Abstract
Migraine, characterized by throbbing cephalalgia, has long been referred to as a “vascular headache” disorder. This stems from the putative role of the extracerebral dural and meningeal vasculature as the end organ of migraine pain, following release of inflammatory neurotransmitters from trigeminal nerve endings. Recognition of migraine as a risk factor for stroke has raised questions about the role of cerebral vessels. Associations of migraine with variant angina, Raynaud’s disorder, retinopathy, livedo reticularis and pre-eclampsia has led to speculation of systemic involvement of the vasculature, particularly the endothelium (1).
Endothelial dysfunction, a precursor to atherosclerosis, is characterized by endothelial cell activation (manifested by inflammation and hypercoagulability) and vascular reactivity impairment. In the investigation of Perko et al. (2), the authors explore systemic endothelial-dependent vasodilatation in migraine patients and healthy controls. The experiments involve end-diastolic measurements of the brachial artery diameter before and after blood pressure cuff–induced hyperemia. Flow-mediated dilatation (FMD) refers to the percentage increase in diameter between the two measurements, and low values reflect impaired reactivity, a functional component of endothelial dysfunction. The authors found no differences in FMD between migraineurs and controls, nor between migraineurs with and without aura, suggesting a lack of systemic endothelial involvement. A strength of the study lies in the careful selection of the population, excluding those with hypertension, hypercholesterolemia, diabetes, abnormal body mass index (BMI) and known cardiovascular disease. Associated with vascular risk factors and predictive of vascular ischemic events, endothelial dysfunction has been referred to as “the ultimate risk of the risk factors” (3). Migraine, particularly with aura, has also been associated an elevated cardiovascular risk profile (4). By excluding subjects with or at risk of vascular disease, the design best allows for examination of the unconfounded migraine-endothelial function relationship. The authors go a step further by excluding those with increased carotid intima-media thickness (IMT), a structural marker of subclinical atherosclerosis. Some studies have shown a correlation between IMT and FMD (5), whereas others have not (6). If these structural and functional markers are related, exclusion of one of these markers may bias against finding differences between groups for the other.
How do the findings of Perko, et al. of non-altered FMD in migraineurs fit in with the accumulating literature on this topic? As the authors summarize in their well-researched discussion, their findings concur with those of one study (7) but conflict with the results of several others (8–11). A recently published study, similarly measuring FMD and IMT in migraine, excluded persons with known risk factors for vascular disease, except for elevated BMI (12). There were no differences in FMD between migraine (with and without aura) patients and controls, except in a subgroup with transformed migraine. FMD did, however, correlate inversely with carotid IMT, which was increased in all of the migraine subgroups. This could be expected to occur if IMT was an earlier marker of disease, but wall thickening is usually found to follow rather than precede endothelial dysfunction.
In light of the discrepant results from the handful of published studies, it is interesting to reflect on the findings of Napoli et al., who investigated vascular reactivity in migraine by examining both the endothelial and vascular smooth muscle cell (VSMC) components (13). They used plethysmography measurements of forearm blood flow (FBF) during brachial artery infusions of acetylcholine (Ach), an endothelium-dependent vasodilator, and nitroprusside, a vasodilator directly acting on VSMC. In addition to this they measured forearm production of NO, a reflection of endothelial activation, and cyclic guanosine monophosphate (cGMP), produced by VSMC. In patients with migraine, the vasodilating effect of ACh was markedly reduced, as was the dose-response curve to nitroprusside. The fact, however, that during ACh infusion, NO release from the endothelium was similar in migraineurs and controls, and cGMP release from VSMC was decreased in migraineurs, led the authors to attribute the impaired vascular reactivity in patients with migraine to VSMC, rather than endothelial, dysfunction.
Perko et al. raise the possibility that impaired vascular reactivity, if not apparent systemically, may still occur within the intracranial circulation, as has been suggested by some, but not all, transcranial Doppler (TCD) studies (14–16). Congruent with either theory are the studies of circulating endothelial markers. Circulating endothelial progenitor cells (EPCs), for instance, which are released from the bone marrow into the circulation to replace dysfunctional endothelium, were reduced in number and function in migraine, particularly migraine with aura (17). In a study of endothelial activation, Tietjen et al. found that compared to controls, women with migraine had evidence of oxidative stress (lower nitrate/nitrite levels) and of both a pro-coaguable (higher levels of von Willebrand factor [vWF] activity and tissue plasminogen activator [tPA] antigen) and pro-inflammatory (higher levels of C-reactive protein [CRP]) milieu during the interictal period (18). The finding of elevated CRP in migraine has been reported in a number of other studies as well (12,19–21). Recent reports of migraine-associated elevation of endothelin-1 (ET-1) (12), a vasoconstrictive peptide produced by (VSMCs), calls to mind earlier findings from an in vivo rat model showing ET-1 to be a potent inducer of cortical spreading depression, the neurophysiological correlate of migraine aura (22). Ictal studies of migraine that document increased vWF, ET-1, soluble intercellular adhesion molecule, tissue necrosis factor, transforming growth factor–beta 1 and matrix metalloproteinase-9 alternatively suggest that migraine attacks may directly activate the endothelium (23–27).
To reconcile the disparate conclusions from investigations of endothelial activation and vascular reactivity, we turn to the few migraine studies incorporating both components of endothelial dysfunction. Tietjen et al. found no difference in cerebrovascular reactivity between migraineurs and controls (unpublished data), although in migraineurs vWF activity levels were inversely related to reactivity as determined by TCD. Hamed et al. reported that in migraine, ET-1 correlated with both IMT and FMD, even though FMD did not differ between migraine and control groups (12). This raises the possibility that markers of endothelial activation are more sensitive than vascular reactivity to early changes of endothelial dysfunction. Prospective studies in migraine are needed to answer this question. Despite the conclusions that Perko et al. have made based on their well-executed study, the story of migraine and endothelial dysfunction still may have a number of additional chapters before it is complete.
Disclosure
The author has received research grants from GlaxoSmithKline and is on the Advisory Board of MAP Pharmaceuticals.