Front. Med. Frontiers in Medicine Front. Med. 2296-858X Frontiers Media S.A. 10.3389/fmed.2021.644871 Medicine Mini Review Associations Between Stevens–Johnson Syndrome and Infection: Overview of Pharmacoepidemiological Studies Imatoh Takuya 1 2 * Saito Yoshiro 2 1Division of Cohort Research, Center for Public Health Sciences, National Cancer Center, Tokyo, Japan 2Division of Medicinal Safety Science, National Institute of Health Sciences, Kanagawa, Japan

Edited by: Mayumi Ueta, Kyoto Prefectural University of Medicine, Japan

Reviewed by: Yoshito Zamami, Tokushima University, Japan; Jaya P. Gnana-Prakasam, Saint Louis University, United States

 *Correspondence: Takuya Imatoh timato@ncc.go.jp

This article was submitted to Ophthalmology, a section of the journal Frontiers in Medicine

 26 03 2021 2021 8 644871 22 12 2020 04 03 2021 Copyright © 2021 Imatoh and Saito. 2021 Imatoh and Saito

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

Stevens–Johnson syndrome and toxic epidermal necrolysis (SJS/TEN) are classified as type B adverse drug reactions, and are severe, potentially fatal rare disorders. However, the pathogenesis of SJS/TEN is not fully understood. The onset of SJS/TEN is triggered by the immune system in response to antigens with or by drugs. As activation of the immune system is important, infection could be a risk factor for the onset of SJS/TEN. Based on the hypothesis that infections induce the onset of SJS/TEN, we conducted pharmacoepidemiological investigations using two spontaneous adverse drug reaction reporting databases (Japanese Adverse Drug Event Report database and Food and Drug Administration Adverse Event Reporting System) and Japanese medical information database. These data suggest that infection could be a risk factor for the development of SJS/TEN. In this mini-review, we discuss the association between infection and the development of SJS/TEN.

 Stevens–Jonhson syndrome toxic epidermal necrolysis infection pharmacoepidaemiology real world evidence 19mk0101085j0503 Japan Agency for Medical Research and Development10.13039/100009619 Introduction

Stevens–Johnson syndrome (SJS) and the related disease toxic epidermal necrolysis (TEN) are two of the most serious adverse reactions caused by drugs. SJS is characterized by high fever, malaise, and a rapidly developing, blistering exanthema of macular papules and target-like lesions, accompanied by mucosal involvement. According to the World Allergy Organizations (WAO) definition of 2014, SJS/TEN are defined by the degree of skin detachment: SJS/TEN are defined as skin involvement of <10 and >30%, respectively, and that of SJS/TEN overlap as 10–30% skin involvement.

SJS/TEN are known as idiosyncratic drug reactions. They are rare and unpredictable. Therefore, serious idiosyncratic drug reaction is one of the cause of withdrawal of drug from the market. SJS/TEN are rare but life-threatening severe cutaneous adverse reactions, which are mainly induced by a variety of drugs. SJS/TEN occur in all ages, races, and sexes (1), with incidences ranging from 0.4 to 1.2 and 1.2 to 6 per million person-years. A USA-based study analyzing nationwide inpatient records from 2009 to 2012 calculated an incidence per million inhabitants of 8.61–9.69 for SJS, 1.46–1.84 for SJS/TEN overlap, and 1.58–2.26 for TEN (2). Epidemiological studies have calculated a TEN incidence rate of 1.2 per 1 million inhabitants per year for France (3) and of 0.93 per 1 million inhabitants per year for Germany (4). However, Frey et al. recently estimated that Asian patients were at a two-fold higher risk of SJS/TEN when compared with Caucasian patients (5). The incidence of SJS ranged from 3.3–4.1 per million people per year in Taiwan between 2000 and 2008 (6). Approximately 800–1,000 cases of SJS and 500–700 cases of TEN are reported annually in Japan (7).

