Advanced stroke imaging leads to new management challenges

The patient was a married engineer and father of three. He presented to the emergency room with a global arterial circulation deficit, aphasia, and hemiplegia. An earlier noncontrast CT of the brain had found no signs of hemorrhage. An MR scan revealed a considerable perfusion defect caused by a carotid artery occlusion. He arrived 1.5 hours too late, however, for intravenous tissue plasminogen activator thrombolytic therapy.

“What am I going to do?” said Dr. Howard A. Rowley, chief of neuroradiology at the University of Wisconsin in Madison. “Say ‘Sorry, I can't do anything more, we are going to give you an aspirin and send you to rehab?’ I don't think so.”

Though anecdotal, the account reflects a situation seen daily at both small hospitals and large stroke care centers across the U.S. About 750,000 stroke cases will be diagnosed this year, according to the National Stroke Association. The condition has become a leading cause of death and disability in the U.S. and Europe. But even though intravenous tPA was approved more than a decade ago, only about 4% of patients actually get the drug-mostly those who can make it to the ER within three hours of stroke onset.

“We are not doing very well with this disease,” Rowley said. “Maybe it's time to think outside the box.”

A growing number of recently published studies may be persuading researchers and clinicians about the possibility of identifying salvageable brain tissue beyond the established three-hour window for lytic therapy. Rowley referred to the anecdotal case described above to illustrate this point. His institution's 15-minute MR scanning protocol found a perfusion/diffusion mismatch in the engineer's brain that helped guide further management. The patient underwent intra-arterial therapy and returned home the next day with no neurological deficit.

“It would have been immoral and unethical not to treat this patient, or at least offer him this treatment option. Was that the right thing to do? I don't know, but it was certainly sensible based on what we knew about him and what we knew about opening vessels as late as six hours after occlusion,” Rowley said.

Not everyone agrees with this assessment.

“I would not be treating him,” said Dr. William J. Powers, chair of neurology at the University of North Carolina School of Medicine in Chapel Hill.

Powers has declined to treat this type of patient on several occasions. In the absence of controlled clinical trial data, cases such as this ought to be seen as experimental and not done unless somebody signs an informed consent form, he said.

Many debates such as this surround imaging and management of acute stroke patients. A panel of stroke imaging experts, including Rowley and Powers, discussed these controversies at the 2007 RSNA meeting. But far from conveying an aura of conflict, the conversation suggested the field is moving on strongly toward groundbreaking developments and, possibly, new clinical standards.

PENUMBRA IMAGING

Noncontrast CT remains the imaging gold standard for patients presenting with signs of acute stroke. But a growing body of clinical literature suggests a more prominent role for MRI in diagnosing stroke and strokelike conditions.

With both champions and detractors, the concept of ischemic penumbra has gained momentum in the imaging community. In theory, the ischemic penumbra represents a zone of reduced blood flow surrounding the infarcted core. Brain tissue in this zone will die if left untreated. The best known MRI-based penumbral imaging technique is the diffusion/perfusion mismatch. Diffusion-weighted im­aging reveals the infarcted core, while perfusion-weighted imaging shows an estimate of cerebral blood flow reduction. The mismatch between the two imaging techniques reveals the area of potentially salvageable ischemic brain tissue.

Available studies have shown MRI to be superior to noncontrast CT for the diagnosis of acute ischemic stroke and even intracranial hemorrhage (Lancet 2007;369: 293-298). Imagers caution, however, that many of these comparisons pitch advanced MR against plain CT. New developments in multislice CT angiography and perfusion imaging could level the field of penumbra imaging and help CT retain an edge in areas like hemorrhagic stroke diagnosis.

Despite its sophistication, advanced neuroimaging of ischemic brain tissue has its downsides. One is that too many models have never been compared or validated in large trials, said Dr. Max Wintermark, an assistant professor of radiology at the University of Cali­fornia, San Francisco. Making things even more complicated is the lack of standardization of many perfusion parameters. Mean transit times, cerebral blood flow, and cerebral blood volume estimates may vary from one manufacturer's image postprocessing software to another's (see accompanying article).

“You have to remember that you can have the same CBF map, but-depending on which software it comes from-the values are not necessarily going to be the same. As a result, you cannot necessarily apply the same threshold of the model you want to use,” Wintermark said.

Being able to demonstrate preventable infarction does not prove clinical utility. The only way to prove that is by demonstrating that patients who had this type of imaging applied to their care actually do better. That is not easy to do, Powers said.

Only a randomized controlled trial that evaluates imaging's contribution and/or adverse effects to therapy could validate the concept of looking for salvageable tissue or penumbra, according to Powers. Imaging by itself could have adverse effects on the outcome, by delaying treatment, for instance.

A good study design should test the hypothesis of the end value and real costs of imaging, he said.

“The only available treatment we have for acute stroke is IV tPA under three hours. And it requires noncontrast CT. That's the type of imaging we should be doing to guide our therapy in acute stroke. Anything else is based on unproven hypotheses and therefore should be performed not as part of clinical practice, but only as part of approved clinical studies,” Powers said.

BRAIN VERSUS HEART

Stroke prevention represents a key part of the care equation but one that lags seriously behind diagnosis and treatment. Hypertension reduction and conservative medical treatment could probably prevent about 75% of strokes every year, according to Dr. Michael H. Lev, an associate professor of ra­diology at the University of California, Los Angeles.

Sroke has a multiple set of causes. About a fifth of them are hemorrhagic, while the rest are ischemic. About 30% of strokes involve a large neck or intracranial vessel, such as a carotid bifurcation, and 30% are embolic, usually due to atrial fibrillation, according to the clinical literature. Recently, the issue of vulnerable plaque in neurovascular disease has become just as prominent as plaque in the coronaries. Degree of stenosis remains the established standard for deciding who gets medical treatment, stenting, or surgery. In spite of this, some phenomena still bewilder investigators. Why, for instance, do some patients with irrelevant stenosis get strokes, while many patients with severe stenosis do not? How can physicians effectively stratify that risk?

Those questions could also be connected with interest in investigating plaque. Tissue-specific biomarkers, not general serum markers, may prove useful in the future for risk prediction and as indicators of the type of inflammatory change that leads to hot plaque and emboli. Inflammatory enhancement on MR has been shown to correlate to a certain degree with symptomatic and even asymptomatic patients. Prelim­inary MRI work with magnetic monocrystalline iron oxide nano­particles looks promising.

Although it seems that early claims about imaging of vulnerable plaque were a bit overrated, the future points in the direction of targeted tissue-specific molecular imaging of inflammation to determine degree of stenosis and stroke risk, according to Lev.

A relevant note to consider is the need to set the focus on the carotid arteries and the vascularity of the brain regarding vulnerable plaque apart from other vascular areas. A critical look shows that a great deal of the literature on the carotids links to references from the coronary arteries, which is incorrect, he said.

“The big take-home message is that the heart is not the brain, that coronary arteries are not carotid arteries,” Lev said.

Stroke imaging experts will surely face many more issues and challenges, some of which are hard to predict. The traditional imaging management model seems to be evolving, however, from a time-oriented one into a tissue-oriented, or better, an individual-oriented one. Advanced CT and MR imaging techniques, separately or together, will continue to play an important role in screening, diagnosis, and management of acute stroke. Stroke prevention may one day occupy a more prominent place in the nation's clinical agenda. 

Mr. Abella is associate editor of Diagnostic Imaging.