Abstract

Immune checkpoint inhibitors (ICIs) are associated with peculiar adverse events related to the mechanism of action. Less than 1% of patients treated with ICIs develop autoimmune encephalitis. The aim of this study was to compare the frequency of encephalitis development due to ICIs with encephalitis due to other drugs using the Japanese Adverse Drug Event Report (JADER) database and Bayesian confidence propagation neural networks for signal detection. Data from the JADER database from April 2004 to December 2020 were downloaded via the Pharmaceuticals and Medical Devices Agency (PMDA) website. The Information Component (IC) values were calculated as an index of signal detection based on the Bayesian method. The lower bound of the 95% credible interval (CI) of the IC values for atezolizumab and pembrolizumab were greater than 0 in most of the periods. Thus, encephalitis occurred more frequently for atezolizumab and pembrolizumab than for other drugs. For nivolumab and ipilimumab, a significant signal was detected only for recent data. In contrast, the lower bounds of the 95% CIs for avelumab and durvalumab were smaller than 0 in most of the periods because encephalitis was seldom reported for avelumab and durvalumab. We showed that encephalitis occurs more frequently for atezolizumab, pembrolizumab, nivolumab, and ipilimumab compared with the frequency for other drugs. The time of onset varied widely, and patients should be monitored for more than 1year after the last administration of ICIs.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call