Endovascular Interventions in Acute Ischemic Stroke

Curr Atheroscler Rep (2016) 18:40 DOI 10.1007/s11883-016-0588-z

CLINICAL TRIALS AND THEIR INTERPRETATIONS (J. KIZER, SECTION EDITOR)

Endovascular Interventions in Acute Ischemic Stroke: Recent Evidence, Current Challenges, and Future Prospects Ramana Appireddy 1 & Charlotte Zerna 1 & Bijoy K Menon 2 & Mayank Goyal 3,4

# Springer Science+Business Media New York 2016

Abstract After many years of clinical research, endovascular thrombectomy has been conclusively proven to be an effective treatment in acute ischemic stroke. The evidence is compelling; however, it is generated in high volume stroke centers with stroke expertise. Challenges remain ahead on translating and implementing this evidence in routine clinical care across the world. The current evidence has opened up avenues for further research and innovation in this field. In this review, we will discuss the evolution of evidence on endovascular thrombectomy followed by a discussion of challenges and future prospects in this exciting field of stroke care.

Keywords Endovascular . Stroke . Thrombectomy

This article is part of the Topical Collection on Clinical Trials and Their Interpretations * Mayank Goyal [email protected] 1

Calgary Stroke Program, Department of Clinical Neurosciences, University of Calgary, Foothills Medical Centre, 1403 29th St NW, Calgary, AB, Canada

2

Calgary Stroke Program, Department of Clinical Neurosciences, Hotchkiss Brain Institute, Cumming School of Medicine, University of Calgary, Foothills Medical Centre, 1403 29th St NW, Calgary, AB, Canada

3

Department of Radiology, University of Calgary, Foothills Medical Centre, 1403 29st NW, Calgary, AB, Canada

4

Seaman Family MR Research Centre, Foothills Medical Centre, 1403-29th St NW, Calgary, AB T2N 2T9, Canada

Introduction The treatment for acute ischemic stroke changed radically in 2015 with the publication of multiple randomized controlled trials evaluating endovascular therapy [1–5, 6••]. This resulted from years of research into various prognostic indicators of stroke involving clinical presentation, imaging features, and workflow of acute care, coupled with technical innovations in both stroke imaging and mechanical thrombectomy. Now that we have the evidence showing safety and efficacy of endovascular therapy (EVT) in acute ischemic stroke (AIS) patients with proximal anterior circulation occlusion, it is time to apply this into routine clinical practice to improve the outcomes of this devastating disease. In this review, we discuss the evolution of the EVT for stroke over the years and the accumulated evidence from clinical trials, followed by the challenges faced in implementation of this therapy in routine clinical practice and future prospects in this exiting field.

Recent Evidence The era of randomized trials for endovascular treatment of stroke started in 1998 with the Prolyse in acute cerebral thromboembolism trial (PROACT) soon followed by PROACT II. Both trials looked at the safety and efficacy of intra-arterial (IA) recombinant pro-urokinase (r-proUK) for AIS within 6 h of onset in angiographically identified M1 and M2 middle cerebral artery (MCA) occlusions. The two studies showed higher recanalization rates and better 90-day outcome (modified Rankin scale 0–2) with r-proUK but at a cost of increased hemorrhages. Shortly after, the emergency management of stroke (EMS) bridging trial demonstrated feasibility of bridging IV-IA tissue-plasminogen activator (tPA) therapy within 3 h of stroke onset but only enrolled 35 patients; despite achieving better

