Abstract
Administration of sumatriptan in subarachnoid haemorrhage (SAH) patients, misdiagnosed as migraine patients, may induce symptomatic cerebral vasospasm with potentially dangerous consequences. Over a 5-year period, we observed three patients with a 3-15-year history of migraine, who received sumatriptan for acute headache. Two patients received 6 mg sumatriptan subcutaneously on days 4 and 6, and one patient 3 x 100 mg sumatriptan orally on day 1 after an acute headache episode. In all three cases, an alleviation of headache intensity from severe to moderate was observed. When headache recurred and meningeal signs appeared, SAH was diagnosed by computed tomography in all three cases. No neurological deficits occurred during the further course of the disease. In both patients with a SAH caused by an aneurysm, transcranial Doppler sonography demonstrated vasospasm of the basal cerebral arteries. An antinociceptive effect of sumatriptan can be observed in SAH patients in good clinical condition, which suggests a specific craniovascular antinociceptive action. This may lead to mis- diagnosis as migraine and delayed appropriate diagnosis and treatment.
Introduction
Despite the widespread availability of neuroimaging facilities, subarachnoid haemorrhage (SAH) is occasionally not recognized. One of the most common misdiagnoses is migraine (Table 1). It has been reported from US neurosurgical units that 25% of SAH patients, who were otherwise in good clinical condition, were initially incorrectly diagnosed. This risk is particularly high in patients without meningeal signs during the early phase after SAH and in those with a known history of recurrent primary headache (1, 2).
Clinical characteristics of headache in migraine and subarachnoid haemorrhage
At present, no published data exist on the frequency and effects of triptan treatment in SAH patients, nor has this problem been investigated in any animal model. From clinical observations and animal experiments, the role of sensory vasoactive neuropeptides in cerebral vasospasm after SAH has been discussed (3). Edvinsson et al. (4) hypothesized that the reduction of the release of calcitonin gene-related protein (CGRP) by triptans could be very dangerous due to induction or augmentation of vasospasm (VSP) after SAH, because CGRP plays an important role as a potent vasodilator, producing strong and sustained vasodilation. White and Sjaastad suggested testing the involvement of CGRP in situations of dangerous VSP in animal experiments or in cases of misadministration of sumatriptan after SAH (5).
We describe three SAH patients who received sumatriptan before the definite diagnosis was established.
Patients (see also Table 2)
Summary of demographic, clinical, sonographic and angiographic findings in two patients with a history of migraine without (patients 1 and 2) and with aura (patient 3) and a new, unrecognized subarachnoid haemorrhage, who received sumatriptan
F, Female; M, male; AcoA, A. communicans anterior, ICA, internal carotid artery; SAH, subarachnoid haemorrhage; VSP, vasospasm; TCD, transcranial Doppler sonography.
One patient had a history of migraine with aura, two without aura according to the International Headache Society criteria (6).
Case 1
A 40-year-old female patient with a 4-year history of recurrent headache without aura came to the emergency room after a severe attack of holocephalic headache with nausea. Meningeal signs were absent. No pain relief was achieved after the i.v. administration of 2.5 g novaminsulphon. One hour later she received 6 mg sumatriptan subcutaneously, which resulted in a significant reduction of pain [from 100 mm to 60 mm on a 100 mm visual analogue scale (VAS)]. Six hours later, headache intensity increased again and meningeal signs appeared. A computed tomographic (CT) scan revealed SAH, and a saccular aneurysm of the A. communicans anterior was demonstrated by angiography. Transcranial Doppler (TCD) and angiography showed no VSP. The aneurysm was successfully treated by endovascular coiling and neurological deficits were absent during the entire course of the disease.
Case 2
A 30-year-old male patient with a 3-year history of recurrent headache without aura suffered from recurrent severe headache on days 1 and 4. When headache was still severe on day 6, he had received an oral dose of 100 mg sumatriptan with some improvement of pain. Two further oral doses of sumatriptan some hours later were less effective. The patient was then sent to our emergency department. On admission, he had severe headache and showed meningeal signs. A CT scan disclosed a SAH, and clinically asymptomatic VSP was demonstrated by TCD on day 8. Angiography was performed on day 12, and clipping of a saccular aneurysm of the internal carotid artery was performed on day 13. The further course of the disease was uneventful.
