A whole-genome association study of major determinants for allopurinol-related Stevens–Johnson syndrome and toxic epidermal necrolysis in Japanese patients


Stevens–Johnson syndrome and toxic epidermal necrolysis (SJS/TEN) are severe, cutaneous adverse drug reactions that are rare but life threatening. Genetic biomarkers for allopurinol-related SJS/TEN in Japanese were examined in a genome-wide association study in which Japanese patients (n=14) were compared with ethnically matched healthy controls (n=991). Associations between 890 321 single nucleotide polymorphisms and allopurinol-related SJS/TEN were analyzed by the Fisher's exact test (dominant genotype mode). A total of 21 polymorphisms on chromosome 6 were significantly associated with allopurinol-related SJS/TEN. The strongest association was found at rs2734583 in BAT1, rs3094011 in HCP5 and GA005234 in MICC (P=2.44 × 10−8; odds ratio=66.8; 95% confidence interval, 19.8–225.0). rs9263726 in PSORS1C1, also significantly associated with allopurinol-related SJS/TEN, is in absolute linkage disequilibrium with human leukocyte antigen-B*5801, which is in strong association with allopurinol-induced SJS/TEN. The ease of typing rs9263726 makes it a useful biomarker for allopurinol-related SJS/TEN in Japanese.


Allopurinol is a xanthine oxidase inhibitor that prevents the production of uric acid to reduce plasma uric acid levels to a normal range. It is the most frequently used anti-hyperuricemic agent in the world due to its long-term pharmacological effect.1 However, allopurinol is also one of the most frequent causes of a variety of delayed severe cutaneous adverse drug reactions (SCARs).2 According to spontaneous reports of severe adverse drug reactions to the Ministry of Health, Labor, and Welfare of Japan, allopurinol-related SCARs accounted for about 11% of all reported SCAR cases in Japan in 2008.3 Allopurinol-related SCARs include the drug-induced hypersensitivity syndrome, Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN).4 SJS/TEN are characterized by high fever, malaise and rapid development of blistering exanthema, with macules and target-like lesions, accompanied by mucosal involvement.5 Even though the incidence of SJS/TEN is extremely low, the mortality rate of TEN can be as high as 26%.5 Therefore, SJS/TEN is a serious problem in allopurinol therapy, in spite of the ideal anti-hyperuricemic effect of allopurinol.

Although previous works have suggested that the development of SJS/TEN depends on an immune mechanism involving a drug-dependent cytotoxic cell response against epidermal cells,5, 6 the pathophysiology of SJS/TEN remains largely unknown. Susceptibility to such idiosyncratic reactions is thought to be genetically determined, and familial predisposition to allopurinol-induced SJS/TEN has been reported.6 Therefore, the exploratory studies for genetic risk factors related to SJS/TEN are needed. A strong association has been observed between allopurinol-induced SCAR and the human lymphocyte antigen (HLA) allele B variant (HLA-B*5801) in the Han Chinese in Taiwan7 and in the Thai population.8 These studies showed that the HLA-B*5801 allele is present in all patients with allopurinol-induced SCAR (51/51 of Han Chinese and 27/27 of Thai patients) and in only 12–15% of tolerant patients (20/135 and 7/54, respectively). The odds ratio (OR) was 580 (95% confidence interval , 34–9781; P=4.7 × 1024) for the Han-Chinese data7 and 348.3 (95% confidence interval, 19.2–6336.9; P=1.61 × 1013) for the Thai study.8 Although the association was confirmed in both Caucasian and Japanese subjects,9, 10 the OR in the Han-Chinese and Thai populations were much higher than those in the Caucasian (OR=80) and Japanese (OR=40) groups. These reports indicated that HLA-B*5801 is the valid genetic biomarker for allopurinol-induced SJS/TEN in various ethnic groups, but the mechanisms by which HLA-B*5801 is specifically involved in allopurinol-induced SJS/TEN progression and the strength of the association showed ethnic differences are unknown.

Currently, genotyping by high-density array scanning of the whole genome allows discovery of previously unsuspected genetic risk factors that influence the pathogenesis of serious adverse drug reactions.11, 12, 13 Genome-wide association studies (GWASs) provide opportunities to uncover polymorphisms that influence susceptibility to allopurinol-induced SJS/TEN free of mechanistic hypotheses. Therefore, in addition to HLA-B typing as shown in our previous study,10 we further conducted a retrospective pharmacogenetic case–control study using whole-genome single nucleotide polymorphism (SNP) data from high-density DNA microarrays in order to identify new and effective genetic biomarkers for allopurinol-related SJS/TEN in Japanese patients.

