A half million and $30 billion. That is the estimated number of strokes each year in the United States and the cost these events impose on the health care system. [1] Approximately one third of these cerebrovascular accidents are fatal, rendering stroke the third leading cause of death. [2] An equal percentage necessitate long-term nursing care because of such effects as incontinence, loss of memory, or motor control deficiencies. [3]

Until recently, supportive care was the only option for stroke patients. However, in the 80% of patients in whom cerebral angiography detects arterial occlusions, [4] thrombolytic recanalization before the infarction is complete may reduce the severity of injury, helping more patients return home with little or no residual deficit. Partly as a result of the availability of thrombolytic agents, the mortality rate of stroke has been declining. [6, 7] To obtain these results, however, considerable expertise and prompt action are necessary. The answer: comprehensive stroke care centers.

The Basics of Stroke Centers

Table 1. Imaging capabilities required by a stroke center, as specified by the National Stroke Association.

The specialized care provided by a stroke center requires an extensive commitment of staff, equipment, and money. Sophisticated imaging equipment is considered necessary (Table 1). The emergency department must have Joint Commission on Accreditation of Healthcare Organizations (JCAHO)-compliant equipment for resuscitation and life support. The intensive care unit must be similarly equipped. Cardiac evaluation and monitoring are also required because of the frequent presence of heart disease in the stroke patient and the potential for dysrhythmias. The operating room must be ready to perform carotid endarterectomy, craniotomy, and removal of intracerebral and intracerebellar hematoma. Laboratory services must be available around the clock. Extensive and often specialized supportive care will be necessary in the days following admission. The center must have access to rehabilitation facilities and support services for the patients’ families.

The necessary staffing is similarly extensive. The minimum recommended is 24-hour availability of a nurse coordinator, emergency physician, neuroradiologist, neurologist, neurosurgeon, and vascular surgeon. The stroke team should also include rehabilitation specialists, and support from other fields such as vascular surgery and cardiology is recommended.

Nor is a stroke center’s work confined to the hospital. At a minimum, the center must work with emergency medical services, teaching the technicians how to recognize a likely stroke and what to do about it, including notification of the stroke center and expeditious transport of the victim. The stroke center also takes on the burden of educating the lay public and community physicians. The goal is to establish a community-wide system focused on delivery of state-of-the-art stroke care within the window of opportunity, as well as provide all other services necessary to ensure that the patient suffers as little residual impairment as possible.

Handling the Stroke

William T.C. Yuh, MD, professor of radiology at the University of Iowa, Iowa City, describes the five steps in managing a stroke (Table 2).

Table 2. The five steps in stroke management.

“Usually, by the time the patient is sent to you, the diagnosis has pretty much been made,” he notes. The existence of a stroke and its type (thrombotic or hemorrhagic) must still be confirmed. “You also need to know where the stroke is, how big it is, and how much of the injury is reversible,” Yuh continues.

At most stroke centers, the basic tool for assessing stroke is CT, with some centers using perfusion or xenon imaging. CT is almost 100% sensitive in detecting intracerebral hemorrhage; absence of blood flow is a good indicator of ischemic stroke. One scanner manufacturer recently released image analysis software specific for imaging strokes. The program analyzes arterial inflow, venous outflow, cerebral blood volume and flow, and mean transit time to determine the site, type, and severity of the stroke.

The University of Iowa prefers MRI, obtaining diffusion-perfusion imaging with cerebral blood volume measurement. MRI is more sensitive than CT, although the depiction of hemorrhage may be nonspecific, and is particularly helpful in the patient with cerebellar or brainstem infarction.

Whichever method is used, the images must be obtained quickly and interpreted promptly, preferably by a neuroradiologist.

“A stroke workup sounds simple, but it is complex, which is why it frequently is not done right,” observes John J. Connors III, MD, director of neurointerventional surgery, Inova Fairfax Hospital, Falls Church, Va. “The big bottleneck is the hospital: processing a patient can take until the next day. Intravenous thrombolysis has to be started within 3 hours of the onset of the stroke, and if a patient arrives at the hospital 2 hours after the event, by the time the hospital gets the CT scan and completes the blood work, it is too late. With the stroke patient, you cannot proceed step by step: you must do many things simultaneously. You even need to be willing to bump another patient off the CT scanner so you can get that critical image.”

