UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Optimization of the detection of circulating DNA in pediatric solid tumor patients treated with granulocyte… Toledo, Thyrza May 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_february_toledo_thyrzamay.pdf [ 3.39MB ]
Metadata
JSON: 24-1.0362865.json
JSON-LD: 24-1.0362865-ld.json
RDF/XML (Pretty): 24-1.0362865-rdf.xml
RDF/JSON: 24-1.0362865-rdf.json
Turtle: 24-1.0362865-turtle.txt
N-Triples: 24-1.0362865-rdf-ntriples.txt
Original Record: 24-1.0362865-source.json
Full Text
24-1.0362865-fulltext.txt
Citation
24-1.0362865.ris

Full Text

OPTIMIZATION OF THE DETECTION OF CIRCULATING DNA IN PEDIATRIC SOLID TUMOR PATIENTS TREATED WITH GRANULOCYTE COLONY STIMULATING FACTOR by   Thyrza May Toledo  BMLSc, The University of British Columbia, 2015  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Pathology and Laboratory Medicine)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   December 2017  © Thyrza May Toledo, 2017 ii  Abstract Solid tumor patients are often administered with a drug called Granulocyte-Colony Stimulating Factor (G-CSF). This drug is used to treat chemotherapy-induced neutropenia and/or to mobilize hematopoietic stem cells from the bone marrow into the blood for ease of collection. Previous in vivo studies in mice showed, that G-CSF can increase tumor growth and promote metastasis. As such, this study investigated whether G-CSF can promote tumor growth in children with solid tumors by non-invasively quantifying tumor-derived circulating cell-free DNA (cfDNA) in the plasma. I also investigated whether tumor-derived cfDNA levels in the plasma correlate with tumor-derived genomic DNA (gDNA) levels in the stem cell product, on the day of stem cell collection procedure.  Tumor cfDNA was measured in the plasma of fourteen children with solid tumors, before and after G-CSF treatment, using methylation specific qPCR against the promoter region of the RASSF1a gene. Nine children [three rhabdomyosarcoma, five neuroblastoma and one rhabdoid tumor] had detectable tumor cfDNA in their plasma, which was suggestive of poor clinical outcome. In addition, paired plasma and stem cell products from the day of stem cell harvest were collected from nine children (four neuroblastoma, one glioblastoma, one rhabdoid tumor, two Hodgkin lymphoma and one choroid plexus carcinoma). At the time of stem cell harvest (post G-CSF administration), I found no evidence of tumor gDNA contamination in the stem cell product from all nine children. However, there was evidence of tumor cfDNA in the plasma product from six children, four of whom had known bone marrow involvement. Overall, four of the seven children with no detectable tumor cfDNA in the plasma had a diagnosis of brain cancer. iii  Altogether, tumor cfDNA levels in children with solid tumors can be detected but the current study did not show that levels increased upon G-CSF administration. The presence of tumor cfDNA in the plasma of children with pediatric cancer at the time of stem cell harvest was not consistent with tumor contamination of the stem cell product.   iv  Lay Summary  It is estimated that 1,440 Canadian children will receive a new cancer diagnosis in 2017. Currently, children with cancer are often treated with a drug called granulocyte colony stimulating factor (G-CSF) to prevent infections, a common side effect of cancer treatments, and to collect stem cells which are special cells that can help repopulate blood cells after chemotherapy. This study showed that free-floating cancer DNA can be detected in the blood of children with cancer. G-CSF treatment does not seem to increase the growth of cancers which was measured by the amount of free-floating cancer DNA in the blood. Overall, findings from this research suggest that free-floating cancer DNA is a promising way to detect if there is residual cancer present in the person simply by collecting a tube of blood.     v  Preface Human ethics and biosafety approval  for the project entitled “Circulating Cell-Free Tumor DNA Analysis in Children with Solid Tumors before and after Granulocyte Colony Stimulating Factor Administration” were obtained from the UBC Children’s and Women’s Health Centre of BC Research Ethics Board (H16-01468 and B16-0159 respectively).    Dr. S. Vercauteren and I identified a potential gap in the current knowledge about circulating cell-free DNA and G-CSF treatment. In collaboration with my co-supervisor, Dr. H. Cote and my supervisory committee members, Drs. G. Reid, P. Lange and R. Morin, Dr. S. Vercauteren and I designed this research project. The BC Children’s Hospital BioBank staff obtained written informed consent from the participants, collected the samples along with their associated clinical data, and de-identified all the samples. I performed the DNA extraction, bisulfite treatment and qPCR part of the study. In addition, I analyzed the results and performed the appropriate statistical analysis.  This is original work that has not been previously published.         vi  Table of Contents  Abstract .......................................................................................................................................... ii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi List of Tables ................................................................................................................................ xi List of Figures .............................................................................................................................. xii List of Abbreviations ................................................................................................................. xiv Acknowledgements .................................................................................................................... xvi Dedication ................................................................................................................................. xviii Chapter 1: Introduction ....................................................................................................................1 1.1 Pediatric solid tumors .............................................................................................................1 1.2 Circulating cell-free DNA ......................................................................................................1 1.3 Stem cell harvest ....................................................................................................................3 1.3.1 Granulocyte macrophage colony stimulating factor ........................................................5 1.3.2 Plerixafor .........................................................................................................................6 1.3.3 Granulocyte colony stimulating factor ............................................................................7 1.4 Chemotherapy-induced neutropenia.......................................................................................9 1.5 Rationale and research questions ...........................................................................................9 1.6 Objectives .............................................................................................................................10 1.7 Hypotheses ...........................................................................................................................10 Chapter 2: Optimization of assay conditions .................................................................................11 2.1 Comparison of total cfDNA yields between the QIAamp® and the Maxwell® kit ..............11 vii  2.1.1 Introduction ...................................................................................................................11 2.1.2 Methodology..................................................................................................................12 2.1.2.1 Blood collection ......................................................................................................12 2.1.2.2 Isolating cfDNA from plasma (Maxwell® kit) ........................................................13 2.1.2.3 Isolating cfDNA from plasma (QIAamp® kit) ........................................................15 2.1.2.4 Real-time PCR .........................................................................................................16 2.1.2.5 Quality assessment of total cfDNA .........................................................................17 2.1.2.6 Statistical analysis ...................................................................................................17 2.1.3 Results ...........................................................................................................................18 2.1.4 Discussion......................................................................................................................23 2.1.5 Conclusion .....................................................................................................................26 2.2 Analysis of total cfDNA yields in EDTA and citrate blood samples ...................................26 2.2.1 Introduction ...................................................................................................................26 2.2.2 Methodology..................................................................................................................26 2.2.2.1 Blood collection ......................................................................................................27 2.2.2.2 Isolating cfDNA from plasma .................................................................................27 2.2.3 Results ...........................................................................................................................27 2.2.4 Discussion......................................................................................................................28 2.2.5 