Anaesthesia for endovascular thrombectomy

Learning objectives.

By reading this article, you should be able to:

• •Describe the indications for endovascular thrombectomy in the management of acute ischaemic stroke.
• •Discuss the advantages and disadvantages of available anaesthetic techniques during endovascular thrombectomy and the physiological targets.
• •Evaluate the evidence base and discuss existing controversies for the use of general anaesthesia during endovascular thrombectomy.
• •Recognise the common complications of endovascular thrombectomy.

Key points.

• •In patients with acute ischaemic stroke, i.v. thrombolysis is recommended within 4.5 h of symptom onset.
• •Endovascular thrombectomy for anterior circulation stroke caused by large vessel occlusion is effective and is the new standard of care.
• •Functional outcome is strongly time dependent. Thrombectomy must be initiated as quickly as possible and within 6 h.
• •It is reasonable to select either conscious sedation or general anaesthesia for thrombectomy.
• •Arterial BP should be maintained within the patient's physiological range: within 10% of baseline values.

Stroke is the third leading cause of death and the leading cause of disability in the developed world. Approximately 66% of stroke patients leave hospital with a disability resulting in a huge economic burden and costing the NHS more than £7 billion each year.¹ The pathophysiology of stroke is complex, with acute ischaemic stroke (AIS) accounting for about 85% of cases. In AIS, a critical reduction in blood flow results in cell death within the ischaemic core. The surrounding tissue, or penumbra, can remain salvageable for several hours and prompt restoration of blood flow is crucial.² Until recently, the only option for reperfusion was i.v. thrombolysis. However, approximately half of patients with an anterior circulation stroke will have a large vessel occlusion and only 10–40% of these patients go on to achieve a good clinical outcome with i.v. thrombolysis.³ For this patient group or those unsuitable for thrombolysis, the efficacy of mechanical clot retrieval is now firmly established. The introduction of this new service poses major challenges to most healthcare systems. This article discusses the anaesthetic management of these patients, the controversy surrounding the optimal anaesthetic technique, and the logistical problems of safe service provision.

Endovascular thrombectomy

Mechanical or endovascular thrombectomy (EVT) aims to provide rapid recanalisation of blocked large cerebral arteries. Multiple trials have reported strongly positive findings for anterior circulation EVT, with the proportion of those able to function independently at 90 days increasing by between 19 and 35%. Trials differed in design but most examined the effects of adding EVT to standard medical care with thrombolysis (Table 1).⁴ Several systematic reviews and meta-analyses of randomised controlled trials have combined results with similar positive findings. The Highly Effective Reperfusion Evaluated Multiple Endovascular Stroke Trials (HERMES) collaboration combined patient data from five trials of EVT published in 2015.⁵ Pooling data from 1287 patients (634 in the EVT group and 653 standard care) allowed more precise estimates of overall treatment effect (adjusted common odds ratio for reduced disability 2.49, 95% confidence interval 1.76–3.53; P<0.0001). The modified Rankin scale (mRS) is the most commonly used outcome measure of disability after stroke; using this score, the number needed to treat to reduce disability by at least one level on the mRS for one patient was 2.6. This is a highly significant outcome in a patient group who traditionally had a poor response to medical management with thrombolysis and a high risk of mortality and morbidity. Based on the evidence from these studies, the current indications for thrombectomy have been defined and inclusion and exclusion criteria developed (Table 2).

Table 1.

Trials showing effect of EVT compared with best medical care (IVT) only on good functional outcome (mRS<2) at 90 days.

Trial year	Type size	% With mRS≤2 IVT	% With mRS≤2 EVT plus IVT	Adjusted OR (95% CI) P value
MR CLEAN 2015	RCT, MC n=500	19.1	32.6	OR 2.16 (1.39–3.38)
SWIFT PRIME 2015	RCT, MC n=196	35	60	OR 1.70 (1.23–2.33) P<0.001
THRACE 2015	RCT, MC n=414	42.1	54.2	OR 1.55 (1.05–2.30) P=0.028
EXTEND IA 2015	RCT, MC n=70	40	71	OR 4.2 (1.4–12) P=0.01
ESCAPE 2015	RCT, MC n=316	29.3	53.0	OR 1.7 (1.3–2.2)
REVASCAT 2015	RCT, MC n=206	28.2	43.7	OR 2.1 (1.1–4.0)

