18.1.1 Clinical Presentation

A 76-year-old female developed right hemiparesis and aphasia and was last known well at 9:30 p.m. Emergent medical services arrived at her home at 10:30 p.m. and she was transferred to an outside hospital. She had no significant ischemic stroke risk factors and was on a daily dose of 81 mg of aspirin. Intravenous tissue plasminogen activator (IV-tPA) was administered at 11:58 p.m. Her National Institutes of Health Stroke Scale (NIHSS) score was 15. A CT angiography (CTA) of the head showed a left M1 segment middle cerebral artery (MCA) occlusion (collateral score 3, clot burden score 6). She was transferred to our facility via life flight, arriving at the angiography suite at 1:54 a.m. Due to aphasia, agitation, anticipated patient discomfort, and increased procedural complexity from motion artefact, she was rapidly intubated and groin puncture was performed at 2:05a.m. At 2:23 a.m., she was completely revascularized consistent with thrombolysis in cerebral infarction (TICI) 3.

18.1.2 Imaging Workup and Investigations

A noncontrast computed tomography (NCCT) of the brain demonstrated early changes concerning for acute ischemia in the caudate nucleus and putamen. There was no evidence of hemorrhage. CTA demonstrated an occlusion of the mid M1 segment of the left MCA (Fig. 18.1a, b). There was excellent collateralization to the MCA territory (collateral score: 3, clot burden score: 6).

18.1.3 Diagnosis

Left M1 segment MCA occlusion.

18.1.4 Treatment

Due to persistent neurological deficit despite IV-tPA, presentation within a favorable time window (4 hours and 24 minutes from last known well), presence of large vessel occlusion on CTA, and excellent collateral supply to the occluded territory, the patient was deemed an excellent candidate for mechanical thrombectomy.

18.1.5 Postprocedure Care/Outcome

The patient was extubated at the end of the procedure. At the time, her right hemiparesis had partially resolved. Her subsequent course in hospital was uneventful. A brain MRI the next day showed infarction in the left basal ganglia without hemorrhagic transformation (Fig. 18.1e, f). She was diagnosed with atrial fibrillation and discharged to rehabilitation on Apixaban. At 90 days she had returned to her baseline functional status and scored 0 on the modified Rankin Scale (mRS) without objective neurological deficit.

18.2.1 Early Retrospective Studies

In 2006, Abou-Chebl et al published a case series reporting on the application of coronary interventional techniques, including stenting and angioplasty, for cerebral large vessel occlusion strokes. Most cases (37/40) were performed under conscious sedation. The remaining three patients received general anesthesia due to deafness, severe agitation, and patient movement in acute stroke. The authors advocated for conscious sedation, as it may allow for earlier detection of intraprocedural complications. ¹⁸ Van den Berg et al performed a retrospective analysis of 348 patients undergoing intra-arterial treatment including mechanical thrombectomy and intra-arterial thrombolysis for large vessel occlusion stroke of the anterior circulation. A total of 278 patients received conscious sedation, while only 70 underwent the procedure under general anesthesia. The authors concluded that patients undergoing conscious sedation were significantly more likely to have good outcomes than patients receiving general anesthesia. ¹⁹ However, after adjusting for prespecified prognostic factors, statistical significance was no longer present.

At this point, it has to be considered that retrospective studies are always at risk for selection bias, as severely ill patients with poor prognosis in the first place may be more likely to receive general anesthesia instead of conscious sedation.

18.2.2 Prospective Studies and Randomized Controlled Trials

The Sedation versus Intubation for Endovascular Stroke Treatment (SIESTA) trial, a recent single-center randomized controlled trial conducted in Germany, investigated the improvement of NIHSS 24 hours after intervention as primary outcome parameter. A total of 73 patients were randomized into the general anesthesia group, while a total of 77 were randomized into the conscious sedation group. Eleven (14.3%) patients randomized into the conscious sedation group required conversion to general anesthesia due to severe agitation (7 patients), apnea from sedation bolus, ² respiratory insufficiency, ¹ and direct puncture of ICA. ¹

