Petrous apex’s dural arteriovenous fistula (DAVF) mimicking a pineal region tumour: a rare cause of Hakim triad
- 1 Neuroscience, Neurosurgery, Ospedale dell'Angelo, Mestre, Veneto, Italy
- 2 Neuroradiology, Ospedale dell'Angelo, Mestre, Italy
- Correspondence to Dr Nicola Cavasin; nicola.cavasin@aulss3.veneto.it
Abstract
A dural arteriovenous fistula (DAVF) is a dural-based shunt between meningeal arteries and meningeal veins, sinuses and/or cortical veins; they have been classified and named according to the location and the flow pattern. Petrous apex DAVFs are located where the petrosal vein penetrates the dura mater into the superior petrosal sinus; there are only few cases reported in the literature, they can show an aggressive behaviour (subarachnoid haemorrhage, severe brainstem oedema) with a high mortality rate. The described case is, to the best of our knowledge, the first case of a DAVF presenting with symptoms mimicking idiopathic normal pressure hydrocephalus. After worsening of gait impairment, memory loss and urinary incontinence an urgent CT of the brain showed hydrocephalus and a hyperdense mass in the pineal region mimicking a pineal tumour; an emergent digital subtraction angiogram showed a left petrous apex Borden type III DAVF. A transvenous embolisation was performed obtaining a complete obliteration.
Background
A dural arteriovenous fistula (DAVF) is a pathologic dural-based shunt between the meningeal arteries and meningeal veins or sinus or through the cortical veins.1 2
The most common subtypes of DAVFs are placed within the dura adjacent to the venous sinuses,3 but they might occur at any site, and rare cases of intraosseous-seated fistulas have been reported2–5; they have usually been classified and named according to the location of the involved sinus and their flow pattern (with particular regard to the presence or absence of retrograde flow to the brain6).
The DAVF’s aetiology is often linked with dural sinus thrombosis, venous hypertension, previous craniotomy and trauma, though many lesions are idiopathic, and clinically can manifest with a wide spectrum of symptoms.2 3
We report a singular case of a petrous apex aggressive DAVF, which was initially misdiagnosed as idiopathic normal pressure hydrocephalus (iNPH), for its clinical presentation, then as a pineal region tumour after the first imaging workout.
Case presentation
A man in his 50s was admitted to the neurological service because of a few months’ history of gait impairment (magnetic gait), memory loss and urinary incontinence. The personal clinical history of the patient showed previous Guillain-Barré syndrome, obstructive sleep apnoea syndrome, lymphedema of the right leg due to saphenectomy, hypertension, previous tracheotomy, previous gastric surgery for obesity.
Investigations
Suspecting iNPH, the patient should have undergone an elective cerebral magnetic resonance (MR), but he was prematurely referred to the emergency department because of rapid worsening of the magnetic gait, urinary incontinence, cognitive impairment (mental confusion and memory loss) associated with mild frontal headache with alternating phases of humour deflexion and euphoria. A CT of the brain showed triventricular hydrocephalus with initial cortical sulci effacement and a hyperdense mass in the pineal region mimicking a pineal tumour (figure 1A). The patient was then referred to the neurosurgery department. The neurological examination on admission showed Hakim triad, mild right dysmetria and mild right VII cranial nerve palsy. In order to better define the suspected pineal tumour, an MR of the brain was planned but the patient’s clinical condition suddenly worsened with increasing agitation and sudden loss of consciousness. A new CT scan of the brain was performed and revealed a dramatic intraventricular haemorrhage with signs of intracranial hypertension (figure 1B). The patient underwent an urgent external ventricular drainage and an MR study of the brain. The MR showed a large flow void area, resembling a vascular pouch at three level of the cerebellomesencephalic fissure, with a heterogeneous signal, predominantly hyperintense on T1-weighted images and hypointense on T2-weighted images with an ‘onion-shaped’ hyperintense rim, potentially consistent with the presence of partial thrombosis or wall inflammation (figure 2A); the aqueduct was compressed. An old MR (performed 6 months before and not previously available) retrospectively showed the same left-sided tegmental flow void although smaller and barely visible (figure 2B). A vascular malformation was then hypothesised and an emergent digital subtraction angiogram (DSA) was performed.
(A) CT scan on arrival to another institution showing a hyperdense lesion in the left cerebellomesencephalic fissure with mass effect on the ipsilateral posterior aspect of the tegmental plate and compression of the aqueduct and triventricular hydrocephalus. (B) CT scan after rapid deterioration revealing tetraventricular haemorrhage with signs of intracranial hypertension.
(A) MR, T2-weighted sequence, performed 6 months before in another institution showing a small left tegmental flow void. (B) MR, T2-weighted sequence, on admission showing enlargement of the venous pouch in the cerebellomesencephalic fissure with compression on the aqueduct.
