The Importance of Identifying Responders in Migraine

David Dodick, MD, MS, a neurologist at Mayo Clinic

David Dodick, MD, MS

In September, 2 more agents in the calcitonin gene-related peptide class will be reviewed for approval by the FDA: galcanezumab and fremanezumab.

Establishing which therapy is preferred over another is a difficult task, and an inability and lack of need to commit to head-to-head trials does not ease that difficulty. Although, there are subtle, currently unknown differences in response for patients, leading to headache and migraine specialists, such as David Dodick, MD, MS, of Mayo Clinic, to call for further research into biomarkers by which responders could be identified.

Dodick spoke with NeurologyLive in an exclusive interview to discuss his thoughts on the agents, what their introduction means for the landscape of treatment, and why identifying responders could improve the state of care.

NeurologyLive: If galcanezumab and fremanezumab are granted FDA approval, how does it help to change the landscape of treatment? How are they different?

David Dodick, MD, MS: The distinction between, let’s say, galcanezumab and fremanezumab and erenumab is that they have different targets. Galcanzeumab and fremanezumab, as I’m sure you know, target the peptide itself, whereas erenumab targets the receptor. What that means in terms of efficacy, safety, and tolerability I don’t know. I don’t know if anybody knows but looking at the clinical trials all 3 of these, regardless of whether they’re targeting the peptide or the receptor, appear to be effective, appear to be very well tolerated, and appear to be safe. In other words, blocking the CGRP pathway by and large looks to be a well-tolerated safe and effective therapy.

What it means, the way I look at is, for example, that if you have a receptor-targeted antibody that, let’s say, is not effective for a patient, is it possible that a peptide-targeted antibody could be effective for a patient, or vice versa? Because that means a lot to people who have sort of waited for a therapy like this for a long time—what if it doesn’t work for them? What I’ve been telling patients is it’s conceivable, of course, that if you don’t respond to the currently available treatment that it’s possible that you’ll respond to 1 of the treatments available because the target is different. The pathway they’re blocking this the same, but the target is different. It’s conceivable that, while we’ll have potentially 4 antibodies in this space that block the same pathway, it’s possible that a patient may respond to 1 and not another, and that’s a fantastic thing for patients, physicians, and providers who care for patients to have that flexibility and that option of switching from 1 to another if 1 is not very effective.

Right now, if you asked me to differentiate them on safety, on tolerability, or on efficacy I can’t, because A. I’ve only had clinical experience so far with 1, and B. looking at the clinical trial data, generally speaking, they all appear very comparable in every respect to safety and efficacy. The only difference is fremanezumab may come forward, and time will tell, with a quarterly injection. The option for patients to inject once every 3 months as opposed to once every month. Some of their trials their phase III trials have shown a quarterly injection is obviously better than placebo and looked comparable to a monthly injection. Anything that reduces the frequency of administration would obviously be an advantage for patients. That’s the only distinction, is in the frequency and the delivery of fremanezumab compared to the other 2, galcanezumab and erenumab, which are monthly subcutaneous injections.

Are there any head-to-head trials planned, or needed?

I don’t think that will be necessary. I think that’s an enormous expense and I don’t think you could design a study with a large enough sample to try to show any small differences that might be there between patient groups. If you design a superiority trial to try to show 1 is superior to another, it would take thousands of patients, and I’m not even sure you would be able to show a difference. If you designed a non-inferiority trial and showed that the 2 are equal than we haven’t learned anything. I don’t think that’s where m1y, time, resources, and energy should be committed, in doing head-to-head trials.

We should be identifying biomarkers for responsiveness and non-responsiveness. For example, I would rather know that this is a patient who is more likely to respond to this antibody versus that antibody. First of all, I would like a biomarker to be able to tell me that this patient is very likely, highly likely, to respond to this therapy to blocking the CGRP pathway. Then, going a little bit further, is it possible to dissect that even finer and is there a biomarker to tell me that yes indeed this patient is more likely to respond to a receptor targeted antibody versus a peptide targeted antibody, or vice versa. If we could know that that would help physicians and patients a great deal, rather than spending enormous amounts of time and m1y in trying to pull apart or show superiority of 1 over the other when that’s not even likely to happen.

Is there any way to go about identifying super responders?

If you’re asking me are those patients who are “super-responders”—could they be identified before they actually receive the therapy, I don’t know of any evidence to suggest that they’re readily identifiable. Given my clinical experience thus far with erenumab since it’s been available, we do indeed see super-responders, and then we see people who respond but you might not consider them a super-responder, although the patient themselves might consider themselves a super-responder. Then there are patients, of course, who don’t respond, and so far, I haven’t been able to, in practice, identify who is who. But, that’s not surprising. There aren’t any indicators to predict which patients are going to respond to any of the prophylactic therapies that are conventionally or historically used for migraine prevention. Whether it’s an oral drug like topiramate or amitriptyline, or another biological like onabotulinum toxin, there aren’t yet factors which can predict who’s going to respond to what. I don’t know if we can identify people who are going to respond in general, nor can we hint at which patients are going to be super-responders.

The good news is that thus far, the adverse effect profile appears to be very good with these antibodies. There are no contraindications warnings or precautions with erenumab, so it makes it very easy to use in practice. Now we don’t know, of course, the long-term potential adverse effects of these. Even though 3-year data suggests that the tolerability and safety appear roughly the same as what we’ve seen in clinical trials, we don’t yet know what the real long-term safety of these antibodies will be, and we don’t know what the safety will be in a more diseased population. Patients who have a lot of co-existing medical illnesses—we don’t really know what the safety is in that population because they weren’t studied in the clinical trials. So, there’s a lot yet to know.

However, given the adverse-effect profile, we’ll know, generally speaking, if the patient’s going to have a response within 2 months of therapy. These drugs may have an onset of effect as early as 1 week, certainly within 1 month, and certainly by 2 months. So, it’s not like we’re exposing a patient to a significant risk by giving them 2 injections to see whether or not they’re going to be a responder. If, at the end of the day, we never have a biomarker—because we don’t have any yet—to determine whether a patient’s going to respond or not, it wouldn’t be the end of the world. But it sure would be nice to be able to save the patient just the time required to determine whether or not they’re going to respond when they could have gotten relief really quicker had we known that they’re not going to be a responder to the treatment that we’re putting them on.

Transcript edited for clarity.