Genotyping of S-mephenytoin 4’ -hydroxylation in an extended Japanese population

Objective: To assess the genotype pattern of S-mephenytoin 4’-hydroxylation in an extended Japanese population. Methods: One hundred eighty-six unrelated, healthy Japanese subjects were genotyped for S-mephenytoin 4’-hydroxylase (CYP2C19) according to a genotyping technique to identify the wild-type (wt) gene and two mutations, CTP2Cl9,, in exon 5 and cTIJ2Cl 9d in exon 4. Fourty-six of the 186 subjects genotyped were phenotyped with racemic mephenytoin using the conventional g-hour urine analysis of 4’- hydroxymephenytoin. Reszh: The frequency of poor metabolizers by the genotyping analysis was 18.8% (35 of the 186 subjects), consisting of 12 homozygous for CYP2C19,, (ml/ml), three homozygous for CYP2C19,, (m2/m2), and 20 heterozygous for the two defects (ml/m2). Thus the allele frequencies of CTP2C19,, and CTP2C19,,,2 were calculated to be 0.29 and 0.13 (107 and 46 of the total of 372 alleles), respectively. Among the 46 subjects phenotyped, seven were identified as the poor metabolizers, with a log,, urinary excretion of 4’-hydroxymephenytoin of <0.3% of the racemic dose. These seven subjects were genotyped as the individuals with the ml/ml (two), ml/m2 (four) or m2/m2 (one) allele combination, indicating a complete concordance between the phenotyping and genotyping tests. Conclusion: The present genotyping test confirmed that the frequency of CTP2Cl9 mutant gene ml is about 2.2 times greater than another mutant gene, m2, among Japanese poor metabolizers. The genotyp- ing of CYP2Cl9discriminates between the two S-mephenytoin 4’-hydroxylation phenotypes completely in the Japanese subjects. (Clin Pharmacol Ther 1996;60:661-6.)

Takahiro Kubota, MS, Kau Chiba, PhD, and Takashi Ishizaki, MD Huchioji City and Tokyo, Japan

The genetically determined S-mephenytoin 4’- hydroxylation polymorphism is characterized by two phenotypes, termed as extensive metabolizers and poor metabolizers,1,2 and the poor metabolizer phe- notype results from deficient S-mephenytoin 4’- hydroxylase (CYP2C19) activity.2,” This genetic polymorphism shows a cosegregation with the oxidative metabolism of several clinically import-

From the Center for Molecular Biology and Cytogenetics, SRI, Inc., Hachioji City, and the Department of Clinical Pharmacology, Research Institute, International Medical Center of Japan, Tokyo.

Supported by a grant-in-aid from the Japan Health Science Foun- dation (1-7-1-C) Drug Innovation Science Project (l-2-10), and the Ministry of Human Health and Welfare, Tokyo, Japan.

Received for publication April 15, 1996; accepted July 17, 1996. Reprint requests: Takashi Ishizaki, MD, Department of Clinical

Pharmacology, Research Institute, International Medical Cen- ter of Japan, Toyama 1-21-2, Shinjuku-ku, Tokyo 162, Japan.

Copyright 0 1996 by Mosby-Year Book, Inc. 0009-9236/96/$5.00 + 0 13/I/76667

ant drugs such as diazepam,4,5 imipramine,6,7 omeprazole,8‘10 propranolol,rt and selective seroto- nin reuptake inhibitors.12,13 Furthermore, this phar- macogenetic entity has shown a marked interethnic difference in the incidence of the poor metabolizer phenotype: the poor metabolizer frequency is much greater (18% to 23%) in Japaneser4,r5 than that (3% to 5%) in North American1”4 and European white populations.16

Two different mutation events that cause a defec- tive CW2C19 have recently been described17? de Morais et a1.17 first reported that the principal defect in the poor metabolizer phenotype is a single base pair (G-+A) mutation in exon 5 of CYP2C19 (termed as CyP2C19,,), which creates an aberrant splice site, and showed that seven of 10 American and 10 of the 17 Japanese poor metabolizer subjects are homozygous for this defects. de Morais et a1.l’ also reported a new mutation (termed as

