Migrainomics—identifying brain and genetic markers of migraine

DOI: https://doi.org/10.1038/nrneurol.2017.151

218

2 Mechanisms

Main In the 2015 Global Burden of Disease study [12], migraine was ranked seventh of 310 diseases as a cause of 'global years lived with disability'.

Among neurological disorders, migraine is the leading cause of disability and

accounts for over half of all years lived with disability [12, 13].

These two types account for 90–95% of all migraines [14], and the remaining 5–10% are less common types of migraine with distinctive neurological symptoms. Diagnosis of migraine is, therefore, currently based on retrospective reports of headache characteristics and groupings of their associated symptoms, and medical and neurological examinations and laboratory studies are usually normal [15].

We review the latest results that have been produced in these two exciting research fields and detail what these findings add to our current understanding of migraine mechanisms.

Mining Imaging and Genetic Data Imaging studies, therefore, have great potential to provide insight into the brain mechanisms involved in migraine.

The heritability of migraine (40–60% [16–19]) indicates that genetic factors are important, meaning that the identification of specific genetic variants associated with migraine also has excellent potential to improve our understanding of the aeti- ology of the disease [20].

The greatest progress towards identifying robust and informative markers of

migraine has been made with imaging and genetics.

Imaging of Migraine One difficulty in using imaging in this way is that whether the observed changes are disease causing, disease effects or both is unclear.

Progressive brain changes with increasing age and/or migraine frequency have

been reported, suggesting that the disease can drive brain changes [21, 22].

Imaging markers of brain changes seem to be reversible, or at least change when migraine attacks decrease [21], suggesting that the markers indicate disease activity. Imaging is being used for biomarker development in pain [23–25], and the num- ber of studies of functional, structural and chemical imaging measures in migraine is increasing exponentially.

A growing number of these studies are attempting to identify potential biomark-

ers of migraine disease states or treatment effects.

Imaging the Dynamic Migraine Brain No imaging study has captured brain changes during the premonitory phase of a spontaneous migraine attack, although premonitory symptoms occur in 7–88% [26] of patients with migraine.

Imaging in human translational studies, in which nitroglycerine was used to trig- ger migraine attacks, revealed activation in the posterolateral hypothalamus, the brainstem (including the midbrain tegmental area, the periaqueductal grey and the dorsal pons) and cortical areas (including occipital, temporal and prefrontal areas) [27] during the premonitory phase.

2.1 Genetics

219

In one fMRI study of trigeminal ganglion activation in patients with migraine, an increase in activation at the ganglion in response to nociceptive stimulation indi- cated the preictal phase of a migraine attack [28].

Temporal lobe responses to painful heat were altered during the ictal and interic- tal periods compared with those in healthy controls, suggesting hyperexcitability of this brain region in patients with migraine [29].

Imaging the 'Static' Migraine Brain Interictal changes have been assessed across numerous imaging domains, and the findings have provided insight into migraine-induced changes in brain structure and function.

Substantial changes (both increases and decreases) in grey matter volume have been reported in patients with migraine in multiple cortical [30] and subcortical (cerebellum [31], basal ganglia [32, 33] and periaqueductal grey [34]) regions.

Putative anatomical brain changes in migraine include alterations in dendritic

complexity [35], which contribute to measures of change in grey matter volume.

In a small study, involving 23 patients with chronic migraine, differences in brain structure were detected between patients who responded (n = 11) and patients who did not respond (n = 12) to treatment to reverse chronic migraine [21]: patients who responded exhibited cortical thickening in the primary somatosensory cortex, ante- rior insula, left superior temporal gyrus and left pars opercularis.

Few imaging studies have been conducted to assess the effects of antimigraine drugs on brain function and structure in healthy controls or in patients with migraine.

Genetic Markers of Migraine A 2011 GWAS in six population-based European cohorts that included 2446 patients with migraine and 8534 controls, performed by the Dutch–Icelandic (DICE) migraine genetics consortium, identified no loci that reached genome-wide signifi- cance, although analysis of previously identified candidate genes revealed a modest but significant gene-based association of MTDH with migraine (Bonferroni- adjusted gene-based p = 0.026) [36].

In addition to identifying five new SNP risk loci—rs12134493 near tetraspanin 2 (TSPAN2), rs10915437 near adherens junctions associated protein 1 (AJAP1), rs13208321  in four and a half LIM domains 5 (FHL5), rs4379368  in succinyl CoA:glutarate CoA transferase (SUGCT) and rs10504861 near the matrix metallo- peptidase 16 (MMP16) gene—this study replicated all previously implicated loci associated with MA in the 2010 IHGC GWAS except for the rs1835740 SNP (near MTDH).

Substantial efforts have been made to identify the causal genes at the identified risk loci, and the findings have shown that common migraine is associated with markers related to dysfunction of various pathways, including vascular-related pathways, metal ion homeostasis and neuronal migration.

Integrating Imaging and Genetics Combinational approaches, such as using imaging and genetic markers to classify migraine subtypes and improve the specificity of surrogacy, could improve this pro- cess and contribute to the development of precision medicine.

220

2 Mechanisms

This process can be implemented on a large scale, as demonstrated by the Enhancing Neuro Imaging Genetics Through Meta Analysis (ENIGMA) Consortium [37], and can link brain structure and function to genomic data [38]; this process will be necessary to unite the findings from these fields in migraine research.

As integrative techniques become more widely available, they might provide the means to develop genetic markers for disease prediction, disease resilience or drug effects based on brain imaging metrics.

The recent progress in the imaging and genetics of migraine means the field is in a good position to adopt 'imaging genetics', which could provide an integrated bio- marker of migraine (such as genes that are associated with an imaging phenotype).

Conclusions and Future Prospects Larger GWAS are likely to identify more genetic risk loci for migraine [39]; further increases in GWAS sample size are most likely to come from large commercial and public data sets.

In order to identify the causal variants at migraine risk loci and unravel their functional consequences with respect to migraine pathophysiology, functional char- acterization of genetic risk factors in the context of specific patient subgroups within deeply phenotyped clinical cohorts will be needed.

Similar integration of migraine imaging data with migraine genetics data on a large scale has enormous potential to provide a deeper understanding of migraine mechanisms and guide future research.

The advances in 'migrainomics', together with the emergence of deep correlative phenotyping of patients with migraine [40–42], bring the field to a point at which advanced imaging and genetics should be combined to provide “complementary insights into the complexity and heterogeneity of migraine”.

Acknowledgement A machine generated summary based on the work of Nyholt, Dale R.; Borsook, David; Griffiths, Lyn R. 2017 in Nature Reviews Neurology.

Genetic and biochemical changes of the serotonergic system in migraine pathobiology