 3 days initially until stabilized.29 Biochemically, vitamin responsiveness is indicated by falling homocysteine and methionine levels while remaining on pyridoxine, in the presence of adequate vitamin B12 and folate. While on pyridoxine treatment, the newborn is deemed vitamin responsive when the free homocystine is < 5 μmol/l. However, if the free homocystine and methionine remains persistently elevated or rises while on pyridoxine, the newborn is biochemically pyridoxine non-responsive and is commenced on a methionine-restricted diet.
The status of vitamin responsiveness in a neonate may be difficult to determine on biochemistry alone. While on a trial of pyridoxine, falling homocysteine and methionine levels due to coincidental growth spurt may be mistakenly attributed to vitamin responsiveness.
The dose of pyridoxine used for a newly diagnosed adult patient is 200 mg t.i.d. It is prudent to point out that adequate response to pyridoxine may be masked in the presence of inadequate folate and vitamin B12.30
The pyridoxine dose required to achieve successful biochemical control is titrated against plasma homocysteine levels. The optimal dose is reached when, at a minimum dose of pyridoxine, the plasma homocysteine is at its lowest. Doses of pyridoxine required for a response vary markedly among pyridoxine responders. Barber and Spaeth36 achieved responses to doses of 250–500 mg/day, whereas Gaull et al.37 utilized 800–1200 mg/day of pyridoxine for a similar response. Pyridoxine 500–1000 mg/day should be given for several weeks before a patient is considered unresponsive.38
For neonates diagnosed by newborn screening, the Irish group used a dose of 150 mg/day in three divided doses of pyridoxine and a dose of 100–800 mg/day for adult patients.29 The Dutch group used dosages of 200–500 mg/day of pyridoxine in children and for adults a dose of 750 mg/day.39 The Australian group used a lower dose of 200 mg/day pyridoxine.31
Clinicians have always been mindful of patient safety in the usage of these megadoses of pyridoxine. Vitamin B6 at doses from 200 mg to 6 g has sporadically been considered as a cause of sensory neuropathy.40,41 Remarkably, these side effects have never been observed in patients with homocystinuria who received long-term treatment (up to 24 years) with dosages up to 750 mg.42,43 This is in line with the experience of the three centres, which reported a total of 1314 patient-years of pyridoxine treatment in these patients.29,31,39
A biochemical response to pyridoxine may not occur in a potentially responsive patient if folate depletion is present.44 Patients while on pyridoxine become folate depleted unless supplemented, probably as a consequence of increased flux through the remethylation pathway. The dose of folic acid generally used has been 5 mg/day in adults. Furthermore, to avoid folate refractoriness, it is mandatory to prevent deficiency of vitamin B12, the cofactor in folate-mediated remethylation of homocysteine to methionine.
For those who are responsive to pyridoxine in combination with folic acid and vitamin B12, it remains debatable whether or not long-term treatment with other measures such as methionine-restricted, cysteine-supplemented diet or betaine supplementation provide the optimal biochemical control. Even patients with maximal responsiveness to pyridoxine in the presence of adequate folate and vitamin B12 have reduced tolerance to methionine during methionine loading tests. Pyridoxine in these patients corrects only the biochemical abnormalities in the fasting state.45 Such patients, in theory, may experience abnormal episodic surges in methionine or homocysteine following protein ingestion.38 Hence, it follows that these patients may benefit from some methionine restriction or the use of small frequent feedings.45 It is good clinical practice to give general advice to these patients on reducing the intake of food containing methionine or at least distributing the methionine intake during the day.
Methionine-restricted, cysteine-supplemented diet
A methionine-restricted, cysteine-supplemented diet to correct biochemical abnormalities is used in the treatment of pyridoxine-non-responsive homocystinuria. It was initially devised by Komrower33 and Perry in 196634 and has been extensively reviewed elsewhere.32
The Irish group, which comprises predominantly pyridoxine non-responders detected through newborn screening, provides a comprehensive long-term experience on methionine-restricted, cysteine-supplemented diet.46 The dietary treatment for CβS deficiency consists of providing the components discussed in the following four sections.