40

Page 2 of 11

recanalization rates, it failed to show improvement in clinical outcomes measured by dichotomized mRs [7]. Mortality was higher in the combined IV tPA-IA treatment group. This combined approach was further explored in the Interventional Management of Stroke (IMS) I and II Trial [8, 9]. Both IMS I and II were prospective cohort studies that showed improved clinical outcomes at 3 months without significantly increasing symptomatic intracranial hemorrhage (sICH) rate and confirmed the safety and feasibility of bridging therapy. Development of the flexible and corkscrew-shaped Mechanical Embolus Removal in Cerebral Ischemia (MERCI) retriever device started the epoch of mechanical thrombectomy for acute ischemic stroke. The Food and Drug Administration (FDA) approved the MERCI device in 2004. The MERCI and MULTI-MERCI trials using new generation MERCI retriever were multicenter prospective single-arm studies designed to test the safety and efficacy of using the MERCI retriever in AIS within 8 h of symptom onset [10, 11]. The rates of recanalization with the MERCI retriever were higher (46– 48 %) compared to the placebo arm of PROACT II with better outcomes at 90 days and acceptable safety. Following approval of the MERCI retriever, the Penumbra aspiration system was introduced and was tested in the Penumbra Pivotal Stroke Trial (PPST), a multicenter prospective single-arm study for safety and efficacy. The PPST showed that the system is safe and achieved better recanalization rates than reported for the MERCI retriever [12]. In 2010, the FDA approved the use of Penumbra aspiration system for AIS due to large vessel occlusion within 8 h of symptom onset. Following the encouraging results from these single-arm studies, three randomized controlled trials (RCT)—IMS III, SYNTHESIS, MRRESCUE— ensued, and the results were published in 2010 [13–15]. The results of the above RCTs did not support the expected superiority of endovascular thrombectomy over IV tPA in improving clinical outcomes for AIS with proximal occlusions. Some of the factors that may have contributed to the results include the lack of appropriate imaging-based selection of patients likely to benefit; predominant use of older generation mechanical thrombectomy devices, like the MERCI and Penumbra systems, with inferior recanalization rates as compared with newer stent retrievers; and lower emphasis on workflow in the hospital and target time metrics resulting in longer treatment times [16–20]. Over the last few years, there have been improvements in the device technology resulting in safer and more effective mechanical thrombectomy devices like the stent retrievers. Stent retrievers were originally developed for stentassisted treatment of intracranial aneurysms and were found to be quite effective for thrombectomy in acute stroke [21, 22]. Following initial encouraging case reports and case series, the two retrievers Solitaire FR and TREVO were pitted against the MERCI retriever in the SWIFT and TREVO2 trials [23, 24]. Both of these trials showed that the stent retrievers had higher recanalization rates and better 90-day clinical outcome

Curr Atheroscler Rep (2016) 18:40

compared to MERCI device. Coupled with the better understanding of the limitations of first-generation RCTs, advances in stroke imaging, availability of the newer mechanical thrombectomy devices, and the emphasis on workflow paved way for the success of the recent RCTs [1–5]. The five positive RCTs showing superiority of endovascular therapy for AIS involving the anterior circulation are MR CLEAN, ESCAPE, EXTEND-IA, SWIFT-PRIME, and REVASCAT. The design characteristics of the five trials are presented in Table 1, followed by a summary of the results in Table 2. The outcome assessment was blinded in all the trials. The investigators of the Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands (MR CLEAN) were the first to report the benefit of the endovascular thrombectomy in October 2014. Following this, further enrollment in two other trials, BEndovascular treatment for Small Core and Anterior Circulation Proximal O c c l u s i o n w i t h E m p h a s i s o n Mi ni m i zi n g C T to Recanalization Times^ (ESCAPE) and BExtending the Time for Thrombolysis in Emergency Neurological Deficits – IntraArterial^ (EXTEND-IA) trial were stopped because of positive results at interim analysis. This was soon followed by termination in enrollment in BThe Solitaire with the Intention for Thrombectomy as Primary Endovascular Treatment for Acute Ischemic Stroke trial^ (SWIFT-PRIME) and BRandomized Trial of Revascularization with Solitaire FR Device versus Best Medical Therapy in the Treatment of Acute Stroke Due to Anterior Circulation Large Vessel Occlusion Presenting within Eight Hours of Symptom Onset^ (REVASCAT). Despite variation in the eligibility criteria and trial features, all the trials predominantly enrolled patients with moderate to severe stroke having a small-moderate ischemic core and a target proximal occlusion, and attempted mechanical thrombectomy with a stent retriever. Some of them (ESCAPE and SWIFTPRIME) placed more emphasis on workflow optimization and strict target time metrics to be achieved. Turning to the individual trials, MR CLEAN adopted a pragmatic protocol noted by its choice of vascular imaging for patient selection (CTA/MRA/DSA), allowing multimodal IA treatments (IA thrombolysis and/or mechanical treatment—clot retraction/aspiration/mechanical disruption/use of stent retrievers), enrolling proximal and distal vascular occlusions (distal ICA/MCA M1 and M2/anterior cerebral artery A1 or A2 segments) and inclusion of mild strokes (National Institute of Heath Stroke Score [NIHSS] 2 or more). The investigators targeted a stroke onset to groin puncture time under 6 h. Though there was no predefined cutoff for the amount of baseline ischemic core, the majority of the patients enrolled in the trial had a good ASPECTS score (median of 9 in both arms). The median NIHSS for the control and intervention arms was 18 and 17, respectively, with ∼90 % of the cases having proximal intracranial occlusion in both arms. The median time from tPA to randomization was over 180 min in both

5

ESCAPE

CT/MRI

NCCT/MRI

NCCT MRI (discouraged) NCCT/MRI

NCCT/MRI

Parenchymal imaging

None

None

CTA/MRA CT/MRI

CTA/MRA CTP/MRI (PWI)

