Results of Detailed Investigations Into Stevens-Johnson Syndrome With Severe Ocular Complications

Abstract

Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are acute inflammatory vesiculobullous reactions of the mucosa of the ocular surface, oral cavity, and genitals, and of the skin. Severe ocular complications (SOC) are present in about half of SJS/TEN patients diagnosed by dermatologists. We review our group's findings on the genetic predisposition for and the etiology of SJS/TEN with SOC. We suspected that abnormal innate mucosal immunity, resulting in an anomalous response to commensal bacteria that usually do not elicit such a response, contributes to the ocular surface inflammation seen in SJS/TEN with SOC. We found that cold medicines, including multi-ingredient cold medications and nonsteroidal anti-inflammatory drugs, were the main causative drugs especially in patients with SJS/TEN with SOC. Cold medicine–related SJS/TEN (CM-SJS/TEN) with SOC was strongly associated with HLA-A*02:06 in the Japanese populations, and significantly associated with HLA-B*44:03 in the Japanese and in Indian and Brazilian populations. Single nucleotide polymorphism association analysis showed that the Toll-like receptor 3 (TLR3), prostaglandin-E receptor 3 (PTGER3), and IKZF1 gene were significantly associated with CM-SJS/TEN with SOC and that they could regulate mucocutaneous inflammation including that of the ocular surface. As we found several HLA-SNP sets with a high odds ratio, we postulated that they may help to predict the possible development of SJS/TEN with SOC. From our findings we suggest that besides microbial infection and cold medicines, a combination of multiple gene polymorphisms and their interactions contribute strongly to the onset of CM-SJS/TEN with SOC.

Stevens-Johnson syndrome (SJS) is an acute inflammatory vesiculobullous reaction of the mucosa of the ocular surface, oral cavity, and genitals, and of the skin. In patients with extensive skin detachment and a poor prognosis, the condition is called toxic epidermal necrolysis (TEN). In the acute stage of SJS/TEN, approximately 50% of patients present with severe ocular lesions such as severe conjunctivitis with pseudomembrane and ocular surface epithelial defects.¹ 

Ophthalmologists encounter patients not only in the acute but also the chronic stage, while dermatologists tend to see SJS/TEN patients only in the acute stage. Our ophthalmologic diagnosis of SJS/TEN was based on a confirmed history of acute-onset high fever, serious mucocutaneous illness with skin eruptions, and involvement of at least two mucosal sites including the ocular.^2–8 SJS/TEN patients with severe ocular complications (SOC) in the acute stage often suffer severe ocular sequelae such as vision loss and very severe dry eye that prevent their having a normal life.⁹ 

We defined acute-stage SOC as a condition with severe conjunctivitis with pseudomembrane and epithelial defects on the ocular surface (cornea and/or conjunctiva),¹⁰ and chronic-stage as a condition with ocular sequelae such as severe dry eye, trichiasis, symblepharon, and conjunctival invasion into the cornea.⁹ Ophthalmologists tend to diagnose both SJS and TEN with SOC broadly as “Stevens-Johnson syndrome.”² In this review, I show the results of detailed investigations into SJS/TEN with SOC. 

Abnormal Innate Mucosal Immunity in SJS With SOC

We also observed that SJS/TEN with SOC presented with opportunistic infection of the ocular surface by bacteria, especially methicillin-resistant Staphylococcus aureus (MRSA) and methicillin-resistant Staphylococcus epidermis (MRSE)¹¹; the detection rate of MRSA and MRSE was higher on the ocular surface of SJS/TEN with SOC patients than in individuals with other devastating ocular surface disorders. Our SJS/TEN with SOC patients suffered persistent inflammation of the ocular surface even in the chronic stage and their inflammation was associated with colonization with MRSA and MRSE,² resulting in a hyperinflammatory reaction against commensal bacteria, and decolonization of these bacteria may reduce the inflammatory response. Under normal conditions, colonization with MRSA and MRSE does not elicit such ocular surface inflammation² (Fig. 1). Consequently, we suggested the possibility of an association between a disordered innate mucosal immune response and SJS/TEN with SOC.² 

Figure 1

 

