This phase III trial compares proton craniospinal irradiation (pCSI) to involved-field radiation therapy (IFRT) for the treatment of breast or non-small cell lung cancer that has spread from where it first started to the cerebrospinal fluid filled space that surrounds the brain and spinal cord (leptomeningeal metastasis). Patients with leptomeningeal metastasis (LM) may develop multiple areas of nervous system (neurologic) impairment that can be life-threatening. Radiation therapy (RT) effectively relieves local symptoms due to LM. RT uses high energy radiography (x-rays), particles, or radioactive seeds to kill cancer cells and shrink tumors. IFRT is commonly used to treat symptoms of LM. IFRT is radiation treatment that uses x-rays to treat specific areas of LM and to relieve and/or prevent symptoms. pCSI uses protons that can be directed with more accuracy than x-rays which allows treatment of the entire central nervous system space containing the cerebrospinal fluid (CSF), brain, and spinal cord. The pCSI treatment could delay the worsening of LM. Giving pCSI may be better than IFRT in treating LM in patients with breast or non-small cell lung cancer.
PRIOR TO STEP 1 REGISTRATION
Patients with pathologically (histologically or cytologically) proven diagnosis of breast cancer or NSCLC
Positive CSF cytology for malignancy
CSF cytology with suspicious cells is considered positive; CSF cytology with atypical cells is considered equivocal and not positive
None
Parallel
Biospecimen Collection
Procedure/SurgeryUndergo blood and CSF sample collection
Computed Tomography
Procedure/SurgeryUndergo CT or PET/CT
Involved-Field Radiation Therapy
RadiationUndergo IFRT
Lumbar Puncture
Procedure/SurgeryUndergo LP
Magnetic Resonance Imaging
Procedure/SurgeryUndergo MRI
Positron Emission Tomography
Procedure/SurgeryUndergo PET/CT
Proton Beam Craniospinal Irradiation
RadiationUndergo pCSI
Quality-of-Life Assessment
Other interventionAncillary studies
Overall survival (OS)
Will be event-driven and will be conducted when a total of 88 OS events (from both treatment arms) have been observed. The final analysis (if the trial does not stop early at an interim) is expected to occur roughly 45 months after study activation (including the initial 6-months ramp-up). All analyses (interim or final) will be done on a modified intent-to-treat (ITT) basis such that all randomized patients who have follow-up information will be included in the arm to which they are randomized regardless of what treatment the patients receive. Treatment comparisons will be based on the log-rank test. At the final analysis, if the test has an associated 1-sided p-value of 0.024 or less in favor of proton craniospinal irradiation (pCSI) (equivalently, Z \> 1.98), then the trial will conclude that the pCSI arm results in improved survival over the involved field radiotherapy (IFRT) arm.
Central nervous system (CNS) progression-free survival (PFS)
Will be based on modified ITT, including all randomized patients who have follow-up information. The analysis will be performed according to the treatment arm to which patients were randomized. Analysis for CNS-PFS will consist of estimation of its distribution for each treatment arm via the Kaplan-Meier method and a log-rank test. Statistical power will depend on the total CNS-PFS events accumulated at trial end. Additional analyses may consist of estimating the treatment hazard ratio via the Cox proportional hazards model accounting for prognostic covariates including randomization stratification factors evaluating whether the proportional hazards assumption holds or whether any treatment effect is notably time-varying, and evaluating for potential treatment by prognostic covariate interactions.
Time to CNS progression
Analysis for time to CNS progression will involve comparing the rate of CNS progression in patients who receive pCSI to patients who receive IFRT. The cumulative incidence function (CIF) estimator will be used to estimate the median time to CNS progression in the presence of competing event of deaths in the two arms, separately. The Gray's test will be used to evaluate the difference in the distribution of CNS progression between treatment arms. These results will be interpreted in light of the competing deaths. The comparison of treatment effect on time to CNS progression will involve estimating the cause-specific hazard ratio (CHR) in a Cox proportional hazards model adjusting for patients and disease characteristics (e.g. stratification randomization factors).
Time to radiation-induced CNS necrosis (TTRN)
Analysis will involve comparing the rate of radiation necrosis in patients who receive pCSI to patients who receive IFRT. Death will be treated as a competing risk event in this analysis. The cumulative incidence function (CIF) estimator will be used to estimate the median time to radiation-induced CNS necrosis in the presence of competing event of deaths in the two arms, separately. The Gray's test will be used to evaluate the difference in the distribution of TTRN between treatment arms. These results will be interpreted in light of the competing deaths. The comparison of treatment effect on TTRN will involve estimating the CHR in a Cox proportional hazards model adjusting for patients and disease characteristics (e.g. stratification randomization factors).
Incidence of treatment-related adverse events (AEs)
Will be evaluated by comparative frequency summaries for all adverse event types between arms. AEs will be graded according to Common Terminology Criteria for Adverse Events version 5.0. Comprehensive summaries of all AEs by treatment arm will be generated and examined. Counts and frequencies of worst (highest score) AE per patient will be presented overall and by AE type category, separately by assigned treatment group. The proportion of patients with at least one grade 3 or higher AE will be compared between treatment arms. Similarly, frequencies for specific potentially treatment related AEs where grade 3 or higher events are noted may be compared. Any frequencies to be tested will be evaluated using the chi-square or exact test as appropriate, with two-sided significance level 0.05.
Estimates of the primary outcome treatment effect by sex
Will estimate the primary outcome treatment effect and the corresponding 95% confidence intervals (CIs) by sex.
Estimates of the primary outcome treatment effect by race
Will estimate the primary outcome treatment effect and the corresponding 95% CIs by race.
Estimates of the primary outcome treatment effect by ethnicity
Will estimate the primary outcome treatment effect and the corresponding 95% CIs by ethnicity.
Click on the lead sponsor to view all of their trials.
