More Functional Imaging of Headache

Abstract & Commentary

Source: May A, et al. PET and MRA findings in cluster headache and MRA in experimental pain. Neurology 2000; 55:1328-1335.

In a previous study, may and colleagues re-viewed the early positron emission tomography (PET) findings of hypothalamic activation in patients with cluster headache (CH) (Neurology Alert 1998;17:19-20). The current paper by May et al builds on their important ongoing work. In this study, nine patients with active CH and eight patients with inactive CH were investigated with PET. Nitroglycerin (NTG) was used to induce CH in the nine during active periods. NTG will not activate a CH during a period of remission. PET findings fell into three broad categories: 1) activation in the ipsilateral posterior hypothalamic gray, an area specific to CH but not to other primary headache disorders; 2) activation in areas known to be involved in pain processing such as the cingulate cortex, insula, prefrontal cortex and contralateral thalamus; and 3) activation of extracerebral areas such as large intracranial blood vessels.

In addition, a single patient with spontaneous CH was studied with MRA. During this attack, a significant increase in blood flow in the ipsilateral carotid artery was demonstrated. In fact, ipsilateral carotid vasodilatation also was observed in this case when normal volunteers without CH were studied after local injection of capsaicin injection into the forehead. It appears that the pain, rather than a generator of the pain, drove the vascular changes. By contrast, in CH the primary disorder appears neuronal and instigated in the posterior hypothalamus. This region is most likely given the clinical circadian features of CH.


Functional imaging and the dynamic blood flow changes in the brain that are associated with the primary headache disorders such as CH and migraine are helping to evolve the concept of the vascular headache into something better thought of as the neurovascular headache. Such a concept better explains the fact that neuronal activation occurs prior to the pain and observed vasodilatation. In other words, contrary to previous and widely held dogma, the vascular changes may indeed be epiphenomenon. —Jeffrey Reich