Natural or dietary protein is given in a small amount to provide sufficient methionine, an essential amino acid, for normal growth and development. The amount given is measured in ‘exchanges’ where one exchange contains 1 g of protein (25 mg of methionine). The number of exchanges varies with time in a single patient and also between patients, as this factor is dependent on the methionine tolerance of the individual patient.46 There are increased natural protein requirements during periods of growth spurt in children and during pregnancy as the normal fetus grows, while still maintaining good biochemical control.47 Conversely, the requirement is reduced during periods of intercurrent illnesses and in conditions in which there is catabolic stress.
The aim of treatment is to maintain a plasma free homocystine of < 5 μmol/l and a methionine level of 50–70 μmol/l. In the treatment of these newborns and growing children, giving particular attention to the provision of sufficient methionine will prevent growth retardation due to a deficiency while maintaining good biochemical control.46
As the diet restricts methionine intake, it severely curtails the intake of natural or dietary protein. Hence, it is necessary to provide a synthetic protein supplement depleted of methionine in order to enable normal growth and development. This supplement is enriched with cysteine, which becomes an essential amino acid in CβS deficiency. Several of these methionine-free, cysteine-enriched amino acid mixtures are available commercially for use in patients with β deficiency. Some of these proprietary mixtures are complete with vitamins and minerals, while others are not. When using a mixture that is not complete, supplementation with vitamins and minerals mixes is necessary to prevent deficiencies.
‘Free foods’ – non-protein/methionine-containing food
‘Free foods’ are natural foods that do not contain protein or methionine, for example many vegetables and fruits. These are permitted in unlimited quantities and contribute to the caloric intake of the patient. They also provide variety in the diet.
These are commercially available products that are low in protein and obtainable on prescription, for example low-protein bread, pasta, biscuits and flour. They are allowed in unlimited amounts and are considered as ‘free foods’. They play an important part in the provision of calories and variety for the patient.
The total protein intake is the sum of the natural and synthetic protein intakes. The amount of synthetic protein prescribed is inversely proportional to age and calculated per kilogram of body weight.48
Vitamins and minerals are prescribed to meet the recommended dietary allowances.49 Caloric intakes are based similar to the recommended dietary allowances.49 The provision of adequate calories to meet the energy expenditure of the patient is important in minimizing catabolism and hence hyperhomocysteinaemia.
Betaine is used in late-detected pyridoxine-non-responsive individuals for whom such a diet may be unpalatable and dietary compliance difficult to achieve.35,50–54 It may be used as an adjunct to dietary restriction to obtain the best possible biochemical control.32 Anhydrous betaine is usually prescribed at 3–9 g/day in two or three divided doses in adults;31,50,53 however, the optimal dosing has not been determined.55
Additional treatment with betaine provides an alternative remethylation pathway to convert excess homocysteine to methionine, usually accompanied by a rise in plasma methionine. The resultant hypermethioninaemia has been harmless in the vast majority of cases with 226 patient-years of betaine treatment.32 However, cerebral oedema has been reported in a few cases with hypermethioninaemia > 1000 μmol/l.56–58
Additional adjunct therapy
Additional medical interventions that do not affect the biochemical abnormalities but aim to reduce or eliminate the thrombotic tendencies have also been used.6 These include the avoidance of situations that predispose to an increased risk of thromboembolism and the use of high-dose antithrombotic prophylaxis with heparin or warfarin in conditions with prolonged bed rest. Aspirin, either alone or in combination with dipyridamole, has been used to normalize decreased platelet survival and minimize vascular lesions in patients with HCU.32
Therapeutic aim and biochemical control
The therapeutic aim while receiving homocysteine-lowering treatment has been reported by several centres in the past. The different centres have used different homocysteine fractions as their measurement for biochemical control. It is important to point out that, prior to 1990, measurement of total homocysteine was not possible.