Possibility of an association between a disordered innate mucosal immune response and SJS/TEN with SOC. The ocular surface inflammation of patients with SJS/TEN with SOC may be exacerbated by colonization with MRSA and MRSE, resulting in a hyperinflammatory reaction against commensal bacteria. Decolonization of these bacteria may reduce the inflammatory response

We postulated that a balance between innate mucosal immunity of the ocular surface and pathogenicity of bacteria is important and that when the host innate mucosal immunity is normal, commensal bacteria are in a symbiotic relationship with the host. On the other hand, in the absence of such immunity, commensal bacteria including MRSA and MRSE can become pathogenic.¹² Mucosa harboring commensal bacteria usually does not respond to commensal bacteria; in normal conditions mucosa has no inflammation despite harboring commensal bacteria, however, in patients with ocular-surface SJS/TEN with SOC whose innate mucosal immunity might be compromised, a response may be elicited by commensal bacteria including MRSA and MRSE² and may result in the ocular surface inflammation seen in patients with SJS/TEN with SOC.² 

Cold Medicine and HLA Analysis

While the reported annual incidence of SJS/TEN is only one to six cases per 10⁶ individuals,^1,13,14 its mortality rate is high (3% for SJS and 27% for TEN ).¹⁵ An association with inciting drugs has been documented.^{5,7,8,14,16–20} We reported that 80% of our SJS/TEN with SOC patients developed SJS/TEN within several days after taking cold medicines including multi-ingredient cold medications and nonsteroidal anti-inflammatory drugs (NSAIDs) to combat the common cold.^2,5,8,20 Our Brazilian collaborators also found that more than half of their SJS/TEN with SOC patients had taken cold medicines,²¹ suggesting that such medicines are major causative drugs for SJS/TEN with SOC. Our Thailand collaborators also reported that cold medicines are main causative drugs for SJS/TEN with severe chronic ocular complications: 49 of 71 patients with SJS/TEN with SOC (69%) had a history of taking CM before the onset of SJS/TEN in Thailand.²² Moreover, our Korean collaborators reported that NSAIDs, including cold medicines, might be associated with severe chronic ocular complications in Korean patients with SJS/TEN.²³ 

The pathobiological mechanisms underlying the onset of SJS/TEN with SOC have not been fully identified. The extreme rarity of cutaneous and ocular surface reactions to drug therapies led us to suspect individual susceptibility. Therefore, we analyzed the possible association between human leukocyte antigen (HLA) genotypes and cold medicine–related SJS/TEN (CM-SJS/TEN) with SOC. We found that in the Japanese it was strongly associated with HLA-A*02:06 (patients, n = 151; controls [normal], n = 639; odds ratio [OR] = 5.6 [95% CI, 3.8–8.3]; P = 2.7 × 10⁻²⁰) and significantly associated with HLA-B*44:03 (cases, n = 151; controls [normal], n = 639; OR = 2.0 [95% CI, 1.3–3.1]; P = 1.3 × 10⁻³).⁸ Interestingly, these HLA genotypes were not involved in CM-SJS/TEN without SOC,⁸ suggesting that the genetic predisposition such as the HLA genotype might be different in SJS/TEN patients with/without SOC.⁸ Moreover, HLA-A*02:06 and HLA-B*44:03 are not associated with cold medicine–unrelated (other medicine-related) SJS/TEN with SOC,⁸ suggesting that the genetic predisposition such as the HLA genotype is different among causative drugs, for example, between cold medicine and allopurinol or carbamazepine^5,19,24,25 (Fig. 2). 

Figure 2

 

The elicitation of SJS/TEN with SOC depends on a genetic predisposition and not all drugs are causative. Reproduced with permission from Ueta M, Sawai H, Sotozono C, et al. IKZF1, a new susceptibility gene for cold medicine-related Stevens-Johnson syndrome/toxic epidermal necrolysis with severe mucosal involvement. J Allergy Clin Immunol. 2015;135:1538–1545.e17. © 2015 American Academy of Allergy, Asthma & Immunology. Published by Elsevier, Inc.