CI, confidence interval; ESCAPE, Endovascular Treatment for Small Core and Proximal Occlusion Ischaemic Stroke; EXTEND IA, Extending the Time for Thrombolysis in Emergency Neurological Deficits–Intra-arterial; IVT, i.v. thrombolysis; MC, multicentre; MR CLEAN, Multicenter Randomized Clinical trial of Endovascular treatment for Acute ischemic stroke in the Netherlands; OR, odds ratio; RCT, randomised controlled trial; REVASCAT, Endovascular revascularization with Solitaire^™ Device versus Best medical Therapy in Anterior Circulation Stroke Within 8 hours; SWIFT PRIME, Solitaire^™ With the Intention For Thrombectomy as Primary Endovascular Treatment Trial; THRACE, Trial and Cost Effectiveness Evaluation of Intra-arterial Thrombectomy in Acute Ischaemic stroke.

Table 2.

Indications and inclusion criteria for endovascular thrombectomy.

Inclusion criteria for EVT

NIHSS and mRS

Patients of all ages with anterior circulation stroke Presentation <6 h of symptom onset Inadequate response or contraindications to i.v. thrombolysis Proximal vessel occlusion (internal carotid/M1/M2 segments of MCA) on imaging No new ischaemic changes present on CT/MRI brain Significant new disability (NIHSS>5) Previously independent in ADL (mRS<3)

Measures severity of stroke. Scores are awarded for level of consciousness, vision, sensation, movement, speech, language: 0: No stroke 5–15: Moderate stroke 16–20: Moderate/severe stroke 21–42: Severe stroke mRS Measures the degree of disability using a functional assessment scale: 0: No symptoms 1: No significant disability 2: Slight disability. Independent but unable to carry out all previous activities 3: Moderate disability. Can walk independently but requires help with ADL 4: Moderately severe disability. Not independent with ADL 5: Severe disability. Full time nursing care 6: Death

ADL, activities of daily living; MCA, middle cerebral artery.

Studies demonstrate a clear benefit of EVT in the 6 h after stroke onset with the benefit diminishing as the interval between onset and EVT increases. Beyond 6 h, the impact of intervention is less certain, although the presence of salvageable brain tissue on advanced imaging may indicate benefit. The Diffusion Weighted Imaging or Computerised Tomography Perfusion Assessment with Clinical Mismatch in The Triage Of Wake-Up and Late Presenting Strokes Undergoing Neurointervention With Trevo (DAWN) trial compared EVT plus standard care to standard care alone for patients last known to be well 6–24 h earlier and who had a mismatch between clinical deficit and infarct volume on imaging. The results showed better outcomes for disability at 90 days for EVT plus standard care over medical treatment alone.⁶ However, studies are still required to determine if all patients with delayed presentation require advanced imaging (e.g. perfusion imaging, CT or MRI).

These results have led to a global change in management of AIS. In March 2018, NHS England published a clinical commissioning policy recommending its routine use.⁷ The UK, Europe, and the USA have all published recommendations for the use of EVT in the management of patients with AIS.8, 9, 10 All neuroscience centres should now be planning for the implementation of a 24 h a day, 7 days a week thrombectomy service.

