2023 MSC Poster Contest Presentation: Functional and structural magnetic resonance imaging of migraine and treatment outcomes.

The 2023 MSC Emerging Science Contest for Early-Career Investigators took place on December 13, 2023. Below is a written summary of one of the presentations from the contest. Read about other presentations from the event in our Early-Career Science Library.

Category: Physiology/Diagnosis/Therapeutics

Winner: Chloe Cheung, PhD student, University of Western Ontario, Canada

Title: Functional and structural magnetic resonance imaging of migraine and treatment outcomes.

Hypothesis, methodology, findings and conclusions.

Hypothesis: Migraine is the second most disabling disease globally [GBD 2016 Disease and Injury Incidence and Prevalence Collaborators et al., 2017 Lancet], with an estimated 8.3% or 2.7 million Canadians living with migraine [Ramage-Morin & Gilmour, 2014 Health Rep]. In addition, structural brain abnormalities such as decreased grey matter volume (GMV) [Kim et al., 2008 Cephalalgia], and functional brain abnormalities such as worse cognitive performance [Mathur et al., 2015 Neuroimage Clin], increased brain activity during pain [Moulton et al., 2011 Cereb Cortex], and aberrant resting state functional connectivity [Mainero et al., 2011 Ann Neurol] have been observed in migraine patients. Quality of life is also impaired, as decreased sleep quality [Lateef et al., 2011 Cephalalgia] and comorbidities such as anxiety and depression often occur alongside migraine [Buse 2013 J Neurol]. However, pharmacological treatments for migraine have major drawbacks for certain patients due to minimal symptom improvement [Diener et al., 2011 Curr Pain Headache Rep], challenging side effects [Ong et al., 2018 Drugs; Schürks et al., 2008 Curr Treat Options Neurol], exacerbation of headaches from repeated medication use [Diener et al., 2016 Nat Rev Neurol; Buse et al., 2019 Headache], and being contraindicated by pregnancy [Burch, 2020 Headache]. Therefore, there is a need for effective non-pharmacological treatment options.

Non-pharmacological interventions such as meditation are often implemented by patients due to their low physical and emotional risk, relatively low cost, and their ability to enable patients to take a more active role in their treatment [Wahbeh et al., 2008 Neurology]. A type of meditation called mindfulness-based stress reduction (MBSR) increased quality of life and reduced disability and depression in migraine patients [Wells et al., 2021 JAMA Intern Med], while a prior randomized clinical trial in our lab demonstrated enhanced mindfulness-based stress reduction (MBSR+) meditation significantly reduces headache frequency, making it an effective treatment for migraine [Seminowicz et al., 2020 Pain] with outcomes comparable to standard migraine drugs such as propranolol [Diener et al., 2004 J Neurol].

However, a direct comparison of treatment responders (patients experiencing 50% or greater reduction in headache frequency [International Headache Society et al., 2012 Cephalalgia]) to non-responders has not been conducted, and there are no known predictors to date of treatment response for migraine. Understanding how treatment response is associated with both clinical changes and structural and functional brain changes may uncover potential predictors or biomarkers of treatment response in migraine patients that can aid treatment plan development, as well as implicate brain regions as potential targets for novel migraine interventions that will ultimately produce improved outcomes and quality of life for millions of migraine patients.

We hypothesized that in migraine patients undergoing MBSR+, treatment response is associated with changes in clinical symptoms and brain structure and function.

Methods: Our lab previously collected clinical data from migraine patients randomized to MBSR+ (n = 50) and educational control (n = 48). Response status was also determined, with patients being categorized as responders or non-responders. Clinical data was collected on headache frequency, headache intensity, headache disability, sleep deficits, depression, and anxiety at baseline (pre-treatment), 10 weeks (about halfway through treatment), 20 weeks (post-treatment), and 52 weeks (1 year follow-up). This data was analyzed to identify differences between treatment responders and non-responders in clinical symptoms over time. Neuroimaging data was also collected from the same dataset on structural MRI (GMV) and functional MRI (pain and cognitive task fMRI, rsfMRI) at baseline, 10 weeks, and 20 weeks.

