Eur J Clin Pharmacol
DOI 10.1007/s00228-017-2250-2
PHARMACOGENETICS
Influence of genetic and non-genetic factors on phenytoin-induced
severe cutaneous adverse drug reactions
Kittika Yampayon 1 & Chonlaphat Sukasem 2,3 & Chanin Limwongse 4 & Yotin Chinvarun 5 &
Therdpong Tempark 6 & Ticha Rerkpattanapipat 7 & Pornpimol Kijsanayotin 1
Received: 14 January 2017 / Accepted: 30 March 2017
# Springer-Verlag Berlin Heidelberg 2017
Abstract
Purpose The purpose of this study was to investigate the as-
sociation of genetic factors including variants in HLA-B and
CYP2C genes and non-genetic factors with phenotype-
specific phenytoin (PHT)-induced severe cutaneous adverse
reactions (SCARs) in Thai patients.
Methods Thirty-six PHT-induced SCAR cases (15 Stevens-
Johnson syndrome (SJS) and 21 drug rash with eosinophilia
and systemic symptoms (DRESS)/drug hypersensitivity syn-
drome (DHS)) and 100 PHT-tolerant controls were studied.
Variants in HLA-B, CYP2C9, and CYP2C19 genes were ge-
notyped. Fisher’s exact test and multiple logistic regression
analysis were performed to test the association of genetic
and non-genetic factors with specific type of SCARs.
Results Multiple logistic regression models showed that ge-
netic and non-genetic factors associated with PHT-induced
SCARs were specified to its phenotype. HLA-B*13:01,
Kittika Yampayon, Chonlaphat Sukasem, and Pornpimol Kijsanayotin
contributed equally to this work.
Electronic supplementary material The online version of this article
(doi:10.1007/s00228-017-2250-2) contains supplementary material,
which is available to authorized users.
* Pornpimol Kijsanayotin
pornpimol.k@chula.ac.th
1
Department of Pharmacology and Physiology, Faculty of
Pharmaceutical Sciences, Chulalongkorn University, 254 Phyathai
Rd.,Wangmai, Patumwan, Bangkok 10330, Thailand
2
Division of Pharmacogenetics and Personalized Medicine,
Department of Pathology, Faculty of Medicine Ramathibodi
Hospital, Mahidol University, Bangkok, Thailand
3
Laboratory for Pharmacogenomics, Somdech Phra Debaratana
Medical Center (SDMC), Ramathibodi Hospital, Bangkok, Thailand
HLA-B*56:02/04, CYP2C19*3, and omeprazole co-
medication were strong risk factors of DRESS/DHS (adjusted
OR = 13.29, p = 0.0001; adjusted OR = 56.23, p = 0.0007;
adjusted OR = 6.75, p = 0.0414; and adjusted OR = 9.21,
p = 0.0020, respectively). While CYP2C9*3 and having
Chinese ancestry were significant risk factors of SJS (adjusted
OR = 10.41, p = 0.0042 and adjusted OR = 5.40, p = 0.0097,
respectively). Combined genetic and non-genetic risk factors
optimized sensitivity and increased specificity for predicting
PHT-induced SCARs.
Conclusion This study showed that distinct genetic
markers were associated with phenotype-specific PHT-in-
duced SCARs. Non-genetic factor, omeprazole co-medica-
tion, was strongly associated with PHT-induced DRESS/
DHS in addition to variants in HLA-B and CYP2C genes.
Combined markers may be better predictors for PHT-
induced SCARs.
4
Division of Medical Genetics, Department of Medicine, Faculty of
Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand
5
Neurology Unit, Department of Internal Medicine, Phramongkutklao
Hospital, Bangkok, Thailand
6
Division of Dermatology, Department of Pediatrics, Faculty of
Medicine, Chulalongkorn University, Bangkok, Thailand
7
Division of Allergy Immunology and Rheumatology, Department of
Medicine, Faculty of Medicine, Ramathibodi Hospital, Mahidol
University, Bangkok, Thailand