Conclusion .....................................................................................................................29 2.3 Determination of the lower limit of detection of the methylation specific qPCR  (MS-qPCR) technique ........................................................................................................................29 2.3.1 Introduction ...................................................................................................................29 2.3.2 Methodology..................................................................................................................32 viii  2.3.2.1 Sources of samples ..................................................................................................32 2.3.2.2 Genomic DNA extraction ........................................................................................33 2.3.2.3 Quantification of gDNA ..........................................................................................34 2.3.2.4 Bisulfite treatment ...................................................................................................34 2.3.2.5 MS-qPCR ................................................................................................................35 2.3.3 Results ...........................................................................................................................36 2.3.4 Discussion......................................................................................................................37 2.3.5 Conclusion .....................................................................................................................38 Chapter 3: G-CSF administration in pediatric solid tumor patients ..............................................39 3.1 Assessment of the effects of G-CSF treatment on tumor growth ........................................39 3.1.1 Introduction ...................................................................................................................39 3.1.2 Methodology..................................................................................................................40 3.1.2.1 Consent and ethics ...................................................................................................40 3.1.2.2 Blood collection ......................................................................................................40 3.1.2.3 Isolating cfDNA from plasma .................................................................................40 3.1.2.4 Real-time PCR .........................................................................................................41 3.1.2.5 Bisulfite treatment of cfDNA samples ....................................................................41 3.1.2.6 Methylation specific real-time PCR ........................................................................42 3.1.2.7 Statistical analysis ...................................................................................................43 3.1.3 Results ...........................................................................................................................44 3.1.4 Discussion......................................................................................................................69 3.1.5 Conclusion .....................................................................................................................72 ix  3.2 Comparison of tumor cfDNA levels in the plasma and tumor gDNA levels in the stem cell product ........................................................................................................................................72 3.2.1 Introduction ...................................................................................................................72 3.2.2 Methodology..................................................................................................................74 3.2.2.1 Blood collection ......................................................................................................74 3.2.2.2 Isolating cfDNA from the plasma product ..............................................................75 3.2.2.3 Isolating gDNA from the stem cell product  ...........................................................75 3.2.2.4 Real-time PCR .........................................................................................................76 3.2.2.5 Bisulfite treatment of cfDNA from plasma and gDNA from stem cells .................76 3.2.2.6 Methylation specific real-time PCR ........................................................................76 3.2.3 Results ...........................................................................................................................77 3.2.4 Discussion......................................................................................................................83 3.2.5 Conclusion .....................................................................................................................85 Chapter 4: Overall conclusions ......................................................................................................86 4.1 Summary of research findings ..............................................................................................86 4.2 Strengths ...............................................................................................................................87 4.3 Research significance ...........................................................................................................87 4.4 Limitations ...........................................................................................................................88 4.5 Future research directions ....................................................................................................89 4.5.1 Sample collection ..........................................................................................................89 4.5.2 Quantification ................................................................................................................91 4.5.3 Total cfDNA ..................................................................................................................91 References .....................................................................................................................................92 x  Appendices ..................................................................................................................................102 Appendix A Primer and probe sequences for qPCR ................................................................102 Appendix B Primers used for PCR ..........................................................................................103 Appendix C Standard curve for the semi-quantitative analysis of gDNA contamination .......104 Appendix D Confirmation of promoter methylation status of the RASSF1a gene in the IMR32 cell line and WBCs ...................................................................................................................105       xi  List of Tables Table 2-1 Differences in the QIAamp® kit and Maxwell® kit ...................................................... 12 Table 2-2 Frequency of the RASSF1a promoter methylation in pediatric cancer patients ........... 30 Table 2-3 Serial dilutions of the IMR-32 cells in increasing amounts of WBCs ......................... 33 Table 3-1 Patient characteristics for the pre and post G-CSF study ............................................. 44 Table 3-2 Raw data for before and after G-CSF treatment comparison of cfDNA levels ............ 48 Table 3-3 Days between important clinical events and sample collection ................................... 49 Table 3-4 Pre and post G-CSF tumor cfDNA inter-assay coefficient of variation (CV) determinations ............................................................................................................................... 53 Table 3-5 Patient characteristics for the association study ........................................................... 78 Table 3-6 Total DNA and normal DNA concentrations from plasma and stem cell products from pediatric solid tumor patients ........................................................................................................ 80 Table 3-7 Tumor cfDNA yields in plasma and stem cell products............................................... 82  xii  List of Figures Figure 1-1 Hematopoietic stem cell lineage ................................................................................... 4 Figure 1-2 Proposed mechanism of action of GM-CSF ................................................................. 5 Figure 1-3 Proposed mechanism of action of Plerixafor ................................................................ 7 Figure 1-4 Proposed mechanism of action of G-CSF ..................................................................... 8 Figure 2-1 Promega's Maxwell® RSC cartridges and deck tray ................................................... 13 Figure 2-2 Promega Maxwell® RSC instrument with the loaded deck tray ................................. 14 Figure 2-3 Vacuum manifold of the Qiagen Circulating Nucleic Acid kit ................................... 16 Figure 2-4 Concentrations of PCR amplifiable DNA using three different elution protocol ....... 19 Figure 2-5 PCR amplifiable DNA yields using Qiagen and Promega kits ................................... 20 Figure 2-6 Agarose gel electrophoresis of the isolated cfDNA showing the presence of high molecular weight gDNA fragments .............................................................................................. 22 Figure 2-7 PCR amplifiable DNA yields from EDTA and citrate tubes ...................................... 28 Figure 2-8 Bisulfite treatment and MSP ....................................................................................... 