Abstract: Objective To explore the risk signals of lenvatinib related adverse drug reactions (ADR) and provide reference for clinical safe applications. Methods The report odds ratio (ROR) and proportional reporting ratio (PRR) methods were used to mine the adverse reactions of the suspected drug-lenvatinib-from the FDA’s adverse event reporting system (FAERS) between the first quarter of 2015 and the fourth quarter of 2019. The detection threshold for these adverse reactions was that the number of reports was larger than 10, both ROR and PRR exceeded 2 and the 95% confidence interval (CI) lower limit of the ROR and PRR was above 1. Results During the given period of time, the FAERS database received 7 922 reports of adverse reactions with lenvatinib as the PS, and detected 191 risk signals, involving 18 SOC(systems on chips). There were 52 risk signals that were not mentioned in the manual, which were mainly concentrated in 8 signals of the gastrointestinal system, 7 signals of hepatobiliary diseases, 6 signals of benign, malignant and unspecified tumors (including cysts and polyps), 6 signals of abnormal laboratory examination, and 5 signals of infection and invasion. Among the top 50 adverse reaction signals of lenvatinib in terms of intensity, 12 adverse signals were not mentioned in the manual, mainly distributed in benign, malignant and unspecified tumors (4 signals), hepatobiliary diseases (3 signals). Conclusion The discovery of suspicious ADR signals of lenvastinib can provide reference for the safe use of this drug in treating hepatocellular carcinoma and renal cancer in China.

Key words: lenvatinib, FDA adverse event reporting system, adverse drug reaction, data mining