INTRODUCTION AND TERMINOLOGY — Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe mucocutaneous reactions, most commonly triggered by medications, characterized by extensive necrosis and detachment of the epidermis [1]. Mucous membranes are affected in over 90 percent of patients, usually at two or more distinct sites (ocular, oral, and genital).
According to a widely accepted classification, SJS and TEN are considered a disease continuum and are distinguished chiefly by severity, based upon the percentage of body surface affected by blisters and erosions (table 1) [2,3]:
●SJS is the less severe condition, in which skin detachment is <10 percent of the body surface (picture 1A-C).
●TEN involves detachment of >30 percent of the body surface area (BSA) (picture 2A-D).
●SJS/TEN overlap describes patients with skin detachment of 10 to 30 percent of BSA.
We will use the term "SJS/TEN" to refer collectively to SJS, TEN, and SJS/TEN overlap syndrome.
The pathogenesis, clinical manifestations, and diagnosis of SJS/TEN are discussed in this topic. Treatment, prognosis, and long-term complications are discussed separately. Other drug eruptions are reviewed elsewhere.
●(See "Drug reaction with eosinophilia and systemic symptoms (DRESS)".)
●(See "Acute generalized exanthematous pustulosis (AGEP)".)
●(See "Exanthematous (maculopapular) drug eruption".)
●(See "Fixed drug eruption".)
●(See "Lichenoid drug eruption (drug-induced lichen planus)".)
●(See "Acneiform eruption secondary to epidermal growth factor receptor (EGFR) and MEK inhibitors".)
EPIDEMIOLOGY — Estimates of incidence for Stevens-Johnson syndrome (SJS), toxic epidermal necrolysis (TEN), and SJS/TEN range from two to seven cases per million people per year [4-10]. However, data from a nationwide German registry indicate that the incidence is probably one to two per million per year [4,11]. SJS is more common, outnumbering TEN by as much as three cases to one [4,12]. In the United States, the estimated incidences of SJS, SJS/TEN, and TEN among children between 2009 and 2012 were 5.3, 0.8, and 0.4 cases per million children per year [13].
The incidence of SJS/TEN is much higher among HIV-infected individuals and patients with active cancer than in the general population [14,15]. (See 'HIV infection' below and 'Malignancy' below.)
SJS/TEN can occur in patients of any age. It is more common in women than in men, with a male to female ratio of approximately 1:2 [12].
The overall mortality rate among patients with SJS/TEN is approximately 30 percent, ranging from approximately 10 percent for SJS to up to 50 percent for TEN. Mortality continues to increase up to one year after disease onset.
In a survival analysis of a large cohort of SJS/TEN patients, the overall mortality rate was 23 percent at six weeks, 28 percent at three months, and 34 percent at one year [12]. When distinguishing patients by severity, the mortality rates at six weeks were 12 percent for SJS, 29 percent for SJS/TEN overlap, and 46 percent for TEN and increased at one year to 24 percent for SJS, 43 percent for SJS/TEN overlap, and 49 percent for TEN. Older age (>70 years) and presence of severe comorbidities, but not disease severity, were associated with mortality beyond three months of disease onset. (See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Management, prognosis, and long-term sequelae".)
ETIOLOGY
Drugs — Medications are the leading trigger of Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) in both adults and children. The risk of SJS/TEN seems to be limited to the first eight weeks of treatment. Drugs used for a longer time are unlikely to be the cause of SJS/TEN. The typical exposure period before reaction onset is four days to four weeks of first continuous use of the drug. However, approximately 25 to 33 percent of cases, and probably an even higher proportion of pediatric cases, cannot be clearly attributed to a drug [16,17].
In a large multinational case-control study including 379 cases of SJS/TEN and 1505 controls, the following agents or groups of agents were most commonly implicated (table 2) [18,19]:
●Aromatic antiseizure medications and lamotrigine
●Antibacterial sulfonamides (including sulfasalazine)
●Oxicam nonsteroidal anti-inflammatory drugs (NSAIDs)
In children, the medications most often associated with SJS/TEN are sulfonamide antimicrobials, phenobarbital, carbamazepine, and lamotrigine. A weak association with acetaminophen/paracetamol has also been reported [20,21] but remains doubtful, as these drugs are frequently given to treat the prodromal or early symptoms of the disease (fever, headache, malaise, burning eyes, burning mouth) [22]. Furthermore, a history of previous tolerated use of these drugs makes their association with SJS/TEN even more unlikely.
SJS/TEN is less commonly attributed to penicillins, particularly amoxicillin or ampicillin. When these drugs are implicated, it is important to reexamine the clinical history to make sure that the penicillin was not administered to treat symptoms that were likely manifestations of the prodrome of SJS/TEN, when the more likely causative drug was administered 4 days to 4 weeks prior to the onset of the prodrome.