I will analyze the above neuroimaging data to examine differences between treatment responders and non-responders over various time points in both specific brain regions and whole brain patterns. I will also perform an exploratory mediation analysis to determine the extent to which brain changes directly cause clinical improvements. Finally, to see if the identified differences in brain regions and whole brain patterns predict response in other chronic pain disorders, I will analyze additional patient data from the recently formed ENIGMA Chronic Pain Working Group database [Thompson et al., 2014 Brain Imaging Behav] using the same process described above.

Results: Overall, headache disability was significantly decreased in all meditation and educational control patients at week 10, week 20, and 1 year compared to baseline. Additionally, at 20 weeks, responders had significantly decreased headache disability compared to non-responders in both meditation and educational control groups.

During cognitive challenge, responders had significantly decreased right dorsoanterior insula (R daINS) activity compared to non-responders in both meditation and educational control groups, while during pain onset, meditation responders have significantly decreased R daINS activity compared to educational control responders.

Conclusions: Responders have greater improvement in headache disability as well as reduced R daINS activity compared to non-responders during both cognitive challenge and at onset of pain.

Implications for understanding migraine disease and/or its comorbidities, or how the research holds promise as a new avenue of future migraine study.

The overarching goal of this research is to understand the brain mechanisms that produce improved symptom management in treatment responders. Although it is unlikely MBSR+ will be beneficial for all migraine patients, by defining MBSR+ responders and non-responders and linking clinical and MRI outcomes between those groups, this study will provide insight into the brain changes necessary for treatment response. Compared to healthy individuals, migraine patients have lower grey matter volume (GMV) [Kim et al., 2008 Cephalalgia] in the insula, which is associated with increased lifetime frequency of headaches [Kim et al., 2008 Cephalalgia], abnormal functional connectivity [Mainero et al., 2011 Ann Neurol], reduced cognitive performance [Mathur et al., 2015 Neuroimage Clin], and greater brain activity during pain [Moulton et al., 2011 Cereb Cortex] including in the anterior insula — a brain region important in active control over pain [Freund et al., 2009 Somatosens Mot Res] that is thought to contribute to chronic pain such as migraine if dysfunctional [Wilder-Smith et al., 2004 Gut]. Meditation in healthy individuals has been found to reduce pain and decrease pain activity in the insula [Zeidan et al., 2011 J Neurosci], alongside increasing GMV in the anterior insula [Lazar et al., 2005 Neuroreport; Grant et al., 2010 Emotion]. These observations in healthy individuals would remedy the structural and functional deficits seen in migraine patients, but it is unknown if these changes occur in migraine patients. Determining if meditation is able to reverse the structural and functional deficits seen in migraine patients will further our understanding of treatment response mechanisms in migraine.

Moreover, previous studies have not examined the effects of MBSR+ on migraine brain structure and function. This research could not only contribute to our knowledge of pain mechanisms, but also to the advancement of research on MBSR+ for migraine and other disorders as these findings will likely be extendable to other populations of chronic pain. Therefore, this study will support the expansion of research in meditation-based interventions for chronic pain.

Finally, it is essential that we understand the brain mechanisms of mind-body interventions: therapies focusing on relationships between the brain, mind, body, and behavior, and their effects on health and disease [Wahbeh et al., 2008 Neurology]. This study will have a major impact not only in the use of meditation in migraine, but in foundational knowledge of other mind-body interventions in the treatment of migraine, such as biofeedback, tai chi, acupuncture, hypnosis, and visual imagery [Wahbeh et al., 2008 Neurology; Woods-Giscombé & Black, 2010 Complement Health Pract Rev]. Thus, examining brain mechanisms of meditation could also provide insights into the brain mechanisms of other mind-body interventions to improve current migraine treatment plans and understand the mechanisms by which mind-body interventions achieve effectiveness.