We also analyzed the association between HLA and CM-SJS/TEN with SOC by using samples from individuals with ethnic backgrounds other than Japanese. We found that CM-SJS/TEN with SOC was significantly associated with HLA-B*44:03 in Indians (cases, n = 20; controls [normal], n = 55; OR = 12.3 [95% CI, 3.6–42.0]; P = 1.1 × 10⁻⁵) Thailand people (case n = 49; controls (normal), n = 119; OR = 7.2 (95% CI, 3.06–16.7); P = 5.5 × 10⁻⁶),²² Brazilians, especially Caucasian Brazilian patients (cases, n = 15; controls [normal], n = 62; OR = 6.2 [95% CI, 1.6–23.6]; P = 3.7 × 10⁻³). In ethnic Koreans HLA-A*02:06 was associated with CM-SJS/TEN with SOC (cases, n = 31; controls [normal], n = 90; OR = 3.0 [95% CI, 1.2–7.6]; P = 0.018).⁷ 

Dermatologists and others have reported that allopurinol (a uric acid–lowering drug) and anticonvulsants such as carbamazepine are the main inciting drugs for SJS/TEN.¹⁷ Allopurinol-induced SJS/TEN was strongly associated with HLA-B*58:01 in Han Chinese,²⁶ Caucasian,²⁷ and Japanese²⁸ patients. Carbamazepine-induced SJS/TEN manifested a very strong association with the HLA-B*15:02 allele in Taiwanese Han Chinese patients,²⁹ and the HLA-A*31:01 allele was strongly associated with carbamazepine-induced SJS/TEN in Japanese³⁰ and European³¹ patients. Interestingly, allopurinol can induce SJS/TEN without severe ocular complications,²⁵ and not all patients with carbamazepine-induced SJS/TEN have SOC.³² 

The first HLA analysis of SJS performed by ophthalmologists appeared in 1982.³³ It showed that the level of the HLA-Bw44 antigen, a subgroup of HLA-B12 (serotype), is significantly higher in Caucasians with SJS with ocular involvement than in a control Caucasian population. Dermatologists have also found that the frequency of the HLA-B12 antigen is significantly increased in French SJS/TEN patients whose disorder is clearly drug induced as compared to a French control population; the main causative agents are NSAIDs.³⁴ As the HLA-B*44:03 genotype is included in serotype HLA-B12, it is possible that HLA-B*44:03 is also associated with SJS/TEN with SOC in Caucasians. 

Genetic Predisposition for CM-SJS/TEN With SOC and the Role of Susceptibility Genes

While SJS/TEN with SOC can be induced by certain drugs, not all patients taking these drugs develop the condition. As the incidence of SJS/TEN with SOC is very low, we suspected a genetic predisposition³ and performed single nucleotide polymorphism (SNP) association analysis by using candidate genes associated with innate immunity. Subsequent genome-wide association study (GWAS) identified several susceptibility genes for SJS/TEN with SOC (Fig. 3). We also performed function analysis of some susceptibility genes with mouse models. 

Figure 3

 

Susceptibility genes for SJS/TEN with SOC in the Japanese. To elucidate the pathology, we performed genetic analysis and found that some genes are associated with SJS/TEN with SOC in the Japanese.

Toll-like Receptor-3 (TLR3)

Among TLRs 1–10, TLR3 is expressed most highly in the ocular surface epithelium and more intensely than in mononuclear cells.^9,35 It recognizes dsRNA, a component of the life cycle of most viruses mimicking polyI:C, and it can induce proinflammatory cytokines and IFN-β on the ocular surface.^35–37 Therefore, we first focused on the TLR3 gene. We examined 17 SNPs of TLR3 and found that seven were significantly associated with SJS/TEN with SOC. Next we focused on CM-SJS/TEN with SOC and found that five SNPs (rs4861699, rs6822014, rs3775296, rs5743312, rs3775290) exhibited a significant association.³⁸ 

We also studied a murine model of experimental allergic conjunctivitis to further address the issue of ocular surface inflammation.³⁹ We examined TLR3 gene function in tlr3 knock-out (KO) and tlr3 transgenic (TLR3Tg) mice. Ocular surface inflammation was significantly reduced in tlr3 KO and significantly increased in tlr3 Tg mice.³⁹ Our study of murine contact dermatitis and atopic dermatitis models revealed that, as in the experimental allergic conjunctivitis model, skin inflammation was significantly reduced in tlr3 KO and significantly increased in tlr3 Tg mice.^40,41 We also found that TLR3 was strongly expressed in the skin epidermis. These findings indicate that TLR3 could positively regulate mucocutaneous inflammation of the skin and ocular surface (Fig. 4) and suggest that it might contribute to mucocutaneous inflammation in patients with SJS/TEN with SOC. 

