Vardan Nersesjan1, Oskar McWilliam1, Lars-Henrik Krarup1, Daniel Kondziella2. 1. From the Department of Neurology (V.N., O.M., L.-H.K., D.K.), Rigshospitalet, Copenhagen University Hospital; and Department of Clinical Medicine (D.K.), Faculty of Health and Medical Sciences, University of Copenhagen, Denmark. 2. From the Department of Neurology (V.N., O.M., L.-H.K., D.K.), Rigshospitalet, Copenhagen University Hospital; and Department of Clinical Medicine (D.K.), Faculty of Health and Medical Sciences, University of Copenhagen, Denmark. daniel_kondziella@yahoo.com.
Abstract
OBJECTIVE: To determine the clinical and laboratory features of immune checkpoint inhibitor (ICPI)-associated autoimmune encephalitis (ICPI-AIE), an increasingly recognized adverse event with ICPI treatment. METHODS: We searched PubMed, The Cochrane Library, and Embase for ICPI-AIE cases from the first description in 2015 until January 2020 using standard bibliographic measures including PRISMA guidelines and preregistration with PROSPERO. RESULTS: Thirty-nine studies met inclusion criteria, resulting in 54 patients with ICPI-AIE (mean age 58.6 years; 43% female). Common cancers included melanoma (30%) and non-small cell lung cancer (30%). Brain metastases were found in 16 patients (30%). The most frequent ICPI was nivolumab (61%). Onset of ICPI-AIE occurred after a median of 3.0 treatment cycles, but very early and late presentations were common. Nonlimbic AIE was roughly twice as frequent as limbic AIE (p < 0.05). The most common laboratory abnormalities included bitemporal fluid-attenuated inversion recovery lesions on MRI, continuous slow waves and diffuse slowing on EEG, and monocytic pleocytosis on CSF analysis. Intraneuronal antibodies were more frequent than neuronal surface antibodies and a significant predictor for lack of improvement after first-line immunotherapy (p < 0.05). CONCLUSIONS: ICPI-AIE consists of a heterogenous group of conditions. Neurologists will likely encounter ICPI-AIE more often in the future, but important unresolved questions include the pathophysiologic mechanisms, the epidemiology, and the best treatment approaches associated with ICPI-AIE.
OBJECTIVE: To determine the clinical and laboratory features of immune checkpoint inhibitor (ICPI)-associated autoimmune encephalitis (ICPI-AIE), an increasingly recognized adverse event with ICPI treatment. METHODS: We searched PubMed, The Cochrane Library, and Embase for ICPI-AIE cases from the first description in 2015 until January 2020 using standard bibliographic measures including PRISMA guidelines and preregistration with PROSPERO. RESULTS: Thirty-nine studies met inclusion criteria, resulting in 54 patients with ICPI-AIE (mean age 58.6 years; 43% female). Common cancers included melanoma (30%) and non-small cell lung cancer (30%). Brain metastases were found in 16 patients (30%). The most frequent ICPI was nivolumab (61%). Onset of ICPI-AIE occurred after a median of 3.0 treatment cycles, but very early and late presentations were common. Nonlimbic AIE was roughly twice as frequent as limbic AIE (p < 0.05). The most common laboratory abnormalities included bitemporal fluid-attenuated inversion recovery lesions on MRI, continuous slow waves and diffuse slowing on EEG, and monocytic pleocytosis on CSF analysis. Intraneuronal antibodies were more frequent than neuronal surface antibodies and a significant predictor for lack of improvement after first-line immunotherapy (p < 0.05). CONCLUSIONS: ICPI-AIE consists of a heterogenous group of conditions. Neurologists will likely encounter ICPI-AIE more often in the future, but important unresolved questions include the pathophysiologic mechanisms, the epidemiology, and the best treatment approaches associated with ICPI-AIE.