Several conventional and targeted anticancer therapies have been associated with SJS/TEN, including thalidomide [23], capecitabine [24], afatinib [25], vemurafenib [26], tamoxifen [27], and immune checkpoint inhibitors (ipilimumab, pembrolizumab, nivolumab) [28,29]. However, some of these reports may reflect SJS/TEN-like reactions that differ from SJS/TEN clinically and/or histologically [30].
Mycoplasma pneumoniae infection — Infections, including Mycoplasma pneumoniae infection, are the next most common trigger of SJS/TEN, particularly in children [31-33]. A systematic review of case series and single case reports suggests that M. pneumoniae-associated cases may be characterized more often by moderate to severe involvement of two or more mucosal sites and sparse, or even absent, skin involvement [34,35]. However, it cannot be excluded that these cases of M. pneumoniae-associated mucositis (MIRM) with minimal or absent skin involvement are a separate entity or an atypical form of erythema multiforme major [36].
Other — In over one-third of SJS/TEN cases, no cause can be identified [16]. Rarely reported and debatable causes of SJS/TEN include vaccinations, systemic diseases, contrast medium, external chemical exposure, herbal medicines, and foods [3,37-39]. Whether drugs or drug metabolites that are present in foods as additives or contaminants are implicated in apparently idiopathic cases has not been determined [40].
Cases of SJS/TEN have been reported after bone marrow transplantation as a manifestation of acute graft-versus-host disease rather than as a drug-induced reaction [41-43]. Radiotherapy in addition to treatment with antiseizure medications (eg, phenytoin, phenobarbital, carbamazepine) or amifostine may trigger SJS/TEN, with lesions localized predominantly at sites of radiation treatment [44,45].
RISK FACTORS — Risk factors for Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) include HIV infection, genetic factors, underlying immunologic diseases or malignancies, and, possibly, physical factors (such as ultraviolet light or radiation therapy).
HIV infection — Patients with HIV infection have been reported to have an increased risk of drug reactions and a 100-fold higher risk of SJS/TEN than the general population [14]. The reasons for this susceptibility are not fully understood, although exposure to multiple medications, immune dysregulation, presence of concomitant infections, and polymorphisms of genes involved in specific drug metabolism may contribute [46-51]. A genome-wide association study in a cohort of 151 nevirapine-hypersensitive and 182 tolerant, HIV-infected Malawian adults found that human leukocyte antigen (HLA)-C*04:01 predisposes sub-Saharan Africans to nevirapine-induced SJS/TEN but not to other hypersensitivity reactions [52].
Malignancy — Patients with active malignancy have an increased risk of SJS/TEN [18,53,54]. The risk is highest for patients with hematologic cancers [8,55]:
●An analysis of data from a large cohort of over 100,000 cancer patients identified from electronic health records between 2002 and 2015 identified 20 cases of confirmed SJS/TEN, 30 of possible SJS/TEN, and 12 with previous SJS/TEN. The estimated average annual incidences were 5.7 and 14.9 per 100,000 per year for confirmed and possible cases, respectively [15]. These incidence rates are approximately 30 to 60 times higher than in the general population. The most common trigger were antibiotics, followed by anticonvulsants and antineoplastic agents.
●In a case-control study including 480 validated SJS/TEN patients and 1920 controls from the United Kingdom Clinical Practice Research Datalink, patients with active cancer had a two-fold increased risk of SJS/TEN compared with controls (odds ratio [OR] 2.01, 95% CI 1.27-3.18) [8].
Whether the increased risk is due to the malignancy itself, to the frequent immunocompromised state of cancer patients, or to the increased exposure of these patients to potentially causative medications remains uncertain.
Genetic factors — Although several HLA haplotypes have been implicated in drug-specific susceptibility in certain ethnic groups [56-58], genome-wide association studies on both Asian and European populations have failed to identify highly penetrant genetic risk factors associated with SJS/TEN across multiple drugs [59]. It is possible that multiple genetic factors with small effect size may be implicated.
HLA types — The risk of drug-induced SJS/TEN has been associated with certain human leukocyte antigen (HLA) types (table 3):
●HLA-B*15:02 – Patients with HLA-B*15:02 phenotype are at increased risk for SJS/TEN due to carbamazepine and other aromatic anticonvulsants (eg, oxcarbazepine, phenytoin, phenobarbital) [56,60]. The prevalence of this allele is significant, approximately 7 to 10 percent, in patients of Asian and South Asian ancestry. (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects".)
●HLA-B*15:11 – An association between HLA-B*15:11 and carbamazepine-induced SJS/TEN has also been reported in Asian populations [61]. (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects", section on 'Carbamazepine'.)
