Stand alone mechanical thrombectomy (with penumbra system) for acute ischemic stroke based on MR imaging: Single center experience :Manish Shrivastava1, Sourabh Lahoti2, Darshana Sanghvi3, Annu Aggarwal2, Shirish Hastak2, , Neurololy India

Home

TOPIC OF THE ISSUE: ORIGINAL ARTICLE

Year : 2012 | Volume : 60 | Issue : 4 | Page : 406--414

Stand alone mechanical thrombectomy (with penumbra system) for acute ischemic stroke based on MR imaging: Single center experience

Manish Shrivastava¹, Sourabh Lahoti², Darshana Sanghvi³, Annu Aggarwal², Shirish Hastak²,
¹ Department of Interventional Radiology, Kokilaben Dhirubhai Ambani Hospital, Four Bunglows, Andheri, Mumbai, India
² Department of Neurology, Kokilaben Dhirubhai Ambani Hospital, Four Bunglows, Andheri, Mumbai, India
³ Department of Neuro Radiology, Kokilaben Dhirubhai Ambani Hospital, Four Bunglows, Andheri, Mumbai, India

Correspondence Address:
Manish Shrivastava
Department of Interventional Radiology, Kokilaben Dhirubhai Ambani Hospital, Four Bunglows, Andheri (West), Mumbai
India

Abstract

Background: There is dismal rate of recanalization following intravenous thrombolysis of large vessel occlusive ischemic stroke. Trials on use of mechanical clot retrievers in acute ischemic stroke have used time from onset and clinical deficit at presentation as the main indications for intervention. Materials and Methods: Retrospective analysis of case records of acute stroke seen between May 2009 and October 2011 was done. It revealed 23 patients with acute ischemic stroke treated by mechanical thrombectomy using Penumbra system (PS). We used magnetic resonance (MR) imaging in correlation with clinical presentation to determine patients likely to benefit from recanalization and accordingly offered or at times deferred revascularization. A comparison of approach and outcomes was done with other relevant trials/reports. Results: Recanalization was achieved in all but one patient. Median modified Rankin Scale (mRS) score at 90 days was 2. Good clinical outcome (mRS ≤ 2) was achieved in 56.5% compared with 25% in Penumbra pivotal trial and 36% in multi Mechanical Embolus Removal in Cerebral Ischemia (multi MERCI) trial. All cause mortality was 13.04%. Symptomatic intracerebral hemorrhage (ICH) occurred in two patients (8.6%). Conclusion: Analysis of our results suggests that PS is safe and effective (91.3%) in recanalizing cerebral vessels without concomitant thrombolytics.

How to cite this article:
Shrivastava M, Lahoti S, Sanghvi D, Aggarwal A, Hastak S. Stand alone mechanical thrombectomy (with penumbra system) for acute ischemic stroke based on MR imaging: Single center experience.Neurol India 2012;60:406-414

How to cite this URL:
Shrivastava M, Lahoti S, Sanghvi D, Aggarwal A, Hastak S. Stand alone mechanical thrombectomy (with penumbra system) for acute ischemic stroke based on MR imaging: Single center experience. Neurol India [serial online] 2012 [cited 2023 Mar 29 ];60:406-414
Available from: https://www.neurologyindia.com/text.asp?2012/60/4/406/100704

Full Text

Introduction

Acute ischemic stroke (AIS) caused by large vessel occlusion is associated with high morbidity and mortality. Unlike the western world, where stroke rates are either declining or have plateaued, [1],[2],[3] in Asia the incidence of stroke is rising. Compounding this problem are dismal rates of recanalization following intravenous (i.v) thrombolysis and technical challenges of transarterial interventions. [4] Since good clinical outcomes are related to early and complete revascularization, [5] in recent years, greater research has been focused in this direction. A significant advancement has been the availability of mechanical clot retrievers such as the concentric device (merci-Concentric Medical, Mountain View, CA, USA), [6] Penumbra aspiration system (PS) (Penumbra Inc., Alameda, CA, USA) [7] and retrievable stent (Solitaire Irvine, CA, USA). [8] These devices preclude use of lytic agents, thereby reducing the incidence of reperfusion bleeds and have higher rates of recanalization. There have been multiple trials and publications on use of mechanical clot retrievers in AIS. [6],[7],[8],[9],[10] Mostly, these studies have used time from onset and clinical deficit at presentation as the main indications for intervention. We use magnetic resonance imaging (MRI) in correlation with clinical presentation to get a comprehensive understanding of ongoing hemodynamics. This helped us to determine patients who are likely to benefit from recanalization and accordingly offer or at times defer revascularization. We report our experience of 23 patients with AIS treated using the PS over a period of 30 months and compare our results with other similar studies.

