Secondary Logo

Fever in Stevens–Johnson Syndrome and Toxic Epidermal Necrolysis in Pediatric Cases: Laboratory Work-up and Antibiotic Therapy

Paulmann, Maren PhD; Mockenhaupt, Maja MD, PhD

The Pediatric Infectious Disease Journal: May 2017 - Volume 36 - Issue 5 - p 513–515
doi: 10.1097/INF.0000000000001571
ESPID Reports and Reviews

Fever is a symptom that often accompanies skin eruptions, especially in children. It can be a sign of an infectious condition presenting with exanthems or it may precede an exanthematous eruption. Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe reactions affecting skin and mucosa with blisters and erosions. High fever occurs in these conditions, frequently before the skin and/or mucosa is affected.

From the Department of Dermatology, Dokumentationszentrum schwerer Hautreaktionen (dZh), Medical Center-University of Freiburg, Hauptstrasse 7, 79104 Freiburg, Germany.

Accepted for publication January 30, 2017.

The RegiSCAR-study, in which all cases of the German Registry of severe skin reactions (dZh) are included, was funded by research institutions including DFG and BMBF as well as various pharmaceutical companies that provided unrestricted grants.

The authors have no conflicts of interest to disclose.

Address for correspondence: Maja Mockenhaupt, MD, PhD, Department of Dermatology, Dokumentationszentrum schwerer Hautreaktionen (dZh), Medical Center-University of Freiburg, Hauptstrasse 7, 79104 Freiburg, Germany. E-mail:

Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are rare, severe skin reactions characterized by detachment of the epithelium of skin and mucous membranes.1 These acute life-threatening mucocutaneous reactions are accompanied by fever and flu-like symptoms, for example, sore throat, headache and reduced state of health.1,2

Traditionally, SJS/TEN has been classified into the erythema multiforme (EM) spectrum, because of similar histology, mucosal involvement and presence of target-like lesions. Therefore, many physicians equated EM with mucosal involvement, that is, the major type of EM or EM majus (EMM) with SJS, mainly because the mucosal erosions in both conditions cannot be differentiated. This has led and still leads to confusion, especially when the risk for mortality or recurrence is evaluated. However, in a case–control study, it has been shown that SJS/TEN and EMM are distinct entities differing in their clinical pattern of skin lesions, demographic data and etiology.3 Furthermore, a positive Nikolsky sign is an important clinical clue for differentiating SJS/TEN and EMM.1,2 With the help of a consensus classification, a diagnosis can be determined in more than 90% of the cases.4 SJS and TEN are considered as 1 entity differing in the extent of epidermal detachment related to the body surface area (BSA): SJS, <10% BSA; SJS/TEN overlap, 10%–30% BSA; TEN, >30% BSA. Hemorrhagic mucosal erosions are found in SJS, SJS/TEN overlap, but may be absent in some cases of TEN.5 Lesions in SJS/TEN are widespread, often confluent purpuric macules (spots) or atypical flat targets (without palpable edema or infiltration) on which blisters develop (Fig. 1A). In contrast, in EMM, typical (regular round shape, well-defined border with 3 different concentric zones) and/or atypical raised targets (poorly defined border, 2 zones) appear mainly on the limbs (Fig. 1B), sometimes also on the face and trunk.3,4



Back to Top | Article Outline


The overall incidence of SJS/TEN has been calculated as 1.5–1.8 cases per 1 million inhabitants.6 However, the exact incidence for children is still unknown. Preliminary data of the German Registry of severe skin reactions show that only 7% of the validated SJS/TEN cases collected over 10 years are younger than 18 years of age with an almost equal distribution of gender. The mortality is about 6% and much lower in the pediatric population than in adults, as has been reported in small case series.7 Recurrence of SJS/TEN has not been observed except when the causative drug was taken again. One might expect that SJS/TEN induced by infection tends to reoccur, but this has never been observed in more than 25 years of the German Registry [unpublished data of Dokumentationszentrum (dZh)]. In contrast, recurrence is frequently seen in EM/EMM.

SJS/TEN is mainly drug induced, and in approximately 65%–75% of all cases, a causative agent can be identified.5 In children and adolescents, certain antiepileptics, antibacterial sulfonamides including sulfasalazine and far less frequently antibiotics (cephalosporins, macrolides and ampicillin) have been identified to be the main causative drugs.8,9 The preliminary data of the German Registry show that in no more than 50% of the pediatric cases, a causative drug was found. Many other cases seem to be induced by infection (eg, Mycoplasma pneumoniae, Streptococci, cytomegalovirus, Epstein–Barr virus and other viruses) or have to be considered as idiopathic. In such cases, we suggest to perform laboratory examinations to search for bacterial and viral pathogens. However, in many cases, no pathogen can be detected.