Thrombectomy pathway

Early stroke recognition is essential. Public education has increased awareness of symptoms. Prehospital triage systems enable prompt patient assessment, stabilisation, and rapid transfer. In most countries, regional stroke networks have been developed to ensure that patients are delivered to units best able to provide appropriate specialist care. The receiving hospital must be notified in advance to allow mobilisation of necessary resources before the patient arrives. On arrival in the emergency department (ED), it is important to establish the time the patient was last seen well to estimate time since stroke onset. Patients should have a neurological assessment using the National Institutes of Health Stroke Scale (NIHSS) used to measure the severity of a stroke (Table 2). Baseline investigations include glucose, clotting, urea and electrolytes, troponin (recommended for prognostication), and an ECG. However, these should not delay imaging or thrombolysis. All patients with suspected stroke should have access to imaging within 20 min of arrival in hospital, usually a non-contrast CT scan. The primary aim is to exclude intracranial haemorrhage, or other intracranial pathology such as tumour, which would preclude thrombolysis. Thrombolysis [alteplase (tissue plasminogen activator) 0.9 mg kg⁻¹] should be administered as soon as possible but within 4.5 h. Most EDs have thrombolysis guidelines with indications, contra-indications, and instructions for administration. Initial CT scans are usually non-specific in early AIS and many centres routinely perform CT angiography at the same time to demonstrate the site of vascular occlusion (Fig. 1). Those who might benefit from EVT should be transferred urgently to the nearest comprehensive stroke centre. This decision for transfer is made jointly by the stroke team in the referring hospital and the stroke consultant and neurointerventionalist in the receiving unit. Local protocols are recommended for interhospital transfer. There is recognition that this model (‘drip and ship’) may not be adequate in terms of providing access to EVT for the entire country in the long term. Alternative models have been suggested (e.g. ‘mothership’) where all patients are transferred directly to the specialist centre.

Fig 1.

CT angiogram demonstrating occlusion of the distal left internal carotid artery (carotid T occlusion: occlusion of the terminal carotid and initial segments of both the anterior cerebral and middle cerebral arteries).

Patients transferred from other hospitals are met in the ED by the local stroke team and transferred to the neuroradiology suite. On arrival, the anaesthetist, neuroradiologist, and neurologist assess the patient and decide on the need for additional imaging and the most appropriate type of anaesthesia. This decision is based on the patient's clinical status, the location of vascular occlusion, and any other perceived technical difficulties. The decision for thrombolysis and referral for thrombectomy should be made within 4.5 h of stroke onset, ideally with arterial puncture within 6 h, hence patient evaluation must be done as quickly as possible. However, patients are often elderly with multiple comorbidities. Their neurological status varies but they are frequently unable to give any medical history. Baseline investigations have usually been performed on admission and should accompany the patient. There is rarely time to request additional tests. Occasionally it may be deemed inappropriate to proceed in view of patient comorbidities or progression of the infarct volume. Interventional suites are often distant to the theatre complex and, in the setting of acutely unwell patients, appropriate training and protocols, including for anaesthesia, should be available for those involved in their care.

Performing EVT

Access for EVT usually is via the femoral artery. A guide catheter is advanced into the internal carotid artery and through this, a microcatheter and guidewire are navigated towards the occlusion. The guidewire is removed allowing the stent retriever or alternative device to be advanced within the proximity of the clot. Although stent retrievers are the recommended first line device, a variety of other devices have been developed. Stents can be placed and ‘opened up’ within the occlusion, restoring blood flow, or a retrieval device can be placed distal and withdrawn, trapping the clot. Alternatively, aspiration devices can be placed proximal to the clot. Variable anatomy and underlying pathology will alter the approach and the devices chosen. Occasionally, in the presence of aortofemoral disease, access is obtained via the radial or brachial arteries. Very rarely access is obtained by puncture of the internal carotid artery. There is no consensus on intraprocedural anticoagulation. Some interventionalists use a single dose of heparin to minimise the risk of catheter-related embolism. A relatively common challenge is the presence of a fixed stenosis (e.g. of the internal carotid artery) in addition to the intracranial occlusion. In these cases, a decision must be made as to which lesion to treat first. If a stent is to be left in situ, patients will need dual antiplatelet therapy for 3–6 months. However, the optimal time to start this is uncertain.

The technical goal of EVT is reperfusion (Fig. 2). The extent of revascularisation can be assessed using the modified Thrombolysis in Cerebral Infarction (mTICI) grading system. This uses a scale of 0–3, where 0 is no perfusion and 3 is full perfusion. The partial perfusion category of 2 is subdivided into 2a and 2b. Reperfusion is considered successful by achieving mTICI 2b/3.

Fig 2.

Cerebral angiography (A) before and (B) after thrombectomy. Blood flow is returned to right middle cerebral artery (MCA) territory. The retrieved thrombus is also shown.

