Use of etanercept to treat toxic epidermal necrolysis in a human immunodeficiency virus-positive patient

Toxic epidermal necrolysis (TEN) is an uncommon and severe cutaneous adverse drug reaction that causes disseminated necrosis of epidermal cells and mucocutaneous detachment. Here, we report the case of a 32-year-old man with human immunodeficiency virus infection who presented with generalized violaceous macules and blister formation 4 days after the administration of mefenamic acid and amoxicillin for a dental procedure. Additional symptoms included oral ulcers and conjunctivitis. Results of skin biopsy were compatible with Stevens–Johnson syndrome (SJS). SJS progressed to TEN within 2 days. Etanercept treatment showed a dramatic improvement in the symptoms of mucocutaneous lesions. To our knowledge, this is the first report on the treatment of TEN using etanercept in a human immunodeficiency virus-positive patient.     

The CD40/CD40 ligand system is expressed in the cutaneous lesions of erythema multiforme and Stevens-Johnson syndrome/toxic epidermal necrolysis spectrum

Background Erythema multiforme (EM) and Stevens–Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN) are determined by a dysregulation of cellular immunity. Objectives To evaluate the effector role of cellular immunity and the involvement of the CD40/CD40 ligand (CD40L) system in the pathogenesis of EM and SJS/TEN. Methods Biopsy specimens from eight patients with EM and six with SJS/TEN were stained for immunohistochemical examination using the alkaline phosphatase/antialkaline phosphatase method. The monoclonal antibodies used included those to CD1a, CD4, CD8, CD40, CD40L, CD68, Fas, Fas ligand (FasL) and myeloperoxidase. Results The cellular infiltrate in both EM and SJS/TEN lesions was composed mainly of T lymphocytes and CD68+ macrophages. We also detected large amounts of neutrophils. Fas and FasL were very highly expressed in SJS and TEN, but weakly in EM. CD40 staining was strong in all tissue sections; there were numerous CD40L+ cells in SJS/TEN but much fewer in EM. Conclusions Activated T lymphocytes and macrophages, but also neutrophils, are presumably the main triggers of mucocutaneous damage in the SJS/TEN disease spectrum. The Fas/FasL system is significantly expressed in SJS/TEN lesions, but not in EM, where this apoptotic pathway presumably does not play a pivotal role in the epidermal damage. We suggest that the CD40/CD40L system may represent an important pathway of induction of SJS/TEN lesions, while in EM it would contribute to the immunoinflammation only as a second‐line mechanism.     

Expression of cytokines and chemokine receptors in the cutaneous lesions of erythema multiforme and Stevens-Johnson syndrome/toxic epidermal necrolysis

BACKGROUND: Erythema multiforme (EM) and Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) are caused by a dysregulation of cellular immunity. OBJECTIVES: To evaluate further the potential role of certain cytokines and chemokine receptors in cutaneous lesions of patients affected by EM and SJS/TEN and to establish whether such diseases are polarized preferentially towards a T-helper (Th) 1 or Th2 pattern. METHODS: Biopsy specimens from eight patients with EM, six with SJS/TEN and three healthy controls were stained for immunohistochemical examination using the alkaline phosphatase-antialkaline phosphatase method. The monoclonal antibodies used included those to tumour necrosis factor (TNF)-alpha, interferon (IFN)-gamma, interleukin (IL)-2, IL-5, IL-13, receptor 3 for C-C chemokines (CCR3), receptor 3 for C-x-C chemokines (CXCR3) and CXCR4. RESULTS: The SJS/TEN specimens showed significantly higher expression of all the cytokines and chemokine receptors (except CXCR3) tested than the EM specimens. Both lesional series showed significantly higher expression of all the receptors tested than the healthy control specimens, with the sole exception of a lower expression of CCR3 in EM specimens. Comparison between molecules associated with a Th1 or Th2 response revealed a predominance of Th1 response in EM and no significant imbalance between Th1 and Th2 in SJS/TEN. CONCLUSIONS: We have provided further evidence that TNF-alpha is strongly expressed in SJS/TEN lesions and therefore it may be involved in the epidermal necrosis featured in such diseases. IFN-gamma may play an important role both in EM and SJS/TEN. IL-2, IL-5 and IL-13 may contribute to the cutaneous immunoinflammation in these diseases. Chemokine receptors may be involved strongly in the recruitment of inflammatory cells in lesional skin. In our cases we found a sharp polarization towards a Th1 pattern in EM, while the SJS/TEN lesions showed a mixed Th1/Th2 pattern.     

