Effect of Infectious Diseases on the Pathogenesis of Stevens–Johnson Syndrome and Toxic Epidermal Necrolysis

Abstract

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe cutaneous adverse drug reactions. Recent studies have revealed that the prevalence of SJS/TEN is associated with genetic backgrounds, such as polymorphisms in human leukocyte antigens (HLAs). However, non-genetic factors contributing to the etiology of SJS/TEN are largely unknown. This study aimed to assess the involvement of concurrent infection on the pathological states of SJS/TEN, examining the severity of cutaneous symptoms and ocular involvement as well as the time to onset in drug-induced SJS/TEN patients. We recruited 257 Japanese SJS/TEN patients from June 2006 to September 2013 through a nationwide case collection network and participating hospitals and reviewed the clinical information including patient backgrounds, primary disease and medication status. Association between infection and pathological states of SJS/TEN was assessed using univariate and multivariate analyses. The concurrent infectious group of SJS/TEN patients showed a significantly higher rate of exhibiting severer dermatological and ophthalmological phenotypes and an earlier onset of SJS/TEN than the non-infectious group. Our results suggest that the infection could be a risk factor to cause severer symptoms and earlier onset of SJS/TEN.

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are diseases within the same spectrum of immune-mediated cutaneous adverse drug reactions characterized by the detachment of the epidermis and erosion of the mucous membrane.¹⁾ These disorders are categorized according to the extent of body surface area involved.^2,3) Although SJS and TEN are very rare, the associated mortality rates are very high; recently, the mortality rate was reported to be 34% for SJS/TEN at 1 year after onset in Europe,⁴⁾ and was 3 and 19% for SJS and TEN, respectively, in Japan.⁵⁾ Depending on the mucosal involvements in the acute stage of these diseases, various long-lasting sequelae and complications may develop. Particularly, ocular complications ranging from mild symptoms to very severe symptoms, such as corneal melting and perforation, which sometimes cause blindness, occur in more than 50% of patients.⁶⁾ Therefore, the estimation of severity of cutaneous and ocular symptoms and the therapeutic intervention in early stage are important to avoid lowering the patient’s quality-of-life.

The pathological mechanisms of SJS/TEN remain largely unclear. In the cutaneous lesions of SJS/TEN, soluble Fas ligand (FasL) and granulysin as well as T-helper 1 (Th1) and T-helper 2 (Th2) cytokines produced by activated T cells were strongly expressed, which is considered to cause keratinocyte death.^7,8) Recent studies have revealed that the prevalence of SJS/TEN is associated with genetic backgrounds such as polymorphisms in human leukocyte antigens (HLAs) and a metabolic enzyme, CYP.^9–16) However, not all patients with these high-risk genetic factors develop SJS/TEN after exposure to the risk drugs, suggesting that non-genetic factors could contribute to the SJS/TEN pathogenesis. Recently, White et al. proposed a mechanism (known as heterologous immunity model) that some kinds of virus-specific (memory) T-cells crossreact to self tissue that are modified by a culprit drug.¹⁷⁾ Therefore, infectious status of the patient is considered to be one of the key non-genetic factors for the pathogenesis of SJS/TEN. In this study, in order to investigate possible involvement of concurrent infections in exhibiting severe SJS/TEN phenotypes, we reviewed 257 drug-induced SJS/TEN cases and evaluated the effects of concurrent infections, such as common cold and other acute infections, on the severity of dermatological and ophthalmological symptoms and the time latency between beginning of drug use and onset of SJS/TEN.

MATERIALS AND METHODS

Study Population

This was a retrospective study using clinical information of Japanese patients who experienced SJS/TEN from March 1983 to September 2013, and their clinical data were collected through voluntarily participating universities and hospitals via the nationwide case collection network.¹⁸⁾ This study protocol was approved by the institutional review boards of all participating institutions, and written informed consent was obtained from all the patients.

