Immune-related adverse events (irAEs) commonly involve the gastrointestinal tract, endocrine glands, skin, and liver, and rarely the nervous system. The pathomechanism of irAEs in the nervous system is unclear, and so characterizing these severe toxic effects is a priority, even if irAEs are uncommon in the nervous system. Our patient presented subacute
muscle weakness and
dysesthesia with
colitis as irAEs caused by
pembrolizumab, one of the anti-programmed death-1 (PD-1)
antibodies. Electromyography revealed abundant fibrillations and
fasciculations of upper and lower extremities and severe reduction in motor unit potentials; however,
antineutrophil cytoplasmic antibodies,
rheumatoid factor,
autoantibodies against Hu and Yo, and anti-
ganglioside antibodies, such as GQ1b, were undetectable in the serum. Although he was treated with high-dose
glucocorticoids,
antibiotics, and a monoclonal anti-
tumor necrosis factor alpha (TNFα) antibody, he developed colonic perforation. The total colorectal resection was performed, and the resected colon showed mucosal defect and perforation. He died of
lung aspergillosis. Postmortem examination revealed CD8-positive lymphocyte infiltration around neurons of dorsal root ganglia. The sciatic nerve displayed the widening of myelin laminae and thinning of myelinated fibers but not a decrease in the density of myelinated nerve fibers. In the sural nerve, the density of myelinated fibers slightly decreased, and some fibers showed less densely myelinated laminae. Drug safety information, including previous randomized trials of anti-PD-1 and anti-cytotoxic T-lymphocyte-associated antigen-4 (CTLA-4)
antibodies, showed that patients treated with anti-PD-1
antibodies appeared to have more frequent and severe
peripheral neuropathies compared to those in patients who received anti-CTLA-4
antibodies (1.59% vs. 0.69%; Fisher exact test, P < 0.001; three severe events vs. zero severe events). The present results and drug safety information suggest that the pathomechanism of irAEs caused by anti-PD-1
antibodies is different from that by anti-CTLA-4
antibodies. The neurological irAEs might be clues to solving the pathomechanism of irAEs.