Some drugs have been associated with the onset of SJS/TEN. Phenytoin, carbamazepine, and phenobarbital have been reported as common causes of SJS in a previous case control study (8). Additionally, sulfamethoxazol/trimethoprim, sulfonamides (sulfasalazine, sulfadiazine, sulfadoxine, and sulfafurazole) and oxicam-type non-steroidal anti-inflammatory drugs (NSAIDs) (meloxicam, piroxicam, and tenoxicam) are known to have a high risk of inducing SJS/TEN. A large proportion of suspected drug is comprised cold medicines. In particular, SJS/TEN with severe ocular complications (SOC) associated with cold medicine were reported in several studies (9, 10). In addition to conventional drugs, herbal remedies and new biologicals have been considered as causative agents. Recently, lamotrigine and zonisamide have been reported as major causative drugs of SJS/TEN (7).

The pathogenesis of SJS/TEN is not yet fully understood, but it is thought to be immune-mediated. The role of the T-cell-mediated immune response in the pathogenesis of SJS/TEN has been firmly established. In particular, the human leukocyte antigen (HLA) system plays an important role in the pathogenesis of SJS/TEN, as some drugs may bind directly to the HLA-complex and cause self-reactivity due to the drug-modified HLA-peptide repertoire (11). There are numerous reported associations between some HLA alleles and SJS/TEN (12–14).

In this mini-review, we have focused to analyze on infections, that is, the role of the immune response to viruses and bacteria in SJS/TEN. Previous studies have reported associations between some infections and the onset of SJS/TEN. However, most were clinical studies and case reports (15–19) with very few pharmacoepidemiological studies. Therefore, we conducted a pharmacoepidemiological study to elucidate the association between infection and SJS/TEN using the Japanese Adverse Drug Event Report (JADER) database. We also used the Food and Drug Administration (FDA) Adverse Events Reporting System (FAERS) database to compare these associations in the US, Europe, Japan, and other countries. Based on these findings, we discuss the association between infection and the development of SJS/TEN.

Results of a Pharmacoepidemiological Study Using the JADER Database

First, to elucidate the association between infection and SJS/TEN, we conducted a pharmacoepidemiological study using the JADER database (20).

JADER is a large published spontaneous reporting database for drug adverse reactions that was established by Japan's Pharmaceuticals and Medical Device Agency (PMDA) for pharmacovigilance activities. The dataset can be accessed directly on: http://www.info.pmda.go.jp/fukusayoudb/CsvDownload.jsp (available in Japanese only), and compose of four relational tables: DEMO, DRUG, REAC, and HIST. The tables contain the following information: demographic information of patients including, sex and age (DEMO table), drug information used in patients (DRUG table), adverse drug information that have occurred in patients (REAC table) and primary disease information (HIST table). Each table was connected using the ID number of each recorded case.

We used the November 2014 version of JADER and extracted the records of 177,659 cases of ADRs reported from 2009–2013. The reported cases were classified into three categories (anti-infectious drug group, concomitant infection group, and non-infection group) based on the presence of anti-infectious drugs (either as primary suspected drug or concomitant drug) and infectious disease. We defined the cases whose clinical outcome was “death,” “unrecovered,” or “with sequela” as severe. We assessed the association between SJS/TEN and the presence or seriousness of infection using logistic regression analysis. Logistic regression analysis showed a significant positive association between the infection status and onset of SJS/TEN compared to the non-infection group. Odds ratios (OR) were 2.04 [95% confidence interval (CI) (1.85–2.24)] for the anti-infectious drug group and 2.44 [95% CI (2.21–2.69)] for the concomitant infection group, respectively. Moreover, a significantly positive association between infection and SJS/TEN severity was observed [OR 1.48; 95% CI (1.10–1.98)]. These results suggest that SJS/TEN is strongly associated with infection in this Japanese dataset.

