Recent publication of the Interventional Management of Stroke (IMS III),1 Mechanical Retrieval and Recanalization of Stroke Clots Using Embolectomy (MR RESCUE),2 and SYNTHESIS-Expansion trial3 results in the New England Journal of Medicine prompted skepticism in the lay press about the clinical value of intra-arterial (IA) therapy for the treatment of acute ischemic stroke (AIS). As a professional society representing the subspecialists responsible for the surgical and catheter-based treatment of ischemic stroke, the Joint Cerebrovascular Section of the American Association of Neurological Surgeons (AANS) and Congress of Neurological Surgeons (CNS) endeavors to provide an evidence-based assessment of these recent trials, which may influence medical practice or impact the availability of these treatment options to patients.
The following article examines the results of the IMS III, MR RESCUE, and SYNTHESIS trials in a broader clinical context. First, to what extent did the conduct of these trials comport with modern thrombectomy practice? Second, what lessons emerge from the data in these trials to inform or support elements of that practice? Deconstruction of these 2 fundamental questions requires a characterization of modern endovascular practice along the following dimensions: (1) patient selection, (2) technical performance and adequacy of IA therapy, and (3) systems-based processes in the Comprehensive Stroke Center era. By these standards, the reviewed trials failed to reflect modern thrombectomy practice and render the conclusions of the lay press4 premature.
Taken together, these trials further provide compelling support for preserving the best elements of endovascular practice. Review of the data will demonstrate (1) the absence of competing therapeutic alternatives for patients eligible for nonintravenous recombinant tissue plasminogen activator (rtPA) , (2) the importance of preprocedural arterial imaging to identify and distinguish strokes due to large-vessel occlusions (LVOs), (3) the advantage of endovascular therapy in patients with confirmed LVO by preprocedural imaging, and (4) dramatic improvements in the technical performance and safety profile of modern endovascular devices.
Ultimately, our strongest professional commitment remains to the advancement of medical science through transparent scientific inquiry. We herald the results of IMS III, SYNTHESIS, MR RESCUE, and further prospective randomized trials as steps in that process, and bear the responsibility to our patients to ensure this progress does not become a basis for the denial of endovascular treatment access to stroke patients beyond the narrow cohorts studied in these trials.
CURRENT TREATMENT PRACTICES FOR ACUTE ISCHEMIC STROKE
Before reviewing the trial results, it is important to discuss current stroke treatment practices. At most centers, patients presenting with AIS are rapidly triaged, evaluated by the National Institute of Health Stroke Scale (NIHSS) to objectively quantify the severity of the stroke, and undergo brain imaging (usually computed tomographic [CT] angiography) to rule out intracranial hemorrhage (ICH), determine the extent and status of the ischemic territory, and evaluate for LVO. At this time, intravenous (IV) rtPA (alteplase, Activase/Actilyse) is the only US Food and Drug Administration (FDA)-approved treatment for AIS with class 1 evidence supporting its use5 and represents the standard of care for stroke patients who are candidates. All eligible patients presenting within 4.5 hours of symptom onset and lacking contraindications should be treated with IV rtPA at a standard dose of 0.9 mg/kg. Importantly, the FDA indication for tissue plasminogen activator (tPA) extends only to 3 hours. Broadly applied level 1 evidence supports its use in the 4.5-hour time window.