661

662 Kubota, Chiba, and Ishizaki CLINICAL PHABMACOLOGY &THERAPEUTICS

DECEMBER 1996

CY?‘2C19,,), consisting of a G+A mutation at po- sition 636 of exon 4 of CIT2C19, which creates a premature stop codon. This additional CW2C19 defect accounted for the remaining Japanese poor metabolizer subjects.i8 Hence, de Morais et al.” concluded that CW2C19,1 and CI?‘2C19,, explain 100% of the available Japanese poor metabolizer subjects (34 alleles).

In this study we had two aims. First, we intended to assess the genotype pattern of CW2C19 in an extended Japanese population, because the data re- ported by de Morais et a1.18 were derived from the limited number of Japanese subjects (i.e., 12 Japa- nese extensive metabolizers and 17 Japanese poor metabolizers). Second, we intended to assess the relationship between the genotyping and phenotyp- ing tests of CYP2C19 (i.e., existence of mutant allele ml or m2 versus 4’-hydroxylation capacity of S-mephenytoin) in a selected Japanese group who received both tests.

MATERIAL AND METHODS Subjects. One hundred eighty-six unrelated,

healthy Japanese volunteers (101 men and 85 wom- en; age range, 19 to 61 years) were recruited for the genotyping study. The subjects were interviewed for disease history and were considered to be in good health. They were informed both verbally and in writing about the experimental procedure and the purpose of the study. None of the participants took any drugs during the course of the study. The study protocol was approved by the institutional ethical committee of the International Medical Center of Japan, Tokyo, Japan.

Genotyping. Seven milliliters of venous blood was obtained from all 186 subjects, and deoxyribonucleic acid (DNA) was isolated from peripheral leukocytes with use of an extraction kit (Genomix, Talent, Trieste, Italy). The CYP2C19 wild-type (wt) gene and the two mutated alleles associated with deficient S-mephenytoin 4’-hydroxylation, CW2C19,, and CYP2C19,,, were identified by a polymerase chain reaction (PCR) amplification with use of the allele- specific primers described by de Morais et a1.17,18 with minor modifications: for exon 4, the forward primer (S-AACATCAGGATTGTAAGCAC-3’) anneals in exon 4, 33 base pairs upstream from the exon 4/intron 4 junction, and the reverse primer (5’-TCAGGGCTTGGTCAATATAG-3’) anneals in intron 4, 85 bp downstream from the exon 4- intron 4 junction. For detecting the CY?‘2C19,1 and CIT2C19,,, genomic DNA (200 ng) was amplified

in PCR buffer (10 mmol/L Tris hydrochloride, pH 8.3 and 50 mmol/L potassium hydrochloride, 0.01% gelatin) that contained 200 p.mol/L deoxyribo- nucleoside triphosphate (dNTP) mix (dATP, dCTP, dGTP, and d?TP, Takara Shuzo Co., Ltd., Shiga, Japan), 0.2 p,mol/L concentrations of PCR primers, 1.25 units of AmpliTaq DNA polymerase (Hoffmann-La Roche, Ltd., Base& Switzerland), and 1.5 mmol/L magnesium chloride. Amplification was performed with a Perkin Elmer thermocycler, for 40 cycles consisting of denaturation at 94” C for 1 minute, annealing at 57” C for 1 minute, and exten- sion at 72” C for 2 minutes. An initial denaturation step at 94” C for 5 minutes and a final extension step at 72” C for 5 minutes were also performed. Restric- tion enzyme cleavage was conducted at 37” C for 1 hour after the addition of 25 units of Msp I for CW2C19,, and 25 units of BamHI for CyP2C19,,. The digested PCR products were analyzed on 3% agarose gels and stained with ethidium bromide. Representative genotyping results are shown in Fig. 1.

