On being EBV-negative and having MS


Barts-MS rose-tinted-odometer: ★★★★★

I have been diagnosed with MS and I am Epstein Barr Virus (EBV) negative. Therefore, EBV is not the cause of MS. Correct? I wish it was that simple. 

Firstly, no laboratory test is 100% sensitive and specific. In other words, some people who have negative standard EBV serology may still have the virus, i.e. a false negative result, and some people who have a positive result may not have the virus a so-called false-positive result. A very sensitive assay is one that limits the number of false-negative results, i.e. gets the result correct almost all the time. A very specific assay limits the number of false-positive results and excludes those with infection or disease. Do these terms give you a sense of deja vu? The COVID-19 lab tests have made them part of the public lexicon. 

In this study below we checked out two commercial EBV serology assays and as expected they were not perfect. So yes you can be EBV-antibody negative and still have EBV.

Dobson et al. Comparison of two commercial ELISA systems for evaluating anti-EBNA1 IgG titers. J Med Virol. 2013 Jan;85(1):128-31.

High IgG titers against the Epstein-Barr virus nuclear antigen, EBNA-1, have been strongly correlated with the risk of developing multiple sclerosis. ELISAs are used frequently to measure EBNA-1 titers, however concerns remain regarding the accuracy of results. Ordering absolute results into rank quintiles for analysis may be preferable. Using 120 serum samples, two commercially available ELISAs (produced by DiaSorin and VirionSerion) were compared, both in terms of absolute results and rank quintiles. The positive predictive value of the VirionSerion ELISA was 99.1% when compared to the DiaSorin ELISA, however, the negative predictive value was 64.3%. Sensitivity and specificity were acceptable at 95.5% and 90.0%, respectively. There was poor correlation between absolute results, R(2) = 0.49; and the kappa coefficient for rank quintiles was low at 0.23. Although sensitivity and specificity appear adequate, the poor negative predictive value and kappa coefficient are of major concern. Care must be taken when selecting assays for experimental use.

In a meta-analysis of EBV and MS, we showed that when you use the immunofluorescence assay, which although being very labour intensive is considered the gold standard for diagnosing MS 100% of pwMS were EBV-positive. Interesting? Then on the flip side being EBV immunofluorescence negative was the most powerful predictor of not getting MS. These and other findings are part of the evidence that convinced me decades ago that EBV is the cause of MS. 

Pakpoor et al. The risk of developing multiple sclerosis in individuals seronegative for Epstein-Barr virus: a meta-analysis. Mult Scler. 2013 Feb;19(2):162-6. 

Background: Epstein-Barr virus (EBV) infection is widely considered to be a risk factor for multiple sclerosis (MS). A previous meta-analysis estimated an odds ratio (OR) for MS in individuals seronegative for EBV of 0.06. Given the potential importance of this finding, we aimed to establish a more precise OR for adult and paediatric onset MS in EBV seronegative individuals.

Methods: PubMed and EMBASE searches were undertaken to identify studies investigating the association between MS and EBV. Twenty-two adult and three paediatric studies were included. ORs were calculated using a fixed effects model. A sub-group analysis based on the method of EBV detection was performed.

Results: The OR for developing adult MS in EBV seronegatives was 0.18 (95% confidence interval (CI) 0.13-0.26)) and for paediatric MS was 0.18 (95% CI 0.11-0.30). Sub-group analysis on EBV detection method showed that studies which used immunofluoresence generated an OR=0.07 (95% CI 0.03-0.16); for those that used enzyme-linked immunosorbent assay (ELISA) OR=0.33 (95% CI 0.22-0.50) and for studies which used ELISA and immunofluoresence OR=0.00 (95% CI 0-0.43).

Conclusion: The sensitivity and specificity of the assay used to measure EBV antibody titres have an influence on the association between MS and EBV. Looking at studies where two independent methods are used and therefore are likely to be the most robust, EBV appears to be present in 100% of MS patients. This has implications for future studies of EBV in MS. MS patients without EBV infection, if they truly exist, should be studied in more detail.

Now, what about the diagnosis of MS? 