Intervention arm

Usual medical care Intra-arterial treatment + usual care (with or without (with or without IV tPA) IV tPA) Intra-arterial—IA thrombolysis or mechanical thrombectomy (aspiration, retraction, stent retrieval, wire disruption) Both

Control arm

ICA/M1/M2

Intracranial ICA/MCA M1

Intracranial ICA/MCA M1

Usual medical care Mechanical thrombectomy using (with or without Solitaire stent retriever. IV tPA) IV tPA + usual Mechanical thrombectomy using medical care in Solitaire stent retriever. all IV tPA + usual Mechanical thrombectomy using medical care in Solitaire stent retriever. all

MCA M1 ± M2 with or Usual care (with or Mechanical thrombectomy using without ICA AND without IV tPA) Solitaire stent retriever. Moderate to good collaterals

Distal ICA MCA-M1/M2 ACA-A1/A2

Perfusion imaging Eligible occlusion protocol

CTA/MRA None

CTA

Either of these CTA MRA DSA

Vascular imaging

NCCT non contrast CT, CTP CT perfusion, MRI magnetic resonance imaging, PWI perfusion-weighted imaging, DSA digital subtraction angiography, CTA CT angiography, MRA MR angiography, ICA internal carotid artery, MCA middle cerebral artery, ACA anterior cerebral artery, M1 first segment of the MCA, M2 second segment of the MCA, A1 first segment of the ACA, A2 second segment of the ACA, IV tPA intravenous tissue plasminogen activator

EXTEND IA

SWIFT PRIME

>18 years 12 h

>18 years 6 h

Age limit Time window (from onset)

18– 8h 85 years >7 and < 30 18– 6h 85 years None (eligible for >18 years 6 h IV tPA)

2 or higher

MR CLEAN

REVASCAT >5

NIHSS range

Design features of recent randomized controlled trials of endovascular therapy [1–5]

Trial

Table 1

Curr Atheroscler Rep (2016) 18:40 Page 3 of 11 40

40


NIHSS National Institute of Heath Stroke Score, mRS modified Rankin scale, ICH intracerebral hemorrhage, GA general anesthesia, IV tPA intravenous tissue plasminogen activator.

* a statistically significant difference in mortality with endovascular therapy, ** determined by CT angiogram 2-8 hours after randomization

6.0 0 20 9 40 71 – 35 35

17

13

68

78

210

93

86

60 98

SWIFTPRIME EXTEND IA

98

17

17

100

100

224

57

88

–

36

9

12

0

3.0

9.1 2.7 1.9 3.6 1.9 19* 16 10 18 29 28 53 44 31** – 72 66 51 77 200 269 79 100 73 100 15 17 150 103

17 17

5.2 6.0 2 21 19 33 – 59 – 17

MR 237 CLEAN ESCAPE 165 REVASCAT 103

267

18

91

87

260

–

Endo Control Endo Control Endo Control Endo Control Endo only Endo only Endo Control Endo Control Endo Control Endo only

CT to groin puncture (minutes, median) Onset to groin puncture (minutes, median) NIHSS score IV tPA (%) (median) Patients (n) Trial

Table 2

Results of recent randomized controlled trials of endovascular therapy [1–5]

CT to reperfusion time (minutes)

Favorable reperfusion (%)

mRS 0–2 (%) Mortality (%) Symptomatic GA ICH (%) (%)

