Stevens–Johnson syndrome and toxic epidermal necrolysis: Updates in pathophysiology and management

Abstract

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are life-threatening conditions characterized by extensive detachment of the epidermis and mucous membranes. These severe disorders carry a high mortality rate, and their pathogenesis remains largely unclear. Furthermore, optimal therapeutic strategies for SJS/TEN remain a subject of ongoing debate. Early diagnosis of SJS/TEN is challenging, and reliable biomarkers for diagnosis or severity prediction have not been firmly established. Certain drugs, such as carbamazepine and allopurinol, have shown a strong association with specific human leukocyte antigen (HLA) types. Recently, the potential benefits of HLA screening prior to administering these drugs to reduce the incidence of SJS/TEN have been explored. Epidermal cell death in SJS/TEN lesions is caused by extensive apoptosis, primarily through the Fas–Fas ligand (FasL) and perforin/granzyme pathways. Our findings suggest that necroptosis, a form of programmed necrosis, also contributes to epidermal cell death. Annexin A1, released from monocytes, interacts with the formyl peptide receptor 1 to induce necroptosis. Several biomarkers, such as CC chemokine ligand (CCL)-27, interleukin-15, galectin-7, receptor-interacting protein kinases 3 (RIP3), and lipocalin-2, have been identified for diagnostic and prognostic purposes in SJS/TEN. Supportive care is recommended for treating SJS/TEN, but the efficacy of various therapeutic options–including systemic corticosteroids, intravenous immunoglobulin, cyclosporine, and tumor necrosis factor-α antagonists–remains controversial. Recent studies have investigated the potential benefits of tumor necrosis factor-α antagonists. In this review, we discuss recent advances in the understanding and management of SJS/TEN.

Introduction

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe, life-threatening mucocutaneous reactions characterized by blisters and extensive skin detachment. Drugs and infections, such as those caused by Mycoplasma or the herpes simplex virus, are the primary triggers.^[1]

SJS and TEN are considered part of a disease spectrum differentiated by the extent of skin detachment (SJS: skin detachment area <10%, TEN: >30%, SJS/TEN overlap: 10–30%).^[2] Despite the low incidence rates reported in Japan (2.5 cases per million for SJS and 1.0 case per million for TEN), these conditions have high mortality rates (SJS: 4.1%; TEN: 29.9%).^[3] Moreover, survivors often suffer from severe sequelae, such as blindness.^[1] Therefore, prompt and accurate diagnosis, followed by appropriate treatment, is crucial. However, early diagnosis and severity prediction biomarkers are not yet established, and there is a lack of robust evidence for optimal management strategies for SJS/TEN.

In this review, we describe the clinical features and recent research on the pathomechanisms, diagnostic biomarkers, and management of SJS/TEN.

Clinical Features and Scoring System for Mortality Prediction

The cutaneous manifestations of SJS/TEN include painful erythematous rash, bullae, and erosions predominantly on the face and trunk, spreading to the extremities [Figure 1A, B]. Early lesions appear as round, non-palpable areas with indistinct borders, termed “atypical targets” [Figure 1C]. Skin lesions typically exhibit a positive Nikolsky sign, indicating skin erosion upon gentle pressure. Symptoms such as malaise, fever, and upper respiratory tract issues often precede the rash by a few days. Nearly all patients with SJS/TEN experience mucosal involvement, particularly in the eyes, mouth, and genitalia [Figure 2A, B].^[4] Early ocular involvement varies and may include conjunctival hyperemia, corneal and conjunctival erosion, and pseudomembrane formation. Post-recovery, persistent ocular inflammation can lead to sequelae such as dry eyes, reduced visual acuity, conjunctivitis, corneal erosions, and trichiasis, with severe cases resulting in blindness. Dysuria suggests urethral involvement, and erosions may also occur in the lower respiratory tract and intestines.

Figure 1.

Skin manifestations of SJS/TEN. (A,B) Widespread skin detachment in a TEN patient. (C) Atypical targets. Flat erythematous macules with a darker center. SJS: Stevens–Johnson syndrome; TEN: Toxic epidermal necrolysis.

Figure 2.

Mucosal involvement of SJS/TEN. (A) Affected eye with conjunctivitis. (B) Lip involvement. Severe cheilitis with hemorrhagic crust. SJS: Stevens–Johnson syndrome; TEN: Toxic epidermal necrolysis.

SJS/TEN are primarily drug-induced. Onset typically occurs within 1–2 weeks of drug administration. Infections like Mycoplasma or herpes simplex virus are also significant causes. Commonly implicated drugs include antibiotics, allopurinol, non-steroidal anti-inflammatory drugs (NSAIDs), and antiepileptic drugs.^[5] The frequent causative drugs are listed in Table 1.^[3,5]

Table 1.

Causative drugs of SJS/TEN.

Drug classification	Causative drugs
Antibiotics	Penicillins, cephems, fluoroquinolones, sulfamethoxazole and trimethoprim combination
Xanthine oxidase inhibitor	Allopurinol
Anti-inflammatory drugs	Acetaminophen, loxoprofen, celecoxib, ibuprofen
Antiepileptic	CBZ, lamotrigine, phenytoin, valproic acid
Peptic ulcer agent	Lansoprazole, omeprazole, esomeprazole
Anti-HIV drug	Nevirapine

CBZ: Carbamazepine; HIV: Human immunodeficiency virus; SJS: Stevens–Johnson syndrome; TEN: Toxic epidermal necrolysis.