GA or conscious sedation

The effect of anaesthetic technique on outcomes is controversial. Multiple observational studies have appeared to show worse outcome in patients receiving GA and this was interpreted by many to mean that GA was harmful.¹¹ However, this consensus was based on retrospective reviews containing few details regarding intraprocedural variables such as arterial pressure, respiratory variables, drugs used, or even what constituted anaesthesia. Many studies also suffered from selection bias, with the GA group having higher NIHSS or more severe strokes and, in some, a higher proportion of posterior circulation strokes. Based on the available evidence at the time, consensus guidelines from the UK, Europe and the USA recommended the use of local anaesthesia with conscious sedation (CS) where possible.

More recently, three prospective randomised controlled studies of CS vs GA for EVT have been published. The Sedation vs Intubation for Endovascular Stroke Treatment (SIESTA) trial failed to show an advantage of CS over GA. Interestingly, the proportion of patients with mRS<2 (a secondary outcome) at 90 days was significantly higher in the GA group.¹² The Anesthesia During Stroke (AnStroke) trial randomised patients to GA (propofol/remifentanil for induction with sevoflurane/remifentanil maintenance) or CS (remifentanil infusion) with strict haemodynamic control, normal ventilation and normoglycaemia. There were no differences in early neurological recovery, infarct volume, or anaesthetic or neurointerventional complications.¹³ Finally, the General Or Local Anaesthesia in IntraArterial Therapy (GOLIATH) study randomised patients to CS (fentanyl and propofol) or GA (propofol and remifentanil for maintenance of anaesthesia), again with strict haemodynamic control and normocapnia. There was no difference in the primary outcome measure—infarct volume size. There was a trend towards better clinical outcomes in the GA group with lower mRS scores.¹⁴

However, debate continues. Campbell and colleagues¹⁵ from the HERMES collaboration, recently addressed the role of GA in a meta-analysis, combining individual patient data from seven trials of EVT (although they did not include the three GA vs CS trials discussed above). Patients who had their procedure under GA had worse outcomes even after adjustment for baseline characteristics: for every 100 patients treated under GA vs non-GA, 18 patients would have worse functional outcome, including 10 who would not achieve independence. The authors conclude that GA should be avoided whenever possible. Unfortunately, despite the size and the inclusion of high-quality trials within this meta-analysis, there remains insufficient information about the type of anaesthesia used and intraprocedural variables to give a definitive answer regarding the effect of GA (see Table 3 for a comparison of studies).

Table 3.

Summary of data comparing general anaesthesia to conscious sedation for EVT in acute ischaemic stroke.

Study acronym; reference	Study aim; type; size	Patient population	Study intervention Comparator	End point (95% CI, P value,)	Limitations	Conclusions
SIESTA12	To assess if CS is superior to GA for early neurological improvement in EVT Study type: RCT, Single centre, Size: n=150	Inclusion criteria: NIHSS>10, ICA, M1, <9 h Exclusion criteria: aspiration risk, agitation, difficult airway	Intervention: GA (n=73) Comparator: CS (n=77)	1º end point: NIHSS improvement at 24 h GA 3.2 NIHSS points CS −3.6 NIHSS points mean difference 0.4 points (95% CI, −3.4 to 2.7; P=0.82)	Single centre, small size, early 1º end-point assessment	No advantage for CS over GA in EVT
AnStroke13	Null hypothesis: Anaesthesia technique has no impact on neurological outcome if hypotension is avoided Study type: single centre, open-label RCT Size: n=90	Inclusion criteria: ≥18 yr, anterior circulation occlusion, NIHSS≥10 (right sided), NIHSS≥14 (left sided) occlusion, within 8 h of onset Exclusion criteria: anaesthetic concerns (e.g. airway, agitation, mRS≥4, comorbidity)	Intervention: GA: propofol/ remifentanil, sevoflurane/remifentanil maintenance (n=45) Comparator: CS: remifentanil (n=45)	1º end point: 90 day mRS 0-2 GA 19/45 (42.2%) CS 18/45 (40%) (P=1.00)	Single centre, small size, superiority design not designed to establish non-inferiority	No significant difference between GA and CS in neurological outcome at 90 days
GOLIATH14	Does infarct size depend on type of anaesthesia remifentanil Study type: single centre, open-label RCT Size: n=128	Inclusion criteria: ≥18 yr, anterior circulation occlusion NIHSS≥10, initial infarct volume <70 (n= 65) within 6 h of onset Exclusion criteria: Premorbid state Anaesthetic concerns	Intervention: GA: propofol/Comparator: propofol/fentanyl (n=63)	1º end point: difference in median infarct growth, ml GA 8.2 (2.2–38.6) CS 19.4 (2.4–79) (P=0.10)	Single centre, small size, study may have been underpowered	No significant difference in infarct volume between GA and CS
HERMES group15	EVT vs standard care Study type: Meta-analysis of 7 RCTs Size: n=797	Patients with missing results/who did not have procedure were excluded	Patients with GA identified, baseline characteristics compared with non GA	1º end point: mRS at 90 days NIHSS/common odds ratio (cOR)	None of individual analysed to obtain to investigate GA vs CS, no detail of agents used 236/797 (30%) GA CS vs GA cOR 1.53 (95% CI 1.14–2.04)	Worse outcome with GA vs CS studies designed

