The Minor Alleles HCP5 Rs3099844 A and PSORS1C1 Rs3131003 G Are Associated with Allopurinol-Induced Severe Cutaneous Adverse Reactions in Han Chinese: a Multicentre Retrospective Case-Control Clinical Study
L. Cheng,B. Sun,Y. Xiong,L. Hu,L. Gao,Q. Lv,M. Zhou,J. Li,X. Chen,W. Zhang,H.-H. Zhou
DOI: https://doi.org/10.1111/bjd.16151
IF: 11.113
2018-01-01
British Journal of Dermatology
Abstract:DEAR EDITOR, Allopurinol-induced severe cutaneous adverse reactions (SCARs), which include drug reaction with eosinophilia and systemic symptoms (DRESS) syndrome, Stevens– Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN), are life-threatening conditions and occur almost exclusively in HLA-B*5801 positive individuals in Han Chinese. Recent studies have shed light on epigenetic modifications across autoimmune diseases and have described associations between drug hypersensitivities and single-nucleotide polymorphisms (SNPs). We therefore undertook a study to investigate whether other genetic factors, such as epigenetic variation and SNPs, are associated with allopurinol-induced SCARs in Han Chinese. The study was a multicentre retrospective case–control study. It was registered in the Chinese Clinical Trial Registry (ChiCTR-RCC-12002927) and was approved by the Independent Ethics Committee of the Institute of Clinical Pharmacology, Central South University (CTXY-110011-2). We enrolled 97 participants with allopurinol-induced SCARs (the SCARs group) and 97 who were allopurinol tolerant (the control group) of Han descent. The diagnosis was made by dermatologists according to the RegiSCAR or Roujeau criteria. Details of inclusion and exclusion criteria have been described previously. Genome-scale DNA methylation profiling on an Illumina Infinium HumanMethylation450 platform (Illumina, San Diego, CA, U.S.A.) involving 484 660 CpG dinucleotides in 15 patients with allopurinol-induced SCARs and 20 ageand sex-matched controls was performed. Weighted gene co-expression network analysis (WGCNA) and differentiated methylated positions (DMPs) analysis were used to pick up candidate genes for SNP analysis. TP53AIP1 was among one of the top 10 hub genes in the yellow module after the WGCNA analysis. The genes DUSP22, PDCD1LG2, NOTCH4 and ATP6V1C1 were selected from the top 20 DMPs. TSHZ2 was selected after reviewing its DMP expression data and WGCNA analysis. Other SNPs were selected after reviewing the drug-hypersensitivity reports. Finally, six SNPs selected from methylation profiling and 13 reported drug-hypersensitivity-related SNPs were examined in the SCARs (n = 97, 41% DRESS syndrome, 41% SJS, 6% SJS/ TEN overlap, 11% TEN) and control (n = 97) groups. Genomic DNAs extracted from participants’ peripheral blood were sent for SNP genotyping. SNPs were typed using iPLEX Chemistry (Agena Bioscience, San Diego, CA, U.S.A.) on a matrix-assisted laser desorption/ionization time-of-flight mass spectrometer (MALDI-TOF-MS, Sequenom, San Diego, CA, U.S.A.). A v-test was used for comparison of categorical variables. Linkage disequilibrium (LD) and haplotype analysis were conducted using Haploview (Broad Institute, Cambridge, MA, U.S.A.). Logistic regression analysis with covariates was applied to correct for multiple factors. The corrected P (Pc) values were adjusted by multiplying by 19 using a post-hoc Bonferroni correction. After analysis, five out of 19 SNPs were found to be associated with SCARs with significance (P < 0.05, details available from the authors on request). After logistic regression analysis with age and sex as covariates, four SNPs HCP5 rs3099844 (C>A), PSORS1C1 rs3131003 (A>G), TSHZ2 rs2010156 (C>T) and NOTCH4 rs367398 (G>A) were still found significantly associated with SCARs. Among five differentiated polymorphisms (Table 1), the risk of developing SCARs was 32.41folder higher in HCP5 rs3099844 A carriers than in CC carriers (Pc < 0.0019). The risk of developing SCARs was 17.75-fold higher in PSORS1C1 rs3131003 G carriers than AA carriers (Pc < 0.0019). In the recessive model of PSORS1C1 rs3131003, the GG homozygous carriers were at 3.33-fold higher risk of developing SCARs than individuals with GA + AA (Pc = 0.019). The TSHZ2 rs2010156 T carriers had a lower risk of developing SCARs [odds ratio (OR) 0.27] than CC homozygotes (this finding was no longer significant after correction). Individuals with NOTCH4 rs367398 A variants were at lower risk of developing SCARs in both dominant and recessive models. Results for both dominant and recessive model for PSORS1C1 rs3815087 were not significant. The study confirmed that the minor alleles HCP5 rs3099844 A and PSORS1C1 rs3131003 G are strongly associated with allopurinol-induced SCARs. Our data do not support an association between the other tested SNPs and allopurinol-induced SCARs in Han Chinese. In haplotype analysis, the frequency of the GGA (rs3131003/rs3815087/rs3099844) haplotype at chromosome 6 was significantly higher in the SCARs group than the control group (44% vs. 8%, OR 9.26, Pc = 1.1305 9 10). This shows that the patients carrying the GGA haplotype are 9.26-fold more susceptible to developing allopurinol-induced SCARs. This finding is in agreement with a former result that the chromosome 6 haplotype CACGAC formed by six SNPs