Case 3
A 48-year-old female patient reported sudden-onset severe headache with vomiting while lying in bed. After slight improvement over the next few days, the headache suddenly worsened on day 4, and the patient was brought to our emergency department. She had a 15-year history of migraine with and without aura and an average frequency of five attacks per year. She had been treated routinelywith ergotamines and paracetamol. Upon admission, neurological examination revealed no abnormalities. No pain relief was achieved after the i.v. administration of 2.5 g novaminsulphon over the next 3 h. After receiving 6 mg of sumatriptan subcutaneously, pain intensity fell from 100 to 50 mm on a VAS 1 h after injection, and the patient fell asleep. The next day the headache persisted, but no meningeal signs were detected. At this time a CT scan disclosed a SAH over the left parietooccipital cortex. Angiography was performed on day 5 and no abnormalities were found. TCD demonstrated clinically asymptomatic slight VSP on days 10 and 12. On day 30 TCD was normal.
Discussion
Considering a 1-year migraine prevalence of 11% (7) in comparison with the annual SAH incidence of 10–15 cases/10 000 (8), it is obvious that the differential diagnosis between these two conditions is a common problem in clinical practice.
Since triptans are frequently used to treat severe migraine, the risk of misadministration in patients with SAH in otherwise good clinical condition and absent meningeal signs is evident. However, reports on this topic are, to the best of our knowledge, unavailable in the medical literature, probably due to underreporting.
Two dangerous consequences may arise from the misadministration of sumatriptan in SAH:
Sumatriptan may induce clinically symptomatic cerebral vasospasm. All patients in this study received sumatriptan in the early phase after SAH on days 1, 4 and 6 but remained free of neurological deficits during the phase of drug activity and the further course of the disease. In an attempt to explain this, one has to consider that VSP following SAH is not merely an effector-induced contraction of the smooth muscles in the arterial wall. Morphological, immunological and molecular studies in humans and animals point to inflammatory reactions in the arterial wall as a major cause of contraction, resulting in VSP (3). Reduction of the secretion of proinflammatory substances such as neurokinin A, substance P and CGRP by trigeminal sensory nerves is believed to be one of the possible antimigraine actions of triptans (9). So the anti-inflammatory effect of triptans may counterbalance a simple vasoconstrictive effect. As triptans bear the risk of symptomatic vasoconstriction, clinicians should keep in mind other differential diagnoses in each new headache attack, even in chronic migraineurs, before using them.
Another risk results from the reduction of pain from severe to moderate following the administration of sumatriptan. Assuming a migraine-specific action of sumatriptan, the misdiagnosis as migraine is inevitable. A close temporal relation between cerebrospinal vascular complications and triptan use has been described in several case reports, including spinal cord infarction (10), intracerebral haemorrhage (11), cerebral ischaemia (12) with (10, 12) and without (11) preceding vascular disease. SAH without an apparent source of bleeding on angiography was reported in one patient after sumatriptan application. In this patient a change of headache characteristics after sumatriptan use was observed (11). However, in all reported cases (10–12) neuroimaging studies were done after triptan use, and so far a definite causal relation between triptan use and the cerebrospinal vascular complication has not been established. In our patients a reduction of headache intensity and no change of headache characteristics were found, making a causative relation between sumatriptan use and the occurrence of SAH very unlikely. Asymptomatic vasospasm of the basal cerebral arteries was detected by TCD in two patients as early as 2 and 6 days after sumatriptan application and persisted at least for 8 days. Again, this temporal relation clearly favours a causal relation to SAH and not sumatriptan use.
Our observations indicate an antinociceptive effect of sumatriptan in headache resulting from SAH. The mechanism behind it is probably a specific craniovascular antinociceptive effect (9). Pain signal transmission from the meninges and the cranial arteries to the brain is mediated by the trigeminal nerve. Because of the highly specific trigeminal distribution of the 5HT1D-receptor in the CNS, 5HT1B/1D-receptor agonists are specific craniovascular antinociceptive substances.
With regard to a possible placebo effect of sumatriptan on SAH-related headache, two factors should be considered which render a placebo effect very unlikely: infusions of novaminsulphon preceding the administration of sumatriptan did not relieve pain in two out of three patients, and SAH-related headache does not seem to respond to a placebo effect as much as migraine headache does.
Conclusions
An antinociceptive effect of sumatriptan can be observed in SAH patients in good clinical condition. It is most likely to be explained by its specific craniovascular antinociceptive effect. This may cause misdiagnosis as migraine and a dangerous delay in diagnosis and treatment. Clinically asymptomatic vasospasm, most likely attributable to SAH, was found in two patients but no symptoms occurred. Nevertheless, triptan use carries the risk of arterial vasoconstriction. Attention should be payed to every single headache attack to avoid potential dangerous triptan misadministration in patients with diseases other than migraine.