Keywords Phenytoin . HLA-B . CYP2C . Omeprazole . SJS/
TEN . DRESS/DHS
Introduction
The incidences of severe cutaneous adverse drug reactions
(SCARs) attributable to phenytoin (PHT), one of the most
cost-effective antiepileptic drugs (AEDs), may limit its use.
These SCARs include Stevens-Johnson syndrome (SJS), tox-
ic epidermal necrolysis (TEN), drug rash with eosinophilia
and systemic symptoms (DRESS), and drug hypersensitivity
syndrome (DHS). In Europe, the estimated risks of SJS/TEN
and DRESS/DHS from PHT were 6.9 and 2.3–4.5 per 10,000
new users, respectively [1, 2]. The incidences in Asia were as
high as 2.4 and 2.1 per 1000 users for PHT-related SJS/TEN
and DRESS, respectively [3]. In Thailand, PHT was the most
common cause of DRESS/DHS and in a top five culprit drugs
causing SJS/TEN [4].
Human leukocyte antigen (HLA) is a genetic predisposi-
tion related to hypersensitivity reaction. A strong association
of HLA-B*15:02 with carbamazepine (CBZ)-induced SJS/
TEN is found in people with Han Chinese or Southeast
Asian ancestry [5, 6]; while HLA-A*31:01 is associated with
CBZ-induced DRESS in Han Chinese and Europeans [7]. In
2008, all four PHT-induced SJS/TEN Thai patients carried
HLA-B*15:02 allele [6]. This association of HLA-B*15:02
with PHT-induced SJS/TEN was further found in Han
Chinese and Taiwanese [3, 8, 9]. However, the association
evidence of PHT-induced DRESS/DHS was limited. Distinct
SCAR phenotypes may have different genetic predispositions
across populations.
For drug metabolism variations, CYP2C9 is the major en-
zyme metabolizing up to 90% of PHT to the inactive 5-(4-
hydroxyphenyl)-5-phenylhydantoin (p-HPPH), while
CYP2C19 metabolizes the remaining 10% [10]. The loss of
a function variant, CYP2C9*3, has been associated with PHT-
induced SCARs in Asian population [3, 11, 12]. Drug inter-
action could also be a factor that lowers PHT metabolism
which could further cause SCARs, as seen in lamotrigine-
associated rash [13].
In addition, SJS was frequently found in brain malignancy
patients with co-treatment of PHT and radiation [14]. The
incidence of rash due to PHT in patients who had allergy to
other AEDs was 6.3-fold higher than those who did not [15].
Therefore, PHT hypersensitivity could be attributable to mul-
tiple factors. Non-genetic factors, such as co-medications,
medical treatments, and patient status possibly contribute to
PHT-induced SCARs. However, no association study exam-
ining non-genetic factors has been done. This study aimed to
investigate the association of genetic factors including vari-
ants in HLA-B and CYP2C genes together with non-genetic
Eur J Clin Pharmacol
factors with phenotype-specific PHT-induced SCARs in Thai
patients.
Methods
Cases and controls
All cases and PHT-tolerant controls were recruited from three
medical school hospitals in Bangkok, Thailand, including
Siriraj Hospital, Ramathibodi Hospital, and
Phramongkutklao Hospital. Patients with PHT use found in
medical records between January 2008 and June 2015 were
eligible. Cases were defined as Thai adult patients who had
PHT-induced SCARs; while PHT-tolerant controls were those
with no PHT-induced cutaneous adverse reactions (cADRs)
and had taken PHT for at least 3 months.
Forty-two cases and 100 PHT-tolerant controls were re-
cruited. Based on the participants’ medical records, their de-
mographic and clinical information, laboratory findings, ADR
treatments, onset date, dosage, duration, concomitant drugs,
and history of drug allergies were reviewed. All cases were
further confirmed by the same dermatologist using consensus
diagnostic and phenotyping criteria. PHT has been identified
as the principle cause of skin lesions in all 42 patient cases
according to the Naranjo algorithm [16]. For SJS/TEN cases,
the causality was further assessed by ALDEN [17]. Based on
Roujeau’s criteria [18], SJS and TEN were characterized as
rapidly developing dusky purpuric macules, atypical target-
like lesions, or blisters accompanied by mucosal involvement
with skin detachment. Distinct cases of SJS, SJS-TEN over-
lap, and TEN were differentiated based on the extent of skin
area detachment ranging from less than 10% of the body sur-
face area (BSA), 10 to 30% BSA, and greater than 30% BSA,
respectively [18]. DRESS and DHS were assessed using
RegiSCAR criteria which consist of acute skin rash together
with the involvement of at least one internal organ (e.g., hep-
atitis and nephritis), lymphadenopathy, hematologic abnor-
mality (e.g., eosinophilia and atypical lymphocytosis), and
fever [19]. In addition to above criteria, to be classified as
DRESS cases, the absolute eosinophil counts had to be
1500/μL or greater. Participants’ ancestry information was
self-reported.
Based on the protocol mentioned above, one SJS patient
was excluded because the reaction occurred later than
3 months after the PHT exposure. With another five patients
in the DRESS/DHS group reclassified as PHT-induced
maculopapular eruption, a total of 36 PHT-induced SCARs
cases remained.
Additionally, we obtained 758 and 250 anonymous DNA
samples of the Thai population with pre-determined HLA-B
and CYP2C genotypes, respectively. These samples, provided
by the Laboratory for Pharmacogenomics, Somdech Phra
Eur J Clin Pharmacol
Debaratana Medical Center (SDMC), Ramathibodi Hospital,
Bangkok, Thailand, were collected during August 2011–
June 2014 [20].
This study was approved by the institutional review board
of each hospital, specifically, Siriraj Hospital Institutional
Review Board (471/2557 EC1), Ethics Committee of the
Faculty of Medicine, Ramathibodi Hospital (ID 05-58-17),
and Institutional Review Board of the Royal Thai Army
Medical Department (Q017h/57). Written informed consent
was obtained from each participant.
Blood sampling and DNA extraction
Peripheral blood samples were obtained from all participants.
Genomic DNA was extracted using the QIAamp DNA Blood
Mini Kit (Qiagen, Hilden, Germany) or the MagNA Pure
automated extraction system (Roche Diagnostics, USA).
HLA-B genotyping
HLA-B typing was performed using the polymerase chain
reaction-sequence-specific oligonucleotide probe (PCR-
SSOP) (LABtype®, One Lambda, CA, USA) with the
Luminex method following the manufacturer’s instruction
[21]. The amplicon-probe complexes were measured by the
Luminex® IS 100 flow cytometer (Luminex, TX, USA).
HLA-B alleles were analyzed using HLA Fusion™ 3.0 soft-
ware (One Lambda, CA, USA).
Since the genotyping method used could not distinguish
the HLA-B*56:02 from 56:04 alleles which are in the same
serology group of B56 [22], we pooled these two alleles for
data analysis.
CYP2C9 and CYP2C19 genotyping
The polymorphisms of CYP2C9 and CYP2C19 genes, includ-
ing CYP2C9*2 (rs1799853), CYP2C9*3 (rs1057910),
CYP2C19*2 (rs4244285), CYP2C19*3 (rs4986893), and
CYP2C19*17 (rs12248560), were determined by TaqMan
SNP Genotyping Assays (Applied Biosystems, Foster City,
CA, USA).
Statistical analysis
For the comparisons of demographic and clinical characteris-
tics between PHT-induced SCAR cases and PHT-tolerant con-
trols, Student’s t test or Mann-Whitney U test was used as
appropriate, for continuous variables such as age, PHT dos-
age, and duration, and Pearson’s chi-square test or Fisher’s
exact test was used appropriate for categorical variables such
as gender and birthplace.
The Hardy-Weinberg equilibrium (HWE) was verified for
HLA-B genotypes and all examined SNPs of CYP2C using
PyPop 0.7.0 software (University of California, USA) [23]
and Pearson’s chi-square test, respectively. All HLA-B alleles
and CYP2C variants were in HWE (Supplementary material
Tables 1 and 2).
For the initial bivariate analysis, Fisher’s exact test was first
used to determine the association between phenotype-specific
PHT-induced SCARs and each of individual factors. For a
zero in any cells of a contingency table, Haldane’s modifica-
tion, i.e., adding 0.5 to all cells of a 2 × 2 table, was applied to
estimate the odds ratio (OR) [24]. Candidate genetic factors,
including CYP2C9 and CYP2C19 variants and HLA-B alleles,
were selected based on the following criteria: alleles found in
more than 5% of SCAR cases (Supplementary material
Table 1), and alleles where frequencies in SJS/TEN or
DRESS/DHS cases and PHT-tolerant controls differed by
more than 3%. Attentive non-genetic factors including co-
medications, comorbidity, past medical history, drug allergy
history, and patient’s ancestry were tested for the association
with specific type of SCARs. The association of co-
medications with SCARs was tested only for those with drug
interaction significance levels of 1 to 4 with PHT [25]. All co-
medications were used with unchanged doses for at least
5 days.
Multiple logistic regression analysis with forward stepwise
likelihood ratio method was performed to determine the asso-
ciation of multiple factors with phenotype-specific PHT-in-
duced SCARs. In case of quasi-complete separation, exact
logistic regression was applied to estimate p value and OR
for that variable [26]. Genetic and non-genetic factors that
showed to be risk factors of PHT-induced SJS or DRESS/
DHS from bivariate analysis (p < 0.05) were selected as can-
didate covariates in the multivariate test. Since HLA-B*15:02
has previously shown to be associated with PHT-induced SJS/
TEN, this allele was included in the multivariate analysis [3, 6,
8, 9]. The variance inflation factor (VIF) was assessed for
multicollinearity.
All statistical analyses were performed using SAS
University Edition (SAS Institute Inc., Cary, NC, USA).
ORs with 95% confidence interval (CI) were reported. A
two-tailed p value of <0.05 was considered statistically signif-
icant, and Bonferroni’s correction was applied for multiple
comparisons [27].
Results
Demographic and clinical characteristics of patients
The demographic and clinical characteristics of 36 PHT-
induced SCARs (15 SJS and 21 DRESS/DHS) and 100
PHT-tolerant controls are summarized in Table 1. Liver was
the most common internal organ involvement. All 21 DRESS/
DHS cases had hepatitis. Two or more involved mucosal sites,
 Eur J Clin Pharmacol