●HLA-A*31:01 – Genome-wide studies have also found an association between the HLA-A*31:01 allele and carbamazepine-induced severe cutaneous drug reactions, including SJS/TEN, in persons of Japanese, Indian, and European ancestry [57,62,63]. However, a subsequent multinational study and meta-analysis found that HLA-A*31:01 is strongly associated with carbamazepine-induced drug reaction with eosinophilia and systemic symptoms (DRESS) in European and Asian patients but shows a much weaker association with SJS/TEN [64]. (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects", section on 'Carbamazepine'.)
●HLA-A*24:02 – HLA-A*24:02 has been identified as an additional potential risk factor for SJS/TEN induced by carbamazepine, lamotrigine, and phenytoin in a southern Chinese Han population [65]. (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects".)
●HLA-B*13:01 – HLA-B*13:01 has been found associated with dapsone-induced severe skin reactions, including predominantly DRESS, but also SJS/TEN, in Thai patients [66].
●HLA-B*58:01 – A systematic review and meta-analysis found a significant association between HLA-B*58:01 and allopurinol-induced SJS/TEN in both Asian and non-Asian populations [58]. A subsequent case-control study demonstrated an association between HLA-B*58:01 and allopurinol-induced DRESS or SJS/TEN in a Thai population [67].
Genetic screening — The US Food and Drug Administration has suggested screening patients of Asian and South Asian ancestry for HLA-B*15:02 if use of carbamazepine or oxcarbazepine is under consideration [56,68-70]. One consensus panel has recommended screening all carbamazepine-naïve patients for the HLA-A*31:01 allele prior to starting treatment, regardless of race or ethnicity, since the allele is prevalent in most ethnic groups [71]. (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects", section on 'Carbamazepine' and "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects", section on 'Oxcarbazepine'.)
The clinical utility of testing for the HLA-B*58:01 allele prior to treatment with allopurinol has not been demonstrated in any population [72]. However, the Clinical Pharmacogenetics Implementation Consortium (CPIC) recommends that allopurinol should not be prescribed to patients who are known carriers of HLA-B*58:01 [73]. (See "Pharmacologic urate-lowering therapy and treatment of tophi in patients with gout", section on 'Allopurinol'.)
Genetic polymorphisms — Polymorphisms in the CYP2C19 gene, coding for a cytochrome P450 isoform, may confer an increased risk of SJS/TEN following the administration of phenobarbital, phenytoin, or carbamazepine [74]. Other CYP2C variants, including CYP2C9*3, known to be associated with reduced phenytoin clearance may be associated with an increased risk of severe cutaneous adverse reactions to phenytoin, including SJS/TEN and DRESS [75,76]. Other polymorphisms potentially associated with SJS/TEN include those in the interleukin (IL) 4 receptor gene [77,78], prostaglandin E receptor 3 gene [79], and a group of genes located in the 6p21 region (BAT1, HCP5, MICC, and PSORS1C1), which are in linkage disequilibrium with HLA-B*58:01 [80]. However, larger studies are needed to assess the role of genetic polymorphism in the pathogenesis of SJS/TEN.
Other factors — Other factors that may increase the risk of SJS/TEN include:
●High doses of medications [19].
●Systemic lupus erythematosus – Patients with systemic lupus erythematosus (SLE) appear to be at an increased risk of SJS/TEN [81]. In a retrospective review of 1366 patients with SJS/TEN, the prevalence of SLE was 1.2 percent, whereas the estimated prevalence in the general population is approximately 0.02 to 0.05 percent [82]. However, in lupus patients, the differentiation between SJS/TEN and an extreme phenotype of acute cutaneous lupus ("TEN-like" lupus erythematosus) may be difficult and requires accurate clinicopathologic correlation [82,83].
●Physical stimuli, such as ultraviolet light or radiation therapy, may be cofactors in some cases [84-86].
PATHOGENESIS — The pathologic mechanisms that induce skin damage in Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) are incompletely understood. Early studies of the immunophenotype of lymphocytes detected in the blister fluid of SJS/TEN lesions suggested a cell-mediated cytotoxic reaction against keratinocytes leading to massive apoptosis [87]. Subsequent studies demonstrated that cytotoxic T cells are drug specific, human leukocyte antigen (HLA) class I restricted, and directed against the native form of the drug rather than against a reactive metabolite [88-90].
Drugs can stimulate the immune system by directly binding to the major histocompatibility complex (MHC) class I and the T cell receptor. This results in the clonal expansion of a population of drug-specific cytotoxic T cells that kill keratinocytes directly and indirectly through the recruitment of other cells that release soluble death mediators, including granulysin [91]. (See "Drug allergy: Pathogenesis", section on 'Interaction of drugs with the immune system'.)