Materials and methods

Recruitment of study subjects

A total of 141 Japanese SJS/TEN patients from unrelated families were recruited from July 2006 to April 2010 from participating institutes of the Japan Severe Adverse Reactions (JSAR) research group and through a nationwide blood-sampling network system in Japan for SJS/TEN onset patients, operated by the National Institute of Health Sciences.10 In all, 121 of these patients were diagnosed as defined SJS or TEN by JSAR research group's dermatological experts based on diagnostic criteria4 that are currently used in Japan. Information was collected using a standardized case report form that includes medical records, co-administrated drug records, disease progress and involvement of systemic complications, as well as SJS/TEN treatment. Among the 141 SJS/TEN patients, 20 were diagnosed as probable SJS due to atypical or mild symptoms. TEN and SJS were defined as mucocutaneous disorders characterized by extensive erythema, blisters, epidermal detachment, erosions, enanthema and high fever. SJS was defined as skin detachment of 10% or less of the body surface area, and TEN as skin detachment of more than 10%, excluding staphylococcal scaled skin syndrome.5 In all enrolled cases defined as SJS or TEN, allopurinol was regarded as the drug responsible for SJS or TEN if the onset of SJS/TEN symptoms occurred within the first 2 months of allopurinol exposure. For the retrospective pharmacogenetic case–control study, 991 healthy, ethnically matched subjects in the Tokyo metropolitan area were used as the control group. Healthy subjects were used as the control group instead of allopurinol-tolerant patients because the incidence of SJS/TEN is extremely low (0.4–6 per million per year).3

The ethics committees of the National Institute of Health Sciences, each participating institute of the JSAR research group and the Japan Pharmacogenomics Data Science Consortium (JPDSC) approved this study. Written informed consent was obtained from all cases and ethnically matched controls.

Whole-genome genotyping of SNPs

Genome-wide genotyping of the 14 allopurinol-related SJS/TEN patients and 991 ethnically matched controls was conducted using the Illumina Human 1M-Duo BeadChip (Illumina, San Diego, CA, USA), which contained 11 63 218 SNPs. SNPs were discarded from case–control association analysis if they exhibited a minor allele frequency <0.001 in the control group (2 37 890 SNPs), a call rate <0.95 for each SNP (32 640 SNPs) or a P-value <0.001 in the test of Hardy–Weinberg equilibrium among controls (2 368 SNPs). These quality control steps removed a total of 2 72 897 SNPs. All samples had a call rate for each microarray above 0.99. Sample duplicates and hidden relatedness were investigated on the basis of pairwise identity-by-state analysis via PLINK;14 however, there was no duplicate or hidden relatedness in the samples. This quality-control procedure ensured reliable genotyping data.

HLA genotyping and TaqMan genotyping of SNPs on chromosome 6

HLA A, B and Cw types were determined using sequencing-based methods, as described previously.10 Representative SNPs of 6p21 (rs2734583, rs3099844, rs9263726 and rs3131643) were re-genotyped using TaqMan SNP Genotyping Assays (Life Technologies, Carlsbad, CA, USA) (ID; C__27465749_10, C__27455402_10, C__30352071_10, C__26778946_20) according to the manufacturer’s instruction using 5 ng of genomic DNA. We did not genotype rs9267445 and rs1634776 because TaqMan SNP genotyping assays for these SNPs were not available. Measurement of the linkage disequilibrium (LD) coefficient was performed using the HLA types and 6p21 SNPs of the 141 Japanese SJS/TEN cases and an additional 65 Japanese individuals (non-SJS/TEN patients). The LD coefficient was calculated as previously described.15, 16

Association analysis

Genome-wide SNPs data from allopurinol-related SJS/TEN cases and ethnically matched controls were used for association analysis using the Fisher’s exact test based on the dominant genotype mode and minor allele frequencies of each SNP. Because there are no homozygotes of minor alleles of SNPs, which have significantly related to allopurinol-related SJS/TEN except rs3099844 and rs3131643 in ‘Case group’, other association analysis models such as trend test (Cochran–Armitage analysis) or recessive model analysis were not applied in this study. All association analyses were carried out with PLINK.14 P-values were corrected for multiple testing according to the Bonferroni's correction. P-values <5.62 × 10−8 were regarded as statistically significant.


Characteristics of study subjects

A total of 14 allopurinol-treated Japanese patients, who were diagnosed with definite SJS/TEN were recruited for the whole-genome association study (IDs 1–14 in Table 1). Patients, IDs 1, 2, 3, 9, 10, 13 and 14 were reported in our previous paper.10 After the GWAS, an additional four allopurinol-treated Japanese SJS/TEN patients were recruited for HLA typing (IDs 15–18). Therefore, a total of 18 allopurinol-treated Japanese SJS/TEN patients participated in the study (Table 1). In all, 12 of 18 patients were male and 6 were female, and the average age was 72.3±10.0 (mean±s.d.) years. In all, 12 of 18 cases showed systemic complications of liver and/or renal dysfunction, and most patients had high fever. The average period of SJS/TEN onset after allopurinol treatment was 21.7±11.9 days. Drug-induced lymphocyte stimulation tests were examined in 13 of 18 patients to determine the causative agent; however, in these tests, only two cases (IDs 1 and 5) were positive for allopurinol and only one (ID 16) was positive for oxipurinol, a metabolite of allopurinol. The patient (ID 1) who was positive for the drug-induced lymphocyte stimulation test for allopurinol was also positive for other co-administrated drugs (Table 1). On the other hand, patients who received a patch test showed positive reactions for allopurinol although only two patients were examined (ID 4, 10). The patient who was patch test positive for allopurinol (ID 4) was also patch test positive for other co-administrated drugs (Table 1). Four patients (ID 1, 2, 4 and 14) were co-administrated non-steroidal anti-inflammatory drugs, four (ID 7, 8, 11 and 15) were co-administrated angiotensin II receptor antagonists and three (ID 4, 7 and 17) were co-administrated statin anti-hyperlipemic agents.