The severity-of-illness score for toxic epidermal necrolysis (SCORTEN) is widely employed to predict mortality in SJS/TEN patients.^[6] SCORTEN should be assessed within the first 24 h of admission and re-evaluated on day 3. This scoring system is based on seven independent risk factors: age, malignancy, tachycardia, percentage of epidermal detachment, serum urea, serum glucose, and bicarbonate levels [Table 2]. The presence of more risk factors correlates with higher mortality rates [Table 3].

Table 2.

Risk factors for SCORTEN.

Risk factors	Illustration
Age	Age >40 years
Tachycardia	Heart rate >120 beats/min
Malignancy	Presence of cancer or hematologic malignancy
Epidermal detachment	Epidermal detachment area involving body surface area >10%
Serum urea	Blood urea nitrogen >28 mg/dL (10 mmol/L)
Serum glucose	Blood glucose >252 mg/dL (14 mmol/L)
Bicarbonate	Bicarbonate <20 mmol/L

SCORTEN: Severity-of-illness score for toxic epidermal necrolysis.

Table 3.

Mortality rate in SCORTEN.

Number of risk factors	Mortality rate (%)
0–1	3.2
2	12.1
3	35.3
4	58.3
≥5	90.0

SCORTEN: Severity-of-illness score for toxic epidermal necrolysis.

Recently, our group proposed a novel scoring system, the Clinical Risk Score for TEN (CRISTEN), which predicts early stage mortality using only clinical information, without requiring laboratory data.^[7] CRISTEN includes 10 risk factors such as age, percentage of epidermal detachment, use of antibiotics as causative agents, systemic corticosteroid therapy before onset, and mucosal involvement affecting the ocular, buccal, and genital areas. Underlying conditions such as renal impairment, diabetes, cardiovascular disease, malignant neoplasms, and bacterial infections are also considered [Table 4]. CRISTEN has shown statistically comparable results to previous systems.

Table 4.

Risk factors for CRISTEN.

No.	Parameter	Detailed definition
1	Age	Age over 65 years
2	Epidermal detachment	Epidermal detachment of >10% of BSA
3	Malignant neoplasm	Active phase
4	Diabetes mellitus	Under treatment with medication
5	Renal impairment	Chronic kidney disease
6	Bacterial infection	Pneumonia, sepsis, and urinary-tract infection
7	Cardiac disease	Heart failure, valvular disease, arrhythmias, aortic aneurysms, angina, atrial and ventricular septal defects, and hypertension under treatment
8	Drugs	Antibiotics in the culprit drugs
9	Mucosal damage	Mucosal damage affecting all three of ocular, buccal, and genital mucosa
10	Systemic corticosteroid therapy before the onset of SJS/TEN	Regardless of dose or duration of administration

BSA: Body surface area; CRISTEN: Clinical risk score for toxic epidermal necrolysis; SJS: Stevens–Johnson syndrome; TEN: Toxic epidermal necrolysis.

Determine the Causative Drug

To determine the causative drug, drug rechallenge is the most reliable approach. However, in SJS/TEN cases, it is not allowed due to the risk of life-threatening consequences. In delayed-type hypersensitivity reaction such as SJS/TEN mediated by drug-specific T cells, patch testing is a useful adjunct.^[8] It is performed by applying predetermined dilutions of the compound on intact skin for different periods and assessing for erythema, edema, or vesiculation, all indicative of specific T-cell reactivity. The positive rate of patch test depends on the type of cutaneous adverse drug reactions (cADRs).^[9] Drug-induced hypersensitivity syndrome (DIHS), maculopapular exanthema, and fixed drug eruption (56.3%, 23.9%, and 20.0%, respectively). The positive rate for SJS/TEN is 6.7%. In addition to the disease type, the positive rate also depends on the drug, with contrast media and antiepileptic drugs having high positive rates (41.1% and 41.0%, respectively).

As an in vitro assay, lymphocyte transformation tests (LTT) is used to confirm the causative drug. The LTT measures the proliferation of T cells in response to the drug. Although the sensitivity and specificity of the LTT are limited (sensitivity: 62%, specificity: 91%),^[10] it is useful because LTT is an in vitro assay and not invasive for patients. The sensitivity of the test varied with the causative drugs, clinical phenotypes, and the timing of the test.^[11] In SJS/TEN, it tends to obtain positive reactions when the test was performed at the acute stage of the disease.^[11] Some drugs, such as vancomycin or NSAIDs, may elicit enhanced proliferation even in non-sensitized individuals. Meanwhile, immunosuppressive drugs or cytotoxic chemo-therapy agents interrupt the reaction of LTT.^[12] Of these drugs, LTT may not be performed accurately. Interferon (IFN)-γ enzyme-linked immunoSpot (ELISpot) assay is also used as an in vitro assay to identify causative drugs. IFN-γ ELISpot is a method in which the patient’s PBMCs are stimulated with the causative drug and the number of cells that release IFN is evaluated. IFN-γ ELISpot performed within 30 days of cADR onset revealed high sensitivity and specificity than LTT (sensitivity: 75%, specificity: 95%).^[13]

