Cluster headache and middle meningeal artery dural arteriovenous fistulas: A case report

Abstract

Background

Of the multiple etiologies identified for symptomatic cluster headache, vascular origin is common; however, there are no known reports of arteriovenous (AV) fistulas. Here we report a case with typical presentation of cluster headache, which might be associated with middle meningeal AV fistulas.

Case report

The subject is a 49-year-old man with a 7-year history of episodic left-side cluster headache, consistent with the criteria in the International Classification of Headache Disorders, 3rd edition, beta version (ICHD-3β). Magnetic resonance angiography (MRA) demonstrated dural arteriovenous fistulas at the left posterior fossa, which is supplied mainly by the left middle meningeal artery. After endovascular balloon-assisted embolization, his symptoms have disappeared without relapse for 1.5 years.

Discussion

We suggest that middle meningeal arteriovenous fistulas should be considered as a potential differential diagnosis of secondary cluster headache, even when the initial clinical manifestations and treatment response are typical for primary cluster headache.

Introduction

Cluster headache is a primary headache disorder that manifests as excruciating pain over the ophthalmic branch of the trigeminal nerve with unilateral autonomic symptoms (1). Secondary causes – including arterial aneurysm, arteriovenous malformation, cerebral venous thrombosis, carotid dissection, pituitary tumor, meningioma, epidermoid tumor, and sinusitis – have been suggested to be responsible for symptomatic cluster headache and other trigeminal autonomic cephalalgias in 5% or more of patients (2). Because it can be difficult to differentiate primary versus secondary cluster headache based on clinical manifestations, neuroimaging studies are necessary to exclude secondary causes. Although it is difficult to demonstrate causality of underlying abnormalities and the cluster or cluster-like headaches, interrogating these secondary causes may provide better insight into the pathophysiology of primary cluster headache.

Cluster headache attacks were once regarded as being “vascular” in nature and as being related specifically to vasodilation of the ophthalmic artery on the affected side (3). Waldenlind et al. observed significant dilation of the ophthalmic artery ipsilateral to the pain site in magnetic resonance angiography (MRA) performed during cluster headache attacks (4). However, Dodick et al. performed a meta-analysis of cluster headache magnetic resonance imaging (MRI) studies (1986–1998) and concluded that these blood flow changes were only an epiphenomenon of trigeminal activation (3). Nowadays, cluster headache is considered to be a neurovascular headache associated with activation of the trigeminal-autonomic reflex (3). Trigeminal nociceptors project signals through the trigeminal ganglion to the trigeminocervical complex, with projections to the thalamus, resulting in activation of cortical areas associated with pain transmission, including the somatosensory cortex, insula, and anterior cingulate cortex. A dysfunction or disturbance in the interactions between the posterior hypothalamus and these functionally connected pain areas might be associated with the attacks (5). Concurrent superior salivatory nucleus activation leads to a vasodilator reflex of the middle meningeal artery via parasympathetic innervation, which in turn augments peri-vascular trigeminal nociceptor signaling (5–6).

In the present study, we describe a case of symptomatically typical episodic cluster headache related to dural arteriovenous fistulas fed by the left middle meningeal artery. Our observations provide an interesting impetus for scientists to reconsider the origin of trigeminovascular pain.

Case presentation

A 49-year-old man with a 7-year history of episodic cluster headache was referred to our hospital for headache deterioration. This patient had no history of smoking, hypertension, cardiac disease, head trauma, or surgery. He had been diagnosed as having cluster headache with unilateral acute onset from the left aspect of the nose to the left orbital and retro-orbital region, radiating to the temporal-frontal area with severe pain (intensity of 8 on a 0–10 point numeric rating scale). During attacks, the patient reported experiencing restlessness and unilateral autonomic symptoms, including nasal congestion, lacrimation, conjunctival injection, and ptosis. No auras, such as vision disturbance or sensation change, nor migrainous symptoms, such as nausea, vomiting, photophobia, or phonophobia, were noted.

