Dural Herniation and Cerebrospinal Fluid Leaks

Pathogenesis

A basic knowledge of the embryology of the temporal bone is helpful in the understanding of formation of dural herniation and CSF leaks. There are four ossification centers that form the temporal bone: squamous, tympanic, petrous, and mastoid. Pneumatization follows ossification and continues into adulthood. It consists of a process of marrow resorption, mucosal advancement, and bone remodeling. Disturbances in the normal ossification or pneumatization process may lead to encephalocele formation. Portions of the petrous and squamous ossification centers of the temporal bone ultimately form the roof of the middle ear and mastoid; these areas are the most likely to form encephalocele.

Many authors have reported multiple tegmen defects ( Fig. 21.3 ). Åhrén and Thulin noted a 6% incidence of multiple defects in the tegmen in 94 consecutive cadaveric specimens. Ferguson et al.6 noted a 22% incidence of multiple tegmen defects in 27 preserved dried temporal bones. Lang identified tegmen defects in 20% of 70 temporal bones. Bilateral spontaneous encephaloceles could be predicted from such studies and have been reported by Iurato et al.

The high incidence of tegmen defects is contrasted by the low incidence of spontaneous encephalocele and CSF leak formation. This observation lends credibility to the widely held belief that bony defects are necessary, but not solely responsible, for the formation of encephaloceles. The other factor necessary for encephalocele formation is a pathological process at the dural level. An inflammatory process, chronic suppurative otitis media, and transmastoid surgery are the most common factors in the dura. Other factors, such as benign intracranial hypertension (BIH), obesity, obstructive sleep apnea (OSA), and aging, have also been suggested. ^{, ,}

The association with chronic suppurative otitis media and transmastoid surgery is well documented in the literature. ^{, ,} In the preantibiotic era, meningeal complications from chronic suppurative otitis media were common. The incidence of encephalocele during this period was increased by transmastoid trephination of peridural and brain abscesses. Encephalocele can occur in the face of chronic suppurative otitis media with or without surgery, and with or without cholesteatoma. ^, Paparella et al. reported ten cases of encephalocele all associated with chronic otitis media, six of which had not had previous surgery, and two of which had cholesteatoma. These authors speculated that the inflammatory response led to destruction of the tegmen, followed by extension to the adjacent dura and brain, resulting in herniation. Encephalocele can form when a dural tear occurs at the site of a bone dehiscence from a drill, scalpel, or cautery. Dural injuries may go unrecognized during the surgery; however, if recognized at the time of surgery, they should be repaired to prevent an immediate or delayed problem. Although chronic suppurative otitis media and transmastoid surgery are still a common cause for encephalocele and CSF leak, a decreasing incidence has been cited because of improved surgical techniques, including the operating microscope and the use of drills versus gouges and chisels.

CSF leak secondary to head trauma is well documented. The dura in the skull base that is adherent to the adjacent bone tears along the fracture lines. The CSF leak usually resolves spontaneously or with lumbar drainage within 1 to 2 weeks. Savva et al. managed 26 of 29 patients with CSF otorrhea secondary to head injury successfully with conservative treatment. In contrast, only 1 of 53 patients with surgically induced or spontaneous CSF leak was successfully managed conservatively. Bony dehiscence and dural injury from head trauma can lead to either immediate or delayed encephalocele formation.

Spontaneous encephalocele and CSF leak have been attributed to multiple factors. Arachnoid granulations of the temporal bone may be a cause of spontaneous encephalocele and CSF fistula formation. Normally, the arachnoid granulations protruding into the lumen of venous structures are involved in the resorption of CSF. When not associated with venous structures and surrounded by bone, they may enlarge and cause bone erosion as a result of the intermittent subarachnoid pressure associated with age and physical activity. In a study of temporal bones, Gacek observed that in the tegmen tympani and mastoideum, 22% had tissue on the dural surface and 9% had tissue in the posterior fossa.

Several authors have suggested that BIH might play a role in the development of spontaneous CSF leak and temporal bone encephaloceles. ^{, ,} The indirect indicators associating BIH with spontaneous CSF leaks include morbid obesity, empty sella syndrome, and a high prevalence in women. Recent direct measurements of the intracranial pressures with a lumbar puncture and the opening pressures in patients with spontaneous CSF leak and encephalocele have established an association with BIH. An initial report of 26 patients with spontaneous CSF leaks in which 9 patients consented to a lumbar puncture revealed that 6 patients had opening pressures consistent with BIH, greater than 20.0 cm water (14.7 mm Hg) with a mean of 24.5 cm water (18.0 mm Hg) in a range of 23.5 to 40 cm of water (17.3 to 29.4 mm Hg). Subsequent studies of the lumbar puncture and opening pressures also demonstrated an association with BIH with an incidence of 9% to 69%. ^,

Obesity and OSA are believed to contribute to skull base thinning over time, although the pathophysiology is not well understood. Nelson et al. identified a twofold increase in the incidence of spontaneous CSF leak repair in the United States after comparing the frequency of required craniotomy for the condition between 2002 and 2012. All the patients were overweight with a large percentage having sleep apnea, and the rate of repair was doubled in the region with the highest rate of obesity (Midwest) compared to the region with the lowest (West). These authors also demonstrated that obese patients with spontaneous CSF leak have thinner calvariums compared to the obese controls, suggesting that the body mass index does not appear to be the single predisposing risk factor. In their cohort of patients with leaks, 46% had OSA, which is associated with transient elevations in intracranial pressure.