Clinical Topics & News

Interview with Andrew Solomon, MD, on diagnosing multiple sclerosis


Andrew Solomon, MD, is a neurologist and Associate Professor in the Department of Neurological Sciences and Division Chief of Multiple Sclerosis at The University of Vermont. We sat down to talk with Dr. Solomon about his experience with multiple sclerosis (MS) misdiagnosis and what can be done to improve MS diagnosis going forward.

How prevalent is the misdiagnosis of MS and what are the effects that it has on patients?

The first thing to clarify is what we mean by MS misdiagnosis. In this case, we’re talking about patients who are incorrectly assigned a diagnosis of MS. MS is hard to diagnose. Sometimes we take too long to diagnose people who have MS, sometimes we incorrectly diagnose MS in people who don’t have it.

We don’t have very good data in terms of how frequent misdiagnosis is, but we have some. The earliest data we have is from the 1980s. In 1988 there was a study published involving approximately 500 patients who had been diagnosed with MS in life and subsequently died between the 1960s and the 1980s and had a postmortem exam. 6% of those people who had a diagnosis of MS during their lifetime didn’t actually have MS. 1

In 2012, we did a survey of 122 MS specialists. 2 We asked them if they had encountered patients incorrectly assigned a diagnosis of MS in the past year, and 95% of them had seen such a patient. This was not the most scientific study because it was just a survey and subject to recall bias. Still, many of these MS providers recalled having seen three to ten such patients in the last year where they strongly felt that a pre-existing MS diagnosis made by another provider was incorrect.

Another study was recently published by Dr. Kaisey. 3 She looked at referrals to two academic MS centers on the West Coast, and specifically new patient referrals over 12 months to those two MS centers. She found that out of 241 patients referred to the two MS centers, almost one in five who was referred with a pre-existing diagnosis of MS who came to these MS centers was subsequently found not have MS. This included 17% of patients at Cedars Sinai and 19% at UCLA. That’s an alarmingly high number of patients. And that’s the best data we have right now.

We don’t know how representative that number is of other MS centers and clinical practice nationally, but the bottom line is that misdiagnosis of MS is, unfortunately, fairly common and there’s a lot of risks to patients associated with misdiagnosis, as well as costs to our health care system.

Can you elaborate on the risks to patients?

We published a study where we looked at records from 110 patients who had been incorrectly assigned a diagnosis of MS. Twenty-three MS specialists from 4 different MS centers participated in this study that was published in Neurology in 2016. 4

70% of these patients had been receiving disease modifying therapy for MS, which certainly has risks and side effects associated with it. 24% received disease modifying therapy with a known risk of PML, which is a frequently fatal brain infection associated with these therapies. Approximately 30% of these patients who did not have MS were on disease modifying therapy for in 3 to 9 years and 30% were on disease modifying therapy for 10 years or more.

Dr. Kaisey’s study also supports our findings. She found approximately 110 patient-years of exposure to unnecessary disease modifying therapy in the misdiagnosed patients in her study. 3 Patients suffer side effects in addition to unnecessary risk related to these therapies.

It’s also important to emphasize that many of these patients also received inadequate treatment for their correct diagnoses.

How did the 2017 revision to the McDonald criteria address the challenge of MS diagnosis?

The problem of MS misdiagnosis is prominently discussed in the 2017 McDonald criteria. 5 Unfortunately, one of the likely causes of misdiagnosis is that many clinicians may not read the full manuscript of the McDonald criteria itself. They instead rely on summary cards or reproductions—condensed versions of how to diagnose MS—which often don’t really provide the full context that would allow physicians to think critically and avoid a misdiagnosis.

MS is still a clinical diagnosis, which means we’re reliant on physician decision-making to confirm a diagnosis of MS. There are multiple steps to it.

First, we must determine if somebody has a syndrome typical for MS and they have objective evidence of a CNS lesion. After that the McDonald criteria can be applied to determine if they meet dissemination in time, dissemination in space, and have no other explanation for this syndrome before making a diagnosis of MS. Each one of those steps is reliant on thoughtful clinical decision-making and may be prone to potential error.

Knowing the ins and outs and the details of each of those steps is important. Reading the diagnostic criteria is probably the first step in becoming skilled at MS diagnosis. It’s important to know the criteria were not developed as a screening tool. The neurologist is essentially the screening tool. The diagnostic criteria can’t be used until a MS-typical syndrome with objective evidence of a CNS lesion is confirmed.