IA interventions, conversely, are specifically reserved for those patients with LVO. Although a subset of stroke patients overall, LVO patients experience disproportionate dependency, expense, and death. The natural history of AIS patients with LVO is dismal. The Prolyse in Acute Cerebral Thromboembolism II study,6 a rigorous, prospective study evaluating patients with confirmed large-vessel occlusions (therefore representing an intra-arterial intervention-appropriate population), revealed a 75% poor outcome (modified Rankin score 3-6) and 27% mortality in those not undergoing IA revascularization. Moreover, the FIRST trial7 offers further insight into the natural history of LVO. FIRST is a prospective, multicenter study of a stroke cohort eligible for but untreated by mechanical thrombectomy presenting within 8 hours of symptom onset from a LVO and a NIHSS of at least 10. At 90 days, 80% of patients had a poor outcome, 24 of 59 (40.7%) died, and 6 (10%) patients had an ICH. These data clearly demonstrate that LVO strokes are a devastating disease with a distinct, unforgiving natural history compared with cohorts including all patients with AIS.5 Second, IA treatment with thrombectomy devices is only useful with LVO target identification. Third, revascularization of LVOs with IV rtPA alone is uncommon, ranging from 10% in internal carotid occlusions to approximately 30% in middle cerebral artery occlusions.8
In general, candidates for IA therapies are those that present with documented LVO on angiographic imaging, have high NIHSS (8 or greater), and present within 8 hours of symptom onset. IA therapies are designed to rapidly open the occluded vessel with catheter-based technologies, the most recent of which have excellent revascularization rates that have been demonstrated to be far superior to older technologies in prospective randomized adjudicated trials.9,10
To adequately evaluate the benefit of IA therapies in addition to IV rtPA in a randomized controlled trial, it is essential that several key requirements be met. First, patients must have a documented LVO before randomization, or else the IA treatment will have no capability of contributing to benefit for that patient, and potentially exposes the patient to unnecessary risk. If a significant proportion of patients randomly assigned to IA treatment do not have LVOs, the potential treatment effect across the cohort will be diluted despite an intention-to-treat analysis. Second, all patients who are candidates must receive the established standard of care in both arms. Withholding or providing a lower dose of IV tPA treatment in the IA arm denies these patients the only agent with class I evidence of efficacy. Third, revascularization rates after thrombectomy in the IA arm should be clearly documented and meet current standards. In the section that follows, we will review the individual methodologies and results of the 3 recent published trials. Unfortunately, each of the 3 trials fails to meet these requirements, jeopardizing the generalizability of the results to modern AIS treatment.
BRIEF SUMMARY OF TRIAL RESULTS
The Interventional Management of Stroke (IMS III) Trial1 was a randomized, open-label multicenter study to compare a combined intravenous (IV) and intra-arterial (IA) treatment approach to restoring blood flow to the brain to the current standard FDA-approved treatment approach of giving IV rtPA alone. Both approaches required treatment initiation within 3 hours of stroke onset. A projected 900 subjects with moderate to severe ischemic stroke were to be enrolled at 50+ centers in the United States, Canada, Australia, and Europe. The study began in August 2006 and was halted prematurely on April 9, 2012 after futility analyses. Importantly, the Data Safety and Monitoring Board emphasized that safety concerns between the 2 arms were not the basis for trial suspension.
After randomization of 656 participants, the endovascular group and IV tPA cohorts experienced similar rates of functional independence (modified Rankin Score [mRS] of 0-2: 40.8% vs 38.7%) and mortality (19.1% and 21.6%, P = .52) at 3-month follow-up. Subgroup analysis based on NIHSS severity (NIHSS 8-19 compared with >20) suggested greater therapeutic benefit for endovascular therapy in the more severe stroke category, but did not achieve statistical significance. Rates of symptomatic ICH were additionally comparable between groups (6.2% and 5.9%, P = .83). Critically, there was an observed benefit to IA therapy in those patients with LVO documented on CT angiography (P = .01). Although further trials confirming these findings are necessary, the data herein strongly suggest a benefit to IA therapy in patients with confirmed LVO.
However, only 47% of enrolled patients had CT angiogram imaging at the time of enrollment. Nearly 20% of patients assigned to the IA treatment arm had no LVO or had thrombus that was inaccessible by catheter therapies. Furthermore, the significant majority of patients randomly assigned to the IA arm received less than the standard dose of IV rtPA before undergoing angiography, whereas all patients in the IV arm received standard-of-care doses. Finally, the first-generation thrombectomy devices in the IMS III trial are now known to be inferior to Stentriever devices in terms of recanalization rates, complication profiles, and clinical outcomes; level 1b evidence supports this conclusion. Only 59 patients were treated with modern devices such as the Penumbra aspiration system or stentrievers. Thrombolysis in Cerebral Infarction (TICI) 2b or 3 scores were obtained in only 40% of patients in IMS III, which is dramatically worse than the 68% TICI 2b or 3 reperfusion seen in modern high-quality prospective randomized trials.9 Importantly, IMS III further provided insight into the importance of the TICI 2b recanalization threshold in meaningfully impacting clinical outcomes; modern devices achieve this technical goal with a high degree of success. IMS III thereby guides interventionalists in terms of the technical outcome required for clinical benefit; this represents a seminal contribution on a broad scale.
