Negative Data Matter Too: A Chat with Dan Levy about the Basic Science of Migraine

“I’d like people to know that it’s hard to do basic migraine research, that understanding this condition as a whole is hard, especially if you’re doing physiology. But this shouldn’t stop people from joining the field, which has a lot of things to offer someone who wants to study migraine from different perspectives. There are so many different ways to study this disease. That will make a lot of people happy in their science.”

Editor’s note: This is the seventh in a series of interviews and podcasts with Migraine Science Collaborative editorial board members.

Dan Levy, PhD, is an associate professor, and the principal investigator at the Headache Research Laboratory, Department of Anesthesia and Critical Care, Beth Israel Deaconess Medical Center, Harvard Medical School, US. His work focuses on the sensory innervation of the intracranial meninges and its role in headaches such as migraine headaches. In this interview, he chats with MSC executive editor Neil Andrews to discuss his research, why negative data are important, and some of the challenges and opportunities in the broader migraine field, among other topics. This interview has been edited for clarity and length.

What was your path to becoming a migraine researcher?

During my master’s program I began to study neuropathic pain, and I continued to do so during my PhD. Then, at some point, I was looking for – I wouldn’t say an escape from neuropathic pain – but for a different niche in pain research. I wanted to learn electrophysiology, and I got interested in headache, although not specifically migraine at the time.

My PhD advisor basically told me that there were so many people in the neuropathic pain field and that I should try to find a niche that made sense to me. At that point, we could count maybe on one hand the number of people, at least in the US, who studied headache and migraine preclinically. And together with my advisor, we said, this is a good field for me to enter at this stage of my career.

Dan Levy

I sent a couple of email requests to different people, looking for postdoc positions, and then met with Andy Strassman, who offered me a postdoc position and to teach me electrophysiology. He was the first person to publish electrophysiological work on the dural afferents, and I was very excited to get into this field. This was my entry point into the world of migraine pain.

What is the overall goal of your lab?

My lab is primarily interested in the physiology of the meningeal sensory innervation – how the system works and contributes to headache pain, not only migraine pain but also to other types of headache, for example, posttraumatic headache. The overall goal is to figure out how the system becomes activated, sensitized, and then drives migraine pain and headache. Most of the work that we’ve been doing looks at the inflammation and related processes that drive the system.

Your lab uses animal models. Tell me about that.

For about 20 years we studied anesthetized rats with an open skull and exposed meninges. We would make electrophysiological recordings from trigeminal neurons, trying to understand what happens when you stimulate the neurons, and we did this using models of migraine. Then we figured out that this was not a perfect system, and so we said, let’s try to find a better one. So about five years ago we switched gears to work with mice, trying to do a little bit more digging into the way these meningeal afferents are working in awake animals. And, we decided to adopt some of the advanced methods that people use for neuron imaging, especially calcium imaging.

We developed a method to look at the activity of the afferents in awake, locomoting animals with a closed cranium. We have really exciting data showing that these afferents may not only be nociceptive neurons but may do other things too.

How good are the models in terms of replicating migraine?

My notion of an animal model is that we obviously cannot capture a migraine attack as it occurs in humans. But I’m studying nociception; I’m looking from the perspective of how the primary afferents become activated. So, from that perspective, I think the animal models serve us really well in understanding the neuroscience behind meningeal nociception. I’m trying to stay away from the very complicated issue of what a mouse feels.

What do you see as your most important research findings so far, and why?

The latest findings that are important are our calcium imaging data. But, going back to older literature from my lab, we found that activation of mast cells in the meninges is a way that the system gets activated – an immune or inflammatory response can drive nociception and therefore potentially headache.

Also, I like studies that report negative data, to show that what people have thought before may not be right. Here, we did a study on CGRP [calcitonin gene-related peptide] before CGRP became a huge thing as a therapeutic in managing migraine. Using anesthetized rats, we asked whether CGRP can actually produce meningeal nociception, and we couldn’t get anything to show that it did, at least in males. We also showed that even when CGRP produces vasodilation, the vasodilation itself does not cause any activation of the afferents.

I think this work is telling us that we still do not understand how CGRP works in migraine, and it may not be a direct activator of the system. I’m really proud of this negative data paper.

Your thoughts about negative data make me think of your email signature, which contains a line from the philosopher of science Karl Popper: “The only way to test a hypothesis is to look for all the information that disagrees with it.” This seems like something your lab really puts into practice.

Yes. I’m very skeptical in science, and one of the major things going on in science now is that there’s a research replication crisis. I’m trying to be very careful with experiments, and even if things fail, I think they are important. If you cannot prove or disprove a hypothesis, that is still a finding.