 Results of a Pharmacoepidemiological Study Using a Japanese Medical Information Database

Spontaneous drug adverse reaction databases have considerable biases. Limited information is available in standardized spontaneous reports (21) and only a fraction of adverse drug events are identified and reported (22). There is a growing need for real-world data in medical and healthcare research. Big data in healthcare may be comprised of massive amounts of information from various sources, including electronic health records (EHRs), administrative or claims data, and data from self-monitoring devices. The utilization of electronic medical record databases for drug safety assessment has been extensively discussed (23–25). In Japan, a new medical information database network (designated MID-NET®) to provide real-world data for drug safety assessments was officially launched in April 2018 (26). This network was designed and developed by the Ministry of Health, Labor and Welfare, and the PMDA in collaboration with 23 hospitals of 10 healthcare organizations across Japan. Thus, real-world data is an important source of information for the detection of adverse drug reactions.

We used a medical information database in Japan to confirm the association between infection and the onset of SJS/TEN (27). We used commercial medical information data collected by Medical Data Vision (MDV Co., Ltd). The medical information database covers more than 30 million patients in more than 400 Japanese hospitals and includes health claims, pharmacy claims and the diagnosis procedure combination (DPC) data.

Since antipyretic analgesics, including acetaminophen and loxoprofen sodium, are widely prescribed and are known as major suspected drugs for SJS/TEN (28), we conducted a matched nested case-control study to elucidate the association between concurrent infection and the onset of SJS/TEN in patients prescribed antipyretic analgesics. We used the data of 4,112,055 patients prescribing antipyretic analgesics between January 2014 and December 2015. The presence of International Classification of Diseases, 10th revision (ICD-10) codes for SJS/TEN (L511, L512) was defined as SJS/TEN. 553 (0.01%) were diagnosed with SJS/TEN. Among them, 131 patients who had been prescribed antipyretic analgesics during the month prior to the defined date of SJS/TEN onset were identified. Moreover, to minimize the impact of antibiotics agents, we excluded the patients who had been prescribed top 10 suspected antibiotics and viral and fungal agents (levofloxacin, clarithromycin, amoxicillin, galenoxacin, sefkapen, sulfamethoxazole/trimethoprim, azithromycin, ceftriaxone, cefditoren and vancomycin) during the month prior to the index date.

In a matched nested case-control study, for each case, three controls were randomly matched with the case for age at index date (onset of SJS/TEN) and sex. Infection was defined the diagnosis of infection (bacterial, viral, and fungal infection) during the month prior to the index date, which co-existed with one or more prescription claims for antibiotic agents (antibiotics-bacterial, virus, and fungal infection agents) in the same month. We calculated odds ratios and 95% CIs using conditional logistic regression to estimate the association between SJS/TEN and infection on patients prescribing antipyretic analgesics.

According to conditional logistic regression analysis, patients with infection and prescribing antipyretic analgesics had significantly increased the risk of onset of SJS/TEN [adjusted OR 5.59, 95% CI (2.01–15.51)] (Table 1). Similar to the results of our study using the JADER, It suggests that infection may increase the risk for the onset of SJS/TEN in patients taking antipyretic analgesics.

Odds ratios for the combination between antipyretic analgesics and infection in matched nested case control analysis of SJS/TEN.

Cases Controls Crude Adjusted*
n % n % OR 95% CI OR 95% CI
Antipyretic analgesics/Infection − − 42 32.1 183 46.6 1.00 (Reference) 1.00 (Reference)
− + 18 13.7 58 14.8 1.61 (1.03–2.52) 1.13 (0.43–2.95)
+ − 41 31.3 119 30.3 1.76 (0.75–4.11) 1.61 (1.00–2.61)
+ + 30 22.9 33 8.4 7.42 (3.10–17.79) 5.59 (2.01–15.51)
CCI Low (0–1) 77 58.8 294 74.8 1.00 (Reference) 1.00 (Reference)
Medium (1–3) 37 28.2 68 17.3 2.22 (1.35–3.67) 1.55 (0.88–2.75)
High (3+) 17 13.0 31 7.9 2.17 (1.14–4.13) 1.14 (0.42–2.49)
Corticosteroid use No 65 49.6 308 78.4 1.00 (Reference) 1.00 (Reference)
Yes 65 49.6 60 15.3 6.30 (3.80–10.44) 5.46 (3.20–9.32)

OR, odds ratio; 95% CI, 95% confidence interval; CCI, Charlson comorbidity index.