The Mechanical Retrieval and Recanalization of Stroke Clots Using Embolectomy (MR RESCUE) Trial2 was a multicenter, randomized trial comparing standard medical care with embolectomy in patients presenting within 8 hours with a large-vessel anterior circulation stroke. All patients received preprocedural parenchymal imaging by CT or magnetic resonance imaging to assess penumbral profile; randomization protocols allocated an equivalent number of favorable and unfavorable penumbral profiles to each group. Patients treated with IV rtPA were only enrolled if arterial imaging demonstrated persistence of an LVO.
Of 127 enrolled patients, Rankin scores did not differ between embolectomy and standard medical care (3.9 vs 3.9, P = .99). Embolectomy was not superior to standard medical care in patients with either a favorable penumbral pattern (mean score, 3.9 vs 3.4; P = .23) or a nonpenumbral pattern (mean score, 4.0 vs 4.4, P = .32). Importantly, patients with revascularization did demonstrate mean improvements in 3-month mRS (3.2 [2.6-3.8] vs 4.1 [3.7-4.5], P = .04) and median absolute infarct growth (9.0 vs 72.5 mL, P < .001).
MR RESCUE only included first-generation endovascular thrombectomy technologies, resulting in only 16 of 64 patients (27%) achieving TICI 2b or 3 reperfusion.11 This revascularization rate is dramatically worse than recent SWIFT and TREVO trials using modern thrombectomy technology.
The Intra-Arterial vs Systemic Thrombolysis for Acute Ischemic Stroke (SYNTHESIS) Trial3 was a multicenter Italian trial that randomly assigned 362 patients presenting with an AIS to IV rtPA within 4.5 hours of onset vs IA therapy within 6 hours of onset. No preprocedural imaging was obtained to confirm LVO nor was a lower bound to NIHSS for study inclusion present. Nearly half of enrolled patients had a NIHSS of 10 or less, indicating that many enrolled patients would not meet current standards for deciding to pursue IA therapy. Indeed, 10% of patients in the SYNTHESIS trial randomly assigned to intervention did not harbor an LVO at angiography.12
Moreover, only 165 of the 181 patients randomly assigned to IA therapy received an endovascular procedure. Owing to endovascular treatment differences between the study centers and the United States, 109 patients received IA treatment with pharmacological agents and wire manipulation; only 56 patients received mechanical thrombectomy procedures in the IA arm. Furthermore, IV rtPA was withheld from patients randomly assigned to the IA arm, resulting in the IA arm not receiving appropriate class I therapy. Delays in treatment further biased SYNTHESIS in IA therapy; treatment was initiated approximately 1 hour later in the IA arm compared with the IV arm. These delays carry known clinical repercussions and were statistically significant between the 2 groups. SYNTHESIS trial design introduced additional breaks with conventional endovascular therapy. Patients randomly assigned to the IA arm underwent diagnostic cerebral angiography without antecedent angiographic imaging. If no LVO was identified, IA rtPA was infused into the suspected ischemic territory without a therapeutic target thrombus or occlusion.
By failing to report the revascularization (TICI) success obtained after endovascular intervention, the SYNTHESIS investigators unfortunately limit the application of these trial results to modern practice. Without a sense of technical success in the face of significant anomalies in endovascular practice, the clinical benefit of mechanical thrombectomy cannot be adjudicated. The SYNTHESIS Trial failed to demonstrate a 3-month clinical benefit by mRS, but fell well short of the modern endovascular standard. The absence of LVO confirmation, the withholding of class I indicated IV rtPA, and IA tPA delivery in patients without occlusion all notwithstanding, endovascular therapy had no increase in death or symptomatic ICH compared with IV tPA. The safety of IA therapy as a platform was therefore affirmed.
Patient Selection is Critical
AIS patients presenting with LVO are a unique subset of AIS patients that have a dismal natural history if untreated. It is this patient population for which IA intervention is designed. To adequately assess the benefits of IA therapy, enrolled patients must harbor LVO before enrollment. In IMS III, of the 656 enrolled subjects, only 47% obtained preprocedural CT angiography. Of the 423 subjects who obtained angiographic imaging, 80 did not have LVO and 89 ultimately received no IA intervention. Similarly, the SYNTHESIS Trial did not assess for LVO and did not report the number of patients with LVO. Furthermore, the majority of SYNTHESIS patients had low NIHSS, indicative that many enrolled patients likely did not harbor LVOs. Because documenting LVO is essential before randomization when evaluating the benefits of IA therapies, both IMS III and SYNTHESIS fail to provide an appropriate patient population to adequately study IA treatment.