By directly binding to HLA class I peptide pouch, drugs can also alter the repertoire of peptides recognized as foreign [92] and make the HLA drug complex recognized as foreign [93]. This explains the severity of the reaction and its similarity to acute graft-versus-host disease. (See "Pathogenesis of graft-versus-host disease (GVHD)".)
Mediators of keratinocyte apoptosis — Drug-specific CD8+ cytotoxic T cells, along with natural killer (NK) cells, are thought to be the major inducers of keratinocyte apoptosis [88,89]. Various cytotoxic proteins and cytokines such as soluble Fas ligand, perforin/granzyme, tumor necrosis factor (TNF)-alpha, and TNF-related apoptosis-inducing ligand (TRAIL) have been proposed as mediators for the extensive keratinocyte apoptosis in SJS/TEN. However, it is now accepted that granulysin, a cytolytic protein found in cytotoxic T cells and NK cells, plays a key role in the pathogenesis of SJS/TEN [94].
Granulysin — Granulysin is a cytolytic protein produced and secreted by cytotoxic T lymphocytes, NK cells, and NK/T cells. The gene expression profiling of cells from five patients with SJS or TEN identified granulysin as the most highly expressed cytotoxic molecule [95]. Both fluid and cells from SJS/TEN patient blisters demonstrated cytotoxicity when incubated with keratinocytes, but the effect was reduced by depletion of granulysin. Control fluid or cells from patients with burns did not show such activity. The levels of granulysin in blister fluid correlated with the severity of disease. In addition, injection of granulysin from patient blisters into mouse skin caused dose-dependent blistering and necrosis.
The mechanism through which cytotoxic T lymphocytes and NK cells are stimulated to release granulysin is unknown. The results of one study suggest that the interaction between the CD94/NKG2C receptor on cytotoxic T lymphocytes and HLA-E, an MHC class Ib molecule, expressed by keratinocytes in patients with SJS or TEN may promote degranulation [96].
Interleukin-15 — Serum levels of soluble interleukin (IL) 15 are increased in patients with SJS/TEN and appear to be correlated with disease severity and in-hospital mortality [97]. IL-15 is a widely expressed cytokine produced by many cell types, including immune cells (monocytes, macrophages, and dendritic cells) and keratinocytes [1]. IL-15 has a key role in promoting and maintaining long-lasting cytotoxic T cell and NK cell responses. In in vitro models, the addition of exogenous IL-15 has been shown to increase the secretion of granulysin and, to a lesser extent, of granzyme from blister fluid cells from patients with TEN [97]. In addition, in an in vitro lymphocyte transformation test designed to reveal the sensitization of T cells to a specific drug, the suppression of IL-15 signaling with a monoclonal antibody reduced the drug-mediated lymphocyte activation [97].
These findings suggest that IL-15 may have an important role in the pathogenesis of SJS/TEN. However, further studies are needed to determine whether IL-15 expressed in lesional skin of SJS/TEN patients elucidates the mechanisms of IL-15-driven T and NK cell responses in SJS/TEN.
Other factors — Soluble Fas-ligand, perforin, TNF-alpha, TRAIL, and granzyme B, which are involved in distinct nonapoptotic cell death pathways, are also found in high concentrations in the peripheral mononuclear cells and blister fluid of SJS/TEN patients [98-100]. However, elevations in these mediators are not specific to SJS/TEN.
HISTOPATHOLOGY — The hallmark of Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) is the keratinocyte necrosis, ranging from partial to full-thickness necrosis of the epidermis. In early lesions, apoptotic keratinocytes are scattered in the basal layer of the epidermis, but in established lesions, full-thickness epidermal necrosis and subepidermal bullae may be seen [101]. A scant, perivascular, lymphohistiocytic, inflammatory infiltrate containing a variable amount of eosinophils is present in the superficial dermis (picture 3).
CLINICAL PRESENTATION
Prodrome — Fever, often exceeding 39°C (102.2°F), and influenza-like symptoms precede by one to three days the development of mucocutaneous lesions [102]. Photophobia, conjunctival itching or burning, and pain on swallowing may be early symptoms of mucosal involvement. Malaise, myalgia, and arthralgia are present in most patients.
In some patients, an exanthematous eruption can be the heralding sign of Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN). Signs and symptoms that should alert the clinician to the possibility of SJS/TEN include fever >38°C (100.4°F), mucositis, skin tenderness, and blistering (table 4) [102,103]. (See "Exanthematous (maculopapular) drug eruption", section on 'When to suspect a severe drug reaction'.)