Table 1 Demographic and clinical characteristics of phenytoin-induced SCARs and phenytoin-tolerant controls

Characteristics Case of PHT-induced SCARs PHT-tolerant controls p valuea p valueb

(n = 100)
SJS (n = 15) DRESS/DHS Total SCARs
(n = 21) (n = 36)

Gender, no. (%)

Female 9 (60.0) 13 (61.9) 22 (61.1) 64 (64.0) 0.7795c 1.0000c
Male 6 (40.0) 8 (38.1) 14 (38.9) 36 (36.0)
Age, years
Mean ± SD 57.9 ± 12.7 49.2 ± 19.0 52.9 ± 17.0 49.0 ± 12.3 0.0104d 0.9566d
Birthplace, no. (%)
Central region of Thailand 11 (73.3) 15 (71.4) 26 (72.2) 62 (62.0) 0.5670c 0.4651c
Others 4 (26.7) 6 (28.6) 10 (27.8) 38 (38.0)
PHT exposure
Dosage, median (range), mg/day 300 (200–300) 300 (200–300) 300 (200–300) 300 (100–400) 0.7984e 0.5015e
Duration, median (range), days 26 (7–65) 25 (8–51) 26 (7–65) 1008 (91–8972) <0.0001e <0.0001e
Onset of ADR, median (range), days 21 (7–61) 23 (8–48) 22 (7–61) – –
Internal organ involvement
Liver, no. (%) 8 (53.3) 21 (100) 29 (80.6) – –
AST, IU/L, no. (%)
<100 9 (60.0) 5 (23.8) 14 (38.9) – –
100–499 5 (33.3) 13 (61.9) 18 (50.0) – –
≥500 1 (6.7) 3 (14.3) 4 (11.1) – –
ALT, IU/L, no. (%)
<100 7 (46.7) 0 (0.0) 7 (19.4) – –
100–499 7 (46.7) 18 (85.7) 25 (69.4) – –
≥500 1 (6.7) 3 (14.3) 4 (11.1) – –
ALP, IU/L, no. (%)
<300 14 (93.3) 16 (76.2) 30 (83.3) – –
300–499 1 (6.7) 2 (9.5) 3 (8.3) – –
≥500 0 (0.0) 3 (14.3) 3 (8.3) – –
Kidney, no. (%) 0 (0.0) 1 (4.8) 1 (2.8) – –
Lung, no. (%) 1 (6.7) 0 (0.0) 1 (2.8) – –
Muscle/heart, no. (%) 0 (0.0) 3 (14.3) 3 (8.3) – –
Hematologic abnormality
Absolute eosinophil counts/μL, no. (%)
<700 14 (99.3) 12 (57.1) 26 (72.2) – –
700–1499 1 (6.7) 4 (19.0) 5 (13.9) – –
≥1500 0 (0.0) 5 (23.8) 5 (13.9) – –
Atypical lymphocytosis, no. (%) 0 (0.0) 10 (47.6) 10 (27.8) – –
Mucosal involvement, no. (%)
Oral 15 (100) 3 (14.3) 18 (50.0) – –
Eyes 15 (100) 1 (4.8) 16 (44.4) – –
Genitalia 3 (20.0) 0 (0.0) 3 (8.3) – –

ALP alkaline phosphatase, AST aspartate aminotransferase, ALT alanine aminotransferase, DHS drug hypersensitivity syndrome, DRESS drug rash with
eosinophilia and systemic symptoms, PHT phenytoin, SCARs severe cutaneous adverse drug reactions, SJS Stevens-Johnson syndrome
a
PHT-induced SJS cases and PHT-tolerant controls were compared
b
PHT-induced DRESS/DHS cases and PHT-tolerant controls were compared
c
p Value was calculated by Fisher’s exact test (two-tailed)
d
p Value was calculated by Student’s t test
e
p Value was calculated by Mann-Whitney U test
Eur J Clin Pharmacol
with at least oral and ocular mucosa, were found in all 15 SJS
cases.
Bivariate analysis on genetic factors associated
with phenotype-specific PHT-induced SCARs
Fifty-three HLA-B alleles were characterized from all cases
and PHT-tolerant controls where HLA-B*13:01 was the most
common allele found in DRESS/DHS cases, with an allele
frequency of 28.57% (data not shown). HLA-B*13:01 showed
a strong association with DRESS/DHS when cases were com-
pared with tolerant controls (OR = 6.76, p = 0.0003) and
general population (OR = 7.07, p < 0.0001), even after
Bonferroni’s correction (p < 0.004). HLA-B*56:02/04 alleles
were strongly associated with DRESS/DHS with the highest
odds ratios of 38.03 (p = 0.0046) and 31.42 (p = 0.0006) in
control- and general population-based comparisons, respec-
tively. Regarding drug-metabolizing enzyme variants,
CYP2C19*3 was significantly associated with DRESS/DHS
incident (OR = 4.47, p = 0.0478) whereas CYP2C9*3 was not
(p = 0.6264) (Table 2).
For PHT-induced SJS, a carrier rate of HLA-B*15:02 in
SJS cases (33.3%) was higher than that in tolerant controls
(18.0%) and the general population (14.2%) with no statistical
significance. None of HLA-B alleles showed an association
with PHT-induced SJS (Table 2 and Supplementary material
Table 3). However, CYP2C9*3 was associated with SJS inci-
dents when compared with both tolerant controls (OR = 5.70,
p = 0.0251) and the general population (OR = 3.97,
p = 0.0413) (Table 2).
Bivariate analysis on non-genetic factors associated
with phenotype-specific PHT-induced SCARs
Among concomitant medications, omeprazole was found
more frequent in DRESS/DHS cases than in tolerant controls,
which was statistically significant (OR = 5.50, p = 0.0022). In
addition, Chinese ancestry was associated with an increased
risk of SJS (OR = 3.43, p = 0.0330). History of drug allergies
and other non-genetic factors were not associated with PHT-
induced SCARs (Table 3 and Supplementary material
Table 4).
Multiple logistic regression analysis on the factors
associated with phenotype-specific PHT-induced SCARs
As indicated by bivariate analysis results, further multiple lo-
gistic regression analyses revealed distinctive factors associ-
ated with PHT-induced DRESS/DHS and SJS. Factors signif-
icantly associated with an increased risk of PHT-induced
DRESS/DHS included three genetic factors, namely, HLA-
B*13:01, 56:02/04, and CYP2C19*3 (adjusted OR = 13.29,
p = 0.0001, adjusted OR = 56.23, p = 0.0007, and adjusted
OR = 6.75, p = 0.0414, respectively), and a non-genetic factor,
specifically omeprazole co-medication (adjusted OR = 9.21,
p = 0.0020). These four factors could explain about 50% of
variability in the occurrence of PHT-induced DRESS/DHS
(R2 = 0.486) (Table 4).
Regarding PHT-induced SJS, a final logistic regression
model showed that CYP2C9*3 and Chinese ancestry were
significant risk factors of PHT-induced SJS (adjusted
OR = 10.41, p = 0.0042 and adjusted OR = 5.40,
p = 0.0097, respectively). This model could explain about
20% of the variation in the occurrence of PHT-induced SJS
(R2 = 0.189) (Table 4).
Single and combined markers to predict
phenotype-specific PHT-induced SCARs
Sensitivity and specificity analysis was performed for each
phenotype of SCARs based on the results of association anal-
ysis. The sensitivity and specificity of HLA-B*13:01 allele to
predict PHT-induced DRESS/DHS were 52.4 and 86.0%, re-
spectively. HLA-B*56:02/04 alleles were highly specific
(100%) but with low sensitivity (14.3%). Combining both
HLA-B markers improved sensitivity of testing. The combined
HLA-B*13:01/HLA-B*56:02/04 increased sensitivity of
predicting PHT-induced DRESS/DHS to 66.7%. In addition,
the integration of HLA-B*13:01/HLA-B*56:02/04 and a drug-
metabolizing variant, CYP2C19*3, increased specificity of
the test to 100% with a decreased sensitivity. Adding a non-
genetic factor, omeprazole co-medication, improved sensitiv-
ity of the screening from 14.3 to 28.6% with a comparable
specificity (98%). Furthermore, using the combined HLA-
B*13:01/HLA-B*56:02/04 together with CYP2C19*3/ omep-
razole co-medication showed the highest accuracy in
predicting DRESS/DHS (86.0%) (Table 5).
To predict PHT-induced SJS, CYP2C9*3 and Chinese an-
cestry were significant markers with the sensitivity of 26.7 and
53.3%, respectively. Combining these two markers increased
the sensitivity to 73.3%, with a decreased specificity. Adding
HLA-B*15:02 allele with the combined CYP2C9*3/Chinese
ancestry improved the specificity of testing from 69 to 96%.
Using combined markers showed a better prediction ability
(Table 5).
Discussion
Our findings indicate that the pathological mechanisms of
PHT-induced SCARs could be related to multiple factors in-
cluding genetic variations in the immunologic system and
drug metabolism pathways and certain non-genetic factors.
The variation in HLA allele is a part of genetic factors
involved in different activating immune responses. Our results
suggest that patients carrying HLA-B*13:01 or HLA-B*56:02/
 Eur J Clin Pharmacol