31 Figure 2-9 Lower limit of detection of the MS-qPCR assay ........................................................ 37 Figure 3-1 Total cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients................................................................................................................................ 47 Figure 3-2 Normal cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients................................................................................................................................ 50 Figure 3-3 Tumor cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients................................................................................................................................ 51 Figure 3-4 Timeline of clinical treatment for patient 1................................................................. 55 Figure 3-5 Timeline of clinical treatment for patient 2................................................................. 56 xiii  Figure 3-6 Timeline of clinical treatment for patient 3................................................................. 57 Figure 3-7 Timeline of clinical treatment for patient 4................................................................. 58 Figure 3-8 Timeline of clinical treatment for patient 5................................................................. 59 Figure 3-9 Timeline of clinical treatment for patient 6................................................................. 60 Figure 3-10 Timeline of clinical treatment for patient 7............................................................... 61 Figure 3-11 Timeline of clinical treatment for patient 8............................................................... 62 Figure 3-12 Timeline of clinical treatment for patient 9............................................................... 63 Figure 3-13 Timeline of clinical treatment for patient 10............................................................. 64 Figure 3-14 Timeline of clinical treatment for patient 11............................................................. 65 Figure 3-15 Timeline of clinical treatment for patient 12............................................................. 66 Figure 3-16 Timeline of clinical treatment for patient 13............................................................. 67 Figure 3-17 Timeline of clinical treatment for patient 14............................................................. 68 Figure 3-18 Peripheral blood stem cell collection and transplant flowchart ................................ 73 Figure 3-19 Semi-quantitative gDNA co-isolation check for patient 13 and 17’s cfDNA sample from plasma products collected during stem cell harvest ............................................................. 81  xiv  List of Abbreviations ADC   Adenomatous polyposis coli  ACDA  Acid citrate dextrose solution A  BCCHB British Columbia Children’s Hospital BioBank CBC  Complete blood count  cfDNA Circulating cell-free DNA CT  Computed tomography scan COG5  Conserved oligomeric Golgi complex subunit 5 CXCR4 Chemokine receptor type 4 ddPCR  Digital droplet PCR DMSO  Dimethyl sulfoxide DNA  Deoxyribonucleic acid EDTA  Ethylenediaminetetraacetic acid EGFR  Epidermal growth factor receptor  EWS   Ewing Sarcoma RNA binding protein 1  FDA  Food and Drug Administration FLI1  Friend leukemia integration 1 transcription factor  G-CSF  Granulocyte colony stimulating factor GM-CSF  Granulocyte macrophage colony stimulating factor  gDNA  Genomic DNA GE  Genomic equivalents KRAS  Kirsten rat sarcoma LINE1  Long interspersed nuclear elements 1  xv  MRI  Magnetic resonance imaging PBS  Phosphate buffered saline PBSC  Peripheral blood stem cells PCR  Polymerase chain reaction PET/CT Combined positron emission tomography and computed tomography scan POLR2 RNA polymerase II gene Pro-K  Proteinase K qPCR  Real-time polymerase chain reaction RASSF1a Ras association domain family member 1 gene RNA  Ribonucleic acid RSC  Rapid sample concentrator RT-PCR Reverse transcriptase PCR SD  Standard deviation SDF-1  Stromal cell derived factor 1 SMARCB1  SWI/SNF-related matrix-associated actin-dependent regulator of chromatin subfamily B member 1  TLR3  Toll-like receptor 3 USA  United States of America WBC  White blood cell xvi  Acknowledgements First of all, I would like to express my deep gratitude to my supervisor, Dr. Suzanne Vercauteren, for her valuable mentorship and constant support throughout the completion of my research project and thesis. It has been an interesting and challenging two years but I have learned a lot and have grown more in confidence because of you. Thank you!  I would also like to thank my co-supervisor, Dr. Helene Cote and my supervisory committee members, Drs. Philipp Lange, Ryan Morin and Gregor Reid. With your expertise, insightful inputs and penetrating questions, I was able to deepen my knowledge about the field and complete this proof-of-principle project. My hope is that this project will provide the essential groundwork for researchers to build on for future studies in this field. I gratefully acknowledge the financial support from the Rare Disease Foundation in partnership with the BC Children’s Hospital Foundation through the Rare Disease Foundation microgrant. In addition, I would like to acknowledge the BC Children’s Hospital BioBank which was made possible by a donation from Mining for Miracles through the BC Children’s Hospital Foundation. Special thanks are owed to the “secret superheroes” of the BC Children’s Hospital BioBank - these are the patients and their families who have willingly donated their precious samples to the BioBank to foster research projects like mine. Also, my deep appreciation goes out to the BC Children’s Hospital BioBank staff (Nidhi Arora, Rumbi Chiwaya, Veronica Chow, Ashton Ellis, Stephen Fung, Thomas Soroski, Mandy Suen, Tamsin Tarling, Heather Van Tassel and Adam Velenosi) for their technical and emotional support. All your smiling faces and constant encouragement has helped push me through until the end. Also, I would like to give a heartfelt thank you to Thomas and Veronica, who took the time to proofread my thesis and offer helpful suggestions. xvii  I would like to express my sincere gratitude to all my collaborators. I would like to thank the BC Children’s Hospital Cellular Therapy team (Angela Hall and Elizabeth Lee) for helping me coordinate sample accrual. I would also like to thank the following individuals for their technical expertise, their support and for sharing their lab space and instruments: Emma Hitchcock, Katelin Townsend and Chi Kin Wong from the Gibson lab; Irina Manokhina and Maria Penaherrera from the Robinson lab; and David Chai from the Cytogenetics lab.  Finally, I would like to whole heartedly thank my wonderful parents, family and friends who provided their loving support and inspiration throughout my years of education. I could not have done it without you all!    I hope you all enjoy reading the fruitage of my two and a half years of hard work!     xviii  Dedication  I dedicate this thesis…  To my loving parents who have inspired and supported me throughout my life. Thank you for all the sacrifices that you have made so that I could have a better life and education. I am forever grateful for you.  I love you!   To my best friend, Desiree Baniaga, whose unconditional love helps me to strive to be the best that I can be.  1  Chapter 1: Introduction 1.1 Pediatric solid tumors  It is estimated that 206,200 Canadians will be diagnosed with cancer in 2017, 1,440 of which will be of children and youths under 19 years of age.1 Typically, cancer can be divided into two broad categories: solid and liquid cancer.2 Solid cancers, which make up greater than 80% of all cancer types, refer to tumors that primarily originate in solid organs, while liquid cancers refers to tumors that originate from the blood and disseminate to other parts of the body.2 Children diagnosed with solid cancers undergo invasive procedures for diagnosis and monitoring, such as tissue and bone marrow biopsies. The latter is performed to assess tumor metastasis into the bone marrow. The standard treatment options for these children are surgery, chemotherapy, and radiation therapy, followed by bone marrow rescue using autologous stem cells when clinically indicated.3,4  1.2 Circulating cell-free DNA  Circulating cell-free DNA (cfDNA) refers to the extracellular nucleic acid species that is found in the liquid portion of the blood such as plasma or serum. In a given individual, cfDNA can be composed of normal cfDNA from non-malignant cells, fetal cfDNA from fetal cells and/or tumor-derived cfDNA from tumor cells. cfDNA release into the blood is thought to be due to apoptosis, necrosis or active secretion.5,6  cfDNA released via apoptosis is often observed as whole number multiples of approximately 180 bp fragments which represents the DNA bound to one or more nucleosomes.5,7 However, large molecules of DNA (>1000 bp) has also been observed in plasma which are thought to be cfDNA which were released via necrosis.5,7 However, unlike the larger cfDNA molecules, it is the smaller molecules of cfDNA that are of 2  interest because they often contain tumor-derived aberrations.8,9 Alternatively, the high molecular weight DNA can represent intra-cellular genomic DNA (gDNA) contamination that are released due to mechanical white blood cell (WBC) lysis or during lysis that occurs when blood is allowed to clot.10 cfDNA is useful in cancer studies because it provides insight as to what is happening in an individual. Thus, it is important to minimize the amount of intra-cellular WBC gDNA contamination because it will likely consist of wild-type DNA. For instance, serum appears to be the better sample type to study cfDNA because cfDNA levels are 6-24 times higher in serum samples than in plasma samples.10 However, cfDNA isolated from serum is often contaminated with intra-cellular gDNA from lysed WBC.10 Consequently, plasma samples are preferred when studying cfDNA.  In vivo studies in mice and in humans have shown that tumor cfDNA levels are associated with tumor growth and treatment response.11–13 cfDNA is also a promising biomarker for assessing tumor burden due to the non-invasive nature of sample collection. As a result, multiple sample collections throughout a patient’s treatment and follow up can be easily obtained and analyzed. Aside from quantifying cfDNA, analysis of genetic and epigenetic alterations has also been performed in cfDNA. Sequencing has been successfully used to identify gene mutations and deletions in cfDNA from patients with cancer.14–16 Likewise, methylation specific polymerase chain reaction (MS-PCR) has been used to detect aberrant promoter methylation of tumor suppressor genes in cancer patients. 17–19 The genetic and epigenetic aberrations are confirmed to be of tumor origin through analysis of DNA from primary tumor tissues.14–16 Even though the majority of research using cfDNA is performed on adult cancer, detection and analysis of cfDNA was also shown feasible in pediatric malignancies including solid tumors such as atypical teratoid rhabdoid brain tumor, Hodgkin lymphoma, and neuroblastoma.20–22 In 3  general, healthy children have low amounts of total cfDNA circulating in the blood with concentrations ranging from 0 to 2000 genomic equivalents (GE)/mL of plasma.21,23 In contrast, children with cancer can have 2 to 14 times higher total cfDNA levels.21,23 Levels of total cfDNA also correlate well with disease recurrence.23 Previous studies have shown that there is strong concordance between the mutations found in the tumor tissue and the tumor cfDNA samples.24 For these reasons, total and tumor cfDNA are potentially useful tools for studying cancer.   1.3 Stem cell harvest The most common method for collecting stem cells is called peripheral blood stem cell collection (PBSC) procedure. For this procedure, chemotherapy drugs with or without cytokines such as granulocyte colony stimulating factor (G-CSF), granulocyte macrophage colony stimulating factor (GM-CSF) and/or Plerixafor are used to stimulate the patient’s bone marrow to produce hematopoietic stem cells and mobilize them into the bloodstream, where they could be easily collected.3 In order to ensure that sufficient stem cell yield of the harvested product is reached, mobilized stem cells in the blood are counted using flow cytometry against the CD34 glycoprotein.3,25 After adequate amounts of stem cells are harvested, the patient will then undergo myeloablative chemotherapy and subsequent transplant of the harvested CD34+ hematopoietic stem cells which can then re-populate the bone marrow and give rise to different blood cell types (see Figure 1-1).25,26 4     Figure 1-1 Hematopoietic stem cell lineage Outline of the hematopoietic stem cell lineage with common cell surface antigens. The stem cell products counted using antibodies against the CD34+ antigen likely contains the following cell types: hematopoietic stem cells (HSC), multipotent progenitors (MPP), common lymphoid progenitors (CLP), common myeloid progenitors (CMP), granulocyte-macrophage progenitors (GMP) and/or megakaryocte/erythroid progenitors (MEP). Image provided courtesy of Abcam. Image copyright © 2017265   1.3.1 Granulocyte macrophage colony stimulating factor GM-CSF is a glycoprotein and recombinant forms of it are clinically used as a mobilizing agent for stem cell collection procedures. GM-CSF binds to its receptor on hematopoietic stem cells which subsequently activates various downstream signaling pathways. As Figure 1-2 shows, the end result is the initiation of hematopoietic stem cell survival, proliferation and/or differentiation.27 One disadvantage of the use of GM-CSF with chemotherapy is that it is less efficacious in mobilizing CD34+ stem cells into the blood compared to using G-CSF with chemotherapy.28  Figure 1-2 Proposed mechanism of action of GM-CSF A diagram outlining how granulocyte macrophage colony stimulating factor (GM-CSF) affects hematopoietic stem cells. GM-CSF binds to hematopoietic stem cells through the GM-CSF receptor which ultimately causes the hematopoietic stem cells to survive, proliferate and/or differentiate.    6  1.3.2 Plerixafor Plerixafor (also known as AMD3100 or Mozobil®) is another type of mobilizing agent that is approved by the Food and Drug Administration (FDA) for use during stem cell harvests.3 Plerixafor is effective in that it gives higher stem cell yield and requires fewer days for mobilization.3 The major disadvantage of this mobilizing agent is its high cost; the estimated costs of a 4 day daily injection with Plerixafor versus an 8 day daily injection with G-CSF for a 70 kg individual are $30,220 and $3,770 respectively.29 As Figure 1-3 shows, Plerixafor is thought to work by directly binding to the chemokine receptor type 4 (CXCR4) on the stem cell’s surface thereby preventing the binding of the chemokine called stromal cell derived factor 1 (SDF-1) on the bone marrow stromal cell’s surface.30,31 As a result, instead of staying in the bone marrow, the stem cells are free to circulate into the bloodstream where they could be easily collected.   7   Figure 1-3 Proposed mechanism of action of Plerixafor Mechanism at which Plerixafor mobilizes stem cells from the bone marrow to the blood. Plerixafor binds directly to the chemokine receptor type 4 (CXCR4) on the hematopoietic stem cell’s surface thereby preventing the chemokine called stromal cell derived factor 1 (SDF-1) on the bone marrow stromal cell’s surface. This allows the hematopoietic stem cell to leave the bone marrow niche and enter the circulation for ease of collection. Image reproduced from Fricker et al. with permission from Karger Publishers30    1.3.3 Granulocyte colony stimulating factor G-CSF is a third example of a mobilizing agent that is routinely administered for releasing hematopoietic stem cells into circulation. Even though the exact mechanism of action of G-CSF is not fully understood, analogous to Plerixafor, G-CSF is proposed to affect the SDF-1/CXCR4 axis. For instance, G-CSF is shown to increase the expression of CXCR4 on the stem cells’ surface and decrease the bone marrow stromal cells’ expression of SDF-1.32 However, the latter was thought to be primarily due to the degradation of SDF-1 by the elastase enzyme that is secreted by neutrophils.32 Furthermore, as Figure 1-4 illustrates, it is proposed that G-CSF has an indirect effect on SDF-1 and CXCR4 expressions, by first acting on the bone marrow neural 8  cells, thereby stimulating them to increase their production of catecholamines.33 These catecholamines are thought to increase the expression of CXCR4 on the stem cells’ surface.33 In addition, the catecholamines act on bone marrow stromal cells such as osteoblasts to increase their secretion of SDF-1 into the blood.33 Since SDF-1 functions as a chemokine, this allows hematopoietic stem cells expressing the chemokine receptor, CXCR4, to migrate from the area of low chemokine concentration (the bone marrow) into the area of higher chemokine concentration (bloodstream). Once in the bloodstream, the stem cells can be easily collected.   Figure 1-4 Proposed mechanism of action of G-CSF Schematic representation of how granulocyte colony stimulating factor (G-CSF) mobilizes stem cells from the bone marrow to the blood. It is thought that G-CSF acts on neural cells in the bone marrow causing them to release catecholamines which in turn increases the expression of the chemokine receptor type 4 (CXCR4) on the hematopoietic stem cell’s surface (HSC). The catecholamines are also proposed to increase the secretion of the stromal cell derived factor (SDF1) by osteoblasts into circulation. The SDF1 in circulation acts as a chemokine and thereby recruit HSCs with the chemokine receptor, CXCR4, from the bone marrow into the blood. Image reproduced from Saba et al. with permission from the Taylor & Francis Group33   9  1.4 Chemotherapy-induced neutropenia Myelosuppression is a common side effect of chemotherapy regimens which places patients at risk for infections.34 It was observed that when cancer patients with chemotherapy-induced neutropenia are administered mobilizing drugs like G-CSF or GM-CSF, the patients’ WBC levels are able to effectively go back to normal, thereby reducing the number of days that the patient is neutropenic.34,35  Out of the three aforementioned mobilizing drugs, G-CSF is most commonly used for treating chemotherapy-induced neutropenia due to its effectiveness. For instance, in one randomized control trial that compared the combination of chemotherapy with either G-CSF or GM-CSF, neutrophil counts returned to the normal range faster with G-CSF, requiring only 13 days compared to GM-CSF which required 16 days.36 This is advantageous because it contributes to decreased hospital costs and to an earlier implementation of the patient’s treatment plan.  1.5 Rationale and research questions This research was performed to explore whether G-CSF administration stimulates tumor growth. This is of concern because a significant proportion of children with solid tumors receive G-CSF treatment either to treat neutropenia or prior to stem cell harvest. In addition, this exploratory study investigates whether tumor cfDNA of patients with solid tumors is a marker for the presence of tumor cell contamination in the stem cell product collected after G-CSF stimulation, as this could result in tumor progression and patient relapse following transplantation.  10  1.6 Objectives This research project is divided into the following five objectives: 1. To compare the cfDNA extraction efficiency of Promega’s Maxwell® RSC circulating cell-free DNA plasma kit to Qiagen’s QIAamp® Circulating Nucleic Acid kit. 2. To compare the cfDNA levels in the plasma of healthy volunteers whose blood sample was collected in ethylenediaminetetraacetic acid (EDTA) and citrate tubes. 