Data Collection

Clinical information was collected and diagnosed using a standardized case report form based on the criteria proposed by Bastuji-Garin et al.¹⁹⁾ The concurrent acute infection status, including common cold, herpes, sepsis, pneumonia, infectious enteritis, rhinosinusitis, and urinary infection, was determined by reviewing the case reports, and patients with infections and/or those treated with anti-infectives before SJS/TEN onset were categorized as having SJS/TEN with infection. The patients were also classified based on drug categories of the suspected drugs for SJS/TEN onset. According to the Japanese diagnostic criteria of SJS/TEN, TEN (the severer dermatological phenotype) was diagnosed when the skin detachment was more than 10% of body surface area. Ocular surface involvement was graded from 0 to 3 based on the status of pseudomembrane formation or a defect/erosion of the conjunctiva/corneal epithelia as described in detail previously,¹⁸⁾ and grades 2 and 3 were grouped as severe ocular involvement (SOI).

Statistical Analysis

The association between concurrent infection and the rate of TEN or SOIs was assessed by Fisher’s exact test and multiple logistic regression analysis. Differences in days from initiation of drug administration to the onset of SJS/TEN were assessed by Mann–Whitney U test and a generalized linear model. A multiple logistic regression analysis, Fisher’s exact test, and Mann–Whitney U test were performed using the EZR software (Saitama Medical Center, Jichi Medical University, Saitama, Japan).²⁰⁾ A generalized linear model with a gamma distribution was performed using SAS Enterprise Guide 6.1 (SAS Institute Inc., Cary, NC, U.S.A.).

RESULTS AND DISCUSSION

A total of 257 patients with SJS/TEN were enrolled (128 males, mean age 56.6±21.8 years, median age 61.5 years). Among these, 122 (47.5%) suffered concurrent infectious diseases (Table 1), including common cold (43.4%), herpes (4.9%), sepsis (4.1%), pneumonia (3.3%), infectious enteritis (2.5%), rhinosinusitis (2.5%), and urinary infection (2.5%). The mean age of the participants of the infectious group was significantly lower than that of the non-infectious group, while the sex ratio of both groups was comparable. Antipyretic analgesics (AAs), anti-epileptics, and anti-hyperuricemic drugs are reported relatively frequently as the suspected drugs for SJS/TEN,²¹⁾ and those drugs were administered to 102, 67, and 32 patients, respectively, in this study. The rest of the patients were administered other therapeutic drugs, including antitumor drugs, proton pump inhibitors, and antidiabetic drugs.

Table 1. Patient’s Backgrounds: Drugs Administered and Rates of Infection

Background		Total cases	Infection		p Value†
			−	+
All patients
	No. of cases	257	135	122
	Age (median [IQR]), y	61.5 [39.75–74]	64 [47–75.5]	57 [36–73]	0.013
	Male (%)	50%	48%	53%	0.453
Drug groups
Antipyretic analgesics
	No. of cases	102	30	72
	Age (median [IQR]), y	54 [33.5–72.5]	65 [43.5–76.75]	53 [31.5–67.5]	0.036
	Male (%)	45%	38%	48%	0.386
Anti-epileptics
	No. of cases	67	55	12
	Age (median [IQR]), y	53 [38–65]	54 [40–65]	42 [16.25–66.25]	0.432
	Male (%)	46%	47%	42%	0.761
Anti-hyperuricemic drugs
	No. of cases	32	23	9
	Age (median [IQR]), y	73.5 [60.75–81.75]	75 [63–86]	72 [63–82]	0.378
	Male (%)	66%	70%	56%	0.681

IQR, interquartile range. ^† Mann–Whitney U test and Fisher’s exact test were used for analyses of age and sex, respectively.

To investigate the effect of concurrent infections on the rate of the severe skin phenotype, we compared the rates of TEN between patients with and without infections. The numbers of probable/definite SJS and definite TEN cases were 177 and 80, respectively. The rate of TEN was significantly higher in the infectious group (41.0%) than in the non-infectious group (22.2%), with an odds ratio (OR) of 2.43 (Table 2). Although age distribution differed among the groups (Table 1), the significance still remained after adjusting by age and sex [adjusted odds ratio (aOR)=2.27], suggesting that the difference on TEN ratio was not due to age bias. A previous studies reported that in most cases, the onset of SJS/TEN was observed less than 8 weeks after the initiation of the suspected drugs (hereafter referred to as the <8 weeks criterion).^3,21) We also observed that the rate of TEN remained significantly higher in the infectious group even after excluding patients who did not meet the <8 weeks criterion (aOR=2.39). Among SJS/TEN patients who were administered AAs, the rate of TEN was significantly higher in the infectious group (aOR=4.65), and the difference also remained significant after applying the <8 weeks criterion (aOR=4.08). Similar tendency was observed in anti-hyperuricemia drug-related SJS/TEN patients but was not statistically significant. Conversely, the risk of TEN was comparable between the infectious and non-infectious groups among the SJS/TEN patients who were administered anti-epileptics.

