CLIN. CHEM. 39/9, 1764-1779 (1993) Total Homocysteine in Plasma or Serum: Methods and Clinical Applications Per M. Ueland,”5 Allen2 Total homocysteine Helga Refsum,’ Sally P. Stabler,2 M. Rene Malinow,3 is defined as the sum of all homocys- teine species in plasma/serum, including free and proteinbound forms. In the present review, we compare and evaluate several techniques forthe determination of total homocystelne. Because these assays include the conversion of all forms into a single species by reduction, the redistribution between free and protein-bound homocysteine through disulfide interchange does not affect the results, and total homocysteine can be measured in stored samples. Total homocysteine in whole blood increases at room temperature because of a continuous production and release of homocysteine from blood cells, but artificial increase is low if the blood sample is centrifuged within 1 h of collection or placed on ice. Different methods correlate well, and values between 5 and 15 mol/L in fasting subjects are considered normal. Total homocysteine in serum/plasma is increased markedly in patients with cobalamin or folate deficiency, and decreases only when they are treated with the deficient vitamin. Total homocysteine is therefore of value for the diagnosis and follow-up of these deficiency states and may compensate for weaknesses of the traditional laboratory tests. In addition, total homocysteine is an independent risk factor for premature cardiovascular diseases. These disorders justify introduction of the total homocysteine assay in the routine clinical chemistry laboratory. IndexIng Terms: cobalamin, folate deficiencies . sample handling . reference values thiol compounds hew? disease enzyme metabolism amino acids nutritional status Homocysteine (Hey) determination was introduced into laboratory diagnosis in 1962 when the first patients with the inborn error homocystinuria were described (1, 2).6 These patients excrete large amounts of homocys‘Department of Pharmacology and Toxicology, Armauer Hansens Hus, University of Bergen, N-5021 Bergen, Norway. Fax 47. 5-973115. 2Division of Hematology, Department of Medicine, University of Colorado Health Sciences Center, Denver, CO. 3Oregon Regional Primate Research Center, 505 NW 185th Ave., Beaverton, OR 4Department of Clinical Chemistry, University Hospital, S-22 185 Lund, Sweden. 5Author for correspondence. 6Nonstandard abbreviations: ABD-F, 4-(aminosulfonyl)-7.fluoro-2,1,3.benzoxadiazole-4-su]fonate; GC-MS, gas chromatography-mass spectrometry; Hcy, homocysteine; mBrB, monobromobimane; and SBD-F, ammonium-7-fluoro-2,1,3-benzoxadiazole-4sulfonate. Received December 14, 1992; accepted March 24, 1993. 1764 CUNICAL CHEMISTRY, Vol. 39, No. 9, 1993 Anders Andersson,4 and Robert H. tine into the urine, and the blood concentrations become extremely high (3). The high concentrations could be determined by simple chemical tests (4) or by amino acid analysis. Hey in the acid-soluble fraction of plasma/serum (free Hey) was detected in healthy subjects with the secondgeneration amino acid analyzers that became available in the middle 1970s (5, 6). The first clinical studies on the relation between moderately increased plasma Hey and increased risk for cardiovascular disease, published in the late 1970s and early 1980s, were based on this methodology (7-9). Progress in Hcy research during the last 7 years has been greatly facilitated by the introduction of improved techniques for measuring Hey in plasma and serum (10-19). These methods measure total Hey, which is the sum of protein-bound and free Hey. The main advantage is that stored samples can be analyzed because total Hey is not altered when samples are kept frozen, even for years (B. Israelsson et al., 1992, unpublished). Total Hey is measured in all studies of Hey as a marker of vitamin deficiency states (20,21) and in most studies of Hey and cardiovascular disease (22). To date, -20 clinical studies involving >1800 patients and an equal number of control subjects have demonstrated that a moderate increase of serum/plasma Hey is an independent risk factor for premature cardiovascular disease. Above-normal plasma Hey has been found in -30% of the patients with premature cardiovascular disease who lack the traditional risk factors (22-24). Since 1985, the value of plasma/serum Hey determination in the diagnosis and follow-up of folate or cobal- amin deficiencies has been established. These deficiency states are probably the most frequently encountered causes of marked increases of serum/plasma Hey (2528). Here we review the methodologies for measuring total Hey in plasma/serum and their feasibffity as routine methods in the clinical chemistry laboratory. In the last part of the article, we evaluate total Hey as a marker of human disease, with emphasis on folate and cobalamin deficiencies. BIochemIstry Intracellular formation, metabolism, and release of Hey into the extracellular compartment determine the concentration of Hey in extracellular media (e.g., plasma/serum), which in turn is the basis for measuring Hey as an extracellular marker for human disease. In discussing these processes we will emphasize those that may affect the concentrations of extracellular Hey. Enzymes involved. Hey is formed as a product of the adenosyihomocysteinase (S-adenosylhomoeysteine hydrolase; EC 3.3.1.1) reaction, which is responsible for the removal of S-adenosylhomoeysteine, a product of S-adenosylmethionine-dependent transmethylation (29). Intracellular Hey is either remethylated to methionine, converted to cystathionine, or exported from the cells. The first reaction is catalyzed by the enzyme 5-methyltetrahydrofolate-Hey methyltransferase (methionine synthase; EC 2.1.1.13). This enzyme is ubiquitously distributed in mfimmAlian cells. It requires cobalamin as cofactor and catalyzes a reaction in which Hey remethylation is coupled to the conversion of 5-methyltetrahydrofolate to tetrahydrofolate; it thereby operates at a point of convergence of folate metabolism and the transplasma/serum methylation/transaulfuration pathway. An alternative route of Hey remethylation is confined to the liver: In this reaction, catalyzed by betaine-Hcy methyltransferase (EC 2.1.1.5), betaine serves as methyl donor (30, 31). The vitamin B6-dependent enzyme cystathionine f3-synthase (EC 4.2.1.22) catalyzes the condensation of Hey with serine to form cystathionine. The reaction is irreversible under physiological conditions, and from this point on Hey is committed to the transsulfuration pathway. Cystathionine is further cleaved to cysteine and a-ketobutyrate, catalyzed by another vitamin B6dependent enzyme (y-eystathionase; EC 4.4.1.1); this reaction completes the transsulfuration pathway (31). Enzyme regulation and Hey distribution between pathways. Hey is metabolized by either catabolizing enzymes or methionine-conserving enzymes; the distribution between these competing pathways is determined by the Km for Hey, the regulatory effect of S-adenosylmethionine, and the enzyme concentrations. The differential Km and metabolite regulation are processes that are put into immediate action in response to variable methionine availability, whereas up-regulation of enzyme synthesis is a slow adaptive process (31). Cystathionine -synthase and y-cystathionase are Hey-catabolizing enzymes, for which Km values are >1 mmol/L. Cystathionine /3-synthase is activated by S-adenosylmethionine, and the concentrations of both enzymes increase in response to excess dietary methionine. These properties ensure both immediate and longterm drainage of excess Hey via the tranasulfuration pathway (31). The Hey-remethylating enzymes, 5-methyltetrahydrofolate-Hey methyltransferase and betaine-Hey methyltransferase, have low Km values for Hey (<0.1 mmol/L). The activity is (directly or indirectly) inhibited by S-adenosylmethionine, and increased dietary methionine decreases the activity of the former enzyme. These properties favor methionine conservation at low Hey concentrations(31). Hey export. The release of Hey into the extracellular medium represents the third route of cellular Hey dis- posal and is particularly important in relation to the plasma concentration of Hey. Studies with isolated cells show that Hey export into the extracellular medium reflects an imbalance between Hey production and metabolism (32, 33). When cells are cultured in the presence of excess methionine, Hey export from most cell lines is enhanced (33); this phenomenon resembles the response to the methionine loading test. Pharmacological (34) and cell genetic studies (B. Christensen et al., unpublished) show that the activity of methionine synthase is critical for Hey export at low methionine concentrations, whereas cystathionine /3-synthase activity influences the export at high methionine concentrations. The clinical corollary is that methionine synthase activity determines the concentration of fasting plasma Hey, whereas a defect in cystathionine /3-synthase results in an abnormal response to methionine loading. This model is supported by clinical data showing that fasting plasma Hey is markedly increased in patients with folate (25,26, 35) or cobalaniin deficiency (21,25, 35) but is usually normal in vitamin B6 deficiency (36). Cobalamin-deficient patients and one patient with methylenetetrahydrofolate reductase deficiency had normal increases in plasma Hey after methionine loading (37), whereas patients who were homozygotes (3, 38) or obligate heterozygotes (39, 40) for cystathionine /3-synthase deficiency showed an abnormal response. Different Forms of Hcy In Human Plasma: Implications for Hcy Determination In the first studies of plasma Hey in healthy subjects (5, 41) and in some early clinical studies (20), Hey was measured in the acid-soluble fraction of plasma as Heycysteine mixed disulfide. However, a major fraction of Hey in serum/plasma is associated with plasma protein(s) (42), and probably forms a protein-Hey mixed disulfide with albumin (10). This fraction represents -70% of total Hey in human plasma/serum from healthy subjects (10). Only trace amounts of reduced Hey (11,43,44) and the disulfide homocystine (45) have been demonstrated. The sum of all Hey species in plasma (free plus protein-bound) is referred to as total Hey. The protein binding of Hey in plasma has some unique features. Both experimental (46,47) and clinical studies (48) demonstrate the presence in plasma of binding sites for aminothiols, which interact preferentially with Hey. In contrast to cysteine, binding of Hey to plasma proteins seems to be saturable, with a maximal capacity of -140 imol/L (48). This implies that the free fraction