HLA-B*13 :01 Is a Predictive Marker of Dapsone-Induced Severe Cutaneous Adverse Reactions in Thai Patients

HLA-B*13:01 allele has been identified as the genetic determinant of dapsone hypersensitivity syndrome (DHS) among leprosy and non-leprosy patients in several studies. Dapsone hydroxylamine (DDS-NHOH), an active metabolite of dapsone, has been believed to be responsible for DHS. However, studies have not highlighted the importance of other genetic polymorphisms in dapsone-induced severe cutaneous adverse reactions (SCAR). We investigated the association of HLA alleles and cytochrome P450 (CYP) alleles with dapsone-induced SCAR in Thai non-leprosy patients. A prospective cohort study, 16 Thai patients of dapsone-induced SCARs (5 SJS-TEN and 11 DRESS) and 9 Taiwanese patients of dapsone-induced SCARs (2 SJS-TEN and 7 DRESS), 40 dapsone-tolerant controls, and 470 general Thai population were enrolled. HLA class I and II alleles were genotyped using polymerase chain reaction-sequence specific oligonucleotides (PCR-SSOs). CYP2C9, CYP2C19, and CYP3A4 genotypes were determined by the TaqMan real-time PCR assay. We performed computational analyses of dapsone and DDS-NHOH interacting with HLA-B*13:01 and HLA-B*13:02 alleles by the molecular docking approach. Among all the HLA alleles, only HLA-B*13:01 allele was found to be significantly associated with dapsone-induced SCARs (OR = 39.00, 95% CI = 7.67–198.21, p = 5.3447 × 10−7), SJS-TEN (OR = 36.00, 95% CI = 3.19–405.89, p = 2.1657 × 10−3), and DRESS (OR = 40.50, 95% CI = 6.38–257.03, p = 1.0784 × 10−5) as compared to dapsone-tolerant controls. Also, HLA-B*13:01 allele was strongly associated with dapsone-induced SCARs in Asians (OR = 36.00, 95% CI = 8.67–149.52, p = 2.8068 × 10−7) and Taiwanese (OR = 31.50, 95% CI = 4.80–206.56, p = 2.5519 × 10−3). Furthermore, dapsone and DDS-NHOH fit within the extra-deep sub pocket of the antigen-binding site of the HLA-B*13:01 allele and change the antigen-recognition site. However, there was no significant association between genetic polymorphism of cytochrome P450 (CYP2C9, CYP2C19, and CYP3A4) and dapsone-induced SCARs (SJS-TEN and DRESS). The results of this study support the specific genotyping of the HLA-B*13:01 allele to avoid dapsone-induced SCARs including SJS-TEN and DRESS before initiating dapsone therapy in the Asian population.

HLA-B*13:01 allele has been identified as the genetic determinant of dapsone hypersensitivity syndrome (DHS) among leprosy and non-leprosy patients in several studies. Dapsone hydroxylamine (DDS-NHOH), an active metabolite of dapsone, has been believed to be responsible for DHS. However, studies have not highlighted the importance of other genetic polymorphisms in dapsone-induced severe cutaneous adverse reactions (SCAR). We investigated the association of HLA alleles and cytochrome P450 (CYP) alleles with dapsone-induced SCAR in Thai non-leprosy patients. A prospective cohort study, 16 Thai patients of dapsone-induced SCARs (5

