Key Points
Question
Does proton craniospinal irradiation (pCSI) result in improved survival outcomes compared with involved-field radiotherapy (IFRT) in patients with solid tumor leptomeningeal metastases (LMs)?
Findings
In this randomized clinical trial of 98 patients with solid tumor LM that evaluated radiotherapy approaches, there were superior survival outcomes with pCSI compared with IFRT for patients with non–small cell lung cancer or breast cancer LM.
Meanings
The findings of this trial suggest that pCSI should be considered as a treatment of option of LM if available.
Abstract
Importance
Leptomeningeal metastasis (LM) is associated with limited survival and few treatment options. Photon involved-field radiotherapy (IFRT) is the most common radiotherapy treatment for patients with LM from solid tumors.
Objective
To assess whether proton craniospinal irradiation (pCSI) would result in superior central nervous system progression-free survival (CNS-PFS) compared with IFRT.
Design, Setting, and Participants
A randomized, phase 2 trial of pCSI vs IFRT was conducted between April 16, 2020, and October 11, 2021, and included patients with non–small cell lung cancer and breast cancer with LM. Patients with other solid tumors were also enrolled in an exploratory pCSI cohort.
Intervention
For the randomized groups, after stratifying by histology and systemic disease status, patients were assigned (2:1) to pCSI or IFRT.
Main Outcomes and Measures
The primary end point was CNS-PFS. Secondary end points included overall survival (OS).
Results
Of 98 total patients, 72 individuals (73.5%) were female, and the median (IQR) age was 59 (50-65) years. A total of 42 and 21 patients were randomly assigned to pCSI and IFRT, respectively. At planned interim analysis, a significant benefit in CNS-PFS was observed with pCSI compared with IFRT, leading to the early discontinuation of the trial. In this final analysis, a significant benefit was continually observed in CNS-PFS with pCSI (median, 8.2 months; 95% CI, 6.6-15.3) vs IFRT (median, 2.3 months; 95% CI, 1.2-4.0; P < .001). A statistically significant and clinically meaningful OS benefit with pCSI (median, 11.3 months; 95% CI, 7.5-18.3) vs IFRT (median, 4.9 months; 95% CI, 3.9-15.0; P = .04) was also observed. For the exploratory pCSI cohort (n = 35), the median CNS-PFS was 5.8 months (95% CI, 4.4-9.1) and OS was 7.0 months (95% CI, 5.4-10.6).
Conclusions and Relevance
This randomized clinical trial that assessed the optimal radiotherapy treatment for LM found improved CNS-PFS and OS with pCSI compared with IFRT. The results suggest that pCSI should be considered when available.
Trial Registration
ClinicalTrials.gov Identifier: NCT04343573
This randomized clinical trial examines proton craniospinal irradiation compared with involved-field radiotherapy in improving superior central nervous system progression-free survival for patients with leptomeningeal metastasis.
Introduction
Leptomeningeal metastasis (LM) involves the spread of a cancer into the cerebrospinal fluid (CSF)–filled leptomeningeal space surrounding the brain and spinal cord.1,2 Dissemination of tumor cells through the central nervous system (CNS) results in a diverse presentation of neurologic symptoms that can be debilitating and life-threatening, resulting in substantial morbidity and mortality and requiring multidisciplinary management. The incidence of LM is rising, and the prognosis for patients with LM remains poor, with the median survival of patients often measured in weeks to months with standard treatment.3,4,5
Radiotherapy (RT) has long served as a pillar in managing LM, and involved-field RT (IFRT), such as partial-brain or whole-brain RT and focal spinal RT, has been the standard of care and is effective in palliation of symptomatic or bulky disease.6,7 However, because LM disseminates throughout the entire CNS compartment, IFRT cannot halt the progression of LM along the entire neuroaxis. As a result, IFRT has not been demonstrated to consistently improve survival.6,8,9 However, craniospinal irradiation (CSI) treats the whole leptomeningeal compartment; therefore, it may achieve superior symptom and disease control compared with IFRT.10,11 Due to its physical property in depositing the bulk of its energy at the last few millimeters of its range, proton CSI (pCSI) results in a minimal RT dose beyond the neuroaxis; thus, there are significantly fewer toxic effects compared with photon-based CSI.12
We previously reported in a phase 1b clinical trial that pCSI is safe and that there is promising CNS control for patients with solid tumor LM.13 In a subsequent phase 2 randomized study, we reported superior CNS progression-free survival (CNS-PFS) with pCSI, which led to the early discontinuation of the clinical trial as recommended by the data and safety monitoring committee (DSMC).13 In this article, we report the final results of the randomized phase 2 trial of pCSI vs IFRT in patients with LM.
Methods
We conducted an open-label, randomized, phase 2 trial to assess the efficacy of pCSI in patients with solid tumor malignant neoplasms with LM. Eligibility criteria were reported previously.14 The protocol (Supplement 1) was approved by the institutional review board, registered with ClinicalTrials.gov (NCT04343573), and completed according to the protocol and Good Clinical Practice guidelines. Written informed consent was obtained from the study participants. The study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. Patients were enrolled between April 2020 and October 2021.