The therapeutic aim for the Australian and Dutch groups was a total free homocysteine (non-protein- bound) level < 20 μmol/l, and since 1990 the Dutch group has taken a total homocysteine level < 50 μmol/l as achieving good biochemical control.31,39 The Irish group has used a target level of free homocystine < 5 μmol/l.29 The London and Manchester workers59 have used a level of free homocystine < 10 μmol/l. Together, these five centres provided a total of 2821.6 patient-years of treatment on 158 patients.60
In a newborn screened group of patients, Yap and Naughten29 defined a lifetime median of free homocystine ≤ 11 μmol/l as protective, at least up to 23.4 years, against the overt recognized complications of untreated HCU.
Long-term clinical outcome
From the ever-increasing understanding of the pathophysiology of CβS deficiency, it is known that the characteristic biochemical abnormalities are responsible for the clinical complications. Optimal treatment strategies have been aimed at reducing or normalizing these biochemical abnormalities as previously discussed. It is reasonable to assume that progressive tissue damage has occurred even prior to the appearance of any clinical signs and symptoms. Hence, maximal benefit from homocysteine-lowering therapy may only be derived from detecting the disorder in the newborn period, either by newborn screening or by high-risk screening as a result of a known family history of the disorder.26
The effectiveness of homocysteine-lowering treatment in late-detected patients has been established in several studies.5,31,39 There remains a paucity in the literature of long-term clinical outcome data on HCU patients treated from birth, as few countries screened for β deficiency. Of the increasing numbers of countries screening for HCU, the newborn screened cohort of patients remain relatively young for long-term clinical outcome assessment.
Uniquely, Ireland provides the largest cohort of newborn-screened, early treated and systematically followed up group of HCU patients with one of the longest terms of follow-up in the world.29 The following paragraph will review the long-term clinical outcome data published by Yap on the Irish HCU patients, detected through newborn screening and systematically followed up in a single centre for at least 25 years of the Irish national newborn screening programme since 1971.
The Irish group of 25 HCU patients had a total of 365.7 patient-years of treatment.29 Of the group, 21 (age range 2.5–23.4 years) were detected through newborn screening and commenced treatment within 6 weeks of birth. The remaining four (age range 16.8–20.8 years) were late detected; three were breastfed and one was a vitamin B6 responder. All except one patient were vitamin B6 non-responsive. In the 18 patients screened (mean period of follow-up 14.3 years), early dietary-treated patients who were compliant with treatment, there was no ectopia lentis, no radiological evidence of osteoporosis, no thromboembolic events and no mental handicap with age-appropriate education standards. The remaining three newborn-screened patients (age 12, 16.6, 22.8 years) who were poorly compliant with treatment developed ectopia lentis, osteoporosis and intellectual disabilities, despite the opportunity of newborn screening and an early start to treatment. This study, with 365.7 patient-years of treatment, provided the first evidence that newborn screening, early treatment and good dietary compliance, maintaining a lifetime free homocystine median of ≤ 11 μmol/l, is effective in preventing complications.29
Intellectual and neurological outcome
Intellectual disability remains the most frequent abnormality of the central nervous system and is often the first recognizable sign of CβS deficiency, presenting as developmental delay during the first and second years of life.5 It is, however, not at all a consistent finding, particularly among vitamin B6 responders. Mudd et al.5 documented a very wide range of patients’ intelligence quotients (IQs) (n = 284) from 10 to 138. Vitamin B6 non-responders (n = 115; mean IQ = 57) generally fared significantly worse than vitamin B6 responders (n = 107; mean IQ = 79). The study further documented that only 4% of vitamin B6 non-responders had an IQ ≥ 90 and, when compared with their unaffected siblings, 94% had IQs < 90 and only 6% were comparable.5
General experience with CβS-deficient patients has shown that late treatment rarely, if ever, completely reverses mental impairment,5 although treatment may lead to behavioural improvement.62 Early treatment might, on the other hand, prevent mental damage, but few reports on this exist. The study by Yap et al.63 documented a mean full-scale IQ (FIQ) of 105.8 (range 84–120) in 13 early-treated vitamin B6 non-responders with a total of 187.7 patient-years of treatment, which was significantly better than that in the late-detected or untreated vitamin B6 non-responders (Figure 3). When further compared with their unaffected sibling controls (n = 8), there was no evidence of significant differences in the psychometric parameters measured, except for a higher patient FIQ (P = 0.0397). The study further showed that late treatment is better than no treatment at all in terms of intellectual outcome and that the earlier the commencement of treatment occurs, the better the outcome.