Table 2 Bivariate analysis on genetic factors among phenotype-specific phenytoin-induced SCARs cases, phenytoin-tolerant controls, and general
Thai population

Genetic factors Carrier (%) Cases vs PHT-tolerant controls Cases vs general Thai population

Cases PHT-tolerant controls General Thai population pa,b OR (95% CI) pa,b OR (95% CI)

SJS (n = 15) (n = 100) (n = 758c/250d)

HLA-B
B*13:01 3 (20.0) 14 (14.0) 102 (13.5)c 0.4637 1.54 (0.38–6.14) 0.4430 1.61 (0.45–5.80)
B*15:02 5 (33.3) 18 (18.0) 108 (14.2)c 0.1767 2.28 (0.69–7.48) 0.0544 3.01 (1.01–8.97)
B*56:02/ 04 0 (0.0) 0 (0.0) 4 (0.5)c – – – 1.0000 5.41 (0.28–104.87)
CYP
2C9*3 4 (26.7) 6 (6.0) 21 (8.4)d 0.0251* 5.70 (1.39–23.36) 0.0413* 3.97 (1.16–13.55)
2C19*3 1 (6.7) 5 (5.0) 14 (5.6)d 0.5763 1.36 (0.15–12.48) 0.5929 1.20 (0.15–9.82)
DRESS/DHS (n = 21) (n = 100) (n = 758c/250d)
HLA-B
B*13:01 11 (52.4) 14 (14.0) 102 (13.5)c 0.0003** 6.76 (2.42–18.85) <0.0001** 7.07 (2.93–17.08)
B*15:02 0 (0.0) 18 (18.0) 108 (14.2)c 0.0402* 0.10 (0.01–1.79) 0.0995 0.14 (0.01–2.32)
B*56:02/ 04 3 (14.3) 0 (0.0) 4 (0.5)c 0.0046* 38.03 (1.88–767.19) 0.0006** 31.42 (6.55–150.74)
CYP
2C9*3 2 (9.5) 6 (6.0) 21 (8.4)d 0.6264 1.65 (0.31–8.80) 0.6951 1.15 (0.25–5.27)
2C19*3 4 (19.0) 5 (5.0) 14 (5.6)d 0.0478* 4.47 (1.09–18.36) 0.0399* 3.97 (1.18–13.37)

CI confidence interval, CYP cytochrome P450, DHS drug hypersensitivity syndrome, DRESS drug rash with eosinophilia and systemic symptoms, HLA
human leukocyte antigen, OR odds ratio, p p value, PHT phenytoin, SJS Stevens-Johnson syndrome
*Significant difference at p value <0.05; **significant difference at p value <0.004 (0.05/12) for HLA-B, <0.025 (0.05/2) for CYP2C9, and <0.017 (0.05/
3) for CYP2C19 when adjusted for multiple comparisons using Bonferroni correction
a
The reported data were based on the analysis of the dominant inheritance model for the risk genotypes
b
p Value was calculated by Fisher’s exact test (two-tailed) for the comparisons
c
For HLA-B, data of 758 persons of the general Thai population were obtained from the Laboratory for Pharmacogenomics, SDMC
d
For CYP2C9 and CYP2C19, data of 250 persons of the general Thai population were obtained from the Laboratory for Pharmacogenomics, SDMC

04 have a much higher risk of PHT-induced DRESS/DHS. association of HLA-B*13:01 with SCARs [3]. In addition,
The strong association between HLA-B*13:01 and PHT- our finding is in concordance with a recent phenobarbital
induced DRESS/DHS in this study supports the signaling (PB) study in Thais where HLA-B*13:01 was associated with
finding from a GWAS by Chung et al. that found a weak PB-induced SCARs and DRESS [28]. The association of

Table 3 Bivariate analysis on non-genetic factors between phenotype-specific phenytoin-induced SCARs cases and phenytoin-tolerant controls

Non-genetic factors Number of patients (%) SJS cases vs PHT-tolerant DRESS/DHS cases vs PHT-tolerant
controls controls

SJS cases DRESS/ DHS PHT-tolerant pa OR (95% CI) pa OR (95% CI)

cases controls

(n = 15) (n = 21) (n = 100)

Co-medicationb
Omeprazole 4 (26.7) 9 (42.9) 12 (12.0) 0.2206 2.67 (0.73–9.72) 0.0022** 5.50 (1.92–15.78)
Chinese ancestry 8 (53.3) 4 (19.0) 25 (25.0) 0.0330* 3.43 (1.13–10.41) 0.7794 0.71 (0.22–2.30)

