The case for antivirals targeting lytic EBV infection in MS

Can we afford not to test anti-EBV drugs in MS when the case for doing so is so overwhelming?

An open letter

Dear pharmaceutical or funding agency executives

This MS-Selfie Newsletter is an open appeal to any pharmaceutical or funding agency executives who oversee any anti-viral development or funding programmes. Please don’t forget multiple sclerosis (MS). MS is caused by EBV and the evidence that intermittent or continuous EBV lytic infection is driving the disease is now overwhelming. We need investment in order to test the hypothesis that MS can be treated with antivirals.

Thank you.

Gavin Giovannoni

Photo by Jan Kahánek on Unsplash

The Evidence

Epstein-Barr virus is almost certainly the cause of MS. EBV is necessary, but not sufficient, for someone to develop MS. People who are EBV negative are protected from getting MS; their risk is close to zero. This fact is critical in accepting EBV as the most likely cause of MS. 

People who are EBV-negative and go on to develop MS in the future always seroconvert to becoming EBV-positive before developing MS. The temporal sequence of needing to be exposed to EBV before developing MS is one of the criteria used in proving causation. All of these factors are why there is an urgent need for an EBV vaccine programme to try and prevent MS. I have discussed this before in the MS-Selfie Newsletter ‘Will an EBV vaccine work?’ (12-Jul-2022).  

However, all of these epidemiological observations don’t explain how EBV causes MS. At present, there are several competing theories. The most prominent is the molecular mimicry hypothesis. Molecular mimicry implies that EBV proteins/antigens generate cross-reactive immune responses to self-proteins/antigens that trigger an immune attack on self, which in the case of MS is the myelin-axon unit. Several autoantigens have been shown to have cross-reactive immune responses between self-epitopes (autoantigens) and EBV proteins. However, many of these antigens are not only expressed in the central nervous system but elsewhere, which makes it hard to link the antigens to MS pathology. If these antigens were driving MS, wouldn’t you expect MS-related inflammation to be a multi-system disease?

Another theory is that EBV infection causes dysregulation of the immune system, which is then primed to develop autoimmunity in the future; hence EBV is necessary but not sufficient to cause MS. An argument to support this is the observation that EBV is not only linked to MS but several other autoimmune diseases as well; for example, systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), Sjogren’s syndrome, primary biliary cirrhosis (PBC), inflammatory bowel disease (IBD) and potentially several other autoimmune diseases.

The molecular mimicry and immune dysregulation theories imply that EBV’s role in causing MS is like a hit-and-run accident and that once autoimmunity is primed, targeting EBV will have no further impact on the development or course of MS, which is then driven by other factors unrelated to EBV, for example, genetic factors. This position is not necessarily correct in that EBV replication, what I refer to as latent-lytic cycling, will intermittently or continuously expose the immune system to EBV antigens, which may be necessary to drive both molecular mimicry and/or immune dysregulation. This is why I don’t think we should assume targeting EBV, with either antivirals or immunotherapies, after someone has developed MS won’t work as a therapeutic strategy. I hypothesise this is because all of our licensed disease-modifying therapies (DMTs) in MS, with one exception, teriflunomide, reduce memory B-cell numbers or stop these cells from trafficking into the CNS. Teriflunomide, the exception to the rule, has pan-antiviral activity and may work as a DMT in MS by inhibiting lytic EBV infection. 

The memory B-cell is critical because it is the cell where EBV lives in its dormant or latent phase. EBV, and the other herpes viruses, are very clever viruses from a molecular perspective. They use a molecular programme that allows them to hibernate in B-cells, thereby hiding away from the immune system. This is why herpes viruses are so good at persisting. They then intermittently wake themselves up and trigger a so-called lytic infection cycle during which they make many copies of themselves, which allows the virus to potentially mutate, and spread (saliva, genital fluids, blood), thereby ensuring that the virus survives from an evolutionary perspective. 

An EBV lytic infection cycle allows the immune system to see the virus, which then amounts an immune response to the virus. The immune response to EBV infections involves innate (hard-wired) and adaptive (memory) responses. Stimulating old memory responses and developing new memory responses to other EBV antigens and epitopes is like the booster response to a vaccine. We have evidence that people with active MS have peripheral and CNS activation of the innate immune system. You can measure innate immune activation using biomarkers, i.e. gene signatures in cells in the peripheral blood and/or brain, or by measuring metabolites and cytokines of innate immune cells in the blood, urine and spinal fluid. These are raised in pwMS who have active MS. One particular signature is the so-called type 1 interferon response in people with active MS and SLE. This response typically occurs in response to viral infections. Could this type 1 interferon response in active MS be a biomarker of EBV lytic infection? 

