GA vs. CS for Endovascular Stroke Therapy

Rationale: Stroke is a leading cause of adult morbi-mortality worldwide. Intravenous (IV) recombinant tissue plasminogen activator (rtTPA) was the only therapeutic option for many years. After the publication of five randomized controlled trials, guidelines for stroke treatment were updated and released in 2015. According to these guidelines, in addition to IV rtTPA, endovascular therapy (EVT) for acute ischemic stroke (AIS), secondary to proximal intracranial arterial occlusion, has become established as an international standard of care (Avitsian \& Machado, 2016). Management of patients undergoing stroke is complex due to the association of acute neurological physiological changes and pre-existing co-morbidities. To facilitate efficient and safe delivery of endovascular treatment, patients require anesthetic intervention in the form of sedation or general anesthesia (GA). The choice of anesthetic technique is often guided by individual patient factors and personal clinician practice (Dhakal, Diaz-Gomez, \& Freeman, 2015). Two recent meta-analyses (Brinjikji et al., 2015) have found an association of better functional outcome related to endovascular procedures performed under sedation compared to general anesthesia. The interpretation of current data is challenging as most studies are retrospective and performed under direction of non-anesthetists with poor definition of anesthetic techniques. A prospective controlled randomized trial is required to demonstrate the influence of anesthesia techniques on functional outcome. Prior to commencing this large trial, the investigators will execute a feasibility study (GASTROKE-pilot), which is outlined here. Following confirmation of feasibility, the full prospective randomized controlled trial will take place using the same study methodology.

The objectives of the pilot study are: 1) To determine feasibility of the methods and procedures for a proposed prospective randomized controlled trial. Specifically, to determine if randomization can be reliably performed in a timely fashion to facilitate inclusion in the study. Note: EVT will not be delayed beyond institutional temporal limits to allow recruitment or randomization of potential participants. 2) To determine if anesthesia for EVT can be reliably delivered according to the standardized protocol. 3) To determine if post-treatment follow-up of functional outcome can be reliably performed according to a standardized methodology. If the pilot study demonstrates feasibility, the investigators will proceed with a much larger RCT.

Methodology: Randomization: Patients identified by the Acute Stroke team as candidates for EVT will be referred to the study anesthesia team for consideration as study participants. Only after anesthetic pre-procedural assessment is complete, as per usual standard of care, and clinical uncertainty exists as to the best anesthetic management will the patient be randomized. Sequence generation: Randomization will be performed using REDCap to allow central randomization from any hospital computer with internet access. The blocked randomization sequence list will be created in STATA. The randomization instrument will be created in REDCap separate to the data collection tool to ensure blinding allocation is maintained. All treating consultant anesthetists will have access to the REDCap randomization instrument. User rights will be set so that these clinicians do not have access to participant follow-up data. Details of each procedure (GA or sedation) are outlined below. The attending anesthetist will log-on to REDCap and access the Randomization tool. This process will take 2-3 minutes once trained in the randomization process.