Besides intellectual disabilities, other neurological complications as a result of untreated HCU may include seizures, psychiatric disturbances and dystonia. Few studies have concentrated on the neuropathology and imaging in CβS-deficient patients. The central pathological findings have been described as arteriosclerosis, venous and arterial thrombosis and multiple small infarcts of different ages throughout the brain.64 The corresponding magnetic resonance imaging (MRI) images of cerebral infarction are non-specific; only the young age at presentation is unusual.65 Movement disorders may result from neurochemical changes in the basal ganglia with no abnormalities on computerized tomography (CT)/MRI.66
The Irish study on MRI findings and neurological outcome showed that all newborn-screened patients with good control, vitamin B6 non-responsive HCU patients (n = 8; age > 15 years) had normal MRI of the brain and normal neurological examinations.66,67 The remaining poorly controlled newborn-screened and late-detected patients (total = 15) were found to have a combination of microinfarcts on MRI of the brain and abnormal neurology. However, none had peripheral neuropathy.66,67
Ectopia lentis is a recognized complication of untreated HCU. The natural history for untreated vitamin B6 non-responders is such that 82% will have dislocated lens by age 10 years.5 Mudd et al.5 demonstrated a two-year lag period before appreciable lens dislocation occurred. Cruysberg et al.68 emphasized the diagnostic significance of high degrees of progressive myopia at a young age, due to lens subluxation, in HCU patients. Taylor et al.69 further reported on the detrimental effect of late diagnosis on best corrected visual acuity (BCVA) in patients with HCU.
The Irish ocular outcome studies showed no ocular pathology and a vision of 20/20 bilaterally in the newborn-screened patients (n = 15; 250 patient-years of treatment) with good biochemical control. This was significantly different when compared with the late-detected group (n = 14); P < 0.0001). The poorly controlled newborn-screened patients (n = 6) were also significantly more myopic than the well-controlled newborn-screened group (P < 0.0001). This group of 40 HCU cases studied had a total of 564 patient-years of treatment.70 A further study by the Irish group demonstrated that HCU patients (n = 14; 12 newborn screened) with no ocular pathology had ocular axial lengths comparable with the normal adult eye. It also showed that homocystinuria patients with complete lens dislocation in at least one eye have significantly increased ocular axial length, and this had never been demonstrated before.71
A variety of skeletal abnormalities have been associated with untreated HCU. Kyphoscoliosis, dolichosternomelia, pectus carinatum and genu valgum have all been described. However, osteoporosis remains the most consistent skeletal change, most commonly found in the spine followed by long bones, in untreated patients with HCU.5 The natural history of osteoporosis is such that 50% will have radiological evidence of spinal osteoporosis by age 15 years, and this further increases to 80% by age 30 years.5 Parrot et al.72 reported that spinal and femoral osteoporosis was evident by dual-energy X-ray absorptiometry (DEXA) scans in all of their six late-detected HCU patients (five vitamin B6 non-responsive), which concurs with the natural history. Deficient collagen cross-linking has been hypothesized as the pathophysiology for the development of osteoporosis in β deficiency.73 Lubec and colleagues in 1996 provided further data showing normal plasma parameters for collagen type I and type III with reduced collagen cross-linkers to support this hypothesis.74 Osteoporosis as a complication of untreated deficiency has been documented in several studies with few on the effects of treatment.