In parallel to innate immune changes, you can measure antibody (B-cells) and T-cell responses to EBV in people with MS. Surprise, surprise, pwMS have much higher levels and much broader antibody responses to both latent and lytic EBV proteins. Similarly, the T-cells show the same pattern. Instead of looking at antibody binding to EBV proteins in an immunoassay, for example, an ELISA, you look at the so-called molecular fingerprint of the T-cell receptors. Using very clever computer algorithms, which control for the MHC (major histocompatibility complex), pwMS also have a much broader and deeper T-cell response to EBV.  This tells us that the immune systems of pwMS are seeing EBV a lot and are forcing it to make enhanced innate and memory responses to control the virus. 

In summary, the immunological signature from pwMS tells us that EBV is almost certainly cycling through the latent and lytic phases more than compared to what happens in normal subjects and other control subjects. But do we have virological evidence of this happening? Yes, we do. PwMS are much more likely to shed EBV in their saliva than healthy controls. Salivary shedding is a marker of lytic viral activity in the salivary glands. Increased EBV replication in the salivary glands may be one reason pwMS have enlarged deep cervical lymph nodes. Don’t you think looking in these lymph nodes for evidence of EBV lytic infection would be interesting? 

Outside the salivary glands, the evidence for latent-lytic cycling is more controversial. We don’t necessarily find higher EBV viral loads in the peripheral blood of pwMS. However, when pwMS are treated with AHSCT (autologous haemopoietic stem cell transplantation), EBV reactivation occurs much more frequently than in people with other diseases. In this situation, the virus is found in the peripheral blood. People with MS can’t control the virus when they are severely immunosuppressed. Evidence from studies of the brains of people dying from MS is more controversial. Some studies demonstrate EBV in the brains of pwMS, and others have not been able to reproduce the findings. The methods used to detect EBV generally rely on finding latent infection, and some of these methods may give false positive and false negative results. Sadly this controversy has muddied the waters and is probably the main reason EBV-MS research has not been funded very much in the last two decades. 

Lessons from natalizumab-treated pwMS

How many swallows does it take to make a summer? Natalizumab is a selective adhesion molecule blocker that prevents immune cells from trafficking into the brain. Natalizumab is one of our most effective DMTs, but it comes with a price. By stopping immune cell trafficking into the central nervous system, it prevents CNS immune surveillance, which creates an immunosuppressed compartment that allows viruses to replicate relatively unchecked, mutate and cause CNS infections. This is why PML (progressive multifocal leukoencephalopathy) develops in people who are JC virus positive on natalizumab at a relatively high rate. 

Natalizumab-associated PML has taught us a lot about viruses and CNS infections. A higher antibody index or titre, or a rising titre, are major risk factors for developing PML. This is likely telling us that the JC virus is active and replicating, and by doing this, it is boosting the immune response to its viral proteins. I suspect the compartment the JC virus is replicating in is the CNS. I say this because the strain of the JC virus that causes PML is a mutant strain that acquires the ability to infect glial cells. This process driving the development of a mutant strain is based on evolutionary principles. Mutant strains out-compete non-mutant strains because of their ability to infect glial cells and survive. It is all about survival of the fittest. Mutant PML-causing JCV strains need glial cells to drive their evolution. Could the same process be happening with EBV with the CNS? I suspect yes, but the evidence that EBV infects neurons or glial cells is currently poor. However, this may change over time. I am saying this because we must not accept the current dogma that EBV doesn’t infect other cell types outside B-cells and epithelial cells. We need to look deeper into what is happening in the CNS.

You may be aware two patients who died of natalizumab rebound after the therapy was stopped have been shown to have lytic EBV infection within their brains associated with a massive immune response against the virus and evidence of bystander damage (demyelination and axonal transection). These studies were done with antibodies that detect lytic EBV infection and are unlikely to be false positive. So unlike other people in the field, I refuse to dismiss these cases as unusual. Could all natalizumab rebound be in response to CNS EBV lytic infection? If this is the case, then natalizumab rebound should be reclassified as IRIS (immune reconstitution inflammatory syndrome), not focal MS disease activity. Please be aware that IRIS was initially described in AIDS patients who developed brisk immune responses to opportunistic infections in the CNS when put on antiretroviral therapy; PML IRIS was the prototype. 