ICA, internal carotid artery; M1, M1 segment of middle cerebral artery.

An association between GA and adverse outcome is entirely plausible. The impact of anaesthesia on the neurological outcome, especially at extremes of age, has been the subject of much recent debate. However, GA is not a single entity. There are numerous anaesthetic agents with very different properties both in terms of putative neuroprotective or neurotoxic properties, and effects on physiological functions. Hypotension, hypo- or hypercapnia, or excessive depth of anaesthesia may all compromise an already ischaemic brain. Appropriate assessment, monitoring, and manipulation of cardiovascular and respiratory variables are essential. There are also potential advantages. GA with tracheal intubation provides airway protection and allows manipulation of end-tidal CO₂. It also provides an immobile patient, reducing the potential for intraprocedural injury and possibly allowing a faster procedure time. Consequently, it is preferred by many neuroradiologists. However, an awake patient allows the intraprocedural monitoring of neurological status, is associated with fewer delays to procedure initiation, and may minimise haemodynamic changes associated with GA.

The latest recommendations from the American Heart Association/American Stroke Association have evaluated the evidence to date and state that it is reasonable to select an anaesthetic technique based on an individualised assessment of patient risk factors, technical performance of the procedure, and other clinical characteristics.

Anaesthesia

As discussed above, the type of anaesthesia chosen should be made on an individual basis. There is no evidence to support the choice of one agent over another. However, institutional protocols can assist in the timely and safe delivery of care. Our anaesthesia protocol, with examples of agents used, is given in Figure 3. As there is limited access to the patient during the procedure, long extensions are recommended for infusions and fluid lines. Standard monitoring includes ECG, pulse oximetry, end tidal CO₂ and arterial BP. Haemodynamic instability can occur under both CS and GA. If the procedure is performed under local anaesthesia or CS, non-invasive BP measurement every 3 min is adequate. Under GA, when haemodynamic instability is more frequent, invasive arterial monitoring is recommended. However, this should not delay the procedure and can be performed by accessing the femoral sheath used by the interventional neuroradiologists. Bladder distension may cause failure of CS. However, this should also not be allowed to delay femoral puncture. As patients will have received thrombolysis, caution must be taken to ensure atraumatic catheterisation.

Fig 3.

Example of protocol for anaesthetic management of patient undergoing thrombectomy. ODP, operating department practitioner.

Local anaesthesia

If the patient is conscious and cooperative, the procedure can proceed under local anaesthesia. Groin puncture, contrast injection, and clot retrieval can be painful and analgesia such as fentanyl 25 μg boluses may be required.

Conscious sedation

There is limited access to the patient and, when this is combined with potential communication difficulties and the possibility of deterioration in neurology during stroke evolution, caution with sedative agents is necessary. Short acting agents that provide analgesia and sedation are recommended. Target controlled infusions of remifentanil with or without propofol can be titrated to effect. Dexmedetomidine has also been used successfully. Benzodiazepines such as midazolam need to be used in combination with an analgesic such as fentanyl. However, their sedative effect is unpredictable and difficult to titrate in this population.

