Childhood epidermal necrolysis and erythema multiforme major: a multicentre French cohort study of 62 patients

Abstract

Introduction The distinction between epidermal necrolysis [EN; including Stevens–Johnson syndrome (SJS), toxic epidermal necrolysis (TEN) and overlap syndrome] and erythema multiforme major (EMM) in children is confusing. We aimed to better describe and compare these entities. Materials and methods This French retrospective multicentre study included children ≤18 years old referred for EN or EMM between 1 January 2008 and 1 March 2019. According to pictures, children were reclassified into TEN/overlap, SJS or EMM/unclassified (SJS/EMM) groups and compared for epidemiological and clinical data, triggers, histology and follow‐up. Results We included 62 children [43 boys, median age 10 years (range 3–18)]: 16 with TEN/overlap, 11 SJS and 35 EMM. The main aetiologies were drugs in EN and infections (especially Mycoplasma pneumoniae) in EMM (P < 0.001), but 35% of cases remained idiopathic (TEN/overlap, 47%; SJS, 24%; EMM, 34%). The typical target lesions predominated in EMM (P < 0.001), the trunk was more often affected in EN (P < 0.001), and the body surface area involved was more extensive in EN (P < 0.001). Mucosal involvement did not differ between the groups. Two patients with idiopathic TEN died. Histology of EMM and EN showed similar features. The recurrence rate was 42% with EMM, 7% with TEN/overlap and 0 with SJS (P < 0.001). Sequelae occurred in 75% of EN but involved 55% of EMM. Conclusion Clinical features of EN and EMM appeared well demarcated, with few overlapping cases. Idiopathic forms were frequent, especially for EN, meaning that a wide and thorough infectious screening, repeated if needed, is indicated for all paediatric cases of EN/EMM without any trigger drug. We propose a comprehensive panel of investigations which could be a standard work‐up in such situation. Sequelae affected both EN and EMM.

Stevens–Johnson syndrome in children

“…Toxicology testing on blood, urine, or hair to confirm drug exposure if the history is unclear [23].…”

“…An important element of follow-up for the patient and family who has survived DEN or RIME is patient and family education about the disease and its trigger or risk of recurrence. For RIME, the risk of recurrence is 8–42% [12,20 ▪ ,23,55 ▪ ]. A recent pediatric retrospective case series showed that counseling about recurrence tended to be poorer for patients with an infectious trigger, whose families should be informed that subsequent infections can provoke the same reaction [20 ▪ ].…”

“…Investigations include assessment of overall status with routine bloodwork and clinical and vital sign monitoring. Tests for infectious triggers are necessary [ MP, Chlamydia pneumoniae and respiratory virus PCR, MP and viral serologies (Epstein Barr virus, cytomegalovirus, human herpes virus 6, parvovirus B19), mucosal swab for herpes simplex virus 1 and 2 and enterovirus PCR, chest X-ray] [15 ▪▪ ,23]. Cytokine profiles and immunological assessment may help to differentiate RIME from DEN.…”

“…Cytokine profiles and immunological assessment may help to differentiate RIME from DEN. Metagenomic and transcriptomic analysis could be future techniques to establish infectious causality [23].…”

“…Immunomodulatory therapy with corticosteroids, cyclosporine, and IVIG may be beneficial in the early inflammatory phase of DEN but the results of most recent pediatric systematic and database reviews are variable and do not clearly support any one of these treatment options [81 ▪ ,82,84]. Recent pediatric retrospective cohorts are similarly inconclusive about the benefit of immunomodulatory therapy [23,59 ▪ ].…”