The billion-dollar question is how did EBV get into the brains of these two pwMS on natalizumab? Was EBV there initially, or did B-cells carry it into the CNS when natalizumab was stopped and allowed to wash out? In support of the former is the increase in T-cells responses to EBV latent and lytic that has been described in pwMS on natalizumab. This indicates that pwMS on natalizumab are boosting their immune responses to EBV, but the cells are trapped in the periphery, waiting to get into the target organ, in this case, the brain and spinal cord. We tried to test this hypothesis a few years ago using antibodies. We did not show increased antibody titres to EBNA-1 and VCA (viral capsid protein), two commonly used diagnostic EBV assays, in pwMS on natalizumab. Our study was small and now needs repeating with more subjects and a wider panel of latent and lytic EBV antigens. 

Anti-CD20 is the most effective treatment to prevent natalizumab rebound. And as anti-CD20 targets mainly B-cells, I and others have assumed that EBV-infected B-cells carried the virus into the CNS. This is called the ‘Trojan Horse hypothesis’. But why would EBV-infected B-cells suddenly reactivate their virus to cause massive lytic infection? Is there some coordinated signal within the MS brain and spinal cord that does this? Natalizumab rebound lesions are distributed across the CNS and appear active (Gd-enhancing) in quite a tight time window. We know the lesions in these two patients had CD4+ve and CD8+ve T-cells, but the latter predominated. Could a population of EBV-directed T-cells be responsible for natalizumab-rebound, and could peripheral anti-CD20 therapies target these cells?

CD20-positive T-cells

A small population of T-cells express CD20, which they are likely to acquire from B-cells by a process called trogocytosis. Trogocytosis is part of an emerging theme of cell-cell interactions within and between species, and it is relevant to host-pathogen interactions in many different contexts. Trogocytosis is when one cell physically extracts and ingests “bites” of cellular material from another cell.

Trogocytosis - T-cells acquire CD20 on their surface after interacting with B-cells, presumably via close interaction during the process of B-cell antigen-presentation. The T-cells expressing CD20 are likely to be pathogenic (anti-EBV) and susceptible to depletion by anti-CD20 therapies, e.g. with ocrelizumab.

T-cells don’t synthesise their own CD20 surface molecules. The T-cells that express CD20 must have recently had close contact with B-cells. We assume this B-cell T-cell interaction was recent and may have occurred because B-cells present antigens to T-cells. There is evidence that these T-cells expressing low levels of CD20 are activated effector T cells primed for battle. As anti-CD20 therapies deplete these cells, a population of CD20-positive T-cells could be responsible for natalizumab rebound and not necessarily B-cells. 

This data shows that natalizumab is not impacting EBV biology (latent-lytic cycling) and that the EBV replication may occur with the CNS. Many more studies need to be done to explore these hypotheses further in pwMS on natalizumab. Natalizumab is the one DMT that keeps giving; giving us opportunities and clues to the pathogenesis of MS and providing us with the right environment to test fascinating hypotheses. 

So when I am asked my opinion on the evidence that EBV is in the brains of pwMS, I sit on the fence and say probably yes. On balance, the evidence supports the CNS in pwMS being infected with EBV.  Please don’t forget that the MS brain is stuffed full of B-cells, including memory B-cells and as about 1 in a million lymphocytes are EBV infected, there will be EBV in the brains of people with MS. This is an important point, simply because any therapeutic strategy targeting EBV infected B-cells needs to be CNS penetrant to clear this reservoir of the virus. This may explain why CD20 therapies are ineffective in targeting CNS resident B-cells on MS in that we don’t get higher enough therapeutic anti-CD20 antibodies into the CNS.

Another clue to excess latent-lytic cycling in pwMS is the observation that cells from the blood of pwMS are more likely to make spontaneous EBV-associated lymphoblastoid cell lines or SLCLs. This occurs in about 60% of pwMS and only 10% of normal controls. In addition to this numeric difference, SLCLs derived from pwMS differ from those derived from normal controls. EBV latency is dysregulated in MS SLCLs with increased EBV lytic gene expression. The latter has been described as being particularly prominent in pwMS with active disease. Is this another clue, albeit a clue from cells cultured outside the body, that pwMS have increased latent-lytic cycling?

Finally, we should not forget infectious mononucleosis (IM) or symptomatic primary EBV infection. IM syndrome occurs due to an aberrant or dysregulated immune response to EBV infection. People with IM control EBV less well. As IM is a risk factor for developing MS, this tells us that people destined to develop MS in the future have a problem controlling EBV from their first exposure to the virus. 