CI confidence interval, DHS drug hypersensitivity syndrome, DRESS drug rash with eosinophilia and systemic symptoms, OR odds ratio, p p value, PHT
phenytoin, SJS Stevens-Johnson syndrome
*Significant difference at p value <0.05; **significant difference at p value <0.008 (0.05/6) for co-medications when adjusted for multiple comparisons
using Bonferroni’s correction
a
p Value was calculated by Fisher’s exact test (two-tailed) for the comparisons
b
Patient could have more than one co-medication
Eur J Clin Pharmacol

Table 4 Multiple logistic regression analysis on genetic and non-genetic factors associated with phenotype-specific phenytoin-induced SCARs (final
logistic regression model)

Factors SJS cases vs PHT-tolerant controls DRESS/DHS cases vs PHT-tolerant controls

b pa Adjusted OR (95%CI)c b pa Adjusted OR (95% CI)c

Genetic factors
HLA-B*13:01 – – – 2.587 0.0001b 13.29 (3.79–56.91)
HLA-B*15:02 – – – – – –
HLA-B*56:02/ 04 – – – 4.029d 0.0007d 56.23 (7.17–∞)d
CYP2C9*3 2.343 0.0042b 10.41 (2.06–55.42) – – –
CYP2C19*3 – – – 1.910 0.0414b 6.75 (1.05–44.74)
Non-genetic factors
Omeprazole co-medication – – – 2.220 0.0020b 9.21 (2.35–42.21)
Chinese ancestry 1.687 0.0097b 5.40 (1.57–21.62) – – –
Constant −1.115 −20.319
Max–rescaled R2 = 0.189 Max–rescaled R2 = 0.486
Model p value = 0.0021 Model p valve <0.0001

Candidate covariates included in the logistic regression analysis of each group comparison were the following: HLA-B*13:01, 15:02, 56:02/04,
CYP2C9*3, CYP2C19*3, using omeprazole co-medication and patient’s ancestry
b logistic coefficient, CI confidence interval, DHS drug hypersensitivity syndrome, DRESS drug rash with eosinophilia and systemic symptoms, OR
odds ratio, PHT phenytoin, SJS Stevens-Johnson syndrome
a
The reported data were based on the analysis of the dominant inheritance model for the risk genotypes
b
p Value was calculated by forward stepwise logistic regression analysis (likelihood ratio method)
c
OR was estimated by logistic regression analysis after being adjusted by other factors included in the final model of each group comparison
d
With a quasi-complete separation of the data, the coefficient (b), p value, and OR were estimated by exact logistic regression [26]

HLA-B*13:01 is not limited only to AEDs; HLA-B*13:01 was
also reported as a risk factor of sulfasalazine-induced DRESS/
DHS and dapsone hypersensitivity syndrome [29, 30]. These
evidences suggest that HLA-B*13:01 may be a universal ge-
netic marker for predicting DRESS/DHS.
Furthermore, HLA-B*56:02/04 may be good genetic pre-
dictors for PHT-induced DRESS/DHS. These alleles were ob-
served only in DRESS/DHS cases, but absent in PHT-tolerant
controls. The carrier rate was 30-fold higher than that in the
general Thai population.
However, the associations of HLA-B*51:01 with DRESS
and HLA-B*38:02, 51:01, 56:02, and 58:01 with SJS/TEN
recently reported by Tassaneeyakul et al. [31] were not found
in this study. The variation of HLA-B markers may be a result
from the different distribution of HLA-B alleles in Thai popu-
lation [32]. About 60% of patients in our study were from the
central region of Thailand while those in the other study were
recruited from local hospitals in the north-east region [31].
Although HLA-B*15:02 was reported as a risk factor of
PHT-induced SJS/TEN [3, 6, 8, 9], significant association of
this allele alone with SJS was not found in this study. HLA-
B*15:02 was observed in 33.3% of SJS patients, which was
comparable to those in previous reports in Taiwanese and Han
Chinese (27–47%) [3, 8, 9]. However, the carrier rates of
HLA-B*15:02 in PHT-tolerant controls were higher in Thai
patients than Chinese patients (14.1–18.0 vs 6.9–8.0%,
respectively) [3, 8]. These differences could be attributable
to other influential ethnicity factors. This suggestion is sup-
ported by the findings that patients who have Chinese ancestry
had a much higher risk of PHT-induced SJS, and a better
prediction was observed when HLA-B*15:02 and Chinese
ancestry were considered together.
Apart from the immune system variability, drug metabo-
lism variation also results in varied drug hypersensitivity. A
loss of function variant CYP2C9*3 is a genetic factor associ-
ated with PHT-induced cADRs in several ethnic groups, such
as Korean, Taiwanese, Malaysian, Japanese, and Thai [3, 11,
12, 31]. Our finding that CYP2C9*3 was associated only with
SJS/TEN is in contrast with a Taiwanese study [3], and similar
to the recent study in Thais [31]. An association between
CYP2C19*3, a poor metabolism variant of CYP2C19, and
PHT-induced DRESS/DHS was observed in our study. This
may be resulted from differences in allele frequencies across
populations [3, 33, 34].
The co-medication of omeprazole was found to increase
the risk of PHT-induced DRESS/DHS for the first time in
our study. A trend of association (p = 0.06) of higher PHT
blood level with either CYP2C9*3/CYP2C19*3 carriers or
omeprazole co-medication was observed (Supplementary
material Fig. 1). The finding suggests that an increased PHT
blood level associated with medication-induced enzyme inhi-
bition is in line with the effect of genetic-influenced poor
 Eur J Clin Pharmacol

Table 5 Characteristics of single and combined markers to predict phenotype-specific phenytoin-induced SCARs in Thai patients

Markers Positive Positive number of pa OR (95%CI) Sensitivity Specificity Accuracy

number of tolerant controls (%) (%) (%) (%)
cases (%)

SJS
Single marker
HLA-B*15:02 5 (33.3) 18 (18.0) 0.1767 2.28 (0.69–7.48) 33.3 82.0 75.7
CYP2C9*3 4 (26.7) 6 (6.0) 0.0251 5.70 (1.39–23.36) 26.7 94.0 85.2
Chinese ancestry 8 (53.3) 25 (25.0) 0.0330 3.43 (1.13–10.41) 53.3 75.0 72.2
Combined markers
CYP2C9*3 or Chinese ancestry 11 (73.3) 31 (31.0) 0.0029 6.12 (1.81–20.74) 73.3 69.0 69.6
CYP2C9*3 or Chinese ancestry 4 (26.7) 4 (4.0) 0.0099 8.73 (1.91–39.90) 26.7 96.0 87.0
AND HLA-B*15:02
CYP2C9*3 and HLA-B*15:02 1 (6.7) 0 (0) 0.1304 20.79 (0.81–535.02) 6.7 100 87.8
Chinese ancestry and HLA-B*15:02 4 (26.7) 4 (4.0) 0.0099 8.73 (1.91–39.90) 26.7 96.0 87.0
Chinese ancestry and HLA-B*15:02 7 (46.7) 10 (10.0) 0.0014 7.88 (2.36–26.32) 46.7 90.0 84.3
OR CYP2C9*3
Chinese ancestry and HLA-B*15:02 1 (6.7) 0 (0) 0.1304 20.79 (0.81–535.02) 6.7 100 87.8
AND CYP2C9*3
DRESS/DHS
Single marker
HLA-B*13:01 11 (52.4) 14 (14.0) 0.0003 6.76 (2.42–18.85) 52.4 86.0 80.2
HLA-B*56:02/04 3 (14.3) 0 (0) 0.0046 38.03 (1.88–767.19) 14.3 100 85.1
CYP2C19*3 4 (19.0) 5 (5.0) 0.0478 4.47 (1.09–18.36) 19.0 95.0 81.8
Omeprazole co-medication 9 (42.9) 12 (12.0) 0.0022 5.50 (1.92–15.78) 42.9 88.0 80.2
Combined markers
HLA-B*13:01 or 56:02/04 14 (66.7) 14 (14.0) <0.0001 12.29 (4.22–35.77) 66.7 86.0 82.6
CYP2C19*3 or omeprazole 12 (57.1) 16 (16.0) 0.0002 7.00 (2.53–19.34) 57.1 84.0 79.3
co-medication
HLA-B*13:01 or 56:02/04 AND 3 (14.3) 0 (0) 0.0046 38.03 (1.88–767.19) 14.3 100 85.1
CYP2C19*3
HLA-B*13:01 or 56:02/04 AND 6 (28.6) 2 (2.0) 0.0003 19.60 (3.62–106.23) 28.6 98.0 86.0
CYP2C19*3 or omeprazole
co-medication

