|Home | About | Journals | Submit | Contact Us | Français|
Migraine often co-occurs with patent foramen ovale (PFO) and some have suggested surgical closure as an efficient treatment for migraine. However, prospective studies do not report radical effect of PFO surgery on migraine. Here we examined the hypothesis that PFO and migraine may co-occur as two independent manifestations of lateralization defect during embryonic development. We measured the absolute displacement of a midline structure, the pineal gland, on brain scans of 39 migraineurs and 26 controls. We found a significant asymmetry of the pineal gland in migraineurs compared with controls.
Our data suggest that migraine's circadian component and its association with PFO may be linked to a lateralization defect during embryogenesis, which could be a result from abnormal serotonin regulation.
Migraine is a disabling neurological disease that affects millions of people worldwide. While some aspects of its pathophysiology are becoming clearer, the underlying cause of migraine still remains a mystery.
A patent foramen ovale (PFO) is a frequent co-morbidity of migraine. The foramen ovale is a hole located in the atrial septum of the heart that remains open during fetal stage to allow fetal blood circulation to bypass the lungs, and that normally functionally closes at birth. In some case, the closure is incomplete, or even absent, creating a right-left shunt.
The co-morbidity of migraine and PFO was originally described in a study by Del Sette et al . Later, Wilmshurst reported that the closure of interatrial shunts in patients for decompression illness, stroke or large septal defect alleviated the pain episodes and frequency of migraine attacks in those who had a migraine history, in some cases even resulting in cessation of the attacks. Several studies have since then reported an increased PFO prevalence in migraineurs, especially in migraine with aura (for review, see )
Speculations have been advanced on the causality of PFO and migraine attacks. Two main hypotheses were put forward: First the shunt could allow micro-emboli to reach the brain circulation and provoke migraine (and white matter lesions as described by Kruit et al ). Alternatively, this shunt could allow substances (serotonin, norepinephrin) to bypass filtration by the lungs  and circulate through the brain, where they might trigger migraine attacks in predisposed subjects. These theories, however, experienced a serious setback after the large, randomized, placebo-controlled MIST study indicated that the closure of shunts does not have a desired effect on migrainous symptoms .
Here we examined the hypothesis that the link between PFO and migraine is not one of causality, but that these two conditions co-occur because they share a common etiology in embryogenesis. We propose that both conditions arise from a lateralization defect early in the fetal development.
Our hypothesis is based on the anecdotal evidence that PFO may result from a lateralization defect in embryogenesis. First, exposure to selective serotonin uptake inhibitors such as paroxetin during the first trimester of pregnancy has been linked to heart malformations , including septal defects . Second, serotonin exerts its effect through nodal signaling and the disturbances in the nodal-pathway have been linked to numerous lateralization defects, including atrioventricular septal defects  . Since serotonin is an important regulator of lateralization in the early development and plays a crucial role in heart morphogenesis , any deviations from the optimal serotonin levels may lead to lateralization defects of varying degrees. The link to migraine comes from the interplay of serotonin and nodal-signaling in brain embryogenesis: it has been shown in the zebrafish that abnormal nodal expression results in a displacement from the midline to a lateral position of the pineal gland .
Based on these observations, we explored the hypothesis that migraine may have a common etiology with PFO in the developmental pathway, and that both conditions may arise from abnormal serotonin levels during embryogenesis. We predicted that migraineurs have a higher incidence of pineal displacement than healthy controls, and that amongst migraineurs those with aura would have more displacement than those without aura.
We measured the distance between the center of the pineal gland in magnetic resonance images of migraineurs and controls. All patients were recruited from headache clinics in the area and by advertisement in the hospital. Each patient was screened with a detailed clinical interview. Exclusion criteria included pregnancy, breast-feeding, claustrophobia and any MRI incompatibility. The Hospital internal review board approved this study. Patients were classified in two groups, migraine with aura (MWA) or migraine without aura (MWoA) following the International Headache Society Classification.
Sixty-five participants were scanned in this study: 21 MWA (11 females, mean age=35.8±12.7), 18 MWoA (11 females, mean age=35.7±6.8), and 26 healthy controls (15 females, mean age=32±8.3). The study was conducted according to the Helsinki Declarations on human experimentation, and was approved by the Institutional Review Boards of the Massachusetts General Hospital.
Brain images were obtained and 3D reconstructed by two high-resolution magnetization-prepared rapid acquisitions with gradient-echoes (MP-RAGE) on a 3.0T Siemens Trio equipped with an 8-channel coil and on 1 1.5T Siemens Allegra equipped with a 23-channel coil (Erlangen, Germany). Both MP-RAGE sequences (1×1×1.3 mm, 128 slices, 256×256 matrix, echo time (TE)=3.45/3.31 ms; repetition time (TR)=2530/2500 ms; flip=7°) were motion-corrected and averaged to create one image volume.
Images were registered in MNI space. The pineal gland was located, and the coordinates of its center were recorded by two observers blinded to the diagnosis.
