Human Immunology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

www.ashi-hla.org

journal homepage: www.elsevier .com/locate /humimm

Rapid Communication

HLA-B⁄1502 is associated with carbamazepine induced Stevens–Johnsonsyndrome in North Indian population

http://dx.doi.org/10.1016/j.humimm.2014.09.0220198-8859/� 2014 Published by Elsevier Inc. on behalf of American Society for Histocompatibility and Immunogenetics.

⇑ Corresponding author. Fax: +91 172 274401.E-mail address: ritu_immunopath@yahoo.co.in (R. Aggarwal).

Please cite this article in press as: Aggarwal R et al. HLA-B⁄1502 is associated with carbamazepine induced Stevens–Johnson syndrome in Northpopulation. Hum Immunol (2014), http://dx.doi.org/10.1016/j.humimm.2014.09.022

Ritu Aggarwal a,⇑, Madhulika Sharma a, Manish Modi b, Vivek kumar Garg b, Manilla Salaria a

a Department of Immunopathology, Post Graduate Institute of Medical Education and Research, Chandigarh, Indiab Department of Neurology, Post Graduate Institute of Medical Education and Research, Chandigarh, India

a r t i c l e i n f o

Article history:Received 8 July 2014Accepted 27 September 2014Available online xxxx

Keywords:HLAIndiaPCRPharmacogeneticsToxic epidermal necrolysis

a b s t r a c t

The evidence of association between HLA-B⁄1502 and anticonvulsant induced Stevens–Johnsonsyndrome (SJS) and toxic epidermal necrolysis (TEN) from the Indian population is scant. Patients witha history of SJS/TEN secondary to carbamazepine or phenytoin were enrolled. The control group com-prised of patients who had received carbamazepine/phenytoin for P6 months without any adverse cuta-neous event. Low-resolution DNA typing for HLA-B and high resolution HLA-B⁄15 typing was performed.Seventeen patients with history of SJS/TEN secondary to carbamazepine (9) or phenytoin (8) and 50 tol-erant controls (carbamazepine-37; phenytoin-13) were enrolled. The mean age of patients and controlswas 33.9 ± 11.6 and 28.1 ± 9.9 years, respectively. HLA-B⁄1502 was observed in 2/9 (22.2%) carbamaze-pine-SJS/TEN patients and none of the 37 carbamazepine tolerant controls (p = 0.035). HLA-B⁄1502 wasnot observed in any of the 8 phenytoin-SJS/TEN patients or the 13 phenytoin tolerant controls. Our datasuggests that HLA-B⁄1502 is a risk factor for carbamazepine induced SJS/TEN. Therefore, HLA-B⁄1502testing should be performed prior to initiating carbamazepine in North Indian population.

� 2014 Published by Elsevier Inc. on behalf of American Society for Histocompatibility andImmunogenetics.

⁄

1. Introduction

Adverse drug reaction as defined by WHO, is a noxious andunintended response to a drug that occurs at a dose normally usedin man [1]. They can range from mild maculopapular exanthema tolife threatening conditions, including, Stevens–Johnson syndrome(SJS) and toxic epidermal necrolysis (TEN). In a recent systematicreview of the published evidence of drug-induced SJS and TEN inIndian population, carbamazepine (18.2%) and phenytoin (13.4%)were commonly implicated drugs [2].

Identification of candidate genes that predispose to anticonvul-sant induced severe cutaneous reactions offers the possibility ofavoiding the drug in genetically susceptible individuals. HLA mol-ecules are crucial for antigen presentation to T cell receptor. TheHLA genes are located at the major histocompatibility (MHC)region on small arm of human chromosome 6p, spanning approx-imately 200 genes. There are more than 10 highly polymorphicHLA genes, including, HLA-A, HLA-B, HLA-C, HLA-DR, HLA-DQand HLA-DP loci. HLA-B is the most polymorphic gene in thehuman genome [3].

Studies have suggested an association between HLA-B 1502and carbamazepine induced SJS in patients of Chinese/Asian eth-nicity [4]. The data from the Indian population is scant. There areno guidelines for screening for HLA-B⁄1502 prior to prescribingcarbamazepine in India. There is urgent need for studies to deter-mine the prevalence of SJS/TEN, and HLA B⁄1502 in various ethnicgroups in Asia and its association with carbamazepine-inducedSJS-TEN in each of these ethnic groups. In view of the significantmorbidity and mortality in SJS-TEN, facilities should be developedfor screening of HLA B⁄1502 before carbamazepine is prescribed inthe susceptible population. The aim was to explore the associationof HLA-B⁄1502 and carbamazepine and phenytoin induced SJS/TEN.