CI confidence interval, DHS drug hypersensitivity syndrome, DRESS drug rash with eosinophilia and systemic symptoms, OR odds ratio, PHT
phenytoin, SJS Stevens-Johnson syndrome
a
p Value was calculated by Fisher’s exact test (two-tailed) for the comparisons between cases and PHT-tolerant controls

metabolism. In addition, as shown in the sensitivity analysis,
adding omeprazole co-medication as a co-marker resulted in
the improvement of sensitivity for predicting PHT-induced
DRESS/DHS.
It is possible that drug concentration and HLA-B alleles
synergistically induced SCARs. This suggestion is supported
by our finding that patients carrying HLA-B*13:01 or HLA-
B*56:02/04, together with CYP2C19*3 or omeprazole co-
medication, were at a higher risk of PHT-induced DRESS/
DHS than patients having risk factors relating only to either
immune system or drug metabolism. Even though there is no
statistically significant difference, PHT blood levels in the
DRESS/DHS group tended to be higher than those in the
tolerant control group (Supplementary material Fig. 2).
Similar finding observed in PHT-induced SJS that combined
CYP2C9*3 and HLA-B*15:02 showed the highest specificity
of 100%, while no statistical significance could be found due
to a small sample size. However, combining two risk factor
types yielded an increased specificity of the test which could
be useful in a predictive screening for PHT-induced SCARs.
Regarding PHT-induced SJS, HLA-B*15:02 was the most
frequent allele found in this group. Although there was no
statistical significance when compared with PHT-tolerant con-
trols, a trend of association between HLA-B*15:02 and PHT-
induced SJS was observed in a comparison with the general
Thai population (p = 0.054). Therefore, HLA-B*15:02 was
also considered in sensitivity and specificity analysis.
According to our findings, two possible mechanisms of
PHT-stimulating immune responses could be proposed.
Firstly, since CYP2C9 and CYP2C19 are primary enzymes
responsible for the first step of PHT transformations [10, 35],
poor metabolizer phenotype and enzyme inhibitor could
Eur J Clin Pharmacol
increase PHT blood level. Thus, it is possible that unchanged
PHT (parent drug) is the key signaling molecule in the im-
mune response. Parent drugs that directly bind to HLA mole-
cules have been reported for CBZ [36]. Secondly, the forma-
tion of the inactive metabolite of PHT, p-HPPH, is thought to
proceed via an arene oxide intermediate which may contribute
to PHT hypersensitivity [37]. CYP2C9 and CYP2C19 are also
involved in the clearance of arene oxide [10, 35]. Therefore,
inhibiting and decreasing function of these enzymes may re-
sult in a clearance reduction and lead to the accumulation of
reactive metabolites. These metabolites could be potential
molecules to initiate immune activation. Similarly,
oxypurinol, an allopurinol metabolite, has been suggested as
a major cause of allopurinol hypersensitivity [38, 39]. With a
different chiral configuration, the formation of (S)-p-HPPH is
predominantly catalyzed by CYP2C9, while CYP2C19 is
more stereoselective for (R)-p-HPPH [40, 41]. p-HPPH may
be generated by different stereoisomers of reactive intermedi-
ates. This may explain the associations of distinct CYP2C
variants with PHT-induced SJS and DRESS/DHS.
Moreover, omeprazole co-medication as a predominant risk
factor of PHT-induced DRESS/DHS could be due to its more
potent inhibition of CYP2C19 than CYP2C9 [42].
Based on the findings up to date including ours, larger
investigations, such as multicenter studies and meta-analysis,
are needed to confirm the associations of genetic and non-
genetic factors with phenotype-specific PHT-induced
SCARs in various populations. Further prospective observa-
tional studies are also required to determine the influence of
CYP2C inhibitors on blood levels of PHT and its metabolites
and their association with SCARs. The pathogenic mecha-
nisms of multiple factors on different types of PHT-induced
SCARs in an HLA-dependent pathway should be studied.
Conclusion
This study showed that distinct genetic markers were associ-
ated with phenotype-specific PHT-induced SCARs. HLA-
B*13:01, HLA-B*56:02/04, and CYP2C19*3 were potential
genetic markers of PHT-induced DRESS/DHS in Thai pa-
tients. The strong association between an enzyme inhibitor
and PHT-induced DRESS/DHS is described for the first time
in this study. CYP2C9*3 and HLA-B*15:02 with Chinese eth-
nicity were found to be associated with PHT-induced SJS/
TEN. Both genetic and non-genetic factors may work in con-
cert in pathogenesis of PHT-induced SCARs, and combined
markers could be better predictors.
Acknowledgments This study was supported by the
Ratchadaphiseksomphot Endowment Fund of Chulalongkorn
University and the 90th anniversary of Chulalongkorn University Fund
(GCUCR1125572100M No.86). The authors would like to thank Assist.
Prof. Charoen Treesak, Department of Clinical Pharmacy, Faculty of
Pharmacy, Srinakharinwirot University, for constructive comments and
manuscript revision as well as Assist. Prof. Chulaluk Komoltri,
Department of Health Research and Development, Siriraj hospital,
Mahidol University, for statistical consultation. Finally, we are grateful
to all patients for participating in this study.
Authors’ contribution Kittika Yampayon’s contribution included
study design, data collection, statistical analysis, and drafting and revising
the manuscript. Chonlaphat Sukasem and Chanin Limwongse contribut-
ed to study design, supervised the study, contributed samples, and revised
the manuscript content. Yotin Chinvarun contributed samples and data
and revised the manuscript content. Therdpong Tempark contributed phe-
notype data of the study participants and revised the manuscript content.
Ticha Rerkpattanapipat contributed samples and revised the manuscript
content. Pornpimol Kijsanayotin contributed to study design, supervised
the study, supported the statistical analyses, and critically revised the
manuscript. All authors read and approved the final manuscript.
Compliance with ethical standards All procedures performed in this
study involving human participants were in accordance with the ethical
standards of the institutional review board of each hospital and with the
1964 Helsinki Declaration and its later amendments or comparable ethical
standards. Informed consent was obtained from all individual participants
included in the study.
Conflict of interest The authors declare that they have no conflict of
interest.
References
1. Knowles SR, Shapiro LE, Shear NH (1999) Anticonvulsant hyper-
sensitivity syndrome: incidence, prevention and management. Drug
Saf 21(6):489–501
2. Mockenhaupt M, Messenheimer J, Tennis P, Schlingmann J (2005)
Risk of Stevens-Johnson syndrome and toxic epidermal necrolysis
in new users of antiepileptics. Neurology 64(7):1134–1138. doi:10.
1212/01.wnl.0000156354.20227.f0
3. Chung W, Chang W, Lee Y et al (2014) Genetic variants associated
with phenytoin-related severe cutaneous adverse reactions. JAMA
312(5):525–534. doi:10.1001/jama.2014.7859
4. Health Product Vigilance Center (HPVC), Food and Drug
Administration of Thailand (2014) Spontaneous reports of adverse
drug reaction 2014. Aksorn graphic and design publication limited
partnership, Bangkok
5. Chung WH, Hung SI, Hong HS, Hsih MS, Yang LC, Ho HC, Wu
JY, Chen YT (2004) Medical genetics: a marker for Stevens-
Johnson syndrome. Nature 428(6982):486. doi:10.1038/428486a
6. Locharernkul C, Loplumlert J, Limotai C, Korkij W, Desudchit T,
Tongkobpetch S, Kangwanshiratada O, Hirankarn N,
Suphapeetiporn K, Shotelersuk V (2008) Carbamazepine and phe-
nytoin induced Stevens-Johnson syndrome is associated with HLA-
B*1502 allele in Thai population. Epilepsia 49(12):2087–2091.
doi:10.1111/j.1528-1167.2008.01719.x
7. Genin E, Chen DP, Hung SI, Sekula P, Schumacher M, Chang PY, Tsai
SH, Wu TL, Bellon T, Tamouza R, Fortier C, Toubert A, Charron D,
Hovnanian A, Wolkenstein P, Chung WH, Mockenhaupt M, Roujeau
JC (2014) HLA-A*31:01 and different types of carbamazepine-induced
severe cutaneous adverse reactions: an international study and meta-
analysis. Pharmacogenomics J 14(3):281–288. doi:10.1038/tpj.2013.40
8. Hung SI, Chung WH, Liu ZS, Chen CH, Hsih MS, Hui RC, Chu
CY, Chen YT (2010) Common risk allele in aromatic antiepileptic-
drug induced Stevens-Johnson syndrome and toxic epidermal

