AUTHOR COMMENTARIES - 2008, From Special Topics

According to Essential Science Indicators^SM from Thomson Scientific, Prof. Dr. Ferrari's record includes 132 papers cited a total of 3,659 times between January 1, 1997 and December 31, 2007—34 of these papers with a total of 1,534 cites are classified in the field of Clinical Medicine, and 80 papers with a total of 1,940 cites are classified in Neuroscience & Behavior.

Prof. Dr. Ferrari is Professor of Neurology at the Leiden University Medical Center in The Netherlands. He is also Chair of the Leiden Center for Translational Neuroscience.

In the interview below, he talks with ScienceWatch.com correspondent Gary Taubes about his investigations into the mechanisms of episodic brain disorders.

What do you consider the fundamental dilemma in understanding migraine?

The main thing that interested me when I first started studying migraine, and it still puzzles me now, is the on-off phenomena of migraine and other paroxysmal brain disorders. They’re episodic—these are disorders that come in attacks. For migraine, you’re completely normal, then suddenly your brain switches off and is abnormal for one to three days. Then it normalizes, and you’re completely fine again. That mechanism of episodic brain disorders is my primary research topic.

Your most influential paper, which was actually not included in this particular analysis because it was published two months before our analysis began, is on the genetics of familial hemiplegic migraine, published in Cell in 1996 (Ophoff RA, et al., "Familial hemiplegic migraine and episodic ataxia type 2 are caused by mutations in the Ca2+ channel gene CACNL1A4," 87[3]: 543-52, 1 November 1996). How did that research begin, and why familial hemiplegic migraine?

That was really coincidental. I still recall the day, vividly. I saw two different patients in my clinic with this syndrome called familial hemiplegic migraine, which is migraine associated with half-sided paralysis. It’s a very severe, rare subtype of migraine. And yet here were these two different patients on the same day, presumably from two different families, although both came from the very same region in the Netherlands. I was struck by that. I thought that it could not be true, that it was two different families. And it turned out that it was actually one very, very large, extensive multi-generational family.

This coincidence suddenly changed into a geneticist's delight for doing linkage studies. And that’s exactly what we did. We started doing linkage analysis with a new method—micro-satellites—together with our genetics department, and particularly Rune Frants, who is a co-author on many of our papers.

How many cases did you find?

We found 20 affected patients in this family.

And you managed to localize the gene?

No, actually. That’s a short story, but the longer story is more interesting. We took much longer than we had hoped to find the linkage so we were scooped by another research group working with a similar family in Paris. They localized the gene to chromosome 19. We weren’t happy about being scooped, but we were happy that the linkage was found. Working from that linkage, we identified the gene two years later. So that’s the short story and that was the first gene identified for migraine, as reported in the 1996 Cell paper.

So what does the gene do and how does it relate to migraines?

That is the really nice longer story. We immediately knew what this gene did. It encodes for the ion-conducting subunit of neuronal calcium channels. So it modulated calcium influx into cells. And by doing that, since it’s a neuronal calcium channel, it controlled and modulated the release of neurotransmitter. That, of course, is extremely important in functional terms.

The next step, which pertains immediately to your question, was the next most important paper in that line of research. That is a Neuron paper in 2004 (van den Maagdenberg AMJM, et al., "A CACNA1A knock-in migraine mouse model with increased susceptibility to cortical spreading depression," 41[5]:701-10, 4 March 2004), which discusses the characterization of a transgenic knock-in mouse. What we did is introduced the human calcium channel mutation into the genome of a mouse, and generated a knock-in mouse with this mutation. The results were very striking: we found a whole series of mechanisms that all seemed to be extremely important for migraine. We feel that was the first migraine mouse model.

Do they have the one-sided paralysis, as well?

Yes, they do, and it is, of course, quite striking. They not only show the half-sided weakness, the paralysis, but also episodes of headache and photophobia—super-sensitivity to light—which is also part of migraine syndrome.

Do you know why the weakness appears on only one side?

That’s a question we cannot answer. We did unravel the mechanism behind the familial hemiplegic migraine, though: it’s due to a phenomenon called cortical spreading depression. When you trigger the brain it responds with a very brief hyper-activation, followed by a complete cessation of neuronal function, and that then spreads along the cortex. It recovers after around an hour. So we showed the mechanism for episodic brain disorders. What we proved was that the migraine mouse models were highly susceptible to cortical spreading depression. You could induce cortical spreading depression far more easily than in the normal mouse, which clearly explained what could happen in humans. We know already in normal, common types of migraine, cortical spreading depression is most likely involved. So that was a wonderful link between the mouse model and the human situation.

At the risk of sounding frivolous, how do you tell if a mouse has a headache?

