The Health Terminology/Ontology Portal (HeTOP), on which the curious can discover information about off-label use, lists 645 medications prescribed for migraine worldwide. Treatments ranging from blood pressure medications to antidepressants, and anticonvulsants to antiepileptics, along with their doses and administrations, are all listed. The number of migraine-indicated medications is 114. Dominated by triptans and topiramate, the list also includes erenumab, the calcitonin gene-related peptide CGRP agonist. The difference in figures between the predominately off label and migraine-approved lists is a good indicator of the struggle that health care providers have had through the years to help their patients.
The idea now is to make that list even longer by finding biomarkers that lead to new therapies.
But first, a conversation about the trigeminal ganglia.
The trigeminal ganglia
The trigeminal ganglia sit on either side of the head, in front of the ears. Their primary role is to receive stimuli and convey it to the brain. The human trigeminal ganglia contain 20,000 to 35,000 neurons and express an array of neuropeptides, including CGRP. Some neuropeptides, like CGRP and pituitary adenylate cyclase–activating peptide 38 ( PACAP38) are vasodilators. Others, like substance P, are vasoconstrictors. Edvinsson and Goadsby discussed in 1994 how CGRP was released simultaneously in those with “spontaneous attacks of migraine.”
Over the past 30 years, researchers in our institution and elsewhere have shown repeatedly that migraine develops in individuals who are exposed to certain signaling molecules , namely nitroglycerin, CGRP, cyclic guanosine monophosphate (cGMP), intracellular cyclic adenosine monophosphate (cAMP), potassium, and PACAP38, among others. Such exposure reinforces the notion that peripheral sensitization of trigeminal sensory neurons brings on headache. The attack could occur due to vasodilation, mast cell degranulation, involvement of the parasympathetic system, or activation of nerve fibers.
Some examples from the literature:
- In our research, results from a small study of patients under spontaneous migraine attack, who underwent a 3-Tesla MRI scan, showed that cortical thickness diminishes in the prefrontal and pericalcarine cortices. The analysis we performed involving individuals with migraine without aura revealed that these patients experience reduced cortical thickness and volume when migraine attacks come on, suggesting that cortical thickness and volume may serve as a potential biomarker.
- A comparison of 20 individuals with chronic migraine and 20 healthy controls by way of 3-Tesla magnetic resonance imaging scans revealed that those with headache appeared to have substantially increased neural connectivity between the hypothalamus and certain brain areas – yet there appeared to be no connectivity irregularities between the hypothalamus and brainstem, which as the authors noted, is the “migraine generator.”
In other words, vasodilation might be a secondary symptom of migraine but likely isn’t its source.
Other migraine makers
Neurochemicals and nucleotides play a role in migraine formation, too:
- Nitric oxide . Can open blood vessels in the head and brain and has been shown to set migraine in motion. It leads to peak headache intensity 5.5 hours after infusion and causes migraine without aura.
- GRP. Gastrin-releasing peptide receptors cause delayed headache, including what qualifies as an induced migraine attack. Researchers also note that similar pathways trigger migraine with and without aura.
- Intracellular cGMP and intracellular cAMP. These 2 cyclic nucleotides are found extensively in the trigeminovascular system and have a role in the pathogenesis of migraine. Studies demonstrate that cGMP levels increase after nitroglycerin administration and cAMP increases after CGRP and PACAP38 exposure.
- Levcromakalim. This potassium channel opener is sensitive to ATP. In a trial published in 2019 , researchers showed that modulating potassium channels could cause some headache pain, even in those without migraine. They infused 20 healthy volunteers with levcromakalim; over the next 5-plus hours, the middle meningeal artery of all 20 became and remained dilated. Later research showed that this dilation is linked to substance P .
Identifying migraine types
Diagnosing migraine is 1 step; determining its type is another.