Adjusted for Charlson comorbidity index and corticosteroid use.

Reprinted with permission from (25). Copyright (2020), The American College of Clinical Pharmacology.

 Results of a Pharmacoepidemiological Study Using FAERS

Based on both JADER and Japanese medical information databases, an association was found between infection and the onset of SJS/TEN. We next conducted a pharmacoepidemiological study using FAERS to elucidate the association between infection and the onset of SJS/TEN in other countries.

The FAERS database contains >14 million adverse event reports, medication error reports and product quality complaints resulting in adverse events that were submitted to the FDA in the USA and other countries and has increased to over 1.8 million reports per year. Case report information is separated into seven tables, which contain details on patient demographics and administrative information (DEMO), medication and biological products used (DRUG), adverse drug reactions (REAC), patient outcomes (OUTC), report sources (RPSR), drug therapy start and end dates (THER), and indication (INDI). The database is made publicly available as a quarterly download on the FDA website.

We obtained data from case reports received by the FDA between January 1, 2007 and June 30, 2017. Duplicate reports were removed, and drug names were modified based on modified World Health Organization (WHO) drug dictionaries. To identify infections, we used the DRUG and INDI files, which include two drug variables and an indications variable, defined as follows: (1) DRUG_CODE: code of medicinal product; (2) ROLE_COD: code for drug's reported role in event, and (3) INDI_PT: PT code of indications. We categorized infection into three categories, defined as follows: (1) primary suspected anti-infectious drug group, (2) concomitant infection group, and (3) non-infection group. We identified individual SJS/TEN based on the MedDRA (version 20.1) preferred terms in the FAERS database. The ORs and 95% CIs were calculated using logistic regression analysis.

We observed significant associations between infections and SJS/TEN in the US and England (Table 2). In particular, a strong association was observed in the US. [US: OR for primary suspected anti-infectious drug group 8.21, 95% CI (7.60–8.87), OR for concomitant infection group 3.69, 95% CI (3.32–4.10); England: OR for primary suspected anti-infectious drug group 3.30, 95% CI (2.82–3.87), OR for concomitant infection group 3.69, 95% CI (2.82–3.98)]. However, there was no association between infection and SJS/TEN in China. [OR for primary suspected anti-infectious drug group 0.89, 95% CI (0.53–1.48), OR for concomitant infection group 1.13, 95% CI (0.59–2.18)].

Odds ratios for the association between infection and SJS/TEN in the FAERS database.

Cases Non-cases Odds ratio (95% CI)
Japan Anti-infectious drug 403 27,513 2.05 (1.81–2.33)
Concomitant infection 361 23,517 2.04 (1.80–2.32)
Non-infection 1,146 177,082 1.00 (Reference)
The United States Anti-infectious drug 1,302 203,782 8.21 (7.60–8.87)
Concomitant infection 526 182,813 3.69 (3.32–4.10)
Non-infection 4,134 4,617,902 1.00 (Reference)
England Anti-infectious drug 268 28,227 3.30 (2.82–3.87)
Concomitant infection 215 21,021 3.35 (2.82–3.98)
Non-infection 599 230,352 1.00 (Reference)
China Anti-infectious drug 21 5,125 0.89 (0.53–1.48)
Concomitant infection 13 2,019 1.13 (0.59–2.18)
Non-infection 121 30,864 1.00 (Reference)
All countries (except Japan) Anti-infectious drug 3,430 383,970 6.92 (6.59–7.28)
Concomitant infection 2,129 318,665 4.73 (4.66–5.24)
Non-infection 8,832 6,535,291 1.00 (Reference)

Adjusted for age (by 20-year-old class) and sex.

