Abstract
BACKGROUND: Eye movements play an essential role in the assessment of the unconscious patient and offer a window to the function of the brain. We review the range of ocular motor and vestibular findings in patients with impaired consciousness and present a practical approach to these patients. METHODS: Based on a structured review of the literature (Pubmed, Embase) 54 suitable citations were identified amongst 4,241 total citations. A manual search of the reference list of selected papers added another 57 papers. Based on these publications the spectrum of eye movement abnormalities in the unconscious patient was characterized. RESULTS: The pattern of eye movement abnormalities seen in the unconscious patient depends on the underlying cause and the extent/location of brain damage. Conjugate eye deviations may be observed with either supratentorial or infratentorial lesions, while disconjugate deviations may indicate superimposed eye muscle palsies or decompensated strabismus. The presence of a full range of spontaneous horizontal, oscillatory eye movements (e.g., ping-pong gaze) in the comatose patient usually indicates bilateral cerebral hemisphere dysfunction. With vertical spontaneous eye movements, the identification of a slower and faster phase helps to distinguish between nystagmus and ocular bobbing and its variants. Combined with absent reflexively-induced eye movements, typical ocular bobbing strongly suggests a structural pontine lesion, whereas other vertical spontaneous eye movement patterns do not predict specific (focal) damage. The reflexive eye movements, i.e., the vestibulo-ocular reflex (VOR), can be assessed in comatose patients either by head rotations, caloric irrigation or galvanic stimulation. Intact slow-phase responses indicate relatively preserved brainstem function and inability to keep the eyes in an eccentric position suggest a deficient velocity-to-position integrator either from brainstem or cerebellar involvement. CONCLUSION: Ocular motor and vestibular testing in unconscious patients offer a unique opportunity to assess both brainstem and cerebellar function and its interplay with higher cortical areas. It may also help predict outcome. Challenges to overcome include a lack of standardized diagnostic approaches to unconscious patients. Quantitative eye movement analysis, based on videooculography (VOG) and artificial intelligence using large multimodal data sets are promising new tools for diagnosis, longitudinal observational studies and prediction of outcome.