That’s an excellent question. It’s a complicated story. We are studying that with Professor Jeff Mogil from McGill University in Montreal, Canada. There are several lines of evidence. Part of that is not yet published—it’s under review. What’s now in the public domain, what we’ve presented at conferences, is that mice have a specific grooming behavior. They start stroking their legs, tail, and body. And they do that in a very specific way. They hardly ever stroke their head, unless they have a headache. You can test that by inducing an artificial headache, causing them pain, and suddenly these mice will start stroking their heads, and not so much the rest of their bodies. When we compare our transgenic mice with normal mice, they have a far higher rate of head stroking. That’s one line of evidence that they actually have headaches.

What’s the significance for other episodic brain disorders? Epilepsy, for instance?

The answer to that is complicated. In epilepsy there is hyper-excitation and synchronization of the brain. Patients then get seizures, often with jerking movements of the arms and legs. Still, there are a lot of commonalities between migraine and epilepsy. First of all, both disorders are often co-morbid; they co-occur far more frequently than you would expect by chance. Secondly, quite a few prophylactic medications are effective in both disorders. Not all, but a few. Thirdly, there are now a number of genes being discovered, including our calcium channel gene, that can cause either migraine or epilepsy or both depending on the mutation.

So there is a clinical similarity, there’s a genetic similarity, and there’s a treatment similarity, and yet they’re different diseases. One of my hypotheses, not proven, is that the difference is in cortical spreading depression. Both start with hyper-excitation; in migraine, it’s followed by depression, the cessation of neuronal excitation; in epilepsy, it continues. After that, it’s all speculation. But, yes, we think our findings are useful for understanding other episodic brain disorders.

In fact, we also found that different mutation in this same gene that causes familial hemiplegic migraine causes a completely different disease called episodic ataxia. That’s an ever rarer disorder, where the patients get episodes several hours long of ataxia, or loss of coordination. They basically look like they’re drunk. It’s a disorder of the cerebellum, the part of the brain that controls coordination. It malfunctions for two to three hours, and it turns out this same calcium channel gene is also responsible for that disorder.

Where do you see your research going from here?

The ultimate goal is to understand why migraine patients get attacks. Why are they completely normal and then suddenly they have this derangement of brain function and normalization? Our hypothesis is that these patients have a reduced trigger threshold and that trigger threshold is defined by genetic factors.

There are several ways of reducing this trigger threshold. The ultimate result is migraine patients more easily get a spreading depression, and then, ultimate goal is to identify treatment targets to develop better prophylactic agents that would be used by patients then on a daily basis to prevent the migraine attacks. That’s my dream, to help develop prophylactic migraine agents. There is an enormous demand for that. There are no good ones available.

What kind of progress do you think you can make in, say, the next five years?

I think it’s realistic to say that in five years we should have several new targets for prophylactic agents. There’s already one drug currently being tested based on this hypothesis, so that is a very realistic prediction.

What’s the most difficult or challenging aspect of migraine research?

First of all, there’s no objective test for migraine. If you deal with cancer, for instance, or stroke or myocardial infarction, you have objective tests; you can have a picture showing what’s wrong. You don’t have that with migraines. You have to rely on the story that the patient tells. And you usually don’t see the patients during the attacks. You see them when they’re healthy and they tell you the story.

Secondly, it’s clear that migraine has a clinical heterogeneity. We probably shouldn’t even speak about migraine but about migraines, plural. There are two main kinds—with aura and without. There are probably more. And that has implications for understanding the underlying mechanisms and for treatments. It’s likely that different treatments should be used for different types of migraines.

A third major complicating factor is that it’s clearly a multifactorial disorder; there are both genetic and environmental factors involved. In that respect, it’s very similar to asthma or hypertension or any of the other common diseases, which are not caused by just one or two genes, but by a combination of different genes, interacting with non-genetic environmental factors. In this sense, it’s useful to say that what we did was study a monogenic subtype of migraine. This is a subtype called familial hemiplegic migraine, and its caused by one gene, and the major step now is to go from monogenic subtypes, of which there are several, to the multifactorial complex disorders, and that’s the big challenge.

What message would you want to give the lay public about your research?

There are a couple of them. The first is that not every headache is a migraine. Migraine is an enormously disabling and debilitating disorder. It can severely affect patients and their families. It ranks in the WHO top 15 of most disabling disorders and is the most costly brain disease to society. Having a migraine attack is among the most debilitating events for humans.

The second is that the underlying mechanism really has nothing to do with stress or other psychological events—it is a purely pathophysiological process. To a scientist it’s extremely interesting and challenging to study; the research goes from basic molecular biology to biochemistry and pharmacology and genetics.

Michel D. Ferrari, M.D., Ph.D.
Leiden University Medical Center
Leiden, The Netherlands

Relevant keywords for this interview: Michel Ferrari, migraine, episodic brain disorders, familial hemiplegic migraine, migraine gene, mouse model, trigger threshold, Migraine & Other Vascular Headaches