95% CI, 95% confidence interval.

 Discussion

We consulted three medical databases JADER, a medical information database in Japan, and FAERS, to analyze the association between infection and SJS/TEN. Based on the available information, there appears to be a strong statistical association between infection and the onset of SJS/TEN, in Western countries, as well as in Japan.

Cutaneous diseases, including drug reactions, are increased up to 100-fold in persons with human immunodeficiency virus (HIV) (29). Infectious agents, such as a variant of Coxsackie virus A6 and Mycoplasma pneumoniae, have been previously reported to cause SJS/TEN (16, 30, 31). Furthermore, 56% of SJS patients had an antecedent upper respiratory tract infection or non-specific viral infection (19, 32). Recently, a study using microbiological analysis reported that a higher proportion of pathogenic microorganisms, including Pseudomonas spp., Staphylococcus spp., Streptococcus spp., and Acinetobacter spp., was observed in the SJS patient group (33). Meanwhile, several pharmacoepidemiological studies have reported that some classes of antibiotics are associated with the occurrence of SJS/TEN (3, 5, 34). Previous studies that used EHRs have reported that more than half of the SJS/TEN cases are attributed to antibiotics (35).

We found that the association between infection and SJS/TEN differed between countries. The frequency of HLA alleles differs among ethnicities. For example, HLA-B*1502 is common in East Asians (6.9%) (36), and the FDA recommends screening this allele in individuals of Asian ancestry before initiating treatment with carbamazepine.

SJS/TEN is thought to be a type IV hypersensitivity reaction, in which a drug or its metabolite stimulates cytotoxic T-cells and helper T cells. The specific molecular mechanisms driving this reaction have only been elucidated in a handful of cases, and the pathogenesis of SJS/TEN is not completely clear. Numerous studies have reported that HLA and T-cell receptors play an important role in the immune mechanisms of SJS/TEN. Several concepts have been proposed to explain the pathogenesis of severe cutaneous adverse drug reactions. In the hapten concept, small molecules called haptens elicit an immune response only when attached to proteins. The “p-i” concept postulates that the suspected drugs can stimulate cells by binding directly and reversibly to immune receptors. The altered peptide repertoire model suggests that drugs alter the antigen by binding to the human leukocyte antigen pocket. Recently, White et al. proposed that some T-cell mediated hypersensitivity reactions likely represent another example of heterologous immunity (37). Another possibility is that innate immunity-related molecules such as toll-like receptors might be activated by the infections and their involvement of specific immune reactions to the drug for SJS/TEN onset. The pathogenesis of SJS/TEN clearly requires further investigation.

There are many case reports and small sample-size clinical studies regarding the association between infection and SJS/TEN. Epidemiological studies with larger sample sizes are extremely rare. Although large-scale epidemiological studies have been performed in the United States, Europe, and Korea (5, 38, 39), little has been reported on the association between infection and SJS/TEN. We conducted large-scale epidemiological studies using Japanese medical data from spontaneous reporting adverse databases and medical information databases. While JADER or FAERS are useful to investigate the risk factors for adverse drug reactions, as the incidences of SJS and TEN are very rare, these databases have some inherent limitations. These limitations include underreporting, biased reporting rates, incomplete patient information, and indeterminate population exposure, and create the need for complementary data sources and methods (40). There are two major limitations that are unique to the FAERS database. First, the drug nomenclature consists of medication brands and trade names, generic or non-proprietary names, and chemical or scientific names with frequent misspellings. Second, there are a lot of duplicates and multiple follow-up reports (41). Consequently, the interpretation of these findings requires careful consideration.

