Share this article on:

Migraine Patients Often Misdiagnosed with MS Based on Evidence of White Matter Hyperintensities: Study Points to Need for Multiple Criteria for Diagnosis

Valeo, Tom

doi: 10.1097/01.NT.0000433895.38794.c2
Back to Top | Article Outline


CORONAL FLAIR image in this 54-year-old female imaged for headaches shows many white matter hyperintensites that met that Barkhof and 2010 McDonald criteria for multiple sclerosis

CORONAL FLAIR image in this 54-year-old female imaged for headaches shows many white matter hyperintensites that met that Barkhof and 2010 McDonald criteria for multiple sclerosis

Investigators reported that 2.4-7.1 percent of 168 headache patients between the ages of 10 and 55 who displayed T2 white matter hyperintensities on a brain MRI met the inclusion criteria for the Barkhof criteria for multiple sclerosis, while 24.4-34.5 percent met the more liberal McDonald criteria.

Neurologists called upon to determine the significance of white matter hyperintensities (WMHs) found on a brain scan may over-diagnose multiple sclerosis (MS) if they rely excessively on two common references — the Barkhof and the 2010 McDonald criteria, which consider the “dissemination in space” of the lesions.

According to a new study in the July issue of the journal Multiple Sclerosis, 2.4-7.1 percent of 168 headache patients between the ages of 10 and 55 who displayed T2 WMHs on a brain MRI met the inclusion criteria for the Barkhof criteria, while 24.4-34.5 percent met the more liberal McDonald criteria. [See sidebar: “Age-modified Barkhof and McDonald Classification Scheme based on Definitions of Juxtacortical and Periventricular Lesions” for descriptions of the two criteria.]

Even after taking into account the fact that MS is found more often in headache patients, this rate of diagnosis was far above the .09 percent incidence of MS found in the general population, according to lead author Bronwyn E. Hamilton, MD, an assistant professor of neuroradiology at the Oregon Health and Sciences University in Portland.

“We know about how frequently MS appears in the population, so we know that a very large proportion of migraine patients with white matter hyperintensities that meet the Barkhof or McDonald criteria are unlikely to have MS,” Dr. Hamilton said. “Yet, a large number of patients have been labeled as having radiologically isolated syndrome (RIS) when all they had was headache.”

RIS, a term introduced in a 2009 paper in Neurology, refers to patients who display WMHs on a brain MRI, and have a high risk of converting to MS based on the shape, size, and location of lesions, even though they experience no other symptoms. (See “Radiologically Isolated Syndrome: Some Second Thoughts.”)

An editorial that ran in the same issue of Neurology by Dennis Bourdette, MD, and Jack Simon, PhD, both of the Oregon Health and Science University, emphatically discouraged treating such patients. “Even patients who show radiologic progression cannot be considered to have MS if they have no symptoms of the illness,” they stated. “For now, it is best to recall the wise advice that we treat the patient, not the MRI scan.”

Back to Top | Article Outline


The research just published in the Multiple Sclerosis Journal, which includes Drs. Bourdette and Simon among the co-authors, provides support for their assertion. The study looked at MRI results from 564 headache patients who displayed no signs of demyelinating disease. The researchers excluded patients younger than 10, who seldom develop MS, and older than 55, who commonly have leukoaraiosis – apparently benign white matter brain lesions that appear on MRI and could have confounded the results.



No caption available

No caption available

Of the remaining 326 – a group that included twice as many women as men – WMHs were found in 168 (51.5 percent). The researchers found that 2.4 percent met the Barkhof “touching” criteria, which gives greater weight to WMHs that contact the ventricular and cortical surfaces, while 7.1 percent met the 3 mm touching criteria, which includes lesions within 3 mm of those surfaces. Also, 24.4 percent met the McDonald touching criteria, while 34.5 percent met the 3 mm touching criteria.

Back to Top | Article Outline


“The authors of this paper are saying that you have to pay attention to the fact that MRI is only a supportive element of an MS diagnosis,” said Mark Steven Freedman, MD, director of the Multiple Sclerosis Research Unit and professor of medicine at the University of Ottawa in Ontario, and a senior scientist in neuroscience at the Ottawa Hospital Research Institute. “No MRI by itself can diagnose MS. Patients must present a clinical story that is compelling. If there is no better explanation for the signs and symptoms, then you consider MS as diagnosis. If the shoe doesn't fit, it's not MS.”

