Diagnosis of pediatric central nervous system tumors using methylation profiling of cfDNA from cerebrospinal fluid

Diagnosis of pediatric central nervous system tumors using methylation profiling of cfDNA from cerebrospinal fluid | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Diagnosis of pediatric central nervous system tumors using methylation profiling of cfDNA from cerebrospinal fluid Lotte Cornelli, Ruben Van Paemel, Maísa Santos, Sofie Roelandt, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4218805/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Pediatric central nervous system tumors remain challenging to diagnose. Imaging approaches do not provide sufficient detail to discriminate between different tumor types, while the histopathological examination of tumor tissue shows high interobserver variability. Recent studies have demonstrated the accurate classification of central nervous system tumors based on the DNA-methylation profile on a tumor biopsy. However, a brain biopsy holds significant risk of bleeding and damaging the surrounding tissues. Liquid biopsy approaches analyzing circulating tumor DNA show high potential as an alternative and less invasive tool to study the DNA-methylation pattern of tumors. In this study, we explore the potential of classifying pediatric brain tumors based on methylation profiling of the cell-free DNA in cerebrospinal fluid (CSF). For this proof-of-concept study, we collected 20 cerebrospinal fluid samples of pediatric brain cancer patients via a ventricular drain placed for reasons of increased intracranial pressure. Analyses on the circulating cell-free DNA (cfDNA) showed high variability of cfDNA quantities across patients ranging from levels below the limit of quantification to 40 ng cfDNA per milliliter of CSF. Classification based on methylation profiling of cfDNA from CSF was correct for 8 out of 20 samples in our cohort. Accurate results were mostly observed in samples of high quality, more specifically those with limited high-molecular weight DNA contamination. Interestingly, we show that centrifugation of the CSF prior to processing increases the fraction of fragmented cfDNA to high-molecular weight DNA. In addition, classification was mostly correct for samples with high tumoral cfDNA fraction as estimated by computational deconvolution (> 40%). In summary, analysis of cfDNA in the CSF shows potential as a tool for diagnosing pediatric nervous system tumors especially in patients with high levels of tumoral cfDNA in the CSF, however further optimization of the collection procedure, experimental workflow, and bioinformatic approach is required to also allow classification for patients with low tumoral fractions in the CSF. Pediatric oncology DNA methylation liquid biopsy central nervous system tumor precision medicine cerebrospinal fluid Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Intracranial central nervous system (CNS) tumors are one of the leading causes of cancer related death in children after leukemia 1 , 2 . Diagnosis of these brain tumors is complex as they consist of a heterogeneous group of tumors, from slow-growing, low grade lesions to high grade cancers 3 . An accurate and detailed diagnosis is vital for treatment decisions and determines patient outcome 4 , 5 . The current diagnostic procedure of (pediatric) brain tumors requires a tumor tissue biopsy for histopathological investigation 6 – 8 . In this evaluation, inter-observatory variability occasionally results in misdiagnosis 6 – 8 . Diagnostic procedures are evolving from purely histology based methods 9 , to a combined approach where assays that interrogate molecular markers are becoming increasingly important 4 . Indeed, the latest 2021 World health organization (WHO) classification entails tumor types and subtypes that can only be distinguished by combinations of new molecular profiling methods 5 , 10 . Genomic and epigenomic analyses have improved the diagnostic process for many cancer entities. More specifically the tumor DNA methylation profile is shown to be tissue specific and therefore is a powerful tool for tumor classification 11 – 14 , as very convincingly shown for brain tumor (sub-)classification 11 . For brain cancer patients with delicate tumor location, performing a biopsy or resection can hold disproportional risks 15 – 18 . In the last decade, the potential of the use of liquid biopsies has become evident, emerging as a novel and valuable approach to molecularly investigate the tumor in a minimally invasive manner 19 , 20 . Tumoral biomolecules, including circulating cell-free DNA (cfDNA), RNA and proteins 15 , 18 , 20 are released from different locations within the tumor, and therefore contain molecular information while avoiding sampling bias often seen in tissue biopsies 15 . These molecules are found in biofluids surrounding the tumor, including blood, urine, cerebrospinal fluid, and others 20 . Several studies have shown that the amount of circulating tumor DNA (ctDNA) in the blood is limited in patients with intracranial tumors due to retention by the blood-brain-barrier 18 , 21 – 23 . In these studies, cerebrospinal fluid is suggested as a superior source for ctDNA. The extraction of CSF is often an uncomfortable and invasive procedure, requiring a painful lumbar puncture to acquire fluid from the space surrounding the spinal cord. However, for pediatric patients, CSF is collected during standard diagnostic procedures and collection for cfDNA would require no additional procedures. These patients often present with symptoms of increased intracranial pressure due to CSF flow obstruction requiring emergency placement of an external ventricular drain to remove the excess fluids 24 – 26 . For tumors with risk of cerebrospinal fluid metastasis, an additional lumbar puncture is often performed after removal of the tumor, to examine the CSF for the presence of circulating tumor cells 24 , 27 – 29 . Published cfDNA studies on CSF have mostly focused on tumor follow-up rather than classification. For example, cfDNA is used for mutation detection where tumors are detected based on the previously defined patient/tumor specific mutations 30 – 35 . Since mutations frequently fail to differentiate between tumor subtypes, Li and colleagues instead used whole genome bisulfite sequencing (WGBS) based DNA methylation profiling and hydroxymethylation profiling through anti-CMS immunoprecipitation sequencing to subtype and monitor medulloblastoma tumors 36 . A recent study utilized nanopore technology to investigate tumor classification. Here, Afflebach et al. conducted nanopore sequencing on cerebrospinal fluid cell-free DNA (CSF-cfDNA) to classify tumors based on methylation patterns. However, they were only successful in detecting ctDNA based on methylation profiling in 15 out of 178 samples. Out of these 15 samples, accurate classification was achieved for 13 of them 37 . Although promising, the authors describe a minimal input 1 ml CSF and 5 ng cfDNA, which is challenging to collect in some CSF samples. Additionally, samples in this study were lost after failing the technical pass of 100 000 reads, or were unsuited for methylation profiling when they did not cover a minimal of 1000 CpG’s 37 . Considering the significant expense associated with the use of WGBS and the fact that nanopore sequencing of cfDNA has yielded successful results only in a subset of samples, we have generated proof-of-concept for the use of an alternative technology that allows accurate and minimally-invasive classification of pediatric brain tumors. More specifically, in this study, we use cell-free reduced representation bisulfite sequencing (cfRRBS) that allows methylation profiling of low amounts of fragmented DNA, such as cfDNA 38 . The method involves a step to enrich the more relevant methylated regions, i.e. the CpG rich regions, resulting in a reduction of sequencing cost per sample compared to WGBS. Before, we (Van Paemel et al.) have demonstrated the potential of cfRRBS for accurate cfDNA methylation-based diagnosis of pediatric solid tumors 39 , by achieving a correct classification rate of 94% in high-quality samples, mostly cfDNA from blood plasma samples. In this study, we explore the diagnostic potential of cfRRBS followed by DNA methylation signal deconvolution for tumor fraction estimation on cfDNA isolated from cerebrospinal fluids in pediatric brain tumor patients. Material and methods Patients and samples This study was approved by the ethical committee and informed consent was obtained from all patients and/or their representatives. Pediatric patients presenting with a central nervous system tumor at Ghent University Hospital from February 2020 to July 2023 (n = 19; Age range 3 months to 16 years old) that required ventricular drainage were included. Samples were collected from patients that were pathologically diagnosed with medulloblastoma (n = 6), pilocytic astrocytoma (n = 6), ependymoma (n = 3; two samples were collected from the same patient), choroid plexus papilloma (n = 2), diffuse midline glioma (n = 1), adamantinomatous craniopharyngioma (n = 1), and atypical teratoid/rhabdoid tumor (n = 1). Complete patient and sample data is available in the supplemental Tables 1 and 2. CSF collection Determined by the amount that could safely be sampled, 1.5 to 20 ml of cerebrospinal fluid was collected. For the first 12 patients, samples were separated into two aliquots. One aliquot was centrifuged for 10 minutes at 1900 g and the supernatant transferred to a clean 15 ml falcon tube. The other aliquot was processed without additional interventions. For the following patients, the complete CSF sample was centrifuged. Both the unprocessed CSF and the supernatant after centrifugation were stored at -80°C CSF of 13 patients was centrifuged immediately and stored within four hours after collection. For the remaining patients (n = 7/20), samples were temporarily frozen at the operation room at -20°C before any processing. These samples were subsequently stored at -80°C and centrifuged on the day they were further processed. Cell-free DNA extraction and quality control CSF samples were thawed to room temperature and cell-free DNA was extracted using a Maxwell RSC LV ccfDNA kit (Promega). Depending on sample availability, between 1 and 7.5 ml of CSF was used as input for cfDNA extraction. in order to reach the minimal input for this protocol, one sample with a volume below 2 ml was supplemented with PBS (1X, Gibco) to a total volume of 2 ml, prior to addition of binding buffer. The extraction was performed according to the manufacturers guidelines and the resulting cfDNA was eluted in 75 µl of elution buffer supplied in the kit (Promega). cfDNA concentration was measured using Fluoroskan™ Microplate Fluorometer (Thermo Fisher Scientific) according to the manufacturer’s instructions. Size distribution profiles were obtained using Agilent Tapestation with the Cell-Free DNA ScreenTape kit. CfDNA was defined as DNA fragments with lengths ranging between 70 and 700 bp and high molecular weight DNA (HMW-DNA) as DNA fragments with lengths exceeding 700 bp. Isolated DNA was stored at -20°C until further processing. DNA isolation from surgical tumor biopsy samples An aliquot of isolated genomic DNA from a tumor tissue biopsy was obtained for 11 of the 19 patients. DNA was extracted starting from 3 to 15 paraffin embedded (FFPE) tissue slides according to the manufacturer’s instructions using the Qiamp DNA FFPE Tissue kit (QIAGEN). DNA was stored at 4°C until processing. cfRRBS library preparation Isolated DNA was processed using cell-free reduced representation bisulfite sequencing as previously described 40 . For the initial 12 patients, DNA from whole CSF as well as the centrifuged aliquot was used as input for cfRRBS. When available (n = 29/32), 10 ng of DNA was used as input as described in the protocol. Samples with a DNA concentration below 0.2 ng/µL were concentrated via vacuum centrifugation (SpeedVac, Thermo Fischer Scientific) at 45°C. According to the protocol, 0.01 ng lambda spike-in was added to the samples. After library amplification, DNA was purified using SPRI bead size selection (AMPure XT beads – NEB), with 2.5x proportion of bead to sample volume. The libraries were quantified and checked for the presence of adapter dimers via the Kapa library quantification kit for Illumina platforms (Kapa Biosystems). The length profile was visualized via Fragment Analyser (Advanced Analytical Technologies). Samples were pooled equimolarly to a total concentration of 4 nM. Final concentration of the pooled samples was verified using the Kapa library quantification kit for Illumina platforms (Kapa Biosystems). Sequencing quality control and mapping Samples were sequenced on a NovaSeq 6000 instrument using a NovaSeq SP kit (paired-end, 2 × 50 cycles), supplemented with 3% phiX and a loading concentration between 750 and 800 pM. Samples from different donors and tubes were mixed to avoid sequencing batch effects. BCL files were demultiplexed and quality checked as previously described by Van Paemel et al. 39 . Updated modules were used for demultiplexing (bcl2fastq v2.20), adapter removal (Trim Galore v0.6.6, and CutAdapt), mapping (SAMtools v.1.14, and Bismark v.0.23.1), read counting (Picard tools v.2.21.6). Visualization of the results was done with R version 4.3.2 and ggplot2 v 3.4.4. We obtained on average 21,7M reads, and a minimum of 7M reads per sample. Mapping efficiency was on average 52%, as to be expected for cfRRBS data; bisulfite conversion was at least 95.9% for all samples, and exceeded 98% for 25/32 samples. Full QC report of the samples is shared in