The study did not find a statistically significant difference in NIHSS improvement between general anesthesia (–3.2 points) and conscious sedation (–3.6 points). Postthrombectomy complications were more frequent in the general anesthesia group for postoperative hypothermia (32.9 vs. 9.1%, p < 0.01), delayed extubation (49.3 vs. 6.5%, p < 0.01), and pneumonia (13.7 vs. 3.9%, p < 0.05). The unadjusted 3-month results showed a higher functional independence among the patients treated with mechanical thrombectomy under general anesthesia versus conscious sedation (37 vs. 18.2%, p < 0.01). However, it warrants mention that the SIESTA trial was primarily designed to evaluate short-term outcomes. ¹⁵

Comparable results were reported by Loewhagen et al in a randomized controlled trial named the AnStroke trial, ¹⁶ as well as in a non-randomized prospective study by Slezak et al ¹⁷ The primary outcome measure of both studies was the mRS 90 days after mechanical thrombectomy. While Slezak at al found that patients receiving general anesthesia had higher NIHSS scores at 24 hours postprocedure (14 vs. 9, p < 0.01), and were more prone to develop pneumonia (25.3 vs. 16.5%, p < 0.05), the functional outcomes at three months were equivalent. The AnStroke trial described no statistical difference in any of the measured outcome parameters but reported a higher incidence of pneumonia in the general anesthesia group. ¹⁶ A recently published post hoc analysis of the Solitaire with the Intention for Thrombectomy as Primary Endovascular Treatment (SWIFT-PRIME) trial focused on the outcomes of different anesthesia techniques used in mechanical thrombectomy. Primary outcomes included 90 days mRS, time to treatment initiation, and rates of successful recanalization. The study reported lower rates of functional independence after 90 days and higher incidence of pneumonia in the general anesthesia group (p = 0.05), while successful revascularization rates and the time to treatment initiation were comparable. ^{3 ,​ 20} The recent General or Local Anesthesia in Intra-arterial Therapy (GOLIATH) trial investigated the infarct growth on serial MRI 48 to 72 hours after mechanical thrombectomy as primary endpoint. Prior to data collection, authors hypothesized that patients under conscious sedation would have less infarct growth compared to patients undergoing general anesthesia before mechanical thrombectomy. After 72 hours, no statistically significant difference was found (p = 0.10). The general anesthesia group had a volume growth of 8.2 mL (2.2–38.6), while the conscious sedation group had a growth rate of 19.4 mL (2.4–79.0). The successful reperfusion rate, on the other hand, was significantly higher in the general anesthesia group (76.9 vs. 60.3%; p = 0.04).

Currently, there are several ongoing randomized controlled studies with the aim to compare the outcomes of conscious sedation and general anesthesia. The Sedation versus General Anesthesia for Endovascular Therapy in Acute Ischemic Stroke (SEGA) trial has a planned enrollment of 260 patients. The primary outcome parameter of this study is 90-day mRS. This study is expected to be completed by the end of 2020. ²¹ Another ongoing randomized controlled trial primarily investigating the 90-day mRS is the General Anesthesia versus Sedation during intra-arterial treatment for Stroke (GASS) trial. This study is enrolling 350 patients with an estimated completion date of December 2019. ²²

The Anesthesia Management in Endovascular Therapy for Ischemic Stroke (AMETIS) trial is primarily assessing neurological outcome at day 90 post–mechanical thrombectomy or serious complication within 7 days postprocedure after general anesthesia versus conscious sedation. This study is enrolling 270 participants and is planned to be completed in January 2020.

The majority of recently published prospective and randomized controlled trials showed comparable outcomes for both anesthesia techniques in terms of 24-hour NIHSS improvement and 90-day functional independence. However, a consistent complication of general anesthesia was the development of pneumonia. While this complication did not appear to affect the overall functional outcomes or mortality of patients, it should be considered when making a decision for a sedation technique.