Cutaneous lesions — The skin lesions typically begin with ill-defined, coalescing, erythematous macules with purpuric centers, although many cases of SJS/TEN may present with diffuse erythema (picture 1A, 2A, 2C-D) [104,105]. The skin is often tender to the touch, and skin pain can be prominent and out of proportion to the cutaneous findings.
Lesions start on the face and thorax before spreading to other areas and are symmetrically distributed [106]. The scalp is typically spared, and palms and soles are rarely involved [107,108]. Atypical target lesions with darker centers may be present. As the disease progresses, vesicles and bullae form, and within days the skin begins to slough.
Nikolsky sign (ie, the ability to extend the area of superficial sloughing by applying gentle lateral pressure on the surface of the skin at an apparently uninvolved site) may be positive. The Asboe-Hansen sign or "bulla spread sign" (a lateral extension of bullae with pressure) may also be present. The ultimate appearance of the skin has been likened to that of extensive thermal injury [109]. (See "Approach to the patient with cutaneous blisters", section on 'Nikolsky sign'.)
Mucosal lesions — Mucosal involvement occurs in approximately 90 percent of cases of SJS/TEN and can precede or follow the skin eruption [110]. Painful crusts and erosions may occur on any mucosal surface [102,111].
Oral — The oral mucosa and the vermilion border are almost invariably involved, with painful hemorrhagic erosions covered with a grayish-white membrane (picture 1C-D). Stomatitis and mucositis lead to impaired oral intake with consequent malnutrition and dehydration.
Ocular — Ocular involvement is reported in approximately 80 percent of patients. The most common change in the eyes is a severe conjunctivitis with a purulent discharge (picture 1E), but bullae may develop. Corneal ulceration is frequent, and anterior uveitis or panophthalmitis may occur. Pain and photophobia are accompanying symptoms.
A simple grading system based upon the presence of conjunctivitis, corneal or conjunctival epithelial defect, and pseudomembrane formation has been proposed to assess the severity of acute ocular involvement and guide the therapeutic choice [112]:
●No ocular involvement – 0 (none)
●Conjunctival hyperemia – 1 (mild)
●Either ocular surface epithelial defect or pseudomembrane formation – 2 (severe)
●Both ocular surface epithelial defect and pseudomembrane formation – 3 (very severe)
The eye changes may regress completely, but at least 50 percent of patients have late eye sequelae including pain, dryness, and scarring with the development of synechiae between the eyelids and conjunctiva [113,114]. (See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Management, prognosis, and long-term sequelae", section on 'Eyes'.)
Urogenital — Urethritis develops in up to two-thirds of patients and may lead to urinary retention. Genital erosions are frequent. In women, vulvovaginal involvement may present with erosive and ulcerative vaginitis, vulvar bullae, vaginal synechiae, and may lead to long-term anatomic sequelae. These include labial and vaginal adhesions and stenosis, obstructed urinary stream and urinary retention, recurrent cystitis, or hematocolpos [115-119]. Vulvovaginal adenosis (presence of metaplastic cervical or endometrial glandular epithelium in the vulva or vagina) also has been reported in women with SJS/TEN [120-122].
Other — Pharyngeal mucosa is affected in nearly all patients; tracheal, bronchial, and esophageal membranes are less frequently involved [123,124]. Intestinal involvement is rare [125].
Laboratory abnormalities — Hematologic abnormalities, particularly anemia and lymphopenia, are common in SJS/TEN [106]. Eosinophilia is unusual; neutropenia is present in approximately one-third of patients and is correlated with a poor prognosis [106,126]. However, the administration of systemic corticosteroids can cause demarginalization and mobilization of neutrophils into the circulation, and this may obscure neutropenia.
Hypoalbuminemia, electrolyte imbalance, and increased blood urea nitrogen and glucose may be noted in severe cases, due to massive transdermal fluid loss and hypercatabolic state. Serum urea nitrogen >10 mmol/L and glucose >14 mmol/L are considered markers of disease severity [127]. Mild elevations in serum aminotransferase levels (two to three times the normal value) are present in approximately one-half of patients with TEN, while overt hepatitis occurs in approximately 10 percent [111].
CLINICAL COURSE — The acute phase of Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) lasts 8 to 12 days and is characterized by persistent fever, severe mucous membrane involvement, and epidermal sloughing that may be generalized and result in large, raw, painful areas of denuded skin. Re-epithelialization may begin after several days and typically requires two to four weeks [128]. Skin that remained attached during the acute process may peel gradually, and nails may be shed.
