Leptomeningeal Metastasis in Solid Tumors - Toward a New Era of CSF-Directed Precision Oncology

Maciej M Mrugala^1*Katherine B Peters²

• ¹Mayo Clinic Arizona, Scottsdale, United States
• ²Duke Cancer Institute, Durham, United States

A complementary case report by Li and colleagues presents a unique clinical narrative: a 63-year-old woman with LM from BRAF non-V600E mutant NSCLC experienced 24 months of survival following treatment with intrathecal chemotherapy with pemetrexed and trametinib, an orally available inhibitor of MEK1 and MEK2 (4). This outcome is extraordinary in the context of EGFR/ALK-wild-type NSCLC, where median survival in LM typically ranges from 2 to 4 months.The case report is important not only because of its duration of disease control but also because it demonstrates the clinical utility of CSF-based molecular profiling. While the primary tumor lacked actionable targets, CSF next-generation sequencing (NGS) revealed multiple non-V600E BRAF mutations and CDKN2A loss-mutations rarely addressed in standard NSCLC care. This critical result allowed clinicians to pursue a rational off-label treatment with trametinib, which appeared to stabilize the disease for nearly a year. This patient's course also highlights the inadequacies of traditional diagnostic tools. MRI failed to identify LM definitively; diagnosis relied on CSF cytology and pressure measurement. This finding reiterates the urgent need for improved CSF diagnostics-not just to confirm LM, but to inform treatment decisions based on evolving tumor genomics. In a comprehensive review, Dr. Elena Pentsova outlines how CSF-based liquid biopsies-specifically circulating tumor cells (CSF-CTCs) and circulating tumor DNA (CSF-ctDNA)-are transforming the diagnosis and management of CNS metastases (5). Traditional tools, such as MRI and CSF cytology, suffer from low sensitivity and specificity. CSF liquid biopsy offers a minimally invasive and biologically informative alternative.Studies show that CSF-CTC counts (≥3 cells/3ml) can identify LM with high diagnostic accuracy (93% sensitivity, 95% specificity). Meanwhile, CSF-ctDNA can reveal tumor mutations, resistance mechanisms, and clonal divergence from primary tumors. In CNS metastases, plasma ctDNA is often undetectable due to the BBB, making CSF a uniquely valuable biospecimen.Moreover, changes in CSF-CTC burden and ctDNA variant allele frequencies may serve as real-time biomarkers of disease progression and treatment response, which arecritical capabilities in a disease that evolves rapidly and is difficult to monitor. As CSF liquid biopsy technologies mature, they promise to enhance not only diagnosis but also personalized treatment planning and clinical trial eligibility. Perhaps the most paradigm-shifting insights come from a retrospective analysis of 26 patients with breast cancer LM by Kumthekar and colleagues , revealing frequent discordance between HER2 status in the primary tumor and in CSF tumor cells (6). Using the CNSide™ platform, HER2 amplification was identified in 35% of patients whose primary tumors were initially classified as HER2-negative. This phenomenon termed "HER2 flip", thushighlights the dynamic and compartmentalized nature of tumor evolution, especially in the CNS.Serial testing in 14 patients showed that HER2 status can change over time, even during treatment. These fluctuations had direct therapeutic consequences, prompting the initiation or cessation of HER2targeted therapies, such as intrathecal trastuzumab. In patients with HER2-positive CSF tumor cells, treatment led to tumor burden reduction and clinical stabilization. Moreover, CNSide demonstrated significantly higher sensitivity than traditional cytology, detecting tumor cells in 100% of initial CSF samples compared to just 65% with cytology. Beyond detection, CNSide also quantifies tumor burden and enables HER2 biomarker analysis, offering a comprehensive diagnostic snapshot that cytology cannot provide.This study makes a strong case for routine CSF biomarker testing in LM,regardless of the primary tumor's profile. It also challenges the longstanding notion that metastatic disease mirrors its origin in actionable features. As a result, it opens new therapeutic doors for patients previously considered ineligible for targeted treatment. What unites these four studies is a shared vision of CSF as more than a diagnostic fluid-it is a window into tumor biology, evolution, and treatment response. In both lung and breast cancer, molecular profiling of CSF is revealing actionable mutations, resistance patterns, and tumor heterogeneity that are not captured by plasma or primary tissue testing. Whether guiding the use of MEK inhibitors in BRAFmutant NSCLC, determining eligibility for HER2-targeted therapies, or monitoring treatment response, CSF analysis is emerging as the cornerstone of CNS-directed precision oncology.Furthermore, the growing body of evidence supports a more aggressive, multimodal treatment paradigm for LM, integrating intrathecal chemotherapy, targeted agents, and serial molecular surveillance. This approach moves beyond the nihilism that often surrounds LM and offers a roadmap for extending survival and improving quality of life. While these findings are compelling, the path forward requires rigorous validation. Most data are retrospective or anecdotal, and sample sizes remain small. Prospective, multicenter trials with standardized protocols are necessary to confirm efficacy, assess long-term toxicity, and determine the optimal treatment sequences.Additionally, broader access to CSF-based diagnostics,including NGS, ctDNA, and CTC platforms,must be prioritized. Currently, such technologies are restricted to specialized centers and remain cost-prohibitive for many institutions. Investment in infrastructure, reimbursement, and education will be critical to democratizing these advances. Together, these studies chart a hopeful trajectory in the management of leptomeningeal metastases. They call for a redefinition of what is possible whether it is clinically, diagnostically, and biologically,in a disease long considered untreatable. By embracing CSF as a source of real-time, compartment-specific tumor intelligence and integrating targeted therapies that reach the CNS, oncology is finally turning the tide in LM. The future of LM care is no longer defined solely by palliation;it is being reshaped by precision.