In the study below approximately 1 in 5 people diagnosed with MS don’t have MS. This figure is much higher than in previous studies. I usually quote a large Danish post-mortem study that suggests only 1 in 20 people with MS (pwMS) are misdiagnosed. It is important to realise that there is no one test that can be done to diagnose MS. MS is diagnosed by combining a set of clinical and MRI findings, electric or neurophysiological investigations and laboratory tests. If these tests fulfil a set of so-called MS diagnostic criteria the Healthcare professional (HCP) or neurologist makes a diagnosis of MS.

The underlying principle of making a diagnosis of MS is showing dissemination of lesions in space and time and excluding other possible diagnoses that can mimic MS. The diagnostic criteria have evolved over time from being based purely on clinical finding to the more recent criteria that include evoked potentials, spinal fluid analysis and MRI to help confirm dissemination in time and space.

Dissemination in time means two attacks or MS lesions occurring at least 30 days apart. Dissemination in space means lesions occurring in different locations, for example, the optic nerve and spinal cord.

The electrical or neurophysiological tests are called evoked potential (EPs) and test electrical conduction in a particular neuronal pathway. They can be useful to show the effects of lesions in pathways that are not evident on neurological examination or seen on MRI. The EPs can also show slow electrical conduction which is one of the hallmarks of diseases that affect myelin, the insulation of nerves that are responsible for speeding up electrical conduction.

The laboratory tests are typically done to exclude other diseases that can mimic MS. One test that is useful in helping to make the diagnosis of MS is examining the spinal fluid for the presence of oligoclonal immunoglobulin G or IgG bands (OCBs), which are the fingerprint of a specific type of immune activation within the central nervous system (CNS). 

The OCB fingerprint is relatively specific for the diagnosis of MS in the correct clinical context. Please note OCBs can are found in infections of the nervous system and other autoimmune diseases, therefore, the presence of OCBs are not diagnostic on their own.

Please note being EBV seropositive is currently not part of the diagnosis of MS so you can be diagnosed with MS and still be EBV-negative. What we don’t know is whether or not EBV-negative MS is biological MS, i.e. the same disease as EBV-positive MS. This is something I have been wanting to study for a long time. 

I have spent some time explaining this all to you as we neurologists get the diagnosis wrong approximately 5% of the time and if this paper below is correct maybe in even a higher number of patients. In other words, at least 1 in 20 people who have a diagnosis of MS in life don’t have MS when their brains are studied at postmortem.

Why is getting the correct diagnosis of MS so important? Firstly, some of the treatments for MS have life-threatening complications; you don’t want to expose people without MS to these complications. Some diseases that mimic MS can be made worse by MS DMTs. This latter is particularly relevant for NMO or neuromyelitis optic. Patients with NMO misdiagnosed as having MS get worse on many of the MS DMTs. Finally, a diagnosis of MS has many psychological, social, financial and economic implications for people. Just having a diagnosis of MS, even if you turn out to have benign MS in the future, has implications for the person concerned. For example, it may affect your life choices and may impact your ability to get insurance cover to name to obvious examples. I would, therefore, advise you to make sure you have MS and not an MS mimic. 

The most common MS mimics:

  1. Cerebrovascular disease
  2. Acute disseminated encephalomyelitis or ADEM
  3. Neuromyelitis optica or NMO
  4. Behcet’s syndrome
  5. Migraine
  6. Sarcoidosis
  7. SLE or systemic lupus erythematosus
  8. Antiphospholipid antibody syndrome
  9. Leukodystrophies

The evolving definition of MS based on diagnostic criteria:

Clinical criteria only:

  1. Schumacher, et al. Problems of Experimental Trials of Therapy in Multiple Sclerosis: Report by the Panel on the Evaluation of Experimental Trials of Therapy in Multiple Sclerosis. Ann N Y Acad Sci 1965;122:552-68.

Clinical, EPs and CSF analysis:

  1. Poser, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 1983;13:227-31.

Clinical, EPs, CSF analysis and MRI:

  1. McDonald, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121-7.
  2. Polman, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Criteria”. Ann Neurol 2005;58:840-6.
  3. Polman, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol. 2011;69:292-302.
  4. Thompson et al. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. Lancet Neurol. 2018 Feb;17(2):162-173.