COMPLICATIONS — In severe cases with extensive skin detachment, acute complications may include massive loss of fluids through the denuded skin, electrolyte imbalance, hypovolemic shock with renal failure, bacteremia, insulin resistance, hypercatabolic state, and multiple organ dysfunction syndrome. Abdominal compartment syndrome secondary to excessive replacement fluid therapy has been reported in a few patients [129]. (See "Abdominal compartment syndrome in adults".)
Infections — Patients with Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) are at high risk of bacterial infection. Sepsis and septic shock, most often caused by Staphylococcus aureus and Pseudomonas aeruginosa, are the main causes of death in these patients. In a study of 179 patients with SJS/TEN, bacteremia was detected in 48 (27 percent) [130]. The main pathogens implicated were S. aureus, P. aeruginosa, and Enterobacteriaceae organisms. The risk of bacteremia was higher in patients older than 40 years, in those with skin detachment on more than 30 percent of body surface area (BSA), and in those with greater than 10,000 white blood cell count. (See "Sepsis syndromes in adults: Epidemiology, definitions, clinical presentation, diagnosis, and prognosis".)
Pulmonary complications — Pulmonary complications (eg, pneumonia, interstitial pneumonitis) are frequent. Presenting symptoms include cough and elevated respiratory rate. Strict clinical surveillance is necessary for these patients, due to the risk of progression to acute respiratory distress syndrome. Acute respiratory failure requiring mechanical ventilation has been reported in approximately 25 percent of patients with SJS/TEN [123]. (See "Acute respiratory distress syndrome: Clinical features, diagnosis, and complications in adults".)
Gastrointestinal complications — Gastrointestinal complications may result from epithelial necrosis of the esophagus, small bowel, or colon. Diarrhea, melena, small bowel ulcerations, colonic perforation, and small bowel intussusception have been reported in a few patients [131-134].
Disseminated intravascular coagulation — There are several reports documenting disseminated intravascular coagulation (DIC) in patients with SJS/TEN [135,136]. In a series of 150 patients with SJS/TEN, of whom 32 (21 percent) had laboratory evidence of DIC, DIC was strongly associated with increased risk of gastrointestinal bleeding; respiratory, renal, and liver failure; infection; bacteremia; and death [135].
LONG-TERM SEQUELAE — Long-term sequelae of Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) are discussed separately. (See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Management, prognosis, and long-term sequelae", section on 'Long-term sequelae'.)
PATIENT EVALUATION AND DIAGNOSIS
Clinical findings and history — There are no universally accepted diagnostic criteria for Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN), and histologic findings are neither specific nor diagnostic. Despite these limitations, the diagnosis of SJS or TEN would be appropriate in a patient with the following clinical features [137] (see 'Clinical presentation' above):
●A suggestive history of drug exposure or febrile illness. Drug exposure commonly precedes the onset of symptoms by one to four weeks (average 14 days), but re-exposure may result in onset of symptoms in as little as 48 hours [138].
●A prodrome of acute-onset febrile illness and malaise.
●A painful rash that progresses rapidly.
●Erythematous macules, targetoid lesions, or diffuse erythema progressing to vesicles and bullae (picture 1A).
●Positive Nikolsky sign and/or "bulla spread sign."
●Oral, ocular, and/or genital mucositis with painful mucosal erosions (picture 1C, 1E).
●Necrosis and sloughing of the epidermis of varying degree (picture 2C).
Assessment of drug causality — For patients suspected to have SJS/TEN, the identification of the causative drug is essential because early withdrawal of the offending agent may improve the prognosis [139]. In addition, the identification of the culprit drug is of paramount importance to prevent re-exposure in patients recovering from SJS/TEN.
The assessment of drug causality is based upon detailed history and clinical judgement. Information about the drugs that are most frequently associated with SJS/TEN is helpful (table 2). An algorithm of drug causality for epidermal necrolysis (ALDEN) has been developed as a tool for rapid assessment of drug causality in patients presenting with SJS/TEN, particularly in those exposed to multiple medications [16]. Each potentially offending drug is assigned a score from -11 to 10 based upon six parameters (table 5):
●Time delay from initial drug intake to onset of reaction
●Probability of drug presence in the body on the index day
●A previous history of exposure to the same drug, with or without reaction
●Presence of the drug beyond the progression phase of the disease
●Drug notoriety as a cause of SJS/TEN based upon previous studies
●Presence or absence of other etiologic causes
The score is categorized as very probable (≥6), probable (4 to 5), possible (2 to 3), unlikely (0 to 1), and very unlikely (≤0).
Assessment of severity — The severity and prognosis of SJS/TEN depends upon the amount of skin detachment or "detachable" skin (ie, skin with positive Nikolsky sign). The extent of epidermal detachments should be estimated and recorded as the percentage of the body surface area (BSA) involved.
