Acute impairment of saccadic eye movements is associated with delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage

OBJECTIVE Delayed cerebral ischemia (DCI) causing cerebral infarction remains a significant cause of morbidity and mortality following aneurysmal subarachnoid hemorrhage (aSAH). Early brain injury in the first 72 hours following rupture is likely to play a key role in the pathophysiology underlying DCI but remains difficult to quantify objectively. Current diagnostic modalities are based on the concept of vasoconstriction causing cerebral ischemia and infarction and are either invasive or have a steep learning curve and user variability. The authors sought to determine whether saccadic eye movements are impaired following aSAH and whether this measurement in the acute period is associated with the likelihood of developing DCI. METHODS As part of a prospective, observational cohort study, 24 male and female patients (mean age 53 years old, range 31-70 years old) were recruited. Inclusion criteria included presentation with World Federation of Neurosurgical Societies (WFNS) Grades 1 or 2 ("good grade") aSAH on admission and endovascular treatment within 72 hours of aneurysmal rupture. DCI and DCI-related cerebral infarction were defined according to consensus guidelines. Saccadometry data were collected at 3 time points in patients: in the first 72 hours, between Days 5 and 10, and at 3 months after aSAH. Data from 10 healthy controls was collected on 1 occasion for comparison. RESULTS Age-adjusted saccadic latency in patients was significantly prolonged in the first 72 hours following aSAH when compared with controls (188.7 msec [95% CI 176.9-202.2 msec] vs 160.7 msec [95% CI 145.6-179.4 msec], respectively; p = 0.0054, t-test). By 3 months after aSAH, there was no significant difference in median saccadic latency compared with controls (188.7 msec [95% CI 176.9-202.2 msec] vs 180.0 msec [95% CI 165.1-197.8 msec], respectively; p = 0.4175, t-test). Patients diagnosed with cerebral infarction due to DCI had a significantly higher age-adjusted saccadic latency in the first 72 hours than those without infarction (240.6 msec [95% CI 216.7-270.3 msec] vs 204.1 msec [95% CI 190.7-219.5 msec], respectively; p = 0.0157, t-test). This difference was more pronounced during Days 5-10 following aSAH, the peak incidence for DCI (303.7 msec [95% CI 266.7-352.7 msec] vs 207.6 msec [95% CI 193.7-223.6 msec], respectively; p < 0.0001, t-test). A binary generalized linear model showed that latency in the first 72 hours was the only significant predictor of cerebral infarction (p = 0.0185). CONCLUSIONS This is the first study to use saccadometry to measure the saccadic latency of eye movements in patients with aSAH during the acute period following aneurysm rupture. The results showed that median saccadic latency is associated with the risk of developing cerebral infarction due to DCI and may act as a potential objective biomarker to guide the need for intensive care admission and treatment. Future studies will look to formally validate saccadic latency as a biomarker of DCI in a larger cohort and assess whether the addition of saccades improves current clinical models for predicting patients at risk.