The DSA revealed the presence of a left petrous apex DAVF with main feeders coming from the left meningohypophyseal trunk and accessory feeders from the left internal maxillary artery, middle meningeal artery, occipital artery and ascending pharyngeal artery (figure 3A–C). The single venous drainage was through the lateral mesencephalic vein and the vein of Galen; just before entering the vein of Galen, in the tegmental area, the venous collector showed a huge ectasia with irregular surface and a small bleb, responsible for the bleeding, whereas the calibre of the distal portion of the vein was slightly narrower than the proximal with a relative stenosis just before the confluence in the Galen vein (figure 3D). These features were consistent with a Borden type III DAVF that has a high rebleeding and complication rate and deserves a radical treatment.
Cerebral digital subtraction angiogram showing a left petrous apex dural arteriovenous fistula with main feeders coming from the left meningohypophyseal trunk (A) and accessory feeders from the left internal maxillary artery, middle meningeal artery, occipital artery and ascending pharyngeal artery (C,D) and single venous drainage through the lateral mesencephalic vein and the vein of Galen with a huge ectasia in the tegmental area (B).
Treatment
The case was discussed in the multidisciplinary team and an endovascular approach was chosen. Our first choice was a transarterial embolisation with embolic material through dural branches of the left external carotid artery that we considered safer than the ones coming from the carotid syphon (ie, the meningohypophyseal trunk, figure 4A); we thought we could find an accessible route to get as close as possible with the microcatheter’s tip to the fistulous point to tolerate a safe reflux. The patient was put under general anaesthesia. No intravenous heparin was administered at the beginning of the procedure. A long sheath was positioned at the distal portion of the left common carotid artery and a guiding catheter was brought at first in the internal maxillary artery through which we accessed with a detachable tip microcatheter the middle meningeal artery (figure 4B). We could not get close enough to the fistula with the microcatheter’s tip, therefore the injection of embolic material did not reach the target despite quite a long reflux. We then accessed the occipital artery with the guiding catheter and selectively reached with another detachable tip microcatheter the stylomastoid branch (figure 4C) and again injected embolic material. But we failed once more to reach the fistula. All the other feeders were judged too tiny and tortuous to be used to perform a successful and safe embolisation, so we decided to change strategy and use the venous route that was wider and less tortuous. We reached the jugular bulb with the distal tip of the long sheath and the torcular with the guiding catheter. We then navigated through the straight sinus, the vein of Galen and the lateral mesencephalic vein and accessed the venous portion close to the fistula (figure 4D) with a double marker microcatheter. We then embolised a short portion of this vein with five detachable coils with occlusion of the fistulous point and complete disappearance of the DAVF. No complications occurred.
Digital subtraction angiogram (DSA), embolisation procedure. (A) The meningohypophyseal trunk from the left carotid syphon was not deemed suitable for a safe embolisation. (B) The middle meningeal artery and (C) the stylomastoid branch from the left external carotid artery were catheterised but the embolic material did not reach the fistula. (D) Transvenous route (thin arrows) to reach the deep portion of the venous outlet (thick arrow). (E) The cast of coils in the venous deepest portion (thin arrows). (F) Final DSA showing complete angiographic disappearance of the dural fistula.
Outcome and follow-up
Despite prompt treatment, the patient experienced a short vegetative status and died 4 months later because of kidney insufficiency sustained by severe sepsis.
Discussion
DAVFs represent 15% of all intracranial vascular malformations7 and most of them are acquired lesions.
The DAVFs pathogenesis is thought to be a multistep phenomenon instigated by vein thrombosis and resultant obstruction of a dural sinus; it is often secondary to a hypercoagulable status induced by infection, trauma or iatrogenic stimulus such as surgical resection of lesions involving sinuses.8 Two hypotheses could explain the DAVF formation after dural sinus thrombosis: (a) local hypoxia triggers the production of angiogenic factors, with the formation of aberrant vessels and abnormal direct arteriovenous (AV) shunts9 10; (b) impaired venous drainage increases dural sinus pressure, which causes compensatory dilatation of physiologic shunts connecting the thrombosed sinus and the extracranial arterial system.9 11 Both hypotheses have in common an increase in venous pressure conducting the recanalisation of the dural sinus and arterial shunting; consequently, it can lead to impaired outflow of cortical veins, causing ischaemic or haemorrhagic injuries.9 11
As the venous drainage pattern is the most important prognostic factor, most of the DAVFs classifications are based on the description of this feature to determine the risk of an aggressive clinical course and the indication for treatment.6 Deep-seated DAVFs involve the skull base dura, the tentorium and the dural sinuses. They frequently have an aggressive nature with respect to risk of haemorrhage and corresponding neurologic deficits caused by fistula drainage into leptomeningeal veins12–14; in this case, treatment is indicated.
