Phase I Dose Escalation Trial of Hypofractionated Radiosurgery for Large Intact or Resected Brain Metastases

Abstract

Purpose/Objective(s): Leptomeningeal metastasis (LM) is associated with limited survival and treatment options for patients with solid tumor malignancies. While photon involved-field radiotherapy (IFRT) is effective for local palliation, it lacks durability. Craniospinal irradiation (CSI), in contrast, treats the entire central nervous system (CNS) compartment, thus potentially improving disease control. We previously reported the safety and tolerability of proton CSI (pCSI) in a completed phase IB study, and we are now reporting the mature survival outcomes of these patients. Materials/Methods: We enrolled 24 patients with solid tumor LM to receive hypofractionated pCSI of 30 Gy (RBE) in 10 fractions in this phase I prospective trial between June 2018 and May 2019 (NCT03520504). The primary endpoint is dose-liming toxicity (DLT) within 1 month of treatment. Secondary endpoints, including time to CNS progression (CNS TTP), CNS progression-free survival (PFS), and overall survival (OS), were assessed using Kaplan-Meier analysis with median survival estimates and 95% confidence intervals (CI) reported. Correlations between cerebrospinal fluid (CSF) circulating tumor cells (CTCs) and survival outcomes were evaluated using Kaplan-Meier analysis. An optimal cutoff for CTCs was determined through maximally selected rank statistics using CNS PFS.Results: Of the 24 patients enrolled, 3 did not complete treatment due to systemic disease progression and were excluded from this analysis. The majority of patients had metastatic lung (52%) and breast (33%) malignancies. The median age was 52 years (range 30-67 years) and the median Karnofsky performance status was 70 (range 60-90). The majority of patients had LM that progressed on other therapies prior to enrollment (57% vs. 43% of patients with newly diagnosed LM), and all patients with molecularly driven disease had CNS progression on small-molecule tyrosine kinase inhibitors prior to enrollment. Only 1 patient was censored at 24 months of follow-up, while 20 patients were followed until death with a median OS of 9 months (95% CI 6-22 months). The median CNS TTP was 8 months (95% CI 7-18 months), with 5 (24%) patients with durable CNS control for > 12 months after pCSI. The median CNS PFS was 7 months (95% CI 5-13 months). In the 18 patients with CSF CTCs evaluation at baseline, having a pre-pCSI CSF CTCs < 41/3mL was associated with improved CNS TTP (19 vs. 8 months, P < 0.01), CNS PFS (18 vs. 6 months, P < 0.01), and a trend toward improved OS (21 vs. 8 months, P = 0.06). Conclusion:We reported that pCSI is a safe treatment for patients with solid tumor LM. In this analysis, we demonstrated durable disease control and prolonged survival in some patients. Lower CSF CTCs at baseline, a potential biomarker of response, was associated with improved survival outcomes. An ongoing randomized phase II trial will determine the efficacy of pCSI compared to photon IFRT in patients with solid tumor LM.