supplementary table 3 . Development of a reference set for computational deconvolution The methylation profiles of 2801 brain tumors generated on Illumina array 450K platform, encompassing 81 different brain tumor entities, was obtained from Capper et al. 11 . Seven of the tumor entities in this reference data were excluded, as they are not recognized by the WHO CNS5 (2021), namely, infantile hemispheric glioma, high grade neuroepithelial tumor with BCOR alteration, high grade neuroepithelial tumor with MN1 alteration, anaplastic pilocytic astrocytoma, Ewing sarcoma family tumor with CIC alteration, central neurocytoma, and plasmacytoma. All other 2629 samples covering 74 entities were used to build the brain tumor reference dataset. The reference set was adjusted to allow deconvolution of cfRRBS data by only considering the CpGs in regions that overlap in the cfRRBS and array data as described by Van Paemel et al. 39 . In house data of healthy plasma cfDNA were also included in the reference dataset, as well as published methylation data of prepuberal white blood cells (WBC, n = 52) 41 as some samples with red discoloration are assumed to contain low volumes of contaminating blood (due to placement of the ventricular drain). Computational deconvolution of cellular fractions Tumoral fractions were estimated using Methatlas 42 , a non-linear least square based method. Reference and test samples were grouped in 14.103 clusters, and a median value for the methylation status was calculated for each cluster. The beta values of all clusters were used for deconvolution. The tumor classification of the test samples was defined based on the highest estimated fraction in the sample excluding non-tumoral (white blood cell, plasma, and control brain tissue) fractions. Full deconvolution output from both CSF-cfDNA and FFPE samples are added to supplementary table 4 . Copy number profiles Copy number aberrations were inferred from the cfRRBS data of both CSF liquids and tissue biopsies. We used WisecondorX ( https://github.com/CenterForMedicalGeneticsGhent/WisecondorX ) to detect copy number aberrations after mapping to the bisulfite converted genome. The binsize was set at 400 kb. 43 Samples were normalized with an in house dataset of cfRRBS data from healthy volunteers. Results Reference dataset for computational deconvolution of pediatric brain tumor fractions For our reference dataset, we modified a published Illumina array dataset to align with the genomic regions covered in cfRRBS data (details in M&M). In the reference set, we also included cfRRBS data of blood plasma cfDNA from non-cancerous volunteers, as well as white blood cells because some CSF samples present with contaminating blood cells. We performed UMAP dimensionality reduction on the cfRRBS reference dataset to visualize the grouping/clustering of the tumor entities based on their methylation profile (Fig. 1 A). While most tumor entities can be clearly distinguished in this plot, similar as the published visualization by Capper et al. 11 we observed that the low grade glioma clusters overlap with other tumor entities (Fig. 1 B). This optimized dataset is used as reference for computational deconvolution of pediatric brain tumor fractions in the next paragraphs. Correct tumor classification using cfRRBS on pediatric brain tumor tissue DNA To validate our deconvolution based classification pipeline, we first applied it on cfRRBS profiles generated on genomic DNA isolated from pediatric brain tumor tissue of 11 patients. For 8 out of 11 tumors, the highest estimated tumor fraction corresponds with the histopathological diagnosis. The highest tumor fraction estimated using deconvolution for the medulloblastoma tumor (MB; n = 4), ependymoma tumor (EPN; n = 2) and choroid plexus papilloma tumor samples (PLEX; n = 1) corresponds with the histopathological diagnosis and assigned subclass (Fig. 2 ). However, for pilocytic astrocytoma (LGG-PA; n = 4) only one out of 4 tumors are classified correctly. Given interobserver variability is reported for histopathological diagnoses, a pathologist re-examined these 3 cases but excluded any misdiagnosis. The incorrect classification can be explained by the fact that the DNA methylation profile of LGG-PA cases is not distinct enough as observed in Fig. 1 B. Circulating cfDNA in cerebrospinal fluid We collected CSF samples of 19 pediatric patients presenting with CNS tumors. For one patient, two CSF samples were collected at the moment of two consecutive relapses. In this cohort, we considered these two samples as independent due to our primary emphasis on the sample quality features. All samples were collected in plastic containers without preservatives. When possible, samples were processed immediately (n = 13), if not they were frozen at -20°C in the operation room (n = 7). Previous studies have shown that high molecular weight DNA (HMW-DNA) originating from white blood cells can interfere with cfRRBS and deconvolution by diluting the tumoral signal 39 . For that reason, CSF was centrifuged with the aim of removing cell debris that could contribute to high molecular weight DNA contamination. To check the effect of the centrifugation, for 12 patients we also stored an aliquot of uncentrifuged material. Next, cfDNA was isolated, concentration quantified and fragment length analyzed. We observed a high degree of variability in the total amount of cfDNA, and fragment profile between patients. Samples from three patients showed a circulating cfDNA (70–700 bp length) concentration that was below the limit of detection by Tapestation visualization. Compared to centrifuged CSF, whole CSF samples showed a significantly lower cfDNA concentration on total DNA fraction (Fig. 3 , p-value 0.04814) pointing at more high molecular weight DNA contamination. The total yield of cell-free DNA is not significantly different between whole and centrifuged CSF (p-value = 0.4962; figure in supplemental information) indicating that centrifugation primarily decreases the presence of HMW-DNA by removing cells and thus preventing cell lysis in the sample, and has minimal effect on the fragmented cfDNA. Based on these results, we decided to centrifuge the CSF before processing for the other 8 samples (sample 13 to sample 20). Figure 3 shows the percentage of circulating cfDNA over total DNA that was isolated from CSF after centrifugation of each patient. Deconvolution of tumor fractions in cfDNA from CSF Following quality control, all samples were processed with cfRRBS library preparation, sequencing, and deconvolution to estimate the tumoral and healthy cfDNA fractions. The estimated tumor type (i.e. entity with the highest predicted tumor fraction after using deconvolution) corresponded with the histopathological diagnosis for 8 out of 20 cases: medulloblastoma (4/5), ependymoma (2/3), choroid plexus papilloma (1/2), atypical teratoid rhabdoid tumor (1/1). We noted higher fractions of fragmented cfDNA on total DNA, and a higher estimated tumor fraction according to deconvolution in samples that were correctly classified. Although the relatively small patient cohort does withhold us from defining validated cut-off values, we see that most samples with a cfDNA/total DNA fraction below 40% and/or an estimated tumor burden below 30% are classified incorrectly (10/14). Samples with cfDNA/total DNA fraction of at least 40% and tumor burden of at least 30% are all classified correctly (6/6). Copy number profiling from cfRRBS data From cfRRBS data, also copy number aberration profiles can be extracted. For 11 donors, we were able to compare copy number profiles from tumor and cfDNA. Although data is more noisy compared to whole genome sequencing methods, we observe aberration in 3 of these patients. For samples with lower estimated tumoral fractions, such as case 1 illustrated in Fig. 3 , we could not identify any aberrations in the CSF-cfDNA. Case 5 shows aberrations that correlate well to the ones in the tumor tissue. Interestingly, for 2 of these patients (case 2 and case 5) we observed additional copy number alterations in the liquid samples compared to the tumor suggestive for intratumoral heterogeneity. Discussion In this proof-of-concept study we present an analytical and computational pipeline to classify pediatric CNS tumors using a DNA methylation assay that was developed for fragmented cfDNA isolated from liquid biopsies, and that works on both DNA from fresh frozen and paraffin embedded tissues. Tumor type and tumor fraction are estimated using computational deconvolution based on a reference dataset containing published methylation data of brain tumor tissues complemented with healthy cfDNA profiles. We found that the tumor classification correlates well with histopathological diagnosis for good quality cfDNA samples. We identified several pitfalls of our approach related to CSF collection and CSF characteristics, as well as opportunities for improvement which require further validation on larger patient cohorts before clinical implementation. In summary, 8 out of 20 samples from pediatric CNS cancer patients are classified correctly using cfDNA from CSF. All samples with high cfDNA/total DNA fraction were classified correctly (6/20). In most misclassified samples, we observe increased fractions of HMW-DNA (length over 700 bp) among the isolated DNA in CSF, with 14 samples showing a cell-free DNA fraction lower than 0.5. Additionally, the more than half of the samples exhibiting a cfDNA yield below 5 ng. The scarcity of cfDNA in CSF is not emphasized in similar studies, yet it is notable that published articles working on cfDNA from CSF for tumor classification and follow-up almost exclusively focus on higher grade tumors 36 , 37 , 44 – 50 . Although ctDNA levels are not defined by tumor grade alone, it is an important variable in the release of fragmented DNA 51 . Next to the more urgent clinical need for those more aggressive tumors, lack of studies on lower grade tumors most probably indicates the challenges in obtaining sufficient ctDNA material. The Afflerbach study 37 , in contrast, does include low and high grade tumors, and indeed underscores the low abundance of circulating tumor DNA in the CSF. Due to the low sample input and quality, only 12% of the total cohort passes the minimal technical requirements to allow for methylation profiling and tumor classification. Overall, our cfRRBS approach could successfully generated methylation profiles on all included sample and was able to correctly estimate tumor diagnosis in 30% of the samples. Although obtaining sufficient circulating tumoral DNA still appears challenging for lower grade tumors. In the paragraphs below, we discuss the different factors that impact the performance of our classification approach, including high molecular weight contamination, tumoral cfDNA fractions and reference dataset. In previous cfRRBS studies we have already highlighted the importance of assessing the cfDNA/HMW-DNA fraction in each sample 39 . HMW-DNA, that is also processed in the cfRRBS library preparation will dilute the signal of the cfDNA and might lead to misclassification of samples in case of high fractions of HMW-DNA. The HMW-DNA is likely derived from cells that are damaged during ventricular drain placement or white blood cells in blood-contaminated samples. In this study, we show that centrifugation of a fresh CSF sample before DNA isolation improves sample quality in many cases. Freezing the CSF prior to centrifugation results in lysis of the cells and thus release of HMW-DNA into the fluid. Thus, centrifugation within four hours after collection and before freezing is ideal, but more challenging to implement into clinical practice. In addition, avoiding blood cell contamination will result in better quality samples. Secondly, although yield of circulating DNA relates to variables such as tumor size and aggression 51 , sampling in close proximity of the tumor through a ventricular drain might collect more ctDNA compared to sampling via lumbar puncture. These observations underscore the need for more dedicated studies investigating the pre-analytical variables that can improve the sample quality, as well as more standardized protocols for CSF collection ensuring the standard collection of high-quality samples allowing robust tumor classification. The significance of this is highlighted by the wide variation in published collection protocols, encompassing collection through lumbar puncture, ventricular drainage, or mixed cohorts, with volumes ranging from 200µl to 10 ml. While centrifugation is commonly included in most protocols, there are studies that omit this step 30 , 31 , 33 – 35 , 37 , 45 . The absence of a well-defined profile characterizing the non-tumoral background in cerebrospinal fluid (CSF) poses a limitation in accurately estimating the fraction of a specific tumor type. This challenge becomes particularly pronounced in cases where CSF samples contain lower tumor fractions and higher non-tumoral cfDNA. We observed a better classification accuracy in cfDNA samples with higher estimated tumor fractions. In blood samples with lower tumor fraction, the non-tumoral cfDNA mostly originate from the white blood cells 42 . For CSF however, the origin of the non-tumoral cfDNA is not clearly defined and might originate from WBC, but also from damaged brain tissue due to increased pressure in patients presenting with hydrocephalus. Although the Capper reference data includes various healthy brain tissues, pediatric profiles often differ from adult. To allow proper deconvolution of all the contributing cell types in the CSF-cfDNA samples, a reference dataset encompassing all those cell types is