We also conducted an epidemiological study using a Japanese medical information database. The limitations of real-world evidence studies can include low internal validity, lack of quality control surrounding data collection, and susceptibility to multiple sources of bias for comparing outcomes. As these database were collected retrospectively, there may be some potential confounding factors that were not analyzed. In our medical information database study, even though the patients prescribing top 10 suspected antibiotics and viral and fungal agents in SJS/TEN were excluded, the significant association remained. However, it was not possible to completely distinguish whether the associations were caused by the infection, or were a direct consequence of the antibiotic agents. Hence, additional evidence is needed to elucidate the association between infection and SJS/TEN.

In summary, we reviewed our pharmacoepidemiological studies on the association between infection and SJS/TEN. These studies indicated that there is a strong association between infection and the onset of SJS/TEN, in Japan and other countries analyzed. Overall, infection may be an important risk factor for SJS/TEN worldwide.

 Author Contributions

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

 Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

 References Böttiger LE Strandberg I Westerholm B. Drug-induced febrile mucocutaneous syndrome: with a survey of the literature. Acta Med Scand. (1975) 198:229–33. 10.1111/j.0954-6820.1975.tb19532.x1180130 Strom BL Carson JL Halpern AC Schinnar R Snyder ES Shaw M . A population-based study of Stevens–Johnson syndrome: incidence and antecedent drug exposures. Arch Dermatol. (1991) 127:831–8. 10.1001/archderm.1991.016800500750072036028 Roujeau JC Guillaume JC Fabre JP Penso D Fléchet ML Girre JP. Toxic epidermal necrolysis (Lyell syndrome). Incidence and drug etiology in France, 1981–1985. Arch Dermatol. (1990) 126:37–42. 10.1001/archderm.126.1.372134982 Schöpf E Stühmer A Rzany B Victor N Zentgraf R Kapp JF. Toxic epidermal necrolysis and Stevens–Johnson syndrome: an epidemiologic study from West Germany. Arch Dermatol. (1991) 127:839–42. 10.1001/archderm.1991.016800500830082036029 Frey N Jossi J Bodmer M Bircher A Jick SS Meier CR . The epidemiology of Stevens–Johnson syndrome and toxic epidermal necrolysis in the UK. J Invest Dermatol. (2017) 137:1240–7. 10.1016/j.jid.2017.01.03128202399 Syu FK Pan HY Chuang PC Huang YS Cheng CY Cheng FJ. Incidence of Stevens–Johnson syndrome following combination drug use of allopurinol, carbamazepine and phenytoin in Taiwan: a case–control study. J Dermatol. (2018) 45:1080–7. 10.1111/1346-8138.1452829963717 Hosohata K Inada A Oyama S Niinomi I Wakabayashi T Iwanaga K. Adverse cutaneous drug reactions associated with old- and new- generation antiepileptic drugs using the Japanese Pharmacovigilance Database. Clin Drug Investig. (2019) 39:363–8. 10.1007/s40261-019-00754-z30689189 Rzany B Correia O Kelly JP Naldi L Auquier A Stern R. for the Study Group of the International Case Control Study on Severe Cutaneous Adverse Reactions. Risk of Stevens–Johnson syndrome and toxic epidermal necrolysis during first weeks of antiepileptic therapy: a case-control study. Lancet. (1999) 353:2190–4. 10.1016/S0140-6736(98)05418-X10392983 Ueta M. Results of detailed investigations into stevens-johnson syndrome with severe ocular complications. Invest Ophthalmol Vis Sci. (2018) 59:DES183–DES191. 