Table 1 Summary of clinical characteristics of Japanese patients with allopurinol-related Stevens–Johnson syndrome or toxic epidermal necrolysis

Whole-genome association study of major determinants for allopurinol-related SJS/TEN

A total of 14 allopurinol-related SJS/TEN patients (IDs 1–14), who were diagnosed with definite SJS/TEN, and 991 ethnically matched controls, were genotyped with the use of the Illumina Human 1M-Duo BeadChip containing 11 63 218 SNPs. A series of quality-control steps resulted in the elimination of 2 72 897 polymorphisms. For each SNP, Fisher’s exact tests were performed to compare the dominant genotype distributions and minor allelic frequencies in the allopurinol-related SJS/TEN patients (the case group) versus those in the ethnically matched healthy control group. The resulting P-values were adjusted with the Bonferroni's correction (P<5.62 × 10−8). The distribution of P-values from the Fisher’s exact tests (dominant genotype mode) along each chromosome indicated that 21 SNPs were significantly associated with the cases, all of which were located on the chromosome 6: 6p21.3, 6p22.1 and 6p21.1 (Figures 1a and b). The quantile–quantile (Q–Q) plot for the distribution of P-values showed that observed P-values matched the expected P-values over the range of 0<−log10(p) < 4.0 (Figure 2). A departure was observed at the extreme tail (−log10(p)>4.0) of the distribution of test statistics for the allopurinol-related Japanese SJS/TEN, suggesting that the identified associations are likely due to true variants rather than potential biases such as genotyping error. These SNPs, with their associated genes, are described in Table 2. As is observed in all SNPs in Table 2, minor allele frequencies in the controls were quite small, ranging around 0.5–0.6%. The genotypic distributions of the case and control groups are identical among groups with the same P-value, suggesting that these SNPs might be linked. These SNPs also have ORs that are much higher than the ORs of SNPs commonly observed in sporadic cancer and other complex diseases, suggesting they are of higher penetrance. For example, the most significant SNPs (rs2734583, rs3094011 and GA005234) had an OR of 66.8 (95% confidence interval, 19.8–225.0), and the twentieth most significant SNPs (rs9263827 and rs1634776) had an OR of 60.9 (95% confidence interval, 18.3–202.5). Most SNPs in Table 2 are associated with known or predicted genes; of these, 13 are in known genes. Three SNPs (rs17190526, rs9263726 and rs2233945) were found in PSORS1C1 (psoriasis susceptibility 1 candidate 1), which is considered as one of the potential psoriasis genes.17, 18, 19 The CCHCR1 (coiled coil α helical rod protein 1), which is a regulator of keratinocyte proliferation or differentiation and is overexpressed in keratinocytes in psoriatic lesions,20, 21, 22, 23 contained four SNPs (rs9263745, rs130077, rs9263781 and rs9263785). HCP5 (HLA complex P5), which is involved in hypersensitivity to abacavir,24, 25, 26 had three SNPs (rs3094011, rs3099844 and rs31431643). TCF19 (transcription factor 19), which is a potential trans-activating factor that might play an important role in the transcription of genes required for the later stages of cell cycle progression,27 contained two SNPs (rs9263794 and rs10448701). Two SNPs (rs9263796 and rs9263800) were also found in POU5F1 (POU class 5 homeobox; alternative names for Oct4). BAT1 (HLA-B associated transcript 1) and PSORS1C3 each carried one SNP (rs2734583 and rs9263827). The SNPs, rs1634776 and rs4084090, were located in more than 10 kb away from the HLA-B and HLA-C genes, respectively. Two pseudo genes, MICC (major histocompatibility complex class I polypeptide-related sequence) and PPIAP9 (peptidylprolyl isomerase A (cyclophilin A) pseudogene 9), had one SNP each (GA005234 and rs9267445). Previous report using Han-Chinese patients with allopurinol-induced SCAR indicated rs3117583 of BAT3, rs1150793 of MSH5 and rs2855804 of MICB, which are located in HLA region, showed significant P-values (P<1 × 10−7).7 In this study using Japanese patients, both rs3117583 and rs1150793 showed P=6.34 × 10−3 (allele frequency mode) and P=6.14 × 10−3 (dominant genotype mode). There was no data of rs2855804 in the Illumina Human 1M-Duo BeadChip.

