Perioperative stroke is a devastating yet often under-recognised complication of surgery
Perioperative stroke may be generally define as:
Brain infarction of ischaemic or haemorrhagic aetiology that occurs during surgery or within 30 days after surgery
It can be further subdivided into:
Overt stroke: acute brain infarct with clinical manifestation lasting longer than 24 h
Covert stroke: stroke which isn't recognised at the time of onset because of unappreciated, subtle or misclassified clinical manifestations but is detected on brain imaging done at the time or subsequently
The rate of overt perioperative stroke varies by the studied surgical population
Overall it is described as occurring in 0.1 - 0.8% of non-cardiac, non-neurological surgical patients
The rate of perioperative ischaemic stroke may be increasing over time despite falling rates of perioperative mortality, MACE and non-perioperative stroke
This may be due to an ageing population and/or the increasing prevalence of known risk factors
Timecourse
Peak incidence of perioperative stroke is 1 - 2 days post-operatively
Up to 10% of perioperative stroke occurs intra-operatively
Pathophysiology
Evidence suggests that perioperative stroke is primarily cardio-embolic in origin (29%) or due to large artery atherosclerosis (33%)
A smaller proportion is due to small-vessel or lacunar infarct (7%), but in a quarter of cases the aetiology is non-classifiable
Although hypoperfusion and hypotension may contribute it is rarely the sole pathophysiological factor
The precise pathogenesis of perioperative stroke is, however, not known
Theories include stroke arising due to a combination of
Pre-existing limited cerebrovascular reserve (e.g. due to previous stroke)
Some scoring systems (e.g. ACS-SRC, MICA score) may be highly discriminative for risk of perioperative stroke compared to others, but do not provide a quantitative stroke risk
Prophylactic carotid intervention
Prophylactic carotid artery stenting or endarterectomy is not recommended, although patients with carotid stenosis meeting guidelines for CEA should have vascular review
Anticoagulants
Concern often exists re: stopping anticoagulants in the perioperative period and consequent risk of stroke
Bridging may not be necessary for patients anticoagulated with warfarin for AF with low risk of peri-procedural stroke due to non-inferiority of not bridging and increased bleeding risk (BRIDGE Trial, 2015)
Bridging may be appropriate for patients anticoagulated with warfarin for AF at higher risk of peri-procedural stroke e.g. CHADS2-VASC score >5, recent TIA/CVA, mechanical heart valve, rheumatic heart disease, known cardiac thrombus or previous thrombus whilst anticoagulated
Antiplatelet agents
Evidence from POISE 2 suggested that continuation or initiation of aspirin did not reduce risk of perioperative stroke after non-cardiac, non-neurological surgery
Guidelines do not recommend initiating or continuing aspirin therapy for the prevention of cardiac events unless cardiac stents are present
The exception is in those patients undergoing procedures associated with high risk of bleeding or complications of bleeding (e.g. spinal surgery, some ophthalmological and neurosurgical procedures, Jehovah's witnesses)
Aspirin therapy probably doesn't reduce perioperative stroke risk in non-cardiac patients, although there still may be benefit to continuing aspirin balanced against an increased bleeding risk
For those established on long-term antiplatelet therapy for secondary prevention of stroke, aspirin monotherapy can be continued for most invasive noncardiac procedures except if bleeding risk is high
Beta-blockers
The POISE trial demonstrated initiation of metoprolol pre-operatively increased risk of perioperative stroke (OR 1.26 - 3.74) and mortality
More recent meta-analyses confirms that beta-blocker use did not reduce risk of stroke or improve other perioperative outcomes
Conversely, abrupt cessation of beta-blocker therapy is associated with an increased 30-day mortality (OR 3.93)
Therefore, one should:
Maintain chronic beta-blocker therapy in the perioperative period
Also continue chronic calcium channel blocker, alpha blocker and statin therapy
Not introduce beta-blockers in the perioperative period purely for the purposes of cerebrovascular risk reduction
Timing of surgery
Surgery <3 months post-stroke incurs a 68x higher risk of recurrent stroke regardless of confounding variables including low- or high-risk surgery
The risk of perioperative ischaemic stroke, MACE and mortality post-stroke begins to plateau at about 9 months, although may still trend towards higher risk even beyond 12 months
Therefore, delay elective surgery for at least 9 months after stroke
The timing between stroke and surgery must be considered on an individual basis, balancing risk of further stroke against risk of delaying surgery (e.g. cancer surgery)