necessary. However, pediatric CSF collection is only performed to patients with a (suspected) brain related pathology, thus obtaining a reference sample and DNA methylation profile from pediatric hydrocephalus patients without brain pathologies is almost impossible. One option would be the inclusion of CSF from patients with hydrocephalus caused by a traumatic brain injury. The performance of deconvolution algorithms heavily relies on the choice of reference data. The DNA-methylation based assay for CNS brain tumor diagnosis is utilized in an increasing number of pathological departments and employs the Infinium HumanMethylation450 BeadChip array 11 . This array encompasses 450,000 methylation sites and shows good performance to distinguish between different tumor entities. However, a drawback is that it requires a minimum input amount of 500 ng of DNA, a quantity significantly surpassing the average cell-free DNA (cfDNA) yield from liquid biopsies. To address this challenge, we successfully employed cell-free reduced representation bisulfite sequencing (cfRRBS), an approach tailored for low quantities of highly fragmented DNA, requiring only 10 ng or even less input DNA to generate high quality DNA methylation profiles 38 . We formatted the published 450K array methylation data 11 of CNS tumors to align with the cfRRBS workflow and used it as a reference dataset for deconvolution. A limitation of this approach is that we only use the sites that are covered by both the 450K array and the cfRRBS assay, which is only 13.7% of methylation sites that are covered by the cfRRBS assay. By restricting the number of sites, we noticed that discriminating low grade glioma tumors became more challenging as visualized in the UMAP plot (Fig. 1 ) compared to the published UMAP (ref). Building a (cf)RRBS based reference dataset would enable the utilization of all cfRRBS regions in the deconvolution model and thus increase the available information to discriminate different tumor entities; however, this will come with additional efforts and costs. This problem highlights the trade-off between maximizing data inclusivity and managing data availability or associated financial constraints. In addition to the restricted number of sites, the published version of the classifier shows challenges in discriminating low-grade glioma tumors, resulting in less accurate predictions for this particular subtype 52 . Newer versions of the classifier can improve classification for several challenging tumor types including low grade gliomas, however the reference data of newer versions is not publicly available 52 . Interestingly, the data produced via cfRRBS can also be used for copy number variation (CNV) profiling. Although this data is more noisy compared to dedicated copy number profiling assays such as shallow whole genome sequencing (shWGS), extraction of multiple data layers from cfRRBS reads without requiring new input material is an important asset. Compared to most cfDNA shWGS approaches for CNV analysis, cfRRBS lacks a size separation step and thus also HMW-DNA will be processed 38 resulting in a dilution of the tumoral signal. Indeed, for samples with tumoral fraction below 30% we couldn’t observe any tumor associated aberrations. For the patients with matched tumor and CSF material and higher estimated tumor fractions, we observed some CSF specific aberrations suggesting intratumoral heterogeneity, similar to the results described by Chicard and colleagues 44 . However, it is notable that the lower quality of the CNV profile data limits the number of patients for which the CNV profile can accurately be analyzed. Conclusion Although the presented cfRRBS approach on CSF has several limitations and points that need further optimization, we believe that this approach can become a valuable alternative for cancer patients with tumors located in regions that are too delicate for a surgical biopsy. Validation on larger cohorts is required, still we observed accurate classification for patients with cfDNA fractions higher than 50%. We expect that methylation profiling of cfDNA isolated from liquid biopsies could take an important and complementary position next to standard diagnostic approaches, for example by giving an early diagnosis that can inform oncologists and surgeons in their choice for a treatment strategy. Abbreviations CSF cerebrospinal fluids cfDNA circulating cell-free DNA CNS central nervous system WHO World health organisation cfRNA cell-free RNA ctDNA circulating tumor DNA WGBS whole genome bisulfite sequencing CSF-cfDNA cerebrospinal fluid cell-free DNA cfRRBS cell-free reduced representation bisulfite sequencing HMW-DNA high molecular weight DNA WBC white blood cells LGG low grade glioma MB medulloblastoma EPN ependymoma PLEX choroid plexus papilloma LGG-PA pilocytic astrocytoma ATRT atypical teratoid rhabdoid tumor CPH adamantinomatous craniopharyngioma DMG diffuse midline glioma FFPE formalin fixed paraffin embedded CNV copy number variation shWGS shallow whole genome sequencing ETF estimated tumor fraction Declarations Ethics approval and consent to participate This study is approved by the commission for medical ethics from UZ Ghent (reference ID: EC2019/1514). Written informed consent was obtained from all patients and/or their representatives. Availability of data and materials FastQ files of all samples will be published in an EGA dataset respecting the original informed consent agreements. Competing interests The authors declare that they have no competing interests. References Siegel, D. A. et al. Pediatric cancer mortality and survival in the United States, 2001-2016. Cancer 126 , 4379–4389 (2020). Vassal, G. et al. The SIOPE strategic plan: A European cancer plan for children and adolescents. J. Cancer Policy 8 , 17–32 (2016). Reddy, A. T. & Packer, R. J. Pediatric central nervous system tumors. Curr. Opin. Oncol. 10 , 186–193 (1998). Diamandis, P. & Aldape, K. World Health Organization 2016 Classification of Central Nervous System Tumors. Neurol. Clin. 36 , 439–447 (2018). Louis, D. N. et al. The 2021 WHO Classification of Tumors of the Central Nervous System: a summary. Neuro-Oncol. 23 , 1231–1251 (2021). van den Bent, M. J. Interobserver variation of the histopathological diagnosis in clinical trials on glioma: a clinician’s perspective. Acta Neuropathol. (Berl.) 120 , 297–304 (2010). Sturm, D. et al. New Brain Tumor Entities Emerge from Molecular Classification of CNS-PNETs. Cell 164 , 1060–1072 (2016). Ellison, D. W. et al. Histopathological grading of pediatric ependymoma: reproducibility and clinical relevance in European trial cohorts. J. Negat. Results Biomed. 10 , 7 (2011). Louis, D. N. et al. The 2007 WHO Classification of Tumours of the Central Nervous System. Acta Neuropathol. (Berl.) 114 , 97–109 (2007). Louis, D. N. et al. cIMPACT-NOW update 6: new entity and diagnostic principle recommendations of the cIMPACT-Utrecht meeting on future CNS tumor classification and grading. Brain Pathol. Zurich Switz. 30 , 844–856 (2020). Capper, D. et al. DNA methylation-based classification of central nervous system tumours. Nature 555 , 469–474 (2018). Priesterbach-Ackley, L. P. et al. Brain tumour diagnostics using a DNA methylation-based classifier as a diagnostic support tool. Neuropathol. Appl. Neurobiol. 46 , 478–492 (2020). Jaunmuktane, Z. et al. Methylation array profiling of adult brain tumours: diagnostic outcomes in a large, single centre. Acta Neuropathol. Commun. 7 , 24 (2019). Capper, D. et al. Practical implementation of DNA methylation and copy-number-based CNS tumor diagnostics: the Heidelberg experience. Acta Neuropathol. (Berl.) 136 , 181–210 (2018). Bonner, E. R., Bornhorst, M., Packer, R. J. & Nazarian, J. Liquid biopsy for pediatric central nervous system tumors. Npj Precis. Oncol. 2 , 1–9 (2018). Jackson, R. J. et al. Limitations of stereotactic biopsy in the initial management of gliomas. Neuro-Oncol. 3 , 193–200 (2001). Hamisch, C., Kickingereder, P., Fischer, M., Simon, T. & Ruge, M. I. Update on the diagnostic value and safety of stereotactic biopsy for pediatric brainstem tumors: a systematic review and meta-analysis of 735 cases. J. Neurosurg. Pediatr. 20 , 261–268 (2017). Yi, Z., Qu, C., Zeng, Y. & Liu, Z. Liquid biopsy: early and accurate diagnosis of brain tumor. J. Cancer Res. Clin. Oncol. (2022) doi:10.1007/s00432-022-04011-3. Gelineau, N. U. et al. Case series on clinical applications of liquid biopsy in pediatric solid tumors: towards improved diagnostics and disease monitoring. Front. Oncol. 13 , (2023). Van Paemel, R. et al. The pitfalls and promise of liquid biopsies for diagnosing and treating solid tumors in children : a review. Eur. J. Pediatr. 179 , 191–202 (2020). Izquierdo, E. et al. Droplet digital PCR-based detection of circulating tumor DNA from pediatric high grade and diffuse midline glioma patients. Neuro-Oncol. Adv. 3 , vdab013 (2021). Cantor, E. et al. Serial H3K27M cell-free tumor DNA (cf-tDNA) tracking predicts ONC201 treatment response and progression in diffuse midline glioma. Neuro-Oncol. noac030 (2022) doi:10.1093/neuonc/noac030. Bettegowda, C. et al. Detection of circulating tumor DNA in early- and late-stage human malignancies. Sci. Transl. Med. 6 , 224ra24 (2014). Chang, C. H., Housepian, E. M. & Herbert, C. An Operative Staging System and a Megavoltage Radiotherapeutic Technic for Cerebellar Medulloblastomas. Radiology 93 , 1351–1359 (1969). Wong, T.-T., Liang, M.-L., Chen, H.-H. & Chang, F.-C. Hydrocephalus with brain tumors in children. Childs Nerv. Syst. 27 , 1723–1734 (2011). Pilotto, C. et al. Risk Factors of Persistent Hydrocephalus in Children with Brain Tumor: A Retrospective Analysis. Pediatr. Neurosurg. 56 , 205–212 (2021). Cohen, N. R., Phipps, K., Harding, B. & Jacques, T. S. Is CSF cytology a useful diagnostic procedure in staging paediatric CNS tumours? Cytopathology 20 , 256–260 (2009). Balhuizen, J. C., Bots, G. T. A. M., Schaberg, A. & Bosman, F. T. Value of cerebrospinal fluid cytology for the diagnosis of malignancies in the central nervous system. J. Neurosurg. 48 , 747–753 (1978). Rahimi, J. & Woehrer, A. Chapter 35 - Overview of cerebrospinal fluid cytology. in Handbook of Clinical Neurology (eds. Kovacs, G. G. & Alafuzoff, I.) vol. 145 563–571 (Elsevier, 2018). Kojic, M. et al. Efficient detection and monitoring of pediatric brain malignancies with liquid biopsy based on patient-specific somatic mutation screening. Neuro-Oncol. 25 , 1507–1517 (2023). Cheng, L. et al. Detection of Glioma-Related Hotspot Mutations Through Sequencing of Cerebrospinal Fluid (CSF)-Derived Circulating Tumor DNA: A Pilot Study on CSF-Based Liquid Biopsy for Primary Spinal Cord Astrocytoma. Neurospine 20 , 701–708 (2023). Liu, A. P.-Y., Northcott, P. A., Robinson, G. W. & Gajjar, A. Circulating tumor DNA profiling for childhood brain tumors: Technical challenges and evidence for utility. Lab. Invest. 102 , 134–142 (2022). Panditharatna, E. et al. Clinically Relevant and Minimally Invasive Tumor Surveillance of Pediatric Diffuse Midline Gliomas Using Patient-Derived Liquid Biopsy. Clin. Cancer Res. 24 , 5850–5859 (2018). Wang, Y. et al. Detection of tumor-derived DNA in cerebrospinal fluid of patients with primary tumors of the brain and spinal cord. Proc. Natl. Acad. Sci. 112 , 9704–9709 (2015). Chicard, M. et al. Cell-Free DNA Extracted from CSF for the Molecular Diagnosis of Pediatric Embryonal Brain Tumors. Cancers 15 , 3532 (2023). Li, J. et al. Reliable tumor detection by whole-genome methylation sequencing of cell-free DNA in cerebrospinal fluid of pediatric medulloblastoma. Sci. Adv. 6 , eabb5427 (2020). Afflerbach, A.-K. et al. Classification of Brain Tumors by Nanopore Sequencing of Cell-Free DNA from Cerebrospinal Fluid. Clin. Chem. hvad115 (2023) doi:10.1093/clinchem/hvad115. Koker, A. D., Paemel, R. V., Wilde, B. D., Preter, K. D. & Callewaert, N. A versatile method for circulating cell-free DNA methylome profiling by reduced representation bisulfite sequencing. 663195 Preprint at https://doi.org/10.1101/663195 (2019). Van Paemel, R. et al. Minimally invasive classification of paediatric solid tumours using reduced representation bisulphite sequencing of cell-free DNA: a proof-of-principle study. Epigenetics 16 , 196–208. Koker, A. D., Paemel, R. V., Wilde, B. D., Preter, K. D. & Callewaert, N. cf-RRBS protocol. protocols.io https://www.protocols.io/view/cf-rrbs-protocol-pc6dize (2020). Almstrup, K. et al. Pubertal development in healthy children is mirrored by DNA methylation patterns in peripheral blood. Sci. Rep. 6 , 28657 (2016). Caggiano, C. et al. Comprehensive cell type decomposition of circulating cell-free DNA with CelFiE. Nat. Commun. 12 , 2717 (2021). Krueger, F. & Andrews, S. R. Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications. Bioinformatics 27 , 1571–1572 (2011). Chicard, M. et al. Whole-Exome Sequencing of Cell-Free DNA Reveals Temporo-spatial Heterogeneity and Identifies Treatment-Resistant Clones in Neuroblastoma. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 24 , 939–949 (2018). Pan, L. et al. Non-invasive epigenomic molecular phenotyping of the human brain via liquid biopsy of cerebrospinal fluid and next generation sequencing. Eur. J. Neurosci. 52 , 4536–4545 (2020). Zuccato, J. A. et al. Cerebrospinal fluid methylome-based liquid biopsies for accurate malignant brain neoplasm classification. Neuro-Oncol. 25 , 1452–1460 (2022). Sun, Y. et al. Exploring genetic alterations in circulating tumor DNA from cerebrospinal fluid of pediatric medulloblastoma. Sci. Rep. 11 , 5638 (2021). Liu, A. P. Y. et al. Serial assessment of measurable residual disease in medulloblastoma liquid biopsies. Cancer Cell 39 , 1519-1530.e4 (2021). Orzan, F. et al. Liquid Biopsy of Cerebrospinal Fluid Enables Selective Profiling of Glioma Molecular Subtypes at First Clinical Presentation. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 29 , 1252–1266 (2023). Escudero, L. et al. Circulating tumour DNA from the cerebrospinal fluid allows the characterisation and monitoring of medulloblastoma. Nat. Commun. 11 , 5376 (2020). Avanzini, S. et al. A mathematical model of ctDNA shedding predicts tumor detection size. Sci. Adv. 6 , eabc4308 (2020). Drexler, R. et al. Unclassifiable CNS tumors in DNA methylation-based classification: clinical challenges and prognostic impact. Acta Neuropathol. Commun. 12 , 9 (2024). Additional Declarations No competing interests reported. Supplementary Files 20240403Supplementaryfigures.docx Supplemenaryfile1LenghtProfilescfDNA.pdf Supplementarytable1Patientinformation.xlsx Supplementarytable2SampleInformation.xlsx Supplementarytable3SequencingQC.xlsx Supplementarytable4DeconvolutionResults.xlsx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 21 May, 2024 Reviews received at journal 13 May, 2024 Reviews received at journal 06 May, 2024 Reviewers agreed at journal 06 May, 2024 Reviewers agreed at journal 30 Apr, 2024 Reviewers invited by journal 30 Apr, 2024 Editor assigned by journal 15 Apr, 2024 Submission checks completed at journal 15 Apr, 2024 First submitted to journal 04 Apr, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4218805","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":291328836,"identity":"d8c254eb-77d9-4242-9654-4c8c91413417","order_by":0,"name":"Lotte Cornelli","email":"","orcid":"","institution":"Ghent University","correspondingAuthor":false,"prefix":"","firstName":"Lotte","middleName":"","lastName":"Cornelli","suffix":""},{"id":291328837,"identity":"32e1a07f-89ba-4d8a-815a-2a0fe1039667","order_by":1,"name":"Ruben Van Paemel","email":"","orcid":"","institution":"Ghent University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Ruben","middleName":"Van","lastName":"Paemel","suffix":""},{"id":291328838,"identity":"9a49b6ee-d30c-427e-b46e-ffdaeda58dcb","order_by":2,"name":"Maísa Santos","email":"","orcid":"","institution":"VIB-UGent Center for Medical Biotechnology","correspondingAuthor":false,"prefix":"","firstName":"Maísa","middleName":"","lastName":"Santos","suffix":""},{"id":291328839,"identity":"f144d7ad-a1c7-435b-9c5a-12a417b43acd","order_by":3,"name":"Sofie Roelandt","email":"","orcid":"","institution":"VIB-UGent Center for Medical Biotechnology","correspondingAuthor":false,"prefix":"","firstName":"Sofie","middleName":"","lastName":"Roelandt","suffix":""},{"id":291328840,"identity":"a05364d9-0ef4-4f2d-a0e3-adf4979dc1d8","order_by":4,"name":"Leen Willems","email":"","orcid":"","institution":"Ghent University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Leen","middleName":"","lastName":"Willems","suffix":""},{"id":291328841,"identity":"17429f44-14c2-4fbd-8662-72feabec20da","order_by":5,"name":"Jelle Vandersteene","email":"","orcid":"","institution":"Ghent University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jelle","middleName":"","lastName":"Vandersteene","suffix":""},{"id":291328842,"identity":"d948d985-69dc-4b5f-a494-9912633d7086","order_by":6,"name":"Edward Baert","email":"","orcid":"","institution":"Ghent University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Edward","middleName":"","lastName":"Baert","suffix":""},{"id":291328843,"identity":"b998aaec-9650-4eeb-9229-516a124191d3","order_by":7,"name":"Liselot M. Mus","email":"","orcid":"","institution":"Ghent University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Liselot","middleName":"M.","lastName":"Mus","suffix":""},{"id":291328844,"identity":"2e20ccd4-9aff-4e48-9dd9-1107d1988d8a","order_by":8,"name":"Nadine Van Roy","email":"","orcid":"","institution":"Ghent University","correspondingAuthor":false,"prefix":"","firstName":"Nadine","middleName":"Van","lastName":"Roy","suffix":""},{"id":291328845,"identity":"68250198-b44d-4bea-a21e-31eec1ce63a8","order_by":9,"name":"Bram De Wilde","email":"","orcid":"","institution":"Ghent University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Bram","middleName":"","lastName":"De Wilde","suffix":""},{"id":291328846,"identity":"3cea7797-a55b-46ab-a186-953007270c69","order_by":10,"name":"Katleen De Preter","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABHklEQVRIiWNgGAWjYBACCQYGxgNg1gGoCD+UTsCthZkBSQtQnWQDhEGCFoMDBLRItp8/cOBDxR0GvuO9xz58/GGTb3wj/eLjwh8Mefw4tEjzJDMcnHHmGYPkmXPJM2ckpFluu5FTbDwjgaEY4kJMIMeQzHCYt+0wg8GNHGNmnoTDBmY3ctKkeRIYEjccwKGF/zHD4b//gFruvzFm/pPw38B4Rk76b5CW/Ti0SEsAbWFsANnCY8zMkHDAwEAi/Rgz2BZc3p/x2OBgz7HDPJJn8pIZe9KSDSTOvGGW5kmTKJbAYYvE+cSHD37UHJbjO372MMMPGzsD/vb0h595bGzy+HF4HwZ4wAjKNmAAJwvCAK6F/QExykfBKBgFo2DkAAACz2Bm2IMtUAAAAABJRU5ErkJggg==","orcid":"","institution":"VIB-UGent Center for Medical Biotechnology","correspondingAuthor":true,"prefix":"","firstName":"Katleen","middleName":"","lastName":"De Preter","suffix":""}],"badges":[],"createdAt":"2024-04-04 15:42:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4218805/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4218805/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":54914229,"identity":"faacc338-e036-4132-8209-7534130419c1","added_by":"auto","created_at":"2024-04-18 13:50:18","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":617560,"visible":true,"origin":"","legend":"\u003cp\u003eA: UMAP visualization of the DNA methylation reference dataset that is built for computational deconvolution of pediatric brain tumor classification. Only data from regions covered by both cfRRBS and Illumina 450k arrays are included. The two sample groups supplemented to the published dataset of Capper et al are indicated with an * B: Zoom-in on the low grade glioma (LGG) clusters that overlap with other tumor types.\u003c/p\u003e","description":"","filename":"Figure1UMAP.png","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/7db037d5f2e68b56a1cdf046.png"},{"id":54914233,"identity":"52714183-e567-409d-8e90-e75dca9ece87","added_by":"auto","created_at":"2024-04-18 13:50:18","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1603172,"visible":true,"origin":"","legend":"\u003cp\u003eVisualization of the estimated tumor fraction according to computational deconvolution based on DNA methylation profiles of the tumor tissue samples. The fraction of the histopathological diagnosis is indicated in BROWN; the other estimated tumor fractions that do not correspondent with the histopathological diagnosis are indicated in ORANGE and the non-tumoral fraction in BEIGE. Samples are classified correctly when the highest estimated tumoral fraction corresponds to the diagnosis, indicated with an asterisk. Table with full classification results for each sample is available in supplemental data.\u003c/p\u003e","description":"","filename":"Figure2cfDNACSF.png","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/5f0aef00ec7e86779cf00e2e.png"},{"id":54915394,"identity":"23fcb675-3205-40e0-a55b-b6edbecfb93d","added_by":"auto","created_at":"2024-04-18 14:06:18","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":146046,"visible":true,"origin":"","legend":"\u003cp\u003ePercentage of cfDNA (70 bp-700 bp) over total DNA isolated from CSF. A. cfDNA fraction over total DNA in whole CSF (n=12) versus centrifuged CSF samples (n=20). B. cfDNA fraction over total DNA per tumor type for centrifuged samples. Included tumor types are atypical teratoid rhabdoid tumor (ATRT), adamantinomatous craniopharyngioma (CPH), diffuse midline glioma (DMG), ependymoma (EPN), low grade glioma tumors (LGG), medulloblastoma (MB), choroid plexus papilloma (PLEX).\u003c/p\u003e","description":"","filename":"Figure3DeconvolutionTumor.png","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/3f08774600c2bb4f38b34647.png"},{"id":54914228,"identity":"88b54d93-8a4b-4974-99fc-fec4a3a9fd6c","added_by":"auto","created_at":"2024-04-18 13:50:18","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":73933,"visible":true,"origin":"","legend":"\u003cp\u003ePlot of the cfDNA fraction based on length profiling before library preparation, and the estimated tumor burden after deconvolution. Samples that score high on both these variables show a classification that corresponds with the pathological diagnosis. Correct classification in BLUE, incorrect classification in ORANGE.\u003c/p\u003e","description":"","filename":"Figure4DeconvolutionCSFcfDNA.png","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/5041aaaf944cb4500ac2db88.png"},{"id":54915395,"identity":"f1dcd8d2-a5f5-49d2-bf10-df80bb040db8","added_by":"auto","created_at":"2024-04-18 14:06:18","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":641551,"visible":true,"origin":"","legend":"\u003cp\u003eDNA copy number profiles of four included patients, with corresponding estimated tumor fractions (ETF) and cfDNA fraction of the CSF-cfDNA. Overlapping profiles of the tumor formalin fixed paraffin embedded (FFPE) material in BLUE and CSF material in ORANGE show both corresponding aberrations as well as some heterogeneity.\u003c/p\u003e","description":"","filename":"Figure5CopyNumbers.png","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/b3a16874262e9e7fea60e285.png"},{"id":54916107,"identity":"2138fe09-3759-407a-b44b-603447b9a2d3","added_by":"auto","created_at":"2024-04-18 14:14:22","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3029734,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/8aaa43ac-77ad-46c8-9e67-e9fb23125da0.pdf"},{"id":54914977,"identity":"2d7ae2b9-6aa0-4199-a68a-1b6958fbef7b","added_by":"auto","created_at":"2024-04-18 13:58:18","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":203156,"visible":true,"origin":"","legend":"","description":"","filename":"20240403Supplementaryfigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/0afec65ff811a054d8204d13.docx"},{"id":54914236,"identity":"ceefd967-b315-4c8e-951c-10e6bdc48e46","added_by":"auto","created_at":"2024-04-18 13:50:18","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":4443825,"visible":true,"origin":"","legend":"","description":"","filename":"Supplemenaryfile1LenghtProfilescfDNA.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/62ec2f79b8ea2dcd5f69a142.pdf"},{"id":54914235,"identity":"28607660-ec8a-4cfd-bcad-cbc22485374e","added_by":"auto","created_at":"2024-04-18 13:50:18","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":45584,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarytable1Patientinformation.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/2ecd77a085ffd95375f02c3f.xlsx"},{"id":54914979,"identity":"f2ca4219-24e3-4827-a796-7d75a7f28fa8","added_by":"auto","created_at":"2024-04-18 13:58:18","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":44207,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarytable2SampleInformation.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/063cfd90b6826c426c604d7b.xlsx"},{"id":54914240,"identity":"fe1e7019-6c2a-4822-ba20-e871a5001b48","added_by":"auto","created_at":"2024-04-18 13:50:19","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":74063,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarytable3SequencingQC.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/a1fc624d2ea97c24a303f33d.xlsx"},{"id":54914239,"identity":"3266a3ea-b1fc-4e53-b5aa-b43c15e279dd","added_by":"auto","created_at":"2024-04-18 13:50:19","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":116729,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarytable4DeconvolutionResults.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4218805/v1/7aff214a2d46f2c62b753063.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Diagnosis of pediatric central nervous system tumors using methylation profiling of cfDNA from cerebrospinal fluid","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIntracranial central nervous system (CNS) tumors are one of the leading causes of cancer related death in children after leukemia\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Diagnosis of these brain tumors is complex as they consist of a heterogeneous group of tumors, from slow-growing, low grade lesions to high grade cancers\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. An accurate and detailed diagnosis is vital for treatment decisions and determines patient outcome\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. The current diagnostic procedure of (pediatric) brain tumors requires a tumor tissue biopsy for histopathological investigation\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. In this evaluation, inter-observatory variability occasionally results in misdiagnosis\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Diagnostic procedures are evolving from purely histology based methods\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e, to a combined approach where assays that interrogate molecular markers are becoming increasingly important\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Indeed, the latest 2021 World health organization (WHO) classification entails tumor types and subtypes that can only be distinguished by combinations of new molecular profiling methods\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Genomic and epigenomic analyses have improved the diagnostic process for many cancer entities. More specifically the tumor DNA methylation profile is shown to be tissue specific and therefore is a powerful tool for tumor classification\u003csup\u003e\u003cspan additionalcitationids=\"CR12 CR13\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, as very convincingly shown for brain tumor (sub-)classification\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFor brain cancer patients with delicate tumor location, performing a biopsy or resection can hold disproportional risks\u003csup\u003e\u003cspan additionalcitationids=\"CR16 CR17\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. In the last decade, the potential of the use of liquid biopsies has become evident, emerging as a novel and valuable approach to molecularly investigate the tumor in a minimally invasive manner\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Tumoral biomolecules, including circulating cell-free DNA (cfDNA), RNA and proteins\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e are released from different locations within the tumor, and therefore contain molecular information while avoiding sampling bias often seen in tissue biopsies\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. These molecules are found in biofluids surrounding the tumor, including blood, urine, cerebrospinal fluid, and others\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Several studies have shown that the amount of circulating tumor DNA (ctDNA) in the blood is limited in patients with intracranial tumors due to retention by the blood-brain-barrier\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. In these studies, cerebrospinal fluid is suggested as a superior source for ctDNA. The extraction of CSF is often an uncomfortable and invasive procedure, requiring a painful lumbar puncture to acquire fluid from the space surrounding the spinal cord. However, for pediatric patients, CSF is collected during standard diagnostic procedures and collection for cfDNA would require no additional procedures. These patients often present with symptoms of increased intracranial pressure due to CSF flow obstruction requiring emergency placement of an external ventricular drain to remove the excess fluids\u003csup\u003e\u003cspan additionalcitationids=\"CR25\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. For tumors with risk of cerebrospinal fluid metastasis, an additional lumbar puncture is often performed after removal of the tumor, to examine the CSF for the presence of circulating tumor cells\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003ePublished cfDNA studies on CSF have mostly focused on tumor follow-up rather than classification. For example, cfDNA is used for mutation detection where tumors are detected based on the previously defined patient/tumor specific mutations\u003csup\u003e\u003cspan additionalcitationids=\"CR31 CR32 CR33 CR34\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Since mutations frequently fail to differentiate between tumor subtypes, Li and colleagues instead used whole genome bisulfite sequencing (WGBS) based DNA methylation profiling and hydroxymethylation profiling through anti-CMS immunoprecipitation sequencing to subtype and monitor medulloblastoma tumors\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. A recent study utilized nanopore technology to investigate tumor classification. Here, Afflebach et al. conducted nanopore sequencing on cerebrospinal fluid cell-free DNA (CSF-cfDNA) to classify tumors based on methylation patterns. However, they were only successful in detecting ctDNA based on methylation profiling in 15 out of 178 samples. Out of these 15 samples, accurate classification was achieved for 13 of them\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Although promising, the authors describe a minimal input 1 ml CSF and 5 ng cfDNA, which is challenging to collect in some CSF samples. Additionally, samples in this study were lost after failing the technical pass of 100 000 reads, or were unsuited for methylation profiling when they did not cover a minimal of 1000 CpG\u0026rsquo;s\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eConsidering the significant expense associated with the use of WGBS and the fact that nanopore sequencing of cfDNA has yielded successful results only in a subset of samples, we have generated proof-of-concept for the use of an alternative technology that allows accurate and minimally-invasive classification of pediatric brain tumors. More specifically, in this study, we use cell-free reduced representation bisulfite sequencing (cfRRBS) that allows methylation profiling of low amounts of fragmented DNA, such as cfDNA\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. The method involves a step to enrich the more relevant methylated regions, i.e. the CpG rich regions, resulting in a reduction of sequencing cost per sample compared to WGBS. Before, we (Van Paemel et al.) have demonstrated the potential of cfRRBS for accurate cfDNA methylation-based diagnosis of pediatric solid tumors\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e, by achieving a correct classification rate of 94% in high-quality samples, mostly cfDNA from blood plasma samples. In this study, we explore the diagnostic potential of cfRRBS followed by DNA methylation signal deconvolution for tumor fraction estimation on cfDNA isolated from cerebrospinal fluids in pediatric brain tumor patients.\u003c/p\u003e"},{"header":"Material and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatients and samples\u003c/h2\u003e \u003cp\u003eThis study was approved by the ethical committee and informed consent was obtained from all patients and/or their representatives. Pediatric patients presenting with a central nervous system tumor at Ghent University Hospital from February 2020 to July 2023 (n\u0026thinsp;=\u0026thinsp;19; Age range 3 months to 16 years old) that required ventricular drainage were included. Samples were collected from patients that were pathologically diagnosed with medulloblastoma (n\u0026thinsp;=\u0026thinsp;6), pilocytic astrocytoma (n\u0026thinsp;=\u0026thinsp;6), ependymoma (n\u0026thinsp;=\u0026thinsp;3; two samples were collected from the same patient), choroid plexus papilloma (n\u0026thinsp;=\u0026thinsp;2), diffuse midline glioma (n\u0026thinsp;=\u0026thinsp;1), adamantinomatous craniopharyngioma (n\u0026thinsp;=\u0026thinsp;1), and atypical teratoid/rhabdoid tumor (n\u0026thinsp;=\u0026thinsp;1). Complete patient and sample data is available in the supplemental Tables\u0026nbsp;1 and 2.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eCSF collection\u003c/h2\u003e \u003cp\u003eDetermined by the amount that could safely be sampled, 1.5 to 20 ml of cerebrospinal fluid was collected. For the first 12 patients, samples were separated into two aliquots. One aliquot was centrifuged for 10 minutes at 1900 g and the supernatant transferred to a clean 15 ml falcon tube. The other aliquot was processed without additional interventions. For the following patients, the complete CSF sample was centrifuged. Both the unprocessed CSF and the supernatant after centrifugation were stored at -80\u0026deg;C CSF of 13 patients was centrifuged immediately and stored within four hours after collection. For the remaining patients (n\u0026thinsp;=\u0026thinsp;7/20), samples were temporarily frozen at the operation room at -20\u0026deg;C before any processing. These samples were subsequently stored at -80\u0026deg;C and centrifuged on the day they were further processed.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eCell-free DNA extraction and quality control\u003c/h2\u003e \u003cp\u003eCSF samples were thawed to room temperature and cell-free DNA was extracted using a Maxwell RSC LV ccfDNA kit (Promega). Depending on sample availability, between 1 and 7.5 ml of CSF was used as input for cfDNA extraction. in order to reach the minimal input for this protocol, one sample with a volume below 2 ml was supplemented with PBS (1X, Gibco) to a total volume of 2 ml, prior to addition of binding buffer. The extraction was performed according to the manufacturers guidelines and the resulting cfDNA was eluted in 75 \u0026micro;l of elution buffer supplied in the kit (Promega).\u003c/p\u003e \u003cp\u003ecfDNA concentration was measured using Fluoroskan\u0026trade; Microplate Fluorometer (Thermo Fisher Scientific) according to the manufacturer\u0026rsquo;s instructions. Size distribution profiles were obtained using Agilent Tapestation with the Cell-Free DNA ScreenTape kit. CfDNA was defined as DNA fragments with lengths ranging between 70 and 700 bp and high molecular weight DNA (HMW-DNA) as DNA fragments with lengths exceeding 700 bp. Isolated DNA was stored at -20\u0026deg;C until further processing.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eDNA isolation from surgical tumor biopsy samples\u003c/h2\u003e \u003cp\u003eAn aliquot of isolated genomic DNA from a tumor tissue biopsy was obtained for 11 of the 19 patients. DNA was extracted starting from 3 to 15 paraffin embedded (FFPE) tissue slides according to the manufacturer\u0026rsquo;s instructions using the Qiamp DNA FFPE Tissue kit (QIAGEN). DNA was stored at 4\u0026deg;C until processing.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003ecfRRBS library preparation\u003c/h2\u003e \u003cp\u003eIsolated DNA was processed using cell-free reduced representation bisulfite sequencing as previously described\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. For the initial 12 patients, DNA from whole CSF as well as the centrifuged aliquot was used as input for cfRRBS. When available (n\u0026thinsp;=\u0026thinsp;29/32), 10 ng of DNA was used as input as described in the protocol. Samples with a DNA concentration below 0.2 ng/\u0026micro;L were concentrated via vacuum centrifugation (SpeedVac, Thermo Fischer Scientific) at 45\u0026deg;C. According to the protocol, 0.01 ng lambda spike-in was added to the samples. After library amplification, DNA was purified using SPRI bead size selection (AMPure XT beads \u0026ndash; NEB), with 2.5x proportion of bead to sample volume. The libraries were quantified and checked for the presence of adapter dimers via the Kapa library quantification kit for Illumina platforms (Kapa Biosystems). The length profile was visualized via Fragment Analyser (Advanced Analytical Technologies). Samples were pooled equimolarly to a total concentration of 4 nM. Final concentration of the pooled samples was verified using the Kapa library quantification kit for Illumina platforms (Kapa Biosystems).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eSequencing quality control and mapping\u003c/h2\u003e \u003cp\u003eSamples were sequenced on a NovaSeq 6000 instrument using a NovaSeq SP kit (paired-end, 2 \u0026times; 50 cycles), supplemented with 3% phiX and a loading concentration between 750 and 800 pM. Samples from different donors and tubes were mixed to avoid sequencing batch effects. BCL files were demultiplexed and quality checked as previously described by Van Paemel et al.\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Updated modules were used for demultiplexing (bcl2fastq v2.20), adapter removal (Trim Galore v0.6.6, and CutAdapt), mapping (SAMtools v.1.14, and Bismark v.0.23.1), read counting (Picard tools v.2.21.6). Visualization of the results was done with R version 4.3.2 and ggplot2 v 3.4.4. We obtained on average 21,7M reads, and a minimum of 7M reads per sample. Mapping efficiency was on average 52%, as to be expected for cfRRBS data; bisulfite conversion was at least 95.9% for all samples, and exceeded 98% for 25/32 samples. Full QC report of the samples is shared in supplementary table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eDevelopment of a reference set for computational deconvolution\u003c/h2\u003e \u003cp\u003eThe methylation profiles of 2801 brain tumors generated on Illumina array 450K platform, encompassing 81 different brain tumor entities, was obtained from Capper et al.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Seven of the tumor entities in this reference data were excluded, as they are not recognized by the WHO CNS5 (2021), namely, infantile hemispheric glioma, high grade neuroepithelial tumor with BCOR alteration, high grade neuroepithelial tumor with MN1 alteration, anaplastic pilocytic astrocytoma, Ewing sarcoma family tumor with CIC alteration, central neurocytoma, and plasmacytoma. All other 2629 samples covering 74 entities were used to build the brain tumor reference dataset. The reference set was adjusted to allow deconvolution of cfRRBS data by only considering the CpGs in regions that overlap in the cfRRBS and array data as described by Van Paemel et al.\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. In house data of healthy plasma cfDNA were also included in the reference dataset, as well as published methylation data of prepuberal white blood cells (WBC, n\u0026thinsp;=\u0026thinsp;52)\u003csup\u003e41\u003c/sup\u003e as some samples with red discoloration are assumed to contain low volumes of contaminating blood (due to placement of the ventricular drain).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eComputational deconvolution of cellular fractions\u003c/h2\u003e \u003cp\u003eTumoral fractions were estimated using Methatlas\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, a non-linear least square based method. Reference and test samples were grouped in 14.103 clusters, and a median value for the methylation status was calculated for each cluster. The beta values of all clusters were used for deconvolution. The tumor classification of the test samples was defined based on the highest estimated fraction in the sample excluding non-tumoral (white blood cell, plasma, and control brain tissue) fractions. Full deconvolution output from both CSF-cfDNA and FFPE samples are added to supplementary table \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003e4\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eCopy number profiles\u003c/h2\u003e \u003cp\u003eCopy number aberrations were inferred from the cfRRBS data of both CSF liquids and tissue biopsies. We used WisecondorX (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/CenterForMedicalGeneticsGhent/WisecondorX\u003c/span\u003e\u003cspan address=\"https://github.com/CenterForMedicalGeneticsGhent/WisecondorX\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to detect copy number aberrations after mapping to the bisulfite converted genome. The binsize was set at 400 kb.