Zn vivophenotyping. Among the 186 Japanese sub- jects genotyped as described above, 46 were pheno- typed for their individual capacity to 4’-hydroxylate S-mephenytoin with the amount of 4’- hydroxymephenytoin excreted in the &hour urine after taking an oral dose of 100 mg of racemic mephenytoin (Mesantoin, Sandoz Inc., East Hanover, N.J.). The extensive metabolizer and poor metabolizer phenotypes were classified by use of an antimode of 0.3 in the log,, percentage of the postdose 8-hour urinary excretion of 4’- hydroxymephenytoin (on a molar basis relative to the 100 mg racemic dose of mephenytoin). An individual with a log,, urinary 4’- hydroxymephenytoin excretion of CO.3 was considered to be a poor metabolizer of S-mephenytoin as reported from our laborato- ry- 5,7,15 Urinary 4’-hydroxymephenytoin was mea- sured by means of the capillary gas chromatogra- phy with nitrogen phosphorus detection.15

RESULTS AND DISCUSSION This is the largest population genotyping study on

CYP2C19 for both CIT2C19,1 and CYP2C19,, in light of the recent two population studies where 103 black Zimbabwean Shona subjects” and 140 Ger- man subjects2’ have been genotyped only for the wt and the intron 4-exon 5 G+A splice site mutation (ml) of CYP2C19. DNA obtained from the 186 Japanese subjects was amplified with use of the

CLINICAL PHARMACOLOGY & THERAPEUTICS 1’OLUME 60, NUMBER h

A Msp I Digestion

M wthvt wtlml ml/ml

210- 162- - 169 bp

- 120 bp

Kubotu, Chiba, and Ishizaki 663

B BamH I Digestion

M wtlwt wtlm2 m2lm2

310-

%= W-

118- -119bp - 93bp

Fig. 1. Polymerase chain reaction (PCR) analysis of the exon 5 (A) and exon 4 (B) for the CYP2Cl9 gene in six representative Japanese subjects. A, Shows the PCR amplification of exon 5 digested with Mqr I for CYP2C19,,. B, Shows the PCR amplification of exon 4 digested with BamHI for CYP2Cl9,,. The predicted base pair (bp) sizes of the digested deoxyribo- nucleic acid (DNA) fragments for various genotype patterns are shown on the right-hand side. The sizes of the molecular weight markers (M) are shown on the left-hand side. PCR products from DNA in individuals with the wild-type (wt) allele(s) are cleaved by the restriction enzyme, whereas those in homozygous individuals with the mutation (ml/ml or m2im2) lack the il/lsp I (A) or BamHI (B) site, respectively, and show a single band. The predicted sizes of the digested DNA fragments for heterozygous individuals are shown in A (wt/ml) and in B (wt/m2), respectively.

specific primer and digested with Msp I for CYP2C19,, and with BamHI for CYP2C19,,. Six different allelic band patterns were observed, as shown in Fig. 1. The results of the genotype analysis are summarized in Table I.

Sixty-five (34.9%) of the 186 subjects were ho- mozygous for the wt allele in both exon 4 and exon 5 (wtiwt), 63 (33.9%) were heterozygous for the ml mutation (wt/ml), and 23 (12.4%) were heterozy- gous for the m2 mutation (wt/m2), indicating that the frequency of the heterozygous extensive me- tabolizer genotype (46.3%) is more prevalent than the frequency of the homozygous extensive metabo- lizer genotype among the Japanese subjects.