Figure 1

Genome-wide association study of allopurinol-related Stevens–Johnson syndrome or toxic epidermal necrolysis. Each dot represents a single nucleotide polymorphism (SNP). The x axis: the position of the SNP on chromosomes. The y axis: the −log10 of Fisher's exact test P-values (dominant genotype mode) of the SNP in the case–control association study. SNPs with P-values <5.62 × 10−8 are highlighted in red. (a) Whole genome. (b) Chromosome 6.

PowerPoint slide

Figure 2

Quantile–quantile plot of Fisher's exact test statistics obtained from the genome-wide association study for allopurinol-related Stevens–Johnson syndrome or toxic epidermal necrolysis under dominant genotype mode. The solid red line represents the null model where observed Fisher's exact test values match the expected values. The dots represent observed versus the expected values from the case–control study.

PowerPoint slide

Table 2 The association of single nucleotide polymorphism with allopurinol-related Japanese patients with Stevens–Johnson syndrome or toxic epidermal necrolysis

HLA types of allopurinol-related SJS/TEN patients

Classical class I HLA types (A, B and Cw) of allopurinol-related SJS/TEN patients were determined because the HLA-B*5801 type is associated with allopurinol-related SCARs in Han Chinese,7 Caucasians9 and Japanese10 (Table 3). In this analysis, four patients with allopurinol-related SJS/TEN (IDs 15–18), who were recruited after BeadChip analysis, joined the case group (total of 18 allopurinol-related SJS/TEN patients). Eight cases of HLA-A*3303 (allele frequency=22.2%), 10 cases of HLA-B*5801 (allele frequency=27.8%) and 10 cases of HLA-Cw*0302 (allele frequency=27.8%) were found in 18 allopurinol-related SJS/TEN patients (Table 3). By comparison, the allelic frequencies of HLA-A*3303, HLA-B*5801 and HLA-Cw*0302 were 7.9%, 0.6% and 0%, respectively in Japanese general population (Tables 4a, 4b and 4c). The OR of HLA-A*3303 was calculated as 3.32 (Table 4a). The OR of HLA-B*5801 was calculated as 62.8 (Table 4b), which was a little larger than the previously reported OR in Japanese patients.10 HLA-Cw*0302 also showed significant association with allopurinol-related SJS/TEN (Table 4c). HLA-A*3303 and HLA-Cw*0302 are in LD with HLA-B*5801 in the Japanese although the general frequency of HLA-A*3303 is higher than other two types. Other HLA-A, B and Cw types, which were not listed in Tables 4a, 4b and 4c, showed very low frequencies in the general Japanese population, or were not found in 18 allopurinol-related SJS/TEN patients.

Table 3 HLA types and representative genotypes in 6p21 of allopurinol-related Japanese patients with Stevens–Johnson syndrome or toxic epidermal necrolysis
Table 4a Association between HLA-A alleles and allopurinol-induced Stevens–Johnson syndrome or toxic epidermal necrolysis
Table 4b Association between HLA-B alleles and allopurinol-induced Stevens–Johnson syndrome or toxic epidermal necrolysis
Table 4c Association between HLA-Cw alleles and allopurinol-induced Stevens–Johnson syndrome or toxic epidermal necrolysis

LD of HLA-B*5801 with SNPs on chromosome 6

We compared the genotypic distributions of six SNPs, which were significantly associated with SJS/TEN (Table 2), with HLA types because these SNPs are located near the HLA-B gene. These 6 SNPs listed in Table 3 represent 21 SNPs in Table 2 because the other 15 SNPs are in absolute LD with 1 of the 6 SNPs. Representative six variants of the significant SNPs on chromosome 6 were found in all of the SJS/TEN patients who carried the HLA-B*5801 (10 patients) (Table 3). Therefore, in order to evaluate LD in the Japanese population, LD coefficients (D′) were calculated between classical class 1 HLA types and six representative SNPs at 6p21, using the HLA-type and SNPs genotype data of 206 Japanese individuals, including 141 SJS/TEN cases and an additional 65 non-SJS/TEN Japanese subjects. As shown in Tables 5 and 6 representative SNPs on chromosome 6 showed LD for the HLAs. In particular, three SNPs (rs2734583, rs9267445 and rs9263726) showed a strong linkage with HLA-B and Cw alleles (Table 5). LD between six representative SNPs in 6p21 and HLA-B*5801 are shown in Table 6. A novel observation was the absolute LD (D′=1, r2=1) between rs9263726 in PSORS1C1 and the HLA-B*5801 allele.