INTRODUCTION
Dapsone (4, 4'-diaminodiphenylsulfone, DDS) is wildly used for treatment of infection and inflammation including of leprosy, Pneumocystis jiroveci pneumonia (PJP), or Toxoplasma gondii encephalitis in human immunodeficiency virus (HIV) prophylaxis, neutrophilic dermatoses, dermatitis herpetiformis, and autoimmune bullous disease (1). However, the most frequent adverse drug reactions of dapsone are dose-dependent adverse effects (hemolytic anemia and methemoglobinemia) and rarely dose-independent adverse effects (dapsone hypersensitivity syndrome) (2). Dapsone hypersensitivity syndrome (DHS) or dapsone-induced hypersensitivity reactions (DIHRs) is a lifethreatening drug reaction and usually manifested between the 4 and 6 weeks after initiation of treatment. The clinically characterized through fever, rash, hepatitis or systemic involvement, lymphadenopathy, and abnormal hematologic system (eosinophilia or atypical lymphocytosis) (3). This entity is also termed DHS and DIHRs has been considered a manifestation of drug reaction with eosinophilia and systemic symptoms (DRESS). There was found approximately 0.5-3.6% of patients treated with dapsone have been reported to develop DHS and the mortality rate of 9.9% (4). Especially, about 2% of leprosy patients treated with dapsone have a DHS and 12.5% of mortality (5,6). According to data from the King Chulalongkorn Memorial Hospital, Thailand reported during 2004-2014, dapsone is the 5 th ranked common culprit drug causing DRESS in Thai patients (7).
Severe cutaneous adverse drug reactions (SCARs) is a type of adverse drug reactions (ADRs) that remains a rare but potentially severe life-threatening adverse effect and major problems for both clinical treatment and pharmaceutical industry (8). SCARs comprise a heterogeneous groups of distinct clinical manifestation, including of Stevens-Johnson syndrome (SJS), toxic epidermal necrolysis (TEN), drug reaction with eosinophilia and systemic symptoms (DRESS), and acute generalized exanthematous pustulosis (AGEP) (9). Clinical characteristic of SJS, SJS/TEN overlap, and TEN are acute and rapid progression of mucous detachment and systemic symptoms. They are differentiated by the severe of skin detachment, involving <10% of body surface area (BSA) in SJS, 10-30% of BSA in SJS/TEN overlap, and >30% of BSA in TEN (10). According to the RegiSCARs study, SJS has a mortality rates in the range from about 10% and more than 40% for TEN (11). The main causes of SJS-TEN are medicines and risk factors such as HIV infection, renal disease, liver disease, and active systemic autoimmune disease (12). Drug reaction with eosinophilia and systemic symptoms (DRESS) are characterized by a skin rash usually occurring more than 2 weeks after drug initiation with fever, hepatitis or internal organ involvement, lymphadenopathy, and hematological abnormalities (eosinophilia or atypical lymphocytosis) (13). The mortality rate of DRESS is approximately 10% (14).
Although the exact mechanism of SCARs remains unclear, numerous studies have described the associations between human leukocyte antigen (HLA) and cytochrome P450 genes with the specific drug hypersensitivity reaction (15,16). For example, HLA-B*15:02 with carbamazepine-induced SJS/TEN is recommended for Han Chinese, Malaysia, India, and Thailand (17)(18)(19)(20). On the contrary, HLA-A*31:01 is the main genetic determinant for carbamazepine-induced SJS, TEN, and DRESS in Japanese and Europeans (21,22). Thus, HLA-B*15:02 is phenotype-specific with carbamazepine-induced SJS/TEN in each population. Additionally, there were important discovered the drug metabolism enzymes of phenytoin-induced SJS-TEN. The metabolize processes of phenytoin to p-HPPH (inactive form), arene oxides were cause of phenytoin hypersensitivity reactions by poor metabolizer (PM) alleles of mutation CYP2C9 gene consist of CYP2C9*2 and CYP2C9*3 in Asian (23,24).
In previous studies, only HLA-B*13:01 was strongly associated with DHS in leprosy Han Chinese (odds ratio 122.1, p-value = 6.038 × 10 −12 and odds ratio 20.53, p-value = 6.84 × 10 −25 ) and dapsone-induced DRESS in non-leprosy Thais (odds ratio = 60.75, p-value = 0.0001) (25)(26)(27). Furthermore, Dapsone is metabolized through acetylation and N-hydroxylation. In human study, they found a relation between the rate of Nhydroxylation and clearance of dapsone by cytochrome P450 (28). Genetic polymorphisms of CYP2C9, CYP2C19, and CYP3A4 influenced the dapsone metabolism and cause of DHS through by DDS-NHOH (dapsone hydroxylamine) (29). Nevertheless, there are no data describing whether HLA class I, II alleles and cytochrome P450 is a valid marker for prediction of dapsone-induced SCARs in non-leprosy patients in addition to HLA-B*13:01. Consequently, the aim of this study was to investigate the contributing pharmacogenetics markers association between HLA class I, II, cytochrome P450, and dapsone-induced SCARs in Thai non-leprosy patients.

