Detection of cell‐free DNA fragmentation and copy number alterations in cerebrospinal fluid from glioma patients

Abstract Glioma is difficult to detect or characterize using current liquid biopsy approaches. Detection of cell‐free tumor DNA (cftDNA) in cerebrospinal fluid (CSF) has been proposed as an alternative to detection in plasma. We used shallow whole‐genome sequencing (sWGS, at a coverage of < 0.4×) of cell‐free DNA from the CSF of 13 patients with primary glioma to determine somatic copy number alterations and DNA fragmentation patterns. This allowed us to determine the presence of cftDNA in CSF without any prior knowledge of point mutations present in the tumor. We also showed that the fragmentation pattern of cell‐free DNA in CSF is different from that in plasma. This low‐cost screening method provides information on the tumor genome and can be used to target those patients with high levels of cftDNA for further larger‐scale sequencing, such as by whole‐exome and whole‐genome sequencing.

copy number alterations (SCNAs) without prior knowledge of tumor mutations. In addition, a fragmentation pattern was identified that correlates with detection of SCNAs. This paper provides an improved low-cost screening method for patient samples. The manuscript is largely technically sound, but a few minor comments should be addressed as detailed below.
1. In the figure legend for Fig. 1C, the authors reference sWGS from plasma and urine samples, but this data is not included in the figure. Please include this sWGS data for comparison.
2. The authors demonstrate a relatively clear negative correlation between the amplitude of 10 bp periodic oscillations and the levels of SCNAs (Fig. 2C). The authors should comment on potential explanations for why this trend is observed. Mouliere et al conducted a study regarding cell-free DNA cerebrospinal fluid. The manuscript deals with a highly interesting topic since the analysis of cell-free tumor DNA is currently one the hottest topics in the field of oncology and data on brain tumors are still sparse. However, proximal sampling by the analysis of CSF might increase sensitivity. Overall speaking, the study technically sounds. The manuscript does contain some interesting data and they are on most parts well-presented, however there are some concerns that need to be addressed. 1) The title ". . . using short cfDNA" is somehow misleading, since short cfDNA is not specifically enriched with the presented methods.
2) The authors present data from quite a small cohort of 13 patients and in only 5 of them SCNA could be detected. The author should indicate further analysis approaches for those patients where no cftDNA was detected.
3) In general, the results& discussion section seems a bit minimalistic and could be presented in a more comprehensive way. 4) Moreover, there are some discrepancies regarding the analysis of tumor samples: Line 102: Here the author state that tumor material was availa-ble only from one patient, although two sentences before that the claim that the highest concentration of cfDNA corresponded to three patients with SCNAs in the tumor. Pretty much at the beginning it says that the levels of cftDNA in CSF were not directly correlated to the tumor volume, lateron the detection and confirmation of the SCNA in the CSF sample was influenced by the size of the tumor (in addition to the level of cfDNA, and the glioma grade).
5) The authors observed a shift toward shorter fragment sizes of cfDNA in CSF, which was previously reported to be associated with an enrichment of tumor-derived fragments. Moreover, a negative correlation between the amplitude of the 10 bp periodic oscillations and the levels of SCNAs in these CSF samples was reported. This is somehow contradictory and the author should furthers explain this observation or at least present a hypothesis for it. 6) Figure 2 shows different size distributions of plasma, urine, and CSF. The author should include these data in the results and discussion section. This is an interesting paper that describes a new strategy to detect copy number alterations in glioma via shallow whole-genome sequencing (sWGS) of cell-free DNA in cerebrospinal fluid (CSF). In addition to analyses of somatic copy number alterations (SCNA), the paper provides some insight in to the fragment length profiles of cell-free DNA in CSF. It reads well and the conclusions are generally well-supported. I am happy to recommend this paper for publication in EMBO Molecular Medicine, and think it will be of interest to the readership of this journal. I have a few comments/suggestions that should be readily addressable.
1. Sequencing data needs to be made available, if possible in an open-access repository.
2. On page 3, lines 90-95, a statement is made regarding the sensitivity of the approach taken in this study to detect glioma, a combination of sWGS and SCNA analysis, and methods reported in the literature. The authors compare the detection rate to numbers achieved with tumor-guided methods, and more expensive approaches. There are several issues with this performance comparison: first, there is no direct comparison made with the approaches reported previously, second, the detection rate (39%) is significantly lower than what was achieved by Wang et al, third, the detection rate is likely strongly dependent on the tumor type and volume, as is also clear from Fig.1B. The authors should qualify the statement on page 3, or perform additional experiments and conduct a formal performance comparison.
3. A rarefaction analysis (detection rate vs seq depth) would be helpful to gain insight in the relationship between the cost of the assay (as determined by the depth of sequencing) and the performance of the assay in detecting glioma. No SCNA signature is detected in 8/13 samples. Is this due to technical limitations, or are these tumor cases that do not display SCNA? Would 10x deeper sequencing uncover additional features? 4. On p5 lines 145-150, a novel fragmentation signature is reported that may provide an alternate way to detect the presence of tumor DNA in CSF. This is quite compelling. The authors should provide a few ideas for the origin of this signature (negative correlation between the amplitude of 10 bp oscillations in the distribution of fragment sizes and the levels of cfDNA). Is a similar feature observed for tumor DNA in other bodily fluids? The latter question can be addressed with an analysis of seq data for plasma DNA in other tumor settings available in public repositories, for example. in this manuscript that shallow whole-genome sequencing (sWGS) can be used with cell-free tumor DNA from cerebrospinal fluid (CSF) to detect somatic copy number alterations (SCNAs) without prior knowledge of tumor mutations. A In addition, a fragmentation pattern was identified that correlates with detection of SCNAs. This paper provides an improved low-cost screening method for patient samples. The manuscript is largely technically sound, but a few minor comments should be addressed as detailed below.
We thank the reviewer for their comments.
1. In the figure legend for Fig. 1C, the authors reference sWGS from plasma and urine samples, but this data is not included in the figure. Please include this sWGS data for comparison.
The sWGS data for the corresponding plasma and urine samples of this patient exhibited a copy number neutral profiles with no SCNAs, therefore we have not included them in the detailed Fig 1C. These SCNAs plots are now available in Suppl. Fig. 2.
2. The authors demonstrate a relatively clear negative correlation between the amplitude of 10 bp periodic oscillations and the levels of SCNAs (Fig. 2C). The authors should comment on potential explanations for why this trend is observed.
We have developed more this analysis in the text and commented in the discussion so that it now reads: "Thus an overall decrease in the peak fragment size was associated with a reduction in the amplitude of the sub-nucleosomal peaks. The origin of the 10 bp oscillatory pattern is believed to be due to variable accessibility of the DNA due to its winding around the histone cores (Jiang & Lo, 2016). Alternative nuclease activities in cancer or alternate mechanisms of DNA release may produce an alteration in this fragmentation pattern." 3. The data provided in Supplemental Fig. 2 are very informative and important for interpretation of the paper; thus, this figure should be included in the main text.
We thank the reviewer for this suggestion. This figure is now included in the text as Figure 2.
4. In Supplemental Fig. 2B, tumor size is missing units.
We have added the unit (mm) to the figure.
Referee #2 (Comments on Novelty/Model System for Author): Mouliere et al conducted a study regarding cell-free DNA cerebrospinal fluid. The manuscript deals with a highly interesting topic since the analysis of cell-free tumor DNA is currently one the hottest topics in the field of oncology and data on brain tumors are still sparse. However, proximal sampling by the analysis of CSF might increase sensitivity. Overall speaking, the study technically sounds. The manuscript does contain some interesting data and they are on most parts wellpresented, however there are some concerns that need to be addressed.
We thank the reviewer for their comments.
1) The title ". . ." using short cfDNA" is somehow misleading, since short cfDNA is not specifically enriched with the presented methods.