\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e Samples were normalized with an in house dataset of cfRRBS data from healthy volunteers.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003e \u003cb\u003eReference dataset for computational deconvolution of pediatric brain tumor fractions\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFor our reference dataset, we modified a published Illumina array dataset to align with the genomic regions covered in cfRRBS data (details in M\u0026amp;M). In the reference set, we also included cfRRBS data of blood plasma cfDNA from non-cancerous volunteers, as well as white blood cells because some CSF samples present with contaminating blood cells. We performed UMAP dimensionality reduction on the cfRRBS reference dataset to visualize the grouping/clustering of the tumor entities based on their methylation profile (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). While most tumor entities can be clearly distinguished in this plot, similar as the published visualization by Capper et al.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e we observed that the low grade glioma clusters overlap with other tumor entities (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). This optimized dataset is used as reference for computational deconvolution of pediatric brain tumor fractions in the next paragraphs.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eCorrect tumor classification using cfRRBS on pediatric brain tumor tissue DNA\u003c/h2\u003e \u003cp\u003eTo validate our deconvolution based classification pipeline, we first applied it on cfRRBS profiles generated on genomic DNA isolated from pediatric brain tumor tissue of 11 patients. For 8 out of 11 tumors, the highest estimated tumor fraction corresponds with the histopathological diagnosis. The highest tumor fraction estimated using deconvolution for the medulloblastoma tumor (MB; n\u0026thinsp;=\u0026thinsp;4), ependymoma tumor (EPN; n\u0026thinsp;=\u0026thinsp;2) and choroid plexus papilloma tumor samples (PLEX; n\u0026thinsp;=\u0026thinsp;1) corresponds with the histopathological diagnosis and assigned subclass (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). However, for pilocytic astrocytoma (LGG-PA; n\u0026thinsp;=\u0026thinsp;4) only one out of 4 tumors are classified correctly. Given interobserver variability is reported for histopathological diagnoses, a pathologist re-examined these 3 cases but excluded any misdiagnosis. The incorrect classification can be explained by the fact that the DNA methylation profile of LGG-PA cases is not distinct enough as observed in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eCirculating cfDNA in cerebrospinal fluid\u003c/h2\u003e \u003cp\u003eWe collected CSF samples of 19 pediatric patients presenting with CNS tumors. For one patient, two CSF samples were collected at the moment of two consecutive relapses. In this cohort, we considered these two samples as independent due to our primary emphasis on the sample quality features. All samples were collected in plastic containers without preservatives. When possible, samples were processed immediately (n\u0026thinsp;=\u0026thinsp;13), if not they were frozen at -20\u0026deg;C in the operation room (n\u0026thinsp;=\u0026thinsp;7). Previous studies have shown that high molecular weight DNA (HMW-DNA) originating from white blood cells can interfere with cfRRBS and deconvolution by diluting the tumoral signal\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. For that reason, CSF was centrifuged with the aim of removing cell debris that could contribute to high molecular weight DNA contamination. To check the effect of the centrifugation, for 12 patients we also stored an aliquot of uncentrifuged material.\u003c/p\u003e \u003cp\u003eNext, cfDNA was isolated, concentration quantified and fragment length analyzed. We observed a high degree of variability in the total amount of cfDNA, and fragment profile between patients. Samples from three patients showed a circulating cfDNA (70\u0026ndash;700 bp length) concentration that was below the limit of detection by Tapestation visualization. Compared to centrifuged CSF, whole CSF samples showed a significantly lower cfDNA concentration on total DNA fraction (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, p-value 0.04814) pointing at more high molecular weight DNA contamination. The total yield of cell-free DNA is not significantly different between whole and centrifuged CSF (p-value\u0026thinsp;=\u0026thinsp;0.4962; figure in supplemental information) indicating that centrifugation primarily decreases the presence of HMW-DNA by removing cells and thus preventing cell lysis in the sample, and has minimal effect on the fragmented cfDNA. Based on these results, we decided to centrifuge the CSF before processing for the other 8 samples (sample 13 to sample 20). Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows the percentage of circulating cfDNA over total DNA that was isolated from CSF after centrifugation of each patient.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eDeconvolution of tumor fractions in cfDNA from CSF\u003c/h2\u003e \u003cp\u003eFollowing quality control, all samples were processed with cfRRBS library preparation, sequencing, and deconvolution to estimate the tumoral and healthy cfDNA fractions. The estimated tumor type (i.e. entity with the highest predicted tumor fraction after using deconvolution) corresponded with the histopathological diagnosis for 8 out of 20 cases: medulloblastoma (4/5), ependymoma (2/3), choroid plexus papilloma (1/2), atypical teratoid rhabdoid tumor (1/1).\u003c/p\u003e \u003cp\u003e We noted higher fractions of fragmented cfDNA on total DNA, and a higher estimated tumor fraction according to deconvolution in samples that were correctly classified. Although the relatively small patient cohort does withhold us from defining validated cut-off values, we see that most samples with a cfDNA/total DNA fraction below 40% and/or an estimated tumor burden below 30% are classified incorrectly (10/14). Samples with cfDNA/total DNA fraction of at least 40% and tumor burden of at least 30% are all classified correctly (6/6).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eCopy number profiling from cfRRBS data\u003c/h2\u003e \u003cp\u003eFrom cfRRBS data, also copy number aberration profiles can be extracted. For 11 donors, we were able to compare copy number profiles from tumor and cfDNA. Although data is more noisy compared to whole genome sequencing methods, we observe aberration in 3 of these patients. For samples with lower estimated tumoral fractions, such as case 1 illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, we could not identify any aberrations in the CSF-cfDNA. Case 5 shows aberrations that correlate well to the ones in the tumor tissue. Interestingly, for 2 of these patients (case 2 and case 5) we observed additional copy number alterations in the liquid samples compared to the tumor suggestive for intratumoral heterogeneity.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this proof-of-concept study we present an analytical and computational pipeline to classify pediatric CNS tumors using a DNA methylation assay that was developed for fragmented cfDNA isolated from liquid biopsies, and that works on both DNA from fresh frozen and paraffin embedded tissues. Tumor type and tumor fraction are estimated using computational deconvolution based on a reference dataset containing published methylation data of brain tumor tissues complemented with healthy cfDNA profiles. We found that the tumor classification correlates well with histopathological diagnosis for good quality cfDNA samples. We identified several pitfalls of our approach related to CSF collection and CSF characteristics, as well as opportunities for improvement which require further validation on larger patient cohorts before clinical implementation.\u003c/p\u003e \u003cp\u003eIn summary, 8 out of 20 samples from pediatric CNS cancer patients are classified correctly using cfDNA from CSF. All samples with high cfDNA/total DNA fraction were classified correctly (6/20). In most misclassified samples, we observe increased fractions of HMW-DNA (length over 700 bp) among the isolated DNA in CSF, with 14 samples showing a cell-free DNA fraction lower than 0.5. Additionally, the more than half of the samples exhibiting a cfDNA yield below 5 ng. The scarcity of cfDNA in CSF is not emphasized in similar studies, yet it is notable that published articles working on cfDNA from CSF for tumor classification and follow-up almost exclusively focus on higher grade tumors\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e,\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e,\u003cspan additionalcitationids=\"CR45 CR46 CR47 CR48 CR49\" citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. Although ctDNA levels are not defined by tumor grade alone, it is an important variable in the release of fragmented DNA\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. Next to the more urgent clinical need for those more aggressive tumors, lack of studies on lower grade tumors most probably indicates the challenges in obtaining sufficient ctDNA material. The Afflerbach study\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e, in contrast, does include low and high grade tumors, and indeed underscores the low abundance of circulating tumor DNA in the CSF. Due to the low sample input and quality, only 12% of the total cohort passes the minimal technical requirements to allow for methylation profiling and tumor classification. Overall, our cfRRBS approach could successfully generated methylation profiles on all included sample and was able to correctly estimate tumor diagnosis in 30% of the samples. Although obtaining sufficient circulating tumoral DNA still appears challenging for lower grade tumors.\u003c/p\u003e \u003cp\u003eIn the paragraphs below, we discuss the different factors that impact the performance of our classification approach, including high molecular weight contamination, tumoral cfDNA fractions and reference dataset.\u003c/p\u003e \u003cp\u003eIn previous cfRRBS studies we have already highlighted the importance of assessing the cfDNA/HMW-DNA fraction in each sample\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. HMW-DNA, that is also processed in the cfRRBS library preparation will dilute the signal of the cfDNA and might lead to misclassification of samples in case of high fractions of HMW-DNA. The HMW-DNA is likely derived from cells that are damaged during ventricular drain placement or white blood cells in blood-contaminated samples. In this study, we show that centrifugation of a fresh CSF sample before DNA isolation improves sample quality in many cases. Freezing the CSF prior to centrifugation results in lysis of the cells and thus release of HMW-DNA into the fluid. Thus, centrifugation within four hours after collection and before freezing is ideal, but more challenging to implement into clinical practice. In addition, avoiding blood cell contamination will result in better quality samples. Secondly, although yield of circulating DNA relates to variables such as tumor size and aggression\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e, sampling in close proximity of the tumor through a ventricular drain might collect more ctDNA compared to sampling via lumbar puncture. These observations underscore the need for more dedicated studies investigating the pre-analytical variables that can improve the sample quality, as well as more standardized protocols for CSF collection ensuring the standard collection of high-quality samples allowing robust tumor classification. The significance of this is highlighted by the wide variation in published collection protocols, encompassing collection through lumbar puncture, ventricular drainage, or mixed cohorts, with volumes ranging from 200\u0026micro;l to 10 ml. While centrifugation is commonly included in most protocols, there are studies that omit this step\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e,\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe absence of a well-defined profile characterizing the non-tumoral background in cerebrospinal fluid (CSF) poses a limitation in accurately estimating the fraction of a specific tumor type. This challenge becomes particularly pronounced in cases where CSF samples contain lower tumor fractions and higher non-tumoral cfDNA. We observed a better classification accuracy in cfDNA samples with higher estimated tumor fractions. In blood samples with lower tumor fraction, the non-tumoral cfDNA mostly originate from the white blood cells\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. For CSF however, the origin of the non-tumoral cfDNA is not clearly defined and might originate from WBC, but also from damaged brain tissue due to increased pressure in patients presenting with hydrocephalus. Although the Capper reference data includes various healthy brain tissues, pediatric profiles often differ from adult. To allow proper deconvolution of all the contributing cell types in the CSF-cfDNA samples, a reference dataset encompassing all those cell types is necessary. However, pediatric CSF collection is only performed to patients with a (suspected) brain related pathology, thus obtaining a reference sample and DNA methylation profile from pediatric hydrocephalus patients without brain pathologies is almost impossible. One option would be the inclusion of CSF from patients with hydrocephalus caused by a traumatic brain injury.