Materials and Methods

Retrospective analysis case records of acute strokes presenting to the emergency room of the institute from May 2009 to October 2011 revealed 23 patients treated by mechanical thrombectomy. Information documented from all the 23 patients included: Demographics (age, sex, ethnicity); history (times of stroke onset, hospital arrival, initiation of procedure, and time to recanalize [Thrombolysis in Cerebral Ichemia (TICI) 2 or 3]); clinical examination (National Institute of Health Stroke Scale [NIHSS] score at presentation and discharge); MRI features, adverse effects (symptomatic intracerebral hemorrhage [ICH] was defined as increase in NIHSS by 4 points), and follow up. Algorithm for approach to patients of AIS followed at our hospital is depicted in [Figure 1]. {Figure 1}Mechanical thrombectomy was offered to patients with clinical diffusion mismatch, defined as NIHSS score >8 with small infarct on Diffusion weight imaging DWI, or diffusion perfusion mismatch irrespective of time of onset or NIHSS score at presentation.

MRI protocol for AIS comprised of diffusion (DW) and susceptibility weighted (SW) images, time of flight angiogram and in select patient's perfusion imaging using dynamic susceptibility contrast (DSC) technique. Imaging was performed on a 3 Tesla magnet (Verio, Siemens, Germany). For each patient, the acute infarct was described by location (arterial territory) and size (small, medium, or large). Infarcts involving less than one-third of the arterial territory were recorded as small, between one-third and two-third of the territory as medium and more than two-third of the territory as large. The SWI were reviewed for presence or absence of hemorrhage. MR-angiograms (MRA) were assessed for status of arteries (occlusion, stenosis, or normal caliber). Perfusion study was performed using dynamic susceptibility contrast imaging. Images were rapidly obtained during the 1 st pass circulation of I.V. gadolinium through the brains tissue capillary bed. A pressure injector was used. Postprocessing was performed using Siemen's 'Syngo' platform for each vessel of the brain a MR signal intensity-time curve was generated. Cerebral blood volume (CBV) was determined by integrating the area under the curve. Mean transit time (MTT) was time between arterial inflows and venous out flow. Finally, cerebral blood flow (CBF) was determined by replacing other values in the central volume equation (CBF = CBV/MTT). Color maps of the hemodynamic parameters were generalized. In those cases where perfusion imaging was performed, the relative cerebral blood flow (rCBF) and MTT maps were compared with the diffusion defect. Diffusion-perfusion mismatch was defined as a visually perceived difference of more than 30%.

As part of follow-up, each patient was clinically examined at 30 and 90 days from stroke onset. Functional outcome was measured by NIHSS and modified Rankin Scale (mRS). Good outcome was defined as mRS ≤2 at day 90 while drop of NIHSS by ≥4 was termed as improvement.

Procedure

All procedures were performed on single plane machine (ZeeArtis, Siemens, Germany) under general anesthesia. For access, triple axial system through femoral artery was used in patients of anterior circulation stroke. A 80- cm long sheath (arrowflex) was placed in the common carotid artery/proximal internal carotid artery (ICA) through which 6F Neuron guiding catheter (Penumbra Inc., Alameda, CA, USA) was navigated in the distal cervical/petrous ICA. An angiogram was obtained to confirm the site of occlusion and to demonstrate collateral circulation. Penumbra reperfusion catheter (Penumbra Inc., Alameda, CA, USA) was navigated over a micro guide wire (Transend 0.14; Boston Scientific, Natick, MA, USA) to the proximal end of the thrombus. Following this, separator wire (Penumbra Inc., Alameda, CA, USA) was passed through reperfusion catheter in to the thrombus to macerate it simultaneously suction (20 inches of Hg) was applied by suction pump attached to the proximal end of reperfusion catheter. This was continued till the vessel recanalized (TICI 2 or 3) or 30-40 min which ever was earlier. In the interim, control angiograms were performed every 5 min. For the posterior circulation, double axial system was used (no long sheath). A total of 3000 units of intravenous heparin were given at the beginning of each procedure. In case of underlying stenotic lesion, angioplasty/stenting was performed. [11] All patients who underwent stenting were given abxicimab/eptifibatide injections followed by loading doses of antiplatelets (aspirin 325mg and clopidogrel 600 mg). A computed tomography (CT) scan was obtained immediately postprocedure to check for hemorrhage.