Patients with metastatic non–small cell lung cancer (NSCLC) or breast cancer were randomly assigned in a 2:1 ratio to pCSI vs IFRT, as stratified by histology results and systemic disease status. Patients with other solid tumor histologies were enrolled to an exploratory pCSI group without random assignment. Protocol RT was delivered at a dose of 3 Gy ×10 daily fractions. Patients who received pCSI were treated with pencil beam scanning proton therapy15 to the entire CNS compartment. Patients who received photon IFRT for bulky and/or symptomatic disease underwent 3-dimensional photon planning. All patients received memantine prophylaxis.16
The primary end point was CNS-PFS, defined as from the time of randomization to CNS disease progression or death among patients with NSCLC and breast cancer. CNS progression was defined as new neurologic deficits that are not related to therapeutic intervention and/or progressive radiographic change using the Leptomeningeal Assessment in Neuro-Oncology scale17 and/or new positive CSF cytology results in the setting of a previously negative cytology. The secondary objectives were to (1) compare overall survival (OS) of pCSI and IFRT for patients with NSCLC or breast cancer LM, (2) evaluate CNS-PFS and OS for the patients in the exploratory cohort, (3) characterize patient-reported outcomes using the MD Anderson Symptom Inventory (MDASI)–Brain Tumor Module (MDASI-BT)18 and MDASI-Spine Tumor Module (SP),19 and (4) characterize neurocognitive outcomes. The correlation of outcomes with CSF and plasma circulating tumor DNA (ctDNA) was performed as exploratory objective.
Survival Outcomes
Enrollment of 81 patients with NSCLC and breast cancer LM would provide a 1-sided ∝ of .025 and 0.8 power based on a stratified log-rank test for an estimated 3 months CNS-PFS for IFRT and 6 months for pCSI. An interim analysis was planned after half of the expected events were observed. The Lan-Demets spending function was used. At the interim analysis, if P < .002 favored pCSI, the trial would stop for efficacy. If P > .29, the trial would stop for futility. At the final analysis, pCSI would be declared superior if P < .02. Accrual of the study was discontinued in October 2021 as recommended by the DSMC at the interim analysis due to the study meeting its primary end point.14
All accrued patients were analyzed following an intent-to-treat analysis when the last enrolled patient completed protocol follow-up at 12 months after enrollment. Patients continued standard-of-care follow-up thereafter. Patients were censored at the time of study withdrawal regardless of treatment initiation. The Kaplan-Meier method was used to estimate CNS-PFS and OS from randomization for patients with NSCLC or breast cancer or enrollment for patients with all other solid tumor histologies. Time to CNS progression from random assignment was also evaluated, with the CNS disease progression as the event and death without CNS disease progression as the competing event. A stratified log-rank test and Gray test were used to compare CNS-PFS, OS, and time to CNS progression between patients with NSCLC and breast cancer randomized to receive pCSI and IFRT. A nonstratified log-rank test was also performed on CNS-PFS and OS as a sensitivity analysis. Multivariable Cox proportional-hazards regression models were performed to examine associations of CNS-PFS and OS with treatment, age, Karnofsky performance status score, histology results, systematic disease status, baseline magnetic resonance imaging positivity for LM, and baseline brain metastases status. Backward variable selection based on a z test was applied, with systemic disease status always included in the models. For analyses that included the variable that did not satisfy the proportional-hazard assumption, the weighted Cox regression model20 was performed, with a weighted mean hazard ratio (HR) estimated. All statistical tests were performed in R (version 4.3.2; R Foundation for Statistical Computing) and R packages coxphw, lme4, and emmeans. Statistical significance was set at P < .05.
Patient-Reported and Neurocognitive Outcomes
For patient-reported outcomes, the MDASI-BT and MDASI-SP questions were organized into 4 main groups of interest (Table 1). Linear mixed-effects modeling (LMM) was used to evaluate group score change over time between treatment groups. The square root transformation was applied to a group score that considered skewed distributions. The models included fixed-effect treatment groups and random intercept and time effects. An interaction term between the treatment group and time was included. In the results, P values that tested linear time effect in each treatment group and P values that tested group score differences between treatment groups across all points were reported.
Table 1. Patient-Reported Outcome and Neurocognitive Function Assessments.
| Patient-reported outcome and neurocognitive function | Assessments |
|---|---|
| Patient-reported outcome | |
| Brain symptoms inventory | MD Anderson Symptom Inventory–Brain: symptoms (Q1-Q22), severity (Q1-Q13), brain tumor specific (Q14-Q22), interference (Q23-Q28) |
| Spine symptoms inventory | MD Anderson Symptom Inventory–Spine: symptoms (Q1-Q18), severity (Q1-Q13), spine tumor specific (Q14-Q18), interference (Q19-Q24) |
| Neurocognitive function | |
| Executive function | Trail Making Test parts A and B, Controlled Oral Word Association Test |
| Verbal memory | The Hopkins Verbal Learning Test—Revised: Total, Delayed Recall, Discrimination Index |
| Attention and working memory | Longest Digit Span Forward, Longest Digit Span Backward, Longest Number Sequencing (WAIS-IV), Brief Test of Attention |
| Self-reported scales | The Center for Epidemiological Study–Depression, Functional Assessment of Chronic Illness Therapy-Fatigue Subscale, version 4 |
Abbreviations: Q, question; WAIS-IV, Wechsler Adult Intelligence Scale, Fourth Edition.
