Timothy C. Hain, MD • Page last modified: July 26, 2020
CGRP is a 37 amino acid peptide first described in 1983. It plays a role in pain transmission and CGRP infusion can trigger migraine. As lots of things can trigger migraine such as nearly any vasodilator, one should not confuse this with concluding that CGRP is critical to migraine. The primary site of action for CGRP is thought to be the trigeminovascular system, located outside the blood-brain barrier, where CGRP interacts with trigeminal afferents and the meningeal blood vessels to release nitric acid and more CGRP. It was hoped that blocking CGRP will disrupt this cycle, and reduce head pain mediated through this circuit. (Parikh and Silberstein, 2018).
As of early 2020, there were 5 new CGRP drugs. They seem to be generally rather weak and expensive.
The literature on CGRP is exploding right now, and as of 2/2020 there are at least 390 relevant papers in PubMed, quite a few considering their relative small role in migraine management. .
From reported data and clinical experience as of 2020, these drugs appear to be weak prevention medications --they are not "magic bullet" drugs that "take out" migraine, but rather are add-on prevention drugs that put another straw on the Migraine camel's back. They are "low risk" -- "low gain" drugs. This lack of magic bullet status makes sense in as much as migraine is not a homogeneous condition, and it would be hard to imagine that a single approach could make a big difference. This is the usual problem with disorders that are defined by symptoms. It also appears, at least right now (in 2020), that these drugs were hugely "over-hyped".
The two new "abortive" CGRP drugs, also appear to be weak, adding just a little bit to migraine management. Compared to our older drugs such as the triptans, these drugs seem to be much less powerful.
So as of 5/10/2020, we think that the new CGRP drugs add little to migraine care, and are "another straw to put on the camel's back", but in general they have been disappointing.
These are the CGRP drugs currently approved:
Injection, long acting.
Oral, short acting
CGRP is elevated in jugular vein blood during acute migraine and cluster headaches. According to Durham (2006), "Studies in cultured trigeminal neurons demonstrate that CGRP is released from trigeminal ganglia cells, that CGRP transcription is increased under conditions mimicking neurogenic inflammation, that migraine pharmacotherapies can both reduce CGRP release and inhibit CGRP transcription, and that tumor necrosis factor-alpha (TNF-alpha), an endogenous inflammatory mediator implicated in migraine, can stimulate CGRP transcription. Together, the results suggest that, in migraine, activation of trigeminal nerves release CGRP and other peptides that cause the release of proinflammatory mediators. These mediators further increase CGRP synthesis and release over hours to days in correspondence with the 4- to 72-hour duration of a typical migraine episode. The increased CGRP synthesis and release might be mediated by activation of mitogen-activated protein kinase pathways, which, in turn, can be modulated by endogenous inflammatory substances such as TNF-alpha and affected by drugs such as sumatriptan."
Thus CGRP interacts with TNF-alpha, a cytokine, which is thought to modulate Meniere's disease. One would think that TNF-alpha inhibitors used for autoimmune disease, such as rheumatoid arthritis, might thus then reduce migraine as well. It is well known that migraine and Meniere's disease are associated with each other. Perhaps these drugs also work in a subset of Meniere's disease. Time will tell. The triptans do not appear to work in Meniere's disease, so this conjecture may fail.
Again from Durham (2016), " Our findings provide evidence that proton regulated release of CGRP from trigeminal neurons utilizes a different mechanism than the calcium and synaptosomal-associated protein 25-dependent pathways that are inhibited by the antimigraine therapies, rizatriptan and onabotulinum toxin A. " This might thus provide a reason that onabotulinum toxin works (modestly) in migraine.
CGRP receptors are found in other places than the head, including peripheral, enteric, and central nervous system, as well as the cardiovascular system, and they are also active in wound healing and other physiologic functions.
There is no current evidence for hepatotoxicity (which was the problem with previous types of CGRP receptor antagonists).
CGRP is a vasodilator. Blocking CGRP may block the protective role of CGRP in preventing stroke and heart attack. (Kee et al, 2018).As small stroke like lesions are more common in persons with migraine, this may turn out to be a long term issue.
We probably will not discover the long term side effects of CGRP inhibitors until a decade of use occurs.