A two-tailed t-test, Welch corrected was performed in the absolute distance in millimeters from the midline between migraineurs and controls. ANOVA were computed for the comparison between the three groups. One-tailed Mann-Whitney test was used to compare migraineurs subgroups, to take into account the non-Gaussian distribution of the data.
We found significant differences in the absolute pineal displacement from the midline in migraineurs compared with controls. Migraineurs (Mean, SEM [mm]): 0.65 ± 0.07, N=39; Controls: 0.38 ± 0.06, N=26; P value= 0.006 (t=2.85; df=62).
When performing an ANOVA on the three groups separately, we found that the means were significantly different (F=3.84, P=0.03). Pineal gland was significantly more lateralized in MWA compared with controls (P=0.049), and in MWoA compared with controls (P=0.012) but no difference was present between MWA and MWoA (P=0.47).
We considered pineal displacement as an index of lateralization defect during embryogenesis . We found a significant difference between migraineurs and controls in the amount of pineal displacement from the midline, supporting our hypothesis of a common developmental origin for migraine and PFO. Our second prediction, that the pineal displacement should be greater in MWA than MWoA, however, was not fulfilled and needs to be addressed in further studies.
A number of disorders, including bipolar disorders and cluster headache, exhibit seasonal behavior, and the role of the pineal gland in the chronobiology of our organism is well known. Bipolar disorders are frequently associated with migraine. Cluster headache is noteworthily accompanied by a higher prevalence of PFO .
Pineal dysfunction may contribute to migraine. Migraine has a strong chronobiological component: attacks usually take place between 4 AM and 9 AM, and migraine is linked to impaired sleep quality and disturbances in the sleep-wake cycle in children. In Arctic areas, migraine with aura has been reported to be more prevalent in the light season than in the dark.
The circadian clockwork, which generates endogenous rhythmic phenomena, relies on light-dark transitions. Light itself may play a role in migraine chronobiology because migraineurs have a different sensitivity of S-cones, the detectors of short wavelength as blue . Short-wavelength visible or blue light is the most efficient to suppress melatonin production by the pineal gland .
The pineal produces melatonin from serotonin, itself a derivative from tryptophan. A disturbance in the process of tryptophan to melatonin biosynthesis could lead to an imbalance of the two substances, leading to abnormal levels of both melatonin and serotonin. Serotonin production is increased by light , while melatonin secretion is inhibited by light .
Serotoninergic neurotransmission seems to be altered in migraine [17,18], and low serotonin may facilitate the activation of the trigeminovascular nociceptive pathway ; levels of melatonin on the other hand are pronouncedly reduced in migraineurs . These observations suggest that the control of serotonin-melatonin axis is disturbed both in migraine and bipolar disorder and may explain their co-morbidity.
Phosphatidylinositol 3-kinase (PI3K) is an enzyme mediating the effects of serotonin through 5-HT1A receptor activation. During embryogenesis, PI3K inhibits Nodal-induced cell differentiation , and abnormal levels of PI3K could lead to lateralization defect including PFO and pineal displacement. This in turn could later in life result the abnormal serotonin-melatonin pattern observed in migraine.
Serotonin receptors are the target of numerous anti-migraine drugs. Selective serotonin 1B/1D receptor agonists triptans act in the dorsal raphe, the periacqueductal gray, and the trigeminal nucleus caudalis . In addition, new data suggest a potential effect of triptans in the ventroposteromedial nucleus of the thalamus acting through 5-HT(1A/1B/1D) mechanisms .
Cortical spreading depression if the probable mechanism of migraine aura . The blockade of cortical spreading depression can be achieved by the action on inhibitory 5-HT1A receptors . In addition to stopping migraine attacks, dihydroergotamine is an efficient drug for migraine prophylaxis. Through their actions at 5-HT1A autoreceptors (in the dorsal raphe nucleus) and heteroreceptors (notably in the hippocampus), dihydroergotamine and its metabolite can exert an inhibitory influence on neuronal excitability which might contribute to their antimigraine prophylactic efficiency .
In conclusion, our study indicates that there might be a common embryologic origin between PFO and migraine, and these two ailments may not have a causal relationship. It is not known yet whether the abnormal serotonin or melatonin levels can induce the migrainous attacks directly or whether they are surrogate rather than specific markers of a deeper underlying pathology. Further studies should address our hypothesis of PI3K abnormalities in migraine and other chronobiological disorders.
We want to thank Dr Alexandre DaSilva for his assistance in data collection.
Work supported by NIH grant 5PO1 NS 35611-09
Publisher's Disclaimer: This PDF receipt will only be used as the basis for generating PubMed Central (PMC) documents. PMC documents will be made available for review after conversion (approx. 2-3 weeks time). Any corrections that need to be made will be done at that time. No materials will be released to PMC without the approval of an author. Only the PMC documents will appear on PubMed Central -- this PDF Receipt will not appear on PubMed Central.