Subarachnoid haemorrhage with negative tomographic angiography without vasospasm: is it necessary to perform angiographic controls?

Abstract

In about 15% of cases, the lesion causing subarachnoid haemorrhage (SAH) is not detectable by imaging studies, and some controversy exists as regards how the diagnosis should be performed in these patients. To analyse the diagnostic performance of digital subtraction angiography (DSA) carried out on a significant cohort of patients with SAH and negative tomographic angiography, without cerebral vasospasm, based on the distribution of bleeding in a simple CT scan. A retrospective analysis is performed of the centre's radiological database over an eight-year period. The inclusion criteria were patients with SAH of a perimesencephalic or diffuse distribution with a negative tomographic angiography result, without cerebral vasospasm, with at least one DSA control of the brain vessels. A total of 125 patients met the inclusion criteria: 61 had a perimesencephalic SAH and 64 had a diffuse SAH. The first DSA revealed vascular injury in seven patients (5.6%): one was in the perimesencephalic SAH group (1.6% yield) and six in the diffuse SAH group (9.4% yield). A second DSA was performed on 86 patients, a third on 28 and a fourth on five of them. The absence of vasospasm in the first DSA performed in patients with SAH and negative tomographic angiography could be a useful criterion for establishing an optimal follow-up protocol in these patients, without exposing them to unnecessary risks and radiation.