of Hey increases more than protein-bound Hey does when total Hey is markedly increased. During the brief hyperhomocysteinemia induced by methionine loading, only a moderate, transient increase in the free/ bound ratio occurs. This indicates that a rapid equilibrium exists between free and protein-bound fractions in vivo (20). The association of Hey with plasma proteins has important implications for Hey measurements. Free Hey becomes progressively bound to plasma protein(s) ex vivo. Such redistribution takes place within 24 h at room temperature. Also, most Hey in stored plasma is CUNICAL CHEMISTRY. Vol. 39, No. 9, 1993 1765 protein-bound. Reliable measurement of free Hey requires immediate acid treatment and centrifugation of plasma (49). This is impractical in the clinical setting, valent metals and inhibits oxidation. Hey is also kept in its reduced state by the presence of other thiols or by resupplementation of reductant (43). In the GC-MS as- and determination doned. say, inclusion of deuterium-labeled Hey as the internal standard corrects for reoxidation (12). Using a different sulfur compound as the internal standard does not seem feasible because the rate of oxidation of the internal standard may differ from that of Hey (54). Derivatization and detection. All methods except those based on eleetrochemical detection depend on derivatization of Hey. Precolumn derivatization with fluorogenie reagents for thiols followed by HPLC has become increasingly popular (55). Useful reagents must form Hey adducts with sufficient fluorescent yield to measure Hey at picomolar concentrations or less. Furthermore, the ideal reagent should be nonfluorescent, contain no fluorescence impurities, and react rapidly and specifically with Hey and other thiols to form stable products. No reagent meets all these requirements, but among the fluorogeme reagents available, bimanes and halogenosulfonylbenzofurazans have been found practical (Figure 1). mBrB couples rapidly with thiols at pH 8.0 at room temperature to produce a highly fluorescent thioether (56). One drawback is that the reagent itself, the hydrolysis products, and the impurities are fluorescent (51). These materials give rise to several reagent peaks that may interfere with Hey determination (14). Removal of excess reagent is included in some methods (15, 56). SBD-F and ABD-F are halogenosulfonylbenzofurazans that have been used for determination of thiols, including Hey (57). ABD-F reacts quantitatively with thiols, including Hey, at 50#{176}C at pH 8.0-9.5 for 5-10 mm (58), whereas the less-reactive SBD-F requires more drastic conditions (pH 9.5 and 60#{176}C for 1 h). SBD-F is not reactive towards of free Hey has largely been aban- General ConsIderatIons of Methodology Notes on chemicals. Some batches of DL-Hey contain large amounts (about 30%) of impurities, whereas L-homocystine from several suppliers seems to be pure. S-Adenosythomocysteine hydrolase from rabbit eiythrocytes used for the radioenzymie assays is occasionally available from Sigma Chemical Co. (St. Louis, MO) or can be prepared from liver of the mouse (50) or rat. Any adenosine deeminase present in the enzyme preparation is inhibited by 2’-deoxycoformyein (Pentostatin; Parke-Davis Research Labs., Ann Arbor, MI). Deuterated Hey used as internal standard in the gas chromatography-mass speetrometry (GC-MS) assay (12) is available from CJ) Isotopes (Va#{224}dret’il, Quebec, Canada), as are stable isotope-labeled forms of other metabolites that can be assayed with this technique. Monobromobimane (mBrB) contains fluorescent impurities, the amount of which varies from one batch to another. This reagent is occasionally available from Calbiochem-Behring Diagnostics (La Jolla, CA), but can regularly be supplied by Molecular Probes (Eugene, OR); the product of Molecular Probes is also of relatively high purity. 4-(Aminosulfonyl)- and axnmonium-7-fluoro2,1,3-benzoxadiazole-4-sulfonate (ABI)-F and SBD-F, respectively) are products of Wako (German Branch, Neuss, Germany). Reduction. Determination of total Hey in plasma/sethe reduction of the disulfide bond between Hey and other thiols or albumin. Reductants must be prepared immediately before use, and attention should be paid to stability during automated unattended sample processing. The selection of reduetant depends on the separation and detection system used. Sullhydryl-containing reducing agents such as dithioerythritol, dithiothreitol, and mercaptoethanol liberate Hey from various disulfIdes. These reductants form adducts with thiol-specific reagents and therefore may consume derivatization reagent (51). Sodium or potassium borohydride is a potent reduetant. The reduction takes a few minutes at high concentrations (1.4 mo]/L) (14), but requires up to 30 mm (15) or heating (52) at lower concentrations (40-100 mmol/L). Formation of gas and foaming during the reaction may impose practical problems, especially with regard to automatization, but this can be overcome by adding a surface-active agent, e.g., octanol. Tri-n-butylphosphine does not react with thiol-specific reagents nor does it form gas during reaction; however, it consumes fluorogenic reagent (mBrB) both at 50#{176}C and at room temperature (53) and is an irritant with an unpleasant odor. Reoxidation. Hey formed during the reduction step can be reoxidized before derivatization or detection, and variable reoxidation is a source of erratic results. Thiol oxidation is catalyzed by many transition metals (54); inclusion of EDTA in the reaction mixture chelates dirum requires 1766 CUNICALCHEMISTRY,Vol.39, No. 9, 1993 P 0 RSB )-N--k.. . e 470 am RSH2C mBrB SR N RSH \ 0 gL3SOJuS eai 510am S03-NH4 SBD-F SR RSH N N SO2NH2 ABD-F Fig. 1. Fluorogenic thiol(RSH)-specffic reagents for precolumn de- rivatization of Hcy The excitation and emission wavelengths for the derivatives are listed amines (59), whereas ABD-F also derivatizes amines but with a low fluorescence yield (57). The low reactivity of SBD-F may cause some problems because of Hey reoxidation, and assays based on this agent may be difficult to automatize. However, unreacted benzofurazans are not fluorescent, the thiol adducts are stable, and no fluorescent hydrolysis products are formed. Their use results in clean chromatograms with no reagent peaks. The formation of tert-butyldimethylsilyl derivatives for GC-MS (12) is not affected by traces of moisture, and these derivatives are stable at room temperature for many weeks. Automatization automatization put is required and 8ample output. A high degree of is advantageous, and high sample outfor methods used in routine laboratory diagnosis. Most of the present tion involve manual methods for total Hey determinaprocessing of the samples before chromatography. A fully automated procedure carried out in a sequential manner implies no centrifugation or filtration of sample, and the reduction and derivatization of Hey must be rapid. The combined use of NaBH4 and mBrB fulfills these criteria and has been successfully implemented in a fully automated procedure (14), as will be described below. Development of an automated procedure based on derivatization with a halogenosulfonylbenzofurazan is possible and should be undertaken. The long reaction time and heating required because of the low reactivity of SBD-F may cause problems, whereas ABD-F may be a suitable reagent. Assay procedures based on electrochemical detection can also be fully automated as long as attention is paid to detector stability and maintenance. Sample output depends on sample processing and chromatographic retention time. Manual processing of a large number of samples followed by a rapid chromatographic step gives a high sample output, as demonstrated for GC-MS (12) and for assays based on electrochemical detection (13) or fluorescence detection with use of SBD-F (18). The long analysis time of conventional amino acid analyzers, including analyzers optimized to detect Hey (16), restricts sample output. Codetermination of other metabolites. Some assays measure metabolites in addition to Hey; this may provide useful data for interpretation of the results. Methylmalonic acid and 2-methylcitric acid in serum are specific markers for cobalamin deficiency and may serve to differentiate hyperhomocysteinemia caused by cobalamin deficiency from that due to other causes (21, 60). The concentration of cysteine in serum shows a complex relation to Hey, both in homocystinuries (48,61,62) and healthy persons (44, 63); in one recent study, the ratio between Hey and cysteine correlated better with serum folate than did plasma Hey itself (64). Because methionine intake causes an increase in methionine concentration in blood, which is followed by a transient hyperhomocysteinemia, knowledge of the methionine concentration may be particularly useful when evaluating the results of a methionine loading determination of eystathionine, test (20). Simultaneous betaine, and N,N-dimethylglycine may provide additional information on metabolic alterations caused by folate and cobalamin deficiency and those caused by inborn errors of metabolism (65, 66). Measurement of N-methylglycine may provide a specific marker for folate deficiency (66). Methods Radioenzymic Assays Total Hey in human plasma was first determined with a radioenzymatic assay based on the conversion of Hey to S-adenosythomocysteine in the presence of [‘4C]adenosine and S-adenosythomocysteine hydrolase. Dithioerythritol was used as reductant, and radioactive S-adenosythomocysteine was quantified by HPLC and scintillation counting (10). This assay can also be carried out with unlabeled adenosine and ultraviolet detection at 254 nm. In two modified assays, S-adenosythomocysteine is isolated and quantified by thin-layer chromatography (67) or paper chromatography (17) instead of HPLC. The radioenzymic techniques are sensitive and specific and can also measure Hey in urine. The modifications based on thin-layer chromatography or paper chromatography require inexpensive equipment and can be established in any laboratory regardless of resources available. The disadvantages are that tedious manual processes, e.g., enzyme incubation, protein precipitation, neutralization, and centriftigation, are required. Another drawback is the limited range of this assay, owing to the consumption of radioactive adenosine present in the assay mixture. Gas Chromatography-Mass Spectrometry The GC-MS method for total Hey developed by Stabler et al. (12) has been modified recently (65) and is much simpler than the initial version. The steps before GC-MS are addition of deuterated internal standards in a single pipetting, addition of the reductant dithiothreitol