10.1167/iovs.17-2353730481825 Tangamornsuksan W Chanprasert S Nadee P Rungruang S Meesilsat N Ueta M . HLA genotypes and cold medicine-induced Stevens-Johnson syndrome/toxic epidermal necrolysis with severe ocular complications: a systematic review and meta-analysis. Sci Rep. (2020) 10:10589. 10.1038/s41598-020-67610-532601360 Lee HY Chung WH. Toxic epidermal necrolysis: the year in review. Curr Opin Allergy Clin Immunol. (2013) 13:330–6. 10.1097/ACI.0b013e3283630cc2 Wang Q Sun S Xie M Zhao K Li X Zhao Z. Association between the HLA-B alleles and carbamazepine-induced SJS/TEN: a meta-analysis. Epilepsy Res. (2017) 135:19–28. 10.1016/j.eplepsyres.2017.05.015 Park HJ Kim YJ Kim DH Kim J Park KH Park JW . HLA allele frequencies in 5802 koreans: varied allele types associated with SJS/TEN according to culprit drugs. Yonsei Med J. (2016) 57:118–26. 10.3349/ymj.2016.57.1.11826632391 Dean L. Carbamazepine Therapy and HLA Genotype. Bethesda, MD: National Center for Biotechnology Information (2012). Sah R Neupane S Khadka S Poudyal S Paneru HR Sah R . A case study of Stevens–Johnson syndrome-toxic epidermal necrolysis (SJS-TEN) overlap in Mycoplasma pneumoniae -Associated tracheobronchitis. Case Rep Infect Dis. (2019) 2019:5471765. 10.1155/2019/547176531275674 Fournier S Bastuji-Garin S Mentec H Revuz J Roujeau JC. Toxic epidermal necrolysis associated with Mycoplasma pneumoniae infection. Eur J Clin Microbiol Infect Dis. (1995) 14:558–9. 10.1007/BF0211344223381713 Forman R Koren G Shear NH. Erythema multiforme, Stevens–Johnson syndrome and toxic epidermal necrolysis in children: a review of 10 years' experience. Drug Saf. (2002) 25:965–72. 10.2165/00002018-200225130-0000612381216 Aihara Y Ito S Kobayashi Y Aihara M. Stevens–Johnson syndrome associated with azithromycin followed by transient reactivation of herpes simplex virus infection. Allergy. (2004) 59:118. 10.1046/j.1398-9995.2003.00336.x14674950 Choy AC Yarnold PR Brown JE Kayaloglou GT Greenberger PA Patterson R. Virus induced erythema multiforme and Stevens–Johnson syndrome. Allergy Proc. (1995) 16:157–61. 10.2500/1088541957786668478566721 Imatoh T Sai K Fukazawa C Hinomura Y Nakamura R Okamoto-Uchida Y . Association between infection and severe drug adverse reactions: an analysis using data from the Japanese Adverse Drug Event Report database. Eur J Clin Pharmacol. (2017) 73:1643–53. 10.1007/s00228-017-2320-5 Strom BL Stephen E Kimmel SH. Pharmacoepidemiology. 6th ed. Hoboken, NJ: Wiley (2019). Jha AK Laguette J Seger A Bates DW. Can surveillance systems identify and avert adverse drug events? A prospective evaluation of a commercial application. J Am Med Inform Assoc. (2008) 15:647–53. 10.1197/jamia.M263418579834 Sherman RE Anderson SA Dal Pan GJ Gray GW Gross T Hunter NL . Real-world evidence—What is it and what can it tell us? N Engl J Med. (2016) 375:2293–7. 10.1056/NEJMsb160921629733448 Platt R Wilson M Chan KA Benner JS Marchibroda J McClellan M. The new sentinel network–improving the evidence of medical-product safety. N Engl J Med. (2009) 361:645–7. 10.1056/NEJMp090533819635947 Behrman RE Benner JS Brown JS McClellan M Woodcock J Platt R. Developing the sentinel system—a national resource for evidence development. N Engl J Med. (2011) 364:498–9. 10.1056/NEJMp101442721226658 Yamaguchi M Inomata S Harada S Matsuzaki Y Kawaguchi M Ujibe M . Establishment of the MID-NET® medical information database network as a reliable and valuable database for drug safety assessments in Japan. Pharmacoepidemiol Drug Saf. (2019) 28:1395–404. 