Computed tomographic angiography versus digital subtraction angiography for the postoperative detection of residual aneurysms: a single-institution series and meta-analysis.

Publication/Presentation Date

5-1-2012

Abstract

BACKGROUND: Computed tomographic angiography (CTA) has recently emerged as a non-invasive alternative to digital subtraction angiography (DSA) for the detection of residual cerebral aneurysms (RA).

OBJECTIVE: To compare the diagnostic accuracy of CTA with the current 'gold standard', DSA, in the postoperative detection of RA.

METHODS: Patient data from this single institution were prospectively gathered, and imaging results retrospectively blinded and analyzed. Between 2001 and 2005 eligible patients received microsurgical repair of cerebral aneurysms and were evaluated postoperatively by DSA and CTA. These single-institutional data were compiled with qualified studies published from 1997 to 2009, and a meta-analysis was performed.

RESULTS: This institutional series reports sensitivity, specificity, positive (PPV) and negative predictive values (NPV) of 100%. Eleven studies met the inclusion criteria for the meta-analysis. A total of 427 patients with 513 aneurysms were included, with 61 RA detected by DSA and 40 detected by CTA. Unweighted analysis resulted in pooled sensitivity of 73.8%, specificity of 96.3%, PPV of 91.0% and NPV of 86.1%. Stratified analysis of studies using 16-slice CTA versus 2D DSA reported pooled sensitivity of 92.6%, specificity of 99.3%, PPV of 95.8%, and NPV of 97.8%.

CONCLUSIONS: This meta-analysis supports CTA as an acceptable modality for postoperative detection of RA, although DSA remains the gold standard. By implementing multidetector CTA technology in experienced centers, the sensitivity and specificity of CTA may approach that of traditional DSA for detecting RA. As a cost-effective, non-invasive modality, CTA is a promising alternative to DSA for initial and long-term evaluation of RA.

Volume

4

Issue

3

First Page

219

Last Page

225

ISSN

1759-8486

Disciplines

Medical Sciences | Medicine and Health Sciences | Neurology

PubMedID

21990495

Department(s)

Department of Medicine

Document Type

Article

Share

COinS