Phase II randomized study comparing proton craniospinal irradiation with photon involved-field radiotherapy for patients with solid tumor leptomeningeal metastasis.

2000 Background: Leptomeningeal metastasis (LM) is associated with limited survival and treatments. Photon involved-field radiotherapy (IFRT) is the standard of care radiotherapy (RT) but benefits are limited. We hypothesized that proton craniospinal irradiation (pCSI) encompassing the central nervous system (CNS) compartment would result in superior CNS disease control compared to IFRT. Methods: We conducted a randomized phase 2 study comparing pCSI vs. IFRT in patients with non-small cell lung cancer (NSCLC) or breast cancer LM. Eligibility criteria included radiographic and/or cytologic LM and Karnofsky performance status (KPS) ≥ 60. Patients were stratified by histology (breast vs. NSCLC) and systemic disease (active vs. stable) and were randomized in a 2:1 ratio of pCSI:IFRT. Patients with all other solid tumor histologies were enrolled to an exploratory pCSI arm. RT was 3Gy x 10 fractions for all patients. The primary endpoint is CNS progression-free survival (CNS PFS), defined as time from randomization to CNS progression (POD); secondary endpoints include overall survival (OS) and treatment-related adverse events (TAEs). A target of 81 patients to compare pCSI and IFRT was designed with a one-sided alpha of 0.025 and a power of 0.8 based on stratified log-rank test. Analysis is based on intent-to-treat. Results: From 4/2020-10/2021, 42 and 21 patients were randomized to pCSI and IFRT, respectively. Baseline factors were not different: median age was 56 vs. 61 years (p = 0.5); both cohorts included 57% NSCLC and 52% with active systemic disease. At median follow up of 7.1 months, 25 patients had CNS POD (pCSI = 9 [21%], IFRT = 16 [76%]) and 28 died (pCSI = 15 [36%], IFRT = 13 [62%]). At planned interim analysis, significant benefit in CNS PFS was observed with pCSI (median = 7.5 months, 95% CI: 6.6-NA) vs. IFRT (median = 2.0, 95% CI: 1.0-5.1, p < 0.001). As a result, the Data and Safety Monitoring Committee recommended early discontinuation of the trial. In addition, OS benefit with pCSI (median = 8.2 months, 95% CI: 7.4-NA) vs. IFRT (median = 4.9 months, 95% CI: 3.1-NA, p = 0.04) was observed. In a multivariable analysis including age, KPS and stratification factors, CNS PFS and OS benefit for pCSI remained significant. Grade 3 non-heme TAEs occurred in 3 patients with pCSI and 5 with IFRT. For the exploratory pCSI cohort, 35 patients enrolled, the median age was 61, 20 (57%) had active systemic disease and ovarian (7 [20%]) was the most common histology. At median follow up of 9.6 months, 7 (20%) had CNS POD and 20 (57%) died. Median CNS PFS was 5.4 months (95% CI: 4.8-9.1), OS was 6.6 months (95% CI: 5.4-12.1) and 4 patients had Grade 3 TAEs. Conclusions: In this trial, the first randomized study of RT for LM, we demonstrated improved CNS PFS of pCSI compared to IFRT, meeting the primary endpoint. pCSI also had a significant OS benefit. Grade 3 toxicities were comparable. Clinical trial information: NCT04343573.