We have changed the title to "Detection of cell-free DNA fragmentation and copy number alterations in cerebrospinal fluid from glioma patients".
2) The authors present data from quite a small cohort of 13 patients and in only 5 of them SCNA could be detected. The author should indicate further analysis approaches for those patients where no cftDNA was detected. 3) In general, the results & discussion section seems a bit minimalistic and could be presented in a more comprehensive way.
We have now expanded some aspects of the results and discussion. We have added a new figure, and developed figure 3. We try to stick to the report format as well.
4) Moreover, there are some discrepancies regarding the analysis of tumor samples: Line 102: Here the author state that tumor material was available only from one patient, although two sentences before that the claim that the highest concentration of cfDNA corresponded to three patients with SCNAs in the tumor.
The study uses single region tissue samples from tumors in all cases. However we have multiple regions of tissue from within the same tumor for one patient (GB1). This is now corrected in the document, and specified in the figure 2 of the revised manuscript.
Pretty much at the beginning it says that the levels of cftDNA in CSF were not directly correlated to the tumor volume, later on the detection and confirmation of the SCNA in the CSF sample was influenced by the size of the tumor (in addition to the level of cfDNA, and the glioma grade).
The first part was referring to cfDNA and not cftDNA. This is now corrected in the text.
5) The authors observed a shift toward shorter fragment sizes of cfDNA in CSF, which was previously reported to be associated with an enrichment of tumor-derived fragments. Moreover, a negative correlation between the amplitude of the 10 bp periodic oscillations and the levels of SCNAs in these CSF samples was reported. This is somehow contradictory and the author should furthers explain this observation or at least present a hypothesis for it.
The shift towards shorter size was previously reported for plasma, however this has not been reported previously within the CSF. The reduction is the amplitude of the 10bp periodic oscillations is not contradictory with the global shortening as the 2 are distinct fragmentation features of the circulating DNA.
We have expanded figure 3 and added the following text in the manuscript to highlight this point: "Thus an overall decrease in the peak fragment size was associated with a reduction in the amplitude of the sub-nucleosomal peaks. The origin of the 10 bp oscillatory pattern is believed to be due to variable accessibility of the DNA due to its winding around the histone cores (Jiang & Lo, 2016). Alternative nuclease activities in cancer or alternate mechanisms of DNA release may produce an alteration in this fragmentation pattern." 6) Figure 2 shows different size distributions of plasma, urine, and CSF. The author should include these data in the results and discussion section.
The size distribution of plasma, urine and matched CSF samples were available only for patient GB1. This is now specified in the text.