Consider that a person with a posttraumatic headache can have migraine-like symptoms. To find objective separate characteristics, researchers at Mayo Clinic designed a headache classification model using questionnaires, which were then paired with the patient’s MRI data. The questionnaires delved into headache characteristics, sensory hypersensitivities, cognitive functioning, and mood. The system worked well with primary migraine, with 97% accuracy. But with posttraumatic headache, the system was 65% accurate. What proved to differentiate persistent posttraumatic headache were questions regarding decision making and anxiety. These patients had severe symptoms of anxiety, depression, physical issues, and mild brain injury attributed to blasts.
All of which explains why we and others are actively looking for biomarkers.
The biomarkers
A look at clinicaltrials.gov shows that 15 trials are recruiting patients (including us) in the search for biomarkers. One wants to identify a computational algorithm using AI, based on 9 types of markers in hopes of identifying those predictive elements that will respond to CGRP-targeting monoclonal antibodies (mABs). The factors range from the clinical to epigenetic to structural and functional brain imaging. Another registered study is using ocular coherence tomography , among other technologies, to identify photophobia.
Our interests are in identifying CGRP as a definitive biomarker; finding structural and functional cerebral changes, using MRI , in study subjects before and after they are given erenumab. We also want to create a registry for migraine based on the structural and functional MRI findings.
Another significant reason for finding biomarkers is to identify the alteration that accompany progression from episodic to chronic migraine. Pozo-Rosich et al write that these imaging, neurophysiological, and biochemical changes that occur with this progression could be used “for developing chronic migraine biomarkers that might assist with diagnosis, prognosticating individual patient outcomes, and predicting responses to migraine therapies.” And, ultimately, in practicing precision medicine to improve care of patients.
Significant barriers still exist in declaring a molecule is a biomarker. For example, a meta-analysis points to the replication challenge observed in neuroimaging research. Additionally, several genetic variants produce small effect sizes, which also might be impacted by environmental factors . This makes it difficult to map genetic biomarkers. Large prospective studies are needed to bring this area of research out of infancy to a place where treatment response can be clinically assessed. Additionally, while research evaluating provocation biomarkers has already contributed to the treatment landscape, large-scale registry studies may help uncover a predictive biomarker of treatment response. Blood biomarker research still needs a standardized protocol. Imaging-based biomarkers show much potential, but standardized imaging protocols and improved characterization and data integration are necessary going forward.
The patients
The discovery of the CGRPs couldn’t have been more timely.
Those of us who have been treating patients with migraine for years have seen the prevalence of this disease slowly rise. In 2018, the age-adjusted prevalence was 15.9% for all adults in the United States; in 2010 , it was 13.2%. Worldwide, in 2019, it was 14%. In 2015, it was 11.6%.
In the past few years, journal articles have appeared regarding the connection between obesity, diabetes, hypertension, and migraine severity. Numerous other comorbidities affect our patients – not just the well-known psychiatric disorders – but also the respiratory, digestive, and central nervous system illnesses.
In other words, many of our patients come to us sicker than in years past.
Some cannot take one or more medications designed for acute migraine attacks due to comorbidities , including cardiovascular disease or related risk factors, and gastrointestinal bleeding.
A large survey of 15,133 people with migraine confirmed the findings on these numerous comorbidities; they reported that they have more insomnia, depression, and anxiety. As the authors point out, identifying these comorbidities can help with accurate diagnosis, treatment and its adherence, and prognosis. The authors also noted that as migraine days increase per month, so do the rates of comorbidities.
But the CGRPs are showing how beneficial they can be. One study assessing medication overuse showed how 60% of the enrolled patients no longer fit that description 6 months after receiving erenumab or galcanezumab. Some patients who contend with episodic migraine showed a complete response after receiving eptinezumab and galcanezumab. They also have helped patients with menstrual migraine and refractory migraine .
But they are not complete responses to these medications, which is an excellent reason to continue viewing, recording, and assessing the migraine brain, for all it can tell us.