Erythema multiforme,Stevens–Johnson syndrome and toxic epidermal necrolysis: Frozen‐section diagnosis

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) may be fatal. Although classified by body surface area skin detachment, initial stages of both may present with erythema multiforme (EM)‐like lesions. To diagnose and predict disease activity adequately as early as possible for patients revealing EM‐like lesions, we performed frozen‐section diagnosis. Thirty‐five patients clinically diagnosed as EM, SJS or TEN were biopsied to diagnose and predict disease progression within the initial‐visit day. Half of a histological section taken from a lesion was snap‐frozen and immediately cryostat‐sectioned, acetone‐fixed and stained with hematoxylin–eosin. Specimens were examined with light microscopy for presence of epidermal necrosis. A section from unaffected sites was also examined for 11 patients. Specimens were examined with light microscopy for presence of graft‐versus‐host reaction (GVHR)‐like findings: apoptotic keratinocytes and satellite cell necrosis. Epidermal necrosis was seen in nine patients. Initial diagnosis of the nine was one of overlap SJS‐TEN, four of SJS and four of EM, and final diagnosis of those was one of TEN, one of overlap SJS–TEN, four of SJS and three of EM. Dissociation between initial and final diagnosis was seen in three cases. GVHR‐like findings in the epidermis were observed in two patients finally diagnosed as overlap SJS–TEN and TEN. Frozen sections are useful not only to make a diagnosis of erythema multiforme but to assess a potential to exhibit more aggressive clinical behaviors (SJS or TEN).     

Anti-desmoplakin antibodies in erythema multiforme and Stevens-Johnson syndrome sera: pathogenic or epiphenomenon?

The pathophysiology of erythema multiforme (EM), Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN) is unclear. Whether autoantibodies against desmoplakin (Dp) I and II play a pathogenic role or result from an epitope spreading phenomenon is uncertain. Our aim was to characterize the keratinocyte antigens recognized in EM, TEN and SJS. Of 33 patients studied, 2 had TEN, 1 SJS, 9 EM major and 21 EM minor, according to Roujeau's criteria. All sera were studied by indirect immunofluorescence (IIF), immunoblotting and immunoprecipitation. Twenty normal sera were used as controls. 10/33 sera reacted with polypeptides of 215 and/or 250-kDa molecular mass, which co-migrate with Dp I and II as assessed by an anti-Dp I and II monoclonal antibody on IB. In IP, none of the anti-Dp I and -Dp II 10 patient sera immunoprecipitated Dp I and/or II from radiolabeled keratinocyte extracts. Two of 10 patient sera (SJS, EM minor) reacted with DpI and II when denaturated by the IB procedure. The reactivity against intracellular antigens DpI and II as denaturated proteins may result from the epidermal damage produced by aggressive autoreactive T cells, playing therefore only a secondary role in the pathogenesis of the disease.     

Long‐term complications of Stevens–Johnson syndrome/toxic epidermal necrolysis (SJS/TEN): the spectrum of chronic problems in patients who survive an episode of SJS/TEN necessitates multidisciplinary follow‐up

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe adverse reactions to drugs that cause a life‐threatening eruption of mucocutaneous blistering and epithelial sloughing. While the acute complications of SJS/TEN are well described, it is increasingly recognized that survivors may develop delayed sequelae, some of which can be associated with significant morbidity. Studies of long‐term SJS/TEN outcomes mostly focus on mucocutaneous and ocular complications. However, other internal organs, such as the respiratory tract and gastrointestinal tract, can be affected. Psychological sequelae are also frequent following the trauma of widespread epidermal necrolysis. An appreciation of the ‘chronic’ phase of SJS/TEN is needed by clinicians caring for individuals who have survived the acute illness. This review aims to provide an update on the breadth and range of sequelae that can affect patients in the months and years following an acute episode of SJS/TEN.     

