Article Text

Download PDFPDF

Thunderclap headache, reversible cerebral arterial vasoconstriction, and unruptured aneurysms
  1. A B Singhal1
  1. 1Stroke Service, VBK-802, Massachusetts General Hospital, Boston, MA 02114, USA; asinghal{at}
    1. D W Dodick2
    1. 2Department of Neurology, Mayo Clinic Scottsdale, 13400 East Shea Boulevard, Scottsdale, Arizona 85259, USA; dodick.david{at}

      Statistics from

      Request Permissions

      If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

      In his comprehensive review of thunderclap headache,1 Dr Dodick discusses two patients with the triad of thunderclap headache, cerebral arterial vasoconstriction, and unruptured cerebral aneurysms. We recently reported on two very similar patients, in whom the symptoms developed shortly after exposure to commonly used serotonin enhancing drugs.2 The interrelation between thunderclap headache, cerebral arterial vasoconstriction, and unruptured aneurysms is not clear, and in these four patients the aneurysms may well have been incidental findings. However, it is interesting that, in addition to segmental vasoconstriction, cerebral angiograms in patients with the Call-Fleming and some other vasoconstriction syndromes3 can have areas of vasodilatation beyond the normal diameter of the artery. Moreover, patients with stroke associated with the use of vasoconstrictive drugs such as cocaine and “ecstasy” are known to have an unusually high number of aneurysms.4 It is conceivable that patients who develop cerebral vasoconstriction or thunderclap headaches (without subarachnoid haemorrhage) are more likely to harbour aneurysms due to primary or drug induced abnormalities of vessel tone.

      Dr Dodick reviews cases where thunderclap headache was associated with unruptured aneurysms, without cerebral arterial vasoconstriction, and where thunderclap headache was associated with vasoconstriction, without unruptured aneurysms. It should be noted that unruptured aneurysms have been associated with vasoconstriction, without thunderclap headache.5 This point is emphasised by an additional, hitherto unpublished case (courtesy Dr C Miller Fisher) of severe cerebral vasoconstriction, stroke, and death associated with two unruptured and asymptomatic intracerebral aneurysms without thunderclap headache. The patient, a 65 year woman, was admitted in January 1977 with probable Guillain-Barré syndrome. The hospital course was notable for episodic hypertension (maximum blood pressure 200/100 mm Hg). On day 4, she developed cortical blindness, abulia, aphasia, and right hemiplegia. Computed tomography showed infarctions in both occipital lobes and a parasagittal meningioma. A selective cerebral arteriogram showed aneurysms in the anterior communicating and left middle cerebral artery, severe attenuation of proximal intracranial arteries, and “sausaging” of distal arteries. After returning from the arteriogram the patient became obtunded, then deteriorated clinically, and died on day 13. A necropsy showed cerebral oedema with bilateral temporal lobe herniations, infarctions in the inferior cerebellum and both occipital and frontal lobes, a parasagittal meningioma, and two unruptured aneurysms (a 5 × 5 mm anterior communicating aneurysm and a 10 × 7 mm left middle cerebral artery aneurysm). There was no evidence for arterial inflammation. Lymphocytic infiltration was present in the sciatic nerves, consistent with infectious polyneuritis. As stated in the review article,1 it is difficult to account for any mechanism whereby the aneurysms may have precipitated the vasoconstriction.


      Author's reply

      I would like to thank Dr Singhal for his interest and thoughtful insights concerning the review article on thunderclap headache.1 I will address his comments in order.

      Firstly, I had already read with great interest the recent article from Dr Singhal et al2 regarding three patients with thunderclap headache, reversible vasospasm, and ischaemic stroke possibly secondary to exposure to serotonergic medications. He also correctly points out that the unruptured aneurysms found in some patients with thunderclap headache and reversible vasospasm are possibly incidental—a point that I made in the review article. On the basis of the association with unruptured aneurysms or exposure to sympathomimetic and serotonergic medications in some patients with thunderclap headache and vasospasm, he raises the provocative and interesting possibility that patients who develop thunderclap headache (without subarachnoid haemorrhage) are more likely to harbour aneurysms due to primary or drug induced abnormalities in vessel tone.

      There are certainly cases of thunderclap headache with reversible vasospasm that have occurred shortly after exposure to sympathomimetic medications such as cocaine or amphetamines, as well as during hyperadrenergic metabolic states such as eclampsia and hypertensive crisis.3–6 Most of the patients described in the literature, however, did not harbour intracranial aneurysms, and prospective longitudinal studies of patients with non-aneurysmal thunderclap have not found an increased risk of subarachnoid haemorrhage. Ideally, a longer prospective study of patients with thunderclap headache with cerebrovascular imaging or careful assessment of a large group of patients with unruptured aneurysms (such as the international unruptured aneuryms study) for a history of thunderclap headache would be required to address the hypothesis raised by Dr Singhal.

      Dr Singhal also suggests that unruptured aneurysms may present with vasospasm in the absence of a thunderclap headache. The case (courtesy of C Miller Fisher) that he uses to illustrate this point is a very interesting one. While it is certainly possible that the unruptured aneurysms in this case may have given rise to the vasospasm, I believe the vasospasm in this 65 year old woman with Guillian-Barré syndrome was more likely related to the severe labile hypertension—a result of dysautonomia frequently seen in this disease. As alluded to earlier, vasospasm has been well described in patients with acute hypertensive crises such as phaeochromocytoma, eclampsia, and hypertensive encephalopathy. Hypertensive encephalopathy with posterior leucoencephalopathy syndrome (PLES) was recently described in a patient with thunderclap headache,7 and I have just submitted a similar case for publication in a young woman who also had reversible vasospasm in the setting of a hypertensive crisis and PLES. In fact, it is possible that in many cases of drug induced cerebral vasospasm, the effect on vascular tone and calibre may reflect the effect of these sympathomimetic drugs on arterial blood pressure in addition to a direct vasoconstrictive effect of the drugs. Indeed, it would have been of notable interest to know the arterial blood pressure in the patients he described with reversible vasospasm and stroke in patients exposed to serotonergic medications, since the magnetic resonance imaging abnormalities in his patients are very similar to the changes seen in patients with PLES.