Immune Checkpoint Inhibitors in Patients with Pre-existing Neurologic Autoimmune Disorders

The use of immune checkpoint inhibitors (ICIs) for oncologic indications is associated with immune-related adverse events (irAEs). Patients with pre-existing autoimmune diseases are at increased risk of irAEs and have largely been excluded from clinical trials of ICIs. Therefore, there is limited data on the safety of safety of ICIs in patients with pre-existing neurologic autoimmune diseases (nAIDs) such as myasthenia gravis and multiple sclerosis. This review aims to synthesize the literature on the post-marketing experience with ICI in patients with pre-existing nAID and to discuss possible strategies for mitigating the risk of post-ICI nAID relapses.

Patients with pre-existing myasthenia gravis (MG), myositis, and paraneoplastic encephalitis appear highly susceptible to neurologic relapses of their underlying neurologic disorder following ICI initiation; these relapses can cause considerable morbidity and mortality. In patients with multiple sclerosis (MS), the risk and severity of MS relapses following ICI appears to be relatively lower compared to MG. Preliminary evidence suggests that older MS patients with no recent focal neuroinflammatory activity may be safely treated with ICI. Among the several case reports of ICI in patients with a history of Guillain-Barre syndrome (GBS), neurologic worsening was only recorded in one patient who was in the acute phase of GBS at the time of ICI start. Initiating an ICI in a patient with pre-existing nAID involves a complex risk-benefit discussion between the patient, their oncologist, and neurologist. Relevant issues to consider before ICI include the choice of disease-modifying therapy for nAID (if any) and strategies for promptly identifying and managing nAID relapses should they occur. Currently, the literature consists mainly of case reports and case series, subject to publication bias. Prospective studies of ICI in patients with nAID are needed to improve the level of evidence.