3. To determine the lower limit of detection of tumor cells diluted in normal WBCs using a real-time polymerase chain reaction (qPCR) method which determines methylated versus non methylated RASSF1a gene. 4. To determine the level of tumor cfDNA before and after G-CSF stimulation in the plasma of pediatric solid tumor patients. 5. To determine whether the level of tumor cfDNA in the plasma is associated with the level of genetic tumor DNA in the stem cell product and whether cfDNA levels are associated with bone marrow disease status.   1.7 Hypotheses The following hypotheses were tested:  G-CSF administration is associated with increased levels of tumor cfDNA in the plasma of pediatric solid tumor patients  After G-CSF stimulation, the level of genomic tumor DNA in the stem cell product correlates with the level of tumor cfDNA in the plasma product   Tumor cfDNA levels are associated with bone marrow disease status 11  Chapter 2: Optimization of assay conditions 2.1 Comparison of total cfDNA yields between the QIAamp® and the Maxwell® kit 2.1.1 Introduction There are numerous commercially available isolation kits for extracting cfDNA from plasma or serum samples. Currently, in the literature, the most frequently used commercial kit is the Qiagen’s QIAamp® Circulating Nucleic Acid kit. Recently, Promega developed a new method for isolating total cfDNA from plasma using the automated Maxwell® Rapid Sample Concentrator (RSC) instrument. Total cfDNA isolation using the Maxwell® RSC instrument seems advantageous because up to sixteen samples can be processed simultaneously with minimal sample preparation required.37 The main differences between the two kits are shown in Table 2-1. Since maximal total cfDNA yield is desired, especially when using small sample volumes as is often the case for pediatric patients, the total cfDNA yields from plasma using the QIAamp® kit and the Maxwell® kit will be compared.          12  Table 2-1 Differences in the QIAamp® kit and Maxwell® kit QIAamp® Circulating Nucleic Acid kit38 Maxwell® RSC ccfDNA plasma kit37 Manual extraction Automated extraction Silica based column extraction Cellulose based extraction Uses vacuum protocol Uses magnetic beads Process 24 samples simultaneously Process 16 samples simultaneously 120 minutes 80 minutes Use ≤ 5000 µL plasma Use 200 µL – 1000 µL plasma Requires Proteinase K digestion No Proteinase K digestion Uses carrier RNA No carrier RNA 20-150 µL elution buffer 60 µL elution buffer  2.1.2 Methodology 2.1.2.1 Blood collection Fresh blood samples were collected from five healthy adult volunteers, in EDTA tubes. The mean age of the volunteers was 28.8 years old (range: 19 to 41 years old) and the male to female ratio was 2:3. Within one hour of blood collection, the blood was spun at 1500g for 10 minutes at 22°C for plasma separation. Immediately afterwards, a second spin of the plasma using the Eppendorf microcentrifuge 5424 R (Hamburg, Germany) at 16000g for 10 minutes at 22°C was performed. Plasma was then carefully collected and stored at -80°C until total cfDNA extractions were performed.  13  2.1.2.2 Isolating cfDNA from plasma (Maxwell® kit) For each volunteer, 500 µL of plasma was transferred into two centrifuge tubes containing 500 µL of phosphate buffered saline (PBS), to make a total volume of 1000 µL each. Total cfDNA extractions were performed from these two aliquots using two methods. One of the methods involved the automated Maxwell® RSC instrument, and its associated ccfDNA plasma AS1480 kit from Promega using the protocol in the TM454 technical manual.37 As per Figure 2-1, the ccfDNA plasma AS1480 kit cartridges were placed on the deck tray and elution tubes containing 60 µL of elution buffer were placed in the elution tube positions of the deck tray. Plungers were then placed into well #8 of each cartridge and the 1000 µL plasma-PBS mixtures were directly added to well #1 of the cartridges which contain binding buffer.    Figure 2-1 Promega's Maxwell® RSC cartridges and deck tray Image showing the Promega Maxwell® Rapid Sample Concentrator (RSC) robot’s black deck tray with elution tubes in the elution tubes position as well as kit cartridges that contain 8 wells. Well #1 contain binding buffer, well#2 contain the magnetic beads, well#3-7 contain wash buffers and well#8 contain the plunger.    Well # 1 – add sample Deck tray Elution tube with elution buffer Well # 8 – plunger 14  Subsequently, as per Figure 2-2, the deck tray was loaded into the Maxwell® RSC instrument and the ccfDNA plasma AS1480 extraction run was started, which took approximately 70 minutes.    Figure 2-2 Promega Maxwell® RSC instrument with the loaded deck tray Image showing the Promega Maxwell® Rapid Sample Concentrator (RSC) robot with loaded cartridges in the deck tray.   15  2.1.2.3 Isolating cfDNA from plasma (QIAamp® kit) Unlike the Maxwell® kit, the QIAamp® kit uses manual elution of the total cfDNA bound in the silica membrane. Although the standard protocol suggests doing the elution of the total cfDNA using a one step elution, eluting twice either with the same buffer or with fresh buffer can be considered if maximum yield of the total cfDNA is required. Thus, before the total cfDNA isolations using the QIAamp® kit for the kit comparison part of this study were performed, the elution method that gives the highest total cfDNA yields using the QIAamp® kit was determined using 500 µL of plasma from five healthy volunteers. The following three elution methods were compared: elute once with 60 µL, elute twice with the same 60 µL eluate and elute twice with two subsequent 30 µL elution buffers. The total cfDNA isolation using the QIAamp® kit was performed using the protocol on pages 22-25 of the kit handbook.38 100 µL of Proteinase K (Pro-K) and 800 µL of lysis buffer containing 1.0 µg of carrier ribonucleic acid (RNA) were added to the 1000 µL plasma-PBS mixture and was incubated at 60°C for half an hour. After incubation, 1800 µL of binding buffer was added to the mixture. As per Figure 2-3, the mixture was then loaded onto the spin column and pulled through the silica membrane of the spin column using vacuum pressure. The bound total cfDNA on the spin columns were then washed twice with wash buffers and then with 750 µL of 100% ethanol to remove residual impurities. Then, the spin columns were dried on a 56°C heat block for 10 minutes to evaporate excess ethanol. Finally, the bound total cfDNA was eluted by adding 60 µL of elution buffer to the spin columns, incubating for 3 minutes and spinning at 21130g for 1 minute.  The eluted total cfDNA samples were stored at -20°C until they were quantified by qPCR.   16   Figure 2-3 Vacuum manifold of the Qiagen Circulating Nucleic Acid kit This image shows Qiagen’s QIAamp Circulating Nucleic Acid kit’s vacuum manifold with tube extenders to handle large volumes of solutions, silica membrane columns, Vacvalves for controlling sample flow rates and tubings/connectors for connecting into the vacuum pump.   2.1.2.4 Real-time PCR Frozen total cfDNA extracts were thawed out and quantified by qPCR using the 7500 Fast Real-Time PCR instrument and software v.2.3 from Applied Biosystems® (Massachusetts, USA). The thermal cycling condition used was 95°C for 3 minutes followed by 45 cycles of 95°C for 5 seconds and 60°C for 30 seconds. Each reaction contained 12.5 µL of the PrimeTime® Gene Expression Master Mix from Integrated DNA Technologies (California, USA), 1.5 µL of the forward and reverse primers (300 nmol/L concentration each), 1.25 µL of the TaqMan probe for the RNA polymerase II gene (POLR2) at a concentration of 200 nmol/L, and 8.2 µL of cfDNA. The primer pair and probe sequences were obtained from Mussolin et al.39 and are found in Appendix A. A standard curve was prepared using a 10-fold serial dilution of the PowerQuant™ human male gDNA standard from Promega that was first diluted to a Spin column Vacuum pressure in Add sample mixture (plasma, PBS, Pro-K, lysis buffer with carrier RNA, binding buffer) Vacvalves 17  concentration of 758 GE/µL using the PowerQuant™ dilution buffer. For each qPCR run, the standard and unknown samples were quantified in triplicate. A no-template control composed of nuclease free water was included in every run. The intra-assay variation of the qPCR was evaluated with pooled cfDNA samples tested 10 times in the same qPCR run and determined to be 10%. The inter-assay variation of the qPCR was evaluated with pooled cfDNA samples tested in each qPCR run for 10 different runs and it was determined to be 23%.   2.1.2.5 Quality assessment of total cfDNA PCR was performed on the isolated total cfDNA samples using the Veriti 96-Well thermal cycler from Applied Biosystems (California, USA). The thermal cycling condition used was 95°C for 3 minutes followed by 45 cycles of 95°C for 5 seconds and 60°C for 60 seconds. Two sets of primers were used. One primer was used to target the 94 bp amplicon of the Conserved oligomeric Golgi complex subunit 5 (COG5), used to later visualize the cfDNA in the extracted samples. Another primer was used to target the 944 bp fragment of the Toll-like receptor 3 (TLR3) used to later visualize the presence of the high molecular weight gDNA contamination in the samples. The primer sequences were obtained from Dr. R. Morin (Simon Fraser University) and are found in Appendix B. Visual analysis of the isolated total cfDNA was performed through gel electrophoresis on a 1% agarose gel run at 100 volts for 60 minutes.  2.1.2.6 Statistical analysis Total cfDNA concentrations were expressed in terms of GE per ml of plasma. In order to convert DNA concentrations from ng/ml of plasma into GE/ml of plasma, it was assumed that there are 6.6 pg of DNA for each diploid cell.40 Data was expressed as the average ± standard error of the mean of the three independent replicates. The program used for statistical analysis was the GraphPad Prism software from GraphPad Prism Incorporation (La Jolla, California, USA). 