A subset of patients who received pCSI completed standardized neurocognitive tests across multiple domains21,22 and self-reported mood and fatigue scales23,24 (Table 1) at baseline and 3, 6, and 12 months post-pCSI. Means across the 4 points were estimated for each neurocognitive test score and self-reported score using LMM predicting the score by point. A random intercept was included to account for multiple assessments per patient. Each model included all patients who completed neurocognitive testing on at least 2 points. These models accounted for multiple assessments per patient, as they provided unbiased estimates in the presence of missing data under the assumption that data were missing at random.25 Mean changes between pairs of points were similarly estimated from the LMM and tested for statistical significance using model-based contrasts.
ctDNA
CSF and plasma ctDNA was collected at baseline and 3-month intervals following completion of treatment. All ctDNA samples underwent targeted sequencing via the Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets, a hybridization, capture-based next-generation sequencing assay covering the protein-coding exons of 505 genes.26 Somatic variant calls were made based on paired germline tissue samples, which were available for all except 3 patients who were instead compared with a pooled normal reference. Variants were independently called as previously described.27
Results
Patients were enrolled before the DSMC’s recommendation of early discontinuation of the trial (median [IQR] age, 59 [50-65] years). Forty-two patients with NSCLC and breast cancer were randomized to pCSI and 21 were randomized to IFRT, and 35 patients with other solid tumor histologies were enrolled to the exploratory pCSI group (Figure 128). The baseline patient and tumor characteristics were balanced between the randomized groups. Most patients (86 [88%]) had newly diagnosed LM, and patients with disease-harboring targetable variants all had LM progression while receiving first-line targeted therapy (Table 2).
Figure 1. Trial Profile28.
CSI indicates craniospinal irradiation; IFRT, involved-field radiotherapy; NSCLC, non–small cell lung cancer.
Table 2. Clinical Characteristics of 98 Patients.
| Characteristic | No. (%) | ||
|---|---|---|---|
| Randomized pCSI | Randomized photon IFRT | Exploratory pCSI | |
| No. of patients | 42 | 21 | 35 |
| Age at registration, median (range), y | 57 (37-79) | 61 (31-75) | 61 (27-77) |
| Sex | |||
| Female | 34 (81) | 18 (86) | 20 (57) |
| Male | 8 (19) | 3 (14) | 15 (43) |
| KPS, median (range) | 80 (60-90) | 80 (60-90) | 80 (60-90) |
| Histology | |||
| NSCLC | 24 (57) | 12 (57) | NA |
| EGFR-positive | 12 (29) | 7 (33) | |
| Targeted therapy before enrollment | |||
| Osimertinib | 12 (29) | 7 (33) | |
| ALK/ROS1-positive | 2 (5) | ||
| Targeted therapy before enrollment | |||
| Alectinib | 1 (2) | NA | |
| Lorlatinib | 1 (2) | ||
| Breast | 17 (43) | 9 (43) | |
| ERBB2-positive | 6 (14) | 4 (19) | |
| Targeted therapy before enrollment | |||
| Tucatinib | 2 (5) | 1 (5) | |
| Trastuzumab emtansine | 2 (5) | ||
| Trastuzumab deruxtecan | 1 (5) | ||
| Lapatinib | 1 (2) | 1 (5) | |
| Trastuzumab | 1 (2) | 1 (5) | |
| Ovarian | NA | NA | 7 (20) |
| Esophageal | 6 (17) | ||
| Melanoma | 6 (17) | ||
| Colorectal | 5 (14) | ||
| Head and neck | 3 (9) | ||
| Pancreatic | 2 (6) | ||
| SCLC | 2 (6) | ||
| Anal | 1 (3) | ||
| Biliary | 1 (3) | ||
| Prostate | 1 (3) | ||
| Unknown primary | 1 (3) | ||
| Systemic disease status at enrollment | |||
| Active | 22 (52) | 11 (52) | 20 (57) |
| Stable/none | 20 (48) | 10 (48) | 15 (43) |
| Baseline evaluation | |||
| Positive MRI | 38 (91) | 21 (100) | 33 (94) |
| Brain only | 9 (21) | 7 (33) | 8 (23) |
| Spine only | 3 (7) | 3 (14) | 5 (14) |
| Both brain and spine | 26 (62) | 11 (52) | 20 (57) |
| Positive cytology | 28 (67) | 11 (52) | 23 (66) |
| LM status | |||
| Newly diagnosed | 35 (83) | 18 (86) | 33 (94) |
| Known LM with progression | 7 (17) | 3 (14) | 2 (6) |
| Parenchymal brain metastases at enrollment | |||
| Yes | 28 (67) | 15 (71) | 12 (34) |
| No | 14 (33) | 6 (29) | 23 (66) |
| Median lines of prior systemic therapy received for metastatic disease (range) | 2 (0-8) | 2 (0-8) | 1 (0-5) |
| Photon IFRT fields | |||
| WBRT | NA | 9 (43) | NA |
| Focal spine RT | 1 (5) | ||
| WBRT + focal spine RT | 8 (38) | ||
| Description of focal spine RT fields | |||
| Cauda equina | NA | 6 (29) | NA |
| Cervical and upper thoracic spinal cord | 4 (19) | ||
| Lower thoracic spinal cord | 1 (5) | ||
Abbreviations: ALK/ROS1+, anaplastic lymphoma kinase and c-ros oncogene 1 rearrangements; EGFR, epidermal growth factor receptor gene exon 19 or L858R variants; IFRT, involved-field radiotherapy; KPS, Karnofsky performance score; LM, leptomeningeal metastasis; MRI, magnetic resonance imaging; NA, not applicable; NSCLC, non–small cell lung cancer; pCSI, proton craniospinal irradiation; RT, radiotherapy; SCLC, small cell lung cancer; WBRT, whole-brain radiotherapy.
