乌鸦传媒

WASHINGTON -- Physicians trying to figure out the cause of an unexplained stroke need to take a lesson from former Secretary of Defense Donald Rumsfeld, according to .

In particular, Rumsfeld's reference to "unknown unknowns" -- things we don't know that we don't know -- is a useful idea, said Elkind, professor of neurology and epidemiology at Columbia University in New York City, at a meeting here last week on cryptogenic stroke sponsored by Medtronic, the American Heart Association, and the American Stroke Association.

"In many cases, we just may not have looked for those things that could have caused the stroke," he said.

Cryptogenic stroke is defined as "a brain infarct not attributed to a definite source of large-vessel atherosclerosis, cardioembolism, or small vessel disease; it can be in the presence of an extensive work-up, incomplete evaluation, or more than one cause," Elkind said.

A more recent evolution of the concept is ESUS, for "embolic stroke of undetermined [or unknown] source." This definition, he continued, implies that "the patient has had a full evaluation to show they don't have [a] small-vessel stroke, no evidence of extracranial or intracranial atherosclerosis, no major-risk cardioembolic source, and they have no other specific cause" such as a dissection.

Cryptogenic stroke accounts for around 200,000 to 250,000 strokes per year, and minorities such as African Americans and Hispanics are disproportionately affected, Elkind said. Although common causes of stroke, such as atrial fibrillation (Afib), are well-known, "more recent evidence suggests there probably are a number of other atrial abnormalities also associated with stroke, such as other arrhythmias, blood markers like BMP, an enlarged left atrium, and abnormalities on EKG," he added.

If the initial evaluation of a stroke patient -- which typically includes a medical history, physical examination, blood tests, an EKG, and imaging of the brain, blood vessels, and heart -- doesn't yield a cause of a stroke, then more workups are called for. These tests for "known unknowns" could include transesophageal echocardiography, more prolonged cardiac monitoring, additional imaging of blood vessels, and further lab testing.

And there is a huge list of possible additional causes that might pop out at that evaluation, including migraine, genetic disorders, non-stenosing plaques, infections, sleep apnea, cardiac diseases, patent foramen ovale, and hypercoagulable states, Elkind stated.

Negative Effects

Not knowing the cause of a stroke can have a negative effect on patients and physicians alike, said , director of stroke research at Northwestern University Feinberg School of Medicine in Chicago. "We get lots of different things thrown at us, like 'You don't know what caused my stroke?' That's damning; 'how come you don't know something as simple as a cause of something as important as a stroke'?" he said.

Or patients will ask, "Am I a ticking time bomb?" or "Should I go get a second, third, or fourth opinion?," he continued. "Maybe you don't know [the reason], but someone else does ... If you don't know what caused it how can you tell [a patient] what to do?"

Sometimes patients will say, "'You looked for everything and I must not have some of the bad causes of stroke. I guess that's a good thing!'" Prabhakaran said. "I try to play up on that... It's the notion of half empty or half full. If we didn't find [the cause], you perhaps now have gotten yourself into a lower risk category -- you don't have atrial fibrillation, you don't have a carotid lesion that needs to be treated. I think that's the way we need to communicate."

He noted that part of the problem with cryptogenic stroke is that "there really isn't a concise diagnostic evaluation that every patient is supposed to get; there's so much variability based on where you are."

"There's really an opportunity here for this group ... to help standardize what is the expectation and what tests should be offered to these patients," Prabhakaran said. "Are there determinants of that? Are there access issues related to race, ethnicity, or income that play a role in determining who gets what tests?"

Role of Afib

, director of the cardiac arrhythmia service at St. Vincent Hospital in Indianapolis, discussed the complicated role of atrial fibrillation. "Just because you identify atrial fibrillation on the workup doesn't mean it was the cause of the stroke," he pointed out. On the other hand, if atrial fibrillation is the cause, it's sometimes difficult to detect since patients don't always feel it.

Prystowsky suggested that patients hospitalized for cryptogenic stroke who might have undiagnosed atrial fibrillation should be discharged with a monitor. "I don't understand why [cryptogenic stroke] patients don't go home with a monitor; why do you want to wait 3 months? If you haven't figured it out in the hospital and you're suspicious, you should get them on a monitor at the time of discharge."

There are also causes of stroke that no one really thinks about, said , professor of neurology at the University of Texas Southwestern Medical Center in Dallas. "We do a whole-body scan with a person who has a cryptogenic stroke and [often pick up] cancer," he said. "That's often what causes a hypercoagulable state."

Drug abuse is also not tested for often, he said, adding that when he worked at Northwestern University, "I had an edict ... Everybody who walks in with a stroke or TIA [transient ischemic attack] gets a [toxicology] screen." The oldest positive tox screen he saw was for an "80-something" patient who was using cocaine.

Role of Stroke Centers

Due to the complexity of cryptogenic stroke, "most patient benefit from a workup at a comprehensive stroke center" (CSC) such as at Southwestern, Alberts said. "The thoroughness of the workup and accuracy of diagnosis is going to vary greatly if the patient is not evaluated properly."

, executive vice president for healthcare quality evaluation at the Joint Commission, in Chicago, agreed. "One of goals should be to minimize cryptogenic strokes by getting a proper workup. But if that not done, [you] need referral to CSC."

But , chair of the clinical policies committee at the American College of Emergency Physicians, was concerned about the effects of such a recommendation. "It seems like we must consider appropriate use of resources," he said.

"A lot of what we talk about is endless workup. If you had a tiny stroke, a tiny on an MRI, that's going to initiate a cascade of events; there is a cost to this. Where does cost-effectiveness come in?"

"We would hope some of these diagnostic pathways would have stops," Baker responded. "If you hit an extremely high level of confidence, you may not need to go through a full workup for every patient." However, he added, "All of this is probably going to get dramatically worse in the next few years because of high-deductible health plans; we are going to do expensive workups and people will leave and not be able to afford their medications."

Terminology was another issue debated at the meeting. The term "cryptogenic stroke" should go away, , chief of cardiology at Northwestern University Feinberg School of Medicine, told 乌鸦传媒. Instead, it should be "replaced with a specific cause for stroke or ESUS, because at least it would implicate a treatment," he said.

Alberts, for his part, objected to the use of ESUS. "That's sort of a misnomer because it implies we know it's an embolic stroke. How do you know that?" he said. "When does ESUS become just SUS?"

Comment