Evaluating the incidence and utility of microscopic metastatic dissemination as diagnosed by lumbar cerebro-spinal fluid (CSF) samples in children with newly diagnosed intracranial ependymoma

Ependymomas are the third most common central nervous system (CNS) tumor in childhood. After resection, the standard evaluation of intracranial ependymomas includes a post-contrast spine MRI and a lumbar CSF sample to evaluate for metastasis. Although it is estimated that 10-30% of patients will present with metastatic disease, it is unclear what percentage of metastatic disease is solely identified microscopically via lumbar cytology versus that identified as bulky disease on post-contrast spinal MRIs. We retrospectively evaluated all patients at our institution with intracranial ependymoma diagnosed between January 1991 and June 2008 in an effort to evaluate prognostic factors, survival outcomes and incidence of metastatic disease. Sixty-one evaluable patients were identified: 46% were male and the mean age at diagnosis was 64 months (2.04-196.92). The most common tumor location was the posterior fossa (77%) and 64% of patients achieved a gross total resection with initial surgery. Five-year event-free and overall survivals were 39 ± 7% and 87 ± 4%, respectively. Approximately 10% of patients presented with bulky metastatic disease as seen on post-contrast spine MRI. No patient whose spine MRI was negative for tumor had positive lumbar CSF cytology. These data highlight the rarity of patients who present with microscopic metastatic disease noted on lumbar cytology alone and suggest that evaluation of lumbar cytology may not be useful in patients with negative post-contrast spine MRIs. Future prospective trials may be helpful in validating this conclusion before eliminating this procedure as part of the standard evaluation in newly diagnosed ependymoma patients.