Surgical Resection and Posterior Fossa Location Increase the Incidence of Leptomeningeal Disease in Patients Treated with Stereotactic Radiosurgery for Brain Metastases

Abstract

Purpose/Objective(s): Stereotactic radiosurgery (SRS), with or without prior surgical resection, has emerged as the primary and most efficacious alternative to whole-brain radiation therapy (WBRT) in the treatment of brain metastases. Leptomeningeal disease (LMD) is a severe, but rare, occurrence following local therapy. Growing evidence suggests an increased risk of LMD in the setting of prior surgical resection and posterior fossa location. The purpose of this study is to assess the factors predicting greater risk of LMD in a large cohort of consecutively treated patients who had not received WBRT. Purpose/Objective(s): Three hundred six patients treated with single-fraction SRS without upfront WBRT at a single institution between 2001 and 2013 were identified. Ninety-five (31%) patients had prior surgical resection. One hundred and sixteen (38%) patients had metastatic disease located in the posterior fossa. LMD was determined by chart and imaging review. A logistic regression model was constructed to assess the association of prior surgical resection, posterior fossa location, size of largest metastasis, number of metastases, and type of primary with the incidence of LMD. Results: With a median follow-up time of 9.7 months, LMD occurred in 31 (10%) patients at a median time of 12.3 months after SRS. Median overall survival after developing LMD was 2.3 months. A greater incidence of LMD was present for patients who underwent surgical resection (16.8%) compared to those treated with SRS alone (7.1%) (P = 0.009). The incidence in patients with posterior fossa metastases was 12.9% vs. 8.4% in patients with only supratentorial disease (P = 0.2). Upon evaluating the influence of histology, patients with breast cancers had a LMD rate of 27.8% vs. 7.8% when comparing to patients with all other primary sites (P < 0.001). In the logistic regression model, prior surgical resection (P = 0.044) and breast primary (P = 0.002) were both associated with statistically significant odds ratios (ORs), 2.59 and 4.14, respectively, for developing LMD. A posterior fossa metastasis was associated with an OR of 2.18 (P = 0.076). LMD risk did not correlate with either the number of metastases or size of largest metastasis. When repeating the analysis in only patients with posterior fossa metastases, prior surgical resection was an even stronger predictor of LMD (OR = 5.88, P = 0.009). Breast was excluded from subset analysis due to insufficient number of cases. Conclusion: Prior surgical resection and breast cancer primary place patients at increased risk of developing LMD. Surgical resection is an even greater risk factor for patients with posterior fossa metastases. Further investigation is necessary to determine whether SRS alone, cavity-directed SRS, or surgical resection followed by WBRT is most appropriate in this patient group.