Oculomotor abnormalities in patients with cerebellar ataxia, neuropathy, vestibular areflexia syndrome (CANVAS) reflect midline cerebellar impairment

Abstract Background Biallelic intronic repeat expansions in the RFC1 gene are associated with the cerebellar ataxia, neuropathy, and vestibular areflexia (bilateral vestibulopathy) syndrome (CANVAS). Oculomotor abnormalities may serve as a state marker exclusively reflecting the midline cerebellar involvement in CANVAS, i.e., independent of the combined disorders affecting the patient’s postural control stance and gait. Methods Slow and fast eye movements of 15 CANVAS patients and 14 healthy subjects were compared using a high-resolution video-based eye tracker allowing to record visually-guided saccades, gaze-holding function and smooth pursuit paradigms. Scores of cognitive impairment were related to oculomotor performance. Results Saccades (latency, metria, velocity) were normal. Small amplitude omnidirectional gaze-holding deficit was found in 70% of patients, with downbeat nystagmus (60%) being more common than upbeat nystagmus (13%). Latency of initial acceleration of smooth pursuit was prolonged and there was severe impairment of smooth pursuit eye movements. Montreal Cognitive Assessment (MoCA) scores were lower in patients and correlated with saccade and pursuit latency and initial acceleration. Disease duration and vestibulopathy correlated with no oculomotor abnormalities. Conclusion Cerebellar oculomotor dysfunction affected mainly smooth pursuit and gaze holding function at eccentric gaze positions, but it did neither comprise spontaneous nystagmus nor saccade abnormalities. Prolonged latencies in initial pursuit acceleration might be related to the patient’s cognitive decline, but normal saccade latencies point to cerebellar oculomotor neurodegeneration. Smooth pursuit impairment was not related to disease duration and vestibulopathy, contrariwise pursuit impairment became worse with larger functional impairment. Oculomotor abnormalities in CANVAS are in line with midline cerebellar impairment without evidence for extracerebellar (brainstem) brain involvement.