Page 4 of 11

the arms. Though this may be related to the workflow across multiple centers, other possibilities could be recanalization of distal occlusions following IV tPA in the initial screen for eligible patients, leading to delayed randomization or screen failure. Overall, despite broad inclusion criteria (clinical and imaging), the majority of the patients enrolled in the trial tended to have severe strokes with a small ischemic core and target proximal occlusion, a favorable profile. There was a consistent beneficial shift for all categories of the modified Rankin scale in favor of the intervention, but no significant reduction in mortality. The adjusted odds ratio was 1.67 (95 % confidence interval [CI] 1.21 to 2.30). This resulted in a number needed to treat (NNT) of 7. MR CLEAN had similar serious adverse events in both the endovascular treatment and standard care group [1]. The ESCAPE trial used multimodal imaging to choose patients with small ischemic core, defined by an ASPECTS ≥ 6, on the non contrast head CT and the need for demonstration of proximal intracranial anterior circulation occlusion along with moderate or good collateral status on single/multiphase CTA. Distinctively, a lot of emphasis was put on workflow optimization at each center to achieve a picture-to-puncture time of ≤60 min and picture-toreperfusion time of ≤90 min. The majority of subjects (≈75 %) were initially treated with IV rtPA, but randomization was performed without waiting for IV rtPA response. The ESCAPE trial also noted a statistically significant improvement in disability across the ordinal mRS scale (common odds ration [OR] 2.6; 95 % confidence interval [CI] 1.7 to 3.8; P < 0.001) favoring the intervention and resulting in a NNT of 3.1 [3]. The trial also showed a 3-month mortality reduction in the endovascular treatment group at 90 days (10.4 vs. 19.0 % in the control group; P = 0.04). Endovascular treatment in this trial was as safe as standard of care with a risk for sICH of about 3 % in both arms. Interestingly, there was no heterogeneity noted in pre-specified subgroup analyses by age, sex, site of occlusion, IV tPA administration, NCCT ASPECTS, or time from stroke onset to randomization, thus suggesting that endovascular treatment benefited across all subgroups. Post hoc analysis of the ESCAPE trial data showed that there was no additional benefit in patients with poor collaterals on imaging. EXTEND-IAwas a small phase 2b trial that enrolled 70 out of 100 planned subjects. The trial randomized eligible patients with AIS due to proximal intracranial occlusion with salvageable brain tissue based on CT perfusion imaging (CTP) and receiving treatment with IV tPA to undergo endovascular thrombectomy or continue standard of care. EXTEND-IA chose a combined tissue and clinical endpoint—reperfusion on 24-h perfusion imaging and early neurological improvement (NIHSS score 0–1 at day 3 or 8-point reduction in NIHSS score between baseline and day 3). Endovascular therapy resulted in increased reperfusion at 24 h (P < 0.001) [2].


The trial showed a statistically significant benefit for improved functional outcome in an ordinal analysis of the score on the modified Rankin scale at 90 days (common OR 2.0; 95 % CI 1.2 to 3.8; P = 0.006). The NNT for endovascular treatment within EXTEND-IA was 3. In the SWIFT-PRIME trial, all eligible patients with acute ischemic stroke and proximal intracranial occlusion and small ischemic core receiving IV tPA were randomized to receive endovascular thrombectomy using the Solitaire stent retriever or standard of care. The small core was defined as those showing target mismatch penumbral profile on CTP but was later modified to use small-moderate ischemic core for logistic reasons to enhance trial enrollment. Thrombectomy treatment within the SWIFT-PRIME trial was associated with a favorable shift in the distribution of global disability scores on the modified Rankin scale at 90 days (P < 0.001 by the Cochran– Mantel–Haenszel test, which was lower than the P value of 0.01 that was specified for early stopping; number needed to treat for one additional patient to have a less-disabled outcome was 2.6) [5]. There was an absolute increase of 25 percentage points in the proportion of patients in SWIFT-PRIME who were functionally independent at 90 days (P < 0.001). The rates of adverse events and sICH did not differ significantly between treatment groups. The most recent results came from the Spanish randomized trial of revascularization with solitaire FR Device versus best medical therapy in the treatment of acute stroke due to anterior circulation large vessel occlusion presenting within eight hours of symptom onset (REVASCAT) [4]. Enrollment was halted after inclusion of 206 out of 690 planned subjects due to loss of equipoise after reporting of positive results from the above named trials. Eligible patients had an upper age limit of 85 years, an NIHSS score of at least 6 points, prestroke mRs of 0 or 1, and endovascular treatment needed to be initiated within 8 h of stroke symptom onset. [17] Even with confirmed anterior circulation occlusion, patients were excluded with imaging evidence of a large ischemic core, as indicated by ASPECTS of less than 7 on NCCT or a score of less than 6 on diffusion-weighted magnetic resonance imaging (DWIMR). REVASCAT reported a common odds ratio of improvement in the distribution of the modified Rankin scale score of 1.7 (95 % [CI] 1.05 to 2.8) favoring thrombectomy [4]. Rates of death, serious adverse events, and sICH were similar in both groups.

Current Challenges Translating Evidence to Medical Practice Now that endovascular therapy is the new standard of care, there are important challenges ahead in translating that evidence into reproducible results in routine clinical practice.