Algorithm of Drug causality in Epidermal Necrolysis (ALDEN), an algorithm for assessment of drug causality in epidermal necrolysis, is used to assess drug causality in SJS/TEN.^[14] ALDEN is an algorithm for retrospective assessment of drug causality and is not used during the acute phase. ALDEN is based on six independent parameters, and the total score varies from −12 to +10. The scores attributed to each drug were pooled in categories aimed at reflecting causality: “very probable”: score ≥6; “probable”: score 4–5; “possible”: score 2–3; “unlikely”: score 0–1; and “very unlikely”: score <0 [Table 5]. This algorithm proved to be complex because it needed to assess pharmacokinetics and drug notoriety. However, ALDEN provides a better ability to discriminate between “culprit” and “innocent” drugs.

Table 5.

Details of the ALDEN.

Criterion	Values	Rules to apply	Score
Delay from initial drug component intake to onset of reaction (index day)	Suggestive +3	From 5 days to 28 days	–3 to 3
	Compatible +2	From 29 days to 56 days
	Likely +1	From 1 days to 4 days
	Unlikely –1	>56 days
	Excluded –3	Drug started on or after the index day
		In case of previous reaction to the same drug, only changes for;Suggestive: +3: from 1 days to 4 daysLikely: +1: from 5 days to 56 days
Drug present in the body on index day	Define 0	Drug continued up to index day or stopped at a time point less than five times the elimination half-life* before the index day	–3 to 0
	Doubtful –1	Drug stopped at a time point prior to the index day by more than five times the elimination half-life* but liver or kidney function alterations or suspected drug interactions† are present
	Excluded –3	Drug stopped at a time point prior to the index day by more than five times the elimination half-life,* without liver or kidney function alterations or suspected drug interactions†
Prechallenge/rechallenge	Positive specific for disease and drug: 4	SJS/TEN after use of same drug	–2 to 4
	Positive specific for disease or drug: 2	SJS/TEN after use of similar‡ drug or other reaction with the same drug
	Positive unspecific: 1	Other reaction after use of similar‡ drug
	Not done/unknown: 0	No known previous exposure to this drug
	Negative –2	Exposure to this drug without any reaction (before or after reaction)
Dechallenge	Neutral 0	Drug stopped (or unknown)	–2 to 0
	Negative –2	Drug continued without harm
Type of drug (notoriety)	Strongly associated 3	Drug of the “high risk” list according to previous case-control studies§	–1 to 3
	Associated 2	Drug with definite but lower risk according to previous case-control studies§
	Suspected 1	Several previous reports, ambiguous epidemiology results (drug “under surveillance”)
	Unknown 0	All other drugs including newly released ones
	Not suspected –1	No evidence of association from previous epidemiology study§ with sufficient number of exposed controls‡
		Intermediate score = total of all previous criteria	–11 to 10
Other cause	Possible –1	Rank all drugs from highest to lowest in intermediate score	–1
		If at least one has an intermediate score >3, subtract 1 point from the score of each of the other drugs taken by the patient (another cause is more likely)
Final score –12 to 10

<0, Very unlikely; 0–1, unlikely; 2–3, possible; 4–5, probable; >6, very probable. ^*Drug (or active metabolite) elimination half-life from serum and/or tissues (according to pharmacology textbooks), taking into account kidney function for drugs predominantly cleared by kidney and liver function for those with high hepatic clearance. ^†Suspected interaction was considered when more than five drugs were present in a patient’s body at the same time. ^‡Similar drug = same ATC code up to the fourth level (chemical subgroups). ^§Definitions for “high risk”, “lower risk”, and “no evidence of association”. “High risk” drugs include sulfamethoxazole-trimethoprim, sulfonamide anti-infectives, allopurinol, CBZ, phenytoin, phenobarbital, oxicam-NSAIDs. “Lower risk” drugs include acetic acid NSAIDs, macrolides, quinolones, cephalosporins, tetracyclines, aminopenicillins. Drugs with “no evidence of association” with SJS/TEN include beta blockers, angiotensin-converting-enzyme inhibitors, calcium channel blockers, thiazide diuretics, furosemide, propionic acid NSAIDs, sulfonylurea antidiabetics, insulin. ALDEN: Algorithm of Drug causality in Epidermal Necrolysis; ATC: Anatomical therapeutic chemical; CBZ: Carbamazepine; NSAIDs: Non-steroidal anti-inflammatory drugs; SJS: Stevens–Johnson syndrome; TEN: Toxic epidermal necrolysis.

Genetic Factors

A burgeoning body of evidence suggests a substantial genetic predisposition to cADRs. In 2004, Chung et al^[15] elucidated a robust correlation between human leukocyte antigen (HLA)-B*15:02 and carbamazepine (CBZ)-induced SJS/TEN in a Han Chinese cohort. HLA alleles, categorized into class I and class II, play a pivotal role in presenting antigenic peptides to T cells, thereby initiating an immune response. In a study comprising 44 individuals with CBZ-induced SJS/TEN, all participants were found to possess the HLA-B*15:02 allele. Subsequent investigations have substantiated the association between CBZ-induced SJS/TEN and the HLA-B*15:02 allele across diverse Asian populations, including those in China, Thailand, Malaysia, and India.^[16–27] A recent meta-analysis also revealed that the HLA-B*15:02 significantly increased the risk of CBZ-induced SJS/TEN.^[28]

The association of HLA-B*15:02 extends to other aromatic antiepileptic drugs such as phenytoin, lamotrigine, and oxcarbazepine, albeit with a lower incidence compared to CBZ.^[18,29–33] Notably, no such association has been observed in Japanese, Korean, and European populations,^[34–40] suggesting ethnic variations in HLA-related susceptibilities to SJS/TEN.

