Chen, Chun-BingChun-BingChenWang, Chuang-WeiChuang-WeiWangLu, Chun-WeiChun-WeiLuChen, Wei-TiWei-TiChenZhou, Bing-RongBing-RongZhouCHIA-YU CHUHsu, Shang-FuShang-FuHsuYang, Cheng-TaCheng-TaYangWen-Cheng Chang, JohnJohnWen-Cheng ChangYang, Chan-KengChan-KengYangWang, Chih-LiangChih-LiangWangFang, Yueh-FuYueh-FuFangHsu, Ping-ChihPing-ChihHsuHua, Chung-ChingChung-ChingHuaWu, Chiao-EnChiao-EnWuKo, How-WenHow-WenKoChen, Kun-ChiehKun-ChiehChenYang, Yi-ChienYi-ChienYangTseng, Han-ChiHan-ChiTsengCheng, An-YuAn-YuChengTseng, Li-ChuanLi-ChuanTsengShih, Feng-YaFeng-YaShihHung, Shuen-IuShuen-IuHungHuang, Cheng-YangCheng-YangHuangChung, Wen-HungWen-HungChung2025-07-012025-07-012025-01https://scholars.lib.ntu.edu.tw/handle/123456789/730409Background: Osimertinib, a third-generation epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI), shows superior lung cancer treatment efficacy. However, osimertinib-induced severe hypersensitivity, including Stevens-Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN), is frequently observed in Asian populations and hinders cancer treatment. Objective: We investigated the genetic HLA predisposition and immune pathomechanism of osimertinib-induced hypersensitivity. Methods: We enrolled 17 patients with osimertinib-induced delayed hypersensitivity (seven with severe SJS/TEN and 10 with mild maculopapular exanthema), 98 osimertinib-tolerant subjects, and 2,123 general population controls. We performed HLA genotyping, drug-induced lymphocyte activation test, and surface plasmon resonance assay. Results: HLA-B∗51:02 was present in 83.3% of osimertinib-induced SJS/TEN patients but in only 3.3% of the general population controls (P = 2.8 × 10-7; corrected P = 6.9 × 10-6; odds ratio [OR] = 146), and 0% of osimertinib-tolerant controls (P = 6.5 × 10-8; corrected P = 1.6 × 10-6; OR = 707). The association of HLA-B∗51:01 and HLA-A∗24:02 with osimertinib-induced maculopapular exanthema patients, rather than with osimertinib-tolerant subjects (P = .002, OR = 15.7 for HLA-B∗51:01; and P = .003, OR = 9.5 for HLA-A∗24:02), was identified as a phenotype-specific association. Granulysin, the SJS/TEN-specific cytotoxic protein, was significantly higher in plasma of SJS/TEN patients (39.8 ± 4.5 ng/mL; P < .001) and in in vitro lymphocyte activation test (sensitivity = 83.3%; P < .01) compared with tolerant controls. Patients with osimertinib-induced hypersensitivity appeared to tolerate alternative EGFR-TKIs. Surface plasmon resonance results also confirmed that HLA-B∗51:02 protein has a higher binding affinity for osimertinib and lower or no affinity for other EGFR-TKIs. Conclusions: HLA-B∗51:02 frequently occurs in Asian populations and is strongly associated with osimertinib-induced SJS/TEN. Our findings suggest HLA-B∗51:02 screening as a preemptive test to reduce osimertinib-induced severe hypersensitivity.enHLA-B∗51:02HypersensitivityOsimertinibStevens-Johnson syndromeToxic epidermal necrolysis[SDGs]SDG3journal article10.1016/j.jaip.2024.10.02739505105