Eye Movement and Coma Recovery after Cardiac Arrest
Michael Cronin1, Laura Hwang2, Audrey Wack1, Brian Pugsley3, Mary Russo3, David Greer1, Charlene Ong1
1Boston University School of Medicine, 2Boston University, 3Boston Medical Center
Objective:
To characterize eye movements in comatose patients after cardiac arrest (CA).
Background:
Most patients resuscitated from CA present with altered levels of consciousness due to brain injury. One potential marker of arousal recovery is evidence of eye movements. Brainstem pathways and thalamocortical projections responsible for eye movements share similar anatomic structures responsible for arousal. Our aim is to quantify eye movements using electro-oculography (EOG) data from patients presenting with CA.
Design/Methods:

We are conducting a retrospective study of patients presenting within 24 hours of CA. We identify eye movements on EOG over a period of up to 72 hours, excluding periods of clinical care, sedation, and epileptiform activity. We will study the association of eye movement frequency and the following outcomes by two weeks: 1. Present corneal or pupillary reflex; 2. Eye opening (to stimuli or spontaneously); 3. Command following. Patients are excluded if they have ophthalmoplegia, dementia, or traumatic brain injury. We will compare eye movement frequency in our comatose cohort to a separate cohort of non-comatose patients who underwent sleep studies.


Results:
Our preliminary data consists of 100 patients, mean (± SD) age of 60.2 (± 15.3) years and 38% were female. By 2 weeks, 85% had a pupillary reflex, 75% had a corneal reflex, 43% opened eyes to stimuli, and 21% followed commands. Of 151.5 hours of EOG time across four patients, we preliminarily found eye movements present in 0.56% of total time, with a median of 0.73% and range of 0% to 2.29%. This is comparable to our cohort of 47 patients who underwent sleep studies, with eye movements present in 3.2% of N2, 2.9% of N3, and 19.8% of REM.
Conclusions:
Eye movements in comatose patients may be indicative of recovery potential. Understanding these differences could have prognostic significance and help in triaging care and family discussions.
10.1212/WNL.0000000000203878