and NaOH in a second pipetting, heating at 40#{176}C for 30 min, fractionation of sample on a disposable anion-exchange column, drying, and derivatization with N-methyl -N-(tertbutyldimethylsilyl)trifluoroacetamide. The tert-butyldimethylsilyl derivatives are separated and quantified by capillary GC-MS in the selected-ion monitoring mode, with use of a benchtop GC-MS equipped with an autoeampler. The variable recovery through these steps and the reoxidation of Hey during sample processing are corrected for by including deuterated Hey as internal standard, which results in a highly precise assay. Using semiautomated pipetting equipment, a single technician can process 320 samples in 8 h. The capacity of the automated chromatograph is about 160 derivatized samples per 24 h. Figure 2 shows a selected-ion monitoring trace of Hey and the internal standard in serum. The lack of total automation and the relatively high cost of the equipment are the major disadvantages of CUNICAL CHEMISTRY, Vol. 39, No. 9, 1993 1767 cHrl-cH3 H-N #{182}Hi cHr--CR3 H3 HHII nv CH3ICH3 420 nifz42a3o - 8 I ‘A reoxidation (14). Excess fluorescent ,m 424.20 6.6 mussen (19): They eluted the SBD-F adducts with a methanol gradient and obtained high precision of the assay by including mercaptopropionylglyeine as internal standard, which probably corrects for inaccurate composition of the assay mixture or variable injection volume. However, this compound elutes after 15 min, and significantly prolongs the run time. The SBD-F-based methods are sensitive and specific, and there are no interfering reagent peaks in the chromatogram (Figure 3). Disadvantages include the low reactivity of SBD-F and the long reaction time and high temperature (60 #{176}C) required. The derivatization is carried out manually, but the stability of the adduets allows autoinjection. mBrB has been used for determining Hey in plasma (14, 15). NaBH4 is used as a reductant, and a high concentration is required to obtain maximal fluorescent yield (53). The presence of dithioerythritol during the reduction and derivatization increases the linearity of the assay in the lower range, probably by preventing I . 6.7 6.8 6.8 Retention tine (miii) FIg.2. Determination of Hey in human serum byGC-MS mBrB has been removed by treating the sample with thiol Sepharose (15, 56). Sufficient resolution of the Hey adduct is also obtained after sample clean-up by solid-phase extraction (15) or column switching (14). However, thiol Sepharose treatment and solid-phase extraction are tedious procedures. Column switching requires two solvent-delivery systems, and The structure of the tevf-butyldimethyisllylderivative of Hcy and the site of major fragmentallon during mass spectroscopy are shown. The Hcy part of the formula Is shown In bold type. The uppertrace Is a selected-Ion monitoring recording for the ion m/z 420.3, which represents endo9enous Hcy in serum (6.3 unoUL). The lower trace is a selected-Ion monitoring recording for the ion rm’z424.3, representing the internal standard, which contains four deutetlum atoms. Details are given in references 12 and 65 method. Attractive features are high sensitivity and specificity, codetermination of cysteine, methionine, cystathionine, N,N-dimethylglycine, and N-methylglycine (65, 66), and the fact that the assay has been verified for urine samples (12). With a slight modification, methylmalonic acid (60), 2-methylcitric acid (60), and betaine (66) in serum can also be determined. this Precolumn Derivatization, HPLC, and Fluorescence Detection I Is In the method of Araki and Sako (11), Hey is derivatized with SBD-F, and the Hey, eysteine, and eysteinylglycine adducts are separated and quantified by gradient elution of a reversed-phase column within 12 mm. This method has been improved by Ubbink et al. (18), who obtained baseline separation of these adduets within 6 mm by isocratic elution on a reversed-phase column at pH 2.1. Their system ensures high sample output, but the low pH of the mobile phase increases dissolution of the silica matrix. A silica saturation column mounted in front of the injector or a guard column may reduce this process, and is recommended. Another modification has been constructed by Vester and Ras1768 CUNICAL CHEMISTRY, Vol. 39, No.9, 1993 Retention thne (miii) FIg. 3. Determination of Hey and other thiols in human plasma by automated precolumn denvatization with SBD-F The plasma contained -10 pmol/L Hcy. Details of the chrornatographicconditions are described in reference 19. Cys, cystelne; CysGIy, cystelnyiglycine; OSH, giutathione; IS, internal standard (mercaptopropionylgiyclne) the pressure surge during switching causes column doterioration and short column life. A fully automated method for the determination of plasmR Hey based on NaBH4 reduction and derivatization with mBrB was developed by Refsum at al. (14). To increase the reliability of this method, an improved procedure has been worked out in which the column switching is omitted by taking advantage of the marked pH dependence of the retention of several mBrB derivatives in reversed-phase chromatography. Baseline separation of Hey, cysteine, and cysteinyiglycine is obtained by elutung the column with an ammonium formate buffer adjusted to pH 3.65 (49) (Figure 4). Total plasma Hey has recently been assayed with modifications of the method currently used for amino acid determination, which is based on precolumn de- rivatization with o-phthaldialdehyde followed by HPLC and fluorescence detection (52, 68). Two assays have been published, and in both, plasma is treated with a reductant, Hey is carboxymethylated with iodoacetate before derivatization with o-phthaldialdehyde, and homocysteic acid is used as internal standard. The major differences between these assays are the reductant used and the retention times (10 and 22 miii) for Hey. When 2-mercaptoethanol is used, it reacts with iodoacetate, which must be added in excess (68). However, NaBH4 might increase the complexity of the sample-handling procedure of these assays, particularly the generation of excessive frothing (68). Both techniques include several manual steps (reduction, deproteunization, carboxymethylation, pH adjustment, derivatization), and are not suitable for automation. The performance of the assays is satisfactory. The widespread experience with the o-phthaldialdehyde techniques and the possible determination of other aminothiols and amino acids (including cysteine and cysteinylglycine) are attractive features of these methods (52, 68). HPLC and Electrochemical Detection HPLC coupled to an electrochemical detector is well suited for the determination of biological thiols (45, 69), including Hey (70, 71). This technique has been refined for assay of total Hey in plasma by Malinow et al. (13), who modified the method of Smolin and Sneider (72). NaBH4 is used as a reductant, and no derivatization of sample is required. Thiols in the column effluent are detected by a single gold-mercury electrode, which affords great specificity towards sulihydryl components. This assay has been optimized for the determination of Hey, which elutes at about 4.3 mm (Figure 5), and samples can be assayed every 10 miii (13). The chromatographic system can be modified to resolve cysteine, Hey, and eysteinylglycine (72). Thus, the electrochemical assay offers several attractive features: simple sample processing, specificity, autoinjection, short run time, large sample throughput, and the possibility of determination of other thiols (e.g., cysteine). One major weakness with the electrochemical assay is related to contamination of the flow cell and deterioration I Retention time (rain) FIg. 4. Determination of Hey and other thiols in human plasma by Retention time (miii) FIg. Determination of Hey in human plasma by HPLC and elec- automated precolumn denvatization with mBrB The plasma contaIned -10 tmoUL Hcy. The column is eluted with a mobile trochemical detection phase adjusted to pH 3.65. Other details of the chromatographic conditions described in reference 49. Abbreviations as In FIg. 3 The plasma contaIned -10 pmol/L Hey. Details of the chromatographic conditions are described in reference 13. CysGIy, cystelnylglycine are 5. CUNICAL CHEMISTRY, Vol. 39, No. 9, 1993 1769 of the gold-mercury electrode. Stability of this assay dopends on careful maintenance of the detector system. AminoAcidAnalyzer, Postcolumn Derivatization Determination of total Hey with a conventional amino acid analyzer requires the conversion of the disulfide forms into reduced Hey. 2-Mercaptoethanol (73) and dithiothreitol (16) have been used as reducing agents. Reduced Hey has been S-carboxymethylated by using iodoacetic acid (73). Alternatively, Hey has been chromatographed directly; under those conditions, Hey is protected against reoxidation by the sulfosalieylic acid (74) used for protein precipitation, by the presence of reducing agent in the sample, and by the low pH of the mobile phase. Hey is eluted with a standard program for physiological fluids (75), which also permits quantification of cysteine, methionune, and eystathionine (Figure 6). The elution program is interrupted after the elution of eystathionune and a 16-mm column regeneration period follows. The determination of Hey by a conventional amino acid analyzer has been optimized by Andersson et al. (16). The samples are treated with dithiothreitol (40 mmol/L plasma) and Hey is analyzed in its reduced form. Lack of homocystine disulfide indicates that no reoxidation takes place before or during the chromatography. The method also measures methionine and cystathionine, which elute close to Hey (16). The retention time of Hey (25.6 miii) and the time for column regeneration limit sample capacity, which is low compared with the HPLC assays. The construction and performance of the optimized method are shown in Table 1. The sensitivity limit (-1 mol/L in plasma) is less than that of other methods, but is sufficient for the determination of Hey in 0.5 mL of plasma; the impreei- sion is 5-6% (16). A drawback is the nonseleetivity for sulfur amino acids. In plasma samples frozen at -20 #{176}C for 10 years, nunhydrin-positive material interfered with the Hey peak, and the measured Hey values were about half of those obtained with a specific HPLC method (B. Israelsson et al., 1992, manuscript submitted). This may prohibit using the amino acid analyzer for determination of Hey in stored samples in prospective studies. However, in plasmas stored at -70 #{176}C for several years, no interfering peaks have been observed. The sample preparation is simple, and the technical staff often is well experienced in operating an amino acid