Subarachnoid hemorrhage without evident cause on initial angiography studies: diagnostic yield of subsequent angiography and other neuroimaging tests.

OBJECT: The aim of this study was to assess the diagnostic yield of imaging tests performed in patients in whom the cause of subarachnoid hemorrhage (SAH) had not been demonstrated on initial angiography.
METHODS: By reviewing medical records of 806 patients with SAH who had been admitted during a 6.5-year period, the authors identified 86 in whom initial transfemoral catheter angiography failed to reveal the cause of SAH. Clinical and radiological data were analyzed to determine the diagnostic yield of subsequent catheter angiography, computerized tomography (CT) angiography, magnetic resonance (MR) angiography, and MR imaging of the brain and spine for various subtypes of SAH (bleeding not visualized on CT studies [CT-negative SAH], perimesencephalic SAH, and nonperimesencephalic SAH). Of 41 patients with nonperimesencephalic SAH, 36, 32, and 21 underwent repeated catheter angiography, CT angiography, and MR angiography, respectively; brain MR imaging was performed in 23 patients (18 with Gd and 15 with susceptibility contrast sequences), and spine MR imaging in 17. Of 36 patients with perimesencephalic SAH, 31, 23, and 17 underwent repeated catheter angiography, CT angiography, and MR angiography, respectively; brain MR imaging was performed in 18 patients (17 with Gd and 11 with susceptibility contrast sequences), and spine MR imaging in 14. Of nine patients with SAH not visualized on CT scanning, three, one, and six underwent repeated catheter angiography, CT angiography, and MR angiography, respectively; brain MR imaging was performed in eight patients (five with Gd and three with susceptibility contrast sequences), and spine MR imaging in seven. The cause of SAH could be determined in only four patients, all with nonperimesencephalic SAH. The only test that yielded a diagnosis was catheter angiography (three aneurysms on the second and one on the third angiography, all surgically secured). Diffusion-weighted MR imaging demonstrated small, deep infarcts in five patients.
CONCLUSIONS: Repeated catheter angiography remains the most sensitive test to determine the cause of SAH that is not demonstrated on initial angiography, particularly in the subtype of nonperimesencephalic SAH. Newer, noninvasive imaging techniques provide little diagnostic yield.