In numerous published case reports, antipyretics, analgesics or secretolytics are blamed for inducing SJS/TEN, although 2 large case–control studies could not detect an association.10 This is primarily due to a misunderstanding of how SJS/TEN starts and develops (Fig. 1C). Prodromal symptoms such as fever, sore throat and flu-like symptoms, for which antipyretics, analgesics and secretolytics are taken, appear 1–3 days before the cutaneous and/or mucosal lesions (Fig. 1C; 1-3 D). Frequently, the drugs to treat these prodromal symptoms are accused to have caused the reaction, because they were taken shortly before the mucocutaneous signs appear, which is actually after onset of the disease. This kind of misinterpretation of drug causality is called “protopathic bias.”2 For drugs with a confirmed risk for SJS/TEN, the time latency between beginning of drug use and onset of the reaction varies between 4 days and 4 weeks. Furthermore, SJS/TEN occurs during the first continuous use of the drug (without prior sensitization), whereas earlier well-tolerated use of the medication does not suggest a causal relationship.5,10 In cases where drugs were taken to treat an infection in the corresponding period, the infection, but also the newly introduced drug, may be an etiologic factor (confounding by indication).2

Back to Top | Article Outline


Fever up to 40°C (104°F) or even higher occurs in nearly all patients with SJS/TEN and is independent of the cause (drug vs. infection).2,3 As shown in Figure 1C, fever is also present in patients without infection but with a clear causative drug, which was taken, for example, for 14 days. In such cases, fever is a prodromal or concomitant symptom of SJS/TEN itself. In patients with no relevant drug history, but an infection directly before reaction onset, it is not easy to distinguish whether the fever is due to the infection or part of the reaction itself. However, fever should be treated with antipyretic medication as needed. Concerning the use of antibiotics in SJS/TEN, a more differentiated approach is recommended, because in drug-related SJS/TEN, laboratory values for inflammation parameters (leukocytes, C-reactive protein) are often increased because of the reaction itself. This is an unspecific increase, which should not be treated preventively by antibiotics, because they may mask the occurrence of a real infection as a complication of SJS/TEN.2,5 Parameters like procalcitonin or sedimentation rate are usually normal. In our experience of studying and treating patients with SJS/TEN, fever and increased inflammation parameters disappear or decrease within a few days. If high fever and increased inflammation parameters persist for more than 5 days or clearly increase or rise again, a secondary infection may be present and appropriate investigations should be done, for example, blood culture, radiogram, polymerase chain reaction, procalcitonin and urinary exam. Secondary infections more frequently affect patients with central venous lines or mechanical ventilation and patients in a state of immunosuppression.

In SJS/TEN induced by an infection, the increase of inflammation parameters may be due to the infection and the reaction itself. In such cases, it has to be clarified whether the infection is viral or bacterial. Although a specific pathogen may often not be detected, examinations for the following pathogens should be done by polymerase chain reaction and/or serology: M. pneumoniae, Chlamydia pneumoniae, Streptococcus pneumoniae, Epstein–Barr virus, cytomegalovirus, herpes simplex virus 1 and 2, human herpes virus 6 and 7, influenza A and B, adenovirus, parvovirus and coxsackie virus. In addition, radiograph exam can be helpful, because sometimes an infiltrate of the lung can be detected even when other tests are negative. Any proven bacterial infection should be treated with an antibiotic according to the pathogen and to the resistogram or according to microbiologic treatment guidelines.11 Every antibiotic can be used with 2 exceptions: (1) if an antibiotic is suspected to be the causative drug, this agent should not be administered again and (2) if there is a known drug allergy to a certain antibiotic. In general, it is not necessary to avoid a specific class of antibacterial agents, because cross-reactivity is not a phenomenon observed in SJS/TEN. However, if a certain sulfonamide is thought to be the cause, other antibacterial sulfonamides should be avoided. From our experience, there is no correlation between the intake of specific antibiotic(s) and a prolonged course or a second event of SJS/TEN.

To date, there are no specific laboratory values to determine SJS/TEN, but certain important values for the prognosis of mortality have been summarized in a severity of illness score for SJS/TEN (SCORTEN). SCORTEN shows the best correlation as a severity marker of SJS/TEN within the first 3 days after hospital admission and includes 7 independent equivalent factors: (1) age (≥40 years); (2) heart rate (≥120/min); (3) underlying malignant disease; (4) detachment of the BSA on the first day (≥10%); (5) serum urea (≥10 mmol/L); (6) serum bicarbonate (<20 mmol/L) and (7) serum glucose (≥14 mmol/L).12 The higher the score value, the poorer is the prognosis of the patient and the higher is the risk of death. Although SCORTEN was developed primarily for adults, the prediction of morbidity in children and adolescents is suitable. In a recent study, 4 predictive models were examined not only for mortality, which in children is usually lower than predicted, but also for morbidity (mechanical ventilation, days until wound healing, number of infectious complications, number of acute operations and length of stay). All models were suitable for morbidity, but SCORTEN, specifically designed for SJS/TEN, was considered to be more practicable than the others.13

