Objective: Computed tomographic angiography (CTA) is a rapid and minimally invasive method of detecting intracranial aneurysms. We wished to determine whether CTA could replace digital subtraction angiography (DSA) in the diagnosis and operative planning of ruptured cerebral aneurysms.
Methods: In a prospective study, patients with subarachnoid hemorrhage diagnosed by plain computed tomography underwent CTA, DSA, or both. Computed tomographic scans and CTA studies were first reviewed by the treating surgeon, along with a neuroradiologist, and a decision to proceed to DSA or directly to surgery was made on the basis of the type and quality of information provided by CTA. All patients underwent postoperative DSA.
Results: A total of 173 patients were studied. In 24 patients, both CTA and DSA were negative for a source of subarachnoid hemorrhage. Twelve patients underwent DSA without prior CTA because a technologist capable of performing CTA was not available when the patient was evaluated. Nine patients in poor neurological condition underwent CTA, and all tested positive for aneurysms but died without surgical intervention. Of the 126 patients who underwent CTA and surgery, 65 (52%) also required preoperative DSA. The decision to proceed to DSA after CTA was influenced by aneurysm location; posterior communicating artery (62%) and posterior circulation locations (67-75%) more commonly proceeded to DSA than middle cerebral artery aneurysms (34%; 0.025 > P > 0.01). The sensitivity and specificity of CTA for the detection of all aneurysms, ruptured and unruptured, in the group of patients who underwent both types of angiograms preoperatively were 84 and 100%, respectively. In the group of 61 patients in whom aneurysm surgery was performed on the basis of CTA results alone, the sensitivity and specificity for the detection of all aneurysms, as compared with postoperative DSA, were 90 and 100%, respectively. Missed aneurysms (n = 24) were always small (<4 mm) and were usually found in patients with multiple aneurysms in whom the larger, ruptured aneurysm was identified by CTA. In one patient, the aneurysm missed by preoperative CTA would have resulted in a different operation if detected preoperatively.
Conclusion: It is possible to proceed to ruptured aneurysm repair entirely on the basis of good-quality CTA studies that demonstrate an aneurysm consistent with the pattern of bleeding observed on plain computed tomography (48% of the patients in this series and most common middle cerebral artery aneurysms). However, detection of small unruptured aneurysms in patients with multiple lesions remains a problem.