analyzer, which is part of the routine instrumentation in many hospital laboratories. These facts justify the use of the ninhydrun-based Hey assay in the routine laboratory. Andersson et al. (76) recently described a method for plasma Hey that also measures eysteine, eysteinylglycue, glutathione, and glutamylcysteine. The assay involved reduction of sample with dithiothreitol, separation of the thiols by HPLC, and postcoluinn derivatization with 4,4’-dithiodipyridune and colorimetric detection at 324 nm. The method is characterized by high precision. The baseline fluctuation in this method reflects differences in absorbance between reagent and mobile phase. Therefore, the baseline noise is minimized by cooling the reagent reservoir to avoid hydrolysis and by delivering the postcolumn reagent with low pulsation flow. CorrelationStudies Satisfactory correlation exists between the concentrations of plasma Hey obtained by different methods. Values obtained with an automated mBrB assay (y) (14) have been compared with those obtained (x) with a radioenzymic assay (10) (y = 1.009-x 1.05, n = 160, r = 0.989) (14), an amino acid analyzer (16) (y = 0.85x + 1.88, n = 133, r = 0.975) (14), and electrochemical detection (13) (y = 0.85x + 1.11,r = 0.93).Acorrelation study has also been performed between values obtained with the electrochemical detection/HPLC assay (x) of Malinow et al. (13) and with an amino acid analyzer (16) (y = 0.99x 0.69, n = 133, r = 0.95) (77). Published methods for the determination of total Hey are listed in Table 2, and the performance and practicalities of five methods are compared in Table 1. - VII I - Sample Collection and Stability Effect of FoodIntake Retention Fig. 6. Determination of Hey In time (mm) human plasma with an amino acid analyzer The plasma contained -10 unoVL Hey. The chromatogrephic conditions are In accordance with the standard program of the manufacturer (Blotronlc LC 5000; Blolronic GmbH, Munich,Germany) for amino acids In physiological fluids. Hey, homocystelne; Cysts. cystathionine; Met, methlonine; Val, valine 1770 CLINICALCHEMISTRY, Vol. 39, No. 9, 1993 The data on possible diurnal variation in plasma Hey and its relation to normal food intake are sparse and inconsistent. Malinow et al. (13) found no difference in total Hey concentrations in fasting and nonfasting plasma from 13 subjects; Ubbink et al. (78) found a small but significant decrease in total plasma Hey 2-4 h after consumption of a normal breakfast with the values returning to normal after 8 h. Ubbink et al. suggested that the postprandial decline in plasma Hey could be Table 1. Charactei1stIcs Instrument Requirements, and Performance in Plasma of Five Methods for Determination of Total Homocystelne Assay Method (and Sanipi. ui(,ce) vat, mL GC-MS (12,85) 0.1 Step Plpetting Reduction Solid-phase Dr1ng 0.15 derivatization with SBD-F, HPLC,and fluorescence detection (19) Precolumn derivatizatlon with mBiB, 0.1 HPLC, and Tins, n*t Aad meted - Typ. Vacuurn-dring centrifuge - GC-MS - 30 extraction 5 Dedvalization, heatIng 90 30 Injection interval Precolumn Insment chaot&I 30 20 60 Plpetting Reduction Dedvatization Injection interval 3 3 18 0.2 electrochemical detectIon (13) Aminoacid 0.5 anatyzer(16) Plpetting Reduction Deprotolnlzallon Injection Interval Pipetllng Reduction Deproteinizallon Injection interval 56000 + - 15 Other no1le. dilsnnlnsd Cystathlonine Methionine CV for Hcy, S Cystelne - lLMethyIgIdne N,NDlmethy$gIdne Betairec Meth1malonlc acldc 2-Methylcitric addc Cysteine Cysteinylglycine HPLC pump Autolnjector Detector Integrator 15000 15000 9000 5000 90 HPLC pump Gilson 232 sample injector Detector integrator 15000 17000 80 Cystelne Cysteinylglydne 5000 90 Cysteinylglydne 3-1 15000 7000 5000 65000 24 Methlonine Other amino acids 51 + + + + + fluorescence detectIon (49) HPLC and Capecfty (nwi*sr/24 h) 160 - 7 Pipetting Reduction Deprotelnizatlon Derivatization, healIng Injectioninterval Price (US $) 7000 - HPLC pump - Autolnjector 30 20 - 10 + Detector Integrator Amino acid analyzer 9000 5000 60 30 60 1b + 1lntraassay CV. b interaasay CV. Slight modifications are required for the codeterminatlon of these metabolifes (see references in the text). Table 2. ConstructIon of Assays Reduction agent for Total Hcy In Plasma/Serum Separation principle Ref. Radioenzymatic assays Dithioerythrltol Enzymlc conversion to HPLC Absorbanceat 254 nm or scintillationcounting Scintillationcounting 10 Mass spectrometnj (single-Ionmonitoring) 12, 65 11, 19, 18 HPLC Fluorescence Fluorescence Fluorescence HPLC fluorescence 52 HPLC Fluorescence 68 HPLC Electrochemicaldetection 13, 46 Ionexchange (amino acid analyzer) Ionexchange (amino acid analyzer) HPLC Ninhydrinreaction 73 Ninhydrinreaction 16 Reaction with 76 ,,-Mannvthnmnrvetair,a Dithioerythrltol Enzymicconversionto S-adenosylhomocysteirte Gas chromatography-mass Dfthlothreftol spectrometly te,t-Butyldimethylsllyl derivatizatlon Capillary Precolumn derivatization, HPLC, and fluorescence detection TrI-n-butylphosphlne SBD-F Sodium borohydride Potassium borohydride Potassium borohydnde 2-Mercaptoethanol mBrB mBrB lodoacetic Thin-layer chromatography GC HPLC HPLC acid, 17, 67 14, 43, 49 15 o-phthaldialdehyde lodoacetic acid, o-phthaldlaldehyde HPLC and electrochemical detection Sodiumborohydride None Amino acid analyzer, postcolumn derivatizatlon 2-Mercaptoethanol S-Carboxymethylation lodoacetic acid Dithiothreitol None Dithiothreltol None with 4,4’-dithiodipyrldlne CUNICAL CHEMISTRY, Vol. 39, No.9, 1993 1771 attributed to an increased remethylation rate induced by the other nutrients in the meal. In a study of 11 students, plasma Hcy increased slowly, reaching a maximum increase of about 15% by 6-Sb after dinner (a steak) (A.B. Guttormsen et aL, unpublished results). Such a slow Hey responsetoameal is possibly due to the time it takes to degrade food proteins into free amino acids. After dinner, plasma methionine peaked after 3-4 h. In contrast, after peroral methionine loading, maximum plasma methionine is usually reached within 1-2 h (44, 79), and peak plasma Hey is usually measured after 4-8 h (20, 44, 79, 80). Thus, both after a meal and after a peroral methionine load there is a time difference of about 3-4 h between the maximal plcIRmR concentrations for methionine and Hey. These data indicate that intake of food may affect the concentration of plasma Hey and that its effect, although small, may persist for several hours. Stability of Hcy in Whole Blood Total Hey in serum or plsum’R increases when the separation of serum or plasma from the blood cells is delayed (12,19,49, 78,81,82). Within 1 hof storage of whole blood at room temperature, the phuImR and serum Hey concentrations increase by -10% (19, 49, 78, 82). Stabler et a!. (12) showed that storage of whole blood at room temperaturefor4and24hledtoinereasesinserumHcyof35% and 75%, respectively. The increase is of gimilRr mRgrntude (12, 19, 49), or greater (78, 82) in EDTA plasma These data explain why serum may have a higher Hey content than does optimally prepared plasma (19). During storage of whole blood, the rate of release of Hey from the blood cells is almost constant, independent of the plasma Hey concentration (49). Thus, the Hey release from blood cells may cause particularly large errors in the samples with low (i.e., normal) plasma Hey content, whereas the percent increase is moderate at higher concentrations (49). Similarly, after a peroral methionine load, the percentage of increase during storage of whole blood was less in samples with high plasma Hey (49). The blood cells are the source of the Hey increase storage, because the Hey concentration is stable in plasma or serum (19, 49, 78). Recently, Andersson et a!. showed that the erythrocytes are the major contributors to the artificial increase (82). Because hemolysis does not change or reduce the concentration of Hey in plasma (49, 78), the increase in plasma Hey is presumably due to continuous production of Hey in the erythrocytes, which is then released to the extracellular compartment. These data are consistent with the finding of low intracellular Hey in most cells (83). An artificial increase in total Hey is avoided when blood is put on ice immediately after collection, or when plasma or serum is prepared within 1 h after collection during (12, 19, 49, 78, 81, 82). Plasma Hey is to a large extent bound to (10). Because of postural hemodilution, serum averages 9% more in the blood collected from who are standing than from those in the supine 1772 CUNICAL CHEMISTRY, Vol. 39, No. 9, 1993 albumin albumin subjects position (84). The position may, therefore, affect the measured Hey values, but this has not been evaluated. plasma Stability of Hcy in Plasma Because Hey in plasma or serum is stable for days when kept at room temperature (19,49, 78), transport of unfrozen samples to the laboratory is feasible. Repeated freezing and thawing procedures (12, 19, 78, 81) do not affect the plasma or serum Hey concentration; neither does freezing for a period of several months (19, 85). Recently, Hey concentrations were investigated in samples kept frozen for an average of 10 years and in fresh samples from the same individuals (B. Israelsson et al., submitted). Included in the study were patients (n = 28) who had been healthy at the first investigation, but later suffered a stroke or cardiac infarct, and matched controls (n = 48). In both groups, the concentration of Hey in plasma was significantly correlated between stored and fresh samples. Plasma Hey was somewhat higher in the fresh samples than in the stored ones, but this may be related to the observed effect of the subjects’ age on the plasma concentration of Hey (77). Not only is plasma or serum Hey stable in the frozen state for years, but also the fasting Hey concentration remains fairly constant within most individuals for years (B. Israelsson et al., submitted). The stability of total Hey in whole blood and plasma/ serum under various conditions is summarized in Table 3. Values The normal values for total Hey differ somewhat from one laboratory to another (Table 4), but values between 5 and 15 mol/L are usually considered as normal. The variability may be related to different methodologies (Table 2), differences in sample processing, or the selection of subjects who are under the influence of various factors that affect the concentration of plasma/serum Hey. Total plasma Hey seems to be dependent on age, gender, and, in women, possibly menopausal status (77). The large differences in cysteine-Hey disulfide concentrations between pre- and postmenopausal women reported earlier (6, 86-88) have not been found when measuring total Hey. Total