18.2.3 Meta-analyses

The outcomes of these prospective studies strongly contradict the findings of previous retrospective studies and meta-analyses. Brinjikji et al performed two meta-analyses to evaluate the impact of anesthesia techniques on the outcomes of patients undergoing mechanical thrombectomy for large vessel occlusion strokes. One meta-analysis published in 2015 included nine retrospective studies released between 2010 and 2014 with a total of 1,956 patients, of which 814 received general anesthesia and 1,142 received conscious sedation for mechanical thrombectomy for large vessel occlusion stroke. The results of this meta-analysis indicated that general anesthesia may lead to worse outcomes after mechanical thrombectomy in terms of mortality, respiratory complications, and functional outcomes compared to conscious sedation, while no difference in procedural time was observed. ¹⁹ Similar findings were reported by McDonald et al in a meta-analysis showing that in-hospital mortality, pneumonia, hospital costs, and length of stay were also reduced in conscious sedation compared to general anesthesia. ¹³

In 2017, Brinjikji et al performed another meta-analysis including 22 articles with 3 randomized controlled trials and 19 observational studies published between 2010 and 2017 including 4,716 patients (1,819 underwent general anesthesia and 2,897 non-general anesthesia). After adjusting for the baseline NIHSS scores, the general anesthesia group was associated with lower odds of good clinical outcome (odds ratio [OR], 0.59; 95% confidence interval [CI], 0.29–0.94). Interestingly, when looking at studies that solely used stent retriever or aspiration techniques instead of stenting and balloon dilation techniques, there was no significant difference in good clinical outcome (OR, 0.84; 95% CI, 0.67–1.06). ²³

18.2.4 Decision Making

Since there are currently no distinct guidelines recommending a specific anesthetic type available, a variety of individual factors should be considered before deciding whether mechanical thrombectomy for acute ischemic stroke of the anterior circulation should be performed under conscious sedation or general anesthesia. Both techniques have their advantages and limitations.

18.2.5 Airway Protection

Major indications for general anesthesia include vomiting, decreased consciousness, agitation, or bulbar dysfunction, such as impaired gag reflex, compromising the airway. ^{15 ,​ 24} Although a frequently mentioned concern about general anesthesia is the increased risk of pneumonia, ²⁵ which was apparent in several studies, it did not affect the functional outcome or the overall mortality of patients receiving general anesthesia for mechanical thrombectomy. ^{15 ,​ 16 ,​ 17} In some cases, a conversion from conscious sedation to general anesthesia may be necessary due to increased agitation or intraprocedural complications. ^{16 ,​ 17} Aside from delays of procedural time, these conversions always involve the known risks associated with emergent intubation such as airway injury or aspiration. ^{26 ,​ 27}

18.2.6 Compliance and Patient Movement

The presence of sensory or global aphasia also warrants special precaution, as compliance of these patients will be compromised due to their impaired ability to understand and follow instructions provided to them during the procedure by the interventionalist. ¹¹ Another argument in favor of general anesthesia is substantial reduction of movement in agitated or noncompliant patients. This can potentially lead to motion artifacts ²⁸ and subsequently to compromised safety, and prolonged interventional procedures. It is hypothesized that head motion during the interventional procedure may increase the risk of vascular injury by vessel perforation, leading to intracranial hemorrhages or dissections. ^{29 ,​ 30} To overcome these limitations, Janssen et al advocated for head immobilization using a standard cervical collar during mechanical thrombectomy under conscious sedation. ¹¹

18.2.7 Time from Symptom Onset to Effective Treatment

Another aspect to consider is the timing from stroke onset until recanalization. It is known that timing is crucial in acute ischemic stroke, as ischemic time is proportional to the loss of functional brain tissue. ³¹ While this appears to vary among different institutions, some authors report a significant delay if general anesthesia is used. ^{11 ,​ 32} On the contrary, several large prospective randomized controlled trials did not show a difference in timing between general anesthesia and conscious sedation. ^{15 ,​ 16} In a post hoc analysis of the SIESTA trial, it was found that the time from groin puncture to final angiographic result was reduced in patients under general anesthesia compared to conscious sedation, ¹⁴ possibly due to reduced motion artifact. However, these times may be dependent on the primarily used method at the individual institution. If the anesthesia team and the interventionalists are more experienced in performing endovascular interventions under general anesthesia than under conscious sedation or vice versa, this may influence the time required for these procedures.