Other than HLA-B*15:02, HLA-B*15:11 and HLA-A*31:01 are associated in CBZ-induced cADRs. In the Japanese and Korean population, HLA-B*15:11 is also associated with CBZ-induced SJS/TEN.^[35,38] HLA-A*31:01 has been reported to associate with CBZ-induced cADRs among several populations.^{[26,38,41,42]} In a meta-analysis among several populations, HLA-A*31:01 showed a significant association with CBZ-induced DIHS/drug reaction with eosinophilia and systemic symptoms (DRESS) but not for CBZ-induced SJS/TEN.^[43] In addition, genome-wide association studies for CBZ-induced cADRs in the Japanese population revealed that HLA-A*31:01 was not associated with SJS/TEN, but associated with DIHS/DRESS, maculopapular exanthema, and erythema multiforme.^[44] These studies suggested that the type of cADRs might be determined by genetic variant.

In 2008, the U.S. Food and Drug Administration (FDA) recommended genotyping for HLA-B*15:02 prior to the administration of CBZ.^[45] The implementation of HLA-B*15:02 screening has markedly reduced the incidence of CBZ-induced SJS/TEN.^[46] Similarly in Japan, the effectiveness of HLA-A*31:01 screening for CBZ-induced SJS/TEN was assessed. The incidence of CBZ-induced cADR was significantly reduced when patients who tested negative for HLA-A*31:01 were prescribed CBZ and those who tested positive for HLA-A*31:01 were prescribed an alternative drug.^[47]

Other pharmaceuticals, such as allopurinol and abacavir, also exhibit significant associations with specific HLA alleles. Allopurinol, a primary etiological agent for SJS/TEN, has been linked to HLA-B*58:01 across multiple ethnic groups, including Chinese, Japanese, Korean, Thai, and European populations.^{[35,36,39,48–55]} Genotyping for HLA-B*58:01 has proven effective in preventing allopurinol-induced SJS/TEN. Cost-effectiveness analyses suggest that HLA-B*58:01 screening is economically advantageous.^[56] In the study from the United States suggested that this screening is cost-effective for Asian and African American populations but not for Caucasians or Hispanics.^[57]

Abacavir, used in the treatment of human immunodeficiency virus (HIV) infection, is associated with SJS/TEN in patients carrying HLA-B*57:01.^[58–62] Since 2008, HLA-B*57:01 screening was recommended in clinical care guidelines to reduce the risk of hypersensitivity reaction from abacavir.^[63] It has significantly reduced the incidence of abacavir-induced SJS/TEN.^[64] However, many patients have yet to undergo this screening, and expanding its use is anticipated to further decrease the incidence of abacavir-induced SJS/TEN. Regarding anti-HIV drugs, nevirapine has also been shown to have HLA association. In Malawian cohort, HLA-C*04:01 was found to markedly increase the risk for nevirapine-induced SJS/TEN.^[65]

It has been shown that HLA is associated not only with the onset of cADR but also with the occurrence of complications. In Thai population, HLA-B*44:03 was revealed to strongly associate with acetaminophen-related SJS/TEN patients who developed severe ocular complications.^[66] In addition, HLA-A*02:06 and HLA-B*13:01, HLA-C*14:03 were associated with severe ocular complications in acetaminophen-related SJS/TEN in Japanese patients.^[67,68]

Cytochrome P450 (CYP) enzymes, integral to drug metabolism, also play a crucial role in cADRs. The CYP450 gene family comprises 57 isoforms, each exhibiting distinct functional characteristics. Patients with polymorphisms leading to slow drug metabolism face an elevated risk of adverse drug reactions.^[69] Chung et al^[70] identified significant genetic factors associated with phenytoin-induced SJS/TEN, including 16 notable single nucleotide polymorphisms (SNPs) in CYP2C9. Patients with phenytoin-induced SJS/TEN carrying the CYP2C9*3 allele exhibit delayed phenytoin clearance, thereby exacerbating disease severity.

