Perioperative Stroke: Evaluation and Management

Perioperative stroke is a potentially devastating complication in patients undergoing noncardiac, nonneurological surgery. Although the overall incidence is relatively low (0.1–1.0%), the consequences are severe—with 30-day mortality rates up to 8-fold higher than controls and significantly increased long-term disability. This topic covers preoperative risk stratification, intraoperative prevention strategies, postoperative stroke recognition, and acute treatment considerations based on the 2021 AHA/ASA Scientific Statement.

Definition and Epidemiology

Perioperative stroke is defined as any embolic, thrombotic, or hemorrhagic cerebrovascular event with motor, sensory, or cognitive dysfunction lasting ≥24 hours, occurring intraoperatively or within 30 days after surgery. The vast majority of perioperative strokes are ischemic rather than hemorrhagic (hemorrhagic strokes represent only ≤5% of cases).

Incidence by Surgery Type

The incidence varies substantially by procedure and patient population:

In a large prospective international study of >40,000 patients undergoing noncardiac surgery, the overall rate of perioperative stroke was 0.3%, with higher risk associated with vascular and neurosurgical operations. Notably, between 2004 and 2013, perioperative stroke rates increased from 0.52% to 0.77% despite an overall decrease in other cardiovascular complications.

Silent Cerebral Ischemia

Silent brain infarctions are acute ischemic events that are not clinically apparent, identified only by brain imaging. These "covert" strokes are associated with cognitive decline, dementia, increased risk of future stroke, and increased mortality.

• After carotid endarterectomy (CEA): Up to 17% of patients
• After carotid artery stenting (CAS): 30–50% of patients
• After cardiac surgery: New infarcts on MRI in 25–55% (highest with open valve replacement; up to 90% with transcatheter aortic valve replacement)
• After noncardiac surgery: Up to 10% in a prospective multicenter pilot study of patients >65 years; 7% in a larger study of 1,114 patients (NeuroVISION study)

Silent perioperative strokes carry clinical significance: 1-year risk of cognitive decline is nearly 2-fold higher in patients with silent perioperative stroke compared to those without. Overt stroke and delirium were also higher in the silent cerebral infarct group.

Impact of Perioperative Stroke

• 30-day mortality rates: up to 8-fold higher than controls (absolute rates 21–26%)
• Increased length of stay and discharge to long-term care
• >5 million patients >45 years undergo noncardiac surgery annually in the US; >25,000 may have stroke attributable to perioperative causes

Pathophysiology

The mechanisms of perioperative stroke differ by surgery type and are often multifactorial:

Cardiac Surgery

• Nearly two-thirds are from proximal embolism—direct cardiac/arterial manipulation or bypass pump
• Delayed complications: postoperative atrial fibrillation, myocardial infarction

Noncardiac, Nonneurological Surgery

• Hypotension/low-flow states—reduced cerebral perfusion during anesthesia
• Previously undisclosed large-artery stenosis—unmasked by hemodynamic changes
• Anemia-associated tissue hypoxia—reduced oxygen delivery to watershed zones
• Thromboembolism (cardiac and transcardiac)—paradoxical embolism through PFO
• Fat embolism—especially with orthopedic procedures
• Enhanced coagulability/thrombosis—systemic inflammation, endothelial dysfunction
• Recent antithrombotic cessation—rebound prothrombotic state

Preoperative Risk Stratification

Multiple cardiovascular risk stratification tools have been used to predict perioperative complications. While none were specifically designed for stroke prediction, some show useful predictive accuracy:

Risk Calculators

The AHA/ASA statement recommends using the web-based ACS surgical risk calculator to identify patients with elevated stroke risk, as it exhibited the best overall predictive accuracy in a large retrospective cohort of >540,000 noncardiac surgery patients (overall stroke incidence 0.27%).

