Stroke Imaging Breaks the Speed Barrier

Endovascular treatment has emerged as the best thing since intravenous tissue plasminogen activator (IV t-PA) for acute ischemic stroke after five recent studies showed positive outcomes. Stent retrievers starred in the trials, with imaging playing a critical supporting role. But in clinical practice, which neuroimaging method best identifies likely beneficiaries is up for debate.

The fab five 

It all started with MR CLEAN. MR CLEAN (Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands) found that acute ischemic stroke patients who had proximal intracranial occlusions and received endovascular therapy along with standard care were more likely to achieve functional independence by three months than counterparts randomized to standard care alone (N Engl J Med 2015;237:11-20). The positive results, published in December 2014, prompted interim analyses of several other studies under way that compared endovascular treatment and usual care to usual care alone.

In the beginning: Almost 50% of patients treated with tPA alone in the National Institutes of Neurologic Disorders and Stroke (NINDS) trial had achieved essentially full recovery (N Engl J Med. 1995 Dec 14. 333(24):1581-7). However, subgroup analyses of the NINDS data showed that patients with severe strokes had only an 8% likelihood of achieving clinically significant improvement with tPA (Stroke. 1997 Nov. 28(11):2119-25). The poor outcome in these patients inspired the search for acute-stroke treatments that are more effective than tPA.

Three randomized clinical trials skidded to a halt in enrollment due to efficacy: ESCAPE (Endovascular Treatment for Small Core and Anterior Circulation Proximal Occlusion with Emphasis on Minimizing CT of Recanalization Times); SWIFT PRIME (Solitaire with the Intention for Thrombectomy as Primary Endovascular Treatment); and EXTEND IA (Extending the Time for Thrombolysis in Emergency Neurological Deficits-Intra-Arterial). The Spanish study REVASCAT (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) stopped recruitment for ethical reasons, based on the findings of these studies.

The trials shared several overarching features. They primarily relied on stent retrievers for endovascular therapy and all required imaging to identify patients with proximal large vessel occlusions who were likely to benefit or not from endovascular treatment. The back-to-back results released over the next few months in favor of endovascular therapy cheered neurologists and catalyzed medical societies to integrate the evidence into practice.

“The American Heart Association decided about January to redo the guidelines because they knew there would be new data,” recalls William J. Powers, MD, chair of the neurology department at the University of North Carolina in Chapel Hill and chair of the writing committee that updated stroke guidelines this year. “The original timeline had them to be published in September but because of the importance of this, it was pushed way up to be done in June.”

The update added several Class I Level of Evidence A recommendations, the top tier in the guideline hierarchy. It supported the use of stent retrievers in patients if they met a number of additional criteria and recommended using swift noninvasive vascular imaging if a physician was considering the use of endovascular therapy. Based on the imaging modalities chosen in the trials, it allowed CT, CT angiography, MR and MR angiography but cautioned that more research was needed to show the benefits of CT perfusion and other advanced imaging.

Learning from the past  

It all may have started with MR CLEAN, but with a strong assist from previous studies that fell short of their goal of moving the needle in stroke care. IMS III (Interventional Management of Stroke III), which compared endovascular therapy plus IV tPA to IV tPA alone, was stopped early in 2012 due to futility. The next year, MR RESCUE (Mechanical Retrieval and Recanalization of Stroke Clots Using Embolectomy), a four-arm study designed to identify patients by CT or MR with large areas of salvageable brain tissue and a small infarct core and show they would benefit from endovascular treatment, proved neutral; embolectomy failed to best standard care in patients with or without favorable features.

But the two studies provided three valuable lessons for researchers in the next round of studies. Time, imaging and the device itself are all important for a good outcome. “We went back to the data and asked, why did we fail?” says Reza Jahan, MD, an interventional neuroradiologist at the University of California, Los Angeles Stroke Center who was part of both MR RESCUE and SWIFT PRIME study teams. “We began to see a couple of things. The two most important issues we learned were, No. 1, that all patients were treated very late—hours after their stroke... The second thing was that the device we had available at the time was not as good as we thought it was.”

MR RESCUE required that patients with an acute ischemic stroke caused by a large-vessel occlusion undergo baseline multimodal CT or MR imaging of the brain to assess the presence of favorable penumbral pattern vs. a nonpenumbral pattern. They then were stratified to undergo either mechanical embolectomy or standard medical care. The trial specified that patients get randomized within eight hours of the onset of their stroke.

The study enrolled 118 patients between 2004 and 2011. “This was a time when it was very difficult for people to do sophisticated multimodal CT and MR imaging,” Jahan says. “It took them a long time.” The mean time to enrollment was 5.5 hours and the mean time to groin puncture was 6.2 hours. “You are six hours out of your stroke and you are just beginning treatment. That means that by the time you open up the clogged artery may be seven to eight hours after stroke onset. So one thing we learned is that you opened up these arteries but you are too late. Brain tissue destined to die probably already died.”

In addition, an analysis of catheter angiograms revealed that the embolectomy devices used in MR RESCUE opened arteries but not necessarily the entire vascular bed, Jahan says. Newer generation devices were in the approval process when the trial ended.

IMS III brought home the importance of imaging’s role in patient selection. When the trial launched in 2006, few hospitals used CT angiography to assess acute stroke patients. Participating centers instead included patients with a baseline National Institutes of Health Stroke Scale (NIHSS) score of 10 or higher, a measure of neural deficit that carried some likelihood of a major occlusion. As more centers adopted CT angiography, the protocol was amended to include identification of an occlusion by CT angiography in patients with NIHSS scores of 8 or 9.