The correct evaluation of the extent of lesions may be difficult in areas with spotty lesions. It is useful to remember that in both children and adults the surface of the patient's hand, including palm and fingers, corresponds to approximately 1 percent of the total BSA. (See "Assessment and classification of burn injury", section on 'Extent of burn injury'.)
The prognosis of individual patients can be rapidly evaluated in the early stages of disease by applying a prognostic scoring system called SCORTEN, based upon seven clinical and laboratory variables (table 6) [127]. (See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Management, prognosis, and long-term sequelae", section on 'Assessment of prognosis'.)
Skin biopsy — A skin biopsy for routine histopathologic examination and possibly direct immunofluorescence is useful in confirming the diagnosis and excluding other conditions that may mimic SJS/TEN. An appropriate sample may be obtained by performing a large (>4 mm) punch biopsy or a deep shave biopsy ("saucerization").
In the early stages of disease, apoptotic keratinocytes are scattered in the basal layer of the epidermis, and there is a perivascular, mononuclear, inflammatory infiltrate in the papillary dermis composed primarily of T lymphocytes [101]. This infiltrate is not diagnostic and may be seen in a wide variety of conditions, including a simple drug-induced exanthem. As the lesions progress, frank subepidermal bullae develop, with full-thickness epidermal necrosis (picture 3).
Direct immunofluorescence is always negative.
Laboratory and imaging studies — The laboratory and imaging evaluation of patients presenting with SJS/TEN includes:
●Complete blood count with differential, metabolic panel (ie, glucose, electrolytes, blood urea nitrogen, creatinine, calcium, total protein, albumin, alkaline phosphatase, alanine aminotransferase, aspartate aminotransferase), erythrocyte sedimentation rate, and C-reactive protein.
●Bacterial and fungal cultures should be performed from blood, wounds, and mucosal lesions. Because of the high risk of bacterial superinfection and sepsis in these patients, cultures should be repeated throughout the acute phase of the disease.
●In children, polymerase chain reaction and/or serologies for M. pneumoniae infection should be obtained in the early stage of disease and three weeks later. (See "Mycoplasma pneumoniae infection in children", section on 'Diagnosis'.)
●A chest radiograph should be obtained in all patients, due to high risk of pneumonia and interstitial pneumonitis. (See 'Pulmonary complications' above.)
Investigational tests — Candidate serum markers of SJS/TEN, including soluble Fas ligand, soluble CD40 ligand, granulysin, interleukin (IL) 15, and high-mobility group box 1 protein (HMGB1; a nonhistone nuclear protein released by necrotic and apoptotic cells), have been evaluated in a few small studies [97,98,140-142]. However, further studies are necessary to determine whether these markers may be helpful in the early diagnosis of SJS/TEN. (See 'Pathogenesis' above.)
DIFFERENTIAL DIAGNOSIS — The differential diagnosis of Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) includes:
●Erythema multiforme – Erythema multiforme usually presents with typical target lesions or raised, atypical, targetoid lesions that are predominantly located on the extremities (picture 4). Bullae and epidermal detachment are usually limited and involve less than 10 percent of the body surface area (BSA) (table 1). In contrast with SJS/TEN, erythema multiforme is associated with infection with herpes simplex virus in approximately 90 percent of cases and only rarely with drugs. (See "Erythema multiforme: Pathogenesis, clinical features, and diagnosis".)
●Erythroderma and erythematous drug eruptions – The generalized and symmetric, maculopapular erythema of a drug eruption can mimic early SJS/TEN. However, exanthematous drug eruptions lack mucosal involvement and the prominent skin pain of TEN. Histology shows only a mild interface dermatitis with a perivascular, inflammatory infiltrate of lymphocytes, neutrophils, and eosinophils. (See "Erythroderma in adults" and "Exanthematous (maculopapular) drug eruption".)
●Acute generalized exanthematous pustulosis – Severe cases of acute generalized exanthematous pustulosis (AGEP) may be difficult to differentiate from SJS/TEN (picture 5). AGEP typically develops within a few days of exposure to the offending drug, most often a beta-lactam antibiotic, and resolves without treatment in one to two weeks after drug discontinuation. The histologic hallmark of AGEP is a spongiform, subcorneal, and/or intraepidermal pustule (picture 5). (See "Acute generalized exanthematous pustulosis (AGEP)".)
●Generalized bullous fixed drug eruption – Generalized bullous fixed drug eruption is an extremely rare form of fixed drug eruption characterized by widespread red or brown macules or plaques with overlying, large, flaccid bullae (picture 6). In contrast with SJS/TEN, mucosal involvement is usually absent. Resolution generally occurs in one to two weeks after drug discontinuation. (See "Fixed drug eruption".)