Significant advancements in imaging have occurred since initiation of IMS III that have dramatically refined patient selection for IA therapies. Most centers now use CT or MR angiographic imaging that clearly delineates the status of the involved territory, penumbra, and collaterals. Alberta Stroke Program Early CT Scores, CT perfusion or MR diffusion imaging are routinely used before intervention to determine and exclude those patients with completed infarcts. However, patients in IMS III were excluded based on noncontrast CT imaging if they demonstrated a large area of hypodensity consistent with a completed infarct, but other early signs of infarction based on CT were not exclusionary. In fact, approximately 43.1% of IA therapy and 41% of control patients included in IMS III had Alberta Stroke Program Early CT Scores of 7 or less, indicating that many of these patients likely had early completed infarcts at the time of intervention. Inclusion of patients with completed infarcts in IMS III further diluted any potential benefit seen from IA therapies. Finally, the use of preprocedural CT angiography corresponded with a reduced time from IV tPA to start of thrombectomy by 20 minutes in IMS III (P < .01). Though this observed reduction was uncontrolled and may represent confounding covariates, this epiphenomenon strongly suggests that efficient preprocedural angiographic imaging may not increase time to recanalization. Moreover, the use of preprocedural CT angiography reduced time from groin puncture to start of thrombectomy by 10 minutes in IMS III (P < .01); transfer status or other unaccounted for confounders are unlikely to account for this observed benefit.
From a public health standpoint, it remains noteworthy that the patient population deemed eligible for treatment within the IMS III methodology only represents about 1% to 7% of all patients with AIS.13-15 Those patients presenting outside of the 4.5-hour window, those with “wake up” strokes, and those with contraindications to IV rtPA, such as recent surgery, have IA therapies as the only viable treatment option.
Revascularization Is Necessary
There is strong evidence suggesting that revascularization is associated with better functional outcomes and lower mortality.16 With advancements in device technologies, revascularization has markedly improved over the past 15 years.17 Recent technology has proved superior to older devices, such as the Merci retriever, in prospective randomized trials.9,10 In these studies, TIMI Grade 2 or 3 recanalization was seen in 80% of patients after Solitaire thrombectomy,10 and TICI 2 or 3 was obtained in 86% of patients after Trevo thrombectomy, with 68% having TICI 2b or higher recanalization.9
When evaluating the technical results of the IMS III, SYNTHESIS, and MR RESCUE trials, it becomes clear that all 3 trials have procedural results that fall well below current revascularization standards. In IMS III, a small minority of patients were treated with modern devices such as the Penumbra aspiration system or stentrievers, leading to TICI 2b or 3 scores in only 40% of patients in IMS III. Similarly, in MR RESCUE, the vast majority of devices used were first-generation endovascular thrombectomy technologies, resulting in only 27% achieving TICI 2b or 3 reperfusion. In fact, because of the poor technical performance of the first-generation devices in MR RESCUE, the IA arm had the same reperfusion as the medical arm. In the SYNTHESIS Trial, thrombectomy devices were only used in about one-third of patients in the IA arm, and revascularization was not reported in the results. The failure of all 3 of these trials to reproduce current methods of practice or modern standards of revascularization almost certainly resulted in the potential benefit of IA intervention being diluted. More importantly, these studies lack of successful revascularization strongly limits extrapolation of these trials to modern practice.
This principle is further supported by a subset of patients revascularized within the MR RESCUE trial and IMS III. In MR RESCUE, those patients with a favorable penumbral profile that achieved reperfusion in either arm enjoyed significantly reduced infarct progression over those that did not (9 mL vs 72.5 mL, P < .001). In IMS III, there was a strong concordance between adequacy of reperfusion and 3-month clinical outcome. These findings further suggest that adequate revascularization is critical to stroke volume and functional outcome.
In each of the 3 trials, investigators concluded that IA intervention was not superior to medical management for the treatment of AIS based upon the absence of benefit in functional outcomes between groups. As previously discussed, methodological failures to ensure adequate patient selection and poor technical success rates make generalization of these conclusions to modern endovascular AIS treatment obsolete.