Severe cutaneous reactions caused by barbiturates in seven Iranian children

Background  The severe adverse cutaneous reactions of erythema multiforme (EM), Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are rare mucocutaneous diseases associated with significant morbidity and mortality. The most common cause is antiepileptic drugs, particularly carbamazepine and lamotrigine, as well as the barbiturates group (phenobarbital and phenytoin). In this article, we present seven children with severe adverse cutaneous reactions caused by barbiturates.
Case Reports  The age of the affected children was between 2 and 11 years and they all had a history of taking barbiturates. Their symptoms started 1–3 weeks after the initiation of barbiturates, including a prodrome characterized by 2–3 days of malaise, fever, cough and anorexia, after which the skin and mucosal lesions appeared and worsened. The skin lesions varied from rash to large bullae, plus different forms of mucous membrane involvement. The offending drugs (barbiturates) were stopped immediately and care was largely supportive.
Conclusion  As a result of the morbidity and/or mortality associated with EM, SJS and TEN, physicians should keep in mind their differential diagnosis when cutaneous reactions are observed in patients undergoing barbiturate therapy. Furthermore, although TEN and SJS are life-threatening diseases, early detection and appropriate care can lead to a decrease in the incidence of death. The strategies described here seem to be successful and safe because, despite the serious conditions, our patients responded well. All survived.     

Erythema multiforme, Stevens–Johnson syndrome and toxic epidermal necrolysis in northeastern Malaysia

Background Previous studies have reported that drugs and infections are common causes of erythema multiforme (EM) and Stevens–Johnson syndrome (SJS). Toxic epidermal necrolysis (TEN) is mainly related to drugs. No study has been conducted in Kelantan, the northeastern state of Malaysia, to assess these cutaneous reactions. Methods A retrospective study of all hospitalized cases of EM, SJS, and TEN was conducted covering an 8-year period from 1987 to 1994. Results There were four cases (13.8%) of EM, 22 cases (75.9%) of SJS, and three cases (10.3%) of TEN. Drugs as a definitive cause was observed in one case (25%) of EM, 12 cases (54.5%) of SJS, and two cases (66.7%) of TEN. Drugs as a probable cause was observed in seven cases (31.8%) of SJS and one case (33.3%) of TEN. The male to female ratio was equal in EM and SJS. Antiepileptics were the commonest culprits, followed by antibiotics. One patient died of SJS and one patient died of TEN, giving mortality rates of 4.5% and 33.5% respectively. Fever was noted in 18 patients (62.1%). Leukocytosis was noted in 10 patients (34.5%), and nine patients (31.0%) had elevated liver transaminase enzymes. No significant correlation was noted between these biochemical changes and cutaneous eruption. Secondary infections were observed in 11 patients (37.9%): Staphylococcus aureus was the commonest isolated organism. Conclusions This study shows that drugs remain the commonest culprit in SJS and TEN. Despite adequate treatment, the mortality rate remains high, especially in TEN. These findings are similar to those of other reported studies.     

Recent advances in the genetics and immunology of Stevens-Johnson syndrome and toxic epidermal necrosis

Stevens-Johnson syndrome (SJS) and toxic epidermal necrosis (TEN) are rare but life-threatening severe cutaneous adverse reactions (SCARs), which are majorly (65-75%) induced by a variety of drugs. SJS/TEN could be recognized as SCARs or drug immune reactions, if the reactions are elicited by drugs. The recent studies suggested that SJS/TEN is a specific immune reaction initiated by the cytotoxic T lymphocytes (CTLs) via human leukocyte antigens (HLAs)-restricted pathway. The patho-mechanism involving HLA-restricted presentation of a drug or its metabolites for T-cell activation is supported by the findings of strong genetic associations with HLA alleles (e.g. HLA-B*15:02 and carbamazepine-SJS/TEN, and HLA-B*58:01 and allopurinol-SJS/TEN). However, the genetic associations of SJS/TEN or drug induced cutaneous immune reactions are complex, which are drug specific and ethnicity specific. The genetic polymorphisms and diversity of HLA alleles may provide different binding affinities for drug antigens to launch the activation of specific CTLs responses, further leading to the unique clinical manifestations in SJS/TEN. Fas-FasL and perforin/granzyme B have been advocated mediating the epidermal necrosis in SJS/TEN. Our recent study showed that granulysin, a cytotoxic protein produced by CTLs or natural killer (NK) cells, is the key mediator for disseminated keratinocyte death in SJS/TEN. From the point of view of a physician, the profounder understanding of the genetic predisposition and patho-mechanism we discover, the better strategies for prevention, clinical management, and therapeutic methods of SJS/TEN we can develop in the near future.     

