Dural sinus thrombosis: value of venous MR angiography for diagnosis and follow-up.

Abstract

The purpose of this study was to determine the value of venous MR angiography as the sole procedure for the diagnosis and follow-up of dural sinus thrombosis. Forty-two patients with clinical findings suggestive of dural sinus thrombosis were examined with venous MR angiography and spin-echo MR imaging. Maximum-intensity-projection reconstructions and individual sections of the MR angiograms were examined for direct and indirect signs of dural sinus thrombosis by assessing flow signal from the major sinuses, the jugular bulb, upper jugular veins, ascending cortical veins (occlusion or increased flow due to formation of collaterals), diploe (emissary) veins and extracranial veins, and the deep subcortical veins (Galen's and internal cerebral veins). Direct signs of dural sinus thrombosis on MR angiograms included lack of typical high flow signal from a sinus that did not appear aplastic or hypoplastic on single sections from MR angiography and the frayed appearance of the flow signal from a sinus after recanalization. Indirect signs of dural sinus thrombosis included evidence of formation of collaterals, unusually prominent flow signal from deeper medullary veins, cerebral hemorrhage, visualization of emissary veins, and signs of increased intracranial pressure. When available, conventional angiograms were evaluated by analogous criteria as appropriate. Digital subtraction or cut-film angiograms were available for correlation in nine patients. In nine patients, MR angiography was repeated up to eight times during the course of follow-up. The results of MR angiography for all patients were compared with results in 10 control subjects. Confirmation of the diagnosis of dural sinus thrombosis was based either on conventional angiographic findings or on the changes seen in follow-up examinations. Dural sinus thrombosis could be ruled out in 25 of the 42 patients on the basis of clinical and MR angiographic findings. In 17 patients with MR angiographic findings that indicated dural sinus thrombosis, conventional angiography confirmed the diagnosis in nine patients, and changes seen on repeat MR angiograms during follow-up confirmed the diagnosis in nine patients as well (one patient's diagnosis was confirmed by both techniques). Individual frames from two-dimensional fast low-angle shot sequences allowed direct visualization of thrombus. Limited spin-echo sequences as performed here provided inconsistent findings and were insufficient for diagnosis. In the 10 control subjects, attenuation of flow signal was seen in the torcular Herophili in all studies; one subject had a nonpathologic variant of the sinojugular system. MR angiography is the technique of choice for diagnostic evaluation and follow-up of dural sinus thrombosis, and it is reliable as the sole examination for this condition. When MR angiographic findings are unremarkable and other abnormalities must be ruled out, routine spin-echo MR imaging should be performed.