Pathogenesis and Diagnostic Biomarkers

Immunopathogenesis

SJS/TEN are traditionally regarded as T-cell-mediated disorders. T cells are activated through the binding of drugs to T cell receptors (TCRs) presented by antigen-presenting cells (APCs). Currently, three primary hypotheses elucidate the mechanisms of T cell activation [Figure 3]:^[71] hapten/pro-hapten model, pharmacological interaction (p-i) concept, and altered peptide model. Most drugs and their metabolites function as pro-haptens and do not act as haptens independently. They gain immunogenicity by covalently binding to carrier proteins (hapten antigens). These hapten antigens form complexes with HLA molecules in APCs and are recognized by TCRs, triggering drug-specific T cell activation. In this model, antigenic drugs are covalently attached to peptides presented by HLA molecules to TCRs.^[72–75] However, several drugs, including lamotrigine, celecoxib, CBZ, and sulfamethoxazole, can non-covalently bind directly to HLA and/or TCRs, a mechanism known as the p-i concept.^[76–80] Generally, HLA polymorphisms depend on the structure of antigen-binding cleft. It has been documented that unmodified abacavir binds to HLA-B*57:01 antigen-binding cleft, altering its shape and chemistry and thereby changing the repertoire of endogenous peptides that can bind to HLA-B*57:01 (altered peptide model).^[81,82] The TCR profile is also crucial in the pathogenesis of SJS/TEN. Ko et al^[83] reported that 84% of patients with CBZ-associated SJS/TEN had VB-11-ISGSY clonotype, but the clonotype absent in CBZ-tolerant patients.^[84] This clonotype specificity has also reported in oxypurinol-induced SJS/TEN.^[85] Recently, Pan et al^[86] employed next-generation sequencing to explore the TCR repertoire and identified a public αβTCR from the cytotoxic T cells of patients with CBZ-induced SJS/TEN, capable of binding CBZ and mediating an immune response.

Figure 3.

Models of T cell activation in SJS/TEN. (A) Hapten/pro-hapten model. Drugs or drug metabolites form a complex with carrier proteins and are presented as HLA. (B) p-i concept. Drugs directly bind to HLA and TCR non-covalently. (C) Altered peptide model. Drugs bind to the pocket of HLA resulting alteration of HLA-binding peptide repertoire. APC: Antigen-presenting cells; HLA: Human leukocyte antigen; SJS: Stevens–Johnson syndrome; TCR: T cell receptor; TEN: Toxic epidermal necrolysis.

In the initial stages, cytotoxic CD8+ T cells predominantly infiltrate epidermis and blister fluid, while CD4+ T cells primarily infiltrate in the dermis.^[87,88] Monocytes are also present in the epidermis of TEN patients, and during the later stages, a decrease in lymphocytes alongside an increase in monocytes is observed. Tohyama et al^[89] demonstrated that monocytes play a crucial role in keratinocyte death, likely by activating the cytotoxicity of CD8+ T cells. Elevated levels of soluble IL-2 receptors have been observed in the blister fluid and serum of SJS/TEN patients.^[90] Soluble IL-2 receptors serve as markers of T cell activation, underscoring the significance of activated cytotoxic CD8+ T cells in the keratinocyte death of SJS/TEN.

Keratinocyte death

The epidermal damage observed in the skin lesions of SJS/TEN patients is attributed to apoptotic mechanisms.^[91] Apoptosis in SJS/TEN is mediated by cytotoxic CD8+ T cells via the Fas-Fas ligand (FasL) pathway, the perforin/granzyme pathway, or through granulysin [Figure 4].

Figure 4.

Apoptosis pathway in SJS/TEN. In SJS/TEN, cytotoxic T cell induces apoptosis through the FasL pathway, perforin/granzyme pathway, or granulysin. FasL: Fas ligand; SJS: Stevens–Johnson syndrome; TEN: Toxic epidermal necrolysis.

FasL produced by cytotoxic CD8+ T cells and natural killer (NK) cells binds to Fas receptors on target cells. This binding activates the caspase cascade, culminating in cell apoptosis.^[92] Under normal conditions, Fas is present on the cell surface, while FasL is expressed intracellularly. FasL is translocated to the cell surface when the cell is primed for self-destruction.^[93] Viard et al^[94] demonstrated that FasL is present on the cell surface of keratinocytes in TEN patients, a feature not observed in keratinocytes from patients with maculopapular drug reactions. Moreover, elevated levels of soluble FasL (sFasL) were detected in the serum of TEN patients, and sFasL also has the capacity to mediate apoptosis.^[95]

We demonstrated that serum FasL levels are elevated in TEN patients.^[96,97] In this study, it is shown that peripheral blood mononuclear cells (PBMCs) produce sFasL upon exposure to causative drugs. sFasL released from PBMCs binds to Fas expressed on keratinocytes, inducing apoptosis. This finding suggests that increased serum sFasL levels may serve as a valuable diagnostic marker for SJS/TEN. However, a definitive correlation between sFasL levels and disease severity has not yet been established.^[97,98]

Nassif et al^[99,100] highlighted the significance of the perforin/granzyme pathway. Upon recognizing a target cell, the cytotoxic CD8+ T cells release perforin and granzyme B.^[92] Their study reported that mononuclear cells present in the blister fluid of SJS/TEN patients reveled cytotoxic effects when exposed to a causative drug. This cytotoxicity is inhibited by a perforin/granzyme pathway inhibitor, suggesting that this pathway is responsible for the epidermal damage seen in the skin lesions of SJS/TEN.^[99,100]

In 2008, Chung et al^[101] revealed the apoptosis-inducible effect of granulysin in keratinocyte of SJS/TEN. Granulysin, a pro-apoptotic protein, facilitates cell-mediated cytotoxicity without direct cell-to-cell interaction. Elevated levels of granulysin are detected in SJS/TEN blister fluid. Granulysin is released from blister cells in skin lesions of SJS/TEN, including NK cells and cytotoxic CD8+ T cells. The severity of the skin symptoms correlates with serum granulysin levels. Our research also identified that granulysin could be an early diagnostic marker for SJS/TEN.^[102] However, serum granulysin is also increased in patients with DIHS/drug reactions with eosinophilia and systemic symptoms, another category of severe cADRs.^[103] Therefore, granulysin is not a specific biomarker for SJS/TEN.