Mycoplasma pneumoniae–Associated Mucocutaneous Disease in Children: Dilemmas in Classification

There is controversy regarding precise definitions for Stevens–Johnson syndrome (SJS) and erythema multiforme (EM) major because of overlap in clinical presentations. SJS and EM major associated with Mycoplasma pneumoniae have been reported to occur in children, but Mycoplasma is more commonly reported with SJS. We sought to further characterize Mycoplasma‐associated mucocutaneous disease. Through retrospective chart review over 10 years, six children hospitalized with a diagnosis of SJS who also tested positive for Mycoplasma infection were reviewed. Using documented physical examinations and photographs, diagnoses of SJS or EM major were retrospectively made based upon cutaneous lesional morphology employing the classification system proposed by Bastuji‐Garin et al. The majority of patients were boys, with limited acral cutaneous lesions. All patients required prolonged hospitalization because of mucosal involvement and had good short‐term outcomes. When the classification system was retrospectively applied, five of the six patients were reclassified with a diagnosis of EM major instead of SJS. Children with Mycoplasma‐associated EM major and SJS in our small retrospective series appeared to have significant mucosal involvement but more limited cutaneous involvement with lesional morphology, which is more characteristic of EM major.     

Long-term outcome of patients with severe cutaneous adverse reactions

Visceral involvement associated with Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) and drug-induced hypersensitivity syndrome/drug reaction with eosinophilia and systemic symptoms (DIHS/DRESS) is well documented. However, little is known about the long-term outcomes of severe drug eruptions due to a lack of long-term follow-up. Long-term sequelae may arise in patients who survive the acute complications of severe drug reactions. In SJS/TEN, extensive scarring that result from the healing of mucocutaneous ulcerative lesions may interfere with organ function. Severe sequelae include visual impairment and pulmonary obliterative disease that impair patients' quality of life. In DIHS/DRESS, recent observations suggest that fulminant type 1 diabetes mellitus (FT1D) and autoimmune diseases such as autoimmune thyroiditis and lupus erythematosus can occur after a disease-free period of several months to years. Thus, DIHS/DRESS may lead to the development of autoimmune diseases, which may be overlooked. Dermatologists need to be aware of the sequelae that may arise following resolution of severe cutaneous adverse reactions and should be vigilant for manifestations of autoimmune disease during follow-up.     

Critical factors differentiating erythema multiforme majus from Stevens-Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN)

Erythema multiforme majus (EMM) and Stevens-Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN) are severe cutaneous reactions characterised by targetoid erythematous lesions and mucocutaneous involvement. The initial skin manifestations are similar, making early diagnosis difficult. We retrospectively reviewed 36 cases of EMM and 18 cases of SJS/TEN and also evaluated 6 patients with unclassified EMM. 13 patients in the EMM group and 16 patients in the SJS/TEN group presented with a high fever (>38.5?C; p<0.001). Two or more mucous membranes were affected in 6 patients in the EMM group and 18 patients in the SJS/TEN group. Significantly more SJS/TEN than EMM patients had high levels of C-reactive protein and severe hepatic dysfunction. Thirteen EMM and 13 SJS/TEN cases were caused by medications/drugs. Skin biopsy samples showed stronger mononuclear cell infiltration in the EMM than in the SJS/TEN group (p<0.001). The mean dose of initial systemic corticosteroid used to treat EMM was lower than that used to treat SJS/TEN. No patients died in either group. Clinically, the unclassified cases mostly behaved like EMM. The results of our investigation suggest that EMM and SJS/TEN are distinct conditions and they help in differentiating these syndromes at an early stage.     

Antiretroviral‐induced toxic epidermal necrolysis in a patient positive for human immunodeficiency virus

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are two variants on a spectrum of severe systemic hypersensitivity characterized by blistering maculopapular lesions and desquamation of the skin and mucus membranes. Although several causative agents, including infections, have been reported for SJS/TEN, medications remain the most common cause. We report the case of a 42‐year‐old man with human immunodeficiency virus (HIV) who developed TEN 4 months after starting treatment with darunavir and abacavir. The patient presented with upper body lesions, oral mucosal ulcerations, and impending airway compromise. He was intubated and admitted to the burns unit. Score for Toxic Epidermal Necrolysis (SCORTEN) was 5, with > 90% predicted mortality. However, after intravenous immunoglobulin and supportive treatment, the patient made a remarkable recovery. Abacavir and darunavir may be associated with SJS/TEN. TEN should be considered a risk for patients with HIV and should be monitored for cutaneous eruptions for several months after changes in treatment regimen.     

