Endovascular stroke therapy – a new era Michael D. Hill, Mayank Goyal, and Andrew M. Demchuk Calgary Stroke Program, Department of Clinical Neurosciences, Department of Radiology, Hotchkiss Brain Institute, University of Calgary, Calgary, AB, Canada Endovascular treatment of acute ischemic stroke is a therapy with a visible effect. With reperfusion, patients can get better on the angiogram suite table. We know that in the right patient, this therapy can be highly effective because our patients are hemiple- gic, and then they are not. They come in on a stretcher and they walk out of hospital back into their lives.Yet we have not been able to convincingly and consistently prove this effect in randomized trials. Why? Intra-arterial thrombolysis for ischemic stroke was reported in 1958 (1), but came of age with the PROACT-2 trial in 1999 (2). It was waylaid by manufacturing and business events that prevented progression to a second trial and licensing of pro-urokinase. The use of intra-arterial tissue plasminogen activator (tPA) was never an on-label treatment, but was well accepted in the stroke com- munity, particularly as part of the Interventional Management of Stroke (IMS) trials (3–5). The US regulatory and subsequent remuneration decisions to accept and pay for the MERCI retriever for ischemic stroke treat- ment had global influence (6). This decision simultaneously impeded randomized clinical trials enrollment and launched a wave of innovation in intracranial catheter development. Without this decision, more randomized clinical trials comparing endo- vascular vs. control would have been completed faster, but we may not have had such a rapid cycle of innovation of new devices culminating in the revolutionary stentrievers. The evolution of imaging has played a prominent role in isch- emic stroke management. With fast computed tomography (CT) helical acquisition, it is possible to rapidly obtain vascular imaging in stroke syndrome presentations. In Calgary, we have found that this is incredibly impactful in both major and minor stroke and hemorrhage, such that we do this routinely now on all stroke syndrome patients. These data are very rapidly obtainable (in minutes), similar across platforms and CT manufacturers, and reproducibly interpreted. We have further modified the technique to obtain a multiphase CT angiography (CTA), providing time- resolved images that assess collateral filling. While there have been improvements in CT perfusion, it remains vulnerable to patient motion and is less comparable and standardized across platforms. Magnetic resonance (MR) imaging provides greater spatial reso- lution and physiological information, but simply takes too much time; in acute stroke, the role of MR will be clinical situations where time is not limited. A meta-analysis including 395 patients showed a beneficial effect of intra-arterial thrombolysis, driven by PROACT-2 and MELT studies (7). Multiple cohort studies showed that recanali- zation with endovascular thrombectomy was associated with better outcome, even in later (<eight-hours from onset) time windows after stroke onset. New devices were developed, includ- ing suction thrombectomy, mechanical thrombectomy, and then the stentrievers (8–10). Two trials compared stentrievers to the MERCI retriever and showed reperfusion (TICI 2b/3) rates of up to 69%, again associated with better outcome (11,12); a multi- center prospective cohort study showed reperfusion (TICI 2b/3) rates of 79% (9). Then three randomized trials completed in late 2012 and pub- lished in early 2013 were all neutral on the primary outcome (13–15). The largest was IMS-III, but even within IMS-III there are trends in certain subgroups favoring endovascular therapy (13,16). Overall, these results introduced substantial uncertainty into the stroke community. While IMS-III, MRRESCUE, and SYNTHESIS were state of the art at the time they were designed, the evolution of imaging and catheters outpaced the trials (17) and many judged that the results were not reflective of current practice (18,19). Multiple trials were launched in the wake of these results. Most of these have focused on patient selection with CTA to identify intracranial occlusion and use of modern stentrievers. To varying degrees, the trials have focused on brain parenchymal imaging and speed of treatment. MRCLEAN is the first trial to report its results in the era of stentrievers. The trial enrolled patients within a six-hour from onset time window and proven intracranial occlusion. There were no parenchymal imaging selection criteria. Most (∼90%) patients were concurrently treated with intravenous (IV) tPA, and about half were transferred from a primary center after IV tPA to a tertiary center for randomization and half were treated primarily at the tertiary center. The trial was a study of patients with symp- tomatic confirmed intracranial artery occlusion and there was no Correspondence: Michael D. Hill – Departments of Clinical Neurosciences, Medicine, Radiology and Community Health Sciences, Hotchkiss Brain Institute, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada Mayank Goyal – Departments of Radiology, Clinical Neurosciences, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada Andrew M. Demchuk – Departments of Clinical Neurosciences, Radiology, Hotchkiss Brain Institute, Cumming School of Medicine, University of Calgary, Calgary, AB, Canada Conflict of interest: The authors are the Co-PIs of the ESCAPE trial. Funding: ESCAPE was funded by a consortium including grants from Covidien, Heart & Stroke Foundation Alberta, Alberta Innovates Health Solutions, Department of Clinical Neurosciences, Department of Radiol- ogy, University of Calgary, Hotchkiss Brain Institute, Calgary Stroke Program, Alberta Health Services. Statement of Authorship: M. D. H. wrote the first draft. All authors con- tributed to the final manuscript. DOI: 10.1111/ijs.12456 Leading opinion © 2015 World Stroke Organization 278 Vol 10, April 2015, 278–279