Twenty (10.8%) of the 186 subjects were het- erozygous for the ml and m2 mutation (ml/m2), 12 (6.4%) were homozygous for the ml (ml/ml), and three (1.6%) were homozygous for the m2 (m2/m2), indicating that the frequency of the heterozygous poor metabolizer genotype is greater than that of the homozygous poor metabolizer genotype, with the ml/ml being more prevalent than the m2/m2 pattern. These genotype patterns observed among the Japanese subjects are in agreement with those

Table I. Genotype analysis of 186 Japanese subjects with respect to the CYP2C19 gene

Genotype * No. of subjects 9% of subjects

wtlwt 65 34.9 wtlm I 63 33.9 wtim2 23 12.4 mllm2 20 10.8 ml/ml 12 6.4 m2lm2 3 1.6

~~ ~~.._~~ *wt, Wild-type: ml, CYP2CIY mutation in exon 5: m2, CYpZC19 muta-

tion in exon 4.

reported in the study of 17 Japanese poor metabo- lizers by de Morais et al.‘s Thus our population genotyping study indicated that the frequency of the poor metabolizers is l&8%, which is compatible with the frequency reported from the previous Jap- anese population phenotyping studies,‘4”5 but much greater than that from North American’Z’4 and Eu- ropean white populations’~16320 and a Zimbabwean populationl”

We observed that CYP2C19,, and CYP2C19,, accounted for 107 (28.7%) and 49 (13.2%) of the

664 Kubota, Chiba, and Ishizaki CLINICAL PHARMACOLOGY &THERAPEUTICS

DECEMBER 1996

I”

14

12

10

8

6

4

2

0 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Log, o percentage of 4’-hydroxymephenytoin in urine

Fig. 2. Frequency distribution of S-mephenytoin 4’-hydroxylation in 46 Japanese subjects in relation to the individual genotype patterns. The postdose &hour urinary 4’- hydroxymephenytoin excretion differentiated the extensive metabolizer phenotype (n = 39) from the poor metabolizer phenotype (n = 7). The homozygous (wthvt; n = 23) and heterozygous (wt/ml plus wt/m2; n = 16) extensive metabolizer genotypes overlapped the one histogram modality of the extensive metabolizer phenotype on the right-hand side, suggesting that the genotyping cannot discriminate between the two extensive metabolizer phenotype subgroups. This appears to be the case in the limited number of the homozygous (ml/ml plus m2/m2; n = 3) and heterozygous (ml/m2; n = 4) poor metabolizer subjects, as shown in the distribution on the left-hand side.

372 alleles, respectively. Therefore, according to the Hardy-Weinberg equation, the frequency of the re- cessive gene (q) controlling the poor metabolizer status of S-mephenytoin 4’-hydroxylation is esti- mated to be 0.42. The frequency of CY?‘2C19,, allele is calculated to be 2.2 times higher than that of CIT2C19,, allele in our poor metabolizer subjects (i.e., 0.29 and 0.13, respectively).

Of the 186 Japanese subjects genotyped, 46 were phenotyped by measuring the &hour urinary excretion amount of 4’-hydroxymephenytoin among whom seven were classified as the poor metabolizers with a log,, percentage excretion of the 4’-hydroxy metabo- lite of racemic mephenytoin of ~0.3 (Fig. 2). Their genotypes included two homozygous for CYP2C19,, (ml/ml), one homozygous for CP2C19, (m2/m2), and four heterozygous for the two mutations (ml/r&), indicating that the CYP2C19 defect in exon 5 (ml) and in exon 4 (m2) thus accounts for 100% of the defective allele in the Japanese poor metabolizer-phenotyped subjects as reported by de Morais et all8 The remain- ing 39 extensive metabolizers showed three different

allelic band patterns, as listed in Table II. When the frequency distribution of the urinary 4’- hydroxymephenytoin excretion observed in the 46 in- dividuals is plotted in relation to the homozygous (wt/wt, ml/ml, or m2/m2) and heterozygous @t/ml, wt/m2, or ml/r&) genotype patterns, the distribu- tion histogram as depicted in Fig. 2 was obtained. The mean (+SD) or individual urinary 4’- hydroxymephenytoin excretion data obtained from the six genotype subgroups are also listed in Table II.

