Dural arteriovenous shunt of the left transverse sinus

This 64 year old patient had 6 month history of progressive headaches followed by loss of vision.
At ophthalmological examination bilateral papilledema was found. Initial MRI examination revealed an empty sella. The diagnosis of normal pressure hydrocephalus was made.
As the patient's symptomatology rapidly progressed she was transferred to another hospital where an emergency CT scan was performed. Non-contrast enhanced images were read as normal, however intravenous contrast injection was carried out which showed abnormal vascular structures in the left parieto-occipital region.
MRI examination was repeated and included MRA at this time. Based on this study, the diagnosis of a dural arteriovenous fistula of the left transverse sinus was made. Subsequent catheter based cerebral angiography confirmed the diagnosis.
As an initial therapeutic measure, endovascular treatment was attempted by a venous approach (from the contralateral jugular vein through the torcular) and consisted of deposition of coils in the left transverse sinus. This treatment proved to be inadequate and hence was followed by surgical resection of much of the involved sinus resulting in progressive improvement in vision and resolution of headache.

(Dural AVS of the transverse sinus, 0.5 T)

Examination 1
Fig.1 Transverse non-enhanced CT images. This study was interpreted as normal.
Fig.2 Transverse contrast-enhanced CT images. Marked enhancement of the tentorium and the cortical veins in the left parieto-occipital region is noted. This was felt to reflect abnormal venous drainage (venous congestion) necessitating further evaluation by MRI-MRA.
Fig.3 Transverse proton-density and T2-weighted fast spin echo images. Note the empty sella, possibly reflecting chronically raised intracranial pressure.
Fig.4 Transverse proton density weighted fast spin-echo images. The abnormal vascular structures in the left parieto-occipital region are clearly seen.
Fig.5 Sagittal 2D PC MR angiograms of the left cerebral hemisphere with different velocity encoding values (from left above to right below: 10 cm/s, 20 cm/s, 40 cm/s, 80 cm/s). Optimal visualization of both the arterial and venous structures is obtained with the 40 cm/s Venc value. This is selected for the subsequent 3D PC MRA acquisition.
Fig.6 Coronal targeted MIP reconstructions with slight angular differences of the posterior fossa vascular structures from a 3D PC MRA acquisition data set. Partial visualization of the left transverse sinus and non-visualization of the left sigmoid sinus are demonstrated. Note also the segmental signal drop-out of the right transverse sinus (arrow) which should not be mistaken for thrombosis. In fact, this phenomenon is frequently and typically seen at the junction of the vein of Labbe and most probably reflects signal loss due to phase dispersion in turbulent flow.
Fig.7 Transverse targeted MIP reconstruction of the posterior fossa vascular structures from a 3D PC MRA acquisition data set. Same observations as on Fig. 6.
Fig.8 Sagittal targeted MIP reconstructions of the left (left) and right (right) cerebral hemispheres from a 3D PC MRA acquisition data set for comparison. The partial thrombosis of the transverse sinus and occlusion of the sigmoid sinus on the left side are demonstrated. An abnormal transcranial vascular supply from the left external occipital artery to the patent segment of the left transverse sinus is also suspected. Note the unusual cortical venous connection between the left transverse sinus and the superior sagittal sinus via the vein of Labbe and the vein of Trolard (arrows), serving as an accessory collateral venous drainage channel.
Fig.9 Lateral (above) and a-p (below) DSA images after selective injection of the left external carotid artery. A complex dural arteriovenous shunt with a thrombosed left transverse sinus is recognized, hence the MRI-MRA diagnosis is confirmed. As predicted, the arterial feeders arise from the external carotid artery branches and the shunt is drained towards the contralateral transverse sinus, via the torcular of Herophilus. The upward flow direction in the previously described temporo-parietal collateral venous channel is also appreciated.
Fig.10 Lateral DSA image after selective injection of the left external carotid artery (above) and sagittal targeted MIP image (below) from a 3D PC MRA acquisition data set of the dAVS of the left transverse sinus for comparison, showing good agreement between the imaging modalities.
Fig.11 DSA image (a-p view) after selective injection of the left external carotid artery (above) and coronal targeted MIP reconstruction (below) from the 3D PC MRA acquisition data set for comparison, showing good agreement between the two imaging modalities. Note again the appearance of the junction of the vein of Labbe with the right transverse sinus where the inflow of non-opacified blood results in incomplete segmental visualization of the transverse sinus. This is also seen as a segmental intravascular signal drop-out on the MR angiogram, but in this case is probably due to phase dispersion related to turbulent flow.
Fig.12 Lateral (above) DSA image after selective injection of the left internal carotid artery and lateral (middle) and a-p (below) images after selective injection of the right internal carotid artery. Direct tentorial meningeal branches arising from the internal carotid arteries also contribute to the blood supply of the dAVS. Vascular structures of such a small diameter and slow flow are impossible to identify on MR angiograms and therefore were not seen on the MR angiograms in this case.

Examination 2 (after surgical excision of the left transverse sinus)
Fig.13 Transverse proton density weighted fast spin echo images showing early post-operative changes in the left occipital region.
Fig.14 Transverse 2D PC survey MR angiogram. Non-visualization of most of the left transverse sinus, consistent with the surgical resection. A small residual sinus segment is suspected anteriorly (arrow).
Fig.15 Coronal (above) and transverse (below) targeted MIP reconstructions of the posterior fossa vascular structures from a 3D PC MRA acquisition data set confirming the presence of a small residual transverse sinus segment. Nevertheless, this result was felt to be satisfactory enough, as the remaining fistula appeared to be hemodynamically insignificant.

The ESNR CD-Rom Series