In recent years, the role of exosomes has been attracting attention. Exosomes are small vesicles with a diameter of 50–150 nm containing DNA, miRNA, mRNA, and proteins. Exosomes are secreted by cells and play an important role in cell-to-cell communication including immune response, signal transduction, and antigen presentation. Exosomes are involved not only in the maintenance of homeostasis but also in the pathogenesis of various diseases such as cardiovascular diseases, neurodegenerative diseases, and malignancies.^[104] Zhang et al^[105] revealed that the expression of miRNA, miR-375-3p, was markedly elevated in the plasma exosomes of SJS/TEN patients, which promotes epidermal apoptosis induction by downregulating the expression of the X-linked inhibitor of apoptosis (XIAP).

In 2014, our research elucidated that necroptosis, induced by the interaction between annexin A1 and formyl peptide receptor 1 (FPR1), associates with keratinocyte death in SJS/TEN.^[106] Necroptosis represents a form of programmed cell death that manifests with morphological characteristics akin to necrosis. Unlike apoptosis, necroptotic cells release damage-associated molecular patterns (DAMPs), which include various pro-inflammatory cytokines, thereby inciting inflammation. In contrast, apoptotic cells are rapidly phagocytosed by macrophages and subsequently degraded within phagolysosomes, a process devoid of inflammatory response.^[107] Generally, necroptosis is induced by the stimulation of tumor necrosis factor (TNF)-α when apoptosis is inhibited. Upon TNF-α stimulation, receptor-interacting protein kinases 1 (RIP1) and receptor-interacting protein kinases 3 (RIP3) undergo phosphorylation and assemble into a “necrosome” complex. Additionally, the mixed lineage kinase domain-like (MLKL) pseudokinase is recruited to the necrosome and phosphorylated by RIP3. The phosphorylated MLKL (pMLKL) then translocates to the plasma membrane, instigating cell death.^[107] Supernatant from PBMCs exposed to the causative drug in SJS/TEN patients has been observed to induce cytotoxicity in SJS/TEN keratinocytes. This cytotoxic effect is inhibited by a specific inhibitor of RIP1, necrostatin-1. Keratinocyte in SJS/TEN skin lesions exhibits significant expression of FPR1. Lipocalin-2 released from CD8+ T cells initiates the formation of neutrophil extracellular traps (NETs). Neutrophils undergoing NETosis release LL-37, which induces FPR1 expression in keratinocytes.^[108] The interaction of annexin A1, secreted by monocytes, with FPR1 facilitates necrosome formation [Figure 5]. Inhibition of necroptosis effectively prevents SJS/TEN-like responses in a murine model of SJS/TEN.^[106,109] These findings suggest that necroptosis plays an essential role in the pathogenesis of SJS/TEN.

Figure 5.

Necroptosis pathway in SJS/TEN. Drug-stimulated monocytes secrete annexin A1. Annexin A1 binds to FPR1 and RIP1 and RIP3 form necrosome and MLKL is phosphorylated by RIP3. pMLKL is located on plasma membrane and induces cell death. FPR1: Formyl peptide receptor 1; MLKL: Mixed lineage kinase domain-like; pMLKL: Phosphorylated MLKL; RIP1: Receptor-interacting protein kinases 1; RIP3: Receptor-interacting protein kinases 3; SJS: Stevens–Johnson syndrome; TEN: Toxic epidermal necrolysis.

TNF-α also plays an important role in keratinocyte death in SJS/TEN. When TNF-α binds to TNF receptor 1, apoptosis is induced if caspase 8 is activated, while necroptosis is induced if caspase 8 is inactivated.^[107] CTLs release TNF-α in addition to granulysin and sFasL. Activated keratinocytes and macrophages also release TNF-α. In fact, high concentrations of TNF-α are found in the serum and blister fluid of SJS/TEN patients.^[110,111] TNF-α is involved not only in this direct induction of cell death but also indirectly. TNF-α upregulates granulysin promoter activity and increases granulysin mRNA expression.^[112] In SJS/TEN patients, TNF-α and IFN-γ also upregulate FasL expression and induce keratinocyte apoptosis via upregulation of iNOS.^[113]

Keratinocyte death in SJS/TEN is a mixture of apoptosis and necroptosis, and it is of interest because it has not been clarified which form of cell death is the key phenomenon in the pathogenesis of SJS/TEN, and how the balance between apoptosis and necroptosis is regulated. In addition, recently, other types of cell death including pyroptosis and ferroptosis have been investigated. It has not been analyzed whether these cell deaths are involved in the pathogenesis of SJS/TEN.

Future research is expected to elucidate the pathogenesis of cell death in SJS/TEN, which may lead to the development of new therapeutic agents.