The histogram suggested that the homozygous and heterozygous extensive metabolizers genotyped overlap the individual values for the urinary 4’- hydroxymephenytoin excretion and therefore that the genotype patterns cannot differentiate one from another extensive metabolizer phenotype in the Jap- anese subjects. This also appears to be the case in the limited number of the Japanese poor metabo- lizers genotyped and phenotyped, as shown in Fig. 2. The mean 5 SD values for the urinary 4’- hydroxylation excretion as percentage of the dose (log,,%) were 34.44% 2 7.08% (1.53 +- 0.09) in the

CLINICAL PHARMACOLOGY & THERAPEUTICS VOLUME 60, NUMBER 6 Kubota, Chiba, and Ishizaki 665

Table II. Genotype analysis of 46 Japanese subjects with respect to the CYP2C19 gene who were phenotyped with racemic mephenytoin

Urinary 4’-hydroxymephenytoin Genotype * No. of extensive metabolizers No. of poor metabolizers f (Yo dose)$

wtlwt 23 0 34.44 !I 7.08 wtlml 11 0 32.78 t 5.41 wtlm2 5 0 34.39 2 5.03 mlJm2 0 4 1.13 ? 0.69 ml/ml 0 2 0.65, 1.86 m2lm2 0 1 1.15

wt, Wild-type; ml, CYP2C19 mutation in exon 5; mZ, CYP2C19 mutation in exon 4. tIndividuals were phenotyped by use of the postdose S-hour urinary excretion of 4’.hydroxymephenytoin on a molar basis relative to the 100 mg racemic dose

of mephenytoin.5~7~15 $Mean 5 SD in the cases of more than four individuals included. Otherwise, the individual values are listed.

homozygous (wt/wt) extensive metabolizers (n = 23) 33.28% f 5.35% (1.52 + 0.07) in the heterozy- gous (wt/ml and wt/m2) extensive metabolizers (n = 16) 1.22% 2 0.50% (0.05 + 0.19) in the homozy- gous (ml/ml and m2/m2) poor metabolizers (n = 3) and 1.13% 2 0.69% (-0.02 t 0.24) in the het- erozygous (ml/m2) poor metabolizers (n = 4). As expected from the data shown in Fig. 2, there was no significant difference in the mean values either be- tween the two extensive metabolizer groups or be- tween the two poor metabolizer groups, although the urinary 4’-hydroxymephenytoin excretion differs significantly @ < 0.001) with use of the Mann- Whitney test between the extensive metabolizer and poor metabolizer groups.

Our observation that the genotype patterns can- not separate the homozygous from the heterozygous extensive metabolizers in terms of the metabolic capacity of S-mephenytoin by the urinary phenotyp- ing test is incompatible with the two recent studies in which the mean S/R metabolic ratio values of mephenytoin differed significantly between the Zim- babwean homozygous (wt/wt) and heterozygous (wt/ ml) extensive metabolizersi’ and between the Ger- man homozygous (wt/wt) and heterozygous (wt/ml) extensive metabolizers. ” The reason for this discor- dance remains totally obscure at present.

Our population genotyping study provided un- equivocal evidence that a complete ascertainment of the CYP2C19-related poor metabolizer phenotype can be made by the two simple DNA PCR tests in Japanese subjects. In contrast, however, the predict- able sensitivity of the genotyping test for identifying the two mutations, CYP2C19,1 and CyP2C19,,, has been reported to be 83% in nine white Ameri- can poor metabolizer subjectsls and 94% in 16 Dan- ish poor metabolizer subjects,” suggesting the pos-

sibility that other mutation(s) of CyP2C19 may exist for the unidentified 5% to 15% white poor metabo- lizers of S-mephenytoin 4’-hydro@ation. Neverthe- less, the PCR and restriction enzyme procedures to detect CYP2C19,, and CYP2C19,, mutations can now be used for the clinical pharmacologic research of drugs that are metabolized by this isozyme”*~** and for which therapeutic level monitoring is impor- tant for clinical efficacy or toxicity.23-25 Thus the cyP2C19 genotyping test will prove to be useful in pharmacoepidemiologic and pharmacokinetic stud- ies of patients exposed to drugs or substrates of CP2C19. In addition, the genotyping test will also be useful in clinical studies of the possible involve- ment of CyP2C19 in succeptibility to some sponta- neous disorders, including cancer.26-28