What the evolving definition of MS tells us is that the diagnosis and hence the disease MS as we currently define it is a moving target. In other words, someone 10 years ago who do not fulfil the diagnosis of MS, i.e. didn’t have the disease, maybe diagnosed today as having MS. How can this be? This is why I would prefer to use a biological definition of MS. Yes, I am currently working on a paper that sets out the principles for redefining MS as a biological disease.

So what then do I do at the moment if I have MS and I am EBV negative? Until we prove EBV is the cause of MS and include EBV in the diagnosis I don’t think knowing if you are EBV positive or negative makes any difference to the diagnosis of MS and its management.

However, I would like to challenge the status quo. Can we really continue to ignore the evidence linking EBV to MS? Don’t we owe it the next generation of pwMS to act on this information ASAP? Is anyone prepared to donate several million dollars to a consortium to EBV treatment and prevention trials in MS, i.e. the Charcot Project

Kaisey et al. Incidence of multiple sclerosis misdiagnosis in referrals to two academic centers. Mult Scler Relat Disord. 2019 May;30:51-56.

BACKGROUND: Multiple Sclerosis (MS) specialists routinely evaluate misdiagnosed patients, or patients incorrectly assigned a diagnosis of MS. Misdiagnosis has significant implications for patient morbidity and healthcare costs, yet its contemporary incidence is unknown. We examined the incidence of MS misdiagnosis in new patients referred to two academic MS referral centers, their most common alternate diagnoses, and factors associated with misdiagnosis.

METHODS: Demographic data, comorbidities, neurological examination findings, radiographic and laboratory results, a determination of 2010 McDonald Criteria fulfillment, and final diagnoses were collected from all new patient evaluations completed at the Cedars-Sinai Medical Center and the University of California, Los Angeles MS clinics over twelve months.

RESULTS: Of the 241 new patients referred with an established diagnosis of MS, 17% at Cedars-Sinai and 19% at UCLA were identified as having been misdiagnosed. The most common alternative diagnoses were migraine (16%), radiologically isolated syndrome (9%), spondylopathy (7%), and neuropathy (7%). Clinical syndromes and radiographic findings atypical for MS were both associated with misdiagnosis. The misdiagnosed group received approximately 110 patient-years of unnecessary MS disease-modifying therapy.

CONCLUSION: MS misdiagnosis is common; in our combined cohort, almost 1 in 5 patients who carried an established diagnosis of MS did not fulfill contemporary McDonald Criteria and had a more likely alternate diagnosis.

CoI: multiple

Twitter: @gavinGiovannoni                                              Medium: @gavin_24211

About the author

Prof G

Professor of Neurology, Barts & The London. MS & Preventive Neurology thinker, blogger, runner, vegetable gardener, husband, father, cook and wine & food lover.


Leave a Reply to Frederick Gay Cancel reply

  • Thank you for writing this, i feel partially responsible due to my comments on your earlier post and do support the hypothesis and would love to be involved im any research in anyway possilbe.

  • Not using this question to diagnose anything but what would you like to incorporate to the MS diagnosis criteria? Seems people focus on D3 more than other potentially more useful biomarkers. Also it seems a lot of neuros forget the cornerstone of McDonald’s which is “eliminate other diseases”. Honestly, it feels like MS is the autism of neuros. See brain damage, can’t bother to look hard enough, MS! It doesn’t help that it is the most common neuro disease amongst younger demographics. Feels like a self-fulfilling prophecy doesn’t it?

  • Regarding the mis- and overdiagnosis of MS: I can’t help thinking there is an important mix of 1) findings on the importance of early treatment, 2) allowing patients to have medication that is covered by insurance or otherwise, and 3) (therefore) also relaxing diagnostic criteria.
    As you point out, the diagnosis of MS has lots of consequences (albeit not all so dramatic as dying from PML without having MS). I would like to add: it’s also very hard to shake off that diagnosis, especially if you’re not going for a second opinion right away.

  • Well to be fair to my neuro’s I think I had the full work out. Mri, repeated again with contrast and. LP. I had so many blood tests it hurt to pull the hand brake up on my car. Diagnosis took 18 months from “first” presenting symptom. I personally haven’t heard (amongst the ms forums and local groups that I am aware of) of someone been misdiagnosed with ms more how long it takes to get a diagnosis of ms when someone is suffering with symptoms and not knowing what is causing them.

