By Dara Jamieson, MD

Associate Professor of Clinical Neurology, Weill Cornell Medical College

Dr. Jamieson reports she is a consultant for Roche.

SYNOPSIS: During spontaneous attacks of migraine with visual aura, magnetic imaging studies indicate that the blood-brain barrier remains intact and the pons is activated.

SOURCES: Hougaard A, Amin FM, Christensen CE, et al. Increased brainstem perfusion, but no blood brain barrier disruption, during attacks of migraine with aura. Brain 2017;140;1633-1642.

Moskowitz M. Holes in the leaky migraine blood-brain barrier hypothesis? Brain 2017:140;1537-1539.

Focal neurological symptoms precede approximately 20% of attacks of migraine-like head pain. The mechanism of this migraine aura is presumed to be related to cortical spreading depression (CSD), which is the propagation of neuronal depolarization, accompanied by sequential fluctuation in cerebral blood flow (CBF) and alteration in the blood-brain barrier (BBB) leading to efflux of inflammatory neuropeptides. The stimulation of the trigeminovascular system activates brainstem nuclei with triggering and perpetuation of the pain phase of a migraine headache.

The hypothesis of this study was that attacks of migraine with a typical visual aura would be accompanied by an increase in BBB permeability and in-brain hyperperfusion (increased CBF). Dynamic contrast-enhanced high-field (3 T) magnetic resonance imaging was used to measure BBB permeability and tissue perfusion simultaneously in multiple cerebral and brainstem areas during spontaneous attacks of migraine with aura. Eleven females and eight males underwent magnetic resonance imaging after the aura had resolved. No individuals were scanned while actually experiencing the auras, which lasted from 18-60 minutes. Headache intensity during scanning ranged from 0 to 9, with a mean of 4.6 on a 10-point pain scale. One person had an aura without headache. Imaging occurred between one and 22 hours (mean 7.6 hours) after the onset of an aura, that was visual in all subjects and sensory and/or aphasic in five individuals. The authors chose this time period from aura onset to scanning based on animal studies showing BBB breakdown from three hours, with a peak at 24 hours, following CSD. Scanning also was performed on a day remote from the attack of migraine with aura. Ten anatomic regions of interest on imaging were placed in the bilaterally in the calcarine sulci, lower pons, and in periventricular white matter areas in each vascular territory.

There were no statistically significant differences in BBB permeability (Ki) on migraine aura attack days as compared to attack-free days in any of the regions of interest in either the affected or the non-affected hemispheres. Mean CBF values were compared between attack and attack-free days in the 19 individuals with migraine aura. On a migraine attack day, CBF was increased in the calcarine cortex bilaterally, more on the affected side. An increase of CBF also was found in the posterior periventricular white matter only on the side of the affected hemisphere and in the lower brainstem bilaterally, greater on the perceived aura side. No differences in mean brain perfusion on attack and attack-free days were seen in the anterior and middle cerebral artery territories. No correlations were found between changes in Ki or CBF and the clinical measurement of time from symptom onset or headache intensity. The authors concluded that the findings of hyperperfusion in the brainstem after the aura, with preservation of the BBB, “contradict the preclinical hypothesis of cortical spreading depression-induced blood-brain barrier disruption as a possible mechanism linking aura and headache.”

Michael Moskowitz, MD, wrote a commentary on this research emphasizing the historical importance of the BBB in the explanation of migraine pathophysiology and noting that a permeable barrier may both allow noxious chemicals to trigger pain in the brain and facilitate drugs interacting with central nervous system receptors. How then to explain the finding that the BBB is not impaired after an attack of migraine without aura? Dr. Moskowitz noted that the imaging studies did not find disruption of BBB function “within the time window identified on the basis of CSD in an animal model.” However, he notes that the disruption of the BBB in human migraine may have a different time sequence, not measured by the current investigation.


Hougaard et al did not find differences in BBB permeability within hours to a day following a migraine aura, as compared to attack-free days, challenging the association of aura-associated CSD with the breakdown of the BBB. The other notable finding of this study, the activation of the pons during the migraine head pain after the aura, confirms other studies showing brainstem activation and reinforces the importance of brainstem nuclei in the activation and perpetuation of migraine pain.

This study, deflating the importance of the BBB in the pathophysiology of a migraine aura, indicates the complexity of the cascade of events occurring from the onset of the neurological symptoms of an aura to the triggering of head pain. As pointed out by Dr. Moskowitz in his commentary, disruption of the BBB is not necessary for the therapeutic benefit of CNS impenetrable agents, such as some 5-HT1 receptor agonists and CGRP receptor antagonists, in aborting migraine pain. These efficacious agents appear to target sites outside the BBB, decreasing its therapeutic importance in migraine. The timing of the scanning used to detect changes in the BBB may be critical. There are flaws in extrapolating animal models of migraine phenomena on to the human timeline; and serendipitous brain imaging during a migraine event has added immensely to the understanding of migraine pathophysiology. Before completely discounting the role of the BBB in the migraine aura, more imaging studies, with smaller and varied regions of interest, need to be performed during or in closer temporal proximity to the aura.