In what way was the previous criteria flawed?

I wouldn’t say any of the MS diagnostic criteria were flawed; they have evolved along with data in our field that has helped us make diagnosis of MS earlier in many patients. When using the criteria, it’s important to understand the types of patients in the cohorts used to validate our diagnostic criteria. They were primarily younger than 50, and usually Caucasian. They had only syndromes typical for MS with objective evidence of CNS damage corroborating these syndromes. Using the criteria more broadly in patients who do not fit this profile can reduce its specificity and lead to misdiagnosis.

For determination of MRI dissemination in time and dissemination space, there are also some misconceptions that frequently lead to misdiagnosis. Knowing which areas are required for dissemination in space in crucial. For example, the optic nerve currently is not an area that can be used to fulfill dissemination in space, which is a mistake people frequently make. Knowing that the terms “juxtacortical” and “periventricular” means touching the ventricle and touching the cortex is very important. This is a mistake that’s often made as well, and many disorders present with MRI lesions near but not touching the cortex or ventricle. Knowing each element of our diagnostic criteria and what those terms specifically mean is important. In the 2017 McDonald criteria there’s an excellent glossary that helps clinicians understand these terms and how to use them appropriately. 5

What more needs to be done to prevent MS misdiagnosis?

First, we need to figure out how to better educate clinicians on how to use our diagnostic criteria appropriately.

We recently completed a study that suggests that residents in training, and even MS specialists, have trouble using the diagnostic criteria. This study was presented at the American Academy of Neurology Annual meeting but has not been published yet. Education on how to use the diagnostic criteria, and in which patients to use the criteria (and in which patients the criteria do not apply) is important, particularly when new revisions to the diagnostic criteria are published.

We published a paper recently that provided guidance on how to avoid misdiagnosis using the 2017 McDonald criteria, and how to approach patients where the diagnostic criteria didn’t apply. 6 Sometimes additional clinical, laboratory, or MRI evaluation and monitoring is required in such patients to either confirm a diagnosis of MS, or determine that a patient does not have MS.

We also desperately need biomarkers that may help us diagnose MS more accurately in patients who have neurological symptoms and an abnormal MRI and are seeing a neurologist for the first time. There’s research going on now to find relevant biomarkers in the form of blood tests, as well as MRI assessments. In particular, ongoing research focused on the MRI finding we have termed the “central vein sign” suggests this approach may be helpful as a MRI-specific biomarker for MS. 7,8 We need multicenter studies evaluating this and other biomarkers in patients who come to our clinics for a new evaluation for MS, to confirm that they are accurate. We need more researchers working on how to improve diagnosis of MS.


1. Engell T. A clinico-pathoanatomical study of multiple sclerosis diagnosis. Acta Neurol Scand. 1988;78(1):39-44.

2. Solomon AJ, Klein EP, Bourdette D. "Undiagnosing" multiple sclerosis: the challenge of misdiagnosis in MS. Neurology. 2012;78(24):1986-1991.

3. Kaisey M, Solomon AJ, Luu M, Giesser BS, Sicotte NL. Incidence of multiple sclerosis misdiagnosis in referrals to two academic centers. Mult Scler Relat Disord . 2019;30:51-56.

4. Solomon AJ, Bourdette DN, Cross AH, et al. The contemporary spectrum of multiple sclerosis misdiagnosis: a multicenter study. Neurology. 2016;87(13):1393-1399.

5. Thompson AJ, Banwell BL, Barkhof F, et al. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. Lancet Neurol . 2018;17(2):162-173..

6. Solomon AJ, Naismith RT, Cross AH. Misdiagnosis of multiple sclerosis: impact of the 2017 McDonald criteria on clinical practice. Neurology. 2019;92(1):26-33.

7. Sati P, Oh J, Constable RT, et al; NAIMS Cooperative. The central vein sign and its clinical evaluation for the diagnosis of multiple sclerosis: a consensus statement from the North American Imaging in Multiple Sclerosis Cooperative. Nat Rev Neurol. 2016;12(12):714-722.

8. Sinnecker T, Clarke MA, Meier D, et al. Evaluation of the central vein sign as a diagnostic imaging biomarker in multiple sclerosis [published online ahead of print August 19, 2019] . JAMA Neurology. 2019: doi: 10.1001/jamaneurol.2019.2478.

Next Article: