CSF Leak Presenting with Recurrent Meningitis
In this second group of patients, CSF leak is suspected because of recurrent meningitis in the absence of rhinorrhea. In this setting, the first line approach entails both endoscopy and imaging. BTP is, of course, unhelpful.
Endoscopy, CT and/or MR may succeed in demonstrating lesions both in post-traumatic and spontaneous fistulae. Compared to patients with CSF leak and rhinorrhea, the main difference consists in the absence of imaging findings of active leak (i.e., fluid collection within sinonasal cavities or focal mucosal thickening adjacent to bone dehiscences or to skullbase fractures).
Being based on indirect findings, high-resolution CT may be positive in inactive and active leaks with similar rates because it does not depend on the amount of CSF leakage at the time of investigation (Fig. 7.10).
Recently, a three-dimensional (3D), heavily T2-weighted sequence (3D CISS, constructive interference in steady state), has been proposed for CSF fistula detection (Jayakumar et al. 2001). 3D-CISS, being heavily T2-weighted with better CSF-brain-bone-air contrast, and allowing thinner (sub-millimeter) sections and multiplanar reconstructions, seems to be ideally suited for the demonstration of CSF leak (Fig. 7.11). Its very short TE enables a minimization of signal loss from magnetic susceptibility effects at the air-bone interface. The main advantage of this sequence is its ability to reduce artifacts from CSF pulsations. Unfortunately, in CISS sequence fat tissue appears moderately bright, so that anatomical details at the skull base/CSF interface may be obscured. Jayakumar et al. (2001) reported a sensitivity of 100% in a series of six patients: larger series have to be provided to establish the role of this promising technique.
If both endoscopy and non-invasive imaging are negative, the fluorescein test is considered more useful than invasive cisternographic techniques, because their sensitivity is insufficient in patients with inactive leak.
In the case of a positive fluorescein test, endoscopy may or may not precisely identify the fistula, as the defect can be located beyond diagnostic endoscopy limits. Microendoscopic surgery with fluorescein injection is indicated.
A negative fluorescein test makes the diagnosis of CSF leak unlikely.
Fig. 7.10a,b. Traumatic (due to car accident) CSF fistula. Coronal high resolution CT (a), with reformatted sagittal section (b). On the left side, the posterior aspect of the fovea ethmoidalis is not visible (arrows): a soft tissue density is clearly demonstrated (arrowhead) strictly adjacent to the large bone defect
Fig. 7.10a,b. Traumatic (due to car accident) CSF fistula. Coronal high resolution CT (a), with reformatted sagittal section (b). On the left side, the posterior aspect of the fovea ethmoidalis is not visible (arrows): a soft tissue density is clearly demonstrated (arrowhead) strictly adjacent to the large bone defect a
Fig. 7.11a-e. Small traumatic (due to car accident) meningocele through the right cribriform plate. Coronal TSE T2 (a); coronal (b) and sagittal (c-e) images from a CISS sequence. a The hyperin-tense signal of CSF appears to continue with the hyperintense signal within the right ethmoid cells, through a presumed bone defect in the right cribriform plate, raising the suspicion of a meningocele. However, it is impossible to distinguish the CSF possibly present in the ethmoidal cells from the thickened mucosa and/or sinus secretions. Only the left olfactory bulb is seen (arrow). b On coronal CISS sequence - 0.7 mm of thickness, obtained at the same level as (a) - the small meningocele (white arrows) shows the same signal as intracranial CSF, brighter than mucosal thickening. The contralateral olfactory bulb is demonstrated (black arrow). c-e The CISS sagittal reformatted sections show the small meningocele "neck" (arrows)
Was this article helpful?