Skip to main content

Supplementary material of article entitled: Neurological adverse events of immune checkpoint inhibitors: A systematic review of 428 cases

Cite this dataset

Marini, Alessandro (2021). Supplementary material of article entitled: Neurological adverse events of immune checkpoint inhibitors: A systematic review of 428 cases [Dataset]. Dryad.


Objective: To define the clinical characteristics, management, and outcome of neurological immune-related adverse events (n-irAEs) of immune checkpoint inhibitors (ICIs).

Methods: Systematic review of the literature following the PRISMA guidelines.

Results: A total of 694 articles were identified. Two hundred fifty-six articles, with 428 individual patients, met the inclusion criteria. Reports regarding neuromuscular disorders (319/428, 75%) were more frequent than those on central nervous system (CNS) disorders (109/428, 25%). The most common n-irAEs reports were: myositis (136/428, 32%), Guillain-Barré syndrome and other peripheral neuropathies (94/428, 22%), myasthenic syndromes (58/428, 14%), encephalitis (56/428, 13%), cranial neuropathies (31/428, 7%), meningitis (13/428, 3%), CNS demyelinating diseases (8/428, 2%), and myelitis (7/428, 2%). Other CNS disorders were detected in 25/428 (6%) patients. Compared to the whole sample, myasthenic syndromes were significantly more Ab-positive (33/56, 59%; p<0.001). Anti-PD-1/PD-L1 were more frequent in myasthenic syndromes (50/58, 86%; p=0.005) and less common in meningitis (2/13, 15%; p<0.001) and cranial neuropathies (13/31, 42%; p=0.005). Anti-CTLA-4 ICIs were more frequent in meningitis (8/13, 62%; p<0.001) and less common in encephalitis (2/56, 4%; p=0.009) and myositis (12/136, 9%; p=0.01). Combination of different ICIs was more frequent in cranial neuropathies (12/31, 39%; p=0.005). Tumor types were highly heterogeneous, but melanoma was more frequent in patients with peripheral neuropathies (64/94, 68%; p=0.003) and less common in those with encephalitis (19/56, 34%; p=0.001). The highest mortality rate was reached in myasthenic syndromes (28%).

Conclusion: Considering the increasing use of ICI therapy in the forthcoming future, this information can be valuable in assisting neurologists and oncologists in early n-irAEs
diagnosis and treatment.


Syndrome references were collected by means of a comprehensive search in PubMed, PubMed Central (PMC), Cochrane, Scopus, and Web of Science; additional articles were retrieved by searching through the reference list of all articles. References are divided by syndromes, as stated in the article. 

Supplementary tables file contains tables with distributions (value, percentages) and statistical analysis (with p values) of associations between ab positivity, ICIs used, outcome, with different syndromes. 

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist represents an evidence-based minimum set of items for reporting in Systematic Reviews. PRISMA flow diagram shows the process for selection of articles/patients, from the initial search to final selection (with reasons for exclusion).