Table 5 The linkage disequilibrium between HLA types and representative single nucleotide polymorphisms on 6p21 of 206 Japanese individuals
Table 6 The linkage disequilibrium between representative single nucleotide polymorphisms on 6p21 and HLA-B*5801 of 206 Japanese individuals


In order to explore new genetic biomarkers associated with the occurrence of allopurinol-related SJS/TEN Japanese patients, we conducted a GWAS using 890321 SNPs from patients with allopurinol-related SJS/TEN and an ethnically matched control group. The GWAS data indicated that most SNPs significantly associated with allopurinol-related SJS/TEN are located on or close to genes that overlap the 6p21 region, especially the genes neighboring HLA-B. There was no significantly associated SNP in any other region of the genome (Figures 1 and 2 and Table 2), indicating that the 6p21 region has the most important role in the progress of allopurinol-related SJS/TEN. We expected to find SJS/TEN-associated SNPs, which are unrelated to HLA-B*5801 from this GWAS study because the association of HLA-B*5801 with SJS/TEN is incomplete (10/18) in Japanese patients in contrast to Han Chinese7 and Thai patients.8 However, most of significant SNPs were closely linked with HLA-B*5801 (Table 6). Previous studies have indicated that a SNP (rs2395029) in the HCP5, which is on 6p21.3, is strongly associated with human immunodeficiency virus-1 set points,28, 29, 30 abacavir-induced hypersensitivity24, 25, 26 and flucloxacillin-induced liver injury.31 This SNP is in strong LD with HLA-B*5701 in Caucasians.25 Another SNP in 6p21 in PSORS1C1, a psoriasis-susceptibility candidate gene, was related with psoriasis in Swedish and Canadian populations17, 18 and exhibits LD with HLA-Cw*0602 in Canadian populations.18 These reports suggest that SNPs located in 6p21 link with a specific type of classical class I HLA that could be an alternative biomarker for the physiological phenomenon. Therefore, we examined the LD between these SNPs, shown in Table 2, and HLA-B*5801, which has been regarded as a genetic biomarker of SJS/TEN not only in Han Chinese,7 but also in Caucasians9 and Japanese.10 We found that all of the Japanese patients with the allopurinol-related SJS/TEN who had the HLA-B*5801 (10 patients) also had variant SNPs of genes that are located in 6p21, including BAT1, HCP5, PPIAP9, PSORS1C1 and HLA-B (Table 3). The analysis of the LD coefficients between SNPs located in 6p21 and HLA types in the Japanese population indicated that these SNPs are in strong LD with HLA types (Table 5), and an absolute LD between rs9263726 in PSORS1C1 and HLA-B*5801 was observed in the Japanese population (Table 6). These results mean that all subjects (14 individuals including 10 with allopurinol-related SJS/TEN) who carry HLA-B*5801 are in complete accord with all subjects with minor A allele of rs9263726 in the Japanese population. Therefore, rs9263726 in PSORS1C1 is an alternative biomarker for HLA-B*5801 in the Japanese population. Conventional genotyping of rs9263726 based on allelic discrimination offers several advantages over HLA-B typing, which is determined by genotyping of several SNPs forming the HLA-B*5801 haplotype. Various broadly used technologies (for example, TaqMan genotyping) allow the standardized identification of two distinct alleles in one reaction tube, limiting the risk of contamination and allowing high-throughput genotyping with high sensitivity and specificity. In addition, the test is largely independent of both the performance of and interpretation by laboratory personnel. SNP genotyping is also less time consuming and cheaper than sequence-based HLA typing, and it does not require specialized laboratories. Therefore, the easy detection of these SNPs has a practical and economical advantage in clinical application for predicting the onset of allopurinol-related SJS/TEN. Although the previous report revealed that three SNPs in HLA region strongly associated with allopurinol-related SCAR in Han Chinese,7 the two SNPs analyzed by the Illumina Human 1M-DUO BeadChip showed only weak association in the Japanese. This ethnic difference might be due to the difference of LD.

The functional analysis of genes that carry these SNPs—including HCP5, BAT1, PSORS1C1, CCHCR1, TCF19 and POU5F1—in the pathogenesis of allopurinol-related SJS/TEN might be useful for determining their relevance. CCHCR1 is a regulator of keratinocyte proliferation or differentiation and is overexpressed in keratinocytes in psoriatic lesions.20, 21, 22, 23 TCF19 is a potential trans-activating factor that could play an important role in the transcription of genes required for the later stages of cell cycle progression.27 Possible psoriasis candidate genes near HLA-B include PSORS1C1,17, 18, 19 CCHCR1,22, 23 and POU5F1.32, 33 Mutations in BAT1 may be associated with rheumatoid arthritis.34, 35, 36 HCP5 encodes an endogenous retroviral element mainly that is expressed in immune cells and there is evidence that the SNP in this gene is protective against human immunodeficiency virus-1 infection.37, 38, 39 The functions and relevance of these genes suggest that the pathogenesis of allopurinol-related SJS/TEN might involve not only an immune system disorder, but also processes of cell proliferation and differentiation.