Established Risk Factors

• Prior stroke or TIA—the single strongest predictor
• Advancing age
• Renal disease
• Atrial fibrillation
• Myocardial infarction within 6 months
• Hypertension, COPD, current smoking, female sex, diabetes mellitus
• Emergency surgery or high-risk procedures (thoracic, head and neck, intra-abdominal, vascular, transplant, orthopedic)

Preoperative Stroke Prevention Strategies

Timing of Surgery After Stroke

The timing of surgery relative to a prior stroke is a critical determinant of perioperative stroke risk. Data from the Danish national health care database demonstrates a striking temporal relationship:

Extracranial Carotid Artery Stenosis

Management of carotid stenosis in the perioperative setting depends on symptom status and stenosis severity:

• Symptomatic, high-grade (>70%) stenosis with stroke/TIA within 6 months: Strongly consider revascularization (CEA or CAS) before planned surgery
• Symptomatic, moderate (50–69%) stenosis: Consider revascularization if surgical risk <6%
• Asymptomatic, high-grade stenosis: Uncertain benefit of preoperative revascularization—guidelines recommend consideration for CEA if perioperative stroke/MI/death risk <3%, or CAS if stenosis >60% by angiography
• Routine carotid screening in asymptomatic patients without symptoms is not recommended

CAS carries a slightly higher risk of stroke, while CEA has a slightly higher risk of MI. The CREST-2 trial (ongoing) is testing whether revascularization plus optimal medical therapy is superior to optimal medical therapy alone for asymptomatic high-grade stenosis.

Intracranial Stenosis

Symptomatic intracranial stenosis (50–99%) carries a stroke risk of 15% over the first year. It is routinely managed medically with antithrombotic therapy and risk factor modification. No evidence supports prophylactic stenting in the preoperative setting. A small retrospective study of 38 patients with severe intracranial stenosis undergoing surgery reported a perioperative stroke rate of 6%.

Patent Foramen Ovale (PFO)

• PFO is present in ≈25% of the general population
• A retrospective study of >150,000 patients found that PFO diagnosis before surgery significantly increased the likelihood of perioperative ischemic stroke
• Perioperative strokes in PFO patients were more severe and more likely to involve large vessel territory
• Hip arthroplasty patients with PFO/ASD had perioperative stroke risk 29× greater (7.14% vs 0.26%)
• If PFO closure is already planned, do not delay it for elective surgery; urgent/emergency surgery should not be delayed to address PFO

Perioperative Medication Management

Beta-Blockers

The POISE trial (2008) was a landmark study that unexpectedly showed harm from perioperative metoprolol:

• Patients receiving metoprolol had significantly higher overall mortality and were twice as likely to have a perioperative stroke (1.0% vs 0.5%; HR 2.17, 95% CI 1.26–3.74; P=0.005)
• Significant hypotension (SBP <90 mmHg) was associated with a 2-fold higher odds of stroke
• Hypotension and bradycardia were the likely mediators

Current guidance: Continue beta-blockers in patients already on them. Do not initiate beta-blockers on the day of surgery. It may be reasonable to start beta-blockade in advance for high-risk patients.

Statins

• Perioperative statin continuation is recommended for patients already taking statins
• Perioperative statin use is associated with reduced postoperative MI, all-cause mortality, and new-onset atrial fibrillation
• No studies have shown a significant decrease in perioperative stroke specifically with statin use

Antithrombotic Medications

Managing antithrombotics perioperatively requires balancing thromboembolic risk against surgical bleeding risk:

Thromboembolism Risk Stratification

For patients on anticoagulation, perioperative management should be guided by the thromboembolic risk category:

Intraoperative Management

Blood Pressure Management

Maintaining adequate end-organ perfusion is the cornerstone of anesthetic management for stroke prevention. Drops in blood pressure are very common during surgery, with MAPs falling below 20% of baseline in up to 90% of surgical cases.

• Hypotension was associated with a 2-fold higher risk of ischemic strokes in a large cohort study (POISE data)
• MAPs <65 mmHg lasting >10 minutes are associated with moderate increases in end-organ injury and mortality (OR/RR/HR 1.4–2.0)
• Maintaining SBP within 10% of baseline was associated with a 30% reduction in postoperative organ dysfunction in one multicenter RCT

Perioperative Goal-Directed Therapy

Rather than targeting specific BP numbers alone, goal-directed hemodynamic management optimizes end-organ perfusion and oxygen delivery:

• A Cochrane meta-analysis (2013, 31 RCTs, 5,092 patients) showed that goal-directed therapy reduced complications by 32%, including 29% reduction in renal impairment and 49% reduction in respiratory failure
• OPTIMISE trial: Cardiac output-guided therapy during major GI surgery showed a 6.8% lower absolute risk of complications (not statistically significant, P=0.07)
• FEDORA trial: Nearly 50% fewer moderate/severe postoperative complications in low-moderate risk noncardiac surgery patients with goal-directed therapy
• None of these trials showed significant reductions in strokes specifically, likely due to the low incidence

