Review: Computed tomographic angiography and magnetic resonance angiography accurately detect intracranial aneurysmsPDF
ACP J Club. 2001 May-June;134:113. doi:10.7326/ACPJC-2001-134-3-113
White PM, Wardlaw JM, Easton V. Can noninvasive imaging accurately depict intracranial aneurysms? A systematic review. Radiology. 2000 Nov;217:361-70. [PubMed ID: 11058629] (All 2001 articles were reviewed for relevancy, and abstracts were last revised in 2007.)
In patients with suspected intracranial aneurysms, how do 3 noninvasive diagnostic methods (computed tomographic angiography [CTA], magnetic resonance angiography [MRA], and transcranial Doppler ultrasonography [DUS]) compare with intra-arterial digital subtraction angiography (DSA)?
Studies were identified by searching MEDLINE and EMBASE/Excerpta Medica (1988 to 1998). Reference lists of studies and review articles were checked, and journals not indexed in MEDLINE or EMBASE/ Excerpta Medica that appeared in the reference lists were hand searched.
Studies in any language were selected if they were published between 1988 and 1998, a noninvasive imaging examination was compared with DSA for diagnosing intracranial aneurysms, and ≥ 10 patients received both examinations.
2 reviewers independently reviewed the studies that met initial criteria by using a standardized quality assessment form with 26 to 27 items relevant to studies of diagnostic performance, including study design and examination methods, image-review process, and presentation of results. Studies had to score ≥ 50% to be included in the meta-analysis.
38 studies (1765 patients) met the inclusion criteria and scored ≥ 50% on the quality assessment scale. 14 studies compared CTA with DSA, 18 compared MRA with DSA, 2 compared both CTA and MRA with DSA, and 4 compared DUS with DSA. Blinding of reviewers was explicitly stated in 34 (89%) of the studies. CTA and MRA each did well; the receiver operating characteristic (ROC) curves did not differ for either on a per-patient (P = 0.411) or per-aneurysm (P = 0.09) basis. Insufficient data existed to construct an ROC curve for DUS. The test characteristics for CTA, MRA, and DUS are in the Table.
In patients with suspected intracranial aneurysms, computed tomographic angiography and magnetic resonance angiography perform similarly and with high sensitivity and specificity.
Sources of funding: British Brain and Spine Foundation and Medical Research Council (UK).
For correspondence: Dr. P.M. White, Department of Clinical Neurosciences, University of Edinburgh, Bramwell Dott Building, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, Scotland, UK. FAX 44-131-332-5150.
Table. Pooled test characteristics for detecting intracranial aneurysms*
|Test||Basis||Number of studies||Sensitivity (95% CI)||Specificity (CI)||+LR||-LR|
|CTA||Per patient||11||92% (89 to 95)||94% (88 to 99)||15.8||0.09|
|Per aneurysm||16||90% (88 to 92)||86% (79 to 91)||6.32||0.12|
|MRA||Per patient||18||87% (84 to 90)||92% (88 to 94)||10.3||0.14|
|Per aneurysm||18||87% (84 to 90)||95% (91 to 97)||16.6||0.14|
|DUS||Per aneurysm||2||82% (67 to 92)||70% (35 to 93)||2.73||0.14|
*CTA = computed tomographic angiography; DUS = transcranial Doppler ultrasonography; MRA = magnetic resonance angiography. Diagnostic terms defined in Glossary; -LR calculated from data in article.
Most of the studies identified for the review by White and colleagues were those with a high prevalence of aneurysm. In these cases, investigation almost always involves the use of angiography to identify aneurysms. Angiography is especially useful in planning endovascular treatments. In the acute phase of subarachnoid hemorrhage, MRA and CTA are not useful for identifying small intracranial aneurysms because such aneurysms can be obscured by subarachnoid blood. If multiple aneurysms exist, however, CTA and MRA can often help identify which aneurysm bled.
In assessing asymptomatic patients at increased risk for intracranial aneurysms, the situation is less clear. As the authors point out, with decreasing prevalence of aneurysms, the sensitivity and specificity of noninvasive testing is likely to be lower than that reported in their review. The risk for rupture in the International Study of Unruptured Intracranial Aneurysms (ISUIA) was lower than previously thought, particularly for small aneurysms (1). Furthermore, the morbidity and mortality associated with treatment of asymptomatic aneurysms are not insubstantial. The ISUIA study showed a combined morbidity and mortality of 13.1% to 15.7%, although this risk may be reduced with endovascular techniques (2). Asymptomatic patients should be assessed at a specialist center, where patient attitudes and such factors as age, smoking history, and hypertension that influence the risk for rupture can be considered, rather than introducing blanket screening that is unlikely to be cost-effective (3, 4).
Developments in noninvasive imaging—including the use of contrast-enhanced, 3-dimensional subtraction MRA; multislice CTA; and contrast-enhanced power DUS—are likely to further increase the accuracy of noninvasive vascular imaging. These methods need to be rigorously evaluated in the presence of a symptomatic aneurysm and in the screening setting.
Steven M. Thomas, MBBS, MSc
Sheffield Vascular Institute
Sheffield, England, UK
1. International Study of Unruptured Intracranial Aneurysms Investigators. Unruptured intracranial aneurysms-risk of rupture and risks of surgical intervention. N Engl J Med. 1998;339:1725-33. [PubMed ID: 9867550]
2. Brilstra EH, Rinkel GJ, van der Graaf Y, van Rooij WJ, Algra A. Treatment of intracranial aneurysms by embolization with coils: a systematic review. Stroke. 1999;30:470-6. [PubMed ID: 9933290]
3. Yoshimoto Y, Wakai S. Cost-effectiveness analysis of screening for asymptomatic, unruptured intracranial aneurysms. A mathematical model. Stroke. 1999;30:1621-7. [PubMed ID: 10436111]
4. Crawley F, Clifton A, Brown MM. Should we screen for familial intracranial aneurysm? Stroke. 1999;30:312-6. [PubMed ID: 9933265]