The CGRP drugs compete with a very large number of other migraine prevention drugs. Hopefully as they are working on a different mechanism than other drugs, they will be effective in situations where these other drugs have failed. As noted above, these drugs may share a mechanism with Botox treatment. It will be interesting to see if the effect of Botox and CGRP drugs are additive or not.
Almost all of these drugs are very "hit or miss". Mostly they seem to be "miss". A typical study reports that they work 10-20% of the time. This has also been our experience in trying these drugs in patients.
Eptinezumab has an IV mode of delivery -- this one is not self-administered. In early studies, about 33% of patients taking the 300 mg dose reached the primary endpoint of a 75% reduction in migraine days, compared to 21% of those taking a placebo. So in other words, out of 100 patients, there were about 10 more patients that had a 75% reduction, than patients with placebo. Viewed from this perspective, this drug is not that different in efficacy than all of the other migraine prevention drugs, but of course, the advantage is that it may "pick off" another 10% of the refractory migraine patients. The most common adverse effects in trials were URI and UTI -- i.e. infections. This drug does not appear to be especially effective as results suggest a 1 day/month reduction in migraine frequency. Eptinezumab has a 32-day half-life, but is intended for quarterly administration. This presumably would result in levels in the proportion of 4:2:1 for the first 3 months -- i.e. a sawtooth profile of drug efficacy. It is a little difficult to see the rationale for using an IV drug, similar to other SQ drugs, that works 10% of the time.
erenumab, uniquely, is an anti-CGRP receptor antibody. It is not directed against CGRP, but against the receptor for CGRP. Compared to placebo, this drug was reported to reduce mean headache days by 6.6 (from 18), compared to a 4.2 day reduction in placebo. Converting this to percentages (not the same as the above calculation), the difference in days between active drug and placebo is 2.4, and that divided by 18, is again about a 10% difference. The safety profile of this drug, at least over short periods of time, is similar to placebo. An open label trial (i.e. not controlled) suggested roughly a 50% reduction in migraine days/month, for a dose of 70 mg subcutaneous every 4 weeks. Another phase-2 trial suggested a mean reduction of 2.5 headache days/month compared to placebo. Erenumab has a 28-day half-life (i.e. half is still around after 28 days)
Erunamab was approved by the FDA in May of 2018. It appears to be priced at about $575/month (if you pay out of pocket). As of September 2018, it could be obtained as samples, and ordered through a specialty pharmacy. It can be injected at home. There are two dose sizes -- 70 mg and 140 mg.
We have encountered some complaints from patients that the drug is difficult to obtain. One of our patients dropped Aimovig because of lack of insurance coverage and switched to Ajovy. We also had one patient point out that the Aimovig privacy notice involves quite a bit of signing away of medical record privacy. If you want nearly free drug, you must allow Novartis access to your medical record.
Robbins and Phenicie(2018) reported on 100 patients in an uncontrolled study, and noted constipation occurred in about 18%. The largest single group were 45% of patients that reported 30 to 70% improvement. They observed that the results of Aimovig were similar to results of Botox. We are surprised that Aimovig did so well, and hope that there will be more reports.
A recent update was made to the Aimovig prescribing label as of April 30th 2020. The label discusses watching for hypertension (high blood pressure), which has occasionally been substantial and required hospitalization. From the companies surveillance, this appears to be very rare.
There are now two routes to obtaining Aimovig. An ordinary prescription is now often accepted, for those with commercial health insurance. Otherwise, one can apply to the company using this link: Process to try the first CGRP drug (Aimovig).
Results were similar -- a mean reduction of 4.6 headache days, which differed from placebo (2.5 headache days) (VanderPluym et al, 2018). The population of 297 patients tested had "high frequency" or "chronic migraine" -- or in other words, people with more than 8 headaches/month. Thus drug was injected quarterly in a rather large amount -- 225 or 675 mg. It is also subcutaneous. The most common adverse effect was injection site pain. It is now available in two regimens -- monthly (225 mg) or quarterly (675 mg). Anjovy has an estimated 31-39 day half life. Silberstein et al (2018) reported that decreases in headache, nausea, vomiting, photophobia and phonophobia occurred within the first week of treatment.
Ajovy was approved in September of 2018. Some insurances seem to cover it. We have had a little better results with this drug than with Aimovig, but our experience is not large.
This drug is another CGRP antibody drug. Again, good results were reported, in a very confusing way. Emgality's seems to reduce migraine days/month, compared to placebo, on average by about 2 days. This is in the same general ballpark as all of these other drugs.
