Background: In patients with SAH and multiple aneurysms, the ruptured lesion must be identified to prevent recurrent bleeding.
Aim of the study: To assess the diagnostic value of non-enhanced computed tomography (NECT) in identifying the rupture site in patients with subarachnoid haemorrhage (SAH) and multiple aneurysms.
Material and methods: We included patients with SAH revealed by NECT and multiple aneurysms detected on computed tomography angiography (CTA) in whom a ruptured aneurysm was identified during neurosurgery. Two radiologists predicted the location of the ruptured aneurysm based on the distribution of the SAH and location of intracerebral haematoma (ICH) by NECT.
Results: Eighty-three patients with a mean age of 55.7 ± 14.4 years were included. Ruptured aneurysms were significantly larger (mean size 7.7 ± 4.7 mm) than unruptured aneurysms (mean size 5.9 ± 4.5 mm; p = 0.014). Interobserver agreement was 0.86 (p < 0.001). Overall sensitivity and specificity of radiological prediction were 78.3% (95% CI, 68.6%-87.1%) and 96.4% (95% CI, 94.3%-97.8%) respectively. Overall PPV and NPV were 78.3% (95% CI, 67.6%-86.3%) and 96.8% (95% CI, 94.8%-98.1%) respectively. The sensitivity and PPV for aneurysms in the anterior communicating, anterior, and middle cerebral arteries appeared to be significantly higher than in other locations (p = 0.015 and 0.019 respectively). Analysis of independent predictive factors of correct radiological location revealed that ICH predisposes to a correct radiological diagnosis with an odds ratio of 8.57 (95% CI, 1.07-68.99; p = 0.03).
Conclusions: NECT has a high diagnostic value in identifying the source of bleeding in patients with multiple aneurysms for anterior circulation aneurysms, especially with coexisting ICH. For other locations, NECT is not reliable enough to base treatment decisions upon.
Keywords: intracranial aneurysm; multidetector computed tomography; observer variation; subarachnoid haemorrhage.