Table 2. Evaluation of Risk of TEN and SOI as Well as Days to Onset in the SJS/TEN Patients with Infections

	TEN/Total (%)		OR	95% CI	p Value	aOR†	95% CI†	p Value†
	Infection −	Infection +
All patients	30/135 (22.2)	50/122 (41.0)	2.43	1.41–4.18	0.0014**	2.27	1.30–3.96	0.0038**
<8 weeks	26/114 (22.8)	46/111 (41.4)	2.40	1.34–4.27	0.0031**	2.39	1.32–4.33	0.0042**
Drug groups
Antipyretic analgesics	4/30 (13.3)	31/72 (43.1)	4.91	1.55–15.50	0.0067**	4.65	1.40–15.40	0.0120*
<8 weeks	4/24 (16.7)	29/65 (44.6)	4.03	1.24–13.10	0.0206*	4.08	1.21–13.80	0.0235*
Anti-epileptics	15/55 (27.3)	4/12 (33.3)	1.33	0.35–5.09	0.6740	1.27	0.33–5.00	0.7280
<8 weeks	13/47 (27.7)	4/12 (33.3)	1.31	0.34–5.09	0.6990	1.22	0.30–4.92	0.7780
Anti-hyperuricemic drugs	5/23 (21.7)	5/9 (55.6)	4.50	0.87–23.30	0.0734	4.57	0.82–25.50	0.0832
<8 weeks	3/18 (16.7)	4/8 (50.0)	5.00	0.78–32.10	0.0898	7.04	0.67–74.40	0.1050
	SOI/Total (%)		OR	95% CI	p Value	aOR†	95% CI†	p Value†
	Infection −	Infection +
All patients	26/116 (22.4)	50/114 (43.9)	2.70	1.53–4.79	0.0007**	2.57	1.41–4.71	0.0022**
<8 weeks	22/100 (22.0)	46/106 (43.4)	2.72	1.48–5.00	0.0013**	2.43	1.29–4.57	0.0062**
Drug groups
Antipyretic analgesics	5/30 (16.7)	39/70 (55.7)	6.29	2.16–18.30	0.0008**	7.59	2.23–25.80	0.0012**
<8 weeks	3/24 (12.5)	35/64 (54.7)	8.45	2.29–31.20	0.0014**	8.08	2.09–31.10	0.0024**
Anti-epileptics	10/44 (22.7)	5/10 (50.0)	3.40	0.82–14.20	0.0926	3.15	0.71–14.00	0.1310
<8 weeks	10/40 (25.0)	5/10 (50.0)	3.00	0.72–12.60	0.1320	2.84	0.64–12.60	0.1710
Anti-hyperuricemic drugs	6/22 (27.3)	5/8 (62.5)	4.44	0.80–24.60	0.0876	6.01	0.90–40.10	0.0642
<8 weeks	6/18 (33.3)	5/7 (71.4)	5.00	0.74–33.80	0.0987	7.67	0.83–71.10	0.0729
	Days from initiation of drug to onset of SJS/TEN: Median [IQR] (N)#		p Value (Mann–Whitney U test)			p Value† (Gamma regression)
	Infection −	Infection +
All patients	17 [11–34] (121)	6 [2–17] (115)	1.49E–07**			1.51E–12**
<8 weeks	15 [10–27.25] (114)	6 [2–15.5] (111)	1.41E–07**			7.63E–05**
Drug groups
Antipyretic analgesics	8 [2–18] (21)	3 [1–8] (59)	0.0243*			1.84E–11**
<8 weeks	6.5 [2–14] (18)	3 [1–7] (57)	0.0920			0.0833
Anti-epileptics	23.5 [14–43] (48)	18 [10–31.5] (11)	0.2930			0.0144*
<8 weeks	18 [14–30] (45)	18 [10–27] (11)	0.5430			0.2487
Anti-hyperuricemic drugs	17 [13–25] (15)	16 [10.5–21] (7)	0.7510			0.9363
<8 weeks	16.5 [12.5–21.5] (14)	16 [10.5–21] (7)	0.9400			0.7872