Typically, his attacks occurred at around 1 pm and 2 am and lasted for 30 ∼ 45 minutes. The frequency of the headaches could be up to 4–5 times a day. Each bout lasted for a month or two, with 15–20 headache days in a month. The first bout occurred when he was 42 years old. In the following years, cluster headache bouts occurred mostly twice a year, that is, in May to June or November to December. At times, there was only one bout in a year, but these bouts still preferentially occurred in these months. The duration of bouts experienced increased over time, with the latest bout lasting for more than 3 months. The same unilateral left-sided headache characteristics persisted, albeit with an increase in attack frequency to daily.

Cluster headache was diagnosed according to the International Classification of Headache Disorders, 3rd edition, beta version (ICHD-3β) (1). After an initial transitional therapy consisting of prednisolone (70 mg/d for 2 weeks), he was placed on verapamil (240 mg twice a day (b.i.d.)), dihydroergotamine (DHE) (5 mg b.i.d.), and lithium carbonate (300 mg b.i.d.). These medications completely abolished the patient’s acute attacks when he was regularly taking them. Oral DHE was used twice a day regularly rather than as acute abortive treatment. The patient had tried oxygen therapy as an abortive treatment in his previous bouts, which was also effective. However, prior to being referred to our hospital, he had tried a high flow of 100% oxygen inhalation, but it could only partially alleviate the pain in the most recent bout. Although the therapy suppressed the patient’s headaches, the pain recurred whenever the patient attempted to discontinue these medications. This was in great contrast to his treatment response for previous bouts, in which he could easily discontinue the medications within 1–2 months.

We arranged a brain MRI to exclude secondary causes of his cluster headache. The MRI revealed dural arteriovenous fistulas of about 1.3 cm in diameter and enlarged cortical venous drainage. Angiography demonstrated aggressive dural arteriovenous fistulas (Cognard type 4) at the left posterior fossa near the left transverse sinus, supplied mainly by the left middle meningeal artery with cortical venous reflux (Figure 1(a)). Angiography did not reveal any obvious abnormality in the sellar region or cavernous sinus.

Figure 1.

(a) Preoperative angiography revealed aggressive dural arteriovenous fistulas (Cognard type 4) of about 1.3 cm in diameter at the left posterior fossa near left transverse sinus, majorly supplied by the left middle meningeal artery with cortical venous reflux and enlarged cortical drainage. (b) Post-transarterial embolization angiography suggests the occlusion of dural arteriovenous fistulas by balloon-assisted ONYX (eV3, Inc. Irvine, CA. USA) embolization.

The patient was treated immediately after MRA with endovascular embolization. The aggressive dural arteriovenous fistulas were occluded completely by balloon-assisted ONYX (eV3, Inc. Irvine, CA. USA) embolization (Figure 1(b)). The patient’s headaches remitted following embolization, on the same day, and did not recur after discontinuing preventive medications. A follow-up MRA at 3 months demonstrated no recurrence of the dural arteriovenous fistulas. We followed the patient’s post-surgery clinical condition for 1.5 years, during which we observed complete remission without relapse.

Discussion

In this report, we describe the case of a patient who presented with typical episodic cluster headache features, but with a potential secondary cause, namely, middle meningeal arteriovenous fistulas. The patient’s headache attack phenomenology fulfilled the ICHD-3β criteria for episodic cluster headache (1). Moreover, he responded well to conventional preventive treatment for cluster headache for all his previous bouts. These medications remained effective for the most recent bout except that he was not able to be weaned from them this time. His symptoms were relieved dramatically following embolization, suggesting a pathophysiological link between the fistulas and the headache attacks.