Figure 4

 

Positive regulation of mucocutaneous inflammation by TLR3. (a) TLR3 is one of the Toll-like receptors (TLRs) that are very important in innate immunity. (b) TLR3 positively regulates ocular surface inflammation. Reproduced with permission from Ueta M, Uematsu S, Akira S, Kinoshita S. Toll-like receptor 3 enhances late-phase reaction of experimental allergic conjunctivitis. J Allergy Clin Immunol. 2009;123:1187–1189. © 2009 American Academy of Allergy, Asthma & Immunology. Published by Elsevier, Inc. (c) TLR3 also positively regulates skin inflammation. Reproduced with permission from Nakamura N, Tamagawa-Mineoka R, Ueta M, Kinoshita S, Katoh N. Toll-like receptor 3 increases allergic and irritant contact dermatitis. J Invest Dermatol. 2015;135:411–417. © 2015 The Society for Investigative Dermatology, Inc. Published by Elsevier, Inc.

Prostaglandin E Receptor-3 (PTGER3)

Our GWAS using the Affymetrix GeneChip mapping 500 k array set²⁰ (Affymetrix, Santa Clara, CA, USA) revealed a weak association between the PTGER3 gene and SJS/TEN with SOC. Analysis of 38 SNPs of the PTGER3 gene with the DigiTag2 assay identified 20 SNPs associated with SJS/TEN with SOC.⁴² We found that the association with the PTGER3 gene was stronger in CM-SJS/TEN with SOC than in SJS/TEN with SOC. PTGER3 SNP rs1327464 (G versus A) was most significantly associated with CM-SJS/TEN with SOC; the OR for the major allele was 0.232 (95% CI, 0.1–0.4) (P = 7.92 × 10⁻¹⁰).⁴³ An association was also found in Korean patients.⁴³ 

EP3 is the protein of the PTGER3 gene; it is one of four receptors (EP1, EP2, EP3, and EP4) of prostaglandin E₂. Cold medicines such as NSAIDs (e.g., ibuprofen and loxoprofen) and cold medicine ingredients (e.g., acetaminophen) suppress the production of prostanoids including PGE₂ (Fig. 5a). PGE₂ acts on EP3 in the airway epithelium and negatively regulates inflammation in a murine asthma model.⁴⁴ Elsewhere we reported that EP3 also negatively regulated ocular surface inflammation in a murine allergic conjunctivitis⁴⁵ and skin inflammation in a murine contact dermatitis model.⁴⁶ We suggested that the suppression of PGE₂ production by cold medicines might contribute to the onset and pathogenesis of CM-SJS/TEN with SOC^2,20 because PGE₂ acts on EP3 and negatively regulates mucocutaneous inflammation^44–46 (Fig. 5b). 

Figure 5

 

Negative regulation of mucocutaneous inflammation by EP3. (a) EP3 is one of the prostaglandin (PG) E₂ receptors. Cold medicines such as acetaminophen and NSAIDs suppress prostanoids including PGE₂. (b) PGE₂ acts on EP3 on the ocular surface epithelium and negatively regulates ocular surface inflammation. Reproduced with permission from Ueta M, Matsuoka T, Narumiya S, Kinoshita S. Prostaglandin E receptor subtype EP3 in conjunctival epithelium regulates late-phase reaction of experimental allergic conjunctivitis. J Allergy Clin Immunol. 2009;123:466–471. © 2009 American Academy of Allergy, Asthma & Immunology. Published by Elsevier, Inc. (c) EP3 protein expression is strongly downregulated in the conjunctival epithelium of SJS/TEN with SOC patients. Reproduced from Ueta M, Sotozono C, Yokoi N, Inatomi T, Kinoshita S. Prostaglandin E receptor subtype EP3 expression in human conjunctival epithelium and its changes in various ocular surface disorders. PLoS One. 2011;6:e25209. © 2011 Ueta et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License.

