Usefulness of time-resolved MR angiography in spinal dural arteriovenous fistula (SDAVF)-a systematic review and meta-analysis

Abstract

Spinal dural arteriovenous fistulas (SDAVFs) constitute the most common type of spinal vascular malformations. Their diagnosis requires spinal digital subtraction angiography (DSA), which is time-consuming, requires catheterizing many vessels, and exposes patient to a high radiation and contrast doses. This study aims to evaluate the usefulness of time-resolved MR angiography (TR-MRA) in SDAVF diagnosis. We performed a systematic review of the PubMed and EMBASE databases followed by a meta-analysis. TR-MRA was an index test, and spinal DSA was a reference. Of the initial 324 records, we included 4 studies describing 71 patients with SDAVFs. In 42 cases, TR-MRA was true positive, and in 21 cases, it was true negative. We found 7 false-positive cases and 1 false negative. TR-MRA allowed for shunt level identification in 39 cases. Of these, the predicted level was correct in 23 cases (59%), to within 1 level in 38 cases (97.4%) and to within 2 levels in 39 cases (100%). The diagnostic odds ratio was 72.73 (95% CI [10.30; 513.35]), z = 4.30, p value < 0.0001. The pooled sensitivity was 0.98 (95% CI [0.64; 1.00]), and the pooled specificity was 0.79 (95% CI [0.10; 0.99]). The AUC of the SROC curve was 0.9. TR-MRA may serve as a preliminary study to detect SDAVFs and localize the shunt level with sensitivity and specificity as high as 98% and 79%, respectively. Unless the TR-MRA result is unequivocal, it should be followed by a limited spinal DSA.

Keywords: Diagnosis; Digital subtraction angiography; Magnetic resonance angiography; Spinal dural arteriovenous fistula.

Fig. 1

Literature review search strategy. MR, magnetic resonance

Fig. 2

Forest plot of time-resolved MR angiography in spinal dural arteriovenous fistula diagnostic odds ratio. MR, magnetic resonance

Fig. 3

The funnel plot shows publication bias. Standard error of the diagnostic odds ratio (DOR) is plotted against the DOR. DOR is a measure of the effectiveness of a diagnostic test. Markers represent individual studies

Fig. 4

Forest plot of time-resolved MR angiography in spinal dural arteriovenous fistula sensitivity (A) and specificity (B). MR, magnetic resonance

Fig. 5

Summary receiver-operating characteristic (SROC) curve showing the relationship between the false-positive rate on the x-axis and sensitivity on the y-axis. The point estimate of the pair of sensitivity and false-positive rate is shown by a black, open circle. The red, dashed line represents the 95% confidence region. Blue triangles represent input data (results from particular studies)

Fig. 6

Images of the described illustrative case before (a–d) and after (e, f) surgical treatment. An initial sagittal T2-weighted MR image (a) displays thickened torturous vessels on the posterior surface of the thoracic spinal cord (a, thin arrows) and hyperintensive spinal cord (a, big arrow), indicative of myelopathy: these findings are suggestive of spinal dural arterio-venous fistula (SDAVF). A multiplanar reconstruction (MPR) in the coronal plane of time-resolved MR angiography (TR–MRA, b) reveals a SDAVF, which is consistent with spinal digital subtraction angiography (DSA, c). A maximum intensity projection (MIP) reconstruction in the oblique plane (d) before and a subtracted source image (e) after the surgery demonstrate the resolution of a SDAVF, further highlighted on the volume rendering (VR) reconstruction (f). A circle marks a SDAVF on b–d and its reduction on e, f. MR, magnetic resonance