Three classification systems have been developed to describe DAVFs. The first classification system, established by Djindjian and Merland in 1978,15 is based on the venous drainage of the fistulae; Borden et al’s16 system is based on the site of cortical venous drainage as well as whether there are single or multiple fistulae present in the nidus. The Cognard system6 is an adaptation of Djindjian and Merland’s15 classification incorporating dural sinus flow directionality (retrograde or anterograde) and venous flow architecture.
DAVFs clinical presentation ranges from an incidental finding to severe neurologic deficits secondary to haemorrhage, ischaemia or venous congestion17; the spectrum of presentations depends on the pathophysiologic characteristics of the AV shunting, such as the recruitment of parenchymal venous drainage: headache, pulsatile tinnitus, haemorrhage, seizures and focal deficits.6
Petrous apex DAVFs, also known as superior petrosal sinus or tentorial apex DAVFs, are located where the petrosal vein penetrates the dura mater into the superior petrosal sinus18–20; it accounts for 26% of tentorial DAVFs, representing 3.6%–5% of all cerebral DAVFs20 21; their incidence is low, and there are only few cases reported in the literature.18 19 22 Because petrous apex DAVFs usually present retrograde drainage into the petrosal vein and its tributaries, they are usually classified as Borden type III. Among DAVFs, they show the most aggressive neurologic behaviour including subarachnoid haemorrhage (SAH) and severe brainstem oedema due to venous hypertension. They are burdened with a high mortality rate.20 21
Recently, Li et al 23 proposed a classification of the petrous apex DAVFs based on the vascular architecture, focusing the extent and position of venous ectasia: type I—no venous ectasia; type II—venous ectasia with normal vein proximal to the fistula; type III—venous ectasia at the site of the fistula. The author showed that the draining vein morphologic changes were related to clinical characteristics, treatment and prognosis.23 While type I fistulas have generally fewer feeders and low blood flow and are easily treated both by surgical and endovascular approach, type III can seldom be completely and safely cured with a single-step procedure and need combined treatment. According to this classification our patient presented a type II petrous apex DAVF with drainage in the veins of the posterior mesencephalic compartment.
With special regard to the clinical presentation, the described case is, to the best of our knowledge, the first case of a DAVF presenting with symptoms mimicking iNPH. Li et al 23 reported, in a surgical series composed of 64 cases, that non-haemorrhagic neurologic defects are the most frequent symptoms, associated with brainstem oedema in 57.8% of cases. In 18.75% of cases, patients showed intermittent headache, in 6.3% trigeminal neuralgia, SAH occurred in 12.5% of cases and 4.7% were asymptomatic. Stiefel et al 8 presented a series of four cases: one patient showed intracerebral haemorrhage, one was admitted to the hospital with pulsatile tinnitus and two patients presented sudden facial nerve and trochlear nerve palsy, respectively.
The DAVFs’ haemorrhage rate and rebleeding rate may reach 6%–10% and 35%–46% per year, respectively.21 24 The reported annual mortality and morbidity rates of untreated or partially treated aggressive DAVFs were up to 10.4% and 15.0%, respectively.8 20 25–27 Consequently, complete obliteration of the lesion is of paramount importance, but it also represents a medical challenge: the strong tendency of subtotally occluded DAVFs to recur (up to 43%) is explained by reclutation of more feeding arteries exacerbating the hypertension of the draining vein and increasing the chances of critical complications (up to 55%).22 24 25
Various treatment options have been proposed, alone or in combination. The complete interruption of the AV shunt represents the best treatment: transarterial and/or transvenous endovascular embolisation is the first choice, surgical occlusion of the venous drainage or surgical occlusion of feeding arteries or draining veins, with or without complete excision of the fistula-carrying dural segment, is the main alternative treatment choice, stereotactic radiosurgery has been proposed.28–32
The peculiar features of the reported case were the symptoms of onset, quite uncommon for this DAVF subtype and the radiological features. The presence of a conspicuous expansile lesion in the pineal region misled the initial diagnosis leading to think about a pineal tumour.
Learning points
-
The differential diagnosis between idiopathic normal pressure hydrocephalus and pineal tumour is often clinically challenging and potentially misleading.
-
A thorough interpretation of a complete imaging workout is mandatory to fulfil the right diagnosis and plan the proper and timely treatment.
-
The reported case shows a unique presentation of a petrous apex dural arteriovenous fistula (DAVF), a rare entity characterised by aggressive behaviour that led to dramatic consequences in a few months. We suggest to take into account the petrous apex DAFVs as a possible cause of Hakim triad.
Ethics statements
Patient consent for publication
Footnotes
-
Contributors BG: analysed clinical data, reported the data found in the literature and wrote paper. NC: reported treatment and described pictures. MR: planning and interpretation. SM: acquisition of data, planning and supervision.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Case reports provide a valuable learning resource for the scientific community and can indicate areas of interest for future research. They should not be used in isolation to guide treatment choices or public health policy.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2022. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer