Differential Diagnosis of Multiple Sclerosis

Fred Lublin, MD: After the 2005 criteria came out, and each of these criteria have always had the caveat that there be no better diagnosis—every diagnostic criteria for multiple sclerosis [MS] going back, I don’t know how far now, has always had that caveat because dissemination in time, dissemination in space is characteristic of multiple sclerosis, but not unique to multiple sclerosis. And so after 2005, there was from the community a cry saying, well enough of this caveat, give us some guidelines, and that led to a 2008 paper. The lead author was David Miller, MBChB, FRACP, from your institution, Wallace, National Hospital for Neurology and Neurosurgery. Among other things, it provided some very useful differential diagnosis information in the form of charts that had high probability, median probability, low probability of red flags, of things to be watched for that, if you see those, then you ought to rethink your diagnosis of multiple sclerosis independent of whether someone fulfilled these criteria. That was published in the Multiple Sclerosis Journal in December 2008, but it had interesting things like a red flag if you have a persistently enhancing lesion, one ought to think of something else. If you have joint disease concurrently or skin rashes or things of that sort, that one ought to think of something else. That was already 12 years ago. Do you think you see a need to update those?

Patricia K. Coyle, MD: I do think there might be a real value to readdress that. For example, we are pretty routinely now in first attack checking for IgG [immunoglobulin G] to aquaporin-4 and IgG to MOG [myelin oligodendrocyte glycoprotein], particularly in the optic neuritis, transverse myelitis. But if we’re trying to define the MOG-antibody associated syndrome, then it’s reasonable to do that, although we might have a debate about that. There would be some help in a re-analysis of what might be recommended evaluations and work-ups for ruling out other conditions.

Wallace Brownlee, MBChB, PhD, FRACP: I agree entirely with what Patricia said, but I’d also highlight how, in recent years, it’s become clear that the conditions that are most likely to be mistaken for multiple sclerosis, and multiple sclerosis is likely misdiagnosed, are actually other quite common conditions. For example, migraine, functional neurological disorders, nonspecific white matter lesions on MRI that you might see in people with vascular disease, for example. While it is helpful to emphasize the rare or the less common conditions that can mimic multiple sclerosis, all neurologists should be vigilant to the fact that common neurological disorders can also be included in the differential diagnosis of MS.

Fred Lublin, MD: This has become something that developed more concern about it, and when we were doing the 2017 criteria, we spent a reasonable amount of time talking about this concern of a misdiagnosis. And there have been a number of good papers on this over the last several years, highlighting what you said, Wallace: that there were common things, migraine and vascular disease, and psychiatric disease, and such like that. One of the big drivers of the error was misapplying the MRI criteria. They’ve been laid out, for example, that a periventricular lesion ought to touch the ventricle and the juxtacortical lesions should touch the cortex if not include it entirely, and things like that. Not because MS can’t produce lesions that don’t do that, but because the basis of the criteria was on lesions that did do that.

Patricia K. Coyle, MD: They also reemphasize the size: 3 mm or greater.

Fred Lublin, MD: Misdiagnosis was an issue, and then almost shockingly, when one looked at how the misdiagnosed patients were treated, they were treated quite aggressively with disease modifying therapies, which we’re all happy to use in patients with MS, but not so happy to see them used in patients who don’t have MS.

Sven Meuth, MD, PhD: Since we have the new definition of the criteria with all these items you already mentioned, do you still think that we have a high percentage of misdiagnosed patients?

Fred Lublin, MD: I suspect we do. I haven’t seen it updated in the last year, but part of the problem is that when you write these criteria, we’ve made them easier and easier to do. The downside of easier criteria are that they can be more easily misapplied. The statistics on the positive predictive value, the negative predictive value, and the accuracy all come from highly qualified MS centers. But our intention for the criteria is that they be used by all clinicians who are going to be diagnosing multiple sclerosis, and that’s where the problem comes in. Even more than clinically, I think the MRI misinterpretation is most troublesome.