plasma Hey is decreased by 30-50% during pregnancy and is normalized within 2-4 days postpartum (89, 90). Figure 7 shows the frequency distribution curve for total plasma Hey in a screened population, showing a skewed distribution of values. Reference Table 3. StabIlity of Total Hcy In Whole Blood and Plasma/Serum Sample Whole blood Plasma/serum Sterage condftlens Room temp. 0-2#{176}C Room temp. 0-2 #{176}C -20#{176}C Data from references 12, 19,49,78, StabIlity <1 h 4-12 h At least 4 days Several weeks Years 82, and unpublished results. Table 4. Normal Values for Total Hcy In Human Men, a - 3(XJO Plasma/Serum a Age, yearsb Hcyfractlon n Value, gLmol/l Ref. ± 0.48 10 M 25-55 l8Free Bound 2.27 F 25-55 16 Free 1.95 ± 0.56 M <30 Bound 30-39 40-49 50-59 60-69 F <30 30-39 M+F F M9 F” 40-49 50-59 60-69 18-65 19-55 62.9 ± 10.8 5 Bound=totald 14 25 26 23 9 Bound=total” 8 ± 38 50 Total 99 Total 45 Total 10.0 45 Total 20-39 40-49 F 20-39 24 2.0 ± 0.9 8.9 ± 3.4 2.0 ± 1.0 9.8 ± 3.0 2.1 ± 0.8 10.0 ± 2.8 11.26 ± 3.71 10.15 ± 4.99 8.01 ± 1.76 9.8 ± 1.6 6.8 ± 1.2 13.3 ± 4.7 9.7 ± 3.3 12.4 ± 2.9 9.1 ± 3.0 50-69’ 37 Free Total 92 Total 167 Total 25 Total 8 Total 8 Total 195 Total 52 Total 40 Total 16 Total Total 63.4±11.5 M+F M 22-46 F 25-60 M M+F 25-60 J300#{149} 200100 0.7 ± 2.4 ± 21 104 127 7’ 0.9 ± 2.6 ± 0.7 118 128 19 129 Total Hcy in Laboratory Diagnosis Cobalamin Deficiency The classical view of cobalainin (vitamin B12) deficiency is that it usually develops at age >60 years; usually presents with megaloblastic anemia, macrocy- 50 Fig. 7. Frequency distribution curve for total plasma Hey screened In 3000 men, ages 40-42 years This population was part of a survey conducted In the county of Hordaland, Norway, by the National Health Screening Service In collaboration wIth the University of Bergen. The samples were assayed reference49 20-60 52 1-38 M, male; F, female, b Range, or mean ± SD. #{176} Hcy equivalents, mean ± SO, except as noted. d5jJ Hcy in storedsamples, because free Hcy becomes assoo ated with plasma protein(s). #{149} Mean ± 2 SD after log normalization to correctfor skewness towards higher values. The samples were allowed to stand at room temperature for 1-4 h before centrifugatlon. ‘Same as exceptthat samples were centrifuged wIthin 1 h of collection. 9Normotensive. “Hypertensive. ‘PosfrnenopausaL 1Blacktilbe. M+F 10 20 30 40 TotaLplasma homocysteine (jtmol/L) 0 ± 1.4 ± 4.6 ± lIJ8pniol/L 4.48 jm,oiiL Median = 10.70 pmolIL 95th percenllle -17.6 pmoi/L 126 3.2 11.7 ± 2.4 22 Free 56.0±11.3 ± Free 40-49 F 8.1 Free Total M 8.82 ± 3.82 9.20 ± 3.62 5.4-1 6.2 5.0-1 3.9 5.8 ± 0.9 Total 26 Free Total 15 Free 33 7.50 7.8 10.9 2.1 9.7 2.5 10.4 2.8 Total 50-69 8.06 1.28 ± 2.32 ± 2.00 ± 2.02 ± 2.32 ± 2.02 ± 1.64 7.00 ± 1.94 Total M 8.84 7.26 24 30 Total 63.3 9.44 ± 2.62 ± Mean SD= 400 6.51 ± 1.35 7.29 6.82 8.92 500 wIth the method desc.lbed in and hypersegmented neutrophils; and is often accompanied by neurological abnormalities. In these cases, the hematological picture was taken as an important diagnostic feature (91). The neurological abnormalities have long been considered to be confined to advanced deficiency, and to occur only rarely in the absence of anemia or other hematological findings. Recent clinical research has demonstrated that the classical view is wrong and that cobsdtiniin deficiency exists in a large number of patients with subtle biochemical changes and no hematological abnormalities (92-94); moreover, neuropsychiatric disorders due to cobalamin deficiency commonly occur in the absence of anemia and tosis, macrocytosis(93). The established diagnostic tests for cobalamin deficiency have some disadvantages. Serum cobalamin assays do not completely discriminate deficient patients from normal persons because, by definition, 2.5% of norma.! subjects have low values. Furthermore, 5-10% of patients with clinically confirmed cobislamin deficiency have normal values for serum cobalamin (93,95). Other procedures, e.g., bone marrow examination and the cJ.assic Schilling test, may also give false results (21, 94). Many cobalamin-deficient patients fail to absorb protein-bound cobalamin, but their absorption of the free vitamin is normal (96). The deoxyuridine suppression test is too cumbersome to be included in routine laboratory diagnosis (97, 98). Concentrations of total Hey in plasma/serum are increased in most patients with cobalamin deficiency (21, 25,28, 35, 93, 94, 99). In a large population of patients, some of whom have classical cobalamin deficiency, serum cobalamin and plasma Hey concentrations display a hyperbolic relation (Figure 8). The plasma Hey content increases abruptly when serum cobRbmin approaches values that are below normal (<130 pmol/L), but a negative correlation is observed in the normal to CLINICAL CHEMISTRY, Vol. 39, No. 9, 1993 1773 70’ that increases rapidly after perturbation of the intracellular cobalamin function. This is demonstrated in pa- ‘7 50’ 40 20 I: ‘r T 4047 IC’ I-’ 0 70 100 130 200 300 5001000 .1000 0 Serum cobelaznin(pinol/L) 2,3 5 73 10 IS 20 30 >30 Serum folale (nmol/L) FIg. 8. The relation between total Hey and serum cobalamin orserum folate Total plasma Hcy, serum cobalamin, and (or) serum lolate concentrations were determined in 1600 patients admitted to Haukeland University hospital. The patients were divided into groups according to their cobalamin or folate values. Eachgroupis definedby the vitamin concentrations given at the base of each column;thecolumnheight (andbars) indicate the corresponding mean (± SD) Hcy concentrationsforthatgroup.The numbersat the tops of the columns indicatethe numberof patients in each group low-normal range for cobalamin (130-300 pmoWL). This observation may suggest that having serum cobalamin in the normal range does not ensure optimal cobAlamindependent Hey remethylation in tissues. Compared with serum methylmalonic acid, plasma/ serum Hey is not a specific marker for cobalamin deficiency: The plasma concentration of Hey also increases markedly in other conditions, including folate deficiency, which causes indistinguishable hematological abnormalities. The specificity for both tests can be increased by measuring the metabolite concentrations before and after treatment with cobalainin or folic acid, because the concentration of Hey and methyhnalonic acid decreases rapidly towards normal (within 14 days) only after administration of the vitamin that is lacking (21). Diagnosis of the correct vitamin deficiency is very important: If a cobalamin-deflcient patient is mistakenly treated with folate, the hematological abnormalities may improve or completely correct, but the neuropsychiatric abnormalities will continue to progress (21, 100). Clinical studies show that plasma/serum Hey is a sensitive marker of coblilsimin deficiency (101). Hey and (or) serum methylmalonic acid concentrations were increased in 94% of 86 consecutive cobalamin-responsive patients, many of whom lacked one or more of the classic hematological or laboratory abnormalities (94). Furthermore, hematological relapse in patients on infrequent cobalamin maintenance therapy was associated with increased metabolite concentrations in 95% of the cases, whereas the serum cobalamin was low in only 69%. Serum methylmalomc acid was somewhat more sensitive than plasma/serum Hey in this study (95). The sensitivity of plasma/serum Hey in comparison with other diagnostic tests for cobalamin deficiency is an important issue. This question is related to the ability of plasma/serum Hey to reveal an early deficiency state. Total plasma/serum Hey is a responsive marker 1774 CLINICAL CHEMISTRY, Vol. 39, No. 9, 1993 tients being anesthetized with nitrous oxide, which oxidizes cobal(I)amin, a cofactor for the enzyme methionine synthase; after 90 miii of nitrous oxide exposure, a significant increase in total Hey is observed (102). Subtle or atypical cobalamin deficiency seems to be at least as common as the classical deficiency (103) and appears to occur often in the geriatric population (104, 105). Its nature is stifi being defined, and the importance of increased metabolites requires additional investigation. Nevertheless, determination of plasma/serum Hey is both a sensitive indicator of cobalamin depletion and an objective measure of therapeutic response. Folate Deficiency Folate deficiency is often nutritional and usually develops in predisposed subjects such as alcoholics, pregnant women, and patients with generalized malabsorption. Classical signs are macro-ovalocytosis, anemia, and hypersegmented neutrophils (98, 100). Plasma/serum Hey is markedly increased in folatedeficient patients (25, 26, 35, 101). In 19 patients with subnormal concentrations of serum folate (<2 gfL), 137 subjects with low-normal concentrations (2-3.9 g/ L), and 44 subjects with normal concentrations (4-17.9 zgIL), serum Hey was negatively correlated with the serum folate concentration. In 84% of the patients with subnormal serum folate, Hey was 2 SD greater than the normal mean, and up to 70 tmol/L. Of particular interest was the observation that more than half of the patients with low-normal serum folate had increased serum Hey (26). A similar picture is shown for the relation between plasma Hey and serum folate in 1620 patients (Figure 8). This suggests that folate intake that results in a serum folate content in the lower normal range is not necessarily sufficient for optimal remethylation of Hey and that a relative deficiency of tissue folate may exist under these conditions (26). Data from an experimental model involving rats fed a folate-defieient diet show that low-normal and subnormal concentrations of serum folate cause a two- to fourfold increase in serum Hey (106). The possibility that subjects with no clinical or laboratory signs of folate deficiency may actually be deficient in intracellular folate is supported by a decrease in plasma Hey after administration of high doses of folic acid. The effect of folic acid was observed both in hyperhomocysteinemic postmenopausal women (88) and in apparently healthy subjects (85). An increase in plasma Hey reflects intracellular folate deficiency or impaired function of folate-dependent reactions. The responsiveness of this factor is demonstrated by the observation that plasma Hey increases markedly within hours after administration of the antifolate drug methotrexate (80, 107-1 09). Furthermore, the specificity of plasma Hey as an indicator of folate deficiency is increased by the simultaneous determination of serum methylmalonic acid and by the reduction of plasma Hey in folate-deficient subjects within weeks after administration of folic acid but not after cobalamun (21). Plasma Hey as an indicator of folate deficiency may compensate for some defects of the conventional laboratory test. Serum folate does not discriminate between a transitory reduction in folate intake and chronic folate deficiency associated with depletion of tissue folate (110). Furthermore, both serum and erythrocyte folate may be normal in patients with alcoholism and megaloblastic anemia (111). In addition, assay of red cell folate has many technical pitfalls: Both the interlaboratory variation and the coefficient of variation have been high (112). The deoxyuridine suppression test has been regarded as a more reliable method than measurement of either serum or red cell folate for detecting folate deficiency (113), but the complexity of the procedure has prevented widespread use of this technique. The combined use of serum cobalamin, folate, total Hey, and methylmalomc acid concentrations currently provides the maximal utifity in diagnosing and distinguishing between cobalamin and folate deficiency. Any of the appropriate tests can give normal results in some patients with clinically confirmed cobalamun or folate deficiency (94, 95), but the use of the panel makes it possible to make or exclude the diagnosis of cobalamun or folate deficiency in virtually every patient (101). Nevertheless, a logistic regression analysis evaluating each of the four tests would be of interest. Homocystinuria Homocystinuria is a group of inherited disorders characterized by excretion of large amounts of homocystine into the urine. The most common form is cystathionine p-synthase deficiency, which has a prevalence of 1:200 000 worldwide. Rare forms are caused by various defects of Hey remethylation, including 5,10-methylenetetrahydrofolate reductase (EC 1.5.1.20) deficiency and errors of cobalamin metabolism. Common clinical signs in these diverse conditions are mental retardation and premature vascular disease (114). Homocystinuria has been the subject of several recent review articles (3,22, 114). Determination of Hey in plasma and homocystine in urine are laboratory tests for the diagnosis of various forms of homocystinuria. In patients with cystathionine fi-synthase deficiency, increases of plasma Hey up to 500 mol/L have been reported (3). These patients also have large amounts of methionine in plasma and low or lownormal concentrations of cystathionine (65)-important biochemical features that serve to distinguish these patients from patients with inborn errors of Hey remethylation. Patients with 5,10-methylenetetrahydrofolate reduetase deficiency or impaired methylcobalamin synthesis have high concentrations of Hey in plasma and urine, normal or subnormal concentrations of plasma methionine, and high concentrations of serum or plasma eyst.athionune (65). Measurement of methylmalonie acid and 2-methylcitric acid (60) in plasma or urine may differentiate reductase deficiency (with normal concen- of methylmalonic acid and 2-methylcitric acid) from defects in cobalamin metabolism (in which these acid concentrations are increased) (3). Measurement of metabolites in plasma and urine may point to a particular defect of Hey metabolism, but confirmation of the diagnosis requires determination of enzyme activities in cultured fibroblasts. This laborious technique has been established in only a few laborato- trations ries(3). Premature Cardiovascular Disease Studies involving -1800 patients and a comparable number of controls show a statistically significant correlation between hyperhomocysteinemia and premature cardiovascular disease (22). The overall increase in plasma Hey in populations with early cardiovascular disease is -30% over that in healthy subjects, but shows some variability related to the site of the vascular lesions. In coronary artery disease, the ratio for mean plasma Hey in patients vs controls was 1.2 to 1.3; in peripheral and cerebrovascular disease, this ratio varied from 1.5 to 1.8 (24). This distribution of statistical means is related to the observation that the incidence of hyperhomocyst(e)inemia is highest in cerebrovascular disease [-40% (115, 116)], intermediate in peripheral arterial disease [-25% (117)1, and lowest in coronary heart disease [-15% (118)]. Hyperhomocysteinemia is a possible risk factor for premature vascular disease, and is independent of other factors such as hypertension, diabetes, smoking, and plasma cholesterol (20, 22). A causative role of Hey in atherogenesis has not been established. A prospective study showing a relationship between plasma Hey and risk of myocardial infarction has been published recently (119), and another population-based prospective study has been completed (E. Annesen et al., manuscript submitted). These studies confirm that Hey is a strong risk factor for myocardial infarction, particularly in the younger age group. However, no intervention studies, which are critical for the question of the role of Hey in atherogenesis, have been conducted. We propose the tentative recommendation that plasma Hey should be determined in patients with premature (age <55 years) cardiovascular disease, especially in patients who lack other risk factors. The results from ongoing epidemiological and proposed intervention studies may broaden this recommendation in the future. and Drugs Plasma Hey is increased in patients with renal failure. This increase is moderate to marked (up to 50 unolJL) and is positively correlated to serum creatinune (20, 120). A moderate increase of plasma Hey is observed in most patients with psoriasis (81) and in some patients with leukemia (109) or solid tumors (107). Plasma concentrations are low in some patients with Down syndrome (20, 121) and in most patients with hyperthyroidism (Lien et al., unpublished). Some drugs affect the plasma Hey concentration (for a recent review, see 122). The antifolate drug methotrexOther Diseases CUNICAL CHEMISTRY, Vol. 39, No. 9, 1993 1775 ate, at doses of 25 mg (80) to 8 g/m2 (108, 109), and nitrous oxide anesthesia (30, 102) induce rapid and transient increases in plasma Hey. Preliminary data suggest that the antiepileptic drugs phenytoin and carbamazepine induce hyperhomocysteinemia, probably due to interference with folate functions. The cholesterol-lowering regimen of colestipol plus niacin also causes a significant hyperhomocysteinemia (123), possibly through interference with folate absorption. Drugs acting as vitamin B6 antagonists [isoniazide, cycloserine, hydralazine, penicillamine, pheneizine, and procarbazine (124)] would be expected to increase plasma Hey, as demonstrated for the banned drug azauridine (124). High doses of folic acid (5 mg daily) cause variable decreases in high plasma Hey concentrations in patients with renal failure but also in healthy subjects without overt folate deficiency (84, 125). The first doses of leucovorin (70 mg/m2), administered to modulate the effect of 5-fluorouracil, induced a marked decrease in plasma Hey within hours (J. Geisler et a!., unpublished). and ConclusIon It is mandatory that a technique Summary to be established in chemistry laboratory measures all Hey species in plasma/serum, i.e., total Hey. This is a prerequisite a clinical for the technique to be compliant with the routines for sample collection, processing, and transport in the clinical setting. Several new Hey methods have been developed and old techniques have been refined. The practical aspects of these techniques are quite different, as summarized in Table 1. Choice of method depends on the personnel and instrumental resources available, the training of the technical staff, the anticipated number of samples to be analyzed, and special interests concerning the codetermination of other metabolites. Determination of plasma/serum total Hey together with determinations of other metabolites, such as methylmalonic acid, is of value for the diagnosis and follow-up of cobalamin and folate deficiency, especially in cases of subtle or atypical deficiency, where the classical signs are lacking. These conditions seem to be rather common. Thus, the laboratory diagnosis of cobalamin and folate deficiency will probably account for the largest number of plasma/serum Hey determinations. In addition, assessment of risk for cardiovascular disease and clinical evaluation of premature vascular disease may become important applications in the near future. Plasma/serum Hey also plays a pivotal role for the diagnosis of different forms of homocystinuria, but these are rare conditions. References 1. Gerritaen homocysteine T, Vaughn JG, Weisman HA. The identification of in the urine. Biochem Biophys Res Commun 1962; 9:493-6. 2. Carson NAJ, Neil DW. Metabolic abnormalities detected in a survey of mentally backward individuals in Northern Ireland. Arch Dis Child 1962;37:505-13. 3. Mudd SH, Levy IlL, Skovby F. Disorders of transaulfuration. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic basis of inherited disease, 6th ed. New York: McGraw-Hffl, 1989: 693-734. 4. Wannmnker CMD, Weiner M, Giiigliani R, Filho CSD. An improved specific laboratory test for homocystinuria. Cliii Clam Acta 1982;125:367-9. 5. Gupta VJ, Wilcken DEL. The detection of cysteine-homocys. teine mixed disulphide in plasma of normal fasting man. Eur J Clin Invest 1978;8:205-7. 6. Wilcken DEL, GuptaVJ. Cysteine-homocysteine mixed disul. phide: differing plasma concentrations in normal men and women. Cliii Sci 1979;57:211-5. 7. Wilcken DEL, Wilcken B. The pathogenesis of coronary artery disease. A possible role for methionune metabolism. J Cliii Invest 1976;57:1079-82. 8. Wilcken DEL, Reddy SG, Gupta VJ. Homocysteinemia, ischemic heart disease, and the carrier state for homocystunuria. Metabolism 1983;32:363-70. 9. BrattatrOm LE, Hardebo JE, Hultberg BL. Moderate homocysteunemia-a possible risk factor for arteriosclerotic cerebrovascular disease. Stroke 1984;15:1012-6. 10. Refsum H, Helland S, Ueland PM. Radioenzymic determination of homocyateine in plasma and urine. Clin Chem 1985;31: 624-8. 11. Araki A, Sako Y. Determination of free and total homocysteine in human plasma by high-performance liquid chromatography with fluorescence detection. J Chromatogr 1987;422:43-52. 12. Stabler SP, Marcel! PD, Podell ER, Allen RH. Quantitation of total homocysteune, total cyateine, and methionine in normal serum and urine using capillary gas chromatography-mass spectrometry. Anal Biochem 1987;162:185-96. 13. Malinow MR, Kang SS, Taylor LM, Wong PWK, Inahara T, Mukeijee D, et aL Prevalence of hyperhomocyst(e)inemia in patients with peripheral arterial occlusive disease. Circ Res 1989;79: 1180-S. 14. Refaum H, Ueland PM, Svardal AM. Fully automated fluorescence assay for determining total homocysteine in plasma. Clin Chem 1989;35:1921-7. 