Abstract
Case report
A 25-year-old woman suddenly developed a severe headache. The next day, she came to us and reported she had felt back pain for the previous 7 days. Neurological examinations were normal, except for orthostatic headache and tinnitus. Cranial magnetic resonance imaging (MRI) demonstrated typical findings of spontaneous intracranial hypotension (SIH), including pachymeningeal gadolinium enhancement, brain stem sagging, and enlargement of the venous sinuses (Fig. 1), with bilateral subdural hygroma also noted. The cerebrospinal fluid (CSF) opening pressure was unmeasurable and CSF outflow was confirmed by Queckenstedt's test during lumbar puncture. Subsequent radioisotope cisternography demonstrated early bladder filling at 1.5 h after starting the examination, with the tracer diffused into soft tissue above the lupper umbar region in early examinations. Spinal MRI disclosed an extensive CSF collection, which was most prominent at the dorsal thoracic level. In addition, engorgement of the venous plexus with gadolinium enhancement in the epidural space was found throughout the entire spine. At the lumbar level, sagittal fat-saturated T2-weighted MRI demonstrated segmental marginal hyperintensity around low-intensity areas of epidural fat, which corresponded to the leaked CSF signals (Fig. 1). Further, MR myelography demonstrated a collapsed theca and engorgement of the lumbar epidural venous plexus. A perineural cyst was also found. Computed tomography (CT) myelography revealed a bony lesion protruding into the subarachnoid space at the T8–9 intervertebral disc level (Fig. 2). CSF leakage began at this point and was distributed throughout the entire spinal canal, especially in the thoraco-lumbar junction (Fig. 2). The perineural cyst was not opacified in either early or delayed CT myelograms.
(a) Sagittal gadolinium enhanced T1-weighted magnetic resonance imaging (MRI) of the brain demonstrating diffuse pachymeningeal enhancement, brain stem sagging, and enlargement of the venous sinuses and pituitary gland. (b) Sagittal fat-saturated T2-weighted MRI of the lumbar region showing segmental marginal hyperintensity around a low-intensity area of epidural fat. (c) Lumbar MR myelography demonstrating a collapsed theca and engorgement of the epidural venous plexus.
(a) Sagittal reconstruction and (b) axial computed tomography (CT) myelogram images demonstrating an osteophyte invading the subarachnoid space at the T7–8 intervertebral disc level, where cerebrospinal fluid (CSF) began to leak (arrows). CSF leakage is seen as the less opaque contrast material compared with the intrathecal contrast material. (c) Coronal reconstruction image of CT myelogram showing lower thoracic and upper lumbar level with massive CSF leakage.
Conservative treatment for 3 weeks, such as bed rest, hydration and analgesic administration, provided the patient only partial relief from the symptoms. She received an epidural blood patch at T12–L1, but the procedure was stopped when the total amount of injection had reached only 4 ml because of intolerable back pain. Thereafter, her symptoms improved considerably, but did not completely resolve. Three weeks later, a second epidural blood patch in a 20-ml injection was given at L2–3, which provided complete relief. At the 2-year follow-up examination, the patient was free from symptoms and engaged in full-time work. Follow-up cranial and lumbar MRI confirmed the compete disappearance of pretreatment findings of SIH (Fig. 3).
(a) Sagittal gadolinium enhancement T1-weighted magnetic resonance imaging (MRI) of the brain demonstrating disappearance of diffuse pachymeningeal enhancement. Brain stem sagging is not seen. The venous sinuses and pituitary gland appear to be normalized. (b) Sagittal fat-saturated T2-weighted MRI of lumbar area showing the disappearance of pretreatment findings. (c) MR myelography demonstrating enlarged theca as well as perineural cyst compared with pretreatment. Engorgement of the epidural venous plexus is not seen.
Discussion
Spontaneous CSF leakage can occur from a variety of causes, with spontaneous tearing of the nerve root sleeves of the spine the most common. In addition, connective tissue disease and pre-existing cystic lesions, such as arachnoid diverticulum and perineural cysts, have been reported as underlying pathogenic factors (1–3).
More rarely, cases of SIH associated with spinal degenerative lesions including osteophytes and intervertebral discs (four with cervical osteophytes, one with a thoracic osteophyte and two with thoracic disc herniation) have been reported (4–8). The cases with an osteophyte were diagnosed by CT myelography (5, 7, 8), whereas those with a large thoracic disc were diagnosed from MRI (4, 6). Three of the four cases with a cervical osteophyte (7, 8) and one case with a thoracic osteophyte (5) required direct surgical repair. Notably, two of those cases with T7–8 disc herniation did not require direct surgical repair and were successfully treated by use of a targeted epidural blood patch (4, 6), whereas failure of a targeted epidural blood patch was noted in a case with an osteophyte at C5–6 (8). In addition, a lumbar epidural blood patch failed in a case with an osteophyte at C5–6 and obtained a partial response in a case with one at C4–5 (7, 8). After detecting the exact site of the CSF leak, a targeted epidural patch using injection directly at or near the site of CSF leakage should be attempted prior to direct surgical repair. In cases with a solid dural tear from a bony pathology with resultant CSF leakage at the cervical or thoracic level, a targeted epidural patch may be more effective than a lumbar epidural patch (9, 10). Although the present patient was successfully treated by use of an epidural blood patch, fixed and degenerative pathogenesis may cause SIH recurrence over the long term, for which surgical intervention may be required, as reported previously (5, 7, 8).
Based on past reports and results of the present case, CT myelography is superior to MRI techniques to detect a bony pathology. Further, high-resolution multidetector CT can easily cover the entire spinal axis in a very short period of time and produce reconstruction images as with MRI. The importance of CT myelography was previously emphasized for diagnosis of a CSF fistula caused by a cervical osteophyte (7, 8). On the other hand, MRI techniques are able to depict various compensatory changes according to the Monro-Kellie doctrine that are secondary to CSF depletion (11). Pachymeningeal gadolinium enhancement, brain stem sagging and venous sinus enlargement are well-known findings on cranial MRI. Engorgement of the epidural venous plexus, dilation of the perimedullary vein and fluid collection can also be seen on spinal MRI. In the present case, follow-up MRI demonstrated attenuation that was typical of findings for SIH, as the symptoms resolved. Thus, comparison of MRI findings before and after treatment is useful for therapeutic follow-up.
In conclusion, it is important to detect the exact site of CSF leakage for the treatment of SIH, and CT myelography is essential to detect an associated bony pathology. Thereafter, a targeted epidural blood patch should be considered prior to direct surgical repair. Finally, MRI can well depict secondary changes following CSF leakage and is useful for therapeutic follow-up.