2. Materials and methods

The study design was case–control and hospital based. It wasconducted over a 2-year period from February 2012 to February2014. Patients with a history of SJS or TEN secondary to carbamaz-epine or phenytoin were recruited from the Neurology outpatient.SJS was defined as skin detachment of 610% of body surface area.TEN was defined as skin detachment of P30% [5]. The controlgroup comprised of patients who had received carbamazepine orphenytoin for P6 months without any adverse cutaneous event.

Indian

2 Rapid Communication / Human Immunology xxx (2014) xxx–xxx

One milliliter of blood was collected in EDTA from cases as wellas controls and was subjected to DNA extraction using a commer-cially available kit (Qiagen, Hilden, Germany). The quantity of theDNA was estimated by spectrometry and quality was establishedby performing polymerase chain reaction (PCR) for beta actin, ahousekeeping gene. Subsequently, PCR with 48 sequence specificprimers for HLA-B antigen was performed. Cases positive forHLA-B⁄15 antigen were subjected to high resolution HLA-B⁄15 typ-ing using sequence specific primers. Commercially available kitsfor low resolution DNA typing for HLA-B and high resolutionHLA-B⁄15 (Invitrogen, Life Technologies, Carlsbad, CA, USA) wereutilized. The study was approved by the Institutes Ethics Commit-tee. Written informed consent was obtained from each patient/control.

Statistical analysis of the difference in the frequency of HLA-B⁄1502 among patients and controls was performed by Fisher’sexact test (GraphPad Software, Inc., CA, USA). The reported p valuesare two tailed. p values <0.05 are considered significant.

3. Results

Seventeen patients with a history of SJS/TEN secondary to car-bamazepine (9) or phenytoin (8) were enrolled as cases. Fifty anti-convulsant tolerant patients (carbamazepine-37; phenytoin-13)formed the control group. The mean age of the patients and con-trols was 33.9 ± 11.6 years (range: 15–50) and 28.1 ± 9.9 years(range: 15–55), respectively. Among the SJS/TEN patients, 13 werefemales and 4 males (M:F 1:3.2). Among the controls, M:F ratiowas 3:2. Majority of the cases were of SJS – 13 (76.4%), followedby TEN – 3 (17.6%) and a single case of SJS/TEN overlap. Mucosallesions were observed in 8 (47%).

HLA-B⁄1502 was observed in 2 of 9 (22.2%) carbamazepine-SJS/TEN patients and none of the 37 carbamazepine tolerant controls(p = 0.035). Both the patients with HLA-B⁄1502 genotype were het-erozygous. HLA-B⁄1502 was not observed in any of the 8 phenytoininduced SJS/TEN patients or the 13 phenytoin tolerant controls.Other HLA-B genotypes observed in SJS/TEN patients includedHLA-B⁄1505 (carbamazepine) and HLA-B⁄1518 (phenytoin) in oneeach. In tolerant controls, HLA-B⁄1517 was observed in 4 (8%),HLA-B⁄1501 in 2 (4%) and HLA-B⁄1510 in 1 (2%); none had aHLA-B75 genotype. The sensitivity and specificity of HLA-B⁄1502for carbamazepine induced SJS/TEN was 22.2% and 100%,

Table 1Selected studies of association of HLA-B⁄1502 with anticonvulsant induced Stevens–Johns

S.No.

Country, year ofpublication, reference

No. of patients Drug Association with HLA-B⁄

1 Taiwan, 2004 [4] SJS/TEN: 44;tolerantcontrols: 101Controls:93

CBZ 44 (100%); tolerant cont

2 Europe, 2006 [15] SJS/TEN: 12; CBZ No association in Europein those with Asian ance

4 Thailand, 2008 [16] CBZ-SJS: 6PHT-SJS: 4tolerantcontrols: 50

CBZ,PHT

6 (100%): CBZ4 (100%): PHT8 (19%): CBZ-tolerant8 (18%) PHT-tolerant

5 Hong Kong, 2007 [17] CBZ-SCR: 4PHT-SCR: 1tolerantcontrols: 48

CBZ,PHT

4 (100%): CBZ-SCR1 (100%): PHT-SCR

6 India, 2009 [12] SJS: 8controls: 10

CBZ 6 (75%): CBZ Controls: n

7 Japan, 2010 [18] SJS/TEN: 14 CBZ Nil8 Thailand, 2010 [19] SJS/TEN: 42

tolerantcontrols: 42

CBZ 37 (88%): CBZ5 (12%): CBZ-tolerant

SCR, severe cutaneous reaction; CBZ, carbamazepine; PHT, phenytoin; SJS, Stevens–John

Please cite this article in press as: Aggarwal R et al. HLA-B⁄1502 is associatedpopulation. Hum Immunol (2014), http://dx.doi.org/10.1016/j.humimm.2014.0

respectively. The corresponding positive and negative predictivevalue were 100% and 84.1%, respectively.