In conclusion, the results of this GWAS of allopurinol-related SJS/TEN in Japanese patients show that SNPs in genes located in 6p21, which are in LD with HLA-B*5801, are strongly associated with the cutaneous adverse reaction. Therefore, these SNPs, especially rs9263726, prove to be predictors for allopurinol-related SJS/TEN in Japanese, and their genes might be involved in the pathogenesis of allopurinol-related SJS/TEN. The OR of rs9263726 is extremely high from this case–control study and the typing cost of SNP is much cheaper than that of HLA typing. Moreover, the SJS/TEN has a very severe adverse reaction of allopurinol, which is high mortality. Therefore, we believe that the screening of rs9263726 genotype before allopurinol administration is necessary to prevent SJS/TEN in allopurinol-treated Japanese patients, although its allele frequency is very low in the Japanese. Association analyses of other ethnic populations are needed for confirming and comparing the results obtained in this study. In vitro functional studies of these genes are also necessary for identification of the physiological and molecular pathways leading to allopurinol-related SJS/TEN.


  1. 1

    Wortmann RL . Gout and hyperuricemia. Curr Opin Rheumatol 2002; 14: 281–286.

    Article  PubMed  Google Scholar 

  2. 2

    Chung WH, Hung SI, Chen YT . Human leukocyte antigens and drug hypersensitivity. Curr Opin Allergy Clin Immunol 2007; 7: 317–323.

    CAS  Article  PubMed  Google Scholar 

  3. 3

    Tohkin M, Ishiguro A, Kaniwa N, Saito Y, Kurose K, Hasegawa R . Prediction of severe adverse drug reactions using pharmacogenetic biomarkers. Drug Metab Pharmacokinet 2010; 25: 122–133.

    CAS  Article  PubMed  Google Scholar 

  4. 4

    Bastuji-Garin S, Rzany B, Stern RS, Shear NH, Naldi L, Roujeau JC . Clinical classification of cases of toxic epidermal necrolysis, Stevens-Johnson syndrome, and erythema multiforme. Arch Dermatol 1993; 129: 92–96.

    CAS  Article  PubMed  Google Scholar 

  5. 5

    French LE . Toxic epidermal necrolysis and Stevens Johnson syndrome: our current understanding. Allergol Int 2006; 55: 9–16.

    Article  PubMed  Google Scholar 

  6. 6

    Bowman C, Delrieu O . Immunogenetics of drug-induced skin blistering disorders. Part I: perspective. Pharmacogenomics 2009; 10: 601–621.

    CAS  Article  PubMed  Google Scholar 

  7. 7

    Hung SI, Chung WH, Liou LB, Chu CC, Lin M, Huang HP et al. HLA-B*5801 allele as a genetic marker for severe cutaneous adverse reactions caused by allopurinol. Proc Natl Acad Sci USA 2005; 102: 4134–4139.

    CAS  Article  PubMed  Google Scholar 

  8. 8

    Tassaneeyakul W, Jantararoungtong T, Chen P, Lin PY, Tiamkao S, Khunarkornsiri U et al. Strong association between HLA-B*5801 and allopurinol-induced Stevens-Johnson syndrome and toxic epidermal necrolysis in a Thai population. Pharmacogenet Genomics 2009; 19: 704–709.

    CAS  Article  PubMed  Google Scholar 

  9. 9

    Lonjou C, Borot N, Sekula P, Ledger N, Thomas L, Halevy S et al. A European study of HLA-B in Stevens-Johnson syndrome and toxic epidermal necrolysis related to five high-risk drugs. Pharmacogenet Genomics 2008; 18: 99–107.

    CAS  Article  PubMed  Google Scholar 

  10. 10

    Kaniwa N, Saito Y, Aihara M, Matsunaga K, Tohkin M, Kurose K et al. HLA-B locus in Japanese patients with anti-epileptics and allopurinol-related Stevens-Johnson syndrome and toxic epidermal necrolysis. Pharmacogenomics 2008; 9: 1617–1622.

    CAS  Article  PubMed  Google Scholar 

  11. 11

    Wilke RA, Lin DW, Roden DM, Watkins PB, Flockhart D, Zineh I et al. Identifying genetic risk factors for serious adverse drug reactions: current progress and challenges. Nat Rev Drug Discov 2007; 6: 904–916.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  12. 12

    Nakamura Y . Pharmacogenomics and drug toxicity. N Engl J Med 2008; 359: 856–858.

    CAS  Article  PubMed  Google Scholar 

  13. 13

    Daly AK, Day CP . Genetic association studies in drug-induced liver injury. Semin Liver Dis 2009; 29: 400–411.

    CAS  Article  PubMed  Google Scholar 

  14. 14

    Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007; 81: 559–575.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  15. 15

    Zhao JH, Curtis D, Sham PC . Model-free analysis and permutation tests for allelic associations. Hum Hered 2000; 50: 133–139.