Authors’ note added in proof After this manuscript was submitted for publica-

tion, a short communication (Takakubo F, Kuwano A, Kondo I. Evidence that poor metabolizers of (S)-mephenytoin could be identified by haplotypes of CYPZC19 in Japanese. Pharmacogenetics 1996;6: 265-7) was published in which the frequency of the poor metabolizers was 14.6% in the 217 unrelated healthy Japanese genotyped for CY?‘2C19,, and CYP2C19,,. The results were found to be in rea- sonable agreement with the results of this study.

References 1. Wilkinson CR, Guengerich FP, Branch RA. Genetic

polymorphism of S-mephenytoin hydroxylation. Phar- macol Ther 1989;43:53-76.

2. Goldstein JA, de Morais SMF. Biochemistry and mo- lecular biology of the human CYP2C subfamily. Phar- macogenetics 1994;4:285-99.

3. Goldstein JA, Faletto MB, Romkes-Sparks M, Sulli-

666 Kubota, Chiba, and Ishizaki CLINICAL PHARMACOLOGY & THERAPEUTICS

DECEMBER 1996

van T, Kitareewan S, Raucy JL, et al. Evidence that CYP2C19 is the major (S)-mephenytoin 4’- hydroxylase in human. Biochemistry 1994;33:1743-52.

4. Bertilsson L, Henthorn T, Sam E, Tybring G, Sawe J, Villen T. Importance of genetic factors in the regula- tion of diazepam metabolism: relationship to S-mephenytoin, but not debrisoquin, hydroxylation phenotype. Clin Pharmacol Ther 1989;45:348-55.

5. Sohn DR, Kusaka M, Ishizaki T, Shin SG, Jang IJ, Chiba K. Incidence of S-mephenytoin hydroxylation deficiency in a Korean population and the interphe- notypic differences in diazepam pharmacokinetics. Clin Pharmacol Ther 1992;52:160-9.

6. Skjelbo E, Brasen K, Hallas J, Gram LF. The mephe- nytoin oxidation polymorphism is partially responsible for the N-demethylation of imipramine. Clin Pharma- co1 Ther 1991;49:18-23.

7. Koyama E, Sohn DR, Shin SG, Chiba K, Shin JG, Kim YH, et al. Metabolic disposition of imipramine in Oriental subjects: relation to metoprolol a-hydroxylation and S-mephenytoin 4’-hydroxylation phenotypes. J Pharmacol Exp Ther 1994;271:860-7.

8. Andersson T, RegLdh CG, Lou YC, Zhang Y, Dahl ML, Bertilsson L. Polymorphic hydroxylation of S-mephenytoin and omeprazole metabolism in Cauca- sian and Chinese subjects. Pharmacogenetics 1992,2:25- 31.

9. Chiba K, Kobayashi K, Manabe K, Tani M, Kamataki T, Ishizaki T. Oxidative metabolism of omeprazole in human liver microsomes: cosegregation with S-mephenytoin 4’-hydroxylation. J Pharmacol Exp Ther 1993;266:52-9.

10. Balian JD, Sukhova N, Harris JW, Hewett J, Pickle L, Goldstein JA, et al. The hydroxylation of omeprazole correlates with S-mephenytoin metabolism: a popula- tion study. Clin Pharmacol Ther 1995;57:662-9.

11. Ward SA, Walle T, Walle UK, Wilkinson GR, Branch RA. Propranolol’s metabolism is determined by both mephenytoin and debrisoquin hydroxylase activities. Clin Pharmacol Ther 1989;45:72-9.

12. Sindrup SH, Brosen K, Hansen MGJ, Aaes-Jorgensen T, Over@ KF, Gram LF. Pharmacokinetics of citalopram in relation to the sparteine and the mephenytoin oxida- tion polymorphisms. Ther Drug Monit 1993;15:11-7.