We then examined EP3 protein expression on the human ocular surface and found that it was much lower in the conjunctival epithelium of patients with SJS/TEN with SOC than in our control subjects composed of patients with conjunctival chalasis and chemical burns.⁴⁷ These findings led us to suggest that the expression of EP3 is strongly downregulated in the ocular surface of patients with SJS/TEN with SOC⁴⁷ (Fig. 5c) and it might contribute to the ocular surface inflammation of SJS/TEN with SOC. 

Thus we reported that TLR3 could upregulate and that PTGER3 could suppress ocular surface inflammation.^39,45 Considering the opposite roles of the PTGER3 and the TLR3 gene in ocular surface inflammation, we posited an unknown functional interaction between EP3 and TLR3. To investigate this issue we prepared tlr3/Ptger3 double knock-out (DKO) mice, and compared them with wild-type mice and tlr3-KO mice. We found that in the allergic conjunctivitis model, ocular surface inflammation in tlr3/Ptger3-DKO mice was at a level similar to that in tlr3-KO mice and significantly lower than in wild-type mice. On the other hand, upregulation was significantly greater in Ptger3^−/− than wild-type mice, suggesting that EP3 negatively regulates TLR3-dependent ocular surface inflammation.⁴² 

PolyI:C is a TLR3 ligand. In human conjunctival epithelial cells, an EP3 agonist suppressed the production of polyI:C-induced inflammatory cytokines and the mRNA expression, for example, TSLP,⁴² RANTES, IP-10,⁴⁸ and MCP-1.⁴⁹ 

As we found that patients presenting with CM-SJS/TEN with SOC had taken cold medicines to combat the common cold due to viral or mycoplasma infection, we think that not only cold medicines but also some microbial infectious agents such as viruses or mycoplasma may play a necessary trigger role in the onset of SJS/TEN with SOC.² 

IKZF1

We included 117 Japanese CM-SJS/TEN with SOC patients and 691 Japanese controls in our second GWAS that used the Affymetrix AXIOM genome-wide ASI-1 array.⁵ As in our earlier studies, the GWAS Manhattan plot indicated that the HLA-A region exhibited the strongest association with susceptibility to CM-SJS/TEN with SOC.⁵ Moreover, we found that the IKZF1 gene was strongly associated with CM-SJS/TEN with SOC in Japanese patients (Fig. 6a).⁵ When we subjected all Japanese samples (149 patients, 877 controls) and samples from 31 Korean patients and 90 controls, 20 Indian patients and 56 controls, and 39 Brazilian patients and 135 controls to genotyping of SNPs of the IKZF1 gene, we found that meta-analysis of these ethnic groups revealed a significant genome-wide association between CM-SJS/TEN with SOC and IKZF1 (rs4917014 [G versus T], OR = 0.5 [95% CI, 0.4–0.6], P = 8.5 × 10⁻¹¹).⁵ This might suggest IKZF1 as a universal marker for susceptibility to CM-SJS/TEN with SOC.⁵ 

Figure 6

 

Association between IKZF1 gene SNPs and SJS/TEN with SOC. (a) Manhattan plot of our GWAS (117 Japanese CM-SJS/TEN with SOC patients and 691 controls). Reproduced with permission from Ueta M, Sawai H, Sotozono C, et al. IKZF1, a new susceptibility gene for cold medicine-related Stevens-Johnson syndrome/toxic epidermal necrolysis with severe mucosal involvement. J Allergy Clin Immunol. 2015;135:1538–1545.e17. © 2015 American Academy of Allergy, Asthma & Immunology. Published by Elsevier, Inc. (b) Ikzf1Tg mice developed dermatitis and blepharoconjunctivitis. Reproduced with permission from Ueta M, Hamuro J, Nishigaki H, et al. Mucocutaneous inflammation in the Ikaros Family Zinc Finger 1-keratin 5-specific transgenic mice. Allergy. 2018;73:395–404. © 2017 EAACI and John Wiley and Sons A/S. Published by John Wiley and Sons Ltd.