Stevens-Johnson syndrome induced by doxycycline

Stevens-Johnson syndrome (SJS) is an acute mucocutaneous eruption nosologically related to erythema multiforme (EM) and toxic epidermal necrolysis (TEN). Medications are the most common triggering factors for SJS, with anticonvulsants, sulfonamides, penicillins, allopurinol, and nonsteroidal anti-inflammatory drugs (NSAIDs) most commonly implicated. SJS is very rarely associated with tetracyclines. We report a case of doxycycline-induced SJS in a 46-year-old man.     

Efficacy of additional i.v. immunoglobulin to steroid therapy in Stevens–Johnson syndrome and toxic epidermal necrolysis

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are rare and life‐threatening cutaneous adverse drug reactions. While there is no established therapy for SJS/TEN, systemic corticosteroids, plasma exchange and i.v. immunoglobulin (IVIG) have been used as treatment. The efficacy of IVIG is still controversial because total doses of IVIG used vary greatly from one study to another. The aim of this study was to evaluate the efficacy of IVIG, administrated for 5 days consecutively, in an open‐label, multicenter, single‐arm study in patients with SJS or TEN. IVIG (400 mg/kg per day) administrated for 5 days consecutively was performed as an additional therapy to systemic steroids in adult patients with SJS or TEN. Efficacy on day 7 of IVIG was evaluated. Parameters to assess clinical outcome were enanthema including ophthalmic and oral lesions, cutaneous lesions and general condition. These parameters were scored and recorded before and after IVIG. We enrolled five patients with SJS and three patients with TEN who did not respond sufficiently to systemic steroids before IVIG administration. All of the patients survived and the efficacy on day 7 of the IVIG was 87.5% (7/8 patients). Prompt amelioration was observed in skin lesions and enanthema in the patients in whom IVIG therapy was effective. Serious side‐effects from the use of IVIG were not observed. IVIG (400 mg/kg per day) administrated for 5 days consecutively seems to be effective in patients with SJS or TEN. IVIG administrated together with steroids should be considered as a treatment modality for patients with refractory SJS/TEN. Further studies are needed to define the therapeutic efficacy of IVIG.     

Stevens-Johnson syndrome and toxic epidermal necrolysis in patients with lupus erythematosus: a descriptive study of 17 cases from a national registry and review of the literature

Background Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe cutaneous adverse reactions with high morbidity and mortality. Some expressions of lupus erythematosus (LE) may cause enormous difficulties in differentiating them from SJS and TEN by showing large areas of sheet‐like epidermal necrosis. Objective To evaluate clinically and histopathologically probable or definite cases of SJS/TEN with a history of systemic or other LE [(S)LE]. Methods This was a retrospective analysis of validated cases of SJS/TEN with a history of (S)LE, based on a large population‐based national registry. Results Among 1366 patients with SJS/TEN, 17 with a sufficiently documented history of (S)LE and representative histological material could be identified, suggesting a considerable over‐representation of LE in patients with SJS/TEN. Eight of these showed clinically and/or histopathologically some LE‐characteristic features interfering with the diagnosis of SJS/TEN. Differentiation could be elaborated on clinical and histopathological grounds: four patients were classified as SJS/TEN with a preceding (S)LE exacerbation and/or LE‐typical histopathological features, and four as ‘TEN‐like’ (S)LE. Conclusion Most patients with SJS/TEN and a history of (S)LE demonstrate clinical and histopathological properties allowing clear differentiation. However, occasionally acute cutaneous manifestations of (S)LE and SJS/TEN can be phenotypically similar, caused by extensive epidermal necrosis. Although no feature by itself is conclusive, a combination of recent (S)LE exacerbation, evident photodistribution, annular lesions and absent or only mild focal erosive mucosal involvement may favour LE over SJS/TEN clinically. Histopathologically, in particular, junctional vacuolar alteration, and the presence of solitary necrotic keratinocytes at lower epidermal levels, combined with moderate to dense periadnexal and perivascular lymphocytic infiltrates with a variable presence of melanophages, and mucin point to a LE‐related origin.     