Results

A total of 23 patients with AIS were treated by PS [Table 1]. Of the 23 cases, 20 involved anterior circulation (13 ICA and 7 middle cerebral artery [MCA]) while 3 involved posterior circulation (basilar artery, BA). There was total occlusion of vessel (TICI 0) in all but one, who had critical proximal BA stenosis with overlying thrombus. Nineteen patients presented within 8 h of symptom onset while two had unknown time of onset. Median time since onset to hospital arrival was 4 h. Median time from onset of symptoms to start of the procedure ('onset to groin time') was 6 h 20 min. Median time from MRI to groin was 1.5 h. This time was primarily used to explain the procedure (risk, benefit, expected outcome, expense) to patient relatives. Median time needed for recanalization, defined as time from application of suction to achieve TICI score >2 in the target vessel, was 20 min (range: 5-40 min to failure to recanalize in one patient). It took less than 10 min to achieve complete recanalization (TICI 3) in five patients and more than 30 min in two patients while for the remaining patients it took between 10 and 30 min. Median age of patients was 62 years (range: 37-72 years) and they presented with median NIHSS score of 15 (7-21). Of the 23 patients, 16 had clinical diffusion mismatch [Figure 2]. Five patients underwent perfusion imaging. All five patients who underwent perfusion imaging had diffusion perfusion mismatch. Out of these five patients, four had favorable clinical outcome, while one patient had a fatal perfusion bleed.{Figure 2}{Table 1}

Fibrinolytics (intravenous or intra arterial) were not used before or during the procedure in any patient. Successful recanalization (TICI 3) of occluded vessel was achieved in 21 patients. In one patient who had ICA occlusion, vessel could not be recanalized due to extreme tortuosity of ICA in its extracranial and intracranial segments. Stents were used in six patients, four in intracranial vessels, and two in extra cranial ICA. In a patient with MCA occlusion following recanalization by PS, M1 was seen to re-occlude repeatedly possibly secondary to an underlying stenosis. Stent was deployed in this patient to keep the vessel patent. In a patient with BA occlusion, there was thrombus overlying a critical stenosis. He was progressively worsening, hence the thrombus was aspirated and stent deployed. Two patients who had ICA occlusion at its origin, also had tandem occlusions distally (ICA 'T' in one patient and M1-MCA in the other). In both these patients, angioplasty was performed to gain distal access and then tandem occlusions were recanalized using PS, following which stents were deployed across the ICA origins. Similarly in a patient with BA occlusion, a stenotic lesion was noted at the site of occlusion following revascularization. This was stented, achieving a TICI-3 flow but extravazation was seen due to injury to the vessel wall.

Patient 3 in our series had ICA occlusion and died 45 days following successful recanalization due to supraventricular tachycardia. He was recovering well from stroke and NIHSS score had dropped by 7 but succumbed to comorbidities. Patient 4 had left ICA occlusion and presented within 90 min of stroke onset. MRI showed large diffusion defect in MCA territory. Considering that he was not very far from stroke onset and taking into account the possibility of reversal of restricted diffusion in certain instances early in the window period, treatment was offered. Though recanalization was achieved, there was no significant clinical improvement. He required decompressive craniotomy for malignant MCA territory infarct and remained bed bound (mRS-4), further emphasizing that relying on window period alone may be misleading. Patient 11 had left ICA occlusion with medium size infarct and significant diffusion/perfusion mismatch. [11] After recanalization, underlying critical stenosis was seen at ICA origin and treated by stenting. Since he was not on antiplatelets, 5 mg abciximab was given as bolus. Postprocedure CT scan revealed huge parenchymal (corresponding to the infarct on DWI), subarachnoid, and subdural hemorrhage leading to fatally high intracranial pressure. Patient 14 had BA occlusion with multiple foci of restricted diffusion in brainstem and cerebellum. He presented at 6 h from stroke onset and vessel was recanalized by 8 h. After recanalization, underlying stenosis was detected; this was stented. Following stenting a small extravazation was seen in the subarachnoid space, which was self limiting. However, he eventually succumbed to high intracranial pressure due to cerebellar and brainstem infarct.