Association between concurrent infection and the rate of TEN or SOIs was assessed by Fisher’s exact test and a multiple logistic regression analysis. Differences in days from initiation of drug to onset of SJS/TEN was assessed by Mann–Whitney U test and a generalized linear model. TEN, toxic epidermal necrolysis; SOI, severe ocular involvement; SJS, Stevens–Johnson syndrome; OR, odds ratio; CI, confidence interval; IQR, interquartile range. ^† Adjusted for age and sex. ^# For the patients who took multiple culprit drugs, we used the period of the earliest prescribed drug. * p<0.05, ** p<0.01.

We next compared the rates of SOI between the infectious and non-infectious groups. The number of SJS/TEN patients with and without SOI was 76 and 154, respectively. The remaining 27 patients were excluded because SOI information was not described in their case reports. As shown in Table 2, the infectious group developed SOI more frequently (43.9%) than the non-infectious group (22.4%), with an aOR=2.57, and this difference remained significant even after applying the <8-week criterion (aOR=2.43). Among SJS/TEN patients treated with AAs, the infectious group was more likely to experience SOI (55.7%) than the non-infectious group (16.7%; aOR=7.59). This association is consistent with that of a previous study, which investigated only for common cold.¹⁸⁾ This significance remained even after applying the <8 weeks criterion (aOR=8.08). Although tendencies of higher rates of SOI were also observed in the infectious group among the anti-epileptic or anti-hyperuricemic drug-related SJS/TEN patients, these changes were not statistically significant.

Finally, we compared the time to onset from the initiation of drug intake between concurrent infectious and non-infectious groups. As expected from the heterologous immunity model, if memory T-cells are involved in the pathogenesis of the cutaneous adverse reactions, the period from taking drugs to the onset of SJS/TEN would be shorter. For the patients who took multiple culprit drugs, we used the period of the earliest prescribed drug. We found that the median days to onset was significantly shorter in the infectious group of SJS/TEN patients (Median: 6 d) than that in the non-infectious group (Median: 15 d), even after the <8 weeks criterion was applied. Moreover, the difference was statistically significant in the AAs and anti-epileptics-administered groups in total, but these differences were disappeared after application of <8 weeks criterion.

Although we could not completely exclude the possibility that anti-infectives themselves might have directly influenced the severity of SJS/TEN, the same effects of infection still remained in our separate analysis excluding the patients who received anti-infectives before onset of SJS/TEN. Thus, this observation suggests that pathophysiological status induced by infection would be involved in the SJS/TEN development (data not shown). In addition to heterologous immunity model, which suggests the involvement of T-cell cross-reactive nature against virus and drug, infection would affect the total balance of immunity which may cause an undesired immune response, such as autoimmune diseases. Actually, a previous study has shown that regulatory T cells (Tregs) prevent life-threatening skin damage, such as TEN.²²⁾ Therefore, our results suggest that the infection could affect the activity of Tregs and cause more severe symptoms of SJS/TEN, particularly induced by AAs. The symptoms of SJS/TEN drastically change within several days.⁵⁾ Therefore, early therapeutic intervention is needed to stop the progression of the disease and prevent long-term complications. Given that concurrent infections are related to the earlier onset of SJS/TEN, the SJS/TEN with infection may progress easier and faster than the non-infectious counterpart, suggesting that higher control on their symptoms is necessary for the infectious group of SJS/TEN patients.