GA

GA is recommended for all patients with agitation, reduced Glasgow Coma Score (GCS), nausea and vomiting, and posterior circulation stroke. Intraprocedural conversion to GA is associated with a worse outcome. In our institution, we have a low threshold for choosing GA in patients with large dominant hemisphere strokes or complicated anatomy. I.V. anaesthetic agents, such as target controlled infusion of propofol and remifentanil, or inhalation agents such as sevoflurane, are the most popular. Tracheal intubation is recommended for patients receiving a GA and rapid-sequence induction (RSI) of anaesthesia where there is doubt over starvation status.

Physiological targets

Whatever the type of anaesthesia chosen, strict attention should be paid to maintaining physiological targets. Consensus guidelines recommend a target systolic arterial BP (SBP) of 140–180 mm Hg.¹⁶ However, the SIESTA trial maintained SBP 120–185 mm Hg and the Goliath trial SBP>140 mm Hg or MAP>70 mm Hg. Neither group found evidence of adverse neurological outcome using these targets. Whalin and colleagues¹⁷ reported a ≥10% MAP decrease from baseline to be a risk factor for poor outcome and suggested this as a suitable target threshold. It may not be necessary to target an SBP>140 mm Hg in those patients who start with a lower baseline pressure.

There is no evidence to recommend specific agents either to lower or increase BP. We find metaraminol infusions useful in our practice. Bolus doses of ephedrine and labetalol may also be required. SpO₂ should be maintained >94% (there is no role for hyperoxia) and $P{\text{a}}_{{\text{co}}_2}$ maintained between 4.5 and 5 kPa. Hyperglycaemia is common and predicts poor outcome in stroke patients. Hyperthermia is also detrimental to the injured brain and there is currently no evidence of benefit for cooling in EVT. Both normoglycaemia (7.8–10 mmol l⁻¹ or 140–180 mg dl⁻¹) and normothermia should be maintained in all patients.

Postoperative care

There are no data specific to the management of patients immediately after EVT. However, all patients need ongoing neurological and haemodynamic monitoring and in the longer term, the comprehensive package of care offered by a specialist multidisciplinary stroke team. Patients who receive CS are discharged directly to the hyperacute stroke unit (HASU) after the procedure. Those who receive GA without complication are initially discharged to PACU, with care delivered by nurses experienced looking after neurosurgical patients, after which they are discharged to the HASU. There is no evidence-based consensus to guide which patients should be admitted to critical care and this will partly depend on local service provision. In our institute, critical care is reserved for those patients who have complications, arrive intubated, or who have a low GCS before the procedure.

The optimal BP for best outcome in the postoperative period is unknown. Following reperfusion therapy, either thrombolysis or EVT, it should be maintained ≤180/105 mm Hg. There is no evidence for the best agents to achieve this but labetalol, nicardipine, and nitroprusside have been used successfully. Initiating or restarting antihypertensive therapy within the first 48–72 h is safe but not associated with better outcomes.

Pain is rarely an issue in the postoperative period, but patients may complain of headache or discomfort from the groin puncture site. Suitable analgesia should be prescribed and given as required. Simple analgesia such as paracetamol is safe and usually adequate. NSAIDs should be avoided. After thrombolysis, there are increased risks of bleeding at the groin puncture site, intracranial bleeding, and haemorrhagic transformation in the infarcted brain tissue. Opioids should be used with caution because of their sedative properties, which may interfere with monitoring the patient's neurological status after the procedure. Hypoventilation after opioids can lead to hypercapnia, which can also affect brain perfusion in the period when cerebral autoregulation is disordered after a stroke. Antiemetics should be prescribed and administered if required.

A comprehensive care pathway should be initiated immediately on return to the stroke unit to reduce complications. Aspirin is recommended in patients with AIS within 48 h of onset. However, after i.v. thrombolysis, this is usually delayed until 24 h later. Investigation of the likely causative mechanism of stroke is required. Cardiac monitoring is recommended for the first 24 h to screen for atrial fibrillation or other arrhythmias. Further rhythm evaluation or an echocardiogram to assess cardiac structure may be indicated. Patients under the age of 60 y may need a thrombophilia screen or autoantibody and anticardiolipin screen. Secondary prevention and rehabilitation measures should begin as soon as possible.