In addition to these 23 patients who were offered revascularization, a subset of patients presented within the therapeutic window (8 h) had a large diffusion defect and negligible penumbra (no clinical diffusion mismatch or diffusion perfusion mismatch). These patients were not offered revascularization therapy as it was unlikely to be beneficial.

Outcome

Median mRS score at 90 days from stroke onset was 2. Good clinical outcome (mRS ≤2) was achieved in 13 (56.5%) patients. All cause mortality was 13.04% (three deaths). Symptomatic ICH was seen in two patients (8.6%). Of the patients who had good clinical outcome, six had ICA occlusion, four had MCA occlusion, and two had BA occlusion. There were threefatalities, one patient died 45 days after the procedure due to cardiac cause, secondpatient had ICA occlusion and died due to reperfusion bleed immediately after procedure leading to fatal increase in intracranial pressure. In the third case, there was similar increase in intracranial pressure but due to brainstem and cerebellar infarction in a patient with BA occlusion. All the three patients were more than 65 years.

Intra cerebral bleed occurred in six patients but affected clinical outcome in only two patients. Of these one was due to procedure related trauma to arterial wall and was asymptomatic. There were four reperfusion bleeds, out of which one was massive and fatal, the other three were asymptomatic. There was one bleed after deployment of stent in BA which was self limiting and stopped in a few seconds. However, the patient succumbed to progressive rise in intracranial pressure secondary to brainstem and cerebellar infarction.

Discussion

Treatment of large vessel occlusion is complex from decision making to execution. This is related to large thrombus burden resistant to lytic agents [12],[13],[14],[15] and is compounded by technical challenges of endovascular treatment. If adequate revascularization is not achieved, morbidity and mortality are high. [5] However, as novel therapeutic options such as mechanical clot retrievers emerge, our ability to therapeutically revascularize large vessel occlusion continue to improve. [4],[6],[7],[8]

A comparison of different thrombolytic trials is given in [Table 2]. All these trials had different approaches for thrombolysis and the table compares outcomes with individual approach. National Institute of Neurological Disorders and Stroke (NINDS) studied the benefit of r-tPA administered within 3 h of onset of ischemic stroke irrespective of site of occlusion or volume of pretreatment infarction and found no difference in clinical outcome compared with placebo at 24 h but better outcome with odds ratio of 1.7 at 3 months. mRS ≤2 was seen in 38-40% of the patients. Symptomatic ICH was observed in 6.4% patients. [16] The Prolyse in Acute Cerebral Thromboembolism (PROACT) and PROACT II trials studied efficacy of intraarterial pro-urokinase in MCA occlusion and reported recanalization rates of 57% and 66%, respectively with symptomatic ICH in 10.9%. Good clinical outcome was seen in 31% in PROACT and 40% in PROACT II. [17]{Table 2}

Multi Mechanical Embolus Removal in Cerebral Ischemia (multi MERCI) trial studied the safety and efficacy of a mechanical embolectomy device (Merci Retriever) in stroke patients and reported successful recanalization with adjunctive thrombolytic drugs in 69.5% of patients (without adjunctive therapy-57.3%) and procedure related complication rate of 5.5%. Symptomatic intracranial hemorrhage was observed in 9.8% of patients. Good neurological outcome was seen in 36% of patients; while all cause mortality was 34%. [6] Machi et al. reported their experience of 56 patients with the Solitaire FR device. They achieved successful recanalization in 89.2%, had procedure complications in 8.9%, symptomatic intracerebral bleed in 1.7%, and all cause mortality in 7.1% of patients. [8] In Penumbra pivotal trial 81.6% of patients were successfully revascularized. Procedural events were seen in 12.8% and 28% were found to have ICH of which 11.2% were symptomatic. All cause mortality was 32.8% and 25% had mRS ≤2 at 90 days. [9]

In this study good recanalization (TICI ≥2) was achieved in 91.3% of patients as compared with 82% in Penumbra pivotal trial, 89% with Solitaire FR system, 48% in multi MERCI trial, and 66% in PROACT II. Of the 23 patients in our series, 13 (56.52%) had ICA occlusion while in Penumbra pivotal 18%, in Solitaire FR study 25%, and in multi MERCI only 1% of patients had ICA occlusion. In this study good recanalization was achieved despite significantly larger proportion of ICA occlusions. Good clinical outcome, defined as mRS ≤2 at 90 days from stroke onset could be achieved in 56.52% of the patients in our series as against 25% in Penumbra pivotal trial and 36% in multi MERCI. Symptomatic intracranial hemorrhage occurred in two (8.7%) of our patients as compared with 11.2% in Penumbra pivotal and 9.8% in multi MERCI. All cause mortality at 90 days in this series was 13.04% (three patients).