●Phototoxic eruptions – Severe phototoxic eruptions may be confused with SJS/TEN. Important clues to the correct diagnosis include recent sun exposure, known phototoxic properties of certain medications, and location of the lesions on sun-exposed areas. (See "Photosensitivity disorders (photodermatoses): Clinical manifestations, diagnosis, and treatment".)
●Staphylococcal scalded skin syndrome – Staphylococcal scalded skin syndrome (SSSS) is caused by epidermolytic toxins produced by certain strains of staphylococci and is usually seen in neonates and young children. SSSS presents with generalized erythema rapidly followed by the development of flaccid blisters and desquamation (picture 7). The mucous membranes are not involved. Histology reveals sloughing of only the upper layers of the epidermis, in contrast with the subepidermal split with full-thickness epidermal necrosis observed in SJS/TEN. (See "Vesicular, pustular, and bullous lesions in the newborn and infant", section on 'Staphylococcal scalded skin syndrome'.)
●Paraneoplastic pemphigus – Paraneoplastic pemphigus is a rare disorder that can represent the initial presentation of a malignancy or occur in a patient with a known neoplastic process, such as non-Hodgkin lymphoma in adults or Castleman's disease in children. Patients may develop severe mucocutaneous disease with ocular and oral blisters and skin lesions (picture 8A-C). (See "Paraneoplastic pemphigus".)
●Linear IgA bullous dermatosis – Linear IgA bullous dermatosis (LABD) is a rare autoimmune blistering disease that may mimic TEN (picture 9). Histology reveals a subepidermal blister with an underlying, neutrophil-predominant, dermal infiltrate. Direct immunofluorescence shows linear deposits of IgA along the basement membrane. (See "Linear IgA bullous dermatosis".)
●Chikungunya fever – An atypical, SJS/TEN-like form of chikungunya fever characterized by fever and generalized, vesicobullous eruption and superficial erosions has been reported in infants and young children [143,144]. However, in contrast with SJS/TEN, mucosal involvement is generally absent, and the resolution of the skin manifestations occurs in most cases in 4 to 10 days. (See "Chikungunya fever: Epidemiology, clinical manifestations, and diagnosis".)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Drug allergy and hypersensitivity".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topic (see "Patient education: Stevens-Johnson syndrome and toxic epidermal necrolysis (The Basics)")
SUMMARY AND RECOMMENDATIONS
●Definition – Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are rare, severe mucocutaneous reactions triggered, in most cases, by medications and characterized by extensive necrosis and detachment of the epidermis. Cases with less than 10 percent epidermal involvement are classified as SJS; those with 30 percent or more involvement are classified as TEN, and cases with 10 to 30 percent involvement are considered overlap SJS/TEN. However, we use the term "SJS/TEN" to refer collectively to SJS, TEN, and SJS/TEN overlap syndrome. (See 'Introduction and terminology' above.)
●Etiology and risk factors – Medications are the leading trigger of SJS/TEN in both adults and children. Allopurinol, lamotrigine, aromatic anticonvulsants, antibacterial sulfonamides, and "oxicam" or COX-2 inhibitor nonsteroidal anti-inflammatory drugs (NSAIDS) are most commonly implicated (table 2). Mycoplasma pneumoniae infection is the next most common trigger of SJS/TEN, particularly in children. Risk factors for SJS/TEN include HIV infection, genetic factors, concomitant viral infections, underlying autoimmune diseases, and, possibly, physical factors. (See 'Etiology' above and 'Risk factors' above.)
●Clinical presentation – SJS/TEN begins with a prodrome of fever and influenza-like symptoms one to three days before the development of mucocutaneous and skin lesions. The cutaneous eruption typically starts with ill-defined, coalescing, erythematous macules with atypical target lesions (picture 1A, 2A, 2C-D). As the disease progresses, vesicles and bullae form, and within days the skin begins to slough. Mucosal involvement occurs in approximately 90 percent of cases of SJS/TEN and can precede or follow the skin eruption. (See 'Clinical presentation' above.)
●Complications – In severe cases with extensive skin detachment, acute complications may include massive loss of fluids and electrolyte imbalance, hypovolemic shock with renal failure, bacteremia, insulin resistance, hypercatabolic state, and multiple organ dysfunction syndrome. (See 'Complications' above.)
●Diagnosis – The diagnosis of SJS/TEN is based upon clinical and histologic findings in a patient with a history of antecedent drug exposure or febrile illness. Histologic findings on skin biopsy are supportive but not independently diagnostic. (See 'Patient evaluation and diagnosis' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Milton H Nirken, MD, now deceased, who contributed to an earlier version of this topic review.
The UpToDate editorial staff also acknowledges Jean-Claude Roujeau, MD, who contributed to an earlier version of this topic review.
