Conference Coverage

Eye movement, not CT or MRI, rules out posterior stroke

 

Key clinical point: Posterior fossa strokes reveal themselves by eye movement abnormalities, not imaging.

Major finding: The sensitivity of CT for acute posterior strokes was just 30.8%. The sensitivity of MRI was better at 76.4%, but it still missed one in four.

Data source: Meta-analysis of more than 800 patients

Disclosures: There was no industry funding for the work. The senior investigator might profit from commercialization of HINTS training.


 

AT ANA 2017

– When a patient with vertigo and dizziness presents to the emergency department, the first order of business is to figure out if they’re due to benign inner ear problems or a posterior fossa stroke.

Emergency department physicians usually use noncontrast CT (NCCT) to rule out stroke, while neurologists turn to MRI with diffusion-weighted imaging (MRI-DWI) to make the call.

Neither are good enough. The sensitivity of NCCT for acute posterior fossa strokes is just 30.8%. The sensitivity of MRI-DWI is better at 76.4%, but “false negatives [occur] in roughly one [of] four brainstem strokes in the first 48 hours,” according to a meta-analysis from Johns Hopkins University, Baltimore. The study, presented at the annual meeting of the American Neurological Association, involved more than 800 patients in the 14 strongest studies to look into the issue since 1990.

Dr. David Newman-Toker, professor of neurology, ophthalmology, and otolaryngology at Johns Hopkins University, Baltimore
Dr. David Newman-Toker
“We miss a ton of posterior circulation strokes. CT stinks and you shouldn’t use it, and MRI is not as good as you think it is. There’s a big problem out there that has to be solved,” said senior investigator David Newman-Toker, MD, PhD, a professor of neurology, ophthalmology, and otolaryngology at Hopkins, and director of the school’s Armstrong Institute Center for Diagnostic Excellence.

“Anterior circulation strokes are obvious, but posterior circulation strokes are subtle. We’ve surveyed ED physicians, and there’s a certain rate in the population who just don’t understand how bad CTs are for detecting acute stroke.” Meanwhile, “neurologists know” they can’t rely on CTs, “but they think MRIs are good enough. That turns out not to be true either,” he said.

Dr. Newman-Toker’s comments were sparked by a discussion about the meta-analysis, but he spoke from years of work trying to improve the situation. He was clear about what’s at stake: The early recognition of posterior fossa strokes, treatment with thrombolytics and surgical decompression (when warranted), and prevention of a second, larger stroke, can mean the difference between walking out of the hospital and dying or being wheelchair bound for life.

He and his colleagues have developed a way to detect posterior fossa strokes using eye movement abnormalities. They call it HINTS, which stands for “head impulse, nystagmus, and test of skew. “It turns out that the eye movements of inner ear disease look slightly different than the eye movements of brain disease; the subtle differences are enough to distinguish between the two.” When the technique is mastered, “our best estimate is that the sensitivity for posterior circulation stroke is around 99%,” Dr. Newman-Toker said. He is working to get the message out and train people; a video of the technique is online (Semin Neurol. 2015 Oct;35[5]:506-21).

As for the meta-analysis, “none of us were terribly surprised that CT wasn’t much good, and I knew that MRI wasn’t going to be perfect, but it was worse than I expected when we crunched all the numbers,” he said.

It’s no surprise that eye movement trumps imaging. “Physiology beats anatomy in the acute phase. It’s takes a little while for anatomy to change” on imaging, but eye movements change immediately “when patients become symptomatic with dizziness and vertigo,” he said.

Meanwhile, “if you don’t know how to evaluate peoples’ eye movements, I think MRIs are the next best thing,” he said.

There was no industry funding for the work. Johns Hopkins is working with a company called Natus to develop HINTS training. Dr. Newman-Toker might earn royalties, but so far “I haven’t made any money off it, and there’s no document saying I’m ever going to make any money from it,” he said.

   Comments ()

Recommended for You

News & Commentary

Quizzes from MD-IQ

Research Summaries from ClinicalEdge

Related Articles