Correct patient selection is central to achieving good clinical outcomes. Patients most likely to benefit from recanalization therapy are those with a small diffusion defect with a substantially larger diffusion perfusion mismatch on MR imaging [Figure 3]. [18] If MRI shows large diffusion defect in the vascular territory of the occluded vessel, patient is unlikely to benefit from revascularization even in the window period. Moreover, a patient presenting beyond the window period with a small diffusion defect may benefit by revascularization. [18],[19] This is based on the presumption that diffusion-perfusion imaging and not window period, reflects ongoing hemodynamics more accurately. Evidently, earlier a patient presents, higher is the likelihood of existing penumbra, but this may not always be true. Revascularization therapy without appropriate imaging may be superfluous with discouraging final clinical outcomes. The disadvantage of this approach is time lost in imaging; a CT scan with angiogram would take no more than 5 min whereas MRI (stroke protocol) would require on an average 20 min. In our experience, this loss of time, however, is compensated by more informed decision making and the resultant positive influence on final outcome.{Figure 3}

Unlike most other studies, no fibrinolytics were used in our patients in order to limit reperfusion hemorrhage. [20] All studies used time from onset and severity of clinical deficit as the criteria for intervention, [21],[22] while we emphasized on MRI features in deciding the mode of treatment. For example, one patient was found stupourous in the bathroom at 7.30 am and had NIHSS of 17. MR imaging showed small diffusion defect signifying clinical diffusion mismatch. [23] He was offered treatment despite unknown time of onset. NIHSS score dropped to one following mechanical thrombectomy. This case series provides insight into the concept of comprehensive understanding of AIS and its ongoing hemodynamics by MRI which probably allows more appropriate patient selection for revascularization thereby improving eventual clinical outcome.

Analysis of our results suggests that PS is safe and effective in recanalizing cerebral vessels without concomitant thrombolytics. Experience from this series of cases draws attention to importance of imaging and choice of revascularization therapy, particularly correlation of MRI findings with time and clinical presentation along with choosing revascularization therapy depending on site of vessel occlusion rather than time which may improve the final outcome. [24] There may be a subset of patients, who have penumbra even after 8 h of onset of symptoms. [25],[26] Identifying such patients and offering them revascularization is likely to yield better clinical outcome [Figure 4].{Figure 4}