All these pieces of the jigsaw puzzle are crucial because we need to make a case for testing antiviral drugs targeting lytic EBV infection in pwMS. This is why we are currently testing FamV or famciclovir in pwMS. We are also trying to nudge various pharmaceutical companies with EBV antiviral drugs to test them in people with MS. The circumstantial evidence that latent-lytic infection is driving MS is very compelling. Unless we do the necessary experiment of testing antivirals in MS, we will not know the answer. There is an experiment waiting to happen.

Articles of interest

Soldan et al. Unstable EBV latency drives inflammation in multiple sclerosis patient derived spontaneous B cells. Preprint is available at Res Sq.

Epidemiological studies have demonstrated that Epstein-Barr virus (EBV) is a known etiologic risk factor, and perhaps prerequisite, for the development of MS. EBV establishes life-long latent infection in a subpopulation of memory B cells. Although the role of memory B cells in the pathobiology of MS is well established, studies characterizing EBV-associated mechanisms of B cell inflammation and disease pathogenesis in EBV (+) B cells from MS patients are limited. Accordingly, we analyzed spontaneous lymphoblastoid cell lines (SLCLs) from multiple sclerosis patients and healthy controls to study host-virus interactions in B cells, in the context of an individual’s endogenous EBV. We identify differences in EBV gene expression and regulation of both viral and cellular genes in SLCLs. Our data suggest that EBV latency is dysregulated in MS SLCLs with increased lytic gene expression observed in MS patient B cells, especially those generated from samples obtained during “active” disease. Moreover, we show increased inflammatory gene expression and cytokine production in MS patient SLCLs and demonstrate that tenofovir alafenamide, an antiviral that targets EBV replication, decreases EBV viral loads, EBV lytic gene expression, and EBV-mediated inflammation in both SLCLs and in a mixed lymphocyte assay. Collectively, these data suggest that dysregulation of EBV latency in MS drives a pro-inflammatory, pathogenic phenotype in memory B cells and that this response can be attenuated by suppressing EBV lytic activation. This study provides further support for the development of antiviral agents that target EBV-infection for use in MS.

Schneider-Hohendorf et al. Broader Epstein–Barr virus–specific T cell receptor repertoire in patients with multiple sclerosis. J Exp Med. 2022 Nov 7; 219(11): e20220650.

Epstein–Barr virus (EBV) infection precedes multiple sclerosis (MS) pathology and cross-reactive antibodies might link EBV infection to CNS autoimmunity. As an altered anti-EBV T cell reaction was suggested in MS, we queried peripheral blood T cell receptor β chain (TCRβ) repertoires of 1,395 MS patients, 887 controls, and 35 monozygotic, MS-discordant twin pairs for multimer-confirmed, viral antigen–specific TCRβ sequences. We detected more MHC-I–restricted EBV-specific TCRβ sequences in MS patients. Differences in genetics or upbringing could be excluded by validation in monozygotic twin pairs discordant for MS. Anti–VLA-4 treatment amplified this observation, while interferon β– or anti-CD20 treatment did not modulate EBV-specific T cell occurrence. In healthy individuals, EBV-specific CD8+ T cells were of an effector-memory phenotype in peripheral blood and cerebrospinal fluid. In MS patients, cerebrospinal fluid also contained EBV-specific central-memory CD8+ T cells, suggesting recent priming. Therefore, MS is not only preceded by EBV infection, but also associated with broader EBV-specific TCR repertoires, consistent with an ongoing anti-EBV immune reaction in MS.Subscriptions and donations

Paid subscriptions to MS-Selfie are being used to administer the Newsletter and associated MS-Selfie microsite currently in development. At the request of several readers, I have now added the option of making a one-off donation. To keep this initiative open to all readers, I would appreciate it if those who can afford a subscription to please subscribe. For active paying subscribers, thank you; your contribution is much appreciated.

Donations to MS-Selfie

Prof G’s backstory and CV

Prof G’s MS Blog Archive

Conflicts of Interest

Twitter / LinkedIn / Medium

General Disclaimer: Please note that the opinions expressed here are those of Professor Giovannoni and do not necessarily reflect the positions of Barts and The London School of Medicine and Dentistry nor Barts Health NHS Trust. The advice is intended as general and should not be interpreted as personal clinical advice. If you have problems, please tell your healthcare professional, who will be able to help you.