Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) have shared characteristics of erythematous cutaneous reaction with blister formation accompanied by mucosal involvement. Patients with SJS have desquamation of the skin affecting <10% of their body surface area whereas patients with TEN have >30% body surface area involvement. Patients with skin lesions affecting between 10 and 30% of their body surface area are considered to have SJS/TEN overlap (1). These epidermal detachments are considered life-threatening (2) and are most frequently manifested by adverse drug reaction. In addition to occurring as a reaction to drugs, these skin eruptions are also associated with underlying infectious diseases, and include the cutaneous manifestations of disseminated candidiasis (3,4), Mycoplasma pneumoniae (5), Chlamydia pneumoniae (6), cytomegalovirus infection (7) and human immunodeficiency virus (8,9). The mortality rates of these skin eruptions have been reported to range from 16 to 25% (10–13). Treatment is based on symptoms and supportive fluid and electrolyte replacement. Dermal coverage to prevent secondary infection and the loss of fluid are also crucial aspects of treatment. Several immunomodulative therapies have been suggested to treat SJS and/or TEN, particularly glucocorticoids and immunoglobulin. Prognostic factors and scoring systems have been used to define the mortality risk in these patients, including the severity-of-illness score of toxic epidermal necrolysis (SCORTEN) scale (14).

In this study, the clinical manifestations, drug implications, treatment and outcomes of patients with SJS and/or TEN who had been hospitalized over the past 5 years in a tertiary referral care center were retrospectively reviewed and analyzed.

SJS, SJS/TEN overlap and TEN are rare but life-threatening conditions. It is important to recognize the clinical characteristics of the mucocutaneous eruption at early stage due to the high mortality rate, which ranges from 16 to 25% (1,10–12). The most frequent cause of these conditions is medication (15,16). The most common precipitating drug in this study was allopurinol. Anticonvulsants and antibiotics were found to be the second common causative agents in the present study, despite these two groups of medication being reported as the most frequent etiology in a previous study (17). The difference between these results could be explained by the common medications previously reported as having cutaneous adverse reactions now being avoided. With advancements in the identification of specific human leukocyte antigen (HLA) alleles that are associated with drug reactions, screening to identify patients at risk for drug reaction is becoming a part of standard clinical practice in certain academic institutions (18–21). This could be one of the reasons why the incidence of drug reactions toward commonly known medications appears to have declined.

The use of systemic corticosteroid treatment in patients with SJS/TEN is controversial, with concerns regarding an increased rate of infections, the masking of septicemia, and delay of epithelialization (22). By contrast, the benefits of immunosuppressants (including corticosteroids) have been reported as preventing ocular complications (23–26). In the present study, 65% of patients received systemic steroid treatment (range, 1–10 days). Notably, there was no mortality in patients treated with systemic steroids. By comparison, 3 cases of mortality were patients who did not receive steroid treatment. The results of the present study suggest the beneficial effect of systemic corticosteroid use in selected groups of SJS and/or TEN patients.

The SCORTEN scoring system has been used to define the severity of the disease and predict mortality for more than a decade. The results of the present study revealed a lower number of cases of mortality than predicted by SCORTEN score (14). There was no significant difference in SCORTEN score between survival and non-survival groups. This finding emphasizes the limitation of the SCORTEN as previously mentioned in certain review articles (27–29).

The mortality rate in the presented study was 6.9%, which was lower than the rates in previous studies (1,10–12). This could be the result of early diagnosis, the immediate discontinuation of causative medication, supportive medical care and immunosuppressive treatment. The finding in the present study concerning corticosteroid treatment is similar to that of the EuroSCAR-and RegiSCAR studies (13,30), which suggested the beneficial use of corticosteroid in selected subgroups of patients with SJS and/or TEN. To confirm this finding, a future prospective controlled study should be undertaken to evaluate the benefit of corticosteroid treatment in SJS/TEN.

This study has several limitations. Firstly, in terms of the small size of the study population selected from a major tertiary care center. Secondly, the study was designed to include only hospitalized patients. This may not provide the full evaluation in both quality and quantity of management in general, since the majority of the patients with mild forms of disease are treated in primary local hospitals. Thirdly, genetic data was not collected. With the advance and availability of genetic testing, genetic screening may be warranted to assist in the selection of treatment options and prevention. Lastly, this observational study may suggest the beneficial effect of steroids in the treatment of patients with SJS/TEN. However, a further double-blinded placebo control study is required to confirm this suggestion.

The severe forms of mucocutaneous eruptions, SJS and/or TEN, are mostly associated with adverse drug reactions. With early recognition and selected treatments, the mortality rate could be reduced. Improved understanding of clinical presentation and risk factors should help physicians to improve the care of high-risk individuals at an earlier stage. Patient age and the area of mucocutaneous involvement have been identified as significant factors associated with mortality.