In a phase 2b trial, there was a mean reduction of 1.1 headache days per month with 120 mg of this drug. Emgality has an estimated 25-30 day half-life.
In a recent study reported in 2018, there was approximately a decrease of 4 migraine headache days, compared to pre-treatment (Sklijarevski et al, 2018). These patients were not very sick and only were experiencing (on average) about 7 migraine headache days with no treatment. Both the placebo and the active treatment had a large effect -- about 3 of the 4 headache days were placebo response. So this is really a study of persons with relatively mild migraine headaches. I would not say that a 1 day/month change in the # of headaches (compared to placebo), was very large.
I was told by the drug representative (for Lilly), that an ordinary prescription is sufficient for Emgality, but there was also a generous support program for those whose insurance does not cover Emgality.
This drug is very different -- it not only ends with a "pant" instead of a "mab", but it is oral as well as quick to act. Like the other CGRP drugs, it does not appear to be very effective. In the phase 3 double-blind study, a dose of 75 mg resulted in freedom from pain in 21% at 2 hours, compared to placebo (11%) (Croop et al, 2019). This is quite a bit worse than the triptans. In fact, it is not very different from placebo. Our thought is that this drug might be used to figure out if one of the other ones are going to work, as well as in people who can't take more effective abortives (such as the triptans or NSAIDs). This drug seems to improve matters by about 10%.
This oral CGRP Receptor Antagonist or Gepant, was approved by FDA on in Dec 23, 2019. Like rimegepant, this drug is an oral agent. We have not had any clinical experience with this drug as of 5/10/2020. It is available in two doses, 50 and 100. If a patient has commercial insurance, Allergan (the maker) offers programs to make the cost similar to that of other migraine abortives. Similar to all of the other CGRP drugs, this drug seems to just be a little better than placebo, with roughly 34% vs 24.5% reporting much/very better migraine at 2 hours. (Dodick et al, 2020). There does not seem to be a big difference between the two doses (i.e. improvement does not scale with dose). It appears to increase the # of people better at 2 hours by about 10%.
Cluster headache may be a special situation indicating use of CGRP inhibitors. Vollesen et al (2018) reported that CGRP itself provokes headache in persons in their active phase of cluster headache. This suggests that perhaps the CGRP blocking drugs might have a special role for cluster flare-ups. This remains to be established. As CGRP provokes headaches in persons not having cluster headaches as well, perhaps this is really just showing that something makes people more sensitive during the "active" phase, and not that these drugs have a special role. Trial and error will get this figured out.
Tepper et al (2018) reported that "Galcanezumab showed effectiveness in preventing episodic cluster headache" and also "Both galcanezumab and fremanezumab failed to prevent chronic cluster headache". Chan and Goadsby (2020) reported that "Galcanezumab was shown to be efficacious, reducing the frequency of attacks in episodic cluster headache, while fremanezumab failed its primary endpoint in episodic cluster headache. Both fremanezumab and galcanezumab trials in chronic cluster headache were terminated after futility analysis predicting the failure of both trials to fulfil their primary endpoint. "
We would not think that the abortive CGRP drugs, like rimegepant, would be a good idea in cluster, as cluster headaches are both very intense and usually are over very quickly.
Children and CGRP drugs.
Although long-term safety has not been established, experts from the American Headache Society recommended considering anti-CGRP mABS for children with migraine who have more than 8 headache days/month, have a PedMIDAS score of 30 or greater, have failed two or more therapies (drug, non-drug, nutrients), and are past puberty (Szperka et al, 2018). In this writer's opinion, these experts are suggesting that almost anyone might get these drugs. In as much as long term safety has not been established, we think that this recommendation is unreasonable. We think children should have failed at least 2 types of drugs (i.e. antidepressant, anticonvulsant), be extremely impaired (i.e. a higher score on the PedMIDAS), and be very impaired (i.e. not be going to school). We think criteria should be loosened up after 10 years, given that no long term issues have emerged. Note that Robbins and Phencie (2018) also suggested avoiding use in children.
We think it is obvious that these drugs should not be used in pregnant women, until safety has been established. They are simply unknown quantities. We expect data will become available.
As summarized above, we have a new class of migraine drugs, that seem to be weak, even more so than nearly every other migraine prevention drug. Because they are new, we don't know long term risks. Because they are weak, we are balancing here a small benefit with an unknown risk and a large price.
