Safety and efficacy of immune checkpoint inhibitors (ICI) in patients (pts) with pre-existing neurologic autoimmune diseases (NAID) and Parkinson's disease.

Abstract

2653 Background: Patients with pre-existing NAID often flare with ICIs; neurological (neuro) immune related adverse events (irAEs) are often morbid or fatal, and little is known about safety of ICI in pts with NAID and other neuro conditions (e.g. Parkinson’s). Thus, we aimed to determine their safety and efficacy in these contexts. Methods: We retrospectively analyzed 71 pts from 5 institutions receiving ICIs with NAIDs and Parkinson’s disease. NAIDs included multiple sclerosis (MS), myasthenia gravis (MG), inflammatory neuropathy, transverse myelitis, Lambert-Eaton, myotonic dystrophy, and multifocal motor neuropathy. We collected demographics, cancer outcomes, NAID characteristics, and safety outcomes. We used descriptive statistics to analyze treatment outcomes, NAID flares, and irAEs. Results: We collected 71 pts, 40 with NAID and 31 with Parkinson’s. 24 had melanoma, 12 non-small cell lung cancer (NSCLC), 3 small cell lung cancer (SCLC), 6 urothelial/bladder, 4 renal cell carcinoma, and 22 other; 60 had metastatic and 11 had localized disease. Median age at ICI start was 72 years (61% male, 39% female); 8 received combination ICI and 63 received monotherapy. Of 40 pts with NAID, 30% had either NAID flares or neurologic irAEs and 33% experienced other (non-neuro) irAEs (Table). Toxicities were particularly severe in MG; 70% (n=7) had MG flare/neuro irAE, 3 of which were fatal. In addition, 31 had Parkinson’s; of these, 12.9% (n=4) had Parkinson’s flares, 3% (n=1) had neuro-irAEs, and 42% (n=13) had other irAEs. One pt died of grade 5 myocarditis/myositis following combination ICI (Table). 5 MS patients responded to therapy (23%), but none were those with NAID flare/neuro irAE. Both MG patients who had non-neuro irAEs responded to treatment. Conclusions: In this cohort of ICI treated pts with prior neuro disorders, we demonstrated that MG pts have high rates of MG flare/neuro irAEs, hospitalizations, and fatalities, but low response rates. In contrast, other NAID (particularly MS) as well as Parkinson’s appeared to have modest risks of flares/neuro irAEs. Notably, Parkinson’s pts had high rates of non-neuro irAEs and response rates. [Table: see text]