18  Due to our small sample size, the Friedman’s test and the Wilcoxon signed rank test which are both non-parametric tests were performed to compare total cfDNA levels. The Friedman’s test is for comparing more than two paired groups while the Wilcoxon signed rank test is for comparing two paired groups. P-values ≤ 0.05 were considered statistically significant.  2.1.3 Results To ensure that the maximum cfDNA yields can be obtained from patient samples, the total cfDNA yields from the Maxwell® kit were compared with the total cfDNA yields from the QIAamp® kit using plasma samples from five healthy adult volunteers. The elution method that gives the highest cfDNA yield using the QIAamp® kit was first determined. Figure 2-4 shows the total cfDNA concentrations obtained when three elution protocols were compared using the QIAamp® kit: eluting once with 60 µL, eluting twice with the same 60 µL eluate and eluting two times with new 30 µL elution buffer. Eluting twice with the same 60 µL eluate appeared to give a slightly higher mean total cfDNA yield of 2155 ± 526 GE/ml of plasma compared to eluting once with 60 µL which had a mean total cfDNA yield of 2005 ± 563 GE/ml of plasma, or eluting two times with new 30 µL elution buffer which gave a mean total cfDNA yield of 1910 ± 618 GE/ml of plasma. Though the comparisons are not statistically significant (p=0.093), I chose to continue with the elution method that involves eluting twice with the same 60 µL eluate because it gave slightly higher total cfDNA yields. 19   Figure 2-4 Concentrations of PCR amplifiable DNA using three different elution protocol  cfDNA isolation from plasma samples of five healthy adult volunteers using the Qiagen QIAamp® Circulating Nucleic acid kit using three different types of elution protocol. Isolated cfDNA was quantified via qPCR targeting the RNA polymerase II (POLR2) gene without prior PCR amplification. The figure shows the mean of the cfDNA concentrations from each individual with three independent replicates along with their associated standard deviations. The line represents the overall mean for the five volunteers. The statistical comparative test performed was the Friedman’s test (p=0.093).  Next, the two types of cfDNA isolated kits were compared in terms of their total cfDNA yields. As demonstrated in Figure 2-5, the total cfDNA yields from the QIAamp® kit and the Maxwell® kit were 1580 ± 484 and 848 ± 281 GE/ml of plasma respectively. There was no statistically significant difference between the two kits (p=0.063).  20   Figure 2-5 PCR amplifiable DNA yields using Qiagen and Promega kits cfDNA was isolated from plasma samples from five healthy adult volunteers using the Qiagen QIAamp® Circulating Nucleic acid kit and Promega Maxwell® RSC ccfDNA plasma kit. Isolated cfDNA was quantified via qPCR targeting the RNA polymerase II (POLR2) gene with no prior PCR amplification. The figure shows the mean of the cfDNA concentrations from each individual with four independent replicates along with their associated standard deviations. The line represents the overall mean for the five volunteers. The statistical comparative test performed was the Wilcoxon signed rank test.  The intra-assay variation of the total cfDNA extraction using the Maxwell® kit was evaluated eight times with the same batch of pooled plasma sample and it was determined to be 11%. The inter-assay variation of the total cfDNA extraction using the Maxwell® kit was 21  evaluated with pooled plasma samples tested in each run on six different days and it was determined to be 23%. On the other hand, the intra-assay variation of the total cfDNA extraction using the QIAamp® kit was evaluated eight times with the same batch of pooled plasma sample and it was determined to be 7%. The inter-assay variation of the total cfDNA extraction using the QIAamp® kit was evaluated with pooled plasma samples tested in each run on six different days and it was determined to be 27%.   As Figure 2-6 show, high molecular weight gDNA contamination can be observed in all of the isolated cfDNA regardless of the kit used. A semi-quantitative estimate of the amount of gDNA contamination was performed using standards of known concentrations found in Appendix C. It seems that about 6-29 GE of gDNA per ml of plasma is present in the cfDNA isolated using the Maxwell® kit while about 6-727 GE of gDNA per ml of plasma is present in the cfDNA isolated using the QIAamp® kit. Therefore, it seems that there is more gDNA contamination in the cfDNA isolated using the QIAamp® kit.  22   Figure 2-6 Agarose gel electrophoresis of the isolated cfDNA showing the presence of high molecular weight gDNA fragments cfDNA samples isolated using different kits (Qiagen and Maxwell kits), cfDNA samples from different blood anticoagulants (ethylenediaminetetraacetic acid or EDTA and citrate), gDNA from white blood cells (WBC) and gDNA from the IMR-32 Neuroblastoma cells were amplified by PCR using the conserved oligomeric golgi complex subunit 5 (COG5) and toll-like receptor 3 (TLR3) primers. The PCR products were then visualized in a 1% agarose gel which was run at 100V for 60 minutes with the 94bp and 944bp bands representing cfDNA and high molecular weight gDNA respectively.23  2.1.4 Discussion Unlike gDNA which are long molecules of DNA found inside cells such as white blood cells, cfDNA are generally short fragments of DNA that are found outside the cells and are freely circulating in the blood. When a cell lyses, the intra-cellular gDNA molecules are released in the blood and may dilute the total cfDNA molecules which are found in smaller quantities. For this reason, it is important to choose the isolation method that gives the highest amount of and purest total cfDNA. There are numerous in-house methods as well as commercial kits available for isolating total cfDNA from plasma samples. The major disadvantage of in-house methods is that they often require effort and time to optimize in order to obtain high yields and reduce run-to-run variability. In terms of kits available commercially, Qiagen’s QIAamp® Circulating Nucleic Acid kit is the most frequently used because it gives high total cfDNA yields with a variability of 10% and does not inhibit downstream qPCR analysis.41 However, this kit is expensive, has a lengthy isolation protocol (120 minutes), and requires the use of a costly vacuum pump.38 Recently, an automated cfDNA isolation kit from Promega became commercially available. The Maxwell® RSC ccfDNA plasma kit allows the processing of up to 16 samples in only 80 minutes, with fewer sample preparations steps.37 Furthermore, the automated Maxwell® instrument lowers the possibility of human error, which should reduce the variation between runs. Thus far, there have only been two studies that directly compared these two kits in terms of total and tumor cfDNA levels. Pérez-Barrios et al. (2016) and Sorber et al. (2017) both recently published reports showing similar total and tumor cfDNA levels between the Maxwell® and the QIAamp® kits.7,42 Unlike our study where we looked at total cfDNA levels from healthy individuals, Pérez-Barrios and her research team used plasma samples from thirty-two stage III-IV lung cancer patients and one colon cancer patient.7 They observed that the Maxwell® kit gave a median total cfDNA yield 24  of 9470 GE/ml of plasma which was not statistically different from the QIAamp® kit which gave a median total cfDNA yield of 8182 GE/ml of plasma.7 Furthermore, they found that in the four patients analyzed further, the frequencies of the tumor cfDNA with the epidermal growth factor receptor (EGFR) mutation was not different between the QIAamp® kit and the Maxwell® kit.7 Similarly, Sorber et al. used plasma samples from cancer patients, specifically nine pancreatic cancer patients.42  They found that the QIAamp® kit which gave a mean total cfDNA yield of 2825 GE/ml of plasma was not statistically different to the Maxwell® kit which gave a mean total cfDNA yield of 3129 GE/ml of plasma.42 In addition, the tumor cfDNA levels measured by the kirsten rat sarcoma (KRAS)  mutations in the cfDNA were comparable between QIAamp® and Maxwell® kit.42 The findings from both studies are in agreement with our observation that there is no statistically significant difference in the total cfDNA yields obtained from the QIAamp® and the Maxwell® kit. However, I elected to choose the QIAamp® kit for subsequent experiments because it gave slightly higher yields though not statistically different from the Maxwell® kit. Since the p-value was insignificant, it may indicate that we were underpowered to detect significance given the small number of replicates in our study. Unfortunately, upon qualitative assessment of the isolated total cfDNA, it appears that high molecular weight gDNA contamination is present in all the isolated cfDNA. The gDNA contamination occurred at lesser amounts in the samples isolated using the Maxwell® kit compared to the QIAamp® kit, suggesting the former allows the isolation of purer cfDNA. The presence of gDNA contamination in our cfDNA likely accounts for the non-significant differences that I observed in the total cfDNA yields between the two kits. Though Sorber et al. did not clearly outline what spin protocol or how long the blood samples spent at room temperature before processing for plasma, the authors acknowledged that two out of their twenty samples had substantially more 25  total cfDNA which likely was due to gDNA contamination from the buffy coat layer.42 Likewise, Pérez-Barrios et al., who analyzed the isolated cfDNA samples using a Bioanalyzer, observed a high molecular weight DNA (>10000 bp) in 69% of their samples.7 In our study, we processed the blood sample immediately after collection and we performed a double spin protocol to obtain plasma. Of note, even with these precautions, we still observed cellular gDNA contamination. The cellular gDNA is likely derived from normal WBCs. One possible explanation is the disturbance of the buffy coat layer containing the WBCs upon aspiration of the plasma. Alternatively, the double spin protocol implemented in this study is insufficient in removing residual WBCs from the plasma samples. These contaminating WBCs could subsequently lyse thereby releasing their gDNA content. Since the commercial isolation kits cannot differentiate between high molecular weight gDNA from WBCs and the low molecular weight cfDNA of interest, the eluted cfDNA will be contaminated with cellular gDNA. In the future, filtering of the plasma before cfDNA extraction could be performed to efficiently remove residual cells. Further work is needed to accurately quantify and compare how much gDNA contamination is present in the cfDNA samples isolated from both kits.  