Association of Immune Checkpoint Inhibitors With Neurologic Adverse Events: A Systematic Review and Meta-analysis.

Importance: Neurologic adverse events (NAEs) due to immune checkpoint inhibitors (ICIs) can be fatal but are underexplored. Objective: To compare NAEs reported in randomized clinical trials (RCTs) of US Food and Drug Administration-approved ICIs with other forms of chemotherapy and placebo. Data Sources: Bibliographic databases (Embase, Ovid, MEDLINE, and Scopus data) and trial registries (ClinicalTrials.gov) were searched from inception through March 1, 2020. Study Selection: Phase II/III RCTs evaluating the use of ICIs were eligible for inclusion. Unpublished trials were excluded from the analysis. Data Extraction and Synthesis: Two investigators independently performed screening of trials using the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guideline. NAEs were recorded for each arm. Data were pooled using a random-effects model. Main Outcomes and Measures: The risk of NAEs with ICI use compared with any drug regimen, cytotoxic chemotherapy, and placebo. Results: A total 39 trials including 23 705 patients were analyzed (16 135 [68.0%] men, 7866 [33.1%] White). The overall risk of a NAE was lower in the ICI group (risk ratio [RR], 0.59; 95% CI, 0.45-0.77) and in the subgroup of RCTs comparing ICI use with chemotherapy (RR, 0.22; 95% CI, 0.13-0.39). In the subgroup of RCTs comparing ICI with placebo, the overall risk of NAE was significantly higher in the ICI group (RR, 1.57; 95% CI, 1.30-1.89). Peripheral neuropathy (RR, 0.30; 95% CI, 0.17-0.51) and dysgeusia (RR, 0.41; 95% CI, 0.27-0.63) were significantly lower in the ICI group. Headache was more common with the use of ICIs (RR, 1.32; 95% CI, 1.10-1.59). In the subgroup analysis of RCTs comparing ICI use with chemotherapy, peripheral neuropathy (RR, 0.09; 95% CI, 0.05-0.17), dysgeusia (RR, 0.42; 95% CI, 0.21-0.85), and paresthesia (RR, 0.29; 95% CI, 0.13-0.67) were significantly lower in the ICI group. RCTs comparing ICIs with placebo showed a higher risk of headache with ICI use (RR, 1.63; 95%, CI, 1.32-2.02). Conclusions and Relevance: Results of this meta-analysis suggest that the overall risk of NAEs, peripheral neuropathy, and dysgeusia is lower with the use of ICI. When compared with chemotherapy, the overall risk of NAE, peripheral neuropathy, paresthesia, and dysgeusia was lower with ICI use; however, when compared with placebo, the risk of NAEs is higher with the use of ICI.