You mentioned the contribution of the immune system and inflammation to meningeal nociception. You were part of a debate at the most recent American Headache Society annual meeting that discussed whether inflammation is a mechanism of migraine. You argued yes. What’s the most convincing evidence that inflammation is an important contributor to migraine?

From my point of view as a preclinical researcher, the key idea is that the afferents that innervate the meninges are inflammatory sensors. Also, the meninges are a huge inflammatory and immune niche. We still don’t know how everything starts, but if you look at this meningeal niche and the afferents, I think that’s a good indication that inflammation is important.

As I said in the debate, if we believe that activation of the afferents is key to developing migraine pain, then there’s no other way to explain how they become activated during a migraine attack. If there aren’t any immune mediators, I cannot give you a good explanation for how the afferents produce long-term pain during an attack.

Looking at the basic neurobiology of migraine, what are the most important questions for the field to answer, and why?

The key part will be what’s going on in the brain itself – how the brain is involved in all the symptoms of migraine, not just the pain. And how do brain processes interact with the periphery? Is there interaction between the brain and the meninges, for example, and vice versa?

We know there are premonitory symptoms in migraine, and we know some areas of the brain are activated early in the attack. But whether and how this leads to pain is one of the interesting mysteries we need to figure out. What gets activated, and how does this contribute to pain?

Looking at the migraine field more broadly, are there particular research developments that seem particularly promising to you?

In preclinical research, studies that do molecular profiling of the neurons that are involved in headache look like the future. And if you can do molecular profiling in animals and then in humans, we may be getting closer to understanding migraine. In addition, looking at sex differences is a key area of research that we need more of and that maybe can be combined with the molecular profiling studies.

Is there anything going on in the migraine field that you think is moving in the wrong direction?

There is still a lot of work we have to do with behavioral modeling. Relying on allodynia, for example – it’s just the tip of the iceberg in terms of what we can do. Not that I know what to do, as I’m not a behavioral scientist, but I think we may get blocked by looking at things that may not shed light on the origin of migraine or on the breadth of symptoms. Maybe combining different behavioral assessments will help, and also maybe using different animals; we’re still stuck with rats and mice right now.

Also, the research is going where the drugs are. So when the triptans came along, everybody did research on triptans. Now it’s CGRP, and everybody does research on CGRP. That’s not to say this is the wrong direction, but more that we should focus on additional things, like the immune system, for example. But no research is wrong research. It’s just a matter of what you get out of the research and how it moves the field forward – how much does it teach us about migraine?

Where would you like to see the migraine field be 10 years from now?

I would like to see a lot more interaction between basic scientists and clinicians. Then I’d like to see application of clinical knowledge to the animal models – for instance, by looking at clinical imaging and applying that to animal imaging, or getting genetic data from humans and applying that to animals. And I’m sure there are going to be many more technologies developed in the next 10 years that we can apply. But we are still going our separate ways right now.

Along with this, as I mentioned, doing more molecular profiling, and also developing better tools, will get us ahead. Also, sampling from human tissues is being done now but it can go deeper.

My take on migraine is that it’s not one disease. We have two different types, migraine with aura and migraine without aura, but I’m sure there are other types or subtypes that we can figure out. But for that to happen, we have to bring basic scientists and clinicians together.

Why isn’t there a lot of back-and-forth between the clinic and the lab?

There aren’t a lot of scientists who study headache to begin with. The field has grown, but there still aren’t enough basic science researchers. I’d like to see a lot more people coming to the migraine field from different areas, for instance, from the neuroimmunology field. Just think of all the people who are studying the meninges, but not migraine itself. But we have to persuade those people to join the field and direct at least part of their research towards migraine.

What are the prospects for new and effective treatments for migraine?

At present we have triptans and pharmacologic treatment that targets CGRP. These drugs work for some people, but we know there are people who do not respond and we don’t understand why. We need to see the data – and the data are out there – to look at the profile of responders versus non-responders and see what’s different. I’d also like to see a lot more personalized medicine. Considering the sex differences in migraine, maybe women will need different treatments than men.

Is there anything else you’d like to mention about migraine that you think is important for people reading this interview to know?

I’d like people to know that it’s hard to do basic migraine research, that understanding this condition as a whole is hard, especially if you’re doing physiology. But this shouldn’t stop people from joining the field, which has a lot of things to offer someone who wants to study migraine from different perspectives. There are so many different ways to study this disease. That will make a lot of people happy in their science.

Neil Andrews is a science journalist and executive editor of the Migraine Science Collaborative. Follow him on Twitter @NeilAndrews