Cerebrospinal fluid-venous fistula visualisation by intrathecal pressurization: A technical note and illustrative case

Abstract

CSF-venous fistulas (CVFs) are an important cause of spontaneous intracranial hypotension and challenging to diagnose. Lateral decubitus positioning during myelography and a technique called “resisted inspiration” has shown to improve CVF detection. However, the impact of intrathecal pressurization to improve visualization of CVF has mostly been speculated on. In this brief report, we demonstrate how a CVF became progressively more visible only after stepwise intrathecal pressurization: An indication of the importance of pressurization for CVF detection.

Introduction

Spontaneous intracranial hypotension (SIH) typically presents with orthostatic headache but may also lead to various symptoms and serious complications. Next to a ventral (type 1) and lateral (type 2) dural tear, a CSF-venous fistula (CVF, type 3) is a major cause of SIH and occurs in at least 25% of cases.^1,2

The diagnosis of CVF is challenging, which is one reason why this entity has only been known since 2014. ³ In recent years, substantial technical advances have been reported to improve detection of CVF: Compared to prone position, lateral decubitus position increases the diagnostic sensitivity of myelography by up to five times,^4,5 as iodine contrast medium is heavier than CSF and supports CVF outflow. Another technique to better visualize CVF is myelography during inspiration or, as a further refinement, “resisted inspiration” ⁶ : Continuous inspiration against resistance (small straw or slip tip syringe) predominantly decreases intrathoracic venous pressure and moderately increases CSF pressure, resulting in a pressure gradient from the CSF space to the paravertebral veins and promoting CVF outflow during myelography.^7–9 On the other hand, intrathecal pressurization (by volume administration of saline, e.g.) only increases the spinal CSF pressure, but should also result in a pressure gradient with respect to the venous system and thus better visibility of the CVF during myelography. Although the use of intrathecal pressurization is not new and is already in use, the effects on better visualization of CVF have mostly been speculated but not yet illustrated by imaging.

Case description

A 26-year-old woman presented to the emergency department with severe positional headache that had occurred spontaneously 5 days earlier and was distinctly different from a familiar migraine. The patient had no history of trauma or lumbar puncture. After worsening of headaches 2 days later without response to analgesics, she was further evaluated. MRI of the head showed a “Bern SIH score” of three points (narrowing of the suprasellar and mamillopontine distance), suggesting an intermediate probability for the presence of SIH (Figure 1 A, B). ¹⁰ MRI of the spine showed no epidural fluid collection and thus no evidence for a spinal dural tear. The patient underwent a first myelogram which did not identify a CVF, however, subsequent myelograms were performed with pressure augmentation, which revealed a CVF at the Th12/L1 level left. She then underwent CVF-ligation by surgery and developed rebound hypertension postoperatively, which was treated with acetazolamide. One month after treatment, headache improved considerably, but mild non-positional headache persisted.

Figure 1.

Sagittal contrast-enhanced T1 MRI of the head shows a narrowing of the suprasellar (white line in a; 2.5 mm) and mamillopontine distance (white dashed line in A; 5.9 mm), whereas no pachymeningeal enhancement is visible (b). Lateral decubitus CT myelography (LD-CTM) on the left side discloses a conspicuous, tubular-shaped diverticulum (white asterisks in A) without evidence for a CSF-venous fistula (CVF). A rescan in the same session 7min later was identical (not shown). Right-sided LD-CTM on the second day after pre-pressurization with 4 mL of contrast medium before injection of a regular 7–8 ml of contrast shows low contrasted paraspinal veins at the level Th11/12 and Th12/L1 on the left side (open arrows in coronal view in D and axial view in E), remote to the spinal diverticulum (white asterisks in D and E). A rescan in the same session in the lateral decubitus position on the right side after 9 min and on the left side after a further 5 min showed no more contrasted veins (not shown). All images are presented as maximum intensity projection (MIP).

Myelography and pressurization technique

The initial procedural plan was to perform a “same-day bilateral decubitus CT myelography” (CTM) with a divided contrast bolus, as has been described in the literature. ¹¹ Lumbar puncture was performed at L5/S1 level with a 21-ga atraumatic needle. Opening pressure was 7 cm of water. The patient was tilted head down by 7° (on a custom-made table) at the CT scanner, and 7–8 ml of iodine contrast (300 mg iodine/ml) was injected. During the exam, the patient inhaled continuously through a straw (“resisted inspiration”): the first run with a half bolus on the left side down demonstrated no CVF but a suspicious meningeal diverticulum at T12/L1 (Figure 1(C)). After the left decubitus exam, the needle was inadvertently dislodged, therefore, the contralateral decubitus exam was performed on a subsequent day, during which a larger bolus was administered (since more contrast was available for a unilateral CTM): 4 mL of contrast medium was pre-injected to fill the spinal diverticula, followed by regular 7–8 ml of contrast after 1 min. During this exam, on the right side there was suspicious venous filling contralaterally at the Th11/12 and T12/L1 level next to the conspicuous diverticulum on the left (Figure 1(D), (E)). Therefore, the patient was taken to DSM on a third day for pressure-augmented exam. In addition, the DSM with the features of high temporal resolution should clarify which level and side the CVF definitely originated: After pre-pressurization with 10 mL saline, DSM followed immediately after injection of regular 7–8 ml contrast medium (60s DSM run at one frame/sec) at the left side while the patient kept breathing shallowly. This time, a CVF was clearly visible at the Th12/L1 level on the DSM on the left (Figure 2(a), supplementary video), and also on subsequent ultrahigh-resolution cone-beam CT myelography after additional administration of 2 mL of contrast medium, a technique that has been described previously (Figure 2(b) to (d)). ¹²