CONCLUSION

We found that the infectious group of SJS/TEN patients showed a significantly higher rate of more severe dermatological and ophthalmological phenotypes and earlier onset of SJS/TEN than the non-infectious group. Our finding provides clinically important information and would be useful for further research on mechanisms of SJS/TEN pathogenesis and its prevention.

Acknowledgments

The study was partly supported by Grants from the Research on Regulatory Science of Pharmaceuticals and Medical Devices from Japan Agency for Medical Research and Development (AMED), the Japan Research Foundation for Clinical Pharmacology and the Japan Society for the Promotion of Science (JSPS) KAKENHI (26293122, 15K19168, 16K08433).

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Roujeau JC, Stern RS. Severe adverse cutaneous reactions to drugs. N. Engl. J. Med., 331, 1272–1285 (1994).
• 2) Chung WH, Wang CW, Dao RL. Severe cutaneous adverse drug reactions. J. Dermatol., 43, 758–766 (2016).
• 3) Yamane Y, Matsukura S, Watanabe Y, Yamaguchi Y, Nakamura K, Kambara T, Ikezawa Z, Aihara M. Retrospective analysis of Stevens–Johnson syndrome and toxic epidermal necrolysis in 87 Japanese patients—Treatment and outcome. Allergol. Int., 65, 74–81 (2016).
• 4) Sekula P, Dunant A, Mockenhaupt M, Naldi L, Bouwes Bavinck JN, Halevy S, Kardaun S, Sidoroff A, Liss Y, Schumacher M, Roujeau JC, RegiSCAR study group. Comprehensive survival analysis of a cohort of patients with Stevens–Johnson syndrome and toxic epidermal necrolysis. J. Invest. Dermatol., 133, 1197–1204 (2013).
• 5) Aihara M, Kano Y, Fujita H, Kambara T, Matsukura S, Katayama I, Azukizawa H, Miyachi Y, Endo Y, Asada H, Miyagawa F, Morita E, Kaneko S, Abe R, Ochiai T, Sueki H, Watanabe H, Nagao K, Aoyama Y, Sayama K, Hashimoto K, Shiohara T, SJS/TEN Study Group. Efficacy of additional i.v. immunoglobulin to steroid therapy in Stevens–Johnson syndrome and toxic epidermal necrolysis. J. Dermatol., 42, 768–777 (2015).
• 6) Sotozono C, Ang LP, Koizumi N, Higashihara H, Ueta M, Inatomi T, Yokoi N, Kaido M, Dogru M, Shimazaki J, Tsubota K, Yamada M, Kinoshita S. New grading system for the evaluation of chronic ocular manifestations in patients with Stevens–Johnson syndrome. Ophthalmology, 114, 1294–1302 (2007).
• 7) Caproni M, Torchia D, Schincaglia E, Volpi W, Frezzolini A, Schena D, Marzano A, Quaglino P, De Simone C, Parodi A, Barletta E, Fabbri P. Expression of cytokines and chemokine receptors in the cutaneous lesions of erythema multiforme and Stevens–Johnson syndrome/toxic epidermal necrolysis. Br. J. Dermatol., 155, 722–728 (2006).
• 8) Chung WH, Hung SI, Yang JY, Su SC, Huang SP, Wei CY, Chin SW, Chiou CC, Chu SC, Ho HC, Yang CH, Lu CF, Wu JY, Liao YD, Chen YT. Granulysin is a key mediator for disseminated keratinocyte death in Stevens–Johnson syndrome and toxic epidermal necrolysis. Nat. Med., 14, 1343–1350 (2008).
• 9) Chung WH, Hung SI, Hong HS, Hsih MS, Yang LC, Ho HC, Wu JY, Chen YT. Medical genetics: a marker for Stevens–Johnson syndrome. Nature, 428, 486 (2004).