Interventional Treatment

If the stroke is occlusive, the next step is balancing the risks and the benefits of thrombolytic treatment for the patient’s particular situation. Yuh thinks of this step as assigning a grade.

“There are only three outcomes of thrombolysis,” he notes. “You help the patient, you don’t help the patient, or you hurt the patient.” If benefit is the most likely outcome, the grade is +100; if thrombolysis is unlikely to make a difference, the grade is 0, and if the patient is likely to be harmed, it is -100. Therefore, only the patients with a +100 score receive thrombolysis.

“If you have a 35-year-old patient with an ischemic stroke in the dominant hemisphere and involvement of the middle cerebral artery territory, with 85% of the affected area having reversible ischemia, this patient is a +100 — a good candidate for treatment. Fifteen percent of such patients will die, but many of the rest will get well. On the other hand, if the patient is 85 years old with congestive heart failure, diabetes, and a stroke in the nondominant hemisphere, this is -100. Probably the best thing to do is simply to relieve the pressure on the brain.”

The mainstay of early therapy of thrombotic stroke is clot-dissolving enzymes such as tissue plasminogen activator (t-PA). In a widely publicized trial, five patients treated with intravenous t-PA were 30% more likely to have little or no disability at 3 months than those who received placebo. However, there was no difference in the degree of early neurologic improvement in the two groups.

Connors believes that acceptance of thrombolytic agents for stroke treatment has been hampered by the results of this and other similar trials of intravenous therapy and what they imply.

“With IV thrombolysis, health care workers do not see anybody getting better despite their extra efforts,” he notes. “They do see the complications, especially if patients are not properly selected.” For example, the risk of intracranial hemorrhage is considerably higher in stroke patients than in patients receiving the same thrombolytic agents for myocardial infarction (the rate was 16% in one recent study), especially if patients are not chosen according to standard protocols.

However, the reluctance to use the new (and expensive) agents is softening as the results of intra-arterial therapy trials emerge.

“Intra-arterial therapy gives results you see,” Connors stresses. “We deliver the agent directly to the clot, so we need only about 1/100 of the dose needed for intravenous administration. The lower dose reduces both the complications and the cost. We also can physically manipulate the clot. In some recent series, intra-arterial therapy within 6 hours of stroke onset has produced success rates as high as 70%. However, this treatment is labor intensive. The up-front cost also is considerable, although it is recovered in the first few weeks because many patients are cured. From the patient’s standpoint, the results are certainly worth the cost. I think intra-arterial therapy also is a bargain from society’s viewpoint when reduction in long-term costs is taken into account. Of course, long-term costs seldom are considered because typically, they are not the hospital’s concern.”

Whereas the guidelines for intravenous use of thrombolytic agents are formalized, guidelines for intra-arterial therapy are less clear.

“We know that neurons will die in 5 minutes if they do not have an adequate blood supply,” Yuh points out. “But you and I do not have the same collateral circulation. In fact, the collateral supply may be more or less extensive even in different areas of the same brain. One result is that for some people, thrombolysis 3 hours after a stroke begins is too late. It is in cases such as these that we may see intracranial hemorrhage and its consequences but no benefit. However, my team and others have described successful intra-arterial thrombolytic treatment as late as 12 hours. If you use perfusion MRI and compare the stroke site with the contralateral site in the brain, you can determine whether the patient is a treatment candidate even after 3 or 6 hours. If the perfusion is above 35%, the stroke site is not dead. If it is above 50%, the damage probably is reversible. We have treated a few patients as late as 20 hours after stroke onset with good results and without problems.”

To help determine how best to use intra-arterial thrombolysis in the face of these uncertainties, Connors is organizing an outcomes registry.