Survival Outcomes
Of the 63 patients with NSCLC or breast cancer LM, 17 of 42 patients (40%) who received pCSI, and 16 of 21 patients (76%) who received IFRT experienced CNS progression, 25 (60%) and 17 (81%) died at the end of protocol period, respectively. Of the 17 patients who received pCSI and experienced CNS progression, 2 (12%) had clinical, 5 (29%) radiographic, 4 (24%) CSF cytologic, 5 (29%) clinical and radiographic, and 1 (6%) radiographic and cytologic progression. Of the 25 patients who received pCSI and died, 5 (20%) died with CNS progression, 10 (40%) with systemic progression, 6 (24%) with CNS and systemic progression, and 4 (16%) of other causes. Of the 16 patients who received IFRT and experienced CNS progression, 4 (25%) had clinical, 3 (19%) radiographic, 7 (44%) clinical and radiographic, and 2 (12%) clinical and cytologic progression. Of the 17 patients who received IFRT and died, 9 (53%) died with CNS progression, 6 (35%) with CNS and systemic progression, and 2 (12%) of other causes.
The median follow-up was 26.8 months (95% CI, 18.5 to not achieved) with CNS progression or death. At 6 months, patients who received IFRT had significantly higher risk of CNS progression (cumulative incidence rate, 88%; 95% CI, 51%-98%) compared with those who received pCSI (cumulative incidence rate, 22%; 95% CI, 9.5%-38%; P < .001; Figure 2A). The 6-month CNS-PFS was 66% for pCSI patients (95% CI, 51%-84%) and 6.0% for IFRT patients (95% CI, 0.9%-40%). The median CNS-PFS was 8.2 months (95% CI, 6.6-15.3 months) for pCSI and 2.3 months (95% CI, 1.2-4.0 months; P < .001; Figure 2B) for IFRT patients. In multivariate analysis, pCSI (HR, 0.14; 95% CI, 0.06-0.30; P < .001) and stable systemic disease (HR, 0.52; 95% CI, 0.28-0.97; P = .04) were significantly associated with improved CNS-PFS. The 6-month OS was 72% for pCSI patients (95% CI, 58%-89%) and 47% for IFRT patients (95% CI, 28%-78%). The median OS was 11.3 months (95% CI, 7.5-18.3 months) for pCSI and 4.9 months (95% CI, 3.9-15.0 months; P = .04; Figure 2C) for IFRT patients. In multivariate analysis, pCSI (HR, 0.50; 95% CI, 0.27-0.94; P = .03) remained independently significantly associated with improved OS (Table 3).
Figure 2. Treatment Outcomes.
Patients who were randomized to proton craniospinal irradiation (pCSI) had significantly improved time to central nervous system progression (CNS TTP) (A; pCSI: 17 events; photon IFRT: 16 events), central nervous system progression-free survival (CNS-PFS) (B; pCSI: 29 events; photon IFRT: 17 events), and overall survival (C; pCSI: 25 events; photon IFRT: 17 events). IFRT indicates involved-field radiotherapy.
