Closure of Cerebrospinal Fluid Leaks and Repair of Meningoceles/Encephaloceles
Summary
From 1952, when the first endonasal approach for closure of nasal cerebrospinal fluid (CSF) leaks was performed by Hirsch in Austria, and since the late 1980s, endoscopic transnasal approaches have been validated as the most effective and associated with less morbidity. In patients with suspected CSF leaks, confirmation can be gained with the use of beta-trace nasal/blood ratio as well as high-resolution computed tomography (CT) and T2-weighted (or following intrathecal gadolinium injection) magnetic resonance imaging (MRI). When the etiology of CSF rhinorrhea is unclear or in cases of recurrent CSF leaks, evaluation for elevated intracranial pressure (ICP) must be performed. Endoscopic closure of CSF leaks should follow a graduated approach, with the use of autologous grafts and vascularized flaps, as required. There is little evidence to recommend the routine use of lumbar drainage or bed rest postoperatively in uncomplicated cases.
Introduction
A nasal CSF leak describes the escape of CSF from the intracranial cavity through a bony defect within the skull base. It implies a tear or hole of the underlying dura, resulting in a communication between the intracranial contents and the nasal cavity. Leakage of spinal fluid can result in (recurrent) meningitis and for that reason needs to be closed. CSF leaks have been associated with an ~10% risk of developing meningitis per year. Most of the cases of recurrent meningitis are caused by anatomical defect, either congenital or posttraumatic.1
CSF leaks are often classified as traumatic or nontraumatic. Most leaks are traumatic. The leak can be found directly after the trauma but also many years later. A small number of traumatic leaks are iatogenic: intentionally in skull base surgery or usually unintentionally in sinus surgery. Sometimes nontraumatic CSF leaks are referred to as spontaneous; some authors have used the term spontaneous to include CSF leaks associated with multiple etiologies, such as tumors, delayed CSF leak from trauma, and CSF leaks associated with congenital malformations of the skull base.2 We prefer to classify CSF leaks according to their etiology and reserve the characterization of spontaneous or idiopathic to CSF leaks of unknown etiology.3 Other classifications are based on anatomical site. Most commonly, the leak is found at a cribriform plate location (30–40%), followed by the anterior ethmoid (20–30%), sphenoid sinus (15–25%), frontal sinus (< 10%), and posterior ethmoid (< 10%); in a small percentage, there are multiple leaks.4,5 Real idiopathic CSF leaks are rare and are often associated with undiagnosed intracranial hypertension ( Table 36.1 ).
Intracranial Hypertension
Note
Patients with spontaneous CSF leaks without congenital defect and absolutely no history of trauma, as well as patients with recurrent CSF leaks following failed procedures, are increasingly being recognized as suffering from increased intracranial pressure ( Figs. 36.1, 36.2, 36.3, and 36.4 ).
Benign intracranial hypertension (BIH) is characterized by elevated ICP without apparent pathology on conventional intracranial imaging and normal CSF studies. Several pathophysiologic mechanisms have been proposed, including alterations in CSF production and absorption, cerebral edema, endocrine disorders, and cerebral venous hypertension. Patients with idiopathic spontaneous CSF leaks demonstrate increased ICP. Epidemiologically, large numbers of these patients are young, obese women and typically have other symptoms of BIH, such as headaches, pulsatile tinnitus, and balance or visual disturbances. Not all patients, however, fit this description, with some patients presenting with a CSF leak and no other symptoms. In such patients, examination for papillary edema and measurement of CSF pressure is mandatory; CSF pressures > 200 mm H2O are likely to be abnormal.
Tips and Tricks
To make things more complicated, ICP may not be increased preoperatively (as the CSF rhinorrhea provides an outlet for excess CSF), so if the diagnosis is being considered, measurement of CSF pressure must be repeated a few weeks after the closure.
CT scans of patients with BIH can show some characteristic abnormalities. The skull base is often broadly thinned and attenuated. Typically, arachnoid pits, due to bony impressions from arachnoid villi, can be identified along the bony skull base.6 Additionally, almost all of these patients have been described as having pneumatization of the lateral sphenoid sinus recess ( Fig. 36.4 ).
MRI can show empty sella syndrome in patients with both spontaneous idiopathic CSF leaks and BIH. The empty sella is an extension or herniation of the subarachnoid space into the pituitary fossa through an incompetent sellar diaphragm that is commonly associated with clinical manifestations and endocrine abnormalities. The sella has the radiographic appearance of an absent pituitary gland due to filling with CSF ( Fig. 36.5 ). The empty sella may be primary or secondary to a variety of pituitary disorders. It is a probable consequence of long-standing intracranial hypertension associated with a congenital deficiency of the diaphragma sellae.
Diagnosis of Leaks
The accurate diagnosis of a CSF leak is mandatory prior to any potential intervention. A careful history is taken to find potential causes of a CSF leak, such as trauma, congenital malformations, and sinus or other surgery of the head. Questions about the history of meningitis should be asked.
Presenting Signs
Patients who present early after trauma may have pneumocephalus or active CSF rhinorrhea. However, the classic clinical presentation is a patient with intermittent clear nasal discharge, frequently unilateral, sometimes exacerbated by Valsalva maneuver, as well as bending and following physical exercise, but no other exacerbating factors and no associated nasal symptoms. Differentiation between CSF and clear, watery rhinorrhea can be difficult. Along with unilaterality as a sign of CSF, the discharge continues at night. CSF causes a halo sign on the pillow in the morning. Some patients present with recurrent episodes of meningitis and no history of rhinorrhea. In these patients, we place a small Merocel foam pack for 48 hours twice with 1 to 2 weeks in between to try to collect occult CSF and perform careful radiological examination to find a potential leak.