“Many of the patients in IMS III did not have a CT angiogram,” says Mayank Goyal, MD, director of imaging and endovascular treatment at the Calgary Stroke Program in Canada and an investigator in IMS III, ESCAPE and SWIFT PRIME. “They did not have a documented proximal occlusion.”

Without that information, physicians could not know if the clot was amenable to treatment, according to Jahan. “What happened in IMS III was a rather significant percentage of patients had a clot that had dissolved or had gone into small branches” where the device couldn’t reach, he says. “When those patients went to device vs. best medical management, they did equally well and that led to neutralization of results.”

Make it snappy

CT imaging is now routine to rule out hemorrhage before initiating IV tPA. “Imaging is critical to the whole management of stroke,” emphasizes Michael D. Hill, MD, MSc, director of the stroke unit at Foothills Medical Centre in Calgary and a collaborator in ESCAPE and IMS III. “Imaging is our biomarker in stroke. It is the only way to tell whether someone is appropriate for therapy or not.”

But imaging takes time, expertise and availability. The insights gained from the older trials and applied to the newer ones underscored the need for fast, feasible and reliable imaging that facilitated decision-making. MR CLEAN, ESCAPE, SWIFT PRIME and EXTEND-IA used CT angiography to identify patients with proximal intracranial clots but the latter three added the presence of a small ischemic core and adequate collaterals or evidence of salvageable brain tissue at baseline for eligibility.

How they succeeded all the while racing against the clock varied. Physicians in ESCAPE were discouraged from using MR to identify occlusions because MR may take more time than CT. They also were encouraged to employ multiphase CT angiography to determine collateral status. The technique, which is championed by the Calgary group and was adopted by about 60 percent of sites in the trial, provides time-resolved angiograms of the brain’s vasculature. Their other allowed option, conventional CT angiography, captures only a moment in time. Multiphase CT angiography is said to be faster and less variable than CT perfusion, too.

“We in the past have been guilty of trying to get too much information from imaging, of trying to be too fancy,” Hill says. “One of the things we did in ESCAPE was to say ‘OK, what we need is simple, we need fast and we need accurate.’ We used existing technology but in a new way. This multiphase CTA provides us extra information but at a very fast pace. That allowed us to make strong decisions.”

They set a goal of 90 minutes or less from noncontrast CT to first reperfusion, monitored sites and provided feedback to help meet those goals. “The median time between start of IV tPA and successful endovascular treatment in IMS III was over 220 minutes,” Goyal notes. “In ESCAPE, we were able to bring it down to a median 84 minutes.”

Hill credits an approach that applied a general workflow strategy that they then customized to accommodate individual sites. They looked for opportunities to complete tasks in parallel rather than serially and find ways to shorten the gap between stroke onset and treatment, for instance, by bypassing the emergency room and taking the patient directly to a scanner.

And they keep pushing for better speed. “In our setup, we are trying to train people to meet up at the CT scanner and make a decision right at the CT scanner, and not wait for the CT images to come to you,” Goyal says.

Options but few answers

SWIFT PRIME followed a similar approach as ESCAPE except it initially required CT perfusion along with a specialized software to identify patients with salvageable brain tissue. They changed the study protocol midway because of concerns about delays using CT perfusion and to accommodate sites that lacked CT perfusion capabilities. Instead, sites could use the Alberta Stroke Program Early CT score (ASPECTS), a 10-point quantitative scoring system based on noncontrast CT imaging; ASPECTS was used in REVASCAT and ESCAPE as well.

Those sites in SWIFT PRIME that chose the ASPECTS route also performed CT angiography to identify a large vessel occlusion. For sites that preferred to use MR, they reduced the number of required images. In the end, more than 80 percent of participants underwent CT perfusion.

“We really streamlined the imaging to make sure people spent very little time getting necessary information and get that patient to the cath lab as fast as possible,” Jahan says. The time from qualifying image to groin puncture in the endovascular group was 57 minutes. 

By contrast, EXTEND IA called for CT angiography and standardized CT perfusion imaging for patient selection. The trial’s authors noted that 65 percent of patients in MR CLEAN underwent CT perfusion imaging, which they proposed may have been a factor it that study’s success (N Eng J Med online Feb. 11 2015). In EXTEND IA, the median time from initial imaging to groin puncture was 93 minutes.

One review led by Goyal with collaborators from ESCAPE and EXTEND IA gave ESCAPE the win for overall speed. According to their analysis, the time from stroke onset to first reperfusion was shortest for ESCAPE, at 241 minutes. EXTEND IA clocked in at 248 minutes vs. 252 minutes for SWIFT PRIME and 332 minutes for MR CLEAN (Stroke 2015;46:1453-1461).

Powers cautions that none of the trials were designed to prove the necessity of imaging or compare one imaging approach with another. Instead, they strived to identify which patients with acute ischemic stroke should undergo endovascular treatment. “Is this imaging necessary? Which imaging is best? None of them were set up to answer that question,” he says. “They were set up to answer the question: If you take people who look like this, do they do better if they are treated?”

“Time is brain” is the common phrase all the trial investigators cited, based on findings that every minute an acute ischemic stroke patient remains untreated, he or she loses 1.9 million neurons. They recommend hospitals use whichever appropriate modality works best for them, as long as it is fast, safe, accurate and informs decisions. Imaging is one part—an integral part—of a process that requires everyone to work efficiently and in concert to provide speedy care.

“This is not a disease that can be treated by one person alone,” Goyal says. “It requires teamwork and organization.”