Subjects
We enrolled 16 non-leprosy Thai patients with dapsone-induced SCARs consist of 5 SJS-TEN patients and 11 DRESS patients were classified by RegiSCAR criteria. SJS is defined as skin detachment less than 10% of BSA, SJS/TEN overlap has 10-30% of BSA involved, and TEN as skin detachment more than 30% of BSA (30). Moreover, SJS-TEN with severe ocular surface complications (SOC) was diagnosis with history of acute-onset high fever, serious mucocutaneous illness with skin eruption, and the involvement of at least two mucosal sites (oral cavity and ocular surface) (31). DRESS was defined by the triad of skin eruption, hematological involvement, and internal organ involvements according to the RegiSCAR Group Diagnosis Score (13). All patients with dapsoneinduced SCARs were accessed through review of photographs, pathologic slides, and medical records by two dermatologists. Furthermore, there were two cases with SJS-TEN and seven cases with DRESS in the Taiwan population. Forty dapsone-tolerant controls who had been non-leprosy Thai patients and received dapsone more than 6 months without any cutaneous adverse reaction.
All of participants in this study from the Faculty of Medicine Ramathibodi Hospital, Mahidol University, Faculty of Medicine, There were collected the clinical data of dapsone-induced SCARs and controls consist of age, gender, indication for dapsone treatment, dapsone dose (mg/day), co-medication, complete blood cell count (CBC), blood urea nitrogen (BUN), serum creatinine (SCr), aspartate aminotransferase (AST) or serum glutamic oxaloacetic transaminase (SGOT), and alanine aminotransferase (ALT) or serum glutamic pyruvic transaminase (SGPT).

HLA Class I and II Genotyping
HLA class I and II alleles were genotyped using sequence-specific oligonucleotides (PCR-SSOs). Diluted DNA sample was amplified polymerase chain reaction (PCR) by GeneAmp ® PCR System 9700 (Applied Biosystems, Waltham, USA). The PCR product was then hybridized against a panel of oligonucleotide probes on coated polystyrene microspheres that had sequences complementary to stretches of polymorphism within the target HLA class I and II alleles using the Lifecodes HLA SSO typing kits (Immucor, West Avenue, Stamford, USA) and detection by the Luminex ® IS 100 system (Luminex Corporation, Austin, TX, USA). HLA class I and II alleles were performed using MATCH IT DNA software version 3.2.1 (One Lambda, Canoga Park, CA, USA).
In Silico Model of Dapsone, DDS-NHOH, and HLA-B*13:01 Complex and healthy Thai population. Statistical analysis was performed using SPSS version 16.0 (SPSS Inc., Chicago, IL, USA). The association was estimated by calculating the odds ratio (OR) with a 95% confidence interval (CI). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. The corrected P-values (P c ) for the multiple comparison of HLA alleles (16 for HLA-A, 22 for HLA-B, 20 for HLA-C, 18 for HLA-DRB1, 9 for HLA-DQA1, and 11 for HLA-DQB1) were calculated using Bonferroni's correction. P-values were less than 0.05 (two-tailed) was considered to indicate statistically significant.

Clinical Characteristic
The demographic and clinical data of patients with dapsoneinduced SCARs and controls are listed in Table 1. Patients who were diagnosed with SJS, TEN, and DRESS were validated as "probable" and "definite" case by dermatologists using RegiSCAR criteria and all of dapsone-induced SJS-TEN patients without severe ocular complications (SOC). The 16 patients with dapsone-induced SCARs consisted of 10 females (62.5%) and 6 males (37.5%), with a median age of 45 (range 2.5-64) years. Meanwhile, 28 (70%) dapsone controls were females with a median age of 41.5 (range 4-75) years. The median onset time of SJS-TEN and DRESS was 32.5 (14-56) and 31.5 (3-63) days, respectively, after exposure to dapsone. The median onset time of SJS-TEN and DRESS were not significantly different. Dapsone was used among the cases and controls for the HIV prophylaxis (25.00% of cases, 17.50% of controls), systemic lupus erythematosus (SLE) (18.75% of cases, 22.50% of controls), chronic bullous disease of childhood (CBDC) (6.25% of cases, 7.50% of controls), and immune thrombocytopenic purpura (ITP) (18.75% of cases, 2.50% of controls). Eight patients (20.00%) had a previous history of cotrimoxazole-induced hypersensitivity reaction in the dapsone-tolerant group. Dapsone dosages used were 100 mg/day, while two patients (2.5 and 4 years old) received 18 and 16.7 mg/day, respectively. The hematological abnormalities and hepatitis were more prominent among the dapsone cases, as shown in Table 1. Furthermore, the most common of comedication used among the dapsone cases and controls were colchicine, efavirenz, lamivudine, and acyclovir.
When we compared the frequency of HLA-B*13:01 allele of seven Asian patients with dapsone-induced SJS-TEN and dapsone-tolerant control Thais and the general Thai population, HLA-B*13:01 allele was strongly associated with dapsone-induced SJS-TEN among Asians compared to the dapsone-tolerant control Thais (OR: 54.00; 95% CI: 5.12-569. 39               Association Between Dapsone-Induced DRESS and HLA Class I, II Alleles The association of HLA class I and II alleles with dapsoneinduced DRESS were shown in Table 6. We found that 81.82% (9/11) Table 3).
On comparing, 11 dapsone-induced DRESS cases with 40 tolerant controls and 470 general Thai population, the frequencies of HLA-C*03:04, HLA-DQB1*06:01, HLA-B*13:01-C*03:04, and HLA-B*13:01-DQB1*06:01 were significantly associated with dapsone-induced DRESS (p-value < 0.05). However, HLA-B*15:02 allele was not statistically significant association with dapsone-induced DRESS when compared with tolerant controls and Thai population by p-value of 0.1617 and 0.3873, respectively. When the frequencies of HLA alleles in Asian and Taiwanese group were compared with those in Thai dapsone-tolerant controls and the general Thai population, only the HLA-B*13:01 allele was associated with dapsone-induced DRESS ( Table 4) (Figures 1 and 2). The docking results in Table 8 showed that although dapsone likely interacted with both proteins via an insertion of its -NH 2 group into the Fpocket (90. 4

