LAUSR

BACKGROUND In approximately 15% of spontaneous subarachnoid hemorrhage (SAH) patients, no bleeding source is found in the initial imaging. These patients can be categorized as either perimesencephalic (PM-SAH) or non-perimesencephalic (NP-SAH) subarachnoid hemorrhage patients. Follow-up imaging is routinely performed after NP-SAH to detect treatable etiologies; however, the optimal follow-up imaging protocol remains unclear. This study examines the optimal time interval to re-imaging and the performance of magnetic resonance imaging and angiography (MRI/A) in this setting. METHODS In this retrospective study, the records of NP-SAH patients treated at the University Hospital of Zurich (Switzerland) from 2005 to 2018 were analyzed. Clinical and radiological data were collected. Re-imaging data was grouped according to imaging modality and divided into three time-categories after bleeding: short-term (<2 weeks), medium-term (2-8 weeks) and long-term (>8 weeks) re-imaging. RESULTS Eighty-one NP-SAH patients were included. In 8 patients an aneurysm was diagnosed during re-imaging via digital subtraction angiography (9.9% diagnostic yield). Five aneurysms were detected at short-term in 81 patients (6.2% short-term yield) and three at medium-term re-imaging in 27 patients (11.1% medium-term yield). No aneurysms were found after 8 weeks in 56 patients. Five of these 8 patients also received MRI/A re-imaging, which was able to show the aneurysm in all 5 cases. CONCLUSIONS Our study emphasizes the importance of re-imaging in NP-SAH patients, which should be done both at short-term and at medium-term follow-up after the hemorrhage. Long-term re-imaging after 8 weeks might not be of diagnostic benefit. MRI/A might be considered as a possible noninvasive re-imaging modality in this setting.