Current Series: New Paradigms: Preventive Treatment for Chronic Migraine


Stephen Silberstein, MD: There’s an old expression, “Statistics don’t lie, but statisticians do.” Think about this. You walk into a company and you want to say how wonderful this might be and they say, “Three days reduction in mean monthly migraine days? Why do I care?” But think about it. That’s comparing responders to nonresponders. If a person does not respond to the drug, they will not take it. Look at the organization that says whether a drug is worthy or not worthy. They compare the clinical trials and they say, “What is the difference between active drug and placebo,” assuming this is just like somebody who took a drug to prevent a heart attack and was on it for 10 years. The fundamental difference is, if it doesn’t work, you stop it, and they don’t have the sense, the ability, the knowledge, or the brains to figure that out.

When you talk about a 3-day reduction, you’re talking about across all patients. If you look at the 50% of responders who stay in the trial, it’s much higher. Using those 3 days destroys the benefit of the drug. I think we should talk about, in those who respond, how well do they respond, not the average of people.

Andrew Blumenfeld, MD: That’s true. I agree with you that responder rates are a far better way of addressing, for an individual patient, how to think about a medication.

Stephen Silberstein, MD: Right.

Andrew Blumenfeld, MD: Not looking at a population, but individually.

Stephen Silberstein, MD: We should look at the population of responders versus nonresponders.

Andrew Blumenfeld, MD: In this open-label study that we were talking about, it is somewhat biased, because patients who drop out because they’re not responders, or because they may have had [adverse] effects, are not included in this long-term analysis.

Stephen Silberstein, MD: That’s the point I’m making. That is exactly the point I’m making.

Andrew Blumenfeld, MD: I’m agreeing with you, Steve.

Stephen Silberstein, MD: If you don’t respond, you won’t stay on the drug. From a cost-benefit analysis, all the assumptions are wrong. You don’t stay on a drug that doesn’t work, unlike a drug for hypertension or diabetes or long-term benefits. That’s the fundamental difference, and they have to be analyzed differently.

Stephanie J. Nahas, MD, MSEd, FAHS, FAAN: It would be very interesting to take this cohort who have been in the open-label extensions and analyze their months 1 to 3 or months 1 to 6 data, and see how much different it is from the entire cohort.

Stewart J. Tepper, MD: Steve, I would add that you can have it both ways, actually. In the open-label trial, where you enrich it, in a sense, by only looking at responders, you look at the percentage of patients who dropped out due to either adverse events or lack of efficacy. In those you had previously looked at, it’s 4%, 5%, or 6% in each of those for a lack of efficacy or adverse effects, whereas for the topiramate trial, by 12 weeks, 44% had dropped out.

At the same time, we now have a very remarkable study that was also presented at the AAN [American Academy of Neurology], which was a randomized, placebo-controlled trial of eptinezumab for chronic migraine for a year. You can look at a year and compare to placebo at a year. In that study, again, there was an unprecedented responder rate in which 54% of patients had at least a 75% reduction of their mean monthly migraine days at a year. That was not a biased study; that was a placebo-controlled trial. You have it both ways. You can do the analysis both ways.

Stephen Silberstein, MD: I’d like to make our audience aware that the topiramate used in that trial was an immediate-release topiramate, not the extended-release that you talked about, so that tells us nothing about how changing the formulation will lead to better tolerance and better use of the drug. We have to be very careful in this day and age to look at the difference in drugs—the delivery system, the formulation—in terms of trials, and that was immediate-release generic topiramate. I just wanted to point that out to our audience.

Stephanie J. Nahas, MD, MSEd, FAHS, FAAN: To piggyback on what you were saying with this study on 1 year with eptinezumab, in a sense, we’re trying to teach the brain not to have migraine attacks so often. Sometimes I use this analogy with my patients, and it helps them to understand why it may take time and why it may be slower than we want. Some dogs, you can train very quickly to do tricks. Some people learn languages very easily. I do not. It takes me a lot longer to learn a language than somebody who is very facile with them. Some people can learn a language in 6 months, some it takes a year. If you keep studying it, you get better and better, and more and more fluent. The longer you stick with it, and the longer you keep training, the better you’re going to do.