TEN and SJS are acute life threatening conditions. Epidermalnecrosis causes erosion of the mucous membrane, extensivedetachment of the epidermis and severe constitutional symptoms.These conditions are capable of killing or severely disabling previ-ously healthy people. Different anticonvulsant drugs have varyingpotential to cause cutaneous adverse drug reactions. The commondrugs causing life threatening SJS/TEN differ from country to coun-try. In Europe, carbamazepine is the most common causative drugaccounting for 8.2%, followed by phenobarbital (5.3%) and phenyt-oin (5%) [6]. In Asia, particularly, Southeast Asian countries, car-bamazepine accounts for an even higher share of SJS/TEN; 26% inTaiwan, 35.7% in Malaysia and 27.7% in Singapore [7].

Etiology of 57 patients with TEN/SJS, treated over a period of6 years has been previously reported from our centre [8]. Anticon-vulsants were responsible for 44% of life threatening TEN/SJS. Phe-nytoin and carbamazepine was the incriminating drug in 11 (19%)cases, each [8]. In another retrospective study from South India,etiology of 41 patients with TEN/SJS was evaluated [9]. Carbamaz-epine and phenytoin were the incriminating drugs in 18 (32%) and3 (5%) cases, respectively. Authors opined that the increased num-ber of prescriptions of carbamazepine for the control of pain wasthe plausible reason for the increased frequency of SJS/TEN dueto carbamazepine [9].

The mechanism by which drugs result in cutaneous adversereaction is ill understood. Immune reactions can be involved inSJS/TEN; a re-challenge with the same drug typically results insevere symptoms within 2–3 days. In addition, cytotoxic T cellsinfiltrate the skin lesions of SJS/TEN. The blister fluid in the lesionsdemonstrate a predominance of CD8+ phenotype [10]. The obser-vation indicate that CD8+ T cell mediated cytotoxic response arethe major events in SJS/TEN [10].

HLA-B⁄1502 was found in 22.2% patients with carbamazepineinduced SJS/TEN in our study as compared to none of the controls(p = 0.035). The first drug-induced HLA-B⁄1502 allele-associatedcase from India was reported in 2009 [11]. Subsequently the loneseries was reported from the western state of Gujarat; HLA-B⁄1502 was identified in 6 of 8 (75%) patients with carbamaze-pine-SJS/TEN [12]. Several more studies need to be conducted toidentify the frequency of HLA-B⁄1502 in different regions of thediverse population of India. The less strong association observedin our study as compared to the one from Gujarat could possibly

on syndrome and toxic epidermal necrolysis.

1502 Ethnicity

rols: 3%, controls: 8.6% Han Chinese

an population. 4 (100%) associationstry

Germany, France, Vietnam, China,Cambodia and Reunion islandThai population

Han Chinese

il Western India

JapaneseThai population

son syndrome; TEN, toxic epidermal necrolysis.

with carbamazepine induced Stevens–Johnson syndrome in North Indian9.022

Rapid Communication / Human Immunology xxx (2014) xxx–xxx 3

be explained by the varied genetic ancestry. North Indians are con-sidered to be genetically closer to Caucasians with a possible originin Eurasia, whereas the west central India which includes ethnicpopulation of Gujarat is more closely related to other Asian thanEuropean groups [13,14]. Selected studies of association of HLA-B⁄1502 with anticonvulsant induced SJS/TEN are illustrated inTable 1.

A pharmacogenetic study is more likely to yield a positive resultwhen conducted in a population with a high frequency of theincriminated allele [20]. Association of up to 100% has beenreported from selected regions/ethnic populations, e.g., Han Chi-nese and Thai population [4,16]. The strong association pointstowards a direct functional involvement of HLA molecules in thepathogenesis. Greater than 15% of the population is reported posi-tive for HLA-B⁄1502 in Hong Kong, Malaysia, Thailand and parts ofthe Philippines, compared to about 10% in Taiwan and 4% in NorthChina [21]. South Asians, including Indians, appear to have inter-mediate prevalence of HLA-B⁄1502, averaging 2–4%. HLA-B⁄1502is present in <1% of the population in Korea and Japan [21]. HLA-B⁄1502 is largely absent in Caucasians, African-Americans, Hispan-ics, and Native Americans [21].

