.png){kind=icon}
.png){kind=icon}
Return
Nanopore-based liquid biopsy analysis enables preoperative tumor classification in pediatric brain tumors
.png){kind=icon}
OA
AI
N
D
M
P
K
A
J
S
B
A
A
L
J
B
C
L
G
C
D
S
J
DOI：10.1186/s40478-026-02395-8.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Liquid biopsy utilizing circulating tumor DNA (ctDNA) derived from cerebrospinal fluid (CSF) has rapidly evolved towards clinical application in pediatric brain tumors. However, limited information on technical applicability and clinical translation impedes transformation into clinical routine. Herein, we assessed the feasibility and clinical utility of methylation-based CNS-tumor classifiers on pre-/intraoperatively acquired CSF samples integrated with real-world diagnostic measures and respective clinical scenarios. We analyzed 44 samples, with CSF acquired via extra-ventricular drains (EVD), lumbar puncture, Ommaya reservoirs, or intraoperatively. Subsequent cell-free (cf)DNA analysis and tumor classification were performed using Nanopore sequencing utilizing four tumor tissue classifiers (CrossNN, Sturgeon, MethyLYZR, RapidCNS2). Obtained results were integrated with magnetic resonance imaging after blinded review and potential impact on clinical management was assessed by an interdisciplinary panel. CtDNA levels ranged from non-measurable to 5.7 ng/µl and ctDNA fractions from 0.006 to 0.97. Importantly, samples derived from EVD, frequently implanted before major surgery, performed very well, highlighting the reliability of cfDNA based analysis. Regarding preoperatively acquired samples, correct classification rates ranged from 69 to 78%. While most embryonal tumor cases (medulloblastoma, atypical teratoid rhabdoid tumor) were correctly identified, the identification of other diagnostic groups (gliomas, germ cell tumors, ependymoma) remains more challenging. We further show that correct classification correlates with cfDNA content, ctDNA concentration, and peak fragment length. Timeline analysis of the developed workflow indicated generation of results in less than 10 h from time of CSF acquisition. Subsequent, integration with magnetic resonance imaging after blinded review resulted in increased correct diagnostic prediction of 91% of cases. Finally, interdisciplinary review of cases suggested that, for one-third of patients, preoperative knowledge of diagnosis could impact patient management. In summary, preoperative CSF methylation-based diagnosis demonstrates high accuracy and a rapid turnaround in the majority of pediatric brain tumor cases, specifically identifying the majority of medulloblastomas. Adopting preoperative liquid biopsy into standard diagnostic workflows thus harbors potential to revolutionize clinical decision-making and multidisciplinary care strategies.
Keywords:
Pediatric CNS tumors
Liquid biopsies
Methylation-based tumor classification
Cerebrospinal fluid
Nanopore-sequencing
Cell-free DNA
Journal
IF:
5.7
Papers:
2.3K
Citations:
1.0W
Organization
Cited Papers