Table 3. Univariable and Multivariable Cox Proportional Hazards Regression Models With CNS-PFS and OS as Outcomes.
| Variables | HR (95% CI) | P value |
|---|---|---|
| CNS-PFS univariable analysis | ||
| pCSI (reference: photon IFRT) | 0.14 (0.06-0.30) | <.001 |
| Age at randomization, unit = 10 y | 1.10 (0.84-1.43) | .50 |
| Primary disease, breast (reference: NSCLC) | 1.03 (0.57-1.85) | >.99 |
| Stable/nonsystemic disease (reference: active) | 0.52 (0.28-0.96) | .04 |
| KPS score ≥80 (reference: <80) | 0.80 (0.44-1.45) | .50 |
| MRI positivity at baseline | 1.02 (0.3-2.87) | >.99 |
| Brain metastases at baselinea | 1.17 (0.53-2.57) | .70 |
| CNS-PFS multivariable analysis | ||
| pCSI (reference: photon IFRT) | 0.14 (0.06-0.30) | <.001 |
| Stable/nonsystemic disease (reference: active) | 0.52 (0.28-0.97) | .04 |
| OS univariable analysis | ||
| pCSI (reference: photon IFRT) | 0.47 (0.25-0.88) | .02 |
| Age at randomization, unit = 10 y | 1.31 (0.99-1.73) | .06 |
| Primary disease, breast (reference: NSCLC) | 0.90 (0.49-1.67) | .70 |
| Stable/nonsystemic disease (reference: active) | 0.49 (0.26-0.93) | .03 |
| KPS score ≥80 (reference: <80) | 0.85 (0.45-1.61) | .60 |
| MRI positivity for LM at baseline | 0.72 (0.25-2.04) | .50 |
| Brain metastases at baseline | 1.40 (0.70-2.79) | .30 |
| OS multivariable analysis | ||
| pCSI (reference: photon IFRT) | 0.50 (0.27-0.94) | .03 |
| Age at randomization, unit = 10 y | 1.37 (1.03-1.82) | .05 |
Abbreviations: CNS-PFS, central nervous system progression-free survival; HR, hazard ratio; IFRT, involved-field radiotherapy; KPS, Karnofsky performance score; LM, leptomeningeal metastasis; MRI, magnetic resonance imaging; NSCLC, non–small cell lung cancer; OS, overall survival; pCSI, proton craniospinal irradiation.
The weighted mean HR was estimated.
Of the 35 patients with other solid tumor histology LM in the exploratory cohort, all received pCSI, and 13 patients (37%) experienced CNS progression: 2 clinical, 4 radiographic, 2 cytologic, and 5 radiographic and clinical progression. Twenty-seven patients (77%) died: 7 with CNS progression, 11 with systemic progression, 6 with CNS and systemic progression, and 3 of other causes. The median follow-up was 18.2 months with CNS progression or death. At 6 months, 21% (95% CI, 8.4%-38%) of the patients experienced CNS progression. The CNS-PFS at 6 months was 43% (95% CI, 28%-66%), with a median CNS-PFS of 5.8 months (95% CI, 4.4-9.1 months). The OS at 6 months was 54% (95% CI, 38%-76%), with a median OS of 7.0 months (95% CI, 5.4-10.6).
Patient-Reported and Neurocognitive Outcomes
A summary of MDASI groups during treatment and follow-up points is listed in eTable 1 in Supplement 2. There was no significant difference in MDASI group scores by treatment at baseline, during treatment, and at follow-up after protocol therapy while patients remained on the protocol-specified treatment (eFigure 1 in Supplement 2). When evaluating the change in MDASI group scores over time, there was no significant difference between photon IFRT and pCSI, with a trend toward improved interference symptoms (relations with other people, enjoyment in life, mood, walking, activity, and work) with pCSI compared with IFRT (eTable 2 in Supplement 2).
A subset of 12 patients who received pCSI underwent serial neurocognitive testing. For all neurocognitive tests and self-report scales, except Trial Making Test (TMT) parts A and B, 12 patients completed the assessments at baseline, 11 at 3 months, 9 at 6 months, and 4 at 12 months. For TMT-A and TMT-B, 12 patients completed the tests at baseline, 8 at 3 months, 5 at 6 months, and 3 at 12 months. The mean age (SD) was 57 (14) years (range, 31-77 years), 9 (75%) were women, 11 (92%) were right-handed, and 5 (42%) had RT for brain metastases before study enrollment.
LMM analyses showed a significant decline in graphomotor speed (Trail Making Test part A; eFigure 2 in Supplement 2) and verbal recognition memory (Hopkins Verbal Learning Test-Revised Discrimination Index) from baseline to 3 months, with scores returned to and remained stable compared with baseline at 6 and 12 months. Verbal learning (Hopkins Verbal Learning Test–Revised Total Learning) scores were similar at 3, 6, and 12 months, with all lower than baseline. In timed set-shifting (Trails Making Test part B), scores decreased from the baseline across all follow-up assessments, with scores significantly different from baseline at 6 months and 12 months. There was a significant improvement in self-reported symptoms of depression from baseline to 6 months. There was no significant change in attention and working memory or self-reported fatigue between the baseline and the follow-up period.
ctDNA
Baseline CSF ctDNA were available for 24 pCSI and 15 IFRT patients, with 18 pCSI and 8 IFRT baseline CSF samples passing sequencing quality criteria. CSF ctDNA was analyzed in at least 1 follow-up sample for 10 pCSI and 2 IFRT patients. Baseline plasma ctDNA was measured for 20 pCSI and 8 IFRT patients, with 15 pCSI and 7 IFRT samples passing quality criteria. Plasma ctDNA was analyzed in at least 1 follow-up sample for 13 pCSI and 2 IFRT patients.