• 10) Kaniwa N, Saito Y, Aihara M, Matsunaga K, Tohkin M, Kurose K, Furuya H, Takahashi Y, Muramatsu M, Kinoshita S, Abe M, Ikeda H, Kashiwagi M, Song Y, Ueta M, Sotozono C, Ikezawa Z, Hasegawa R, JSAR research group. HLA-B*1511 is a risk factor for carbamazepine-induced Stevens–Johnson syndrome and toxic epidermal necrolysis in Japanese patients. Epilepsia, 51, 2461–2465 (2010).
• 11) Hung SI, Chung WH, Liou LB, Chu CC, Lin M, Huang HP, Lin YL, Lan JL, Yang LC, Hong HS, Chen MJ, Lai PC, Wu MS, Chu CY, Wang KH, Chen CH, Fann CS, Wu JY, Chen YT. HLA-B*5801 allele as a genetic marker for severe cutaneous adverse reactions caused by allopurinol. Proc. Natl. Acad. Sci. U.S.A., 102, 4134–4139 (2005).
• 12) Ozeki T, Mushiroda T, Yowang A, Takahashi A, Kubo M, Shirakata Y, Ikezawa Z, Iijima M, Shiohara T, Hashimoto K, Kamatani N, Nakamura Y. Genome-wide association study identifies HLA-A*3101 allele as a genetic risk factor for carbamazepine-induced cutaneous adverse drug reactions in Japanese population. Hum. Mol. Genet., 20, 1034–1041 (2011).
• 13) Mallal S, Nolan D, Witt C, Masel G, Martin AM, Moore C, Sayer D, Castley A, Mamotte C, Maxwell D, James I, Christiansen FT. Association between presence of HLA-B*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet, 359, 727–732 (2002).
• 14) Illing PT, Vivian JP, Dudek NL, Kostenko L, Chen Z, Bharadwaj M, Miles JJ, Kjer-Nielsen L, Gras S, Williamson NA, Burrows SR, Purcell AW, Rossjohn J, McCluskey J. Immune self-reactivity triggered by drug-modified HLA-peptide repertoire. Nature, 486, 554–558 (2012).
• 15) Kaniwa N, Sugiyama E, Saito Y, Kurose K, Maekawa K, Hasegawa R, Furuya H, Ikeda H, Takahashi Y, Muramatsu M, Tohkin M, Ozeki T, Mushiroda T, Kubo M, Kamatani N, Abe M, Yagami A, Ueta M, Sotozono C, Kinoshita S, Ikezawa Z, Matsunaga K, Aihara M, Japan Pharmacogenomics Data Science Consortium. Specific HLA types are associated with antiepileptic drug-induced Stevens–Johnson syndrome and toxic epidermal necrolysis in Japanese subjects. Pharmacogenomics, 14, 1821–1831 (2013).
• 16) Kaniwa N, Saito Y. Pharmacogenomics of severe cutaneous adverse reactions and drug-induced liver injury. J. Hum. Genet., 58, 317–326 (2013).
• 17) 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., 136, 219–234, quiz, 235 (2015).
• 18) Kaniwa N, Ueta M, Nakamura R, Okamoto-Uchida Y, Sugiyama E, Maekawa K, Takahashi Y, Furuya H, Yagami A, Matsukura S, Ikezawa Z, Matsunaga K, Sotozono C, Aihara M, Kinoshita S, Saito Y. Drugs causing severe ocular surface involvements in Japanese patients with Stevens–Johnson syndrome/toxic epidermal necrolysis. Allergol. Int., 64, 379–381 (2015).
• 19) Bastuji-Garin S, Rzany B, Stern RS, Shear NH, Naldi L, Roujeau JC. Clinical classification of cases of toxic epidermal necrolysis, Stevens–Johnson syndrome, and erythema multiforme. Arch. Dermatol., 129, 92–96 (1993).
• 20) Kanda Y. Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant., 48, 452–458 (2013).
• 21) Mockenhaupt M, Viboud C, Dunant A, Naldi L, Halevy S, Bouwes Bavinck JN, Sidoroff A, Schneck J, Roujeau JC, Flahault A. Stevens–Johnson syndrome and toxic epidermal necrolysis: assessment of medication risks with emphasis on recently marketed drugs. The EuroSCAR-study. J. Invest. Dermatol., 128, 35–44 (2008).
• 22) Azukizawa H, Sano S, Kosaka H, Sumikawa Y, Itami S. Prevention of toxic epidermal necrolysis by regulatory T cells. Eur. J. Immunol., 35, 1722–1730 (2005).