We also observed that both the intra-assay and the inter assay-variations of the isolated cfDNA samples were similar between the two kits. Notably, the inter-assay variations of the Maxwell® kit and the QIAamp® kit were 2x and 3x higher than their corresponding intra-assay variations. Therefore, for any subsequent total cfDNA extractions, we made sure to extract the paired plasma samples (pre and post G-CSF) on the same day to minimize the sample processing variation. 26  2.1.5 Conclusion  We did not find significant differences in the total cfDNA yield between Qiagen’s QIAamp® Circulating Nucleic Acid kit and Promega’s Maxwell® RSC ccfDNA plasma kit. Also, high molecular weight gDNA was co-isolated with the cfDNA of interest though to a lesser degree with the Maxwell® kit. Since we only analyzed a small number of replicates in this study, a larger sized cohort would be of interest to clarify these findings.   2.2 Analysis of total cfDNA yields in EDTA and citrate blood samples  2.2.1 Introduction Blood can be collected in different kinds of anticoagulants. Routinely, if a patient simply requires a complete blood count, his/her blood sample will be collected in an EDTA tube. However, when a patient’s blood is being processed using an apheresis machine during stem cell collection, the anticoagulant typically used is acid citrate dextrose solution A (ACDA). Lam et al. detected comparable cfDNA levels from healthy volunteers in EDTA, citrate and heparin blood tubes after 0, 2 and 6 hours of collection.43 However, in this study, it is important to make pre and post G-CSF comparisons of cfDNA levels without the concern that any significant changes are actually due to the difference in anticoagulants used. As such, total cfDNA levels from blood tubes with different anticoagulants will be verified in this study.  2.2.2 Methodology In order to compare the total cfDNA yields obtained from blood tubes containing EDTA and citrate as anticoagulants, the same methodology as found in section 2.1.2 of this thesis was carried out with the following differences: 27  2.2.2.1 Blood collection Fresh blood samples were collected in paired EDTA and citrate tubes from the same five healthy adult volunteers. 2.2.2.2 Isolating cfDNA from plasma For each volunteer, 500 µL of plasma were transferred into centrifuge tubes which contained 500 µL of PBS to make a total volume of 1000 µL. Total cfDNA extractions were performed using only the QIAamp® kit, following the protocol on pages 22-25 of the kit handbook, as briefly summarized in 2.1.2.3.38 Total cfDNA was eluted twice with the same 60 µL eluate. 2.2.3 Results  cfDNA yields from EDTA tubes were compared to cfDNA yields from citrate tubes using plasma samples from five healthy adult volunteers. As demonstrated in Figure 2-7, the total cfDNA yields from the EDTA and the citrate tubes were 1500 ± 585 and 1227 ± 544 GE/ml of plasma respectively. Thus, the total cfDNA yields for EDTA and citrate samples were comparable (p=0.438).    28   Figure 2-7 PCR amplifiable DNA yields from EDTA and citrate tubes cfDNA was isolated from plasma samples from five healthy adult volunteers obtained in ethylenediaminetetraacetic acid or EDTA and citrate tubes. Isolated cfDNA was quantified via qPCR targeting the RNA polymerase II (POLR2) gene without prior PCR amplification. The figure shows the mean of the cfDNA concentrations from each individual with three independent replicates along with their associated standard deviations. The line represents the overall mean for the five volunteers. The statistical comparative test performed was the Wilcoxon signed rank test.  2.2.4 Discussion Clinically, a variety of anticoagulants which may include sodium citrate, EDTA, and heparin are commonly used to prevent clotting of whole blood samples.44 Our study samples have been treated with two different blood anticoagulants. It was therefore important to investigate whether the anticoagulant affect the total cfDNA concentration differently. There has only been one reported study that has investigates this in a similar fashion. They found no 29  significant difference in the total cfDNA levels of the three anticoagulants: EDTA, citrate and heparin (n=10) which was in agreement with our observation of comparable total cfDNA levels in EDTA and citrate tubes.43   As will be discussed in chapter 3 of this thesis, the exact anticoagulant used to collect the plasma and the stem cell samples on the day of the stem cell harvests from the pediatric solid tumor patients is ACDA. Even though ACDA contains citrate, it also contains dextrose solution A. One limitation of our study is that we did not compare the total cfDNA levels between plasma samples collected in EDTA and ACDA. This should be further investigated to confidently conclude that any observed differences in the samples are not due to the additive used.  2.2.5 Conclusion   Plasma samples with EDTA or citrate as anticoagulants yield similar amounts of total cfDNA.   2.3 Determination of the lower limit of detection of the methylation specific qPCR (MS-qPCR) technique 2.3.1  Introduction Pediatric cancers generally have a quiet genome in terms of mutational status and a specific mutation is not always detectable. In addition, the mutations that are detected in pediatric cancer can be quite heterogeneous across different cancer types.45 This makes the study of cancer using specific mutations challenging. As such, some researchers are using the methylation status of certain tumor suppressor genes as a marker for the presence of cancer cells. One example is the Ras association domain family member 1 (RASSF1a) gene located in the short arm of chromosome 3.46 It has been shown that most patients with malignancies have methylated promoter region of the RASSF1a gene, whereas this methylation pattern is rare in 30  patients without malignancies.47  Table 2-2 summarizes the frequency of the promoter methylation status of the RASSF1a gene in various pediatric cancer types.   Table 2-2 Frequency of the RASSF1a promoter methylation in pediatric cancer patients Type of cancer Sample size (n) Frequency (%) Acute leukemia47 20 15 Ewing’s sarcoma47,48 8, 31 0, 68 Hepatoblastoma47,49 27, 74 19, 34 Hepatocellular carcinoma50 2 100 Hodgkin lymphoma51 8 63 Medulloblastoma47,52 16, 27 88, 81 Melanoma53 20 35 Neuroblastoma47,54,50,55 27, 35, 8, 45 52, 83, 88, 93 Osteosarcoma47 11 0 Retinoblastoma47 17 59 Rhabdomyosarcoma47 18 61 Thyroid carcinoma50 2 100 Wilm’s tumor47,56,57 31, 39, 84 42, 54, 50   31  One approach in detecting and quantifying the promoter methylation status of RASSF1a gene is via MS-qPCR. To prepare for this method of quantification, first, isolated DNA must be treated with sodium bisulfite. As outlined in Figure 2-8, the principle behind this method is that sodium bisulfite converts unmethylated cytosine residues that are adjacent to a guanine residue into uracil residues so that after the PCR reaction, the uracil will be converted to thymine while methylated ones remain unchanged.58 This way, two unique sets of qPCR primers can be prepared; one against the unmethylated DNA and the other against the methylated DNA.58   Figure 2-8 Bisulfite treatment and MSP Principle of bisulfite treatment and methylation specific PCR (MSP). Sodium bisulfite will convert unmethylated cytosines of DNA molecules into uracil and after PCR amplification, the uracil nucleobases will be converted into thymine. On the other hand, methylated cytosines will remain as cytosines after sodium bisulfite treatment and PCR amplification. Consequently, two sets of PCR primers could be used to distinguish between the two species of DNA: one against the methylated DNA molecules and one against the unmethylated DNA molecules. Image reproduced from Zhang et al. with permission from the Royal Society of Chemistry58   Since the number of tumor cells compared to the number of normal WBCs circulating in blood is expected to be low, it is extremely important to have a sensitive detection and 32  quantification method. As such, one of the objectives of this study is to determine the lower limit of detection of the MS-qPCR technique against the RASSF1a gene. 2.3.2 Methodology 2.3.2.1 Sources of samples Since the IMR-32 neuroblastoma cell line has been previously shown to have 100% methylation of the RASSF1a gene, this was used as the source of methylated DNA thereby representing DNA from tumor cells.59 Similarly, since WBCs from healthy individuals have been previously shown to have 100% non-methylation of the RASSF1a gene, this was used as the source of unmethylated DNA thereby representing DNA from normal cells.59 These two findings were also confirmed in this study (Appendix D). One vial containing 4 million cells of the human neuroblastoma cell line, IMR-32, was received as a gift from Dr. G. Reid (British Columbia Children’s Hospital Research Institute) as well as vials containing a total of 0.5 to 1 billion WBCs from tonsil tissues from healthy children were thawed. The contents of the vials were transferred into corresponding centrifuge tubes containing media that was composed of PBS and 10% fetal bovine serum (FBS). The tubes were then spun at 135g for 5 minutes at 22°C and the cell pellets were re-suspended in media. The cells were counted using pocH-100i automated hematology analyzer from Sysmex (Illinois, USA). IMR-32 cells were diluted in increasing amounts of WBCs as shown in Table 2-3. The samples were then spun at 300g for 5 minutes at 22°C and the cells were re-suspended in 500 µL of nuclease free water.     