CVFs in SIH: EID vs. PCD CTM

Spontaneous intracranial hypotension (SIH) is a debilitating condition caused by a spinal CSF leak. One subtype of spinal CSF leak, the CSF-venous fistula (CVF), is a pathologic connection between the CSF-containing nerve root sleeve and an adjacent paraspinal vein. Identifying and localizing the CVF that causes SIH is a critical prerequisite to curative treatment. Visualizing CVFs requires high-quality myelography. Despite improvements in myelographic techniques, CVF identification remains challenging because they are diminutive structures and depend on identifying small volumes of contrast in a paraspinal vein, the "hyperdense paraspinal vein sign."

Compared with EID CT myelography (EID-CTM), PCD-CTM enables much higher spatial resolution and a greater contrast-to-noise ratio, carrying with it the promise of improved detection rates for CVF. Previous publications report early encouraging findings on the benefits of PCD-CTM for CVF detection; however, all prior work has been based on retrospective data. Importantly, these retrospective studies have been biased toward reduced PCD-CTM CVF detection, since only the patients with CVFs that are not definitely identified on EID-CTM undergo PCD-CTM (thus, more challenging cases are completed on the PCD CT). No studies have prospectively demonstrated the additional benefit of PCD-CTM over EID-CTM. Furthermore, no studies have prospectively reported the yield of any type of myelography for CVF detection. Based on the study team's knowledge of these imaging modalities and the imaging appearance of CVFs, PCD-CTM may have a benefit in detecting CVFs that can be missed on EID-CTM, which has been suggested in retrospective studies.

This is a prospective randomized clinical trial designed to assess the diagnostic yield of two different CT scanner types for detecting CSF-venous fistulas (CVFs) during CT myelography. The primary study aim is to compare PCD-CTM to traditional EID-CTM, which is currently the most commonly used technique for CVF detection. Patients meeting inclusion/exclusion criteria will be identified at the Duke CSF Leak Clinic. After informed consent, enrolled participants will be randomized (1:1 allocation) to either PCD-CTM or EID-CTM. CTM technique for both arms will be performed in a standardized manner as part of routine standard of care, as previously published. There will be no additional imaging outside of the current standard of care.

The only intervention in this study is randomization into either of the two imaging arms.

CTM images will be reviewed by the radiologist performing the procedure as part of routine standard of care as well as one additional neuroradiologist. The presence or absence of a CVF will be classified according to the Duke CVF Confidence Score (DCCS). Disagreements will be adjudicated by a third neuroradiologist.