Cytokine profile-guided management of Stevens-Johnson syndrome and toxic epidermal necrolysis (SJS/TEN): A management algorithm useful for guiding the selection of treatment options.

BACKGROUND

Because different immunosuppressive therapies have their own characteristic properties to inhibit/enhance the production of various cytokines in Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN), cytokine profile dynamics could be used to predict and monitor the therapeutic response across different treatments, when assessed at baseline and following therapy.

METHODS

This retrospective analysis was performed to assess how changes in cytokine profiles available in clinics contributed to an optimal therapeutic response in SJS/TEN patients. We included 31 patients with SJS/TEN, treated with standard corticosteroids (n = 9), pulse corticosteroids (n = 11), intravenous immunoglobulins (n = 4), plasmapheresis (n = 6), and supportive therapy (n = 1).

RESULTS

High granulysin (≥ 1.6 ng/ml) and soluble Fas ligand (≥ 2540 pg/ml) levels at baseline were strong predictors for the failure of standard corticosteroid therapy. We found that standard corticosteroid treatment inhibited not only Th1 cytokines but also monocyte-derived cytokines in patients exhibiting an adequate treatment response. Although pulse corticosteroids have a fundamentally similar effect as standard corticosteroids on cytokine profiles, pulse corticosteroids more efficiently reduced monocyte-derived cytokines. Intravenous immunoglobulins suppressed interferon-γ and interleukin (IL)-12p40 production more efficiently compared with standard corticosteroids and other immunosuppressive agents. A profound decrease in granulocyte colony-stimulating factor and IL-10 associated with a marked increase in IL-12 was specifically detected in SJS/TEN patients treated with plasmapheresis.

CONCLUSIONS

We can select the most appropriate treatment option for individual patients following the failure of standard corticosteroids by examining treatment-relevant cytokine profiles before and after treatment. We propose a treatment algorithm that can serve as a decision-making tool for tailored therapy selection.