Therapy of SJS/TEN remains unspecific, and supportive treatment is crucial.1,2,5,11 In drug-related cases of SJS/TEN, the causative drug should be discontinued immediately. In infection-induced cases, the infection requires adequate treatment, as explained above. For topical treatment, antiseptic solutions or gels should be used, and drug-free, nonadherent wet gauze can be applied to areas affected by blisters and erosions. Adequate analgesia should be started, and if necessary, intravenous fluids should be given, and the room temperature should be raised. Furthermore, for mucosal erosions, local antiseptic treatment is recommended. In case of ocular involvement, a daily ophthalmologic consultation is important, because the eye inflammation may result in long-lasting sequelae.1,2 Systemic immunomodulatory therapy with glucocorticosteroids and intravenous immunoglobulins remains controversial, and data for both treatments are ambiguous.1,11 Cyclosporine has been used successfully in the treatment of SJS/TEN, but further studies are needed, especially in children and elderly patients.2

Back to Top | Article Outline


  1. SJS/TEN and EMM are different entities which can be distinguished by the clinical pattern.
  2. SJS/TEN in children and adolescents is often not drug induced and probably caused by infections or unknown factors (idiopathic).
  3. Fever is present and inflammation parameters are increased in SJS/TEN, independent of the etiology (medication, infection and idiopathic).
  4. Supportive therapy is the “gold standard,” systemic immunomodulating therapy remains controversial.
  5. Preventative use of antibiotics is not recommended, but antibiotics should be given in cases with a proven bacterial infection.
Back to Top | Article Outline


1. Mockenhaupt M. Burgdorf WHC, Plewig G, Wolff HH. Severe cutaneous adverse drug reactions. In: Braun-Falco’s Dermatology. 2009:Heidelberg: Springer; 473–484.
2. Paulmann M, Mockenhaupt M. [Schwere arzneimittelinduzierte Hautreaktionen: Klinik, Diagnostik, Ätiologie und Therapie] Severe drug-induced skin reactions: clinical features, diagnosis, etiology, and therapy. J Dtsch Dermatol Ges. 2015;13:625–645.
3. Auquier-Dunant A, Mockenhaupt M, Naldi L, et al; SCAR Study Group. Severe Cutaneous Adverse Reactions. Correlations between clinical patterns and causes of erythema multiforme majus, Stevens-Johnson syndrome, and toxic epidermal necrolysis: results of an international prospective study. Arch Dermatol. 2002;138:1019–1024.
4. Bastuji-Garin S, Rzany B, Stern RS, et al. Clinical classification of cases of toxic epidermal necrolysis, Stevens-Johnson syndrome, and erythema multiforme. Arch Dermatol. 1993;129:92–96.
5. Mockenhaupt M. The current understanding of Stevens-Johnson syndrome and toxic epidermal necrolysis. Expert Rev Clin Immunol. 2011;7:803–813; quiz 814.
6. Rzany B, Mockenhaupt M, Baur S, et al. Epidemiology of erythema exsudativum multiforme majus (EEMM), Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) in Germany (1990–1992). Structure and results of a population-based registry. J Clin Epidemiol. 1996;49:769–773.
7. Hamilton GM, Fish J. Pediatric toxic epidermal necrolysis: an institutional review of patients admitted to an intensive care unit. J Burn Care Res. 2013;34:e351–e358.
8. Quirke KP, Beck A, Gamelli RL, et al. A 15-year review of pediatric toxic epidermal necrolysis. J Burn Care Res. 2015;36:130–136.
9. Levi N, Bastuji-Garin S, Mockenhaupt M, et al. Medications as risk factors of Stevens-Johnson syndrome and toxic epidermal necrolysis in children: a pooled analysis. Pediatrics. 2009;123:e297–e304.
10. Mockenhaupt M, Viboud C, Dunant A, et al. Stevens-Johnson syndrome and toxic epidermal necrolysis: assessment of medication risks with emphasis on recently marketed drugs. The EuroSCAR study. J Invest Dermatol. 2008;128:35–44.
11. Ghislain PD, Roujeau JC. Treatment of severe drug reactions: Stevens-Johnson syndrome, toxic epidermal necrolysis and hypersensitivity syndrome. Dermatol Online J. 2002;6:.
12. Bastuji-Garin S, Fouchard N, Bertocchi M, et al. SCORTEN: a severity-of-illness score for toxic epidermal necrolysis. J Invest Dermatol. 2000;115:149–153.
13. Beck A, Quirke KP, Gamelli RL, et al. Pediatric toxic epidermal necrolysis: using SCORTEN and predictive models to predict morbidity when a focus on mortality is not enough. J Burn Care Res. 2015;36:167–177.

Stevens–Johnson syndrome; toxic epidermal necrolysis; erythema multiforme majus; role of infections

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.