Recurrent variants, median baseline variant allele fraction (VAF; calculated as the mean proportion of sequencing reads that show any alteration), and mean change in VAF at first follow-up measurement were evaluated. For pCSI and IFRT, baseline VAF was not associated with CNS-PFS in CSF (HR, 0.54 [95% CI, 0.18-1.63; P = .27] and 0.48 [95% CI, 0.10-2.25; P = .35], respectively) or plasma (HR, 2.50 [95% CI, 0.69-9.0; P = .16] and 0.35 [95% CI, 0.06-2.00; P = .20], respectively). By patient, more variants that were unique to the CSF (median, 9; range, 2-56) were observed relative to variants unique to the plasma (median, 1; range, 0-15) or shared variants (median, 0; range, 0-15). A decrease in CSF ctDNA VAF and plasma ctDNA VAF at first follow-up was seen in 7 of 10 (70%) and 5 of 13 (38%) patients after pCSI, respectively. Change in VAF was not associated with CNS-PFS in CSF (HR, 0.87; 95% CI, 0.15-5.15; P = .88) or plasma (HR, 0.93; 95% CI, 0.25-3.48; P = .91).
Discussion
In this final analysis of what is to our knowledge the first randomized clinical trial to evaluate RT for patients with solid tumor LM, we observed superior CNS-PFS (median, 8.2 vs 2.3 months) and OS (median, 11.0 vs 4.9 months) with pCSI compared with IFRT for patients with NSCLC or breast cancer LM. On multivariate analysis, pCSI was independently associated with improved CNS-PFS and OS. Patients with stable systemic disease at the time of enrollment had significantly improved CNS-PFS. In the exploratory cohort of patients with LM of other solid tumor histologies, a patient population with very limited CNS-directed treatment options, we observed promising CNS-PFS (median, 5.8 months) and OS (median, 7.0 months) with pCSI. While our study was not designed to compare the randomized and exploratory pCSI cohorts, the difference in CNS-PFS may reflect a more durable response to pCSI in certain histologies and warrants further investigation. In addition to this trial, recent retrospective data have also demonstrated that pCSI may be a promising treatment option for patients with solid tumor LM. Gal et al29 reported outcomes after pCSI for 38 patients with LM (87% of patients with NSCLC or breast cancer), with a median OS of 10.1 months. Lam et al30 reported 45 patients with LM (73% with NSCLC or breast cancer) who were treated with pCSI with a median OS of 13.7 months. These results highlighted the need to address the entire CNS compartment for patients with LM to meaningfully affect CNS disease control and survival. Furthermore, emerging data suggest that photon-based volumetric modulated arc therapy CSI may have an acceptable safety profile31 and the benefit of not being limited to proton facilities, although prospective evaluation is warranted to confirm its safety and efficacy for this patient population.
An important aspect of this trial was the inclusion of patient-reported and neurocognitive outcomes. A trend toward improvement with daily function with pCSI was observed using the MDASI scales, although our ability to measure this was limited by the small number of patients in the analysis, namely those patients who received IFRT and discontinued the study at CNS progression and thus had with no further patient-reported data collected. This also limited our ability to determine whether pCSI resulted in more durable improvement. In terms of neurocognitive function, declines in graphomotor speed, executive function, and verbal memory after pCSI were observed in a subset of patients, which overall was consistent with adverse neurocognitive effects of whole-brain RT.32,33,34 However, we observed no significant change after pCSI in attention and working memory or self-reported fatigue and an improvement in self-reported symptoms of depression. Cumulative adverse effects of prior lines of therapy, concurrent therapy, and systemic and CNS disease may also contribute to neurocognitive decline in these patients. It is imperative for clinical trials for patients with LM to not only assess treatment efficacy but also its effect on quality of life to better understand specific patient needs after treatment. Methods to preserve neurocognitive function, such as hippocampal avoidance, should be explored for select patients.
Limitations
There were several limitations to this study, as previously described, including the relatively small sample size that resulted from early discontinuation of the trial and that the study was not designed to evaluate the effect of pCSI on OS. An ongoing randomized phase 3 trial being conducted through NRG Oncology (NCT06500481) aims to assess whether pCSI improves OS in patients with NSCLC or breast cancer LM and may also provide important data on which patients could benefit most from pCSI.
Conclusions
In this randomized clinical trial, we continued to observe improved CNS-PFS with pCSI compared with IFRT, meeting the primary end point of the study. We also continued to observe an improvement in OS with pCSI. A trend toward improved self-reported daily function was observed for patients who received pCSI. While declines in graphomotor speed, executive function, and verbal memory were observed after pCSI, there was no significant change in attention, working memory, and self-reported fatigue, and an improvement in self-reported symptoms of depression was found after pCSI. The evidence presented in this study suggests that patients should be considered for pCSI if available. An ongoing phase 3 trial is evaluating the effect of pCSI on OS in patients with NSCLC or breast cancer LM.
