Aspects of migraine and patent foramen ovale in ischemic stroke

Sammanfattning: Stroke is one of the main causes of death and disability worldwide, and only in Sweden approximately 25 000 individuals suffer from stroke each year. This thesis focuses on two common conditions; migraine and patent foramen ovale (PFO) and their role in stroke. Migraine is a common primary headache disorder, affecting approximately 11-13 % of the population with a 3:1 female preponderance. One third of the patients have an initial aura of neurological symptoms before the headache, and clinically migraine can be divided into migraine with and without aura. The pathophysiological mechanisms are highly complex, and involve cortical spreading depression (CSD) as the substrate for migraine aura, and activation of the trigeminovascular system causing the headache. PFO is an inborn anomaly, a remnant from the fetal circulation, that is prevalent in approximately 25 % of the population. PFO enables shunting of venous blood to the arterial circulation, bypassing the pulmonary system and enables paradoxial embolization. PFO is associated with ischemic stroke, as well as migraine with aura. The reason for the latter is unknown, but may relate to micro-embolisms through a PFO triggering migraine attacks. For decades, migraine has been suggested as a risk factor for stroke and cardiovascular disease. The risk seems to be mostly related to migraine with aura, female gender and young age. The reasons for this are still unknown and probably multifactorial. Different theories involve increased prevalence of cardiovascular risk factors among migraineurs, co-existence of other co-morbid conditions increasing the risk for stroke (i.e. PFO and cervical artery dissection), and association to endothelial dysfunction with subsequent hypercoagulability and decreased vascular reactivity. Migraine with aura may also be associated with a phenotype, leading to an increased susceptibility for CSD, and an increased sensitivity to cerebral ischemia. This thesis involves four different projects concerning PFO and migraine in relation to stroke. The projects are performed in stroke populations (Study I-III) and in a population-based twin sample (Study IV). In Study I, patients with ischemic stroke investigated with transesophageal echocardiography were included (N=117), and dichotomized depending on the co-existence of a PFO. The prevalence of PFO was 11.9 %. Patients were analyzed regarding cardiovascular risk factors and allele frequency of 100 different genetic markers, previously associated with cerebrovascular disease. Four genetic markers, located in the Prothrombin-, Selectin E- and Apolipoprotein C III- genes, were significantly associated with PFO. The strongest association was for Prothrombin 20210 G/A (p= 0.0049), which is a marker associated with increased risk of venous thromboembolism. There were no differences regarding risk factors in the two groups. In Study II, patients with a diagnosis of TIA, ischemic or hemorrhagic stroke, admitted to the stroke ward during a six-month period, were included (N=175). Prevalence of migraine was investigated using a structured questionnaire, and patients were analyzed regarding cardiovascular risk factors and clinical characteristics of their stroke, depending on co-existing migraine or not. The prevalence of migraine was 20 %, which is comparable to prevalence estimates in the general population. However, migraine with aura was more prevalent than expected (61 % of those with migraine). Migraine was associated with PFO. Most stroke patients had a favorable outcome after stroke and there was no difference in stroke severity depending on migraine status. In Study III, patients with cryptogenic stroke and PFO, planned for closure of their PFO (N=20) or on medical treatment only (N=7), were included. Patients were followed prospectively from baseline to six month after closure. Fifty percent of the patients had co-existing migraine, whereof the majority had migraine with aura (84.6 %). Endothelial dysfunction was assessed at baseline and after one day, one month and six months. A majority of patients had an impaired endothelial function at baseline, but there was no change after PFO closure. In a few patients, migraine frequency was impacted after closure with a distinct increase in migraine attacks after the procedure. At a second, long-time follow-up, the majority of patients were improved regarding frequency of migraine attacks. In Study IV, the risk for stroke related to migraine was investigated in a population from the Swedish Twin Registry (N=53 404). A diagnose of migraine with or without aura was identified through a symptom based algorithm. The twins were followed longitudinally for more than 10 years for the outcome of stroke, using data from national patient registries. There was no general increased risk for stroke associated with migraine, but twins with migraine with aura had a border-significant 27 % increased risk for stroke. Further analysis suggested that this could be impacted by familial factors. However, in comparison with previous studies, our results showed a considerably weaker association to stroke related to migraine.