Nonrandom spatial clustering of spontaneous anterior fossa cerebrospinal fluid fistulas and predilection for the posterior cribriform plate.
Academic Article
Overview
abstract
OBJECTIVE The anterior skull base is a common site for the spontaneous development of meningoceles, encephaloceles, and meningoencephaloceles that can lead to cerebrospinal fluid (CSF) fistula formation, particularly in association with idiopathic intracranial hypertension. In some circumstances the lesions are difficult to localize. Whether all sites in the anterior skull base are equally prone to fistula formation or whether they are distributed randomly throughout the anterior skull base is unknown, although the anterior cribriform plate has been proposed as the most frequent location. The purpose of this study was to identify sites of predilection in order to provide assistance for clinicians in finding occult leaks and increase the understanding of the etiology of this pathology. METHODS The authors performed a retrospective review of a prospectively acquired surgical database of all endonasal endoscopic surgeries performed at Weill Cornell Medical College by the senior authors. Spontaneous CSF fistulas of the anterior skull base were identified. The anatomical sites of the defects were located on radiographic images and normalized to a theoretical 4 × 2 grid representing the anterior midline skull base. Data from the left and right skull base were combined to increase statistical power. This grid was then used to analyze the distribution of defects. Frequency analysis was performed by means of a chi-square test, with a subsequent Monte Carlo simulation to further strengthen the statistical support of the conclusions. RESULTS Nineteen cases of spontaneous CSF fistulas were identified. Frequency analysis using chi-square indicated a nonrandom distribution of sites (p = 0.035). Monte Carlo simulation supported this conclusion (p = 0.034). Seventy-four percent of cases occurred in the cribriform plate (p = 0.086). Moreover, 37% of all defects occurred in the posterior third of the cribriform plate. CONCLUSIONS Anterior skull base spontaneous CSF leaks are distributed in a nonrandom fashion. The most likely site of origin of the spontaneous CSF leaks of the anterior midline skull base is the cribriform plate, particularly the posterior third of the plate, likely because of the lack of significant thick bony buttressing. Clinicians searching for occult spontaneous leaks of the anterior skull base should examine the cribriform plate, especially the posterior third with particularly close scrutiny.
