Log in to MyACC
Perioperative Covert Stroke in Noncardiac Surgery Patients
Study Questions:
What is the relationship between clinically silent acute perioperative brain infarction (“covert stroke”) and incidence of cognitive decline 1 year after surgery?
Methods:
Previous research in nonsurgical settings has shown that brain infarction often precedes long-term cognitive impairment, but more often than not occurs without detection of a coincident, clinically recognized neurologic event. In this large multinational collaboration, the NeuroVISION (Non-cardiac Surgery Patients Cohort Evaluation) investigators have explored the relationship between covert stroke and incidence of cognitive decline 1 year after surgery.
Patients aged ≥65 years undergoing elective noncardiac surgery with ≥2 nights of anticipated hospital stay were enrolled. All underwent an array of neurocognitive assessments, including the Montreal Cognitive Assessment (MoCA) before and 1 year after surgery. Between 2-9 days after surgery, all underwent brain magnetic resonance imaging (MRI) employing specific scanning sequences including diffusion-weighted imaging to enable detection of acute infarction occurring within a prior 10-day period. Covert stroke was defined as MRI-determined acute infarction without concurrent clinical stroke symptoms. Postoperative cognitive decline was defined as ≥2 point decline in MoCA performance at 1 year compared to preoperative baseline. Secondary outcomes included delirium during the first 3 postoperative days, and composite of transient ischemic attack (TIA) or stroke, death, new-onset atrial fibrillation or flutter, and nonfatal myocardial injury or infarction within 1 year.
Results:
A total of 1,116 patients were recruited and underwent preoperative cognitive testing. After surgery, four were excluded from further analysis (two with clinically recognized stroke prior to MRI, and two with uninterpretable MRI findings). Of the remaining 1112, 78 patients (7.0%) had postoperative MRI findings showing acute brain infarction, without reported clinical stroke symptoms; in 10 of these 78 patients, multiple areas of acute infarction were observed. MRI findings consistent with chronic (age-indeterminant) brain infarction had similar prevalence in patients with versus without acute perioperative brain infarction.
At 1 year, cognitive decline was observed in 42% of patients with covert perioperative stroke versus 29% of patients without covert perioperative stroke, a 13% absolute difference, with multiple regression showing an adjusted odds ratio of 1.98 (1.22-3.20). Among secondary findings, delirium within the first 3 postoperative days occurred more often (10% vs. 5%; hazard ratio [HR], 2.24 [1.06-4.73]), and subsequent composite TIA/stroke was more frequent in patients with versus without covert perioperative stroke (4% vs. 1%; HR, 4.13 [1.14-14.99]). No statistically significant difference in frequency of atrial fibrillation, myocardial injury/infarction, or death was observed between patients with versus without covert stroke. Between the 60% of patients receiving general anesthesia versus 40% receiving spinal anesthesia, no difference in cognitive decline was observed.
Conclusions:
Findings from the NeuroVISION cohort demonstrate that most perioperative strokes are asymptomatic or fail to raise clinical suspicion, consistent with observations in nonoperative settings. In 7% of these surgical patients, covert stroke was identified only on the basis of postoperative imaging; although otherwise clinically unrecognized, this condition had far-reaching clinical significance, twice as often associated with perioperative delirium, four times as often associated with subsequent stroke, and independently associated with 1-year cognitive decline. Chosen anesthetic technique showed no association with cognitive outcome.
Perspective:
This study provides unique perspective on incidence and sequelae of perioperative stroke, a condition with profound and lasting impact on our surgical patients and their families. The apparent frequency of perioperative stroke events that escape clinical recognition underscores the need for more sensitive and practical risk reduction strategies. Interpretation of the high rate of 1-year decline in neurocognitive performance within the overall surgical cohort would be facilitated by a parallel study in an age-matched control population.
Clinical Topics: Arrhythmias and Clinical EP, Cardiovascular Care Team, Geriatric Cardiology, Noninvasive Imaging, Prevention, Atrial Fibrillation/Supraventricular Arrhythmias, Magnetic Resonance Imaging
Keywords: Anesthesia, General, Anesthesia, Spinal, Atrial Fibrillation, Brain Infarction, Cognition, Delirium, Diagnostic Imaging, Geriatrics, Ischemic Attack, Transient, Magnetic Resonance Imaging, Myocardial Infarction, Perioperative Period, Secondary Prevention, General Surgery, Vascular Diseases
< Back to Listings