In the pursuit of classifying severe cutaneous adverse reactions

We suggest adding an additional type of lesion to the existing 4 types of lesions of the erythema multiforme (EM), Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN), namely "flat typical target" and call the original typical targets "raised typical target." The EM group would consist of raised typical targets and raised atypical targets, similar to the original definition, and the SJS/TEN group would consist of flat typical targets, flat atypical targets and macules with or without blisters. In our proposed modified classification (Table 1), all the lesions that are found in the EM group are raised, whereas all lesions that characterize the SJS/TEN group are flat, even though they have blisters on them.     

Clinical Pattern of Cutaneous Drug Eruption among Children and Adolescents in

Abstract: Various types of cutaneous drug eruptions and the Incriminating drugs were analyzed tn 50 children and adolescents up to 18 years of age (34 or 65% boys, 16 or 32% girls). Thirteen (26%) patients had a maculopapular rash, 11 (22%) a fixed drug eruption (FDE), 10 erythema multiforme (EM), 6 (12%) toxic epidermal necrolysis (TEN), 5 (10%) Stevens-Johnson syndrome (SJS), 3 (6%) urticaria, and 2 (4%) erythroderma. The Incubation period for maculopapular rashes, SJS and TEN due to commonly used antibiotics and sulfonamides was short, a few hours to two to three days, reflecting reexposure, and for drugs used sparingly such as antiepileptics and antitubereulosis agents, was approximately one week or more, suggesting a first exposure. Antibiotics were responsible for cutaneous eruptions in 27 patients, followed by antlepileptics In 17, analgin in 4, and metronidazole and albendazole in 1 each. Cotrimoxazole, a combination of sulfamethoxazole and trimethoprim, was the most common antibacterial responsible for eruptions (11 patients), followed by penicillin and its semisynthetlc derivatives (8 patients), sulfonamide alone (3 patients), and other antibiotics (4 patients). Antiepileptics were the most frequently incriminated drugs in EM, TEN, and SJS. The role of systemic corticosteroids in the management of SJS and TEN is controversial. We administered prednisolone or an equivalent corticosteroid 2 mg/kg/day for 7 to 14 days. With this dosage the mortality rate in the combined patients with TEN and SJS was 18.2%. Our limited experience suggests that these drugs might still have a role in the management of SJS and TEN In children and adolescents.     

The incidence of erythema multiforme, Stevens-Johnson syndrome, and toxic epidermal necrolysis. A population-based study with particular reference to reactions caused by drugs among outpatients

We carried out a study to estimate the incidence of erythema multiforme (EM), Stevens Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN) requiring hospitalization and to determine which drug therapies were associated with these reactions. We reviewed the clinical records of all patients who were hospitalized with these discharge diagnoses at Group Health Cooperative (GHC) of Puget Sound, Seattle, Wash, from 1972 through 1986. During this 14-year period, an average of about 260,000 persons, with demographic characteristics similar to those of the general population, received their care from GHC, and there were about 25,000 admissions to hospitals per year at the GHC hospitals. Based on International Classification of Diseases-Adapted coding, a total of 61 suspect cases of EM, SJS, or TEN were identified from the computerized hospital discharge file. Based on record review and the application of a uniform set of diagnostic criteria, a total of 37 patients (61%) were classified as having EM, SJS, or TEN. Of these, 16 cases (43%) were attributed to drugs administered to these patients prior to hospitalization. The overall incidence of hospitalization for EM, SJS, or TEN due to all causes was 4.2 per 10(6) person-years. The incidence of TEN alone due to all causes was 0.5 per 10(6) person-years. The incidence of EM, SJS, or TEN associated with drug use were 7.0, 1.8, and 9.0 per 10(6) person-years, respectively, for persons younger than 20 years of age, 20 to 64 years of age, and 65 years of age and older. Drug therapies with reaction rates in excess of 1 per 100,000 exposed individuals include phenobarbital (20 per 100,000), nitrofurantoin (7 per 100,000), sulfamethoxazole and trimethoprim, and ampicillin (both 3 per 100,000), and amoxicillin (2 per 100,000). Overall, our data suggest that cases of EM, SJS, and TEN sufficiently severe to require hospitalization are infrequent among outpatients using well-established drug therapies. A continuing challenge is the evaluation of these severe cutaneous reactions that are associated with newly marketed or less frequently prescribed drug therapies.