Aster Harder

8 CHAPTER 8 180 categories in the PheWAS analyses. These results are consistent with previous reports of a shared etiology and some genetic correlation between migraine and cardiovascular and cerebrovascular endpoints.76-78, 88-92 However, in our analysis, the migraine risk alleles neither consistently increased nor consistently decreased the risk of coronary artery disease or the risk of hypertension. A key role of the central nervous system (CNS) in migraine pathophysiology has emerged from animal models, human imaging, and neurophysiological studies,10, 93 while support for CNS involvement from genetic studies has been more difficult to obtain. A likely reason is the paucity of gene expression data from CNS tissue types, but recently more data have become available, making such studies feasible. Our LDSC-SEG analysis using gene expression data from 13 brain regions showed an enrichment for caudate nucleus in the basal ganglia, and with chromatin-based annotations for five CNS tissue types: dorsolateral prefrontal cortex, neurospheres derived from cortex, fetal brain, germinal matrix and neurospheres derived from ganglion eminence. Alterations in the structure and/or function of several brain regions,93-95 including basal ganglia, cortex, hypothalamus, thalamus, brainstem, amygdala and cerebellum, have been reported for individuals who suffer from migraine, but the cause of these changes is not known. In addition to the support for the hypothesis that both vascular and CNS are important in migraine pathogenesis,8, 93, 96 the tissue enrichment analyses also reported some tissue types of the digestive system as well as ovary at FDR 5%. Given the female preponderance and suggested influence of sex hormones (e.g. menstrual related migraine) in migraine,97-99 the involvement of the ovary is an interesting finding, although the statistical evidence for it currently remains weaker compared to that for the vascular and central nervous systems. A particularly interesting finding in our GWAS was the identification of risk loci containing genes that encode targets for migraine-specific therapeutics. One new locus contains the CALCA and CALCB genes on chromosome 11 that encode calcitonin gene-related peptide (CGRP). CGRPrelated monoclonal antibodies have been successful for the preventive treatment of migraine,100 and they are considered as a major breakthrough in migraine-specific treatments since the development of the triptans for acute migraine over two decades ago. Another new locus contains the HTR1F gene that encodes serotonin 5-HT1F receptor, which is the target of another recent migraine drug class called ditans.101 Ditans provide a promising acute treatment especially for those migraine patients that cannot use triptans because of cardiovascular risk factors.23 These two new GWAS associations near genes that are already targeted by effective migraine drugs suggest that there could be other potential drug targets among the new loci and provide a clear rationale for future GWAS efforts to increase the number of loci by increasing sample sizes further. In addition, GWAS data with migraine subtype information can help prioritize treatment targets for particular migraine symptomatology, such as aura symptoms, that currently lack treatment options. More generally, utilizing genetic evidence when selecting new drug targets is estimated to double the success rate in clinical development.102, 103

RkJQdWJsaXNoZXIy MTk4NDMw