Figure 2.

Lateral decubitus digital subtraction myelography (LD-DSM) on the left side a third day after intrathecal pre-pressurization of 10 mL saline clearly visualizes a CSF-venous fistula (CVF, open arrows in A, supplemental video) originating from the tubular-shaped diverticulum at the level Th12/L1 left (black asterisks in a). Ultrahigh-resolution cone-beam CT myelography in the same session after a further injection of 2 mL contrast medium shows the same CVF (solid white arrow in B) in connection with the diverticulum (white asterisks in b-d) and additionally spreading of contrast to the level Th11/12 left via the internal epidural venous plexus (open arrows in b, c, and d). Remarkably, the diameter of the diverticulum is larger after the pressurization (2.6 mm at the level of the asterisks in B) than before pressure augmentation (1.8 mm at the level of the asterisks in Figure 1). Images of B-D are presented as maximum intensity projection (MIP).

Discussion

We present a patient with a CVF that remained undetected on myelography without intrathecal pressurization but became partially visible after pressurization with 4 mL and completely visible after pressurization with 10 mL volume.

Intrathecal pressurization during myelography to improve detection of CVF has occasionally been described in the literature as an institutional standard (but was not routinely applied at our institution).^1,11,13 Usually 10 mL of saline solution is used for this purpose, although this amount is arbitrary and not based on scientific evidence. Alternatively, Callen et al. described raising pressure to 25–30 cm CSF using saline infusion and a digital manometer. ⁹ Caton et al. reported preliminary results based on craniospinal space compliance curves showing a potential correlation between intrathecal pressure elevation and CVF diagnosis. ⁸ However, to our knowledge, there are no reports demonstrating its efficacy for better visualization of CVF in detail on imaging. Recently, Mark et al. ¹⁴ observed a phenomenon at DSM: The same CVF that was visible on a lateral decubitus DSM performed with pre-pressurization of 5–10 mL saline was not necessarily visible when pre-pressurization was absent in another DSM in that area. The authors hypothesized that a certain threshold of intrathecal pressure may be necessary to “open” a CVF.^8,14 We speculate that the pre-injection of 4 mL of contrast medium at the second myelography in our case (Figure 1(c)) was sufficient to “open” the CVF for only a brief moment, so that the contrast medium draining into the paraspinal veins partially washed out and was no longer connected to the diverticulum from which it originated. However, pressurization with 10 mL saline at the third myelography was sufficient to open the CVF continuously (supplemental video). As the density of the contrast medium in the corresponding diverticulum was already very high in the first CTM (Figure 1(a)), the timing of the examinations and also the injection speeds was comparable, and these factors probably had less influence on the different results of the myelograms. Also, the technique of “resisted inspiration,” which was only used in the first two CTMs, apparently had no crucial impact on the visualization of the CVF (which was not or only partially depicted here compared to the DSM/UHR-CBCT without applying this technique). We believe that it is critical to generate pressurization immediately before or during myelography, otherwise intermediate opening of CVF could result in falling below the required intrathecal pressure. On the other hand, it is likely that some CVFs have no or a low threshold, as we have seen other CVFs open without pressurization. In addition to an individual report, it should be mentioned as a limitation that an intrathecal pressure increase after volume administration was assumed but not actively demonstrated.

With this report, we want to draw attention to the potential importance of pressurization for the detection of CVF and stimulate further research on this topic.

Conclusion

We present a case with SIH in which an underlying CVF became visible on myelography only after stepwise intrathecal pressurization, supporting previous speculations on the importance of pressurization to visualize CVF.

Supplemental Material

Video 1.

Appendix.

Abbreviations

CVF

CSF-venous fistula

SIH

spontaneous intracranial hypotension

DSM

digital subtraction myelography

CTM

computed tomography myelography

Ethics statement

Ethics approval

All procedures performed in the studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Consent to participate

Informed consent was obtained from all individuals included in the study.

Consent for publication

All authors provide consent for publication.

Data availability statement

Deidentified information of the patient can be made available if required.*

ORCID iDs

Niklas Lutzen https://orcid.org/0000-0003-0555-7863

Horst Urbach https://orcid.org/0000-0001-7264-4807