Trial protocol
eTable 1. Summary of MDASI group mean scores during treatment and follow up time points
eTable 2. Time effect on MDASI group scores and treatment modalities
eFigure 1. Comparison of MDASI-BT and MDASI-SP scores. Scores reported on study for patients who received pCSI and IFRT
eFigure 2. Changes in neurocognitive testing overtime. Changes in neurocognitive testing in cohort of patients who received pCSI
Data sharing statement
References
- 1.Wasserstrom WR, Glass JP, Posner JB. Diagnosis and treatment of leptomeningeal metastases from solid tumors: experience with 90 patients. Cancer. 1982;49(4):759-772. doi: 10.1002/1097-0142(19820215)49:4<759::AID-CNCR2820490427>3.0.CO;2-7 [DOI] [PubMed] [Google Scholar]
- 2.Clarke JL, Perez HR, Jacks LM, Panageas KS, Deangelis LM. Leptomeningeal metastases in the MRI era. Neurology. 2010;74(18):1449-1454. doi: 10.1212/WNL.0b013e3181dc1a69 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Omuro AM, Kris MG, Miller VA, et al. High incidence of disease recurrence in the brain and leptomeninges in patients with nonsmall cell lung carcinoma after response to gefitinib. Cancer. 2005;103(11):2344-2348. doi: 10.1002/cncr.21033 [DOI] [PubMed] [Google Scholar]
- 4.Lamba N, Wen PY, Aizer AA. Epidemiology of brain metastases and leptomeningeal disease. Neuro Oncol. 2021;23(9):1447-1456. doi: 10.1093/neuonc/noab101 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Morikawa A, Jordan L, Rozner R, et al. Characteristics and outcomes of patients with breast cancer with leptomeningeal metastasis. Clin Breast Cancer. 2017;17(1):23-28. doi: 10.1016/j.clbc.2016.07.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Buszek SM, Chung C. Radiotherapy in leptomeningeal disease: a systematic review of randomized and non-randomized trials. Front Oncol. 2019;9:1224. doi: 10.3389/fonc.2019.01224 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Barbour AB, Kotecha R, Lazarev S, et al. Radiation therapy in the management of leptomeningeal disease from solid tumors. Adv Radiat Oncol. 2023;9(2):101377. doi: 10.1016/j.adro.2023.101377 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.El Shafie RA, Böhm K, Weber D, et al. Palliative radiotherapy for leptomeningeal carcinomatosis—analysis of outcome, prognostic factors, and symptom response. Front Oncol. 2019;8:641. doi: 10.3389/fonc.2018.00641 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Gani C, Müller AC, Eckert F, et al. Outcome after whole brain radiotherapy alone in intracranial leptomeningeal carcinomatosis from solid tumors. Strahlenther Onkol. 2012;188(2):148-153. doi: 10.1007/s00066-011-0025-8 [DOI] [PubMed] [Google Scholar]
- 10.Devecka M, Duma MN, Wilkens JJ, et al. Craniospinal irradiation (CSI) in patients with leptomeningeal metastases: risk-benefit-profile and development of a prognostic score for decision making in the palliative setting. BMC Cancer. 2020;20(1):501. doi: 10.1186/s12885-020-06984-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.El Shafie RA, Böhm K, Weber D, et al. Outcome and prognostic factors following palliative craniospinal irradiation for leptomeningeal carcinomatosis. Cancer Manag Res. 2019;11:789-801. doi: 10.2147/CMAR.S182154 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Brown AP, Barney CL, Grosshans DR, et al. Proton beam craniospinal irradiation reduces acute toxicity for adults with medulloblastoma. Int J Radiat Oncol Biol Phys. 2013;86(2):277-284. doi: 10.1016/j.ijrobp.2013.01.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yang TJ, Wijetunga NA, Yamada J, et al. Clinical trial of proton craniospinal irradiation for leptomeningeal metastases. Neuro Oncol. 2021;23(1):134-143. doi: 10.1093/neuonc/noaa152 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Yang JT, Wijetunga NA, Pentsova E, et al. Randomized phase II trial of proton craniospinal irradiation versus photon involved—field radiotherapy for patients with solid tumor leptomeningeal metastasis. J Clin Oncol. 2022;40(33):3858-3867. doi: 10.1200/JCO.22.01148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kotecha R, La Rosa A, Mehta MP. How proton therapy fits into the management of adult intracranial tumors. Neuro Oncol. 2024;26(12)(suppl 2):S26-S45. doi: 10.1093/neuonc/noad183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Brown PD, Pugh S, Laack NN, et al. ; Radiation Therapy Oncology Group . Memantine for the prevention of cognitive dysfunction in patients receiving whole-brain radiotherapy: a randomized, double-blind, placebo-controlled trial. Neuro Oncol. 2013;15(10):1429-1437. doi: 10.1093/neuonc/not114 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Le Rhun E, Devos P, Boulanger T, et al. ; European Organisation for Research and Treatment of Cancer Brain Tumor Group Central Nervous System Metastases Committee and the EORTC BTG Imaging Committee . The RANO Leptomeningeal Metastasis Group proposal to assess response to treatment: lack of feasibility and clinical utility and a revised proposal. Neuro Oncol. 2019;21(5):648-658. doi: 10.1093/neuonc/noz024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Armstrong TS, Mendoza T, Gning I, et al. Validation of the M.D. Anderson Symptom Inventory Brain Tumor Module (MDASI-BT). J Neurooncol. 2006;80(1):27-35. doi: 10.1007/s11060-006-9135-z [DOI] [PubMed] [Google Scholar]
