Abstract: Objective: To promote safe upadacitinib prescribing by mining and evaluating the upadacitinib adverse reaction events(ADE) signals from the FDA Adverse Event Reporting System (FAERS). Methods: In the FAERS, “upadacitinib” was used as a search term to screen ADEs. ADE signals were collected from August 2019 to March 2023 and were detected. The analysis method included four algorithms of reporting odd ratio (ROR), proportional reporting ratio (PRR), bayesian configuration promotion neural network (BCPNN), and multi item gamma poisson shrinker (MGPS). ADE reports were described and classified according to the System Organ Class (SOC) and Preferred Term (PT) in the Medical Dictionary for Regulatory Activities (version 24.0). Results: 21 213 ADEs were collected with an upward trend, including 245 cases of tumor and 78 cases of thrombosis. A total of 357 ADE signals were obtained by four methods. The top signal intensities were rheumatoid factor positive, pelvic floor muscle weakness, eczema herpeticum, spinal cord infection, oral lichen planus, of which rheumatoid factor positive, pelvic floor muscle weakness, spinal cord infection, oral lichen planus were not included in the package insert. Reported multiple ADE signals mainly involved arthralgia, novel coronavirus infection, nasopharyngitis, urinary tract infection, of which arthralgia, novel coronavirus infection,  urinary tract infection were not included in the package insert. The ADE signal involved in 15 SOCs, mainly were infections (5490 cases, 37.57%), musculoskeletal and connective tissue diseases (3370 cases, 23.06%). Conclusion: Our research has found new significant ADE signals of upadacitinib such as in COVID-19 and urinary tract infection, etc. In addition, the risk of malignant tumor and thrombosis can not be ignored.

Key words: upadacitinib, FAERS, adverse drug event, data mining, black box warning, pharmacovigilance