Related Experiment Video
Updated: Oct 5, 2026

Detection and Monitoring of Tumor Associated Circulating DNA in Patient Biofluids
Published on: June 8, 2019
Early experience with cerebrospinal fluid cell-free DNA molecular profiling in a neuro-oncology practice
Cecile Riviere-Cazaux1, Rahul Kumar1, Julian S Rechberger1
1Department of Neurological Surgery, Mayo Clinic, Rochester, MN, USA.
Purpose:
Cerebrospinal fluid (CSF) cell-free DNA (cfDNA) is increasingly used as a liquid biopsy in neuro-oncology for diagnosing lesions of uncertain etiology, distinguishing progression from pseudoprogression, and ruling out leptomeningeal disease (LMD). We assessed our institution's early experience with a new CSF cfDNA molecular profiling assay for detecting variant alleles (VA) and copy-number alterations (CNA) across these indications in our neuro-oncology practice.
Methods:
Beginning in August 2021, providers could request research CSF cfDNA sequencing from clinically indicated lumbar punctures. Following extraction, targeted next-generation sequencing (NGS) and/or low-pass whole-genome sequencing (LPWGS) were performed. VAs and CNAs were summarized by indication and integrated with clinical features, cfDNA yield, matched tissue data, and overall survival.
Results:
CSF samples from seventy-six consecutive patients were analyzed and retrospectively classified into three categories: lesions of uncertain etiology (46%), question of CNS progression (21%), or ruling out LMD (26%). Median cfDNA yield was 0.27 ng/mL (IQR = 0.05-0.61 ng/mL). VA and/or CNA detection, hereafter designated "tumor calls," were identified in 20 patients (26%), with 29 patients having negative tumor calls (38%) and 27 patients (36%) having insufficient cfDNA yield/quality. A copy number burden threshold of 7.42 accurately identified tumor-specific CNAs (AUC 0.998) and was associated with poorer survival (HR 3.4, p = 0.006). Overall survival for patients with positive tumor calls was significantly lower than for patients with negative tumor calls or low yield (p = 0.003).
Conclusion:
Despite low cfDNA abundance, CSF NGS and LPWGS provide informative data across neuro-oncologic indications. Further work is underway to optimize the assays for low input CSF cfDNA.