Dr. Cowan, Chief of the division of headache medicine at Stanford, recently wrote a commentary about CGRP antagonists. As Cowan wrote (2018), these drugs will likely be "priced in the range of the currently available onabotulinumtoxin A" -- In other words, rather high, and also require a diagnosis of chronic migraine, and also require trials of 3 other medications. Dr. Cowan suggested that these drugs will be a $4.5 billion product.
Robbins and Phenicie (2018) suggested that use of this family of drugs should be minimized in patients less than 30 years of age due to lack of known long-term risk. Their online article on this is https://www.practicalpainmanagement.com/pain/headache/stake-possible-long-term-side-effects-cgrp-antagonists. They also suggested caution in patients with advancing age (i.e. 70 and above). This would seem to leave only 31-69 where prescriptions might proceed without unusual cautions. Also they suggested extra caution in persons with vascular risk factors, immunological disorders, constipation, as well as a large number of other miscellaneous conditions. They also pointed out many missing pieces of information mainly concerning CGRP's role in normal people. CGRP may be important to keep blood vessels open. It is also involved with healing of ulcers. What might happen in pregnancy ? CGRP suppresses tnf-alpha. Perhaps less CGRP will result in more TNF-alpha and inflammation. We should know the answers to these questions in a few years.
In the author's view, these drugs are worth trying in severely impaired patients, but are not reasonable in most people whose headaches can be managed with conventional treatment. We also expect that it will not work in everybody with "migraine". This is because migraine is not a homogeneous condition, as it is defined by symptoms rather than molecular biology or imaging. This means it is a "wastebasket condition" resembling many other conditions defined by symptoms.
Because of its "wastebasket status", meaning we are not treating a disease but rather a menagerie of disorders, we expect that these drugs in blinded controlled studies, will be less effective than reported in these pre-marketing trials. We also expect that patients will develop blocking antibodies to these proteins, and even if they work for a while, their efficacy will eventually wane. We are not sure whether or not these drugs will affect the propensity of migraine patients to develop white matter lesions -- as CGRP is a vasodilator, blocking CGRP might lead to more white matter lesions. Women with aura are particularly likely to develop white matter lesions.
There are a lot of interesting social questions to consider with the CGRP drugs. These drugs are not saving lives -- they are reducing pain. Who decides how these are paid for ?
- Commercial Insurance coverage right now appears to be working to some extent.. There are rumors in the Neurology community that the large insurance companies view these drugs from the perspective that they are very expensive, and are working out ways to limit their impact on insurance expenses. While one way to look at this is to save costs, one might also think of this as a social policy mechanism, implemented by commercial insurance companies. One mechanism of making it difficult might be to require "Headache certification" to prescribe these drugs. Botox treatment for migraine is somewhat similar to CGRP -- not super effective and very costly. Health insurance companies have already succeeded in making "Botox" a very difficult and somewhat risky process for physician offices. For example, in our neurology office, we learned that we should require the Botox to be prepaid and delivered to the office by insurance, as otherwise the insurance company may cut costs by denying coverage, leaving the doctor's office with the bill. We expect that Botox availability for migraine will go down, as it requires a very large amount of staff time. We write about this not to criticize the insurance company process, but rather to point out the work-arounds that health insurance companies use to allocate limited funding in a regulatory environment that may require unlimited benefits.
- It would seem to us that there should be a proportionality between reduced headache pain and cost. These CGRP drugs are not saving lives -- they are reducing pain. Pain is very important of course, but it is harder to quantify than death. The question then is what is one to four less headaches days/month worth to individuals or the government that funds much of health care ? The list price here appears to be $600/month.
- Another interesting question is -- where should government health care dollars go ? -- if we have $1 billion to spend, who decides what proportion is spent funding treatment of lethal conditions like cancer or Covid-19 vs. migraine treatment ? As noted above, commercial insurance companies often make access decisions like this. Who should make these decisions for publicly funded programs such as Medicare and Medicaid ?
Going back to safety, Loder and Burch (2018) wrote about "who should try new antibody treatments for Migraine". They suggested that they should be avoided in persons with infrequent migraine attacks, as well women who are pregnant or becoming pregnant. They recommend that they be avoided in patients with "an increased risk of stroke and coronary artery disease". We would think this would include many older patients ...
Long term trials are in progress but take .. a long time.