Ikaros, a transcription factor that regulates numerous biological events, is thought to regulate important cell-fate decisions in the development of the adaptive immune system. However, its biological and pathologic functions in mucosal immunity are not clearly understood. Functional analysis of SNPs of the IKZF1 gene revealed that the ratio of the splicing isoforms Ik2/Ik1 may be affected by IKZF1 SNPs significantly associated with susceptibility to CM-SJS/TEN with SOC.⁵ The quantity of the Ik2 isoform is increased in disease-protective genotypes of IKZF1 (rs4917014 G/G and rs10276619 A/A). As Ikaros 2, an Ik2 isoform, lacks DNA-binding ability and seems to be dominant-negative, it is possible that the function of Ikaros, the protein of IKZF1, is enhanced in CM-SJS/TEN with SOC.⁵ 

Elsewhere we suggested that epithelium might contribute to the pathobiology of CM-SJS/TEN with SOC.² We based our suggestion on the observation that TLR3 was strongly expressed on ocular surface epithelial cells^35,37,50 and keratinocytes,⁴¹ and that it regulated ocular surface inflammation³⁹ and dermatitis.⁴¹ EP3 was dominantly expressed on ocular surface epithelium,⁴⁵ the epidermis,⁴⁶ and the airway epithelium.⁴⁴ EP3 on the epithelium has been shown to negatively regulate ocular surface inflammation,⁴⁵ dermatitis,⁴⁶ and airway inflammation.⁴⁴ 

In epidermal keratinocytes and conjunctival epithelial cells, polyI:C, a TLR3 ligand, upregulated the expression of IKZF1 mRNA.⁵¹ We produced K5-Ikzf1-EGFP transgenic (Ikzf1-Tg) mice by introducing the Ik1 isoform into cells expressing keratin 5, which is expressed in epithelial tissues such as the epidermis and conjunctiva.⁵¹ Mucocutaneous inflammation was exacerbated in these mice; they developed dermatitis; some developed blepharoconjunctivitis⁵¹ (Fig. 6b). Histologic analysis showed not only dermatitis but also tissue inflammation in the blepharoconjunctiva, tongue, and paronychia⁵¹ as with patients in the acute state of SJS/TEN with SOC.² 

Our findings suggest that IKZF1 plays a critical role in maintaining mucocutaneous homeostasis² and that it might be implicated in the aggravation of mucocutaneous inflammation seen in the presence of CM-SJS/TEN with SOC.² 

Interaction Between the HLA Genotype and Susceptibility Genes

SNPs are widely used as genetic markers for identifying human disease-susceptibility genes. As gene-gene interactions are more meaningful than major single-locus effects, we investigated the gene-gene interactions involved in the genetic predisposition for CM-SJS/TEN with SOC. We found that in CM-SJS/TEN with SOC there were more than additive effects with respect to HLA-A*02:06 and TLR3 SNPs; HLA-A*02:06 and TLR3 SNP rs3775296T/T (which was in strong linkage disequilibrium with TLR3 SNP rs5743312T/T). Of 133 Japanese patients, 10 (7.5%), but none of the controls, manifested both HLA-A*02:06 and TLR3 SNP rs3775296T/T (P < 0.00005, OR = 37.6 [95% CI, 2.2–647.7] and P < 0.00005, OR = 37.5, respectively, Woolf correction).³⁸ 

TLR3 recognizes viral double-stranded RNA and HLA-A is a component of HLA class I, which resides on the surface of all nucleated cells and alerts the immune system that the cell may be infected by a virus, thereby targeting the cell for destruction. Since the onset of CM-SJS/TEN with SOC was associated not only with certain drugs but also with putative microbial infection, multiplicative interactions between HLA-A and the TLR3 gene might contribute to the characteristic and pathogenic immune response to microbial infection.^2,38,52 