Blood Transfusion Management

Anemia is a risk factor for stroke in the general population but perioperative data are conflicting. The AHA/ASA statement recommends a tiered transfusion approach:

Anesthesia Type

• No clear evidence that general vs regional anesthesia significantly affects perioperative stroke risk
• Anesthetic agents are neither clearly neuroprotective nor neurotoxic in humans
• The Balanced Anesthesia Study (6,644 patients): no difference in stroke between light vs deep general anesthesia
• ENIGMA-II (7,112 patients): nitrous oxide did not increase stroke or cardiovascular complications
• Two large retrospective studies (528,495 and 182,307 patients) suggest neuraxial anesthesia may be weakly associated with fewer strokes in hip surgery
• A meta-analysis of 1.1 million records found no difference in 30-day mortality between neuraxial and general anesthesia for major limb surgery

Ventilation Strategies

• Avoid hyperventilation and hypocapnia—causes cerebral vasoconstriction and reduced cerebral blood flow
• Hypocapnia shifts the cerebral autoregulatory curve downward, reducing perfusion at any given MAP
• Particularly harmful in patients with pre-existing cerebral ischemia
• Use lung-protective ventilation (lower tidal volumes) as part of an overall strategy to reduce pulmonary complications

Summary of Preoperative and Intraoperative Strategies

Postoperative Stroke Recognition

Recognizing stroke in the perioperative setting is uniquely challenging. Between 4% and 17% of all strokes occur in patients already hospitalized, and outcomes are worse than community-onset strokes due to delays in recognition and treatment.

Diagnostic Challenges

• Masking by anesthesia: Residual effects of anesthetics, opioids, and psychoactive medications obscure neurological examination
• Atypical presentations: Mental status changes, agitated delirium, autonomic instability, or delayed emergence may be the only signs
• Exact onset often unknown: Time of "last known well" may be the anesthesia induction time
• In-hospital stroke patients are significantly less likely to receive thrombolytic therapy and experience greater delays to imaging and treatment

Emergent Evaluation: "Code Stroke"

When perioperative stroke is suspected, an institutional stroke code and rapid response stroke team should be activated immediately, following the same time-critical protocols as for community-presenting strokes.

Perioperative In-Hospital Stroke Response Protocol

Acute Treatment of Perioperative Stroke

Mechanical Thrombectomy

Mechanical thrombectomy is the preferred treatment for perioperative stroke patients with large vessel occlusion (LVO), because it does not require systemic anticoagulation and thus minimizes surgical site bleeding risk.

• LVO incidence: >30% in acute ischemic strokes overall; 10.9% in perioperative strokes after cardiac surgery
• Patients with NIHSS >6 and/or cortical deficits should undergo immediate CT angiography and perfusion studies
• Eligible up to 24 hours from symptom onset based on advanced imaging criteria (DAWN, DEFUSE 3)
• Can be performed without systemic anticoagulation
• In a case-control study of 25 perioperative stroke patients, successful reperfusion was achieved in 76% with an 8% symptomatic hemorrhage rate
• High-risk procedures should ideally be performed only at thrombectomy-capable centers, with transfer protocols established for non-capable facilities

Intravenous Thrombolysis (IV tPA)

IV alteplase use in perioperative stroke is more controversial due to the risk of surgical site bleeding:

• Absolute contraindication: Intracranial or intraspinal surgery within the past 3 months
• Several retrospective studies identified patients with recent noncranial surgery who received IV tPA with increased surgical site hemorrhages but no major complications
• In a study of 134 patients treated with IV tPA within 90 days of surgery (including 49 within 1–10 days): surgical site hemorrhage in 7%, serious in only 3%
• Surgical site hemorrhage was more likely with more recent surgery
• Extended-window thrombolysis (4.5–9 hours, imaging-guided) may be considered based on WAKE-UP, EPITHET, ECASS-4, and EXTEND trials, but with extreme caution in surgical patients

Reversal Agents for Emergency Surgery

When emergency surgery is needed in anticoagulated patients or when perioperative hemorrhagic stroke occurs:

Summary: Identification and Treatment of Perioperative Stroke