15. Jacobsen DW, Gatautis VJ, Green R. Determination of plasma homocysteine by high-performance liquid chromatography with fluorescence detection. Anal Biochem 1989;178:208-14. 16. Andersson A, Brattstrom L, Isakason A, Israelsson B, Haltberg B. Determination of homocysteine in plasma by ion-exchange chromatography. Scand J Cliii Lab Invest 1989;49:445-50. 17. Chadefaux B, Coude M, Hainet M, Aupetit J, Kamoun P. Rapid determination of total homocysteine in plasma [Tech Briell. Cliii Chem 1989;35:2002. 18. Ubbink JB, Vermaak WJH, Biasbort S. Rapid high-performance liquid chromatographic assay for total homocysteine levels in human serum. J Chromatogr 1991;565:441-6. 19. Vester B, Rasmussen K. High performance liquid chromatog- raphy method for rapid and accurate determination tame in plasma of homocys- and serum. Eur J Clin Chem Cliii 1991;29:549-54. 20. Ueland PM, Refsum H. Plasma homocysteine, Biochem a risk factor for disease: plasma levels in health, disease, and drug therapy. J Lab Cliii Med 1989;114:473-501. vascular 21. Our original work cited in this review was supported by The Norwegian Cancer Society, the Norwegian Council on Cardiovascular Diseases, the Norwegian Research Council for Science and Humanities, and the Nordic Insulin Foundation (P.M.U., H.R); The National Institute on Ageing (AG-09834) (S.P.S.); The National Institute of Diabetes and Digestive and Kidney Diseases (DK-2 1365) (R.H.A.); The National Institute of Health (RR 0016332) (M.R.M.); and by grants from The Medical Faculty, University of Lund (A.A.). 1776 CLINICALCHEMISTRY,Vol.39, No.9, 1993 Allen RH, Stabler SP, Savage DG, Lindenbaum J. Diagnosis of cobalamin deficiency. 1. Usefulness of serum methylmalonic acid and total homocysteine concentrations. Am J Hematol 1990; 34:90-8. 22. Ueland PM, Refsum H, BrattstrOm L. Plasma homocysteine and cardiovascular disease. In: Francis RB Jr, ed. Atherosclerotic cardiovascular disease, hemostasis, and endothelial function. New York Marcel Dekker, 1992:183-236. 23. Mahnow MR. Hyperhomocyst(e)inemia. A common and easily reversible risk factor for occlusive atherosclerosis. Circulation 1990;81:2004-6. 24. Kang S-S,Wong PWK, Malunow MR. Hyperhomocyst(e)inemia as a risk factor for occlusive vascular disease. Annu Rev Nutr 1992;12:279-98. 25. Stabler SP, Marcell PD, Podell ER, Allen RH, Lindenbaum J. Serum homocysteine (Hcy) is a sensitive indicator of cobalamun (Cbl) and folate deficiency. Blood 1985;66(Suppl):50a. 26. Kang S-S. Wong PWK, Norusis M. Homocysteinemia due to foists deficiency. Metabolism 1987;36:458-62. 27. Hall CA, Chu RC. Serum homocysteine in the routine evaluation of potential vitamin B12 and folate deficiency. Eur J Haematel 1990;45:143-9. 28. Brattstrom L, Israelsson B, Lundgsrde F, Hultberg B. Higher total plasma homocysteine in vitamin B12 deficiency than in heterozygosity for homocystinuria due to cystathionune -synthase deficiency. Metabolism 1988;37:175-8. 29. Ueland PM. Pharmacological and biochemical aspects of S-adenosylhomocysteine and S-adenoaylhomocysteine hydrolase [Review]. Pharmacol Rev 1982;34:223-53. 30. Frontiera MS, Stabler SP, Kothouse JF, Allen RH. Regulation of methionune metabolism: effect of nitrous oxide and excess dietary methionine. J Nutr Biochem 1993;(in press). 31. Finkelstein JD. Methionune metabolism in mammals. J Nutr Biochem 1990;1:228-37. 32. Svardal AM, Djurhuus R, Refsum H, Uelsnd PM. Disposition of homocysteine in rat hepatocytes and in nontransformed and malignant mouse embryo fibroblasts following exposure to inhibitors of S-adenosylhomocysteine catabolism. Cancer Res 1986;46: 5095-100. 33. Christensen B, Refsum H, Vintermyr 0, Ueland PM. Homocysteine export from cells cultured in the presence of physiological or superfluous levels of methionine: methionine loading of nontransformed, transformed, proliferating and quiescent cells in culture. J Cell Physiol 1991;146:52-62. 34. Christensen B, Ueland PM. Methionine synthase inactivation by nitrous oxide during methionune loading of normal human fibroblasts. Homocysteine remethylation as determinant of enzyme inactivation and homocysteine export. J Pharmacol Exp Ther 1993; (in press). 35. Stabler SP, Marcell PD, Podell ER, Allen RH, Savage DG, Lindenbaum J. Elevation of total homocysteine in the serum of patients with cobalamin or folate deficiency detected by capillary gas chromatography-mass spectrometry. J Clin Invest 1988;81: teine and other thiol components in plasma during methionine loading in healthy men. Clin Chem 1992;38:1316-21. 45. Murphy-Chutorian DR, Wexman MP, Grieco AJ, Heininger JA, Glasaman E, Gaull GE, et al. Methionine intolerance: a possible risk factor for coronary artery disease. JAm Coil Cardiol 1985;6:725.-30. 46. Smolin LA, Benevenga NJ. Accumulation of homocyat(e)ine in vitamin B-6 deficiency: a model for the study of cystathionine -synthase deficiency. J Nutr 1982;112:1264-72. 47. Smolin LA, Benevenga NJ. The use of cyst(e)ine in the removal of protein-bound homocysteine. Am J Clin Nutr 1984;39: 730-7. 48. Wiley VC, Dudman NPB, Wilcken DEL. Interrelations between plasma free and protein-bound homocysteine and cysteine in homocystinuria. Metabolism 1988;37:191-5. 49. Fiskeratrand T, Refsum H, Kvalheim G, Ueland PM. Homocysteine and other thiols in plasma and urine: automated determination and sample stability. Cliii Chem 1993;39:263-71. 50. Ueland PM, D#{248}skeland SO. An adenosine 3’:5’-monophosphate-adenosine binding protein from mouse liver. J Biol Chem 1977;252:677-86. 51. Baeyens W, van der Weken G, Lin Lug B, de Moerloose P. HPLC determination of N-acetylcysteine in pharmaceutical preparations after pre-column derivatization with Thiolyte5 MB using fluorescence detection. Anal Lett 1988;21:741-57. 52. Fermo I, Arcelloni C, Devecchi E, Vigano S, Paroni R. Highperformance liquid chromatographic method with fluorescence detection for the determination of total homocyst(e)ine in plasma. J Chromatogr 1992;593:171-6. 53. Baeyens W, van der Weken G, De Moerloose P. Effects of reducing agents on the determination of thiolic compounds in the presence of their disulfides using bimane pre-column derivatization. Chromatographia 1987;23:717-21. 54. Munday R. Toxicity of thiols and disulfides: involvement of free-radical species. Free Rad Biol Med 1989;7:659-73. 55. Ohkura Y, Nolan H. Fluorescence derivatization in highperformance liquid chromatography. Mv Chromatogr 1989;29: 221-58. 56. Fahey RC, Newton GL, Dorian R, Kosower EM. Analysis of biological thiols: quantitative determination of thiols at the picomole level based upon derivatization with monobromobimanes and separation by cation-exchange chromatography. Anal Biochem 1981;111:357-65. 466-74. 36. Miller JW, Ribayamercado JD, Russell RM, Shepard DC, 57. Imai K, Uzu S, Toyo’oka T. Fluorogenic reagents, having benzofurazan structure, in liquid chromatography. J Pharm Morrow FD, Cochary EF, et a!. Effect of vitamin B-6 deficiency on fasting plasma homocysteine concentrations. Am J Cliii Nutr Biomed Anal 1990;7:1395-403. 58. Ling BL, Dewaele C, Baeyens WRG. Micro liquid chromatog1992;55:1154-60. 37. Brattstrom L, Israelsson B, Norrving B, Bergqvist D, Th#{246}nne raphy with fluorescence detection of thiols and disulphides. J J, Hultberg B, Hamfelt A. Impaired homocysteine metabolism in Chromatogr 1991;553:433-9. early-onset cerebral and peripheral occlusive arterial disease59. Imai K. Derivatization in liquid chromatography. Adv Chroeffects of pyridoxine and folic acid treatment. Atherosclerosis matogr 1987;27:215-45. 60. Allen RH, Stabler SP, Savage DG, Lindenbaum J. Elevation of 1990;81:51-60. 38 Brenton DP, Cusworth DC, Gaull GE. Homocystinuria: met2-methylcitric acid I and!! in the serum, urine, and cerebrospinal abolic studies on 3 patients. J Pediatr 1965;67:58-68. fluid of patients with cobalamun deficiency. Metabolism 1993; (in 39. Sardharwalla IB, Fowler B, Robins AJ, Komrower GM. Detecpress). tion of heterozygotes for homocystinuria. Study of sulphur-con61. Wiley VC, Dudman NPB, Wilcken DEL. Free and proteintaining amino acids in plasma and urine after L-methionune bound homocysteine and cysteine in cystathionine beta-synthase loading. Arch Dis Child 1974;49:553-9. deficiency-interrelations during short-term and long-term changes in plasma concentrations. Metab Clin Exp 1989;38:734-9. 40. Beers GHJ, Fowler B, Smals AGH, Trijbels FJM, Leermakers Al, Kleijer WJ, Kloppenborg PWC. Improved identification of 62. Mansoor MA, Ueland PM, Aarsland A, Svardal AM. Homeheterozygotes for homocystinuria due to cyatathionine synthase cysteine and other thiols as components of the metabolic derangement in patients with homocystunuria. Metabolism 1993; (acceptdeficiency by the combination of methionine loading and enzyme determination in cultured fibroblasts. Hum Genet 1985;69:164-9. ed). 41. Wilcken DEL, Gupta VJ. Sulphur containing amino acids in 63. Mansoor MA, Guttormeen AB, Fiskerstrand T, Refsum H, chronic renal failure with particular reference to homocystine and Ueland PM, Svardal AM. Redox status and protein binding of cysteine-homocysteine mixed disulphide. Eur J Cliii Invest 1979; plasma aininothiols during the transient hyperhomocysteinemia 9:301-7. that follows homocysteine administration. Cliii Chem 1993;39: 42. Kong S-S, Wong PWK., Becker N. Protein-bound homocys980-5. t(e)ine in normal subjects and in patients with homocystinuria. 64. Morgan SL, BaottJE, Refsum H, Ueland PM. Homocysteine levels in rheumatoid arthritis patients treated with low-dose Pediatr Res 1979;13:1141-3. 43. Mansoor MA, Svardal AM, Ueland PM. Determination of the methotrexate. Cliii Pharmacol Ther 1991;50:547-56. in vivo redox status of cysteine, cysteinylglycine, homocysteine 65. Stabler SP, Lundenbaum J, Savage DG, Allen RH. Elevation of and glutathione in human plasma. Anal Biochem 1992;200:218serum cystathionine levels in patients with cobalamun and folate 29. deficiency. Blood 1993;81:3403-13. 44. Mansoor MA, Svardal AM, Schneede J, Ueland PM. Dynamic 66. Allen RH, Stabler SP, Lindenbaum J. Serum betaine, NJ’!dimethylglycine and N-methylglycine levels in patients with corelation between reduced, oxidized, and protein-bound homocysCLINICAL CHEMISTRY, Vol. 39, No. 9, 1993 1777 balamin and folate deficiency and related inborn errors of metabolism. Metabolism 1993; (in press). 67. Chu RC, Hall CA. The total serum homocysteine as an indicator of vitamin B12 and folate status. Am J Cliii Pathol 1988;90:446-9. 68. Hyland K, Bottiglieri T. Measurement of total plasma and cerebrospunal fluid homocysteine by fluorescence following high. performance liquid chromatography and precolumn derivatization with ortho-phthaldialdehyde. J Chromatogr 1992;579:55-62. 