Diagnostic biomarkers

Although SJS/TEN are severe diseases, the clinical manifestations in early stage are often indistinguishable from those of erythema multiforme and maculopapular exanthema. Although no reliable biomarkers for diagnosis or severity prediction have been established, several biomarkers have been identified promising diagnostic or prognostic biomarkers for SJS/TEN, which are currently in the research phase and have not yet been applied clinically.

Wang et al^[114,115] revealed serum CCL-27 levels are increased in SJS/TEN patients, correlating with disease activity. CCL-27 is implicated in cutaneous inflammatory diseases by regulating T cell trafficking to the skin.

In addition, CCL-27 was highly expressed in the skin lesions of SJS/TEN patients.^[116] Therefore, it is hypothesized that CCL-27 is produced by keratinocytes in the SJS/TEN skin lesion and released into the circulation.^[117]

Su et al^[118] identified interleukin-15 (IL-15) as a factor related to mortality and severity of SJS/TEN patients. They further revealed that IL-15 exacerbates TEN severity by enhancing T cell- and NK-cell-mediated responses. IL-15 induces the production of TNF-α and downstream cytokines/chemokines,^[119] suggesting that the elevation of numerous cytokines and chemokines in SJS/TEN might be secondary to IL-15 activity.

Through proteomics analysis, we identified galectin-7 as a diagnostic biomarker for SJS/TEN.^[120] We hypothesized that certain soluble factors might be secreted exclusively by drug-specific lymphocytes in SJS/TEN patients, differentiating them from those with non-severe cADRs. These soluble factors could serve as biomarkers for SJS/TEN. PBMCs from SJS/TEN patients were cultured with the causative drugs, and the supernatant was collected for proteomic analysis of the elevated proteins.^[121] Hama et al^[120] concluded that this approach facilitated the identification of novel SJS/TEN-specific biomarkers not associated with the pathogenesis.

Focusing on the mechanisms of epidermal necroptosis, we identified serum RIP3, a key mediator of necroptosis, as a diagnostic and severity marker.^[122] It has been reported that RIP3 expression increases in necroptotic cells.^[123] We demonstrated that RIP3 expression also increases in necroptotic keratinocytes, and serum RIP3 levels are elevated in the acute phase of SJS/TEN patients, potentially correlating with disease activity.

Lipocalin-2 is also a biomarker in relation to the mechanism of necroptosis.^[108] Lipocalin-2 is associated with early stage of necroptosis in SJS/TEN. It is released from CD8+ T cells and induces NETs formation, resulting in FPR1 expression. The serum levels of lipocalin-2 were elevated in early stage of SJS/TEN, but not in other types of cADRs.

Management

In patients with SJS/TEN, the epidermis and mucosal membranes are mainly affected. However, SJS/TEN can also induce various organ complications, such as the kidneys, liver, and respiratory tract. Consequently, a multidisciplinary evaluation and prompt intervention within a specialized hospital setting including intensive care unit or burn care unit are pivotal in enhancing survival outcomes. Initially, immediate cessation of the suspected causative drugs is imperative for SJS/TEN. Furthermore, supportive care, encompassing supplemental oxygen, nutritional assessment, fluid replacement, pain management, and supplemental oxygen is essential. Given that infection resulting from skin detachment is a common complication in SJS/TEN patients, which impairs re-epithelialization and has potential progression to sepsis, meticulous daily skin care is necessary. The blister fluid should be aspirated, and the detached skin retained to cover the dermis as a biologic dressing. However, it is necessary to remove necrotic epidermis. In the areas of skin erosion, non-adherent dressings should be applied. There is no evidence to suggest which dressing is superior, but antibiotic ointments and petrolatum are commonly used.^[124] Antibiotic agent should be administrated if cutaneous infection is clinically suspected.^[125]

The optimal therapeutic strategy for SJS/TEN remains a subject of debate.^[126] In 2016, UK guidelines emphasized the prioritization of culprit drug withdrawal and comprehensive supportive care over systemic treatment due to the lack of evidence supporting the benefits.^[127] However, some studies have reported benefits with the treatment of systemic corticosteroids, intravenous immunoglobulins (IVIGs), cyclosporine, TNF-α antagonists (infliximab and etanercept), and plasmapheresis (PP).^[124,128]

The effectiveness of systemic therapy may be contingent upon the disease phase. For instance, during the acute phase, immunosuppressive treatments may be appropriate due to the occurrence of a robust inflammatory response resembling a “cytokine storm”. Conversely, during the peak period characterized by extensive skin detachment, strong immunosuppressive treatment might impede re-epithelialization and elevate the risk of infection. Previous studies have not adequately considered this aspect, often including results from all phases, thereby leading to inconsistent conclusions. We will introduce each treatment modality in detail below.

Systemic corticosteroids

Previous studies have demonstrated that corticosteroid treatment in SJS/TEN patients is associated with an increased risk of infection and overall complications, including higher mortality rates.^[129–131] Comprehensive analyses and systematic reviews have not shown a survival benefit with the use of systemic corticosteroids.^[132–134]

However, recent investigations suggested a beneficial effect of corticosteroid therapy. A European multicenter retrospective study, along with recent meta-analysis of observational studies, indicated positive effects of corticosteroids.^[128,135] A recent survey in Japan showed that the ratio of expected to observed mortality, calculated using SCORTEN score, was lowest with high-dose of corticosteroid therapy, followed by steroid pulse therapy.^[3] Given that cutaneous infection is a critical concern in the administration of corticosteroids for SJS/TEN patients, short-term corticosteroid use, enhanced infection control, and meticulous wound management are essential to reduce mortality rates.