DISCUSSION
The immunopathogenesis of SCARs are associated with expression of specific HLA allele, T-lymphocyte, structure of drug and peptide molecules (34,35). In this study, we presented the highly specific association of HLA-B    assumptions. Further studies using a large number of samples are required for better comprehension.
In previous study, the detection of HLA-B*13:01-restricted dapsone and metabolite form-responsive CD8+ clones indicates that dapsone hypersensitivity syndrome should be used as an example to discover the structural features of drug, HLA binding and interaction (56). The in silico model suggested that the 5carboxamide group of CBZ might interact with Arg 62 of B pocket of HLA-B*15:02 (binding energy -37.104 kcal/mol) and Asn 63 contributes to the specificity in HLA recognition (57). In this study, we found three amino acid residues on an extra deep sub-pocket on F pocket within the antigen-binding site of HLA-B*13:01 and binding affinity of dapsone and DDS-NHOH for HLA-B*13:01 was much greater than HLA-B*13:02. Additionally, a docking model between dapsone and DDS-NHOH and HLA-B*13:01 allele was found to be appropriate because specific interaction triggers structural changes in the antigen-recognition site, allowing the protein to recognize peptides that are conformationally altered. Specific HLA allele plays a major immunopathogenesis role of drug hypersensitivity reactions, several hypotheses have been proposed to explain the interaction of HLA, drugs, peptides, and T cell (58). In brief, the hapten/prohapten model proposes that a chemically active drug or its metabolite forms a covalent bond with an endogenous peptide and then is intracellularly processed and presented by the particular HLA. While, the direct pharmacological interaction (p-i) model involves a non-covalent and labile interaction of the drug with HLA at the cell surface independent of antigen processing or T cell receptor. Another hypothesis, the altered peptide repertoire model, suggests the drug or its metabolites can bind non-covalent within the pocket of binding groove of certain HLA allele (34,58). Thus, the altered peptide repertoire model involves the binding of dapsone and DDS-NHOH to HLA-B*13:01 allele and explains why the specific HLA-B*13:01 allele is a marker of dapsone-induced SCARs, despite the cytochrome P450 gene is responsible for the metabolism of dapsone to dapsone hydroxylamine.
This study confirms the specific association between HLA-B*13:01 and dapsone-induced SCARs including SJS-TEN and DRESS in the Thai and Taiwanese population. Although HLA-A*24:07, HLA-C*03:04, HLA-DRB1*15:01, and HLA-DQB1*06:01 were associated with dapsone-induced SCARs, none of these associations were considered statistically significant after Bonferroni's correction. Furthermore, there was no association between genetic polymorphisms of CYP2C9, CYP2C19, and CYP3A4 and dapsone-induced SCARs. In addition to the specific interaction of dapsone and DDS-NHOH at the extra deep subpocket around the F pocket on HLA-B*13:01 allele, resulting in a change in the structure of antigen-recognition site of HLA-B*13:01 may induce altered peptides that bind to this HLA allele. Consequently, only HLA-B*13:01 might serve as a pharmacogenetics marker for screening before initiating the therapy with dapsone for the prevention of dapsoneinduced SCARs.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the ethics committee of Ramathibodi Hospital (MURA2016/105), Khon Kaen University (HE510837) and Udon Thani Hospital (22/2563). The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
All authors helped to perform the research. PS's contribution included sample collection, manuscript writing, drafting conception and design, performing procedures, and data analysis. JP, PR, JK, NN, TRu, PK, NS, AM, WA, UK, TT, KW, PJ, NK, TJ, TRe, C-WW, DN, WT, Ma, TRo, MP, and W-HC contributed to sample collection, data analysis and contribution to writing the manuscript. CS contributed to drafting conception, design, and contribution to writing the manuscript. All authors contributed to the article and approved the submitted version.