However, endovascular intervention did in fact demonstrate benefit when used in the correct clinical context within these trials. In IMS III, the investigators had previously specified their intent to use a Van Elteren analysis to evaluate their prespecified subgroup analysis. To this end, they used this analysis to assert the absence of therapeutic benefit to endovascular therapy in the complete IMS III cohort, as well as in the subgroups stratified by NIHSS. However, when patients with a documented LVO based upon preprocedural imaging are analyzed separately, 3-month mRS was higher in those undergoing IA therapies compared with IV rtPA alone (P = .01). Therefore, based upon this analysis, IA therapies provide benefit when LVO was confirmed, supporting that IA therapies are beneficial when used in the appropriate context. This is an important consideration because this population (AIS with LVO) represents the vast majority of patients undergoing mechanical thrombectomy in clinical practice.
Furthermore, patients with higher stroke severities were more likely to receive benefit from IA intervention in IMS III. Twenty-four percent of patients with NIHSS of 20 or greater achieved good outcome in the IA arm compared with only 17% of patients in the medical management arm. Although this analysis was not statistically significant, it further suggests that, when the appropriate patient population is selected, the benefit of IA therapy becomes much more apparent.
Strong Evidence of Safety
There were no differences in symptomatic hemorrhage or mortality in any of the 3 trials between the IA arm and the control groups, demonstrating that IA therapy does not generate added risk to IV rtPA in AIS patients. Despite a reduced IV rtPA dose in IMS III and ∼20% of patients being subjected to angiography without an LVO, endovascular therapy had no increase in death or symptomatic ICH compared with IV rtPA alone. Similarly, withholding IV therapy and simultaneously subjecting all potential IV rtPA patients to IA therapy in SYNTHESIS, including those with minimal deficit and without confirmation of occlusion, resulted in equal efficacy to IV rtPA without significant safety concerns. These trials strongly demonstrate that endovascular intervention is safe, is not associated with increased risk, and performs at least similarly to IV rtPA administration.
IMS III, SYNTHESIS, and MR RESCUE represent the latest trials evaluating the benefit of IA therapy in addition to IV rtPA for AIS. Although each trial failed to demonstrate superiority of IA therapies over medical management, each has significant limitations that prevent its generalization to modern AIS treatment. Most importantly, IMS III and SYNTHESIS failed to adequately identify appropriate candidates for IA therapy by neglecting to identify LVOs before enrollment. Furthermore, each of the 3 trials did not meet current revascularization standards because of the use of predominantly first-generation thrombectomy devices; MR RESCUE, in particular, was hampered by revascularization rates that failed to even exceed the control group.
Despite the acknowledged limitations in revascularization, IMS III subset analysis resulted in a clinical benefit from IA therapy in patients with a confirmed LVO. Thus, these data support the benefit of mechanical thrombectomy in an appropriate target population. This benefit was evidenced despite long lag times to therapy and inferior technical performance of first-generation devices compared with modern endovascular standards. Furthermore, TICI 2b and 3 reperfusion results uniformly corresponded with strong clinical outcomes at 3 months. These data continue to support that IA intervention provides benefit to patients with LVO when revascularization is adequate.
The long patient accrual times of IMS III and MR RESCUE speak to the narrow subset of stroke patients who are ultimately IV rtPA and thus trial eligible. The success of IA therapy in the setting of confirmed LVO is particularly important to the large majority of stroke patients who are not IV rtPA eligible and therefore lack access to any competing therapy. The presented data from the Prolyse in Acute Cerebral Thromboembolism II study and FIRST speak to the terrible natural history of confirmed LVO without intervention.
Finally, the IMS III experience brings into sharp focus the value of preprocedural arterial imaging to identify the subset of patients with treatable LVO who have the potential to benefit from endovascular therapy and inform macrovascular access planning. Far from slowing the treatment process, these patients received care faster than those lacking imaging. Given preprocedural imaging is more commonly obtained at high-volume, experienced endovascular centers, these gains speak to the advantages of centralization and process efficiencies central to Comprehensive Stroke Center efforts.
Overall, despite limitations in design, endovascular technologies, and suboptimal study populations, these reviewed trials provide essential data in the continued refinement of endovascular therapy for large-vessel ischemic stroke. We strongly support further high-quality prospective investigations. In the interim, current data strongly support the reasonable offering of endovascular therapy for patients with LVO.
On behalf of the Joint Cerebrovascular Section of the American Association of Neurological Surgeons (AANS) and Congress of Neurological Surgeons (CNS). Endorsed by the American Association of Neurological Surgeons and Congress of Neurological Surgeons, The Society of Vascular and Interventional Neurology (SVIN), The Society of NeuroInterventional Surgery (SNIS), and The American Society of NeuroRadiology (ASNR).