The Irish group has previously reported no radiological evidence of osteoporosis in their newborn-screened, good-control patient cohort.29 Using DEXA scans for the assessment of osteoporosis, Yap et al.47,75 studied a total of 18 HCU patients with an age range of 15.6–35 years. All the newborn-screened, well-controlled patients (n = 6) showed normal bone mineral densities. Of the remaining 12 patients, five were newborn-screened, poorly controlled patients and seven were late-detected patients; two developed osteoporosis, one from each of these groups. According to the natural history of 50% developing radiological osteoporosis by age 15 years, at least 9 of these 18 patients studied would have developed osteoporosis if they had remained untreated. Instead, only two patients had osteoporosis based on the World Health Organization (WHO) criteria for osteoporosis despite a more sensitive test (P = 0.0275).47,75
Thromboembolic events are the major cause of morbidity and mortality for untreated HCU. The vascular risk is such that half of the patients will have a vascular event before age 30 years and there is predicted to be one event per 25 years at the time of maximal risk.5 Thus, this landmark study established baselines for the evaluation of the effect of treatment and provided suggestive evidence for benefit.5
Up until the late 1990s, effects of treatment on vascular outcomes had initially been reported in single-centre studies on patients with CβS deficiency.29,31 Treatment was shown to be effective not only in patients with isolated CβS deficiency but also in those patients who carried the additional factor V Leiden, a recognized thrombotic risk factor.76 The findings were similar when patients from two additional centres were studied.28 The total number of patients studied thus far, however, remains comparatively small.
An international multicentre collaborative study was carried out to specifically document vascular outcomes and explore the effectiveness of homocysteine-lowering treatment on cardiovascular risk in patients with CβS deficiency.60 This study included 158 patients with 2821.6 patient-years of treatment from five centres: Sydney, Nijmegen, Dublin, Manchester and London. All participating centres have used re cognized treatment regimens with minor modifications.
The natural history predicts one vascular event per 25 years,5 and thus in 2821.6 patient-years of treatment at least 112 vascular events would have occurred if these patients had remained untreated. Instead, only 17 vascular events were recorded in 12 treated patients. This difference is highly significant, with a relative risk of 0.09 [95% confidence interval (CI) 0.036 to 0.228; P < 0.0001]. There was also no evidence of non-homogeneity in the vascular outcomes of each centre despite the minor differences in the treatment regimens used. The documented vascular events included pulmonary embolism, sagittal sinus thrombosis, deep-vein thrombosis, cerebrovascular accidents, abdominal aortic aneurysm, transient ischaemic attacks and myocardial infarction.60 The evidence from this multicentre study conclusively shows that chronic homocysteine-lowering treatment is effective in reducing the vascular risk in HCU patients.
Biochemical control during chronic treatment was also documented and expressed as the mean [± standard deviation (SD)] level of plasma homocysteine.60 Despite the aim of achieving normal levels, the post-treatment levels were several times higher than the mean for the corresponding normal population of each centre (Table 1). The authors concluded that this finding may have relevance to the concept of mild hyperhomocysteinaemia and its supposed associated cardiovascular risk.
|Centres||Mean (± SD) levels of homocystine during chronic treatment (μmol/l)|
|Free Hcy a,b||Total free Hcy [3.3–3.7 ± 1]a,b||Total Hcy [15–19]a,b|
|Sydney||33 ± 17.0|
|Nijmegen||34 ± 13.7||88 ± 34.5|
|Dublin||16.7 ± 11.7||108 ± 55.0|
|Manchester||31.4 ± 35.9|
|London||33.3 ± 29.9|
Undoubtedly, homocystinuria due to CβS deficiency causes an array of devastating complications when left untreated. Many studies have shown over the years that homocysteine-lowering treatment is effective in ameliorating complications and preventing the development of new complications in late-detected patients.
Half a century has passed since the first discovery of HCU in 1962. Much knowledge has been gained into the pathophysiology of this condition but clinical studies into the effectiveness of treatment still lag behind. In this last decade, Yap et al. have systematically documented clinical end points in a group of Irish newborn-screened and treated patients with pyridoxine-non-responsive HCU. The published data, which are reviewed in this paper, comprehensively documented that early homocysteine-lowering treatment is effective in preventing all the recognized complications of untreated HCU. It was shown that good clinical outcomes could be achieved without completely normalizing the homocysteine levels.
Initially thought to be generally a rare, serious and even life-threatening condition, homocystinuria has been proven by molecular studies to be more common than initially assumed. Clinical studies have now confirmed that it is indeed a very treatable inherited condition. The data from the Irish studies point favourably towards advocating newborn screening for homocystinuria, whereby the patients diagnosed, with early treatment, will have the best possible normal clinical outcome.