We present a rare indirect nontraumatic cerebrospinal fluid (CSF) fistula secondary to a fourth ventricle ependymoma. The fistula resulted from rupture of the left temporal horn, distant from the tumour. The fistula was well demonstrated by MRI. High-resolution CT demonstrated a defect in the roof of the sphenoid sinus, but no leakage of CSF was seen on CT cisternography.
Endoscopic sinus surgery (ESS) is a standard treatment for rhinosinusitis, which failed optimum medical therapy. Iatrogenic cerebrospinal fluid (CSF) rhinorrhoea can occur during ESS warrants early repair of the leakage. The common sites for CSF leakage are cribriform plate, fovea ethmoidalis, and anterior ethmoid sinuses. We present five cases of iatrogenic CSF rhinorrhoea due to ESS and its management.
The purpose of this retrospective study is to determine the pattern of cerebrospinal fluid (CSF) rhinorrhoea presenting to our tertiary referral centre in Kuala Lumpur and to assess the clinical outcomes of endonasal endoscopic surgery for repair of anterior skull base fistulas. Sixteen patients were treated between 1998 and 2004. The aetiology of the condition was spontaneous in seven and acquired in nine patients. In the acquired category, three patients had accidental trauma and this was iatrogenic in six patients (five post pituitary surgery), with one post endoscopic sinus surgery (ESS). Imaging included computed tomography (CT) scan and magnetic resonance imaging (MRI). Endoscopic repair is less suited for defects in the frontal sinuses with prominent lateral extension and defects greater than 1.5 cm in diameter involving the skull base. Fascia lata, middle turbinate mucosa, nasal perichondrium and ear fat ('bath plug') were the preferred repair materials in the anterior skull base, whereas fascia lata, cartilage and abdominal fat obliteration was preferentially used in the sphenoid leak repair. Intrathecal sodium flourescein helped to confirm the site of CSF fistula in 81.3 per cent of the patients. Ninety per cent of the patients who underwent 'bath plug' repair were successful. The overall success rate for a primary endoscopic procedure was 87.5 per cent, although in two cases a second endoscopic procedure was required for closure. In the majority of cases endoscopic repair was successful, and this avoids many of the complications associated with craniotomy, particularly in a young population. Therefore it is our preferred option, but an alternative procedure should be utilized should this prove necessary.