Dr Mocco serves a consultant for Endeavor Endovascular and Lazarus Effect. He is an investor in Blockade Medical. Dr Khalessi serves on Clinical Events Committees and provides physician device training for Stryker Neurovascular and Covidien-ev3. These are minor financial conflicts by CNS guidelines. Dr Khalessi further serves on the AHA Writing Group for Extended Use of iv-tPA on behalf of the AANS and the National Steering Committee for Stroke Outcomes for the University Healthcare Consortium (UHC). These Committee assignments are potential related fiduciary but not financial conflicts. The other authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
1. Broderick JP, Palesch YY, Demchuk AM, et al.. Endovascular therapy after intravenous t-PA versus t-PA alone for stroke. N Engl J Med. 2013;368(10):893–903.
2. Kidwell CS, Jahan R, Gornbein J, et al.. A trial of imaging selection and endovascular treatment for ischemic stroke. N Engl J Med. 2013;368(10):914–923.
3. Ciccone A, Valvassori L, Nichelatti M, et al.. Endovascular treatment for acute ischemic stroke. N Engl J Med. 2013;368(10):904–913.
4. Brown D. Long-awaited stroke studies show hopeful new treatment no better than older one. The Washington Post. February 8, 2013.
5. Ciccone A. 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(24):1581–1587.
6. Furlan A, Higashida R, Wechsler L, 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(21):2003–2011.
7. Janardhan V, Giantasio RM, Chen SH, et al.. Abstract 194: Preliminary results from the FIRST trial: a natural history of acute stroke from large vessel occlusion. Stroke. 2013;44:A194.
8. Wolpert SM, Bruckmann H, Greenlee R, Wechsler L, Pessin MS, del Zoppo GJ. Neuroradiologic evaluation of patients with acute stroke treated with recombinant tissue plasminogen activator. The rt-PA Acute Stroke Study Group. AJNR Am J Neuroradiol. 1993;14(1):3–13.
9. Nogueira RG, Lutsep HL, Gupta R, et al.. Trevo versus Merci retrievers for thrombectomy revascularisation of large vessel occlusions in acute ischaemic stroke (TREVO 2): a randomised trial. Lancet. 2012;380(9849):1231–1240.
10. Saver JL, Jahan R, Levy EI, et al.. Solitaire flow restoration device versus the Merci Retriever in patients with acute ischaemic stroke (SWIFT): a randomised, parallel-group, non-inferiority trial. Lancet. 2012;380(9849):1241–1249.
11. Kidwell CS, Jahan R, Gornbein J, et al.. A trial of imaging selection and endovascular treatment for ischemic stroke. Supplemental Appendix. N Engl J Med. 2013;368(10):914–923.
12. Alfonso Ciccone, Carlo Poma Hosp, Mantua, Italy; Luca Valvassori, Boccardi, 2013 Edoardo Boccardi, Roberto Sterzi for SYNTHESIS Investigators “Synthesis Expansion: randomized trial on Primary endovascular treatment vs standard intravenous recombinant tissue plasminogen activator for acute ischemic stroke”, Niguarda Ca’ Granda Hosp, Milan, Italy; Presented at the International Stroke Conference 2013, February 6, 2013, Honolulu, Hawaii.
13. Adeoye O, Hornung R, Khatri P, Kleindorfer D. Recombinant tissue-type plasminogen activator use for ischemic stroke in the United States: a doubling of treatment rates over the course of 5 years. Stroke. 2011;42(7):1952–1955.
14. Morgenstern LB, Staub L, Chan W, et al.. Improving delivery of acute stroke therapy: the TLL Temple Foundation Stroke Project. Stroke. 2002;33(1):160–166.
15. de Los Ríos la Rosa F, Khoury J, Kissela BM, et al.. Eligibility for intravenous recombinant tissue-type plasminogen activator within a population: the effect of the European Cooperative Acute Stroke Study (ECASS) III Trial. Stroke. 2012;43(6):1591–1595.
16. Rha JH, Saver JL. The impact of recanalization on ischemic stroke outcome: a meta-analysis. Stroke. 2007;38(3):967–973.
17. Fargen KM, Meyers PM, Khatri P, Mocco J. Improvements in recanalization with modern stroke therapy: a review of prospective ischemic stroke trials during the last two decades [published online ahead of print]. J Neurointerv Surg. 2012.