13. Kobayashi K, Yamamoto T, Chiba K, Tani M, Ish- izaki T, Kuroiwa Y. The effects of selective serotonin reuptake inhibitors and their metabolites on S-mephenytoin 4’-hydroxylase activity in human liver microsomes. Br J Clin Pharmacol 1995;40:481-5.

14. Nakamura K, Goto F, Ray WA, McAllister CB, Jacqz E, Wilkinson GR, et al. Interethnic differences in genetic polymorphism of debrisoquin and mepheny- toin hydroxylation between Japanese and Caucasian populations. Clin Pharmacol Ther 1985;38:402-8.

1.5. Horai Y, Nakano M, Ishizaki T, Ishikawa K, Zhou HH, Zhou BI, et al. Metoprolol and mephenytoin

oxidation polymorphisms in Far Eastern Oriental sub- jects: Japanese versus mainland Chinese. Clin Phar- macol Ther 1989;46:198-207.

16. Alvan G, Bechtel P, Iselius L, Gundert-Remy U. Hy- droxylation polymorphisms of debrisoquine and me- phenytoin in European populations. Eur J Clin Phar- macol 1990;39:533-7.

17. de Morais SMF, Wilkinson GR, Blaisdell J, Nakamura K, Meyer UA, Goldstein JA. The major genetic defect responsible for the polymorphism of S-mephenytoin in humans. J Biol Chem 1994;269:15419-22.

18. de Morais SMF, Wilkinson GR, Blaisdell J, Meyer UA, Nakamura K, Goldstein JA. Identification of a new genetic defect responsible for the polymorphism of (S)-mephenytoin metabolism in Japanese. Mol Pharmacol 1994;46:594-8.

19. Masimirembwa C, Bertilsson L, Johansson I, Hasler JA, Ingelman-Sundberg M. Phenotyping and genotyp- ing of S-mephenytoin hydroxylase (cytochrome P4502C19) in a Shona population of Zimbabwe. Clin Pharmacol Ther 1995;57:656-61.

20. Brockmoller J, Rost KL, Gross D, Schenkel A, Roots I. Phenotyping of CYP2C19 with enantiospecific HPLC-quantification of R- and S-mephenytoin and comparison with the intron 4lexon 5 G+A-splice site mutation. Pharmacogenetics 1995;5:80-8.

21. Brosen K, de Morais SMF, Meyer UA, Goldstein JA. A multifamily study on the relationship between CYP2CI9 genotype and S-mephenytoin oxidation phenotype. Pharmacogenetics 1995;5:312-7.

22. Gonzalez FJ, Idle JR. Pharmacogenetic phenotyping and genotyping: present status and future potential. Clin Pharmacokinet 1994;26:59-70.

23. Brockmijller J, Roots I. Assessment of liver metabolic function: clinical implications. Clin Pharmacokinet 1994;27:216-48.

24. Dahl ML, Bertilsson L. Genetically variable metabo- lism of antidepressants and neuroleptic drugs in man. Pharmacogenetics 1993;3:61-70.

25. Sohn DR, Shin SG, Ishizaki T. S-Mephenytoin pharma- cogenetics and its clinical implications in Asian ethnic populations. Asia Pat J Pharmacol 1994;9:287-301.

26. May DG, Black CM, Olsen NJ, Csuka ME, Tanner SB, Bellino L, et al. Scleroderma is associated with differences in individual routes of drug metabolism: a study with dapsone, debrisoquin, and mephenytoin. Clin Pharmacol Ther 1990;48:286-95.

27. Flockhart DA, Clauw DJ, Sale EB, Hewett J, Woosley RL. Pharmacogenetic characteristics of the eosinophilia-myalgia syndrome. Clin Pharmacol Ther 1994;56:398-405.

28. Kaisary A, Smith P, Jacqz E, McAllister CB, Wilkin- son GR, Ray WA, et al. Genetic predisposition to bladder cancer: ability to hydroxylate debrisoquin and mephenytoin as risk factors. Cancer Res 1987;47: 5488-93.