Wen‐Hung Chung

Allopurinol, a commonly prescribed medication for gout and hyperuricemia, is a frequent cause of severe cutaneous adverse reactions (SCAR), which include the drug hypersensitivity syndrome, Stevens-Johnson syndrome, and toxic epidermal necrolysis. The adverse events are unpredictable and carry significant morbidity and mortality. To identify genetic markers for allopurinol-SCAR, we carried out a case-control association study. We enrolled 51 patients with allopurinol-SCAR and 228 control individuals (135 allopurinol-tolerant subjects and 93 healthy subjects from the general population), and genotyped for 823 SNPs in genes related to drug metabolism and immune response. The initial screen revealed strong association between allopurinol-SCAR and SNPs in the MHC region, including BAT3 (encoding HLA-B associated transcript 3), MSH5 (mutS homolog 5), and MICB (MHC class I polypeptide-related sequence B) (P < 10(-7)). We then determined the alleles of HLA loci A, B, C, and DRB1. The HLA-B*5801 allele was present in all (100%) 51 patients with allopurinol-SCAR, but only in 20 (15%) of 135 tolerant patients [odds ratio 580.3 (95% confidence interval, 34.4-9780.9); corrected P value = 4.7 x 10(-24)] and in 19 (20%) of 93 of healthy subjects [393.51 (23.23-6665.26); corrected P value = 8.1 x 10(-18)]. HLA alleles A*3303, Cw*0302, and DRB1*0301 were in linkage disequilibrium and formed an extended haplotype with HLA-B*5801. Our results indicated that allopurinol-SCAR is strongly associated with a genetic predisposition in Han Chinese. In particular, HLA-B*5801 allele is an important genetic risk factor for this life-threatening condition.

BACKGROUND: Carbamazepine, an anticonvulsant and a mood-stabilizing drug, is the main cause of the Stevens-Johnson syndrome (SJS) and its related disease, toxic epidermal necrolysis (TEN), in Southeast Asian countries. Carbamazepine-induced SJS-TEN is strongly associated with the HLA-B*1502 allele. We sought to prevent carbamazepine-induced SJS-TEN by using HLA-B*1502 screening to prospectively identify subjects at genetic risk for the condition. METHODS: From 23 hospitals in Taiwan, we recruited 4877 candidate subjects who had not taken carbamazepine. We genotyped DNA purified from the subjects' peripheral blood to determine whether they carried the HLA-B*1502 allele. Those testing positive for HLA-B*1502 (7.7% of the total) were advised not to take carbamazepine and were given an alternative medication or advised to continue taking their prestudy medication; those testing negative (92.3%) were advised to take carbamazepine. We interviewed the subjects by telephone once a week for 2 months to monitor them for symptoms. We used the estimated historical incidence of SJS-TEN as a control. RESULTS: Mild, transient rash developed in 4.3% of subjects; more widespread rash developed in 0.1% of subjects, who were hospitalized. SJS-TEN did not develop in any of the HLA-B*1502-negative subjects receiving carbamazepine. In contrast, the estimated historical incidence of carbamazepine-induced SJS-TEN (0.23%) would translate into approximately 10 cases among study subjects (P<0.001). CONCLUSIONS: The identification of subjects carrying the HLA-B*1502 allele and the avoidance of carbamazepine therapy in these subjects was strongly associated with a decrease in the incidence of carbamazepine-induced SJS-TEN. (Funded by the National Science Council of Taiwan and the Taiwan Drug Relief Foundation.).

The anticonvulsant carbamazepine (CBZ) frequently causes cutaneous adverse drug reactions (cADRs), including maculopapular eruption (MPE), hypersensitivity syndrome (HSS), Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN). We reported that SJS/TEN caused by CBZ is strongly associated with the HLA-B*1502 gene in Han Chinese. Here, we extended our genetic study to different types of CBZ-cADRs (91 patients, including 60 patients with SJS/TEN, 13 patients with hypersensitivity syndrome and 18 with maculopapular exanthema versus 144 tolerant controls). We used MALDI-TOF mass spectrometry to screen the genetic association of 278 single nucleotide polymorphisms (SNPs), which cover the major histocompatibility complex (MHC) region, tumor necrosis factor-alpha, heat shock protein and CBZ-metabolic enzymes, including CYP3A4, 2B6, 2C8, 2C9, 1A2 and epoxide hydrolase 1. In addition, we genotyped 20 microsatellites in the MHC region and performed HLA-typing to construct the recombinant map. We narrowed the susceptibility locus for CBZ-SJS/TEN to within 86 kb flanking the HLA-B gene on the extended B*1502 haplotype, and confirmed the association of B*1502 with SJS/TEN [Pc=1.6x10, odds ratio (OR)=1357; 95% confidence interval (CI)=193.4-8838.3]. By contrast to CBZ-SJS/TEN, HLA-B*1502 association was not observed in the MPE or HSS groups: MPE was associated with SNPs in the HLA-E region and a nearby allele, HLA-A*3101 (Pc=2.2x10, OR=17.5; 95% CI=4.6-66.5), and HSS with SNPs in the motilin gene (Pc=0.0064, OR=7.11; 95% CI=3.1-16.5) located terminal to the MHC class II genes. No SNPs in genes involved in CBZ metabolism were associated with CBZ-induced cADRs. Our data suggest that HLA-B*1502 could contribute to the pathogenesis of CBZ-SJS/TEN, and that genetic susceptibility to CBZ-induced cADRs is phenotype-specific.