Abstract
INTRODUCTION
Multiple epidemiological studies cited in this issue of Headache Currents support the link of migraine and a higher risk for vascular disease, especially ischaemic stroke in women with migraine with aura before age 45 years. The precise mechanism by which migraine predisposes to stroke has not been clearly delineated, although studies show that the risk of stroke in women with migraine is increased with oral contraceptive use and smoking. Other potential factors increasing the risk of stroke in migraineurs include endothelial dysfunction, prothrombotic factors, vasospasm, hyperhomocysteinaemia and possible embolic mechanisms such as paradoxical embolus via a patent foramen ovale (PFO) or related to mitral valve prolapse. Migraine has also been associated with an increased risk of cervicocephalic dissection, which may lead to ischaemic stroke, and migraine frequently occurs years before clinical ischaemic brain disease in individuals with Notch 3 gene mutations associated with cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL). If these potential mechanisms were not complex enough, there is concern that medications such as triptans and ergots, used to treat migraine headaches, may potentially cause ischaemic brain or heart disease (angina and myocardial infarction).
Scher AI, Terwindt GM, Picavet HS, Verschuren WM, Ferrari MD, Launer LJ, Cardiovascular risk factors and migraine: the GEM population-based study. Neurology 2005; 64:614–20.
COMMENTARY
This population-based case–control study from the Netherlands examines the cardiovascular risk profile of individuals with MA, MoA and control subjects. Its findings are that migraineurs, especially those with aura, are more likely to have a higher-risk cardiovascular risk profile than individuals who do not have migraine. This includes a higher likelihood of smoking, reported family history (parents) of early myocardial infarction, unfavourable cholesterol profile, elevated BP and a history of early-onset CHD or stroke. In addition, women with aura were more likely to be using oral contraceptives than women with no history of migraine. The latter have been associated with hypercoagulability. This study also shows that individuals with MA were approximately twice as likely to have an elevated risk of coronary artery disease using the Framingham coronary heart disease risk score. Women with migraine were more likely to have received a diagnosis of gestational hypertension, and there is evidence to link this with hypertension later in life. As the authors point out, it is not possible to determine if migraine modifies cardiovascular risk factors or if cardiovascular risk factors modify the clinical expression of migraine. Nevertheless, this study does show that individuals with migraine (especially MA) are more likely to have risk factors that have been associated with cardiovascular disease. It will be important for the findings in this Dutch population be confirmed in other populations. Meanwhile, those who care for patients with migraine, especially those in primary care, should be aware that there is a likelihood of cardiovascular risk factors being higher than in those individuals without migraine and should appropriately screen these patients for modifiable risk factors. When hypertension, elevated cholesterol and smoking are present, intervention should be undertaken to modify them to reduce the future risk of vascular events.
Wilmshurst P, Nightingale S, Pearson M, Morrison L, Walsh K. Relation of atrial shunts to migraine in patients with ischemic stroke and peripheral emboli. Am J Cardiol 2006; 98:831–3.
This study investigated whether the increased incidence of stroke in young subjects with migraine is because they have an increased prevalence of atrial right-to-left shunts. The investigators report the prevalence of clinically relevant atrial shunts in those with stroke and migraine compared with those with stroke but without migraine, and also in historic control groups of subjects who had MA but no stroke and in population controls. Of 60 consecutive stroke patients, 42 (70%) had large- or medium-sized atrial shunts. Transcatheter shunt closure was performed in 39 patients, of whom 35 had a PFO (mean diameter 9.8 ± 4.1 mm) and four had atrial septal defects. If atrial shunts were unrelated to stroke in patients with migraines, shunt prevalence in those with migraine and stroke would be the same as in those with migraine but without stroke. However, a much greater shunt prevalence was found in those with stroke and MA (84%) than in those with MA but no stroke (38.1%, P < 0.001), population controls (12.2%, P < 0.001) and those with stroke but no migraine (55.6%, P < 0.05). Shunt prevalence was also significantly greater in patients who had stroke and MoA (75%) than in population controls (P < 0.001) and in those with MA but no stroke (P < 0.05). In conclusion, the increased incidence of stroke in subjects with migraine compared with the general population is because they have a higher prevalence of large atrial shunts and hence an increased risk for paradoxic embolism.