Steroid pulse therapy is used in patients with severe cases of SJS/TEN that progress rapidly.^[136] An observational study revealed that the early administration of steroids pulse therapy reduced mortality without increasing infection risk.^[137] Notably, it is reported steroid pulse therapy at disease onset could prevent ocular complications.^[138]

IVIG

IVIG has been widely administrated to patients with SJS/TEN. Although the precise mechanisms of IVIG therapy remain unclear, it has been suggested that IVIG inhibits apoptosis by blocking Fas receptor with anti-Fas antibodies.^[139] Although some case reports have concluded that IVIG does not confer a beneficial effect in reducing mortality,^[140–142] other reports have shown the beneficial effects of IVIG for SJS/TEN patients.^[143–147] The largest retrospective study, the European Study of Severe Cutaneous Adverse Reactions (EuroSCAR), found that IVIG did not improve mortality compared with supportive care alone.^[135] However, recent meta-analyses have shown that high-dose IVIG (<2 g/kg) has a beneficial effect in decreasing the mortality of SJS/TEN.^[148] Recently, there is increasing evidence that combination of IVIG and corticosteroid therapy, rather than IVIG monotherapy, improves the prognosis of SJS/TEN.^[148,149] A retrospective study from China revealed combination therapy was prone to bring a better prognosis for SJS/TEN patients compared with those treated by corticosteroids alone. In addition, the incidence of infections was significantly lower in the IVIG and steroid combined therapy group compared to patients treated with corticosteroids alone.^[150]

Cyclosporine

Cyclosporine, a calcineurin inhibitor, has demonstrated therapeutic efficacy in the treatment of SJS/TEN. This immunosuppressive agent modulates T-lymphocyte-mediated cytotoxicity and inhibits key molecules such as FasL, nuclear factor-kB, and TNF-α.^[151] Several case reports and meta-analyses have indicated that cyclosporine treatment enhances survival rates in patients with SJS/TEN.^[152–159] It is reported that cyclosporine not only improves mortality outcomes but also exhibits minimal adverse effects in the treatment of SJS/TEN.^[160] These findings suggest a promising role for cyclosporine in managing SJS/TEN. Nevertheless, the limited number of patients included in these studies necessitates further research to validate cyclosporine’s efficacy.

Plasmapheresis

Several case series have demonstrated the efficacy of PP in treating SJS/TEN.^[161–166] The primary objective of PP is to eliminate pathogenic factors, including drugs, drug metabolites, and disease-induced cytokines/chemokines, from the patient’s bloodstream. PP sessions are typically conducted every other day or daily. This procedure is generally considered safe, with few adverse side effects. Despite one observational study concluding that PP is ineffective, the overall survival rate in this study was 87.5%.^[167] Narita et al^[168] found that PP was effective in TEN patients who did not respond to supportive or systemic corticosteroid therapies and observed a reduction in serum cytokine levels following PP.

A study suggested that combining PP with IVIG therapy may provide additional benefits.^[169] However, another study reported unfavorable results for the combined therapy, while treatment with PP alone yielded positive outcomes.^[170] Consequently, randomized controlled trials are essential to further elucidate the efficacy of PP in SJS/TEN treatment.

Tumor necrosis factor inhibitors

As previously described, TNF-α is associated with keratinocyte death in SJS/TEN. TNF-α inhibitors have been used and, in some cases, beneficial effects have been suggested. In 2002, Fischer et al^[171] reported a TEN case recovered with a single dose of infliximab. Since then, many cases have been reported in which infliximab was effective in SJS/TEN,^[172–174] and some reports suggest that infliximab in combination with IVIG may be useful.^[175,176] Infliximab has been reported mainly in case reports, and no randomized controlled trials have been conducted. Etanercept is also expected to be a promising treatment. Etanercept improved clinical outcomes and mortality in patients with SJS/TEN in a randomized trial comparing the effects of etanercept and corticosteroids.^[177] Although there was no statistical significance, the actual mortality rate in the etanercept group was lower (8.3%) than that predicted by SCORTEN (17.7%). A recent multicenter observational study showed the efficacy of combination therapy with etanercept and systemic corticosteroids.^[178] Patients who received combination therapy with etanercept and corticosteroids had lower mortality than those with corticosteroid monotherapy and those with IVIG combined with corticosteroids.

Conclusions

This review encapsulates the recent advancements in understanding the pathophysiology, diagnosis, and treatment of SJS/TEN. SJS/TEN is a formidable disease with a high mortality rate, and its diagnostic methods and therapeutic algorithms have not been fully established. However, recent studies have increasingly focused on therapeutic approaches. In particular, TNF-α inhibitors are expected to be a promising treatment. In addition, research on the pathogenesis of SJS/TEN is gradually progressing. As the pathology is elucidated, it is hoped that a novel therapeutic target will be developed.

Funding

This work was supported by Grant-in-Aid for Scientific Research (KAKENHI) (No. 23K15264).

Conflicts of interest

None.