Nasal Endoscopy
Nasal endoscopy can be performed to find from which nasal cavity the CSF is coming; also, the area of the skull base most likely involved can be found by careful observation and suction of CSF. Endoscopy showing normal nasal mucosa can be useful in excluding the diagnosis of rhinitis (although rarely nasal mucosa irritation can occur as a result of chronic, significant CSF leak). Rarely, endoscopy can show the location of the leak and even (in large defects) the dural defect ( Fig. 36.7a, b ; see DVD/Thieme Media Center ). The diagnosis is easier when intrathecal fluorescein has been applied ( Fig. 36.7c ; see DVD/Thieme Media Center ).
Radiology
Imaging studies are crucial in the work-up of a CSF leak. The goals of imaging are to confirm the diagnosis, evaluate for an underlying cause, and localize and characterize the defect site prior to surgical repair.7
Computed Tomography Scan
With the development of high-resolution CT, essentially isotropic datasets can be acquired and, in turn, can be used to generate high-quality multiplanar images that can be viewed in any arbitrary plane. High-resolution multiplanar images, generated from a thin-collimation axial dataset using a bone algorithm, can reliably localize CSF leaks. An active leak does not need to be present to detect the skull base defect on CT scan.
Axial images are considered best for evaluation of the posterior wall of the frontal sinus and the posterior and lateral walls of the sphenoid sinus. Thin coronal sections are the most important for evaluation of the cribriform plates and roof of the ethmoid and sphenoid sinuses. The CT findings suggestive of a CSF leak include a skull base bony defect and an air−fluid level or opacification of the contiguous sinus. However, partial volume averaging can cause both false-negative and false-positive findings. These are minimized by using the thinnest sections possible, but at the expense of a significantly higher radiation dose to the eyes. Sagittal planes are also useful, especially in the most anterior part of the skull base, but here partial volume effects are even more of a problem. All in all, high-resolution CT scanning is a very reliable method, with a sensitivity comparable to the actual place of the leak > 90%.
Magnetic Resonance Imaging
MRI techniques offer noninvasive methods of imaging a CSF leak and are indicated to assess a possible encephalocele or meningoencephalocele. Herniation of the brain parenchyma or meninges through the bony defect may be difficult to differentiate from obstructed secretions on CT scans but is obvious on MRI. MRI after intrathecal gadolinium injection is an easy-to-perform and accurate technique for detection of a dural defect with excellent anatomical detail.8 MRI is performed in our practice if there is a very small osseous defect or if the defect results in a complete opacification of an adjacent sinus, as the latter could indicate the presence of a meningocele or encephalocele, particularly if the soft tissue opacification is lobular or nondependent or if there are multiple potential defects seen on CT (e.g., post-trauma) and it is unclear which defect results in a CSF leak.
Tips and Tricks
MRI is especially helpful in finding the hyperintense signal on T2 images (CSF), then correlating any irregularities with a potential bony defect on a CT scan ( Fig. 36.8 ).
Note
MR cisternography is a recently developed technique to assist in localization of active CSF leaks.
MR cisternography is useful in determining the localization of active CSF leaks. It involves a fast spin echo with fat suppression and image reversal. The MR cisternography algorithm shows the CSF as black against adjacent tissues, which are diminished in intensity.
Beta-2 Transferrin and Beta-Trace Protein
The differentiation of CSF from other forms of watery rhinorrhea can be challenging. Although unilaterality is an important sign, biochemical measurements are needed as proof. Traditionally, glucose was measured. As a general rule, CSF glucose is about two-thirds of serum glucose. However, this method has been abandoned due to low sensitivity and specificity. For more than a decade, assays for beta-2 transferrin or beta-trace protein have been reported in the literature as highly sensitive and specific methods for the detection of CSF in nasal secretions. The protein beta-2 transferrin is found in CSF, as well as in low concentrations in the perilymph in the cochlea and the aqueous and vitreous humor of the eye. This protein is detected by immunoblotting or immunofixation and silver staining, both of which are time consuming and labor intensive. Beta-trace protein is a brain-specific protein with a prostaglandin synthase activity. The major sites of biosynthesis are the leptomeninges and to some extent the choroid plexus. Beta-trace protein shows a CSF/serum ratio of 33, which is the highest of CSF-specific proteins known today. The quantitative beta-trace test is a short, nearly fully automated test based on a latex particle− enhanced immunologic assay using rabbit polyclonal antibodies against human beta-trace with almost no need for manual work. The beta-trace test can be used to reliably diagnose CSF rhinorrhea with higher specificity than the beta-2-transferrin test. Beta-trace concentrations > 0.496 mg/L are highly suggestive of the presence of CSF in examined nasal secretions. Beta-trace should not be used for patients with renal insufficiency and bacterial meningitis, as these conditions substantially increase serum and decrease CSF beta-trace values, respectively. In those cases, beta-2-transferrin test is more reliable. It should be noted that foam packing can be used to collect CSF for beta-trace measurement but one should be careful to use foam packing when measuring beta-2 transferrin because of distortion in the beta-2 transferrin test results through protein adsorption.9 Also there may be a false-positive result in the setting of chronic liver disease or inborn errors of metabolism of glycoprotein. For all these reasons and in all cases, one should measure beta-trace and beta-2 transferrin blood levels concurrently: the ratio of beta-trace or beta-2 transferrin in nasal fluid/blood is much more specific than their levels in the nasal fluid alone.