COMMENTARY
Several studies have demonstrated a higher occurrence of PFO in patients with MA. This study included 60 consecutive patients who presented with ischaemic stroke and underwent contrast echocardiography to seek evidence of possible paradoxic embolism. All patients were <66 years old at the time of presentation or, for those with more than one stroke, the first stroke was at < 66 years (mean age 42.7 ± 13.3 years and an age range of 11–65). Patients were excluded if carotid Doppler showed significant carotid artery disease, if transthoracic echocardiography showed a left-sided cardiac source of emboli, or 24-h ambulatory electrocardiography showed atrial fibrillation or flutter. As outlined above, shunt prevalence was greater in patients with stroke who had migraine, both MA and MoA, compared with population controls.
Although these results suggest that significant right-to-left shunts contribute to the increased stroke risk in individuals with migraine, there are shortcomings in this study. Clearly, peripherally located large vessel cerebral infarctions are more likely to be of embolic origin, but deep small infarctions are often due to small vessel disease. The authors provide no information regarding localization and, indeed, it is not clear whether the patients underwent neuroimaging. The authors do not detail other risk factors for ischaemic stroke such as diabetes mellitus, hypercholesterolaemia and hypertension which might predispose to ischaemic strokes via other mechanisms. In addition, the authors included patients who were >50 years old, when atherosclerotic mechanisms play a larger role in symptomatic cerebrovascular disease than at younger ages. All of these shortcomings may have led to misattribution of strokes to a paradoxical embolism.
Despite the shortcomings, the study lends credence to the hypothesis that paradoxical embolism is likely to account for some of the excess stroke risk in patients with migraine.
Anzola GP, Morandi E, Casilli F, Onorato E. Different degrees of right-to-left shunting predict migraine and stroke: data from 420 patients. Neurology 2006; 66: 765–7.
The authors analyse the extent of right-to-left shunting in patients with migraine, patients with cryptogenic stroke and controls. Patients with both migraine and stroke had larger shunts than those with migraine without stroke (P = 0.038), patients with no migraine with stroke (P = 0.007) and control patients (P < 0.0001). Patients with migraine have overall larger shunts than non-migraineurs, particularly if they have had a stroke. Right-to-left shunting may be causally related to migraine and to the increased stroke risk of migraine.
COMMENTARY
The authors of this retrospective study used contrast-enhanced transcranial Doppler studies to identify evidence of right-to-left shunts in patients and, as in the study by Wilmshurst et al., found that patients with migraine had larger shunts than non-migraineurs, particularly if they had suffered a stroke. They also conclude that large right-to-left shunts are more likely to be associated with stroke in patients with migraine. However, just as in the Wilmshurst study, the study can be criticized for lack of characterization of strokes and lack of control for other stroke risk factors. Nevertheless, the overall findings suggest that some of the excess stroke risk in migraineurs is attributable to larger right-to-left shunts. Even though this study and that of Wilmshurst et al. suggest that right-to-left shunts may contribute to increased stroke risk in migraine, there have been no randomized controlled trials that have demonstrated that closure of a PFO prevents ischaemic stroke in patients with migraine or is an effective treatment for migraine attacks.
Velentgas P, Cole JA, Mo J, Sikes CR, Walker AM. Severe vascular events in migraine patients. Headache 2004; 44:642–51.
COMMENTARY
This large retrospective study of members of a healthcare company in USA assessed whether migraine was associated with stroke or myocardial infarction and found, as in other studies, that migraine is a stroke risk, but no association was found with myocardial infarction. It did not control for independent stroke or cardiovascular risk factors. Importantly, it did show that triptan use was not associated with stroke or myocardial infarction. On the other hand, use of ergots was associated with stroke risk. As clinicians, we are frequently faced with treatment of acute migraine headaches. Triptans, as a group, are the most useful group of agents yet developed for this purpose. The appropriate concern that triptans have been rarely associated with cardiac or cerebral ischaemia often causes hesitation in providing these agents. This study is reassuring that there is no significant association of these agents with vascular events. This suggests that, in general, triptans can be prescribed safely. This should reassure clinicians regarding the safety of this class so that they will prescribe these agents, which have been shown to decrease migraine-related disability. Nevertheless, they are contraindicated in individuals with coronary artery disease, ischaemic cerebrovascular disease, uncontrolled hypertension, and those at high risk for such disorders.