“Pharmaceutical companies are not going to pay for much more research on these issues because stroke treatment is only about 1% of the market for thrombolytic agents. But we need dose-finding studies so we can figure out what works best.” He is collecting data from interventional neuroradiologists on agents, doses, timing, and outcome at two Web sites: www.instor.org; and www.strokeregistry.org.

One of the interventional protocols in use at Connors’ hospital is administration of thrombolytic drug intravenously early in the patient’s course, an approach initially explored in the EMS Bridging Trial. [8] The idea is to begin delivering the thrombolytic agent while the interventional neuroradiology team is being summoned for intra-arterial therapy. In one version, intravenous therapy begins immediately — even in the ambulance — as in some clinical trials in patients with myocardial infarction. In the other version, recognizing that some strokes are hemorrhagic and thus that the patient would be harmed by thrombolysis, the agent is not given until an imaging study has confirmed that the stroke is thrombotic. In either protocol, the patient is then taken for an angiogram and removal of any remaining clot using more thrombolytic drug, physical manipulation, or both. With this approach, more patients can be treated within the window of opportunity.

“Intra-arterial thrombolysis is superior to intravenous thrombolysis,” Connors stresses. “But quick intravenous is better than late intra-arterial.”

Role of Education

The best thrombolytic agent is worthless if it is given after the neurons have died. Thus, stroke experts emphasize the role of stroke centers in public education.

The public tends to have a rather passive attitude toward stroke, believing that little can be done once it occurs. A person who is having a stroke may be so terrified that he or she denies the problem for several hours.

“Patients do not realize that this is their only chance,” Connors reports. “They lie down and wait to get better. When they do not [improve], they call the hospital, but it is too late.”

He would add two steps in front of Yuh’s five:

  • Patients must know their symptoms are an emergency.
  • Patients must be transferred to a hospital immediately.

Realizing the need to educate the public about the symptoms of stroke and their implications, the American Heart Association recently organized the American Stroke Association and began an educational program called Operation Stroke. The work is proving to be a challenge.

“For heart attack, the message is easy to get across: your chest hurts — heart attack! Dial 911,” Connors explains. “But we are having trouble finding a mascot, because a stroke is painless and does not lend itself to pictorial portrayal. We have to convey to the public the idea that if you have trouble seeing or speaking or have numbness, weakness, or many other symptoms, that’s a stroke: you should dial 911. Another problem is that the symptoms of stroke are too numerous to fit on a billboard that you can read as you drive by. Perhaps we should have a symptom of the month and rotate them to get the message across.”

Part 4

Connors and others believe that community education must also include information on stroke prevention such as improving diet and controlling high blood pressure. Also, members of the public tend not to understand that many strokes can be prevented even when there is significant carotid or cerebral atherosclerosis.

“I do intracranial angioplasty,” Connors reports. “Patients are referred to me who have significant symptoms and plaque in a brain artery. But they say, ‘It [symptoms of a transient ischemic attack] has only happened once or twice. There really isn’t anything wrong with me.’ But, of course, there is.”

Preventive education also must reach community physicians. “Most physicians do not know how to evaluate a patient who has had a transient ischemic attack,” Connors notes.


A 1993 meta-analysis9 demonstrated that specialized stroke centers repay their communities with better patient outcomes. These centers live by a motto articulated by Connors: “Time is brain. Everybody has to know that.”

Further Resources

  • An Acute Stroke Toolbox is available on the Web at http://www/stroke-site.org. The site, a cooperative effort of the Brain Attack Coalition (the National Institute of Neurological Disorders and Stroke and several nurse and physician professional societies), provides tools for health care professionals to help them diagnose and treat stroke rapidly. Among the offerings are guidelines, pathways, orders, stroke scales, and public information.
  • On October 15 and 16, a Stroke Therapy Conference will be held at the Ritz-Carlton Hotel in Washington, DC. Part of the conference will be devoted to establishing a stroke center. For further information, contact:
  • Society of Cardiovascular and Interventional Radiology, (703) 691-1805; www.scvir.org.

Judith Gunn Bronson, MS, is a contributing writer for Decisions in Axis Imaging News.