Apart from HLA-B⁄1502, other members of HLA-B75 serotypehave been implicated in carbamazepine-SJS/TEN. Kaniwa et al.from Japan reported association of HLA-B⁄1511 in 4 (29%) patientsof SJS/TEN, however did not detect HLA-B⁄1502 [18]. HLA-B⁄1511belongs to the category of HLA-B75 serotype as does HLA-B⁄1502. The other members include HLA-B⁄1508, HLA-B⁄1505and HLA-B⁄1521. Mehta et al. from India observed HLA-B⁄1508 in1/8 carbamazepine-SJS/TEN patients and HLA-B⁄1521/95 in 2/10normal controls [12]. HLA-B⁄1518 in Japanese and HLA-B⁄1521 inThai population have been reported in carbamazepine-SJS/TEN aswell [19,22]. Wei et al. rationalized that similar amino acidsequence homology may resemble structural features of HLA-B⁄1502 and thus may be able to trigger a similar cutaneous adversereaction to carbamazepine [23]. We did not detect any HLA-B75serotypes, other than HLA-B⁄1502 in selected patients with SJS/TEN. Interestingly, recent studies have demonstrated an associa-tion between HLA-A⁄3101 and CBZ-induced adverse drug reactionsin Caucasian and Japanese populations [24].

The strong associations with HLA alleles support an importantrole for HLA molecules in drug hypersensitivity, and it favors thehapten concept [25]. However, 78% of patients with CBZ-inducedTEN/SJS lacked HLA-B⁄1502 in our study. It is plausible that othersusceptibility genes may be located beyond the range of genotyp-ing screening. It might be possible that the non-covalent directinteraction of the drug or metabolite with matching T cell recep-tors with MHC molecules triggers an immune response [25].

Unlike carbamazepine-SJS/TEN there was no associationbetween phenytoin-SJS/TEN and HLA-B⁄1502 in our study. In astudy from Thailand, Locharernkul et al. observed HLA-B⁄1502 in4 (100%) patients with phenytoin-SJS and in 8 (18%) phenytoin tol-erant controls (p = 0.005) [16]. The M:F ratio of patients with SJS/TEN in our study is 1:3.2. The skewed female majority is likely arandom finding. The M:F ratio of patients with carbamazepine-induced SJS/TEN in the previous study from India was 1:1 [12].The ratio was 2:1 in a European study that comprised patients ofvaried ethnicity [15].

HLA-B⁄1502 is a risk factor for carbamazepine induced SJS/TENin North Indian population. Given the serious and life threateningconsequences of developing SJS/TEN and the availability of alterna-tive drugs, HLA-B⁄1502 testing should be performed prior to initi-ating carbamazepine in India. Physicians in India should be madeaware; facilities for testing HLA-B⁄1502 should to be easilyaccessible and economical.

Please cite this article in press as: Aggarwal R et al. HLA-B⁄1502 is associatedpopulation. Hum Immunol (2014), http://dx.doi.org/10.1016/j.humimm.2014.0

Acknowledgment

The study received funding from the Post Graduate Institute ofMedical Education and Research, Chandigarh.

References

[1] World Health Organization. International drug monitoring: the role of nationalcenters. World Health Tech Rep Ser 1972;498:1–25.

[2] Patel TK, Barvaliya MJ, Sharma D, Tripathi C. A systematic review of the drug-induced Stevens–Johnson syndrome and toxic epidermal necrolysis in Indianpopulation. Indian J Dermatol Venereol Leprol 2013;79:389–98.

[3] Horton R, Wilming L, Rand V, et al. Gene map of the extended human MHC. NatRev Genet 2004;5:889–99.

[4] Chung WH, Hung SI, Hong HS, et al. Medical genetics: a marker for Stevens–Johnson syndrome. Nature 2004;428:486.

[5] Roujeau JC. The spectrum of Stevens–Johnson syndrome and toxic epidermalnecrolysis: a clinical classification. J Invest Dermatol 1994;102:28S–30S.