Page 5 of 11 40

These challenges span a wide variety of issues, and in this review, we will discuss some of the biggest challenges faced by health systems and care providers. A single phrase summarizing these challenges is BHow to get the correct patient to the correct hospital in a timely manner?^ Correct Patient Identification of the correct patient involves multiple processes but starts in the field or first health care facility (first medical contact [FMC]) with the ability to identify patients likely to have a proximal occlusion by the paramedical personnel, as well as physicians other than neurologists. Not all these patients are ideal candidates for endovascular thrombectomy. Factors influencing suitability for the procedure include premorbid functional status, co-morbidities; ability to tolerate the stress of a major stroke; the capacity for recovery; the risk of complications; and the ability to participate in rehabilitation. There are many validated prehospital stroke screens available in the field though they differ in accuracy [25]. Further work is needed to improve these scales, as is the development of acute decision-making algorithms for use by EMS personnel including a combination of easy to use predictors like age, stroke severity, premorbid functional status, etc. Other advances in this field that can likely help in better patient triaging include the availability of tele-stroke facilities in the ambulance/field, use of smart phone technology, point of care blood biomarker facilities, etc.[26, 27]. Transporting a potential patient to the appropriate hospital with advanced imaging and endovascular capabilities in the shortest time possible is the next challenge. Timeliness of transportation is affected by regional health care referral protocols, policies, geography, traffic, and weather. Welldesigned protocols, along with appropriate triaging tools and EMS bypass policies, should be developed to minimize transport times. All of the trials have used advanced imaging criteria for patient selection. Although the criteria differ slightly among the various trials, all allowed identification of patients with two crucial parameters for successful endovascular therapy—favorable ischemic pattern and a proximal arterial occlusion. The favorable ischemic pattern translates to a baseline infarct with a small core and large area of penumbra, meaning salvageable brain that has not yet been irreversibly damaged. This is quite an important metric of brain physiology and is dependent on time from onset as well as on collateral circulation. Parenchymal and/or perfusion imaging using either CT or MRI provides the information to identify a small core. Of the two technologies, CT is widely preferred because of speed, safety, and widespread availability. NCCT of the head provides information on the extent of early ischemic changes, and this can be easily quantified using the ASPECTS scoring system [28]. ASPECTS is a 10-point scoring system for extent

40

Page 6 of 11

of early ischemic changes in the MCA territory [29]. Patients with low ASPECTS score are likely to have poor overall outcome irrespective of the treatment modality; they have an increased risk of hemorrhage after tPA, as well as lower chances of clinical benefit even with successful reperfusion [20, 30–33]. Vascular imaging is central to identification of eligible patients for endovascular therapy, which is reflected by its usage in all the recent trials [1–5]. The presence of a proximal occlusion in the internal carotid artery or middle cerebral artery M1 segment, which is likely to produce moderate to severe clinical deficit and less likely to be dissolved with tPA, is an optimal target for endovascular therapy [34, 35]. Another piece of crucial information obtained with CTA is the estimate of collateral circulation. It consists of an arterial network of leptomeningeal or pial collaterals and Willisian collaterals [36]. The collateral circulation is supporting the penumbral brain tissue and determines the extent of early ischemic changes. Patients with good collateral circulation tend to have lesser extent of ischemic changes early after symptom onset in stroke as compared with poor collateral circulation [37]. This translates to a small core and large penumbra. In addition, CTA provides information related to degree of carotid stenosis and vascular anatomy, which helps in planning the endovascular therapy. Perfusion imaging has a role in the defining the core and penumbra more precisely than with NCCT. Of the various CTP thresholds used for defining core and penumbra, Tmax and CBF have been shown to be more reliable [38, 39•]. The thresholds for the CTP are likely to be dependent on the imaging to reperfusion time rather than purely on onset to reperfusion time [39•]. Correct Hospital The next set of challenges is faced after the patient reaches the hospital with facilities for endovascular thrombectomy, and can be discussed broadly under challenges with regard to workflow in the emergency department, neuroimaging, endovascular suite, and man power. Optimization of workflow by making the best utilization of hospital resources to achieve faster recanalization is the ultimate goal. This requires close scrutiny of all processes, along with coordination and cooperation of the multiple players involved in the workflow [40]. Some of the practices that can improve workflow in the emergency department include adequate pre-notification of the entire team; a quick, focused history and examination; and avoidance of delays from non-urgent investigations like routine laboratory testing, chest X-ray, etc. Ways to optimize the neuroimaging workflow include pre-notification to the CT staff to keep the CT scanner ready, as well as having stroke imaging protocols in place, followed by rapid assessment of imaging to decide on endovascular eligibility within a few minutes of acquisition. Workflow


optimization should focus on parallel processing, especially once the patient is in the CT scanner, where interpretation of images, and decision on reperfusion strategies, including tPA eligibility and endovascular suitability, needs to occur concurrently with patient monitoring and stabilization, along with rapid transport to the endovascular suite if appropriate. This requires multiple members of the stroke team, consisting of the physician(s), nurse(s), porter(s), etc., to work independently in their predefined roles and in a synchronized fashion. Emphasis on teamwork and optimization of workflow with time-based metrics like picture-to-puncture (