We also detected an interaction with additive effects between HLA-A*02:06 and the high-risk genotypes PTGER3 rs1327464 (GA or AA); the combination of PTGER3 rs1327464 (GA/AA) and HLA-A*02:06 had a higher OR (10.8) (95% CI, 3.6–32.1) than each allele alone (P = 2.56 × 10⁻⁷); In the Japanese patients, the OR for PTGER3 rs1327464 GA/AA alone was 4.48 (P = 1.90 × 10⁻⁸); for HLA-A*02:06 alone it was 5.46 (P = 1.39 × 10⁻¹¹).⁴³ In the Korean patients also we found an additive effect between SNPs of the PTGER3 gene and HLA-A*02:06; the combination of PTGER3 rs1327464 (GA/AA) and HLA-A*02:06 exhibited a higher OR (14.2 [95% CI, 3.5–57.7]) than each allele alone (P = 5.58 × 10⁻⁶). The OR for PTGER3 rs1327464 GA/AA alone was 4.07 (P = 0.00101) and for HLA-A*02:06 alone it was 2.50 (P = 0.0412).⁴³ 

These findings suggest that the interaction between HLA-A*02:06 and PTGER3 SNPs might also contribute to the onset of CM-SJS/TEN with SOC.⁴³ 

HLA-A*02:06 with Rec114 rs16957893 CG also exerted more than additive effects in CM-SJS/TEN with SOC (OR = 110 [95% CI, 6.36–1905], P = 4.45 × 10⁻⁸).⁶ This interaction was independent of the interaction between HLA-A*02:06 and TLR3 rs3775296 T/T because Japanese CM-SJS/TEN with SOC patients with both HLA-A*02:06 and Rec114 rs16957893 CG did not harbor TLR3 rs3775296 T/T.⁶ 

In the Japanese, combined genotyping for the associated variants thus far identified (i.e., HLA-A*02:06, Rec114 rs16957893 CG, TLR3 rs3775296 T/T, and PTGER3 rs1327464 GA or AA) may help to predict the risk for CM-SJS/TEN with SOC. 

Strategies for the Future

Since CM-SJS/TEN with SOC is a rare condition with a complex genetic background, it is reasonable to posit the presence of multiplicative interactions between HLA-A and TLR3,^38,52 HLA-A and REC14,⁶ and HLA-A and PTGER3.⁴³ It is possible that multiple susceptibility genes for CM-SJS/TEN with SOC are involved in forming functional networks. We documented that EP3 suppresses the TLR3 ligand and PolyI:C-induced cytokine production,^48,49,53,54 and that IKZF1 was upregulated by the TLR3 ligand polyI:C.⁵¹ An imbalance in these genes may trigger the mucocutaneous inflammation seen in patients with CM-SJS/TEN with SOC (Fig. 7a). 

Figure 7

 

Etiologic mechanisms underlying the development of SJS/TEN with SOC proposed by the author. (a) Multiple susceptibility genes for CM-SJS/TEN with SOC might form functional networks. (b) When the network balance is good, homeostasis is maintained. In the presence of multiple susceptibility gene polymorphisms the network is unbalanced and homeostasis is not maintained. This may lead to the development of CM-SJS/TEN with SOC.

As multiple susceptibility genes for CM-SJS/TEN with SOC may constitute a network, when the network balance is good, homeostasis is maintained. However, when multiple susceptibility genes harbor polymorphisms and the network is out of balance, homeostasis is not maintained, resulting in the development of SJS/TEN with SOC (Fig. 7b). 

As our investigations identified several SNP sets with a high OR, their use may help to alert to the possibility of SJS onset. Our findings reported here suggest that in addition to microbial infections and certain cold medicines, the combination of multiple gene polymorphisms and their interactions contribute strongly to the onset of CM-SJS/TEN with SOC. Abnormal innate immunity might strongly contribute the etiology of SJS/TEN with SOC. 

Acknowledgments

Supported by grants-in-aid from the Ministry of Education, Culture, Sports, Science and Technology of the Japanese government (BioBank Japan Project), by the JSPS Core-to-Core Program, A. Advanced Research Networks, partly supported by grants-in-aid for scientific research from the Japanese Ministry of Health, Labor and Welfare. Funding of the publication fee and administration was provided by the Dry Eye Society, Tokyo, Japan. The Dry Eye Society had no role in the contents or writing of the manuscript. 

Disclosure: M. Ueta, None