gms | German Medical Science

Objective: In approximately 15% of patients with spontaneous subarachnoid hemorrhage (SAH) no bleeding source can be identified in the initial digital subtraction angiography (DSA). Aim of this study was to present our experience with the diagnostic yield of MR-imaging in these "idiopathic" SAH-patients with negative DSA and to ascertain the results of repeat imaging.

Methods: Between 10/2012 and 11/2017 73 patients with non-aneurysmal SAH (NASAH) were treated at our institution. NASAH was defined as primary angiogram-negative, spontaneous SAH. Patient data were prospectively entered in a computerized database. All patients with NASAH diagnosed by computed tomography (CT), or lumbar puncture – perimesencephalic (PM) and non-perimesencephalic (NPM)-SAH were included. We retrospectively analysed the spinal MRI findings and the results of the repeat imagings.

Results: Of the 73 patients with NASAH, 21 were diagnosed with PM-SAH (28,8%), 45 with NPM-SAH (61,6%), and 7 patients with CT-negative SAH (9,6%).

All patients underwent MRI of the cranio-cervical junction, 20 patients (27,4%) additionally underwent holospinal MRI.

MR-imaging of the spinal axis could detect a spinal bleeding source in 2 patients (2,7%): In one case an intraspinal aneurysm, in the other case a perimedullary angioma was found. Both patients initially complained about lumbar pain and had neurological deficits.

Regarding our follow-up examinations, 37 patients (50,7%) underwent a 2nd DSA, 7 patients (9,6%) underwent cMRI/MRA imaging. In none of these patients any bleeding source could be found.

Conclusion: SAH caused by spinal pathology is very rare. Our current data indicates that an expensive routine holospinal MRI in all patients with NASAH may not be recommended as standard procedure. Follow-up imaging should be done symptom-orientated and might be warranted in NPM-NASAH, as these patients seem to have a higher rate of spinal bleeding source, especially when presenting with neurological symptoms.