Article Figures & Data
Figures
- FIG 1.
Normalization of MR imaging findings across time despite persistent spinal CSF leaks. A, Sagittal T1 noncontrast MR imaging in 2018 demonstrates sagging of posterior fossa structures (bracket), engorgement of the pituitary gland, and narrowed suprasellar distance (arrow). B, Axial FLAIR MR imaging in 2018 demonstrates a diffuse, thin subdural collection (arrows). Sagittal (C) and axial (D) T2 MR imaging of the cervical spine in 2018 shows a cervicothoracic ventral epidural fluid collection (arrows). The patient underwent a dorsal nontargeted epidural blood patch in 2018 with partial relief of symptoms. E, Sagittal T1 noncontrast MR imaging in 2023 demonstrates resolution of brain sag, pituitary engorgement, and narrowing of the suprasellar interval. F, Axial FLAIR MR imaging in 2023 with resolution of the subdural collection. Sagittal (G) and axial (H) T2 MR imaging of the cervical spine in 2023 shows persistence of the cervicothoracic ventral epidural fluid collection. I. Intraoperative photograph later in 2023, with a ventral dural defect identified at T2–T3 (arrows). After repair, the patient had substantial symptom improvement, with the Headache Impact Test score improving from 68 to 48 (Headache Impact Test: range, 36–78).
- FIG 2.
Brain MR imaging findings incorrectly reported as normal in a patient with spontaneous intracranial hypotension. A, Sagittal T1 postcontrast MR imaging demonstrates mild narrowing of the suprasellar (1 mm, dotted arrow), mamillopontine (4.8 mm, dashed arrow), and prepontine (3 mm, solid arrow) distances. B, No pachymeningeal thickening or subdural collection on axial T1 postcontrast MR imaging. C, Left-lateral decubitus dynamic CTM detected a CVF arising from the left T7–8 neural foramen (arrow).
- FIG 3.
The Bern score. A, Sagittal T1 noncontrast MR imaging demonstrating the suprasellar interval (solid line, normal, >4 mm), mamillopontine interval (dotted line, normal, >6.5 mm), and prepontine interval (dashed line, normal, >5 mm). B, Sagittal T1 postcontrast MR imaging illustrates a normal flat appearance of the upward margin of the transverse sinus (arrow). C, Coronal FLAIR MR imaging with bilateral subdural fluid collections (arrows). D, Sagittal T1 postcontrast image demonstrates abnormal venous engorgement evidenced by an abnormal convex upward margin (solid arrow), as well as diffuse pachymeningeal enhancement (dotted arrows). The presence of a narrowed suprasellar interval, venous engorgement, or pachymeningeal enhancement are ascribed 2 points each, while a narrowed mamillopontine interval, prepontine interval, or the presence of subdural collections are ascribed 1 point each. A combined score of ≤2 equates to low probability, a score of 3 or 4 equates to moderate probability, and a score of ≥5 equates to high probability of localizing a CSF leak or venous fistula on subsequent myelography.20
- FIG 4.
A 35-year-old man with spontaneous intracranial hypotension who underwent 2 prior nontargeted dorsal epidural blood patches. Sagittal (A) and axial (B) T2-weighted MR imaging sequences demonstrate a ventral epidural fluid collection (arrows). C, Prone dynamic CTM detected contrast extravasating from the subarachnoid into the ventral epidural space at T1-2, consistent with a ventral dural defect (arrow). D, Procedural image from a CT-guided epidural blood and fibrin patch using a 15-cm 22-ga spinal needle via a far-lateral transforaminal approach to target the ventral epidural space adjacent to the defect (arrow). Postinjection sagittal (E) and axial (F) images demonstrate spread of injected blood and fibrin glue along the ventral epidural space (arrows). Posttreatment sagittal T2 (G) and axial 3D T2 fat-saturated MR imaging (H) with resolution of the epidural fluid collection.
- FIG 5.
A 49-year-old man with a history of spontaneous intracranial hypotension, with persistent symptoms after 3 epidural blood patches. A, Axial SWI with hypointense signal along the cerebellar folia (arrows), consistent with superficial siderosis, a rare complication of chronic CSF leak. B, Axial T2-weighted MR imaging demonstrates a ventral epidural fluid collection (arrows). C, Axial noncontrast CT image shows a small osteophyte (arrows) along the ventral canal at T6–7. D, Prone dynamic CTM demonstrates extravasation of contrast from the subarachnoid space into the ventral epidural space at T6–7 (arrows), consistent with a ventral dural defect. E, Intraoperative photograph during repair of the ventral dural defect (arrows) at T6–7, accessed via hemilaminectomy, posterior durotomy, and lateral mobilization of the cord after dentate ligament resection .
Tables
Differential diagnoses in patients presenting with SCSFL
Overlap with SCSFL Key Differences from SCSFL Chronic migraine Headache as primary clinical presentation Often evolves from less-frequent to more-frequent episodes
Less typically orthostatic
Often occur in the morning, rather than worsening throughout the day
Postconcussion headache Both can be precipitated by trauma Headache phenotype like migraine
Typically nonorthostatic
Associated cognitive issues
Postural orthostatic tachycardia syndrome Orthostatic symptoms Defined by changes in hemodynamic parameters on standing
If headache associated, more often migraine-like
Neck stiffness less common
Chiari I malformation Cerebellar tonsillar ectopia may be present in both conditions Brain MR imaging should not demonstrate downward herniation of posterior fossa structures, pituitary/venous engorgement, or pachymeningeal enhancement
Orthostatic headache less common