\u003c/p\u003e \u003cp\u003eThe performance of deconvolution algorithms heavily relies on the choice of reference data. The DNA-methylation based assay for CNS brain tumor diagnosis is utilized in an increasing number of pathological departments and employs the Infinium HumanMethylation450 BeadChip array\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. This array encompasses 450,000 methylation sites and shows good performance to distinguish between different tumor entities. However, a drawback is that it requires a minimum input amount of 500 ng of DNA, a quantity significantly surpassing the average cell-free DNA (cfDNA) yield from liquid biopsies. To address this challenge, we successfully employed cell-free reduced representation bisulfite sequencing (cfRRBS), an approach tailored for low quantities of highly fragmented DNA, requiring only 10 ng or even less input DNA to generate high quality DNA methylation profiles\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. We formatted the published 450K array methylation data\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e of CNS tumors to align with the cfRRBS workflow and used it as a reference dataset for deconvolution. A limitation of this approach is that we only use the sites that are covered by both the 450K array and the cfRRBS assay, which is only 13.7% of methylation sites that are covered by the cfRRBS assay. By restricting the number of sites, we noticed that discriminating low grade glioma tumors became more challenging as visualized in the UMAP plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) compared to the published UMAP (ref). Building a (cf)RRBS based reference dataset would enable the utilization of all cfRRBS regions in the deconvolution model and thus increase the available information to discriminate different tumor entities; however, this will come with additional efforts and costs. This problem highlights the trade-off between maximizing data inclusivity and managing data availability or associated financial constraints. In addition to the restricted number of sites, the published version of the classifier shows challenges in discriminating low-grade glioma tumors, resulting in less accurate predictions for this particular subtype\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Newer versions of the classifier can improve classification for several challenging tumor types including low grade gliomas, however the reference data of newer versions is not publicly available\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eInterestingly, the data produced via cfRRBS can also be used for copy number variation (CNV) profiling. Although this data is more noisy compared to dedicated copy number profiling assays such as shallow whole genome sequencing (shWGS), extraction of multiple data layers from cfRRBS reads without requiring new input material is an important asset. Compared to most cfDNA shWGS approaches for CNV analysis, cfRRBS lacks a size separation step and thus also HMW-DNA will be processed\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e resulting in a dilution of the tumoral signal. Indeed, for samples with tumoral fraction below 30% we couldn\u0026rsquo;t observe any tumor associated aberrations. For the patients with matched tumor and CSF material and higher estimated tumor fractions, we observed some CSF specific aberrations suggesting intratumoral heterogeneity, similar to the results described by Chicard and colleagues\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. However, it is notable that the lower quality of the CNV profile data limits the number of patients for which the CNV profile can accurately be analyzed.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eAlthough the presented cfRRBS approach on CSF has several limitations and points that need further optimization, we believe that this approach can become a valuable alternative for cancer patients with tumors located in regions that are too delicate for a surgical biopsy. Validation on larger cohorts is required, still we observed accurate classification for patients with cfDNA fractions higher than 50%. We expect that methylation profiling of cfDNA isolated from liquid biopsies could take an important and complementary position next to standard diagnostic approaches, for example by giving an early diagnosis that can inform oncologists and surgeons in their choice for a treatment strategy.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eCSF\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;cerebrospinal fluids\u003c/p\u003e\n\u003cp\u003ecfDNA\u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;circulating cell-free DNA\u003c/p\u003e\n\u003cp\u003eCNS\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;central nervous system\u003c/p\u003e\n\u003cp\u003eWHO \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;World health organisation\u003c/p\u003e\n\u003cp\u003ecfRNA\u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;cell-free RNA\u003c/p\u003e\n\u003cp\u003ectDNA\u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;circulating tumor DNA\u003c/p\u003e\n\u003cp\u003eWGBS\u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;whole genome bisulfite sequencing\u003c/p\u003e\n\u003cp\u003eCSF-cfDNA\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;cerebrospinal fluid cell-free DNA\u003c/p\u003e\n\u003cp\u003ecfRRBS\u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;cell-free reduced representation bisulfite sequencing\u003c/p\u003e\n\u003cp\u003eHMW-DNA\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;high molecular weight DNA\u003c/p\u003e\n\u003cp\u003eWBC\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;white blood cells\u003c/p\u003e\n\u003cp\u003eLGG\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;low grade glioma\u003c/p\u003e\n\u003cp\u003eMB\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;medulloblastoma\u003c/p\u003e\n\u003cp\u003eEPN\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;ependymoma\u003c/p\u003e\n\u003cp\u003ePLEX\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;choroid plexus papilloma\u003c/p\u003e\n\u003cp\u003eLGG-PA\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;pilocytic astrocytoma\u003c/p\u003e\n\u003cp\u003eATRT\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;atypical teratoid rhabdoid tumor\u003c/p\u003e\n\u003cp\u003eCPH\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;adamantinomatous craniopharyngioma\u003c/p\u003e\n\u003cp\u003eDMG\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;diffuse midline glioma\u003c/p\u003e\n\u003cp\u003eFFPE\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;formalin fixed paraffin embedded\u003c/p\u003e\n\u003cp\u003eCNV \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;copy number variation\u003c/p\u003e\n\u003cp\u003eshWGS\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;shallow whole genome sequencing\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eETF \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; estimated tumor fraction\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study is approved by the commission for medical ethics from UZ Ghent (reference ID: EC2019/1514). Written informed consent was obtained from all patients and/or their representatives.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFastQ files of all samples will be published in an EGA dataset respecting the original informed consent agreements.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSiegel, D. A. \u003cem\u003eet al.\u003c/em\u003e Pediatric cancer mortality and survival in the United States, 2001-2016. \u003cem\u003eCancer\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e, 4379\u0026ndash;4389 (2020).\u003c/li\u003e\n\u003cli\u003eVassal, G. \u003cem\u003eet al.\u003c/em\u003e The SIOPE strategic plan: A European cancer plan for children and adolescents. \u003cem\u003eJ. Cancer Policy\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 17\u0026ndash;32 (2016).\u003c/li\u003e\n\u003cli\u003eReddy, A. T. \u0026amp; Packer, R. J. Pediatric central nervous system tumors. \u003cem\u003eCurr. Opin. Oncol.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 186\u0026ndash;193 (1998).\u003c/li\u003e\n\u003cli\u003eDiamandis, P. \u0026amp; Aldape, K. World Health Organization 2016 Classification of Central Nervous System Tumors. \u003cem\u003eNeurol. Clin.\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 439\u0026ndash;447 (2018).\u003c/li\u003e\n\u003cli\u003eLouis, D. N. \u003cem\u003eet al.\u003c/em\u003e The 2021 WHO Classification of Tumors of the Central Nervous System: a summary. \u003cem\u003eNeuro-Oncol.\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 1231\u0026ndash;1251 (2021).\u003c/li\u003e\n\u003cli\u003evan den Bent, M. J. Interobserver variation of the histopathological diagnosis in clinical trials on glioma: a clinician\u0026rsquo;s perspective. \u003cem\u003eActa Neuropathol. (Berl.)\u003c/em\u003e \u003cstrong\u003e120\u003c/strong\u003e, 297\u0026ndash;304 (2010).\u003c/li\u003e\n\u003cli\u003eSturm, D. \u003cem\u003eet al.\u003c/em\u003e New Brain Tumor Entities Emerge from Molecular Classification of CNS-PNETs. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e164\u003c/strong\u003e, 1060\u0026ndash;1072 (2016).\u003c/li\u003e\n\u003cli\u003eEllison, D. W. \u003cem\u003eet al.\u003c/em\u003e Histopathological grading of pediatric ependymoma: reproducibility and clinical relevance in European trial cohorts. \u003cem\u003eJ. Negat. Results Biomed.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 7 (2011).\u003c/li\u003e\n\u003cli\u003eLouis, D. N. \u003cem\u003eet al.\u003c/em\u003e The 2007 WHO Classification of Tumours of the Central Nervous System. \u003cem\u003eActa Neuropathol. (Berl.)\u003c/em\u003e \u003cstrong\u003e114\u003c/strong\u003e, 97\u0026ndash;109 (2007).\u003c/li\u003e\n\u003cli\u003eLouis, D. N. \u003cem\u003eet al.\u003c/em\u003e cIMPACT-NOW update 6: new entity and diagnostic principle recommendations of the cIMPACT-Utrecht meeting on future CNS tumor classification and grading. \u003cem\u003eBrain Pathol. Zurich Switz.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 844\u0026ndash;856 (2020).\u003c/li\u003e\n\u003cli\u003eCapper, D. \u003cem\u003eet al.\u003c/em\u003e DNA methylation-based classification of central nervous system tumours. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e555\u003c/strong\u003e, 469\u0026ndash;474 (2018).\u003c/li\u003e\n\u003cli\u003ePriesterbach-Ackley, L. P. \u003cem\u003eet al.\u003c/em\u003e Brain tumour diagnostics using a DNA methylation-based classifier as a diagnostic support tool. \u003cem\u003eNeuropathol. Appl. Neurobiol.\u003c/em\u003e \u003cstrong\u003e46\u003c/strong\u003e, 478\u0026ndash;492 (2020).\u003c/li\u003e\n\u003cli\u003eJaunmuktane, Z. \u003cem\u003eet al.\u003c/em\u003e Methylation array profiling of adult brain tumours: diagnostic outcomes in a large, single centre. \u003cem\u003eActa Neuropathol. Commun.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 24 (2019).\u003c/li\u003e\n\u003cli\u003eCapper, D. \u003cem\u003eet al.\u003c/em\u003e Practical implementation of DNA methylation and copy-number-based CNS tumor diagnostics: the Heidelberg experience. \u003cem\u003eActa Neuropathol. (Berl.)\u003c/em\u003e \u003cstrong\u003e136\u003c/strong\u003e, 181\u0026ndash;210 (2018).\u003c/li\u003e\n\u003cli\u003eBonner, E. R., Bornhorst, M., Packer, R. J. \u0026amp; Nazarian, J. Liquid biopsy for pediatric central nervous system tumors. \u003cem\u003eNpj Precis. Oncol.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 1\u0026ndash;9 (2018).\u003c/li\u003e\n\u003cli\u003eJackson, R. J. \u003cem\u003eet al.\u003c/em\u003e Limitations of stereotactic biopsy in the initial management of gliomas. \u003cem\u003eNeuro-Oncol.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 193\u0026ndash;200 (2001).\u003c/li\u003e\n\u003cli\u003eHamisch, C., Kickingereder, P., Fischer, M., Simon, T. \u0026amp; Ruge, M. I. Update on the diagnostic value and safety of stereotactic biopsy for pediatric brainstem tumors: a systematic review and meta-analysis of 735 cases. \u003cem\u003eJ. Neurosurg. Pediatr.\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 261\u0026ndash;268 (2017).\u003c/li\u003e\n\u003cli\u003eYi, Z., Qu, C., Zeng, Y. \u0026amp; Liu, Z. Liquid biopsy: early and accurate diagnosis of brain tumor. \u003cem\u003eJ. Cancer Res. Clin. Oncol.\u003c/em\u003e (2022) doi:10.1007/s00432-022-04011-3.\u003c/li\u003e\n\u003cli\u003eGelineau, N. U. \u003cem\u003eet al.\u003c/em\u003e Case series on clinical applications of liquid biopsy in pediatric solid tumors: towards improved diagnostics and disease monitoring. \u003cem\u003eFront. Oncol.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eVan Paemel, R. \u003cem\u003eet al.\u003c/em\u003e The pitfalls and promise of liquid biopsies for diagnosing and treating solid tumors in children : a review. \u003cem\u003eEur. J. Pediatr.\u003c/em\u003e \u003cstrong\u003e179\u003c/strong\u003e, 191\u0026ndash;202 (2020).\u003c/li\u003e\n\u003cli\u003eIzquierdo, E. \u003cem\u003eet al.\u003c/em\u003e Droplet digital PCR-based detection of circulating tumor DNA from pediatric high grade and diffuse midline glioma patients. \u003cem\u003eNeuro-Oncol. Adv.