necrolysis in Han Chinese. Pharmacogenomics 11(3):349–356. doi:
10.2217/pgs.09.162
9. Cheung YK, Cheng SH, Chan EJ, Lo SV, Ng MH, Kwan P (2013)
HLA-B alleles associated with severe cutaneous reactions to anti-
epileptic drugs in Han Chinese. Epilepsia 54(7):1307–1314. doi:10.
1111/epi.12217
10. Thorn CF, Whirl-Carrillo M, Leeder JS, Klein TE, Altman RB
(2012) PharmGKB summary: phenytoin pathway. Pharmacogenet
Genomics 22(6):466–470
11. Suvichapanich S, Jittikoon J, Wichukchinda N, Kamchaisatian W,
Visudtibhan A, Benjapopitak S, Nakornchai S, Manuyakorn W,
Mahasirimongkol S (2015) Association analysis of CYP2C9*3
and phenytoin-induced severe cutaneous adverse reactions
(SCARs) in Thai epilepsy children. J Hum Genet 60(8):413–417.
doi:10.1038/jhg.2015.47
12. Lee AY, Kim MJ, Chey WY, Choi J, Kim BG (2004) Genetic
polymorphism of cytochrome P450 2C9 in diphenylhydantoin-
induced cutaneous adverse drug reactions. Eur J Clin Pharmacol
60(3):155–159. doi:10.1007/s00228-004-0753-0
13. Hirsch LJ, Weintraub DB, Buchsbaum R, Spencer HT, Straka T,
Hager M, Resor SR Jr (2006) Predictors of lamotrigine-associated
rash. Epilepsia 47(2):318–322. doi:10.1111/j.1528-1167.2006.
00423.x
14. Ahmed I, Reichenberg J, Lucas A, Shehan JM (2004) Erythema
multiforme associated with phenytoin and cranial radiation therapy:
a report of three patients and review of the literature. Int J Dermatol
43(1):67–73. doi:10.1111/j.1365-4632.2004.01934.x
15. Arif H, Buchsbaum R, Weintraub D, Koyfman S, Salas-Humara C,
Bazil CW, Resor SR Jr, Hirsch LJ (2007) Comparison and predic-
tors of rash associated with 15 antiepileptic drugs. Neurology
68(20):1701–1709. doi:10.1212/01.wnl.0000261917.83337.db
16. Naranjo CA, Busto U, Sellers EM, Sandor P, Ruiz I, Roberts EA,
Janecek E, Domecq C, Greenblatt DJ (1981) A method for estimat-
ing the probability of adverse drug reactions. Clin Pharmacol Ther
30(2):239–245
17. Sassolas B, Haddad C, Mockenhaupt M, Dunant A, Liss Y, Bork K,
Haustein UF, Vieluf D, Roujeau JC, Le Louet H (2010) ALDEN, an
algorithm for assessment of drug causality in Stevens-Johnson
Syndrome and toxic epidermal necrolysis: comparison with case-
control analysis. Clin Pharmacol Ther 88(1):60–68. doi:10.1038/
clpt.2009.252
18. Roujeau JC (1994) The spectrum of Stevens-Johnson syndrome
and toxic epidermal necrolysis: a clinical classification. J Invest
Dermatol 102(6):28S–30S
19. Kardaun SH, Sidoroff A, Valeyrie-Allanore L, Halevy S,
Davidovici BB, Mockenhaupt M, Roujeau JC (2007) Variability
in the clinical pattern of cutaneous side-effects of drugs with sys-
temic symptoms: does a DRESS syndrome really exist? Br J
Dermatol 156(3):609–611. doi:10.1111/j.1365-2133.2006.07704.x
20. Puangpetch A, Koomdee N, Chamnanphol M, Jantararoungtong T,
Santon S, Prommas S, Hongkaew Y, Sukasem C (2015) HLA-B
allele and haplotype diversity among Thai patients identified by
PCR-SSOP: evidence for high risk of drug-induced hypersensitiv-
ity. Front Genet 5(478):1–7. doi:10.3389/fgene.2014.00478
21. Itoh Y, Mizuki N, Shimada T, Azuma F, Itakura M, Kashiwase K,
Kikkawa E, Kulski JK, Satake M, Inoko H (2005) High-throughput
DNA typing of HLA-A, -B, -C, and -DRB1 loci by a PCR-SSOP-
Luminex method in the Japanese population. Immunogenetics
57(10):717–729. doi:10.1007/s00251-005-0048-3
22. Lee KW, Steiner N, Hurley CK (1998) Sequence of an HLA-B56
variant (B*5604) identified in the Korean population. Tissue
Antigens 51(2):210–212. doi:10.1111/j.1399-0039.1998.tb02967.x
23. Lancaster AK, Single RM, Solberg OD, Nelson MP, Thomson G
(2007) PyPop update—a software pipeline for large-scale
multilocus population genomics. Tissue Antigens 69(Suppl 1):
192–197. doi:10.1111/j.1399-0039.2006.00769.x
Eur J Clin Pharmacol
24. Valenzuela C (1993) 2 solutions for estimating odds ratios with
zeros. Rev Med Chil 121(12):1441–1444
25. Tatro DS (2013) Drug interaction facts 2014 : the authority on drug
interactions, 2014th edn. Lippincott Williams and Wilkins,
Philadelphia
26. Allison P (2004) Convergence problems in logistic regression. In:
Altman M, Gill J, McDonald M (eds) Numerical issues in statistical
computing for the social scientist, vol 10. Wiley-Interscience, New
Jersey, pp 238–252
27. Clarke B, Fokoué E, Zhang HH (2009) Multiple testing. In: Clarke