For emergency surgery, a cohort study found urgent (<72hrs post-stroke) surgery had a lower risk of MACE vs. early (4 - 14 days post-stroke) surgery
Monitoring and access
Neurophysiological monitoring (e.g. EEG, evoked potentials, NIRS) has been shown to detect neurological insults (inc. stroke) during some surgeries e.g. CEA, cardiac surgery
However, there is no robust evidence that using such monitors prevents perioperative stroke in non-cardiac surgery
Anaesthetic technique
No robust evidence that choice of anaesthetic technique (general vs. regional) is associated with altered risk of perioperative stroke
Choice of technique should therefore be based on individual patient and surgical factors
Some suggestion that neuraxial anaesthesia may reduce stroke risk in arthroplasty patients, perhaps due to lower blood loss and fewer thromboembolic phenomena
If a neuraxial technique is chosen, the benefit of doing so must be weighed against the risk of discontinuing anti-platelet/-coagulant medication in the perioperative period
No robust evidence demonstrating superiority of either a volatile or TIVA technique, although:
Equipotent doses of sevoflurane lead to greater impairment of the cerebral autoregulatory response to PCO2 than propofol
Sevoflurane impairs autoregulation of cerebral blood flow at ≥1 MAC, whereas propofol only does so at doses of >200μg/kg/min
In the MYRIAD trial, there was no difference in stroke risk between sevoflurane and propofol maintenance in a cohort of patients undergoing cardiac surgery
Haemodynamics
Hypotension is known to affect other major end-organs, although the relationship between intra-operative arterial pressure and risk of perioperative stroke in non-cardiac patients is less clear
Current feeling is that intra-operative hypotension may contribute to risk of perioperative stroke but is often not the driving aetiological factor
For example, the hypotension-avoidance strategy in POISE-3 did not increase risk of MACE inc. stroke vs. a hypertension-avoidance strategy
Neverthless, BP should be maintained as close as practical to preoperative, awake blood pressure
For those at high risk BP should not be allowed to decrease below a MAP of 80mmHg, which is based on the point at which cumulative risk begins to accrue in non-stroke patients
For surgeries where there is a steep gradient between cerebral and central pressure (e.g. shoulder surgery in the beach chair position), consider increasing MAP especially in those at risk of stroke
Other physiological homeostasis
Patients who have experienced previous stroke suffer persistent derangements in cerebral blood flow autoregulation in response to PCO2
Ventilation should therefore be titrated to ensure normocapnoea, as deranged PCO2 can reduce cerebrovascular reserve and increase susceptibility to other insults (e.g. hypotension)
Evidence does not support the use of high inspired oxygen concentrations in stroke patients and therefore oxygenation to maintain saturations >94% is acceptable
Insufficient evidence for specific perioperative glycaemic targets in such patients beyond the standard 6-10mmol/L
Evidence from the SHINE trial demonstrated no difference in 90-day functional outcomes with more aggressive (4-7mmol/L) glycaemic control in stroke patients
Insufficient evidence to recommend a specific red blood cell transfusion target, however:
Receiving a blood transfusion increases risk of stroke post-operatively
Excessively high (high viscosity + reduced CBF) or low haematocrit (reduced DO2) increases risk in patients with acute ischaemic stroke
Maintain ongoing MAP target
Individualised, risk/benefit decision about timing of restarting anti-platelet or anti-coagulant drugs, or starting chemical VTE prophylaxis
Detecting and managing post-operative stroke
Detection and imaging for perioperative stroke is difficult
15% of patients will only present with mental status changes i.e. no appreciable sensorimotor deficit
When detection does occur it is often outside the thrombolysis window
The modified NIHSS score is a practical and reliable screening tool but is not validated for perioperative stroke and may be confounded by the effects of general anaesthesia
Serum biomarkers and neurophysiological investigations for ischaemic cerebral injury are equally not validate in general surgical populations
If stroke is suspected then management includes activation of local stroke protocols, leading to:
Non-contrast CT ± CTA ± CT perfusion
MRI ± diffusion-weighted MRI
Clinical evaluation including thorough neurological assessment and investigations for differential diagnoses
Urgent review by stroke/neurology service
Consideration for mechanical thrombectomy in eligible patients
Covert stroke is associated with increased risk of: