Oncology
META ANALYSIS
● Meta-analysis
CSF ctDNA Status Predicts Survival in Leptomeningeal Metastases: Meta-Analysis
Journal of neuro-oncology
Published March 27, 2026
Sankar Baradwaj Simha, Johnson Drew, Boasiako Paul Antwi, Vargas Luis O, Syed Shoaib, Padova Audrey,…
PubMed ↗
DOI ↗
This meta-analysis aimed to assess the prognostic value of cerebrospinal fluid (CSF) circulating tumor DNA (ctDNA) in patients with leptomeningeal metastases (LMD) from solid tumors. The study included 14 studies with a total of 963 patients evaluating baseline CSF ctDNA and two studies with 26 patients assessing longitudinal ctDNA kinetics. The primary endpoint was overall survival (OS), with progression-free survival (PFS) as a secondary endpoint. The analysis found that patients with adverse baseline CSF ctDNA status, such as EGFR mutation positivity, had significantly worse OS (pooled HR 2.40, 95% CI 1.73-3.33) and PFS (pooled HR 2.45, 95% CI 1.36-4.44). Longitudinal studies indicated that increasing CSF ctDNA variant allele fraction was associated with worse OS (pooled HR 4.11, 95% CI 1.25-13.48). The heterogeneity was moderate for OS (I²=36.6%) and low for PFS (I²=15.5%) and longitudinal OS (I²=8.5%). No specific safety or adverse events were reported in relation to CSF ctDNA testing. These findings suggest that baseline and serial CSF ctDNA measurements could serve as valuable biomarkers for survival outcomes in LMD, potentially guiding therapeutic decisions and complementing clinical and radiographic assessments. The study highlights the need for prospective, standardized, multi-timepoint CSF ctDNA studies to further validate these findings and integrate CSF ctDNA monitoring into clinical practice.
AI Accuracy Review: 9/10
· Auto-published
This research focuses on leptomeningeal disease, a serious condition where cancer spreads to the membranes around the brain and spinal cord. Researchers looked at how a test that measures cancer DNA in cerebrospinal fluid (the fluid surrounding the brain and spine) can help predict how long patients might live and how well their treatment is working. They found that patients with certain DNA changes in their spinal fluid had shorter survival times. Additionally, changes in the amount of cancer DNA over time also suggested worse outcomes. This means that monitoring this DNA could help doctors better understand a patient's condition and adjust treatments accordingly. However, more studies are needed to confirm these findings and to see how best to use this test in everyday care. It's important for patients to know that while this test shows promise, it is still being researched and is not yet standard practice.
What this means for you: A new test measuring cancer DNA in spinal fluid may help doctors track disease and treatment response.
View Original Abstract ↓
PURPOSE: Leptomeningeal disease (LMD) is a devastating complication of advanced solid tumors with limited prognostic and response-assessment tools. Because LMD molecular evolution is frequently compartmentalized behind CNS barriers, cerebrospinal fluid (CSF) circulating tumor DNA (ctDNA) may provide CNS-specific molecular readouts of disease activity. We evaluated whether baseline CSF ctDNA profiles and longitudinal ctDNA kinetics associate with survival in LMD.
METHODS: We performed a systematic review and meta-analysis (PROSPERO CRD420251068543) of Embase, PubMed, Cochrane CENTRAL, WHO ICTRP, ClinicalTrials.gov, Europe PMC, and Web of Science from inception through June 2025. Eligible studies included adults with solid-tumor LMD undergoing anti-cancer therapy with CSF ctDNA assessed at baseline and/or serially and reported associations with overall survival (OS) and/or progression-free survival (PFS). Random-effects models pooled hazard ratios (HRs). Longitudinal studies were analyzed separately due to limited availability.
RESULTS: Fourteen studies (n = 963) evaluated baseline CSF ctDNA and two studies (n = 26) evaluated longitudinal kinetics. Adverse (non-reference) baseline ctDNA status (e.g., EGFR mutation positive vs. negative) was associated with inferior OS (pooled HR 2.40, 95% CI 1.73-3.33; I²=36.6%) and PFS (pooled HR 2.45, 95% CI 1.36-4.44; I²=15.5%). Longitudinally, increasing CSF ctDNA variant allele fraction was associated with worse OS (pooled HR 4.11, 95% CI 1.25-13.48; I²=8.5%). Across longitudinal reports, serial CSF ctDNA measurements tracked progression and response.
CONCLUSION: Baseline and serial CSF ctDNA measurements are associated with survival outcomes in LMD and may complement clinical and radiographic assessment. These findings support prospective, standardized, multi-timepoint CSF ctDNA studies in LMD and warrant CSF-access-enabled monitoring to inform therapeutic adaptation.