    CAS  Article  PubMed  Google Scholar 

  16. 16

    Zhao JH . 2LD, GENECOUNTING and HAP: computer programs for linkage disequilibrium analysis. Bioinformatics 2004; 20: 1325–1326.

    CAS  Article  PubMed  Google Scholar 

  17. 17

    Holm SJ, Carlen LM, Mallbris L, Stahle-Backdahl M, O’Brien KP . Polymorphisms in the SEEK1 and SPR1 genes on 6p21.3 associate with psoriasis in the Swedish population. Exp Dermatol 2003; 12: 435–444.

    CAS  Article  PubMed  Google Scholar 

  18. 18

    Rahman P, Butt C, Siannis F, Farewell VT, Peddle L, Pellett FJ et al. Association of SEEK1 and psoriatic arthritis in two distinct Canadian populations. Ann Rheum Dis 2005; 64: 1370–1372.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  19. 19

    Zhang XJ, He PP, Wang ZX, Zhang J, Li YB, Wang HY et al. Evidence for a major psoriasis susceptibility locus at 6p21(PSORS1) and a novel candidate region at 4q31 by genome-wide scan in Chinese hans. J Invest Dermatol 2002; 119: 1361–1366.

    CAS  Article  PubMed  Google Scholar 

  20. 20

    Suomela S, Elomaa O, Skoog T, Ala-aho R, Jeskanen L, Parssinen J et al. CCHCR1 is up-regulated in skin cancer and associated with EGFR expression. PLoS One 2009; 4: e6030.

    Article  PubMed  PubMed Central  Google Scholar 

  21. 21

    Tiala I, Wakkinen J, Suomela S, Puolakkainen P, Tammi R, Forsberg S et al. The PSORS1 locus gene CCHCR1 affects keratinocyte proliferation in transgenic mice. Hum Mol Genet 2008; 17: 1043–1051.

    CAS  Article  PubMed  Google Scholar 

  22. 22

    Suomela S, Kainu K, Onkamo P, Tiala I, Himberg J, Koskinen L et al. Clinical associations of the risk alleles of HLA-Cw6 and CCHCR1*WWCC in psoriasis. Acta Derm Venereol 2007; 87: 127–134.

    Article  PubMed  Google Scholar 

  23. 23

    Tiala I, Suomela S, Huuhtanen J, Wakkinen J, Holtta-Vuori M, Kainu K et al. The CCHCR1 (HCR) gene is relevant for skin steroidogenesis and downregulated in cultured psoriatic keratinocytes. J Mol Med 2007; 85: 589–601.

    CAS  Article  PubMed  Google Scholar 

  24. 24

    Hughes AR, Mosteller M, Bansal AT, Davies K, Haneline SA, Lai EH et al. Association of genetic variations in HLA-B region with hypersensitivity to abacavir in some, but not all, populations. Pharmacogenomics 2004; 5: 203–211.

    CAS  Article  PubMed  Google Scholar 

  25. 25

    Colombo S, Rauch A, Rotger M, Fellay J, Martinez R, Fux C et al. The HCP5 single-nucleotide polymorphism: a simple screening tool for prediction of hypersensitivity reaction to abacavir. J Infect Dis 2008; 198: 864–867.

    CAS  Article  PubMed  Google Scholar 

  26. 26

    Mallal S, Phillips E, Carosi G, Molina JM, Workman C, Tomazic J et al. HLA-B*5701 screening for hypersensitivity to abacavir. N Engl J Med 2008; 358: 568–579.

    Article  PubMed  Google Scholar 

  27. 27

    Teraoka Y, Naruse TK, Oka A, Matsuzawa Y, Shiina T, Iizuka M et al. Genetic polymorphisms in the cell growth regulated gene, SC1 telomeric of the HLA-C gene and lack of association of psoriasis vulgaris. Tissue Antigens 2000; 55: 206–211.

    CAS  Article  PubMed  Google Scholar 

  28. 28

    Migueles SA, Sabbaghian MS, Shupert WL, Bettinotti MP, Marincola FM, Martino L et al. HLA B*5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors. Proc Natl Acad Sci USA 2000; 97: 2709–2714.

    CAS  Article  PubMed  Google Scholar 

  29. 29

    Altfeld M, Addo MM, Rosenberg ES, Hecht FM, Lee PK, Vogel M et al. Influence of HLA-B57 on clinical presentation and viral control during acute HIV-1 infection. AIDS 2003; 17: 2581–2591.

    CAS  Article  PubMed  Google Scholar 

  30. 30

    Fellay J, Shianna KV, Ge D, Colombo S, Ledergerber B, Weale M et al. A whole-genome association study of major determinants for host control of HIV-1. Science 2007; 317: 944–947.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  31. 31

    Daly AK, Donaldson PT, Bhatnagar P, Shen Y, Pe’er I, Floratos A et al. HLA-B*5701 genotype is a major determinant of drug-induced liver injury due to flucloxacillin. Nat Genet 2009; 41: 816–819.

