Comparison of Time-Resolved and First-Pass Contrast-Enhanced MR Angiography in Pretherapeutic Evaluation of Spinal Dural Arteriovenous Fistulas

S. Mathur; A. Bharatha; T.J. Huynh; R.I. Aviv; S.P. Symons

doi:10.3174/ajnr.A4962

Article Figures & Data

Figures

Tables

Download figure
Open in new tab
Download powerpoint
Fig 1.
Imaging features of SDAVF on TR-MRA. Representative images of the coronal MIP TR-MRA images in early arterial phase (A–C) demonstrate early draining vein (dashed arrow) and smudge of enhancement at the dural sleeve (arrowhead) connected to the radiculomeningeal artery (arrow) with progressive enhancement of the arterialized perimedullary veins (asterisk). Findings can be correlated on oblique frontal projection on DSA (D). E and F demonstrate correlation on source sagittal images; reference line on coronal images is marked by the hollow arrow. The radiculomeningeal artery at the dural sleeve (arrow, E) and prominent draining radicular vein (dashed arrow, F) are seen on sagittal images.
Download figure
Open in new tab
Download powerpoint
Fig 2.
Imaging features of SDAVF on FP-MRA. Coronal (A) and sagittal (B) MIP FP-MRA images demonstrate a prominent arterial network along the dural sleeve (arrowhead) connected to the radiculomeningeal artery (arrow), indicating the level of the fistula. A prominent radicular draining vein (dashed arrow) is noted extending superiorly from the same level. Findings correlated on oblique frontal projection DSA image (C).
Download figure
Open in new tab
Download powerpoint
Fig 3.
Top, Pitfalls in TR-MRA interpretation. Coronal MIP TR-MRA image (A) with enhancing structure at dural sleeve (arrow) mistaken for level of fistula and arterialized intradural vessel (dashed arrow) mistaken for radicular vein. On frontal projection DSA images (B and C) these structures are found to be a venous varix and artery of Adamkiewicz, respectively. Bottom, Pitfalls in FP-MRA interpretation. Sagittal T2WI (A) and enhanced sagittal T1WI (B) show abnormal signal intensity (arrow, A) and mild enhancement (arrow, B) with lower thoracic cord extending to conus. There are enhancing intradural vessels in this region (dashed arrow, B). On sagittal (C and D) and axial (E) FP-MRA images, arterialized structure at dural sleeve (arrow) and arterialized intradural vessel (dashed arrow) are noted. This was found to be the artery of Adamkiewicz on DSA (F). The cord lesion was later diagnosed as astrocytoma on pathology.
Download figure
Open in new tab
Download powerpoint
Fig 4.
“Curvilinear radicular vein at the edge of the FOV”—useful sign to suspect fistula inferiorly. Sagittal T2WI (A) and enhanced sagittal T1WI (B) show abnormal signal intensity (arrow, A) and mild enhancement (arrow, B) within lower thoracic cord extending in to conus with enhancing perimedullary vessels (dashed arrow, C). On sagittal (C) and coronal (D) FP-MRA images, site of fistula inferior to FOV was suspected due to the presence of prominent arterialized vessel at the lower end of the FOV (arrow). Fistula inferior to the FOV was confirmed on DSA (dashed arrow, E) with arterialized radicular vein (arrow, E).

Tables

Figures

View popup

Table 1:

Patient characteristics and frequency of conventional MRI findings

Characteristic	TR-MRA (n = 15)	FP-MRA (n = 31)	P
Age (median [IQR])	57 (54–65)	66 (56–74)	.152
Male sex (no. [%])	9 (60%)	24 (77%)	.219
Days between DSA and MRA (median [IQR])	6 (4–12)	5 (1–10)	.340
DSA evidence of SDAVF	10 (67%)	20 (64%)	.886
Level of fistula identified on DSA			.515
C6–C7	1	0
T1–T3	0	0
T4–T6	2	4
T7–T9	3	4
T10–T12	3	3
L1–L3	1	4
L4–L5	0	1
Sacrum/pelvis	0	4
Conventional MR signs in DSA-positive cases
Cord T2 hyperintensity	10/10 (100%)	20/20 (100%)	1.000
Serpentine flow voids	10/10 (100%)	20/20 (100%)	1.000
Cord enhancement	9/10 (90%)	20/20 (100%)	.333

Note:—IQR indicates interquartile range.

View popup

Table 2:

Diagnostic performance of TR-MRA and FP-MRA compared with DSA, and interobserver agreement

	TR-MRA (95% CI)	FP-MRA (95% CI)
Diagnosis of SDAVF
Sensitivity	100% (59%–100%)	100% (76%–100%)
Specificity	80% (28%–99%)	82% (48%–98%)
Positive predictive value	91% (59%–100%)	91% (71%–99%)
Negative predictive value	100% (28%–100%)	100% (55%–100%)
κ Agreement	1.0 (1.0–1.0)	1.0 (1.0–1.0)
Localization
Reader accuracy to within 1 level of DSA^{^a}	100% (72%–100%)	87% (62%–96%)
κ Agreement	1.0 (1.0–1.0)	0.87 (0.70–1.00)

↵a When SDAVF location within the FOV (n = 15); in 5 cases where the fistula was beyond the FOV, at least 1 of the observers suspected fistula outside FOV on MRA.