Residual flow after cerebral aneurysm coil occlusion: diagnostic accuracy of MR angiography.

BACKGROUND AND PURPOSE The purpose of this study was to estimate the performance measures of MR angiography (MRA) in the diagnosis of aneurysm residual flow after coil occlusion. METHODS Patients having at least 1 cerebral aneurysm treated with coil occlusion were prospectively and consecutively enrolled. Time of flight and contrast-enhanced MRA were performed the same day of the DSA follow-up. The degree of aneurysm occlusion and dimensions of the residual flow were evaluated by independent readers at MRA and digital subtraction angiogram. MRA performance measures were estimated in a cross-sectional analysis and repeated in subgroups of aneurysm sizes and locations. MRA predictive values for recurrence were also estimated using a longitudinal design. RESULTS We obtained 167 aneurysm evaluations for each imaging modality. Class 3 residual flow was seen on digital subtraction angiogram follow-up in 27%. The sensitivity and specificity of MRA was 88% (95% CI, 80-94) and 79% (95% CI, 67-88), respectively. The positive predictive value for a Class 3 recurrence was 67% (95% CI, 51-80) and the negative predictive value was 93% (95% CI, 86-97). Time-of-flight MRA underestimated the length of the residual flow (P=0.039), whereas contrast-enhanced MRA overestimated its width (P<0.0001). MRA sensitivity for a Class 3 residual flow was lower for aneurysms <6 mm (P=0.01). CONCLUSIONS MRA has sufficient accuracy for screening of aneurysm residual flow after coil occlusion. Due to its lower negative predictive value, recurrent aneurysms should be confirmed with digital subtraction angiogram before planning a retreatment. Routine use of MRA to follow small aneurysms should wait better estimation of its performance in this particular subgroup.