[6] Mockenhaupt M, Viboud C, Dunant A, et al. Stevens–Johnson syndrome andtoxic epidermal necrolysis: assessment of medication risks with emphasis onrecently marketed drugs. The EuroSCAR-study. J Invest Dermatol2008;128:35–44.

[7] Hung SI, Chung WH, Chen YT. HLA-B genotyping to detect carbamazepine-induced Stevens–Johnson syndrome: implications for personalizing medicine.Pers Med 2005;2:225–37.

[8] Sharma VK, Sethuraman G, Kumar B. Cutaneous adverse drug reactions:clinical pattern and causative agents – a 6 year series from Chandigarh, India. JPostgrad Med 2001;47:95–9.

[9] Devi K, George S, Criton S, Suja V, Sridevi PK. Carbamazepine – the commonestcause of toxic epidermal necrolysis and Stevens–Johnson syndrome: a study of7 years. Indian J Dermatol Venereol Leprol 2005;71:325–8.

[10] Nassif A, Bensussan A, Dorothée G, et al. Drug specific cytotoxic T-cells in theskin lesions of a patient with toxic epidermal necrolysis. J Invest Dermatol2002;118:728–33.

[11] Shankarkumar U, Shah KN, Ghosh K. Letter: HLA B⁄1502 allele association withoxcarbamazepine-induced skin reactions in epilepsy patient from India.Epilepsia 2009;50:1837–8.

[12] Mehta TY, Prajapati LM, Mittal B, et al. Association of HLA-B⁄1502 allele andcarbamazepine-induced Stevens–Johnson syndrome among Indians. Indian JDermatol Venereol Leprol 2009;75:579–82.

[13] Agrawal S, Srivastava SK, Borkar M, Chaudhuri TK. Genetic affinities of Northand Northeastern populations of India: inference from HLA-based study.Tissue Antigens 2008;72:120–30.

[14] Patel JS, Patel MM, Koringa PG, et al. Human leukocyte antigen alleles,genotypes and haplotypes frequencies in renal transplant donors andrecipients from West Central India. Indian J Hum Genet 2013;19:219–32.

[15] Lonjou C, Thomas L, Borot N, et al. A marker for Stevens–Johnson syndrome:ethnicity matters. Pharm J 2006;6:265–8.

[16] Locharernkul C, Loplumlert J, Limotai C, et al. Carbamazepine and phenytoininduced Stevens–Johnson syndrome is associated with HLA-B⁄1502 allele inThai population. Epilepsia 2008;49:2087–91.

[17] Man CB, Kwan P, Baum L, et al. Association between HLA-B⁄1502 allele andantiepileptic drug-induced cutaneous reactions in Han Chinese. Epilepsia2007;48:1015–8.

[18] Kaniwa N, Saito Y, Aihara M, et al. HLA-B⁄1511 is a risk factor forcarbamazepine-induced Stevens–Johnson syndrome and toxic epidermalnecrolysis in Japanese patients. Epilepsia 2010;51:2461–5.

[19] Tassaneeyakul W, Tiamkao S, Jantararoungtong T, et al. Association betweenHLA-B⁄1502 and carbamazepine-induced severe cutaneous adverse drugreactions in a Thai population. Epilepsia 2010;51:926–30.

[20] Daar AS, Singer PA. Pharmacogenetics and geographical ancestry: implicationsfor drug development and global health. Nat Rev Genet 2005;6:241–6.

[21] Tegretol [package insert]. East Hanover, New Jersey. Novartis PharmaceuticalsCorp; 2013.

[22] Ikeda H, Takahashi Y, Yamazaki E, et al. HLA class I markers in Japanesepatients with carbamazepine-induced cutaneous adverse reactions. Epilepsia2010;51:297–300.

[23] Wei CY, Chung WH, Huang HW, Chen YT, Hung SI. Direct interaction betweenHLA-B and carbamazepine activates T cells in patients with Stevens–Johnsonsyndrome. J Allergy Clin Immunol 2012;129:1562–9.

[24] Niihara H, Kakamu T, Fujita Y, Kaneko S, Morita E. HLA-A31 strongly associateswith carbamazepine-induced adverse drug reactions but not withcarbamazepine-induced lymphocyte proliferation in a Japanese population. JDermatol 2012;39:594–601.

[25] Gerber BO, Pichler WJ. Noncovalent interactions of drugs with immunereceptors may mediate drug-induced hypersensitivity reactions. AAPS J2006;17(8):E160–165.

with carbamazepine induced Stevens–Johnson syndrome in North Indian9.022