Complications

Thrombectomy is now the treatment of choice for patients presenting with AIS attributable to large vessel occlusion, but major complications may occur.¹⁸ A key factor in reducing complications is the experience of the whole multidisciplinary team. Thrombectomy should be performed in units with adequate/high volumes (>200 per annum) and with regular assessment and audit of results. When complications do occur, there must be immediate access to neurointensive care, neurosurgical support, or both. Intracranial haemorrhage has been reported in up to 10% of patients. This may be the result of device-related arterial injury from wires, microcatheters or thrombectomy devices. The newer generation devices have demonstrated superiority in terms of reduced complications compared with older devices. However, intracranial haemorrhage may also occur spontaneously, usually because of haemorrhagic transformation of the area of infarction. Haemorrhagic transformation may also occur in patients who have not received i.v. thrombolysis or EVT. In the event of intraprocedural haemorrhage, heparin, if given, should be reversed with protamine. There is little evidence for the manipulation of BP in this setting. Other catheter-related complications include vessel dissection, vasospasm, and thromboembolic events. Dissection occurs during catheter or guidewire manipulation and can involve any vessel from the puncture site to the intracranial vessels. It may be asymptomatic if localised but it may result in a neurological deficit. It also increases the risk of occlusive or thromboembolic complications. There is no consensus regarding treatment, but options include the use of anticoagulants or dual antiplatelet therapy. Occasionally, balloon angioplasty or stenting may be necessary. Vasospasm typically occurs secondary to vessel irritation and often resolves if the catheter is temporarily withdrawn. Calcium channel blockers such as nimodipine should be used with caution in this situation because of the risks of hypotension. Trauma can occur at the groin access site with injury or damage to adjacent structures, haematoma formation, or infection. Arterial closure devices are safe and effective at stopping bleeding at the access site. Extremely rarely, retroperitoneal haemorrhage, pseudoaneurysm, or limb ischaemia can occur. Adverse events also include reactions to contrast agents, including anaphylaxis, and anaesthetic complications such as respiratory depression, airway obstruction, aspiration, haemodynamic instability, and patient movement.¹⁹ Orolingual angioedema has been reported in up to 5% of patients after thrombolysis. This is usually self-limiting and rarely leads to airway compromise, but may make tracheal intubation challenging.

Some of these complications are life threatening whilst others can have serious sequelae for the patients. Complications can increase the length of stay in hospital, delay rehabilitation, and result in disability or functional impairment requiring various levels of care. It is important for anaesthetists looking after these patients to be aware of and be prepared to manage potential complications.

Conclusions

EVT has been shown to be one of the most powerful interventions in medicine, with no evidence of heterogeneity in treatment effect by age or baseline stroke severity. It has been accepted as a standard of care and is due to be rolled out nationally in the UK as soon as is practical. However, provision of a time-critical 24/7 service presents major challenges. It is estimated that about 10% of stroke patients will be eligible for thrombectomy. As there are approximately 800,000 strokes in the USA and 80,000 in England each year, this represents a significant workload. NHS England suggested that up to 8000 patients per year would be suitable for EVT, but this number may well increase as treatment windows are expanded and patient criteria change. Thrombectomy centres require immediate access to interventional neuroradiologists, anaesthetists (preferably with neuroscience experience), skilled anaesthetic support, appropriately trained nurses, and radiographers. Systems must be in place and monitored to expedite both assessment and intervention, and all patients must be entered on the Sentinel Stroke National Audit Programme database. The logistics of delivering this new service will present major challenges for all involved.

Declaration of interest

There is no declaration of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Biographies

Judith Dinsmore FRCA is a consultant anaesthetist with a specialist interest in neuroanaesthesia at St. George's Hospital where she has helped to develop their service for interventional neuroradiology and awake craniotomies. She is currently the President of NACCSGBI and has represented the society on several national working groups including the production of national consensus guidelines for safe and effective thrombectomy. She is an examiner for the Intercollegiate Specialty Board in Neurosurgery.

Mazen Elwishi FRCA is a consultant anaesthetist at St. George's Hospital with a specialist interest in neuroanaesthesia, who trained in Canada and the UK. He is lead for interventional neuroradiology and has helped to develop the service for thrombectomy.

Pungavi Kailainathan BBiomedSci FANZCA is a locum consultant anaesthetist at Charing Cross Hospital with specialist interests in neuroanaesthesia and pain.

Matrix codes: 1I02, 2A12, 3F00