References

1	Anand K, Chowdhury D, Singh KB, Pandav CS, Kapoor SK. Estimation of mortality and morbidity due to strokes in India. Neuroepidemiology 2001;20:208-11
2	Kaul S, Sunitha P, Suvarna A, Meena AK, Uma M, Reddy JM. Subtypes of Ischemic Stroke in Hyderabad (South India); Data from a hospital based Stroke registry. Neurol India 2002;50(Suppl):S8-S14.
3	Baneerje TK, Mukherjee CS, Sarkhel A. Stroke in urban population of Calcutta: An epidemiological study. Neuroepidemiology 2001;20:201-7.
4	Linfante I, Llinas RH, Selim M, Chaves C, Kumar S, Parker RA, et al. Clinical and vascular outcome in internal carotid artery versus middle cerebral artery occlusions after intravenous tissue plasminogen activator. Stroke 2002;33:2066-71.
5	Rha JH, Saver JL. The impact of recanalization on ischemic stroke outcome: A meta-analysis. Stroke 2007;38:967-73
6	Smith WS, Sung G, Saver J, Budzik R, Duckwiler G, Liebeskind DS, et al. Mechanical thrombectomy for acute ischemic stroke: Final results of the Multi MERCI trial. Stroke 2008;39:1205-12.
7	Bose A, Henkes H, Alfke K, Reith W, Mayer TE, Berlis A, et al. The Penumbra System: A mechanical device for the treatment of acute stroke due to thromboembolism. AJNR Am J Neuroradiol 2008;29:1409-13.
8	Machi P, Costalat V, Lobotesis K, Maldonado IL, Vendrell JF, Riquelme C, et al. Solitaire FR thrombectomy system: Immediate results in 56 consecutive acute ischemic stroke patients. J Neurointerv Surg 2012;4:62-6.
9	The Penumbra Pivotal Stroke Trial Investigators. The penumbra pivotal stroke trial: Safety and effectiveness of a new generation of mechanical devices for clot removal in intracranial large vessel occlusive disease. Stroke 2009;40:2761-8.
10	Higashida RT, Furlan AJ, Roberts H, Tomsick T, Connors B, Barr J, et al. Trial design and reporting standards for intra-arterial cerebral thrombolysis for acute ischemic stroke. Stroke 2003;34:e109-37.
11	Levy EI, Ecker RD, Horowitz MB, Gupta R, Hanel RA, Sauvageau E, et al. Stent-assisted intracranial recanalization for acute stroke: Early results. Neurosurgery 2006;58:458-63.
12	Rubiera M, Alvarez-Sabín J, Ribo M, Montaner J, Santamarina E, Arenillas JF, et al. Predictors of early arterial reocclusion after tissue plasminogen activator-induced recanalization in acute ischemic stroke. Stroke 2005;36:1452-6.
13	Zangerle A, Kiechl S, Spiegel M, Furtner M, Knoflach M, Werner P, et al. Recanalization after thrombolysis in stroke patients: Predictors and prognostic implications. Neurology 2007;68:39-44.
14	Sims JR, Rordorf G, Smith EE, Koroshetz WJ, Lev MH, Buonanno F, et al. Arterial occlusion revealed by CT angiography predicts NIH stroke score and acute outcomes after IV tPA treatment. AJNR Am J Neuroradiol 2005;26:246-51.
15	Rubiera M, Ribo M, Delgado-Mederos R, Santamarina E, Delgado P, Montaner J, et al. Tandem internal carotid artery/middle cerebral artery occlusion: An independent predictor of poor outcome after systemic thrombolysis. Stroke 2006;37:2301-5.
16	Tissue plasminogen activator for acute ischemic stroke: The National institute of neurological disorders and stroke rt-PA stroke study group. N Engl J Med 1995;333:1581-7.
17	Furlan A, Higashida R, Wechsler L, Gent M, Rowley H, Kase C, et al. Intra-arterial prourokinase for acute ischemic stroke: The PROACT II study: A randomized controlled trial. Prolyse in Acute Cerebral Thromboembolism. JAMA 1999;282:2003-11.
18	Fiebach JB, Schellinger PD. MR Mismatch Is useful for patient selection for thrombolysis: Yes. Stroke 2009;40:2906-7.
19	Davis SM, Rother J, Donnan GA. Magnetic resonance in the selection of stroke therapies. The Ischemic Penumbra. USA: Informa Healthcare; 2007. p. 307-16.
20	Singer OC, Berkefeld J, Lorenz MW, Fiehler J, Albers GW, Lansberg MG, et al. MR Stroke Study Group Investigators. Risk of symptomatic intracerebral hemorrhage in patients treated with intra-arterial thrombolysis. Cerebrovasc Dis 2009;27:368-74.
21	Zaidat OO, Suarez JI, Santillan C, Sunshine JL, Tarr RW, Paras VH, et al. Response to intra-arterial and combined intravenous and intra-arterial thrombolytic therapy in patients with distal internal carotid artery occlusion. Stroke 2002;33:1821-26.
22	Arnold M, Nedeltchev K, Mattle HP, Loher TJ, Stepper F, Schroth G, et al. Intra-arterial thrombolysis in 24 consecutive patients with internal carotid artery T occlusions. J Neurol Neurosurg Psychiatry 2003;74:739-42.
23	Prosser JF, Davalos A, Davis SM. Clinical /Imaging mismatch as an index of penumbra. The Ischemic Penumbra. USA: Informa Healthcare; 2007. p. 241-50.
24	Mathews MS, Sharma J, Snyder KV, Natarajan SK, Siddiqui AH, Hopkins LN, et al. Safety, effectiveness, and practicality of endovascular therapy within the first 3 hours of acute ischemic stroke onset. Neurosurgery 2009;65:860-5; discussion 865.
25	Kidwell CS, Warach S. Penumbra imaged with magnetic resonance imaging in Humans. The Ischemic Penumbra. USA: Informa Healthcare; 2007. p. 187-96.
26	Donnan GA, Baron JC, Davis SM, Sharp FR. The ischemic penumbra: overview, definition and criteria. The ischemic penumbra. USA: Informa Healthcare; 2007. p. 7-20.