Blinding: Both patient and treating physician will be aware of the treatment assignment. Post-procedural functional outcomes will be performed by a dedicated clinician blinded to the treatment assignment. Interventions: Physiological Parameters: Target Values (Talke et al., 2014) • Oxygenation: Maintain SpO2 \> 92% and PaO2 \> 60 mmHg. • Ventilation: Maintain normocapnia, PaCO2 35-45 mmHg under GA. Avoid respiratory-induced hypercarbia during sedation • Hemodynamics: Systolic blood pressure \> 140 mmHg and \< 180 mmHg. Diastolic blood pressure \< 105 mmHg. • Temperature: Maintain normothermia, T 35°C - 37°C. • Glucose control: Maintain blood glucose concentration 4.0 - 8.0 mmol/L. If blood glucose levels greater than 8 mmol/L (140 mg/dL) IV insulin infusion to be commenced. Blood glucose concentration to be repeated after 30 minutes. Hypoglycemia as defined by blood glucose \< 3 mmol/L (50 mg/dL) should be treated with IV 10-20% glucose. General Anesthesia Protocol: (Melinda J. Davis, Cynthia R. Campos-Herrera, \& David P. Archer, 2012; Powers et al., 2015; Talke et al., 2014) 1. Monitoring • American Society of Anesthesiologists (ASA) standard (non-invasive blood pressure (BP), electrocardiogram (ECG), oxygen saturation (SpO2), end-tidal carbon dioxide (ETCO2), temperature and neuromuscular monitoring (NMM)) • Invasive BP: direct measure of the arterial BP, with continuous pressure transduction and waveform display is the standard for blood pressure monitoring in AIS. It allows continuous BP assessment and provides reliable vascular access for frequent sampling. However, benefits of invasive arterial monitoring are null if the procedure itself delays reperfusion therapy (Saver, 2006). There is no consensus regarding timing of arterial access in EVT for AIS. This is our protocol: • 1 attempt with the patient awake • Maximum of 2 more attempts with the patient under anesthesia • Stop after a total of 3 attempts, or when the radiologist has access to the femoral artery. 2. Induction of anesthesia. 3. Neuromuscular paralysis. 4. Endotracheal intubation. 5. Positive pressure ventilation: ETCO2 range: 35-40 mmHg until arterial partial pressure of carbon dioxide (PaCO2) is obtained and the gap is measured. 6. Volatile anesthesia with sevoflurane or desflurane to a target age-compensated minimum alveolar concentration (MAC) \> 0.5 but = 1(Sivasankar et al., 2016) 7. Vasopressor support: If the systolic BP decreases more than 20% from pre-anesthetic value or is lower than 140 mmHg, the patient will receive vasopressor support as IV phenylephrine and/or ephedrine, titrated to effect. 8. Hypertension management: For systolic BP \> 180 mmHg, IV labetalol will be administered, titrated to effect. 9. Arterial blood gas sampling to be performed at the earliest convenience after induction of anesthesia. Clinical goals are outlined above. 10. Temperature monitoring wit esophageal temperature probe. 11. Antagonism of neuromuscular blockade at the end of the procedure. The train of four should be = 0.9 prior to extubation (Hassan et al., 2012; Murphy GS, 2015). 12. Transfer the patient to the post-anesthesia care unit (PACU). Sedation Protocol: 1. Monitoring • ASA standard (non-invasive BP, ECG, SpO2, T, respiratory rate (RR), and ETCO2) • Invasive BP: a direct measure of the arterial blood pressure, • Maximum 3 attempts. Stop after a total of 3 attempts, or when the radiologist has access to the femoral artery. 2. Administer oxygen by nasal prongs or facial mask to achieve a target Sp02 = 92%. 3. Commence sedation: Remifentanil: 0.01-0.06 micrograms/kilogram/minute, titrated to effect. (Janssen et al., 2016) 4. Monitor RR. Decrease sedative infusion rate if RR \< 6. 5. Vasopressor support: If the systolic BP decreases more than 20% from pre-anesthetic value or if is lower than 140 mmHg, the patient will receive vasopressor support as IV phenylephrine and/or ephedrine, titrated to effect. 6. Hypertension management: For systolic BP \> 180 mmHg, IV labetalol will be administered, titrated to effect. 7. Arterial blood gas sampling to be performed at the earliest convenience after induction of anesthesia. Clinical goals are outlined above. 8. Temperature monitoring with axillary temperature probe. 9. Transfer the patient to the PACU.

Conversion of sedation to GA: Patients randomized to receive EVT under sedation will be converted to GA in cases of emergency (vascular injury), reduced level of consciousness (GCS\<8), loss of airway reflexed, respiratory failure with rise in EtCO2 and agitation precluding safe conduct of EVT.

Post-Procedural Follow-up:This will be conducted by a clinical research assistant who is blinded to the study intervention. The specifics of the follow-up procedure are detailed below: Day1: The NIHSS clinical scoring tool will be performed on the stoke inpatient unit. This scale is a validated tool to objectively quantify disability following a stroke. The tool entails assessment of level of consciousness, motor function, sensation, coordination, speech and concentration. Currently, this tool is utilised at LHSC as part of the follow-up for AIS post-procedural days 1. Day 1-2: All patients will have a CT angiogram and MRI within 24-48 hours post-procedure as per usual standard of care. From these scans, the Thrombolysis in Cerebral Infarction (TICI) scale and final infarct volume will be calculated. Day 7: The NIHSS will be repeated to quantify ongoing disability. Day 90: As per the usual standard of care, all patient discharged from hospital will have a 3 month post-stroke outpatient follow-up appointment. The mRS and NIHSS will be repeated by the clinical research assistant at the neurological outpatient appointment. Should the appointment be delayed longer than 3 months post-stroke, a telephone call will be made by the clinical research assistant to determine the mRS. If the patient remains an inpatient at 90 days post-stroke, the electronic record will be accessed to gain the required information to calculate the mRS