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, vdab013 (2021).\u003c/li\u003e\n\u003cli\u003eCantor, E. \u003cem\u003eet al.\u003c/em\u003e Serial H3K27M cell-free tumor DNA (cf-tDNA) tracking predicts ONC201 treatment response and progression in diffuse midline glioma. \u003cem\u003eNeuro-Oncol.\u003c/em\u003e noac030 (2022) doi:10.1093/neuonc/noac030.\u003c/li\u003e\n\u003cli\u003eBettegowda, C. \u003cem\u003eet al.\u003c/em\u003e Detection of circulating tumor DNA in early- and late-stage human malignancies. \u003cem\u003eSci. Transl. Med.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 224ra24 (2014).\u003c/li\u003e\n\u003cli\u003eChang, C. H., Housepian, E. M. \u0026amp; Herbert, C. An Operative Staging System and a Megavoltage Radiotherapeutic Technic for Cerebellar Medulloblastomas. \u003cem\u003eRadiology\u003c/em\u003e \u003cstrong\u003e93\u003c/strong\u003e, 1351\u0026ndash;1359 (1969).\u003c/li\u003e\n\u003cli\u003eWong, T.-T., Liang, M.-L., Chen, H.-H. \u0026amp; Chang, F.-C. Hydrocephalus with brain tumors in children. \u003cem\u003eChilds Nerv. Syst.\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1723\u0026ndash;1734 (2011).\u003c/li\u003e\n\u003cli\u003ePilotto, C. \u003cem\u003eet al.\u003c/em\u003e Risk Factors of Persistent Hydrocephalus in Children with Brain Tumor: A Retrospective Analysis. \u003cem\u003ePediatr. Neurosurg.\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 205\u0026ndash;212 (2021).\u003c/li\u003e\n\u003cli\u003eCohen, N. R., Phipps, K., Harding, B. \u0026amp; Jacques, T. S. Is CSF cytology a useful diagnostic procedure in staging paediatric CNS tumours? \u003cem\u003eCytopathology\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 256\u0026ndash;260 (2009).\u003c/li\u003e\n\u003cli\u003eBalhuizen, J. C., Bots, G. T. A. M., Schaberg, A. \u0026amp; Bosman, F. T. Value of cerebrospinal fluid cytology for the diagnosis of malignancies in the central nervous system. \u003cem\u003eJ. Neurosurg.\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 747\u0026ndash;753 (1978).\u003c/li\u003e\n\u003cli\u003eRahimi, J. \u0026amp; Woehrer, A. Chapter 35 - Overview of cerebrospinal fluid cytology. in \u003cem\u003eHandbook of Clinical Neurology\u003c/em\u003e (eds. Kovacs, G. G. \u0026amp; Alafuzoff, I.) vol. 145 563\u0026ndash;571 (Elsevier, 2018).\u003c/li\u003e\n\u003cli\u003eKojic, M. \u003cem\u003eet al.\u003c/em\u003e Efficient detection and monitoring of pediatric brain malignancies with liquid biopsy based on patient-specific somatic mutation screening. \u003cem\u003eNeuro-Oncol.\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 1507\u0026ndash;1517 (2023).\u003c/li\u003e\n\u003cli\u003eCheng, L. \u003cem\u003eet al.\u003c/em\u003e Detection of Glioma-Related Hotspot Mutations Through Sequencing of Cerebrospinal Fluid (CSF)-Derived Circulating Tumor DNA: A Pilot Study on CSF-Based Liquid Biopsy for Primary Spinal Cord Astrocytoma. \u003cem\u003eNeurospine\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 701\u0026ndash;708 (2023).\u003c/li\u003e\n\u003cli\u003eLiu, A. P.-Y., Northcott, P. A., Robinson, G. W. \u0026amp; Gajjar, A. Circulating tumor DNA profiling for childhood brain tumors: Technical challenges and evidence for utility. \u003cem\u003eLab. Invest.\u003c/em\u003e \u003cstrong\u003e102\u003c/strong\u003e, 134\u0026ndash;142 (2022).\u003c/li\u003e\n\u003cli\u003ePanditharatna, E. \u003cem\u003eet al.\u003c/em\u003e Clinically Relevant and Minimally Invasive Tumor Surveillance of Pediatric Diffuse Midline Gliomas Using Patient-Derived Liquid Biopsy. \u003cem\u003eClin. Cancer Res.\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 5850\u0026ndash;5859 (2018).\u003c/li\u003e\n\u003cli\u003eWang, Y. \u003cem\u003eet al.\u003c/em\u003e Detection of tumor-derived DNA in cerebrospinal fluid of patients with primary tumors of the brain and spinal cord. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e112\u003c/strong\u003e, 9704\u0026ndash;9709 (2015).\u003c/li\u003e\n\u003cli\u003eChicard, M. \u003cem\u003eet al.\u003c/em\u003e Cell-Free DNA Extracted from CSF for the Molecular Diagnosis of Pediatric Embryonal Brain Tumors. \u003cem\u003eCancers\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 3532 (2023).\u003c/li\u003e\n\u003cli\u003eLi, J. \u003cem\u003eet al.\u003c/em\u003e Reliable tumor detection by whole-genome methylation sequencing of cell-free DNA in cerebrospinal fluid of pediatric medulloblastoma. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, eabb5427 (2020).\u003c/li\u003e\n\u003cli\u003eAfflerbach, A.-K. \u003cem\u003eet al.\u003c/em\u003e Classification of Brain Tumors by Nanopore Sequencing of Cell-Free DNA from Cerebrospinal Fluid. \u003cem\u003eClin. Chem.\u003c/em\u003e hvad115 (2023) doi:10.1093/clinchem/hvad115.\u003c/li\u003e\n\u003cli\u003eKoker, A. D., Paemel, R. V., Wilde, B. D., Preter, K. D. \u0026amp; Callewaert, N. A versatile method for circulating cell-free DNA methylome profiling by reduced representation bisulfite sequencing. 663195 Preprint at https://doi.org/10.1101/663195 (2019).\u003c/li\u003e\n\u003cli\u003eVan Paemel, R. \u003cem\u003eet al.\u003c/em\u003e Minimally invasive classification of paediatric solid tumours using reduced representation bisulphite sequencing of cell-free DNA: a proof-of-principle study. \u003cem\u003eEpigenetics\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 196\u0026ndash;208.\u003c/li\u003e\n\u003cli\u003eKoker, A. D., Paemel, R. V., Wilde, B. D., Preter, K. D. \u0026amp; Callewaert, N. cf-RRBS protocol. \u003cem\u003eprotocols.io\u003c/em\u003e https://www.protocols.io/view/cf-rrbs-protocol-pc6dize (2020).\u003c/li\u003e\n\u003cli\u003eAlmstrup, K. \u003cem\u003eet al.\u003c/em\u003e Pubertal development in healthy children is mirrored by DNA methylation patterns in peripheral blood. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 28657 (2016).\u003c/li\u003e\n\u003cli\u003eCaggiano, C. \u003cem\u003eet al.\u003c/em\u003e Comprehensive cell type decomposition of circulating cell-free DNA with CelFiE. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 2717 (2021).\u003c/li\u003e\n\u003cli\u003eKrueger, F. \u0026amp; Andrews, S. R. Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1571\u0026ndash;1572 (2011).\u003c/li\u003e\n\u003cli\u003eChicard, M. \u003cem\u003eet al.\u003c/em\u003e Whole-Exome Sequencing of Cell-Free DNA Reveals Temporo-spatial Heterogeneity and Identifies Treatment-Resistant Clones in Neuroblastoma. \u003cem\u003eClin. Cancer Res. Off. J. Am. Assoc. Cancer Res.\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 939\u0026ndash;949 (2018).\u003c/li\u003e\n\u003cli\u003ePan, L. \u003cem\u003eet al.\u003c/em\u003e Non-invasive epigenomic molecular phenotyping of the human brain via liquid biopsy of cerebrospinal fluid and next generation sequencing. \u003cem\u003eEur. J. Neurosci.\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 4536\u0026ndash;4545 (2020).\u003c/li\u003e\n\u003cli\u003eZuccato, J. A. \u003cem\u003eet al.\u003c/em\u003e Cerebrospinal fluid methylome-based liquid biopsies for accurate malignant brain neoplasm classification. \u003cem\u003eNeuro-Oncol.\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 1452\u0026ndash;1460 (2022).\u003c/li\u003e\n\u003cli\u003eSun, Y. \u003cem\u003eet al.\u003c/em\u003e Exploring genetic alterations in circulating tumor DNA from cerebrospinal fluid of pediatric medulloblastoma. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 5638 (2021).\u003c/li\u003e\n\u003cli\u003eLiu, A. P. Y. \u003cem\u003eet al.\u003c/em\u003e Serial assessment of measurable residual disease in medulloblastoma liquid biopsies. \u003cem\u003eCancer Cell\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 1519-1530.e4 (2021).\u003c/li\u003e\n\u003cli\u003eOrzan, F. \u003cem\u003eet al.\u003c/em\u003e Liquid Biopsy of Cerebrospinal Fluid Enables Selective Profiling of Glioma Molecular Subtypes at First Clinical Presentation. \u003cem\u003eClin. Cancer Res. Off. J. Am. Assoc. Cancer Res.\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 1252\u0026ndash;1266 (2023).\u003c/li\u003e\n\u003cli\u003eEscudero, L. \u003cem\u003eet al.\u003c/em\u003e Circulating tumour DNA from the cerebrospinal fluid allows the characterisation and monitoring of medulloblastoma. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 5376 (2020).\u003c/li\u003e\n\u003cli\u003eAvanzini, S. \u003cem\u003eet al.\u003c/em\u003e A mathematical model of ctDNA shedding predicts tumor detection size. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, eabc4308 (2020).\u003c/li\u003e\n\u003cli\u003eDrexler, R. \u003cem\u003eet al.\u003c/em\u003e Unclassifiable CNS tumors in DNA methylation-based classification: clinical challenges and prognostic impact. \u003cem\u003eActa Neuropathol. Commun.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 9 (2024).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"clinical-epigenetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"clep","sideBox":"Learn more about [Clinical Epigenetics](http://clinicalepigeneticsjournal.biomedcentral.com/)","snPcode":"13148","submissionUrl":"https://submission.nature.com/new-submission/13148/3","title":"Clinical Epigenetics","twitterHandle":"@OAgenetics","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Pediatric oncology, DNA methylation, liquid biopsy, central nervous system tumor, precision medicine, cerebrospinal fluid","lastPublishedDoi":"10.21203/rs.3.rs-4218805/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4218805/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePediatric central nervous system tumors remain challenging to diagnose. Imaging approaches do not provide sufficient detail to discriminate between different tumor types, while the histopathological examination of tumor tissue shows high interobserver variability. Recent studies have demonstrated the accurate classification of central nervous system tumors based on the DNA-methylation profile on a tumor biopsy. However, a brain biopsy holds significant risk of bleeding and damaging the surrounding tissues. Liquid biopsy approaches analyzing circulating tumor DNA show high potential as an alternative and less invasive tool to study the DNA-methylation pattern of tumors. In this study, we explore the potential of classifying pediatric brain tumors based on methylation profiling of the cell-free DNA in cerebrospinal fluid (CSF). For this proof-of-concept study, we collected 20 cerebrospinal fluid samples of pediatric brain cancer patients via a ventricular drain placed for reasons of increased intracranial pressure. Analyses on the circulating cell-free DNA (cfDNA) showed high variability of cfDNA quantities across patients ranging from levels below the limit of quantification to 40 ng cfDNA per milliliter of CSF. Classification based on methylation profiling of cfDNA from CSF was correct for 8 out of 20 samples in our cohort. Accurate results were mostly observed in samples of high quality, more specifically those with limited high-molecular weight DNA contamination. Interestingly, we show that centrifugation of the CSF prior to processing increases the fraction of fragmented cfDNA to high-molecular weight DNA. In addition, classification was mostly correct for samples with high tumoral cfDNA fraction as estimated by computational deconvolution (\u0026gt;\u0026thinsp;40%). In summary, analysis of cfDNA in the CSF shows potential as a tool for diagnosing pediatric nervous system tumors especially in patients with high levels of tumoral cfDNA in the CSF, however further optimization of the collection procedure, experimental workflow, and bioinformatic approach is required to also allow classification for patients with low tumoral fractions in the CSF.\u003c/p\u003e","manuscriptTitle":"Diagnosis of pediatric central nervous system tumors using methylation profiling of cfDNA from cerebrospinal fluid","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-18 13:50:13","doi":"10.21203/rs.3.rs-4218805/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-05-21T09:01:19+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-13T10:26:24+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-05-06T16:00:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"55588189801846586188189068166804852720","date":"2024-05-06T14:04:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"142432563052894097722272044005924911910","date":"2024-04-30T14:30:23+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-04-30T14:24:19+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-04-15T10:34:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-04-15T10:33:52+00:00","index":"","fulltext":""},{"type":"submitted","content":"Clinical Epigenetics","date":"2024-04-04T15:39:29+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"clinical-epigenetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"clep","sideBox":"Learn more about [Clinical Epigenetics](http://clinicalepigeneticsjournal.biomedcentral.com/)","snPcode":"13148","submissionUrl":"https://submission.nature.com/new-submission/13148/3","title":"Clinical Epigenetics","twitterHandle":"@OAgenetics","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"afe02b61-70c6-499e-8767-dcd91c58da0e","owner":[],"postedDate":"April 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2024-06-17T08:50:58+00:00","versionOfRecord":[],"versionCreatedAt":"2024-04-18 13:50:13","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4218805","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4218805","identity":"rs-4218805","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}