Detection of Residual Brain Arteriovenous Malformations after Radiosurgery: Diagnostic Accuracy of Contrast-Enhanced Three-Dimensional Time of Flight MR Angiography at 3.0 Tesla.
10.3348/kjr.2009.10.4.333
- Author:
Kyoung Eun LEE
1
;
Choong Gon CHOI
;
Jin Woo CHOI
;
Byung Se CHOI
;
Deok Hee LEE
;
Sang Joon KIM
;
Do Hoon KWON
Author Information
1. Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul 138-736, Korea. cgchoi@amc.seoul.kr
- Publication Type:Original Article
- Keywords:
Brain;
Arteriovenous malformation;
Radiosurgery;
Magnetic resonance (MR);
Angiography
- MeSH:
Adult;
Contrast Media;
Female;
Humans;
Intracranial Arteriovenous Malformations/*diagnosis/*surgery;
Magnetic Resonance Angiography/methods/*standards;
Male;
*Radiosurgery;
Sensitivity and Specificity
- From:Korean Journal of Radiology
2009;10(4):333-339
- CountryRepublic of Korea
- Language:English
-
Abstract:
OBJECTIVE: Although three-dimensional time-of-flight magnetic resonance angiography (3D TOF-MRA) is used frequently as a follow-up tool to assess the response of arteriovenous malformations (AVMs) after radiosurgery, the diagnostic accuracy of 3D TOF-MRA is not well known. We evaluated the diagnostic accuracy of contrast-enhanced 3D TOF-MRA at 3.0 Tesla for the detection of residual AVMs. MATERIALS AND METHODS:This study included 32 AVMs from 32 patients who had been treated with radiosurgery (males/females: 21/11; average patient age, 33.1 years). The time interval between radiosurgery and MRA was an average of 35.3 months (range, 12-88 months). Three-dimensional TOF-MRA was obtained at a magnetic field strength of 3.0 Tesla after infusion of contrast media, with a measured voxel size of 0.40 x 0.80 x 1.4 (0.45) mm3 and a reconstructed voxel size of 0.27 x 0.27 x 0.70 (0.05) mm3 after zero-filling. X-ray angiography was performed as the reference of standard within six months after MRA (an average of two months). To determine the presence of a residual AVM, the source images of 3D TOF-MRA were independently reviewed, focusing on the presence of abnormally hyperintense fine tangled or tubular structures with continuity as seen on consecutive slices by two observers blinded to the X-ray angiography results. RESULTS: A residual AVM was identified in 10 patients (10 of 32, 31%) on X-ray angiography. The inter-observer agreement for MRA was excellent (kappa= 0.813). For the detection of a residual AVM after radiosurgery as determined by observer 1 and observer 2, the source images of MRA had an overall sensitivity of 100%/90% (10 of 10, 9 of 10), specificity of 68%/68% (15 of 22, 15 of 22), positive predictive value of 59%/56% (10 of 17, 9 of 16), negative predictive value of 100%/94% (15 of 15, 15 of 16) and diagnostic accuracy of 78%/75% (25 of 32, 24 of 32), respectively. CONCLUSION: The sensitivity of contrast-enhanced 3D TOF-MRA at 3.0 Tesla is high but the specificity is not sufficient for the detection of a residual AVM after radiosurgery.