    • We have a college who does post-mortems and at least 5% of the people do not have MS when the pathology is examined

      • What do they have? Is getting relapses a hallmark having definitive MS? and say for more experienced Neurologist, say Prof G. Would the misdiagnose rate be the same (something just can’t be avoided with current techniques?) Apologies for so many questions.

  • No mention of PCR and B cells in diagnosing EBV exposure. It would be the most sensitive and specific test for latent and lytic EBV I presume.

  • Don’t we owe it to the next generation? Absolutely….but then you were convinced decades ago. No doubt plenty of others were too. Didn’t we owe it to this one? All the research that goes on in the field of MS. Why are we still here decades later? I know the pharma conspiracy theories but it cannot be just that. Didn’t Michelle Obama’s mum have MS? I would of told my hubbie the first day in office to sort this crime out. I know I am probably terribly naive but I find this just unbelievable

    • I noticed the other paper mentioned on the same link Luis:
      Increased risk if MS following herpes zoster infection. 2011.

      May be the herpes zoster infection is the catalyst that triggers the first MS relapse?
      But the MS was there dormant already?

      • Could be anecdotal

        But i know some friends that got ms one after spinal surgery

        And the other after been bite by a bug

        My doc says that ms is triggers by multiple things not just one

        • When anything happens you can link it back to something, I fell of a bike and broke a tooth, it was the day after we had been to blackpool…Maybe Lancashire gives you broken teeth:-)

  • Thank you Prof G for this super informative post.

    Question – do we expect non-biological MS has a similar disease course as biological MS?

    • The Pakpoor et al paper you quote (? seronegative EBV/MS) in 2013, appropriately ends with the conclusion, ‘MS patients without EBV infection, if they truly exist, should be studied in more detail’.

      Have they? I can find no references on this.

      Also, may I point out for the second time, that this critical antibody study, did not control the validity of the combined tests to be truly negative, in patients known to have avoided infection with EBV.

      A further point, on the pathology. A series of PMs carried out on the brains of patients with no history of MS whatever, revealed classical MS plaques in up to 18%. ( see Colin Adams MS atlas). Wow! What a deceiver MS continues to be!

      • We can’t find them!

        We did try but of the 32 EBV negatives CIS/MS patients we identified 31 came back positive and one was borderline positive with more detailed serology. A German group had the same experience. So in short 100% of MS patients are EBV positive when you use state-of-the-art serology to look for the virus.

        • You did not make it clear that the ‘State of the art tests’ were controlled by known EBV negatives. When you work at this level strict controls are obligatory to confirm that the more sophisticated tests may be much more prone to false positives. This as you will be very well aware is not an excessively niggiling point!
          There is a distinct possibility that certain (MS associated) immune genotypes are significantly more susceptible to EBV. The new Covid variant illustrates the point. A variation on the viral receptor can render particular cells very much more sensitive to infection.
          It is quite possible that within some populations susceptibility to MS is closely linked to EBV susceptibility and this does not necessarily mean that EBV is the cause of MS.

        • I’m not disagreeing but if you take 32 EBV negatives who are non CIS/MS patients, what would their state-of the-art serology look like for the virus?
          Would it also show 100% EBV positive? If so, what does this tell us?

  • Are children with MS also EBV positive? You once mentioned that paediatric MS cases possibly were a problem for the EBV theory. Has this been studied since then?

    Is it that being EBV-negative offers protection against MS but EBV itself does not actively cause MS? The vast majority of people are EBV positive, but just a small proportion have MS

    Your team has tested MS patients and found that they are actually EBV-positive. What if EVERYONE is EBV-positive if the proper sensitive tests are done? What if EBV-negative people don’t exist at all?

    Do you have any figures for people who have MS and are never diagnosed? For example, from MS cases discovered accidentally post-mortem. I think the number of ‘mild’ MS cases that are never seen by a neurologist could be high.

    Finally, does the EBV theory help to explain why ‘autoimmune diseases’ tend to run in families?

  • “… define what is needed to fulfil dissemination in time and space of lesions in the CNS, and stress the need for no better explanation for the presentation”

    In other words, it is MS if it is not any other known thing.

    Good example of a circular definition here.

  • What needs to be looked more carefully is ADEM from coronaviruses. They may be more common than we think and they can mimic CIS.

By Prof G



Recent Posts

Recent Comments