Referee #3 (Remarks for Author):
This is an interesting paper that describes a new strategy to detect copy number alterations in glioma via shallow whole-genome sequencing (sWGS) of cell-free DNA in cerebrospinal fluid (CSF). In addition to analyses of somatic copy number alterations (SCNA), the paper provides some insight in to the fragment length profiles of cell-free DNA in CSF. It reads well and the conclusions are generally well-supported. I am happy to recommend this paper for publication in EMBO Molecular Medicine, and think it will be of interest to the readership of this journal. I have a few comments/suggestions that should be readily addressable.
We thank the reviewer for their comments. 4. On p5 lines 145-150, a novel fragmentation signature is reported that may provide an alternate way to detect the presence of tumor DNA in CSF. This is quite compelling. The authors should provide a few ideas for the origin of this signature (negative correlation between the amplitude of 10 bp oscillations in the distribution of fragment sizes and the levels of cfDNA). Is a similar feature observed for tumor DNA in other bodily fluids? The latter question can be addressed with an analysis of seq data for plasma DNA in other tumor settings available in public repositories, for example.
We have expanded on this analysis (figure 3) and added notably the following text to the manuscript: "Thus an overall decrease in the peak fragment size was associated with a reduction in the amplitude of the sub-nucleosomal peaks. The origin of the 10 bp oscillatory pattern is believed to be due to variable accessibility of the DNA due to its winding around the histone cores (Jiang & Lo, 2016 Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed report from the referee that was asked to re-assess it. As you will see, the reviewer is now supportive, and I am pleased to inform you that we will be able to accept your manuscript pending minor editorial amendments. ***** Reviewer's comments ***** Referee #2 (Remarks for Author): The authors adressed most of my concern! è è è è  common tests, such as t--test (please specify whether paired vs. unpaired), simple χ2 tests, Wilcoxon and Mann--Whitney tests, can be unambiguously identified by name only, but more complex techniques should be described in the methods section;  are tests one--sided or two--sided? Any descriptions too long for the figure legend should be included in the methods section and/or with the source data.
In the pink boxes below, please ensure that the answers to the following questions are reported in the manuscript itself. Every question should be answered. If the question is not relevant to your research, please write NA (non applicable). We encourage you to include a specific subsection in the methods section for statistics, reagents, animal models and human subjects.

definitions of statistical methods and measures:
a description of the sample collection allowing the reader to understand whether the samples represent technical or biological replicates (including how many animals, litters, cultures, etc.).
Please fill out these boxes ê (Do not worry if you cannot see all your text once you press return) a specification of the experimental system investigated (eg cell line, species name).

B--Statistics and general methods
the assay(s) and method(s) used to carry out the reported observations and measurements an explicit mention of the biological and chemical entity(ies) that are being measured. an explicit mention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner.

Data
the data were obtained and processed according to the field's best practice and are presented to reflect the results of the experiments in an accurate and unbiased manner. figure panels include only data points, measurements or observations that can be compared to each other in a scientifically meaningful way. graphs include clearly labeled error bars for independent experiments and sample sizes. Unless justified, error bars should not be shown for technical replicates. if n< 5, the individual data points from each experiment should be plotted and any statistical test employed should be justified the exact sample size (n) for each experimental group/condition, given as a number, not a range; Each figure caption should contain the following information, for each panel where they are relevant:

Captions
The data shown in figures should satisfy the following conditions: Source Data should be included to report the data underlying graphs. Please follow the guidelines set out in the author ship guidelines on Data Presentation.

G--Dual use research of concern
Sequencing data will be available at EGAS00001003255 Sequencing data will be available at ega_box_1048