The Epstein-Barr virus causes multiple sclerosis. So what?

The recent “Epstein-Barr virus causes multiple sclerosis study” is very cool and a model for future epidemiological studies. First, it very strongly suggests that the Epstein-Barr virus (EBV) is behind at least a majority of multiple sclerosis (MS) cases in the US, which is a big deal for a condition which has long had mysterious origins. Also, it proved a causative effect with one very common infection and one relatively rare neurological condition, which is difficult on both counts (for details on how they did this, check footnote [1]).

So, I’d like to applaud this study. I’d also like to ask, “So what?”

I don’t mean to be rude. It really is a cool study, and maybe it doesn’t need a follow-up. Maybe it can just stand on its own, like a cool prog-rock album.

Top Google result for "cool prog-rock album".

But if we did want to follow it up, we’d probably want to follow it up along one of these lines:

1. Changing how we think about preventing MS

2. Changing how we think about the importance of preventing EBV

3. Changing how we treat MS

It’s difficult to know how we’d start with any of these. Let’s start with the top of the list.

This study firmly established that EBV is a necessary condition for at least most MS. Or, in other words, while it’s unlikely you’ll get MS at all, it’s very unlikely you’ll get MS without getting antibodies to EBV first.

This isn’t entirely surprising. Not only had these authors done a suggestive preliminary study with the same data in 2010, we’ve long known that infectious mononucleosis, a manifestation of EBV, is a major risk factor for MS.

More importantly, however, we don’t know the rest of the factors that contribute to MS. 90% of adults in the US have had EBV, while only 0.3% of adults have MS. There must be other factors that make those 0.3% get MS while the other 89.7% do not, but it’s unclear what they are. 

So, the only way we could really change how we prevent MS is by doing something like preventing infection from EBV. While this is a tall order, given the failures we’ve had in vaccinating against EBV so far, we might be able to argue that we now have a stronger reason than before to look for the vaccine.

I say might because we already knew that EBV was bad. Not only did we think it might contribute to MS, there’s also evidence linking it to over 200,000 cases of cancer each year, mononucleosis, chronic fatigue syndrome, and some rare autoimmune conditions. So, now we know for sure that EBV is really bad, whereas before we just knew it was quite bad. That may make a greater economic incentive to develop a vaccine, but it’s hard to say by how much.

"Now I was on board for $80 billion to bomb weddings in Pakistan, but $500k for an EBV vaccine will bankrupt us! We can't afford that!"

So that leaves us with the last point, that this study might change how we treat MS. This is what the study itself seems to point towards, when it points out that anti-CD20 monoclonal antibodies are already known as effective treatments for MS and that they deplete circulating B cells, which is where EBV hides. They also suggest that EBV antivirals might be a treatment for MS.

Here’s where things get a little dicey, though. The authors overstep when they say that anti-CD20 monoclonal antibodies are effective treatments for MS. They’re reasonably effective treatments for a certain type of MS, relapsing remitting MS (RRMS). Now, around 75% of people with MS have RRMS, so it is the most common form of MS by far. But, once we start drilling into the other forms of MS and the times when these treatments don’t work, we can get a better sense of the issue with coming up with better forms of treatment.

There’s always a clash between our neat model of how a disease works and how it actually works in the body. In the EBV-MS paper and its companion commentary, the authors give a neat model of MS that goes like the following: EBV viral protein sequences look like some of the protein sequences of myelin, the insulating sheath that covers neurons. When B cells learn to attack EBV, they occasionally also accidentally learn to attack myelin sheaths. As neurons live in the brian, B cells also become adept at crossing over the blood-brain barrier to attack these myelin sheaths, and then the neurons themselves.

This model has some experimental evidence. But, if you just follow this model, you would be left with the misleading impression that there’s a one way street: EBV infection leads to B cell sensitization leads to myelin sheath destruction leads to neuron death. This is not the case at all. Not only do the vast majority of EBV infections not lead to MS, but the majority of MS cases don’t end up in this one way street sort of scenario. The majority of MS cases are, as mentioned, relapsing remitting: they spend most of their time feeling ok, and then, a few times a year, they feel terrible, as their immune system decides to attack their myelin sheaths again.  Only a minority of cases get progressively worse over time, which is called progressive MS.

This is the opposite of what you would expect based on the model outlined above. More importantly, when we get down to the essential issue of “how do we treat MS”, the only people who are amenable to treatments that deplete B cells are the RRMS people, which is, again, the opposite of what you’d expect. Even those treatments don’t work perfectly, though. They can halve the number of MS attacks, but they can’t stop them completely.

For people with progressive MS, there are currently no great treatments, although I have reasons to believe my company’s lead drug might be a good one [2]. This doesn’t fit in with the model of MS outlined in the paper at all, and also makes you wonder about why EBV leads to RRMS in most people and progressive MS in others, assuming that it is behind both kinds.

So, once again, we have to ask: now what? In the past, epidemiological studies have led us to better water sanitation, smoking restrictions, and pesticide bans. Those were studies with some teeth to them. But for this study, we just have to look at it and say, “Neat”, then put it out of our minds. It’s a pity.

[1] In America, around 90% of people have had EBV by the time they reach adulthood. If this seems surprising to you, keep in mind that if you’ve ever been in close contact with someone who developed mono, you almost certainly had EBV. Meanwhile, only 0.3% of adults have MS.

So, the challenge of this study was to prove that this super common virus is a necessary, but not sufficient condition for developing MS. But how do you do that if almost everyone gets EBV eventually?

Well, the first way they did it was to look at a bunch of blood serum samples collected over decades of an adult’s life in the hopes that it would capture before and after they got EBV, in the hopes that it would reveal MS developing after EBV. In this study, those samples were collected from military personnel, who normally get their blood samples at around age 18. Lo and behold, of the 801 military MS cases with serum samples, only 1 developed MS before developing EBV. The rest developed MS 0-10 years after getting EBV.

So, that’s already pretty suggestive. But, it leaves open the possibility that people just tend to get infected with EBV when they’re younger, and get MS when they’re older. So, the authors’ second piece of evidence was based on the percentage of people who get exposed to EBV versus the percentage of people who develop antibodies to it. They found that only 57% of the non-MS controls who were exposed to EBV developed antibodies (seroconverted), while 97% of the MS cases developed antibodies.

Given that the authors have a good argument as to why antibodies to EBV are the mechanism by which EBV causes MS (which I talk about elsewhere), this seems like some solid proof.

[2] This is a big claim, but I have evidence!

My company’s lead drug is a safer version of cyclosporine. Cyclosporine is a potent immunosuppressant that’s notoriously hard to dose correctly, and overdoses are extremely common.

Multiple sclerosis has long been suspected to be an immune-related disease. As such, pretty much every kind of immunosuppressant has been thrown at MS at one point or another. As such, cyclosporine has been thrown at every kind of MS at one point or another.

Unfortunately, cyclosporine does not work that well in RRMS. It works ok, but not as well as the other drugs and it’s a lot more dangerous. However, what’s interesting is that when they tried it in progressive MS, it actually worked ok, halving the average annual increase in disability score (EDSS). This never ended up being put into clinical practice, though, because of how common severe side effects are for patients on cyclosporine.

My contention is that my version of cyclosporine is way safer than the old version of cyclosporine tested. So, you can get all the benefits with way fewer of the side effects. I’m currently raising money to test this so…wish me luck!