- 19.Cleeland CS, Mendoza TR, Wang XS, et al. Assessing symptom distress in cancer patients: the M.D. Anderson Symptom Inventory. Cancer. 2000;89(7):1634-1646. doi: 10.1002/1097-0142(20001001)89:7<1634::AID-CNCR29>3.0.CO;2-V [DOI] [PubMed] [Google Scholar]
- 20.Schemper M, Wakounig S, Heinze G. The estimation of average hazard ratios by weighted Cox regression. Stat Med. 2009;28(19):2473-2489. doi: 10.1002/sim.3623 [DOI] [PubMed] [Google Scholar]
- 21.Lezak MD. Neuropsychological Assessment. 5th ed. Oxford University Press; 2012. [Google Scholar]
- 22.Wefel JS, Vardy J, Ahles T, Schagen SB. International Cognition and Cancer Task Force recommendations to harmonise studies of cognitive function in patients with cancer. Lancet Oncol. 2011;12(7):703-708. doi: 10.1016/S1470-2045(10)70294-1 [DOI] [PubMed] [Google Scholar]
- 23.Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1:385-401. doi: 10.1177/014662167700100306 [DOI] [Google Scholar]
- 24.Cella DFACIT. Manual of the Functional Assessment of Chronic Illness Therapy. Evanston Northwestern Healthcare; 1997. [Google Scholar]
- 25.Diggle PH, Liang KY, Zeger S. Analysis of Longitudinal Data. 2nd ed. Oxford University Press; 2002. [Google Scholar]
- 26.Cheng DT, Mitchell TN, Zehir A, et al. Memorial Sloan Kettering–integrated mutation profiling of actionable cancer targets (MSK-IMPACT): a hybridization capture-based next-generation sequencing clinical assay for solid tumor molecular oncology. J Mol Diagn. 2015;17(3):251-264. doi: 10.1016/j.jmoldx.2014.12.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wijetunga NA, Goglia AG, Weinhold N, et al. Dynamic mutational landscape of cerebrospinal fluid circulating tumor DNA and predictors of survival after proton craniospinal irradiation for leptomeningeal metastases. Clin Cancer Res. 2023;29(4):775-783. doi: 10.1158/1078-0432.CCR-22-2434 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hopewell S, Chan AW, Collins GS, et al. CONSORT 2025 statement: updated guideline for reporting randomized trials. JAMA. 2025;333(22):1998-2005. doi: 10.1001/jama.2025.4347 [DOI] [PubMed] [Google Scholar]
- 29.Gal O, La Rosa A, Hall MD, et al. Proton craniospinal irradiation for patients with solid tumor leptomeningeal disease: real-world feasibility, toxicity, and outcome analysis. Cancers (Basel). 2025;17(6):1046. doi: 10.3390/cancers17061046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lam K, Nasr LF, Andersen CR, et al. Early outcomes from proton craniospinal irradiation for leptomeningeal metastasis from solid tumors. Adv Radiat Oncol. 2024;10(2):101697. doi: 10.1016/j.adro.2024.101697 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Perlow HK, Matsui JK, Ewing A, et al. Volumetric modulated arc therapy craniospinal irradiation utilizing a vertebral body sparing approach: an alternative approach to improve access and minimize toxicity. Adv Radiat Oncol. 2023;9(3):101424. doi: 10.1016/j.adro.2023.101424 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.DeAngelis LM, Posner JB. Side effects of radiation therapy. In: DeAngelis LM, Posner JB, eds. Neurologic Complications of Cancer. 2nd ed. Oxford University Press; 2009:551-555. [Google Scholar]
- 33.Wefel JS, Kayl AE, Meyers CA. Neuropsychological dysfunction associated with cancer and cancer therapies: a conceptual review of an emerging target. Br J Cancer. 2004;90(9):1691-1696. doi: 10.1038/sj.bjc.6601772 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Raber J, Rola R, LeFevour A, et al. Radiation-induced cognitive impairments are associated with changes in indicators of hippocampal neurogenesis. Radiat Res. 2004;162(1):39-47. doi: 10.1667/RR3206 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Trial protocol
eTable 1. Summary of MDASI group mean scores during treatment and follow up time points
eTable 2. Time effect on MDASI group scores and treatment modalities
eFigure 1. Comparison of MDASI-BT and MDASI-SP scores. Scores reported on study for patients who received pCSI and IFRT
eFigure 2. Changes in neurocognitive testing overtime. Changes in neurocognitive testing in cohort of patients who received pCSI
Data sharing statement
