BACKGROUND: Toxic epidermal necrolysis (TEN) is a very rare but extremely severe drug reaction characterized by widespread apoptosis of epidermis with extensive blisters. We previously found drug-specific cytotoxic CD8 T lymphocytes in the blisters of a single patient. OBJECTIVE: To confirm the role of drug specific cytotoxic lymphocytes in a larger series, to test the cytotoxicity on keratinocytes, and to look for cross-reactivity between chemically related drugs. METHODS: The phenotype of lymphocytes present in the blister fluids of 6 patients with TEN was analyzed by flow cytometry. Cytotoxic functions were tested by chromium release assay on a variety of target cells (autologous or MHC class I-matched EBV-transformed lymphocytes, autologous keratinocytes) after nonspecific (CD3 monoclonal antibody) or specific (suspected and potentially cross-reactive drugs) activation. RESULTS: Blister lymphocytes were CD8 + HLA-DR + CLA + CD56 + . In all 6 cases, they were cytotoxic after nonspecific activation. A drug-specific cytotoxicity was observed in 4 cases (3 related to cotrimoxazole and 1 to carbamazepine) toward lymphocytes. Blister cells also killed IFN-gamma-activated autologous keratinocytes in the presence of drug in the 2 patients tested. Blister cells showed a strong immunoreactivity for granzyme B, and cytotoxicity was abolished by EGTA, but not by anti-Fas/CD95, suggesting perforin/granzyme-mediated killing. By using several sulfonamides for testing the specificity of the drug T-cell receptor interaction, we observed cross-reactivity only between 4 structurally closely related medications. CONCLUSION: These results strongly suggest that drug-specific, MHC class I-restricted, perforin/granzyme-mediated cytotoxicity probably has a primary role in TEN.
BACKGROUND: Toxic epidermal necrolysis (TEN) is a very rare but extremely severe drug reaction characterized by widespread apoptosis of epidermis with extensive blisters. We previously found drug-specific cytotoxicCD8 T lymphocytes in the blisters of a single patient. OBJECTIVE: To confirm the role of drug specific cytotoxic lymphocytes in a larger series, to test the cytotoxicity on keratinocytes, and to look for cross-reactivity between chemically related drugs. METHODS: The phenotype of lymphocytes present in the blister fluids of 6 patients with TEN was analyzed by flow cytometry. Cytotoxic functions were tested by chromium release assay on a variety of target cells (autologous or MHC class I-matched EBV-transformed lymphocytes, autologous keratinocytes) after nonspecific (CD3 monoclonal antibody) or specific (suspected and potentially cross-reactive drugs) activation. RESULTS: Blister lymphocytes were CD8 + HLA-DR + CLA + CD56 + . In all 6 cases, they were cytotoxic after nonspecific activation. A drug-specific cytotoxicity was observed in 4 cases (3 related to cotrimoxazole and 1 to carbamazepine) toward lymphocytes. Blister cells also killed IFN-gamma-activated autologous keratinocytes in the presence of drug in the 2 patients tested. Blister cells showed a strong immunoreactivity for granzyme B, and cytotoxicity was abolished by EGTA, but not by anti-Fas/CD95, suggesting perforin/granzyme-mediated killing. By using several sulfonamides for testing the specificity of the drug T-cell receptor interaction, we observed cross-reactivity only between 4 structurally closely related medications. CONCLUSION: These results strongly suggest that drug-specific, MHC class I-restricted, perforin/granzyme-mediated cytotoxicity probably has a primary role in TEN.
Authors: Olivia A Charlton; Victoria Harris; Kevin Phan; Erin Mewton; Chris Jackson; Alan Cooper Journal: Adv Wound Care (New Rochelle) Date: 2020-01-09 Impact factor: 4.730
Authors: S G Leckband; J R Kelsoe; H M Dunnenberger; A L George; E Tran; R Berger; D J Müller; M Whirl-Carrillo; K E Caudle; M Pirmohamed Journal: Clin Pharmacol Ther Date: 2013-05-21 Impact factor: 6.875
Authors: Ayman Elsheikh; Sidonie N Lavergne; J Luis Castrejon; John Farrell; Haiyi Wang; Jean Sathish; Werner J Pichler; B Kevin Park; Dean J Naisbitt Journal: J Immunol Date: 2010-10-27 Impact factor: 5.422