Symptomatic cluster headaches are often secondary to diseases of the pituitary gland or sellar region. The location of some of these lesions may explain the circadian rhythm of attacks and the seasonality of cluster headache bouts, and also provide a reasonable potential explanation for the mechanism by which the pain could be evoked. Vascular lesions in the proximity of the cavernous sinus could lead to secondary cluster headache, but none related to the middle meningeal artery (which plays an important role in the trigeminovascular system) have been reported (7). Nomura et al. reported three patients with cavernous sinus dural arteriovenous fistula presenting with cluster-like headache, and they suggested that elevation of cavernous sinus pressure may trigger the trigeminal-autonomic reflex and thereby lead to headache attacks (8). Yamada et al. also reported a carotid-cavernous fistula case with presentation of migraine-like and cluster-like headache (9). A review of 63 cases of symptomatic cluster headache studies included five cases of arteriovenous malformation-/cavernous hemangioma (10). These peri-cavernous vascular lesions, similar to sellar and parasellar lesions, might have caused secondary cluster headaches by stretching adjacent pain innervating fibers in the ipsilateral trigeminal nerve. However, this is the first reported instance of cluster headache being attributed to dural arteriovenous fistulas fed by the middle meningeal artery, without involvement of the cavernous sinus or parasellar regions. A previous report demonstrating secondary cluster headache attributed to an arteriovenous malformation fed by the posterior cerebral arteries also supports the notion that sellar or parasellar involvement is not essential for the development of cluster-like headache (11).

The middle meningeal artery, which is a major dural artery, might link to cluster headache pathogenesis in a distinctive way. The middle meningeal arteries are sensitive to stimulation in a way that can evoke pain through innervating trigeminal nociceptor afferents (12). We suggest that middle meningeal artery dural arteriovenous fistulas may increase the pressure and caliber of the middle meningeal artery and its drainage vein, which would stretch the innervating trigeminal nerve and thereby elicit pain directly. It is unknown how vascular lesions may contribute to the cyclical pattern of cluster headache; however, the development of periodicity was reported previously in a patient with secondary cluster headache caused by an arteriovenous malformation (11). Interestingly, that patient responded well to oral dihydroergotamine treatment, the same as ours (11).

As for possible etiopathogenic explanation, we speculate that elevated local venous pressure may produce physiologic arteriovenous shunts that become pathological over time (13). Consequently, outflow obstruction may cause decreased cerebral perfusion and facilitate neoangiogenesis (13). The decreased oxygen supplement in a local cortico-meningeal area might lead to an accumulation of metabolic waste and upregulation of pro-inflammatory cytokines and inflammatory mediators in the angiogenic process, which might irritate perivascular trigeminal nociceptors, thereby activating the trigeminal-parasympathetic reflex and, ultimately, inducing cluster-like headache (Figure 2). The vasoconstrictive effects of dihydroergotamines that we prescribed for the patient might have reduced the shunting flow and thus the middle meningeal arterial pressure temporarily. However, it appears that untreated dural arteriovenous fistulas enlarged over time, progressing to a more severe grade and a worsening of symptoms.

Figure 2.

Proposed mechanism to illustrate the etiopathogenesis.

In summary, cluster headache can be secondary to various lesions, including but not restricted to sellar and parasellar regions. The present case suggests an association of episodic cluster headache with middle meningeal artery dural arteriovenous fistulas. We could not exclude the possibility of coincidence of these two diseases in this patient; however, that possibility seems very unlikely given the low incidences of both diseases. Moreover, the fact that the patient’s headache symptoms disappeared immediately after the fistulas’ treatment is consistent with causality. Although more evidence is needed to support this contention, it may be prudent to include detailed vascular studies in routine surveys of patients presenting with cluster headache or trigeminal autonomic cephalalgia. Moreover, our findings provide a different perspective with which to re-interrogate the role of the vascular contribution to cluster headache pathophysiology.

Consent

Written informed consent was obtained from the patient for the review of his records for this publication. A copy of his written consent is available for review.

Clinical implications

Secondary cluster headache could be associated with dural arteriovenous fistulas fed by the middle meningeal artery.

Neurovascular studies should be considered for patients presenting with trigeminal autonomic cephalalgia to exclude secondary causes.