    CAS  Article  PubMed  Google Scholar 

  32. 32

    Oka A, Tamiya G, Tomizawa M, Ota M, Katsuyama Y, Makino S et al. Association analysis using refined microsatellite markers localizes a susceptibility locus for psoriasis vulgaris within a 111 kb segment telomeric to the HLA-C gene. Hum Mol Genet 1999; 8: 2165–2170.

    CAS  Article  PubMed  Google Scholar 

  33. 33

    Chang YT, Hsu CY, Chou CT, Lin MW, Shiao YM, Tsai CY et al. The genetic polymorphisms of POU5F1 gene are associated with psoriasis vulgaris in Chinese. J Dermatol Sci 2007; 46: 153–156.

    CAS  Article  PubMed  Google Scholar 

  34. 34

    Okamoto K, Makino S, Yoshikawa Y, Takaki A, Nagatsuka Y, Ota M et al. Identification of I kappa BL as the second major histocompatibility complex-linked susceptibility locus for rheumatoid arthritis. Am J Hum Genet 2003; 72: 303–312.

    CAS  Article  PubMed  Google Scholar 

  35. 35

    Kilding R, Iles MM, Timms JM, Worthington J, Wilson AG . Additional genetic susceptibility for rheumatoid arthritis telomeric of the DRB1 locus. Arthritis Rheum 2004; 50: 763–769.

    CAS  Article  PubMed  Google Scholar 

  36. 36

    Quinones-Lombrana A, Lopez-Soto A, Ballina-Garcia FJ, Alperi-Lopez M, Queiro-Silva R, Lopez-Vazquez A et al. BAT1 promoter polymorphism is associated with rheumatoid arthritis susceptibility. J Rheumatol 2008; 35: 741–744.

    CAS  PubMed  Google Scholar 

  37. 37

    van Manen D, Kootstra NA, Boeser-Nunnink B, Handulle MA, van’t Wout AB, Schuitemaker H . Association of HLA-C and HCP5 gene regions with the clinical course of HIV-1 infection. AIDS 2009; 23: 19–28.

    CAS  Article  PubMed  Google Scholar 

  38. 38

    Catano G, Kulkarni H, He W, Marconi VC, Agan BK, Landrum M et al. HIV-1 disease-influencing effects associated with ZNRD1, HCP5 and HLA-C alleles are attributable mainly to either HLA-A10 or HLA-B*57 alleles. PLoS One 2008; 3: e3636.

    Article  PubMed  PubMed Central  Google Scholar 

  39. 39

    Han Y, Lai J, Barditch-Crovo P, Gallant JE, Williams TM, Siliciano RF et al. The role of protective HCP5 and HLA-C associated polymorphisms in the control of HIV-1 replication in a subset of elite suppressors. AIDS 2008; 22: 541–544.

    CAS  Article  PubMed  Google Scholar 

  40. 40

    Tanaka H, Akaza T, Juji T . Report of the Japanese Central Bone Marrow Data Center. Clin Transpl 1996; 9: 139–144.

    CAS  Google Scholar 

  41. 41

    Tokunaga K, Ishikawa Y, Ogawa A, Wang H, Mitsunaga S, Moriyama S et al. Sequence-based association analysis of HLA class I and II alleles in Japanese supports conservation of common haplotypes. Immunogenetics 1997; 46: 199–205.

    CAS  Article  PubMed  Google Scholar 

Download references


We thank all the patients and physicians for their cooperation with this study. We also thank Riken Genesis for the Illumina BeadChip analysis and StaGen for the statistical analysis. This study was supported in part by a Health and Labor Science Research Grant (Research on Advanced Medical Technology) from the Ministry of Health, Labor and Welfare of Japan.

Author information




Corresponding author

Correspondence to M Tohkin.

Ethics declarations

Competing interests

The authors declare no conflict of interest except one member of JPDSC, Mitsubishi Tanabe Pharma, which is a distributor of allopurinol in Japan.


Japan Pharmacogenomics Data Science Consortium (JPDSC)

The Japan Pharmacogenomics Data Science Consortium is composed of Astellas Pharma, Otsuka Pharmaceutical, Daiichi Sankyo, Taisho Pharmaceutical, Takeda Pharmaceutical and Mitsubishi Tanabe Pharma, and is chaired by Ichiro Nakaoka (Takeda Pharmaceutical).

PowerPoint slides

PowerPoint slide for Fig. 1

PowerPoint slide for Fig. 2

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Tohkin, M., Kaniwa, N., Saito, Y. et al. A whole-genome association study of major determinants for allopurinol-related Stevens–Johnson syndrome and toxic epidermal necrolysis in Japanese patients. Pharmacogenomics J 13, 60–69 (2013). https://doi.org/10.1038/tpj.2011.41

Download citation


  • allopurinol
  • Stevens–Johnson syndrome
  • toxic epidermal necrolysis
  • human lymphocyte antigen
  • single nucleotide polymorphism
  • genome-wide association study

Further reading