10.1002/pds.487931464008 Imatoh T Sai K Saito Y. The association between concurrence of infection and the onset of severe eruption or liver injury in patients using antipyretic analgesics: a matched, nested case-control study. J Clin Pharmacol. (2020) 60:1177–84. 10.1002/jcph.161332521061 Liotti L Caimmi S Bottau P Bernardini R Cardinale F Saretta F . Clinical features, outcomes and treatment in children with drug induced Stevens–Johnson syndrome and toxic epidermal necrolysis. Acta Biol Med. (2019) 90:52–60. 10.23750/abm.v90i3-S.816530830062 Coopman SA Johnson RA Platt R Stern RS. Cutaneous disease and drug reactions in HIV infection. N Engl J Med. (1993) 328:1670–4. 10.1056/NEJM1993061032823048413490 Olson D Watkins LKF Demirjian A Lin X Robinson CC Pretty K . Outbreak of mycoplasma pneumoniae-associated Stevens–Johnson syndrome. Pediatrics. (2015) 136:e386–e394. 10.1542/peds.2015-0278 Chung WH Shih SR Chang CF Lin TY Huang YC Chang SC . Clinicopathologic analysis of coxsackievirus A6 new variant induced widespread mucocutaneous bullous reactions mimicking severe cutaneous adverse reactions. J Infect Dis. (2013) 208:1968–78. 10.1093/infdis/jit38323904296 Ginsburg CM. Stevens–Johnson syndrome in children. Pediatr Infect Dis. (1982) 1:155–8. 10.1097/00006454-198205000-00005 Kittipibul T Puangsricharern V Chatsuwan T. Comparison of the ocular microbiome between chronic Stevens–Johnson syndrome patients and healthy subjects. Sci Rep. (2020) 10:4353. 10.1038/s41598-020-60794-w32152391 Mockenhaupt M Viboud C Dunant A Naldi L Halevy S Bouwes Bavinck JNB . Stevens–Johnson syndrome and toxic epidermal necrolysis: assessment of medication risks with emphasis on recently marketed drugs. The EuroSCAR-study. J Invest Dermatol. (2008) 128:35–44. 10.1038/sj.jid.570103317805350 Blumenthal KG Wickner PG Lau JJ Zhou L. Stevens–Johnson syndrome and toxic epidermal necrolysis: a cross-sectional analysis of patients in an integrated allergy repository of a large health care system. J Allergy Clin Immunol Pract. (2015) 3:277–80.e1. 10.1016/j.jaip.2014.10.00225609329 Phillips EJ Sukasem C Whirl-Carrillo M Müller DJ Dunnenberger HM Chantratita W . Clinical pharmacogenetics implementation consortium guideline for HLA genotype and use of carbamazepine and oxcarbazepine: 2017 Update. Clin Pharmacol Ther. (2018) 103:574–81. 10.1002/cpt.100429392710 White KD Chung WH Hung SI Mallal S Phillips EJ. Evolving models of the immunopathogenesis of T cell-mediated drug allergy: the role of host, pathogens, and drug response. J Allergy Clin Immunol. (2015) 136:219–34; quiz 235. 10.1016/j.jaci.2015.05.05026254049 Hsu DY Brieva J Silverberg NB Silverberg JI. Morbidity and mortality of Stevens–Johnson syndrome and toxic epidermal necrolysis in United States adults. J Invest Dermatol. (2016) 136:1387–97. 10.1016/j.jid.2016.03.02327039263 Yang SC Hu S Zhang SZ Huang JW Zhang J Ji C . The epidemiology of Stevens–Johnson syndrome and toxic epidermal necrolysis in China. J Immunol Res. (2018) 2018:4320195. 10.1155/2018/432019530050956 Goldman SA. Limitations and strengths of spontaneous reports data. Clin Ther. (1998) 20(Suppl C):C40–C44. 10.1016/S0149-2918(98)80007-69915089 Wong CK Ho SS Saini B Hibbs DE Fois RA. Standardisation of the FAERS database: a systematic approach to manually recoding drug name variants. Pharmacoepidemiol Drug Saf. (2015) 24:731–7. 10.1002/pds.380526017154

Funding. This study was supported by AMED under Grant Number 19mk0101085j0503.