Among a plethora of treatment options for migraine sufferers, only one class — calcitonin gene-related peptide-inhibiting monoclonal antibodies (CGRP MABs) — is specific to migraine headache prevention; the others are repurposed therapies designed for other indications and are often only modestly effective for migraine prevention.
Cleveland Clinic is a non-profit academic medical center. Advertising on our site helps support our mission. We do not endorse non-Cleveland Clinic products or services Policy
Unfortunately, the four FDA-approved medications in the CGRP MAB class are so expensive that insurance coverage for their use can usually be obtained only after a patient has demonstrated a lack of response to several trials of other therapies — a process that can take months or even years. Moreover, while some patients respond extremely well to CGRP MABs, many derive little benefit.
“Being able to predict who is likely to have a good response to CGRP MABs could be a game changer in migraine therapy,” says Zubair Ahmed, MD, a headache specialist in Cleveland Clinic’s Center for Neurological Restoration. “If we had good data, insurance companies might change their policies to allow patients to start effective treatment sooner.”
With these ends in mind, Dr. Ahmed approached Ignacio Fernandez Mata, PhD, of Cleveland Clinic’s Genomic Medicine Institute, an investigator who specializes in genetic disease associations. Together they developed a research plan and obtained a Cleveland Clinic Co-Laboratories Award providing $100,000 for each of two years. Amira Salim, who is pursuing a doctorate in molecular medicine at Cleveland Clinic’s Lerner Research Institute, joined the team as a data analyst.
“A lot is known about the genetics of migraines, and we also have the tools to characterize the range of response seen for this therapy,” says Dr. Mata. “It’s exciting to apply them to this highly translational research.”
A very effective therapy — for some
The CGRP MABs were approved by the FDA between 2018 and early 2020. Cleveland Clinic experience shows that about 20% of patients have at least a 75% reduction in migraine frequency (“super-responders”), but a disappointingly large share of patients — 42% — have only a 25% or less reduction in migraine frequency (nonresponders), with the remaining patients falling between these groups.
How a patient is likely to respond is currently unknown. A polygenic risk score can help quantify the genetic contribution; such a score was previously tested by other investigators to identify patients likely to respond to triptans, a class of rescue medications used to reduce CGRP levels during an acute migraine attack.
But Drs. Ahmed and Mata believe treatment success is likely to be only partially due to genetic factors. Heritability for migraines is high (40% to 60%) but also complex, with 44 single-nucleotide polymorphisms at 38 distinct loci implicated.
“Demographic and clinical factors are also likely important for therapy response,” says Dr. Ahmed. “A more complete model incorporating such factors has never been developed before for predicting response to migraine pharmaceuticals.”
A model integrating clinical and genetic variables
The investigators’ current research has two primary aims:
1) Identify demographic, clinical, migraine-specific and quality-of-life metrics that correlate with clinical response. Fifty possibly relevant variables are being extracted from the de-identified electronic medical records of about 2,000 Cleveland Clinic patients treated with CGRP MABs for at least three months. The researchers will use these data to build a multivariable clinical prediction model to determine which combination of factors is most associated with clinical response to CGRP MABs.
2) Generate a polygenic risk score to predict response. The team will randomly select 165 super-responders and the same number of nonresponders and then obtain DNA samples to conduct genome-wide genotypes. Single-nucleotide polymorphisms of interest will be based on data from the International Headache Genetics Consortium, a multinational research collaborative that performed the largest migraine genome-wide association study in 2016, using data from about 35,000 patients with migraines. The resulting polygenic risk scores will be analyzed against treatment outcomes, both independently and in combination with the clinical-demographic model developed in the first aim.
“Our goal is to develop a model that can be used by other institutions, with data that are typically and widely collected in healthcare electronic medical records,” says Dr. Mata. “It’s also possible that we will identify factors not typically collected, in which case we may advocate for their incorporation in health records in the future.”
Currently halfway through their grant, Drs. Ahmed and Mata plan to pursue National Institutes of Health funding to further their investigations. They would like to expand their model to include genetic variables linked to migraine treatment response (the current project is using data associated with migraine risk) and to generate cohorts from different institutions for validation of their models.
“Once developed, we expect our model will be generalizable to other migraine treatments, and it would be interesting to see associations that might be identified,” says Dr. Ahmed. “We hope it will allow us to improve on our current trial-and-error system for migraine therapy.”