Fred D. Lublin, MD, a co-author of the 2010 revisions to the McDonald Criteria, agreed. “That is perhaps the most important point of this paper — the McDonald criteria for diagnosing multiple sclerosis are for diagnosing multiple sclerosis, not for the differential diagnosis of multiple sclerosis,” said Dr. Lublin, Saunders Family professor of neurology, and director of The Corinne Goldsmith Dickinson Center for Multiple Sclerosis at the Icahn School of Medicine at Mount Sinai. “The criteria don't distinguish between multiple sclerosis and other disorders that can cause MRI changes in the brain. Even though the people in this study do not meet the criteria for a diagnosis of multiple sclerosis because they never had a clinical event, some of them meet the McDonald-Barkhof MRI criteria, so they should be followed for the development of a clinical event.”



Brian Weinshenker, MD, professor of neurology at the Mayo Clinic in Rochester, MN, another co-author of the McDonald criteria, said he frequently ends up disagreeing with the MS diagnosis applied to patients referred to him who have WMHs on their MRI and display atypical symptoms. “They're basically healthy people with white matter spots, but they're on these $40,000- and 50,000-a-year medications indefinitely,” he said. “Clinicians often feel an imperative to make a diagnosis early and start treatment, but that push to make an early diagnosis often occurs when symptoms and signs are non-specific or even non-existent. Even if a patient meets the Barkhof or McDonald MRI criteria, that is not sufficient to support a reliable diagnosis of MS.”

While some research suggests that the risk of progressive disability can be reduced with early diagnosis followed by disease-modifying therapies, the evidence remains inconclusive, according to Bruce A. Cohen, MD, professor of neurology and clinical neurosciences at the Northwestern University, Feinberg School of Medicine.

“The evidence is either retrospective or limited by potential selection bias,” he said. “Unfortunately the definitive study can't be done because you can't do a 25-year placebo-controlled study of treatment vs. no treatment of MS.”

Back to Top | Article Outline


Clinically isolated syndrome (CIS) refers to people who experience an episode of numbness, vision problems, or some other symptom suggestive of multiple sclerosis (MS). But what if they display nothing more than white matter hyperintensities (WMHs) on MRI that could be consistent with MS?

In an effort to identify such people, who have an elevated risk of developing MS, Darin T. Okuda, MD, and colleagues published a paper in 2009 in Neurology that introduced the term radiologically isolated syndrome (RIS) and proposed formal criteria for identifying it. They studied 44 asymptomatic subjects (41 women and 3 men) who displayed WMHs highly suggestive of MS on brain MRIs done for other reasons, including headache and head trauma. WMHs increased in 59 percent of those imaged, and 10 of the 44 patients developed clinically isolated syndrome or definite MS after an average of 5.4 years.

At the time the Neurology paper was published, Dr. Okuda, now associate professor of neurology at the University of Texas Southwestern Medical Center at Dallas, expressed hope that RIS would “expand the phenotype of at-risk individuals for future demyelinating events,” and it did, which helped spur Dr. Hamilton and her colleagues to undertake their research. “That's the group that stimulated our interest in this,” Dr. Hamilton said of Dr. Okuda and his co-authors. “They said it's probably good to treat these patients with immunological therapy, but our group has been skeptical of that.”

Today, Dr. Okuda agrees with the caution advocated by the authors of the recent Multiple Sclerosis Journal paper. “Their data highlight limitations with the existing dissemination in space criteria for multiple sclerosis,” he said. “We should keep in mind that these criteria were principally designed for application in patients who have experienced a first clinical attack. The authors make the point that we may be over-diagnosing or over-suspecting RIS in headache sufferers with incidental anomalies within the brain.”

Still, Dr. Okuda believes that RIS helps to identify patients with WMHs who should be watched closely. “It is important to highlight that the 2009 RIS criteria extend beyond the presence of MRI anomalies suggestive of MS,” he said. “The criteria incorporate clinical information from subjects.”

—Tom Valeo

Back to Top | Article Outline


•. Liu S, Kullnat J, Bourdette D, Simon J, et al. Prevalence of brain magnetic resonance imaging meeting Barkhof and McDonald criteria for dissemination in space among headache patients. Multiple Sclerosis Journal 2013;19(8):1101–1105.
    •. Okuda DT, Mowry EM, Beheshtian A, et al. Incidental MRI anomalies suggestive of multiple sclerosis: the radiologically isolated syndrome. Neurology 2009;72(9):800–805.
      •. Bourdette D, Simon J. The radiologically isolated syndrome: Is it very early multiple sclerosis. Neurology 2009;72(9):780–781.
        •. Mistry N, Dixon J, Tallantyre E, et al. Central veins in brain lesions visualized with high-field magnetic resonance imaging: A pathologically specific diagnostic biomarker for inflammatory demyelination in the brain. JAMA Neurology 2013;70(5):623–628.
          Neurology Today archive on radiologically isolated syndrome:
            © 2013 American Academy of Neurology