B, Fokoué E, Zhang HH (eds) Principles and theory for data mining
and machine learning. Springer, New York, pp 679–742. doi:10.
1007/978-0-387-98135-2_11
28. Manuyakorn W, Mahasirimongkol S, Likkasittipan P,
Kamchaisatian W, Wattanapokayakit S, Inunchot W, Visudtibhan
A, Wichukchinda N, Benjaponpitak S (2016) Association of HLA
genotypes with phenobarbital hypersensitivity in children.
Epilepsia 57(10):1610–1616. doi:10.1111/epi.13509
29. Zhang FR, Liu H, Irwanto A, Fu XA, Li Y, Yu GQ, Yu YX, Chen
MF, Low HQ, Li JH, Bao FF, Foo JN, Bei JX, Jia XM, Liu J, Liany
H, Wang N, Niu GY, Wang ZZ, Shi BQ, Tian HQ, Liu HX, Ma SS,
Zhou Y, You JB et al (2013) HLA-B*13:01 and the dapsone hyper-
sensitivity syndrome. N Engl J Med 369(17):1620–1628. doi:10.
1056/NEJMoa1213096
30. Yang F, Gu B, Zhang L, Xuan J, Luo H, Zhou P, Zhu Q, Yan S,
Chen SA, Cao Z, Xu J, Xing Q, Luo X (2014) HLA-B*13:01 is
associated with salazosulfapyridine-induced drug rash with eosino-
philia and systemic symptoms in Chinese Han population.
Pharmacogenomics 15(11):1461–1469. doi:10.2217/pgs.14.69
31. Tassaneeyakul W, Prabmeechai N, Sukasem C, Kongpan T,
Konyoung P, Chumworathayi P, Tiamkao S, Khunarkornsiri U,
Kulkantrakorn K, Saksit N, Nakkam N, Satapornpong P,
Vannaprasaht S, Sangviroon A, Mahasirimongkol S,
Wichukchinda N, Rerkpattanapipat T, Tassaneeyakul W (2016)
Associations between HLA class I and cytochrome P450 2C9 ge-
netic polymorphisms and phenytoin-related severe cutaneous ad-
verse reactions in a Thai population. Pharmacogenet Genomics
26(5):225–234. doi:10.1097/fpc.0000000000000211
32. Mahasirimongkol S, Somboonyosdech C, Kumperasart S,
Wattanapokayakit S, Satproedprai N, Inunchot W, Aekplakorn W,
Wichukchinday N (2014) HLA-B allelic distribution in samples
from Thailand National Health Examination Survey. J Health Sci
23(2):191–200
33. Insaro M, Chinvarun Y, Yodnopklao P, Nilayon D, Tantisira MH,
Kijsanayotin P (2011) Association of CYP2C9, CYP2C19 genetic
variants and non-genetic variants with phenytoin blood concentra-
tion in Thai epileptic patients. Thai J Pharmacol 33(1):58–64
34. Sukasem C, Tunthong R, Chamnanphon M, Santon S,
Jantararoungtong T, Koomdee N, Prommas S, Puangpetch A,
Vathesatogkit P (2013) CYP2C19 polymorphisms in the Thai pop-
ulation and the clinical response to clopidogrel in patients with
atherothrombotic-risk factors. Pharmgenomics Pers Med 6:85–
91. doi:10.2147/pgpm.s42332
35. Franco V, Perucca E (2015) CYP2C9 polymorphisms and phenyt-
oin metabolism: implications for adverse effects. Expert Opin Drug
Metab Toxicol 11(8):1269–1279. doi:10.1517/17425255.2015.
1053463
36. Wei CY, Chung WH, Huang HW, Chen YT, Hung SI (2012) Direct
interaction between HLA-B and carbamazepine activates T cells in
patients with Stevens-Johnson syndrome. J Allergy Clin Immunol
129(6):1562–1569. doi:10.1016/j.jaci.2011.12.990
37. Spielberg SP, Gordon GB, Blake DA, Mellits ED, Bross DS (1981)
Anticonvulsant toxicity in vitro: possible role of arene oxides. J
Pharmacol Exp Ther 217(2):386–389
38. Yun J, Mattsson J, Schnyder K, Fontana S, Largiader CR, Pichler
WJ, Yerly D (2013) Allopurinol hypersensitivity is primarily
Eur J Clin Pharmacol
mediated by dose-dependent oxypurinol-specific T cell response.
Clin Exp Allergy 43(11):1246–1255. doi:10.1111/cea.12184
39. Yun J, Marcaida MJ, Eriksson KK, Jamin H, Fontana S, Pichler WJ,
Yerly D (2014) Oxypurinol directly and immediately activates the
drug-specific T cells via the preferential use of HLA-B*58:01. J
Immunol 192(7):2984–2993. doi:10.4049/jimmunol.1302306
40. Mosher CM, Tai G, Rettie AE (2009) CYP2C9 amino acid residues
influencing phenytoin turnover and metabolite regio- and stereo-
chemistry. J Pharmacol Exp Ther 329(3):938–944. doi:10.1124/
jpet.109.150706
41. Yasumori T, Chen LS, Li QH, Ueda M, Tsuzuki T, Goldstein JA,
Kato R, Yamazoe Y (1999) Human CYP2C-mediated
stereoselective phenytoin hydroxylation in Japanese: difference in
chiral preference of CYP2C9 and CYP2C19. Biochem Pharmacol
57(11):1297–1303. doi:10.1016/S0006-2952(99)00034-9
42. Li X-Q, Andersson TB, Ahlström M, Weidolf L (2004)
Comparison of inhibitory effects of the proton pump-inhibiting
drugs omeprazole, esomeprazole, lansoprazole, pantoprazole, and
rabeprazole on human cytochrome P450 activities. Drug Metab
Dispos 32(8):821–827. doi:10.1124/dmd.32.8.821