Leptomeningeal metastases in high-grade adult glioma: Development, diagnosis, management and outcomes in a series of 34 patients

Abstract

Leptomeningeal metastases (LM) in the setting of glioma have often been thought to carry a particularly poor prognosis. We sought to better characterize this phenomenon through a review of patients with glioma seen in our institution over the preceding 10 years. We focus here on 34 cases with LM due to grade III or IV glioma. Over the period in question we estimate a prevalence of almost 4% in those affected by grade IV tumors. Leptomeningeal spread was present at the time of initial diagnosis in 4 patients. In the others, LM occurred at the time of first progression of disease in 17. The median time to development of LM (excluding those where it was present at initial diagnosis) was 16.4 months (95% confidence interval (CI) 8.2-43.9). The median time to further progression of disease following LM was 4.9 months (95% CI 3.1- 6.9). 25 patients were known to have died at the time of writing. Thus median overall survival (OS) was 10.2 months (95% CI 8.8-14.7) following LM. At the time of diagnosis of LM, some form of treatment (chemotherapy and/or radiation vs. no treatment) increased OS (median 11.7 vs. 3.3 months, p<0.001 by log-rank test). Use of radiation therapy (vs. no radiation) also increased OS, although the effect was more modest (7.8 vs. 16.8 months, p=0.07). Higher Karnofsky Performance Status (KPS) at the time of diagnosis of LM was associated with OS (p=0.007, median OS for KPS ≥90 19 months vs. 7.8 for KPS <90). In a two-variable model incorporating the use any treatment (vs. none) and KPS, KPS tended to be a more significant predictor of survival (p=0.22 vs. p=0.06 by likelihood-ratio test). This was also true for radiation (vs. none) and KPS (p=0.27 vs. p=0.02). No significant benefit could be demonstrated for the use of chemotherapy considered alone, either systemic or intrathecal. It should be noted that 4 of 9 patients receiving intrathecal chemotherapy had a ventriculoperitoneal shunt in place during these injections, which may have reduced its effectiveness. © 2014 Dardis, Milton, Ashby and Shapiro.