Epidermal Necrolysis: 60 Years of Errors and Advances

Y.K. Heng; H.Y. Lee; J.-C. Roujeau


The British Journal of Dermatology. 2015;173(5):1250-1254. 

In This Article

Therapeutic Challenges and Errors

The mainstay of treatment of SJS/TEN is still early withdrawal of the culprit drug,[22] management in a reference centre and supportive care. Based on its postulated immune mechanism, the use of various immunomodulating therapies such as pulse corticosteroids,[23] IVIG,[21] ciclosporin,[24] cyclophosphamide,[25] plasmapheresis,[26] TNF inhibitors[27] and others such as detoxifying agents[28] has been reported.

A critical evaluation of such treatments requires the understanding of two concepts in SJS/TEN. Firstly, SJS/TEN is a self-limiting disease. Based on data from the EuroSCAR/RegiSCAR cohort of SJS/TEN, the mean duration from the onset of symptoms to maximum detachment is 8 days (Fig. 1).[29] Therefore, the observed efficacy of an intervention in arresting the disease progression may be merely the natural evolution of disease. Secondly, due to the low incidence of this disease, organizing a randomized controlled trial (RCT) will be challenging.

Figure 1.

Natural history of Stevens–Johnson syndrome/toxic epidermal necrolysis. The delays shown are the mean values observed in two large case–control studies. They were provided only partially in prior publications of methods35 or results that focused on the risks of medications11,12 and on classification.1

To date, there has been only one published RCT evaluating treatment in SJS/TEN. In that trial, thalidomide was the active intervention based on the suspected role of TNF and the anti-TNF effect of thalidomide.[30] The trial was stopped prematurely due to increased mortality in one arm, which, on unblinding, turned out to be the thalidomide arm. Results from this study caution the use of anti-TNF therapy in EN. Corticosteroids and IVIG are the two most common immunomodulatory agents used. In a systematic review of treatment efficacy (including series with 10 or more cases), there was no difference in mortality outcomes [standardized mortality ratio, SMR (95% confidence interval, CI)] between those treated with supportive care: SMR 0·89 (0·68–1·16), systemic corticosteroids: SMR 0·92 (0·53–1·48) and IVIG: 0·82 (0·58–1·12).[31] In another systematic review of IVIG efficacy, which included series with more than eight patients, the pooled odds ratio for mortality was 1·0 (95% CI 0·58–1·75).[32] Following the two published systematic reviews, the largest single-centre cohort of SJS/TEN (n = 64) treated with IVIG again showed that there was no mortality benefit in IVIG [SMR 1·10 (0·62–1·58)]; further subgroup analysis did not reveal any dose-dependent effect.[33] These three studies taken together with the recent understanding that granulysin is the most important mediator should lead to the end of IVIG as a therapeutic modality in SJS/TEN.