69. Richie JPJ, Lang CA. The determination of glut.athione, cyst(e)ine, and other thiols and disulfides in biological samples using high-performance liquid chromatography with dual electrochemical detection. Anal Biochem 1987;163:9-15. 70. Demaster EG, Shirota FN, Redfern B, Goon DJW, Nagasawa HT. Analysis of hepatic reduced glutathione, cysteine and homecysteine by cation-exchange high-performance liquid chromatography with electrochemical detection. J Chromatogr 1984;308:8391. 71. Rabenstein DL, Yamashita GT. Determination of homocysteine, penidilamine, and their symmetrical and mixed disulfides by liquid chromatography with electrochemical detector. Anal Biochem 1989;180:259-63. 72. Smolin LA, Sneider JA. Measurement of total plasma cystsamine using high-performance liquid chromatography with electrochemical detection. Anal Biochem 1988;168:374-9. 73. Kong S-S, Wong PWK, Curley K. The effect of D-penicillamine on protein-bound homocyst(e)ine in homocystinurics. Pediatr Res 1982;16:370-2. 74. Jocelyn PC. Chemical reduction of disulfides. Methods En- zymol 1987;143:246-56. 75. Israelsson B, Brattstrom LE, Hultberg BJ. Homocysteine and myocardial infarction. Atherosclerosis 1988;71:227-34. 76. Andersson A, Isaksson A, Brattstrom L, Hultberg B. Homecysteine and other thiols determined in plasma by HPLC and thiol-specific postcolumn derivatization. Clin Chem 1993;39: 1590-7. 77. Andersaon A, BrattstrOm L, Israelsson B, Isaksaon A, Hamfelt A, Hultberg B. Plasma homocysteine before and after methionine loading with regard to age, gender, and menopausal status. Eur J Clin Invest 1992;22:79-87. 78. Ubbink JB, Vermask WJH, Vandermerwe A, Becker PJ. The effect of blood sample aging and food consumption on plasma total homocysteine levels. Clin Chim Acta 1992;207:119-28. 79. Andersson A, BrattatrOm L, Israelsson B, Isakason A, Hultberg B. The effect of excess daily methionune intake on plasma homocysteine after a methionine loading test in humans. Cliii Chim Acta 1990;192:69-76. 80. Refsum H, Helland 5, Ueland PM. Fasting plasma homocysteine as a sensitive parameter to antifolate effect. A study on psoriasis patients receiving low-dose methotrexate treatment. Clin Pharmacol Ther 1989;46:510-20. 81. Ueland PM, Refsum H, Svardal AM, Djurhuus R, Helland S. Perturbation of homocysteine metabolism by pharmacological agents in experimental and clinical use. In: Aarbakke J, Chiang PK, Koeffier HP, eds. Tumor cell differentiation, biology and pharmacology. Clifton, NJ: Humana Press, 1987:269-78. 82. Andersson A, Isaksson A, Hultberg B. Homocysteine export from erythrocytes and its implication for plasma sampling. Clin Chem 1992;38: 1311.-S. 83. Svardal A, Refsum H, Ueland PM. Determination of in vivo protein binding of homocysteine and its relation to free homocysteine in the liver and other tissues of the rat. J Biol Chem 1986;261:3156-63. 84. Young DS, Bermes EW. Specimen collection and processing. Sources of biological variation. In: Tietz NW, ed. Fundamentals of clinical chemistry. Philadelphia: WB Saunders, 1987:266-86. 85. Brattstrom LE, Israelsson B, Jeppsson J-O, Hultberg BL. Folic acid-an innocuous means to reduce plasma homocysteine. Scand J Clin Lab Invest 1988;48:215-21. 86. Boers GH, Smals AG, Trijl)els FJ, Leermakers Al, Kioppenborg PW. Unique efficiency of methionine metabolism in premenopausal women may protect against vascular disease in the reproductive years. J Clin Invest 1983;72:1971-6. 87. Blom HJ, Boers GHJ, van den Elzen JPAM, van Roessel JJM, Trijbels JMF, Tangerman A. Differences between premenopausal women and young men in the transamination pathway of math- 1778 CUNICALCHEMISTRY,Vol. 39, No. 9, 1993 onune catabolism, and the protection against vascular disease. Eur J Cliii Invest 1988;18:633-8. 88. Brattstr#{246}m LE, Hultberg BL, Hardebo JE. Folic acid responsive postmenopausal homocysteinemia. Metabolism 1985;34: 1073-7. 89. Kang S-S, Wong PWK, Thou J, Cook HY. Preliminsiry report: total homocyst(e)ine in plasma and amniotic fluid of pregnant women. Metabolism 1986;35:889-91. 90. Andersson A, Hultberg B, Brattstrom L, Isaksson A. Decreased serum homocysteine in pregnancy. Eur J Clin Chem Clin Biochem 1992;30:377-9. 91. McRae TD, Freedman ML. Why vitamin B12 deficiency should be managed aggressively. Geriatrics 1989;44:70-9. 92. Cannel R, Sinow RM, Karnaze DS. Atypical cobalamin deficiency-subtle biochemical evidence of deficiency is commonly demonstrated in patients without megaloblastic anemia and is often associated with protein-bound cobalamin malabsorption. J Lab Clin Med 1987;109:454-63. 93. Lindenbaum J, Healton EB, Savage DG, Brust JCM, Garrett TJ, Podell ER, et al. Neuropsychiatric disorders caused by cobalamin deficiency in the absence of anemia or macrocytosis. N Engi J Med 1988;318:1720-8. 94. Stabler SP, Allen RH, Savage DG, Lindenbaum J. Clinical spectrum and diagnosis of cobalamin deficiency. Blood 1990;76: 871-81. 95. Lindenbaum J, Savage DG, Stabler SP, Allen RH. Diagnosis of cobalamin deficiency. II. Relative sensitivities of serum cobalamin, methylmalonic acid, and total homocysteine concentrations. Am J Hematol 1990;34:99-107. 96. Carmel R, Sinow RM, Siegel ME, Salmoff IM. Food cobahimin malabsorption occurs frequently in patients with unexplained low serum cobalainin levels. Arch Intern Med 1988;148:1715-9. 97. Matchar DB, Feusaner JR. Laboratory tests in the diagnosis of vitamin B12 (cobalamin) deficiency. NC Med J 1986;47:118-20. 98. Beck WS. Diagnosis of megaloblastic anemia. Annu Rev Med 1991;42:311-22. 99. Lindenbaum J, Stabler SP, Allen RH. New assays for cobalamin deficiency getting better specificity. Lab Manage 1988;26: 41-4. 100. Allen RH. Megaloblastic anemia. In: Wyngaarden JB, Smith LHJ, Bennett JC, Plum F, eds. Cecil textbook of medicine, 19th ad. Orlando, FL: WB Saunders, 1991:846-54. 101. Savage DG, Lundenbaum J, Stabler SP, Allen RH. Serum methylmalonic acid and total homocysteine in the diagnosis of deficiencies of cobalamin and folate. Am J Med 1993;(in press). 102. Ermens AAM, Refsum H, Ruprecht J, Spijkers LJM, Guttormsen AB, Lindermans J, eta!. Monitoring cobalamin inactivation during nitrous oxide anesthesia by determination of homocysteine and folate in plasma and urine. Clin Pharmacol Ther 1991;49:385-93. 103. Cannel R. Subtle and atypical cobalamin deficiency states. Am J Hematol 1990;34:108-14. 104. Joosten E, van den Berg A, Riezier R, Naurath HJ, Lindenbaum J, Stabler SP, Allen RH. Metabolic evidence that deficiencies of vitamin B12, folate and vitamin B6 occur commonly in the elderly. Am J Clin Nutr 1993;(in press). 105. Pennypacker LC, Allen RH, Kelly JP, Matthews M, Grigsby J, Kaye K, et al. High prevalence of cobalaniin deficiency in elderly outpatients. J Am Geriatr Soc 1992;40:1197-204. 106 Liii J-Y, Kang S-S. Thou J, Wong PWK Homocysteinemia in rats induced by folic acid deficiency. Life Sci 1989;33:319-25. 107. Refsum H, Ueland PM, Kvinnsland S. Acute and long-term effects of high-dose methotrexate treatment on homocysteine in plasma and urine. Cancer Res 1986;46:5385-91. 108. Broxson EH, Stork LC, Allen RH, Stabler SP, Koihouse JF. Changes in plasma methionune and total homocysteine levels in patients receiving methotrexate infusions. Cancer Res 1989;49: 5858-62. 109. Refsum H, in children with chemotherapeutic 1991;51:828-35. 110. Bailey LB. 11. 111. Savage D, 1986;65:322-38. Wesenberg F, Ueland PM. Plasma homocysteine acute lymphoblastic leukemia. Changes during a regimen including methotrexate. Cancer Roe Folate status assessment. J Nutr 1990;120:1508- Lindenbaum in alcoholics. J. Anemia Medicine 112. Brown RD, Jun R, Hughes W, Watman R, Arnold B, Kronenberg H. Red cell folate assay: some answers to current problems with radioasaay variability. Pathologr 1990;22:82-7. 113. Wickramasinghe SN, Saunders JE. Results of three years’ experience with the deoxyuridine suppression test. Acta Haematol 1977;58:193-206. 114. Wilcken DEL, Dudman NPB. Homocystinuria and atherosclerosis [Review]. Monogr Hum Genet 1992;14:311-24. 115. Coull BM, Malinow MR. Beamer N, Sexton G, Nordt F, Garmo P. Elevated plasma homocyst(e)ine concentration as a possible independent risk factor for stroke. Stroke 1990;21:572-6. 116. BrattstrOm L, Lindgren A, Israelsson B, Malinow MR, Norrving B, Upeon B. Hyperhomocysteinemia in stroke. Prevalence, cause and relationship to other risk factors or type of stroke. Eur J Clin Invest 199222:214-21. 117. Molgaard J, Malinow MR, Lassvik C, Holm A-C, Olsson AG. Hyperhomocysteinemim an independent risk factor for intermittent claudicatio. J Intern Med 1992;231:273-9. 118. Malinow MR, Sexton G, Averbuch M, Grossman M, Wilson DL, Upeon B. Homocyst(e)inemia in daily practice: levels in coronary heart disease. Coronary Artery Dis 1990;1:215-20. 119. Stampfer MJ, Malinow MR, Willett WC, Newcomer LM, Upson B, Ulnuinn D, at a!. A prospective study of plasma homecyst(e)une and risk of myocardial infarction in United States physicians. JAm Med Assoc 1992;268:877-81. 120. Soria C, Chadefaux B, Coude M, Gaillard 0, Kamoun P. Concentrations of total homocysteine in plasma in chronic renal failure [Tech Briefi. Cliii Chem 1990;36:2137-8. 121. Chadefaux B, Ceballos I, Hamet M, Coude M, Poissonnier M, Kamoun P, Allard D. Is absence of atheroma in Down syndrome due to decreased homocysteine levels? Lancet 1988;ii:741. 122. Refsum H, Ueland PM. Clinical significance of phannacologlea! modulation of homocysteine metabolism [Review]. Trends Pharinacol Sd 1990;11:411-6. 123. Blankenhorn DR, Malinow ME, Mack WJ. Coleetipol plus niacin therapy elevates plasma homocyst(e)ine levels. Coronary Artery Dis 1991;2:357-60. 124. Droll W, Welch AD. Azaribine-homocystinemia-thrombosis in historical perspectives. Pharmacol Ther 1989;41:195-206. 125. Anonymous. Homocysteine, folk acid, and the prevention of vascular disease. Nutr Rev 1989;47:247-9. 126. Kong S-S, Wong PWK, Cook LW, Norusis M, Messer JV. Protein-bound homocyst(e)ine. A possible risk factor for coronary artery disease. J Clin Invest 1986;77:1482-6. 127. Araki A, Sake Y, Fukushima Y, Matsumoto M, Asada T, Kita T. Plasma sulihydryl-centaining amino acids in patients with cerebral infarction and in hypertensive subjects. Atherosclerosis 1989;79:139-46. 128. Mereau-Richard C, Muller JP, Faivre E, Ardouin P. Rousseaux L. Total plasma homocysteine determination in subjects with premature cerebral vascular disease [Letter]. Cliii Chem 1991;37:126. 129. Ubbunk JB, Vermaak WJH, Bennett JM, Becker PJ, van Staden DA, Bisabort S. The prevalence of homocysteinemia and hypercholesterolemia in angiographically defined coronary heart disease. Kim Wochenschr 1991;69:527-34. CUNICAL CHEMISTRY, Vol. 39, No.9, 1993 View publication stats 1779