An unusual case of CSF leak following post-traumatic rupture of a sacral meningeal cyst

We describe a 43-year-old woman who presented with acute dizziness, rotational vertigo, vomiting, neck pain and non-throbbing posterior headaches after a fall on the buttocks but without direct head injury and no whiplash. Brain imaging was initially performed to rule out the hypothesis of a vertebral artery dissection and did not show any unequivocal brain abnormality (Figure 1). A contrecoup injury with post-traumatic headache might also explain the clinical presentation. However, sagittal brain images covering the upper cervical spine identified an epidural mass (Figure 1(c)), which was further investigated by dedicated spinal magnetic resonance imaging (MRI) (Figures 2 and 3) and interpreted as dilated epidural plexus. Such a significant spinal finding led to consider the alternative hypothesis of intracranial hypotension because dilated veins could contribute to replacing cerebrospinal fluid (CSF) hypovolemia (Monro-Kellie relationship) (1). MR myelography and computed tomography (CT) bone scan confirmed an important sacral CSF leak that was secondary to a sacral fracture in front of a meningeal cyst (Figure 4) (2). Furthermore, the day after MRI diagnosis, the patient finally described orthostatic headache. The clinical outcome was good after seven days of bed rest with oral hyperhydration and acetaminophen, without epidural blood patch.

Figure 1.

(a) axial diffusion-weighted MRI was normal; (b) axial FLAIR MRI showed a slight subdural collection (arrows); (c) sagittal T1-weighted MRI showed mild pituitary enlargement (dashed line, 10 mm maximum height and a mild bulge of the superior margin) and mainly an anterior cervical lesion (arrow); (d) axial three-dimensional (gradient-echo) T1-weighted image obtained a few minutes after intravenous administration of gadolinium showed equivocal dural enhancement (arrows). There was no sagging of the brain or ventricular collapse. MRI: magnetic resonance imaging; FLAIR: fluid-attenuated inversion recovery.

Figure 2.

Sagittal cervical MRI: (a) T1-weighted MRI; (b) fat suppression using STIR MRI; (c) T2*-weighted MRI; (d) T1-weighted MRI obtained a few minutes after gadolinium injection. The premedullary mass was iso-intense on T1-weighted images, hyperintense on T2-weighted images, without clear hypointensity on T2*-weighted images, excluding a post-traumatic epidural hematoma (arrows, (a)–(c)). The strong and homogenous enhancement after gadolinium injection (arrow, (d)) was compatible with a marked dilation of the cervical epidural venous plexus. According to the Monro-Kellie relationship, dilated veins could contribute to replacing CSF hypovolemia and thus engorgement of dural vasculature was highly suggestive of intracranial hypotension. MRI: magnetic resonance imaging; STIR: short tau inversion recovery.

Figure 3.

Axial cervical images centered on C1: (a) T2-weighted MRI; (b) T1-weighted MRI after gadolinium injection. Note the “draped curtain sign” (dashed lines) ascertaining the epidural origin of the lesion. MRI: magnetic resonance imaging.

Figure 4.

(a) axial and (b) sagittal projections of heavily-T2 weighted myelography of the lower spine and (c) sagittal STIR MRI, showing an important sacral meningeal cyst with anterior CSF leak within the pelvis. (d) Axial non-contrast CT scan of the sacrum centered on S3: fracture (arrow) of a very thin cortical bone in front of the cyst. STIR: short tau inversion recovery; MRI: magnetic resonance imaging; CT: computed tomography.

In this particular case, the final diagnosis remained challenging. The clinical presentation was compatible with cerebral concussion and acute post-traumatic headache but the delayed orthostatic headache suggests that the CSF leak might also contribute to the neurological symptoms of our patient (2,3). This case report illustrates that spinal MRI could provide significant added value to brain MRI for the diagnosis of intracranial hypotension by revealing dilated epidural veins and demonstrating CSF leak (4). It might help to deal with the difficulties in establishing a specific diagnosis regarding the spectrum of headaches that occur in a post-traumatic context (3).