Proton Spine-and-Brain Radiation Doubles Survival in Cancer Spread to Brain Linings, Review Finds

Leptomeningeal disease, the spread of cancer cells into the delicate membranes and cerebrospinal fluid that bathe the brain and spinal cord, has long been one of oncology’s most feared diagnoses. Patients with solid tumors that seed this compartment—most often breast cancer, non-small cell lung cancer, and melanoma—have historically survived only weeks to a few months. Now a systematic review published in the Journal of Neuro-Oncology suggests that the choice of radiation technique may matter far more than clinicians have assumed, with proton-based craniospinal irradiation associated with reported median survival more than double that seen with conventional whole-brain radiotherapy.

The review, led by Mishaal Munir and Rupesh Kotecha of the Baptist Health Herbert Wertheim Cancer Institute in Miami, together with colleagues at Florida International University and NYU Grossman School of Medicine, followed the PRISMA 2020 reporting framework. The team searched PubMed for studies of adults with solid-tumor leptomeningeal disease treated with one of three radiation approaches: whole-brain radiotherapy or other non-craniospinal photon techniques, photon craniospinal irradiation, or proton craniospinal irradiation. From 106 initially identified records, the investigators screened 91 after removing duplicates, reviewed 79 full-text reports, and ultimately included 32 unique studies—one of which contributed separate proton and photon craniospinal arms, bringing the total to 33 modality-specific analytic contributions.

The headline finding concerns overall survival. The median of the reported median survival values was 4.5 months across the 21 whole-brain radiotherapy studies, 4.1 months across the 6 photon craniospinal irradiation studies, and 10.7 months across the 6 proton craniospinal irradiation studies. Study-level medians in the proton group ranged from 4.0 to 13.7 months, while the whole-brain group spanned 1.9 to 14.5 months and the photon craniospinal group clustered between 1.9 and 4.8 months. Progression-free survival, reported far less consistently, averaged 3.9 months in the whole-brain group and 4.6 months in the proton group, with no extractable value for photon craniospinal treatment.

Why would protons make such a difference? The answer lies in fundamental radiation physics. Conventional photon beams, whether generated by linear accelerators for whole-brain or craniospinal treatment, deposit dose along their entire path and continue through the body on the exit side. When the target is the entire craniospinal axis—brain, spinal canal, and the cerebrospinal fluid pathways where leptomeningeal disease spreads—photons inevitably irradiate vertebral bone marrow, the heart, lungs, esophagus, and bowel on entry and exit. Proton beams, by contrast, can be shaped to release most of their energy at a precise depth, the Bragg peak, dramatically reducing exit dose to anterior organs and potentially sparing the marrow reservoirs that produce blood cells.

That dosimetric advantage shows up clearly in the toxicity data, though the authors caution that reporting was incomplete and heterogeneous. Among studies with extractable denominators, crude pooled rates of grade 3 or higher hematologic toxicity were 2.8 percent for whole-brain radiotherapy (4 of 145 patients), 80.0 percent for photon craniospinal irradiation (20 of 25), and 59.4 percent for proton craniospinal irradiation (38 of 64). Severe lymphopenia—the depletion of infection-fighting lymphocytes, which is increasingly recognized as a driver of poor outcomes in brain radiotherapy—dominated the craniospinal figures, affecting 80.0 percent of evaluable photon craniospinal patients and 58.6 percent of proton patients, versus 26.7 percent of evaluable whole-brain patients. Severe anemia, neutropenia, and thrombocytopenia were less common but more frequent in the photon craniospinal group.

The review’s context matters enormously here. The proton craniospinal literature is younger and, by design, more selective: these studies generally enrolled fitter patients treated in the modern era, often alongside CNS-active systemic therapies such as targeted agents and intrathecal treatments. The photon craniospinal studies, by contrast, largely date from older treatment periods with less conformal techniques and more protracted fractionation schedules. The authors are explicit that the numerical survival gap does not establish proton superiority over photon craniospinal irradiation, and they frame all cross-modality comparisons as descriptive and hypothesis-generating rather than estimates of comparative efficacy.

Nevertheless, the findings align with a striking randomized signal. In a phase II trial by Jonathan Yang and colleagues, proton craniospinal irradiation improved central nervous system progression-free survival to 8.2 months versus 2.3 months, and overall survival to 11.3 months versus 4.9 months, compared with photon involved-field radiotherapy in patients with breast cancer or non-small cell lung cancer leptomeningeal metastasis—without an increase in serious treatment-related adverse events. The hazard ratio for overall survival in multivariable analysis was 0.50. Yet the comparator in that trial was involved-field radiotherapy, not modern photon craniospinal treatment, leaving the proton-versus-photon craniospinal question formally unresolved. A real-world cohort from Frechette and colleagues reported a median survival of just 3.0 months with proton craniospinal irradiation, underscoring how powerfully patient selection shapes outcomes in this disease.

The whole-brain radiotherapy literature, the largest and most heterogeneous body of evidence, reflects decades of palliative practice. Whole-brain treatment can relieve intracranial symptoms and address coexisting brain metastases, but it does not cover the spinal cerebrospinal fluid compartment where leptomeningeal disease so often progresses. The 4.5-month median survival observed in this review is consistent with the historically grim prognosis documented in prior syntheses. Notably, one study excluded from quantitative pooling—Perlow and colleagues’ vertebral body–sparing photon craniospinal technique—reported no grade 3 or higher toxicity, hinting that hematologic toxicity may be as much a matter of technique as of particle type.

What comes next could settle the debate. The phase III NRG-BN014/RADIATE-LM trial (NCT06500481) is evaluating proton craniospinal irradiation against involved-field radiotherapy in breast cancer and non-small cell lung cancer leptomeningeal metastasis, though its comparator again leaves proton versus modern photon craniospinal irradiation unanswered. The review’s limitations are real: only PubMed was searched, no formal risk-of-bias scoring was applied, most included studies were retrospective, and progression endpoints were inconsistently defined. Still, by placing three radiation modalities in a single descriptive framework and pairing survival with toxicity, the analysis gives oncologists a benchmark for a decision that increasingly demands balancing potential disease control against treatment burden—and it signals that for a disease once considered uniformly terminal, the technology used to deliver radiation may be reshaping what survival looks like.

Subject of Research: Radiotherapy modalities and outcomes in solid-tumor leptomeningeal disease

Article Title: Survival and hematologic toxicity after whole-brain radiotherapy, photon craniospinal irradiation, and proton craniospinal irradiation for solid-tumor leptomeningeal disease: a systematic review

Article References: Survival and hematologic toxicity after whole-brain radiotherapy, photon craniospinal irradiation, and proton craniospinal irradiation for solid-tumor leptomeningeal disease: a systematic review. (n.d.). https://doi.org/10.1007/s11060-026-05776-0

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05776-0

Keywords: leptomeningeal disease, proton craniospinal irradiation, photon craniospinal irradiation, whole-brain radiotherapy, systematic review, overall survival, hematologic toxicity, lymphopenia, brain metastases, radiation oncology, breast cancer, non-small cell lung cancer