33  Table 2-3 Serial dilutions of the IMR-32 cells in increasing amounts of WBCs Samples Contents # of IMR-32 cells # of WBCs 100% methylation 106 0 1:1 dilution 106 106 1:10 dilution 105 106 1:102 dilution 104 106 1:103 dilution 103 106 1:104 dilutiona 103 107 1:105 dilutionb 103 108 1:106 dilutionb 102 108 0% methylation 0 106 a To avoid overloading the instrument, only 50% of the total sample volume which contained 5x106 cells was used for gDNA extraction  b To avoid overloading the instrument, only 1% of the total sample volume which contained 1x106 cells was used for gDNA extraction  2.3.2.2 Genomic DNA extraction Cellular gDNA extraction was performed using the automated Maxwell® RSC instrument and its associated Blood DNA kit from Promega (Madison, Wisconsin) by following the protocol in the TM419 technical manual.60 Briefly, 30 µL of the Pro-K solution and 300 µL of lysis buffer were added to 500 µL of the sample mixtures. Then, the mixtures were incubated at 56°C for 20 minutes. After incubation, as per Figure 2-1, the Blood DNA AS1400 kit cartridges were snapped into position on the deck tray. Elution tubes containing 50uL of elution buffer were placed in the elution tube positions of the deck tray. Plungers were then placed into well #8 of each cartridge and the 830 µL sample mixtures were directly added to well #1 of the cartridges which contained binding buffer.  Subsequently, as per Figure 2-2, the deck tray was loaded into the Maxwell® RSC instrument and the Blood DNA AS1400 extraction run was started which took approximately 40 minutes. After the run was completed, the elution tubes containing the 34  eluted gDNA were spun at 10000g for 2 minutes at 22°C in order to pellet down any remaining magnetic beads which could potentially inhibit downstream qPCR analysis. 2.3.2.3 Quantification of gDNA  To measure gDNA concentrations, 1 µL of the eluted gDNA was added to microcentrifuge tubes containing 199 µL of the QuantiFluor® ONE dsDNA dye from Promega, vortexed briefly and incubated in the dark for 5 minutes at 22°C. This allows the intercalating dye to bind to double stranded DNA. After incubation, the gDNA concentration of the samples was quantified using the Quantus™ fluorometer from Promega. The fluorometer was calibrated using the human cancer cell K562 gDNA. 2.3.2.4 Bisulfite treatment Extracted gDNA samples were treated with sodium bisulfite using the Epitect Bisulfite kit from Qiagen (Hilden, Germany) by following the protocols on pages 16-20 of the kit handbook.61 Briefly, 1 µg (range: 3 µL – 18 µL) of gDNA solutions were transferred into a 200 µL polymerase chain reaction (PCR) tube and nuclease free water was added to make a total volume of 20 µL. Then, 85 µL of bisulfite mixture dissolved in water and 35 µL of DNA protect buffer were added. The mixtures were briefly vortexed and placed in the Veriti 96 well thermal cycler from Applied Biosystems® (Massachusetts, USA) using the following thermal cycling conditions: 95°C for 5 minutes, 60°C for 25 minutes, 95°C for 5 minutes, 60°C for 85 minutes, 60°C for 175 minutes and 20°C overnight. On the following day, purification of the bisulfite treated gDNA samples was performed. For this step, 560 µL of binding buffer was mixed with the bisulfite treated gDNA solutions and vortexed before loading into the Epitect spin columns. After the gDNA have bound to the column, a wash step using 500 µL of wash buffer was performed. Then, 500 µL of desulfonation buffer was loaded to the spin columns and incubated 35  at 22°C. After 15 minutes of incubation, the gDNA were washed twice with the wash buffer. In order to remove excess ethanol, the columns were placed in a 56°C heat block. Finally, the bound bisulfite treated gDNA were eluted twice from the spin columns using the same 50 µL of elution buffer by centrifugation at 15000g for 1 minute. The eluted bisulfite treated gDNA samples were subsequently stored at -20°C until MS-qPCR was performed. The intra-assay variation of the bisulfite treatment was evaluated eight times with the same batch of pooled cfDNA samples from cancer patients and it was determined to be 21% and 23% using unmethylated RASSF1a and methylated RASSF1a qPCR primers respectively. The inter-assay variation of the bisulfite treatment was evaluated with pooled cfDNA samples from cancer patients tested in each run on eight different days and it was determined to be 30% and 26% using unmethylated RASSF1a and methylated RASSF1a qPCR primers respectively.  2.3.2.5 MS-qPCR Frozen bisulfite treated gDNA samples were thawed out and quantified MS-qPCR using the 7500 Fast Real-Time PCR instrument and software v.2.3 from Applied Biosystems® (Massachusetts, USA). The thermal cycling condition used was 95°C for 3 minutes followed by 45 cycles of 95°C for 5 seconds and 60°C for 30 seconds. Each reaction contained 12.5 µL of the PrimeTime® Gene Expression Master Mix from Integrated DNA Technologies (California, USA), 1.5 µL of the forward and reverse primers (300 nmol/L concentration each), 1.25 µL of the TaqMan probe for the methylated or umethylated RASSF1a gene at a concentration of 200 nmol/L, 1.2 µL of nuclease free water and 7 µL of bisulfite treated gDNA solution. The primer pair and probe sequences were obtained from Stutterheim et al.59 and are found in Appendix A. For each MS-qPCR run, the samples were quantified in triplicate. A no template control was included in every run which consisted of nuclease free water. The intra-assay variation of the 36  MS-qPCR assay was evaluated eight times with the same batch of pooled bisulfite treated cfDNA samples from cancer patients and it was determined to be 19% and 28% using unmethylated RASSF1a and methylated RASSF1a qPCR primers respectively. The inter-assay variation of the MS-qPCR was evaluated with pooled bisulfite treated cfDNA samples from cancer patients tested in each run on six different days and it was determined to be 15% and 22% using unmethylated RASSF1a and methylated RASSF1a qPCR primers respectively. 2.3.3 Results The lower limit of detection of the MS-qPCR assay using unmethylated and methylated RASSF1a primers and probes was tested by performing serial dilution of the IMR-32 neuroblastoma cells into increasing amount of WBCs. As seen in Figure 2-9, the assay reached a lower limit of detection of one tumor cell in 104 normal WBCs.      37   Figure 2-9 Lower limit of detection of the MS-qPCR assay The lower limit of detection of the methylation specific quantitative PCR (MS-qPCR) assay was assessed by performing serial dilutions of the IMR-32 neuroblastoma cells (100% methylated of the Ras associated domain family member 1 (RASSF1a)) gene into increasing amount of white blood cells from healthy individuals (100% unmethylated RASSF1a gene). Isolated DNA was subjected to sodium bisulfite treatment and quantified via MS-qPCR using methylated and unmethylated RASSF1a primers. The experiment was performed in three independent replicates. In the figure, the mean of the cycle threshold (Ct) values are displayed with their associated standard deviation.    2.3.4 Discussion Using patient specific primers is the preferred approach to quantify tumor cfDNA due to its high specificity. However, it is time consuming and requires sequencing of each patient’s DNA derived from tumor tissue. Alternatively, qPCR based on the promoter methylation status of a tumor suppressor gene such as RASSF1a can be used, given that promoter methylation of the RASSF1a gene is commonly observed in cancer patients but rarely seen in healthy individuals.62 In addition, because promoter methylation of the RASSF1a gene is observed in various cancers, 38  it can be used to study a wide range of cancer types simultaneously.63 Promoter methylation rates for the RASSF1a gene can vary between different types of cancer as seen in Table 2-2.47–57  However, in order to perform MS-qPCR, cfDNA first needs to undergo sodium bisulfite treatment, a harsh condition that can decrease cfDNA levels dramatically.61 As such, it is extremely important to determine the lower limit of detection of tumor cfDNA of this assay. For this study, it was determined that the MS-qPCR assay for RASSF1a gene methylation had a lower limit of detection of one tumor cell per 10000 normal cells which was different from Stutterheim et al.’s lower limit of detection of one in 100000.59 In our hands, highly diluted samples such as the one in 106 sample showed greater variability, likely due to sampling error. 2.3.5 Conclusion  The MS-qPCR assay for the quantification of cfDNA which either has methylation or no methylation of the RASSF1a gene is able to detect one tumor cell in the presence of 10000 normal cells.  39  Chapter 3: G-CSF administration in pediatric solid tumor patients 3.1 Assessment of the effects of G-CSF treatment on tumor growth 3.1.1 Introduction G-CSF is an effective drug for treating chemotherapy-induced neutropenia, and for preparing for stem cell harvests.34,64 However, its widespread use is concerning because research findings show that Ewing sarcoma and neuroblastoma cell lines as well as patient tumors have the receptor for G-CSF, raising the possibility that G-CSF can stimulate tumor cell growth.65,66 When neuroblastoma cells from the SK-N-SH, SK-N-AS and SH-SY-5Y cell lines were treated with G-CSF in vitro, the cells showed increased proliferation compared to untreated cells.67 Then, when the SK-N-SH cells were treated with G-CSF, increased DNA synthesis and increased invasiveness were observed.67 In terms of in vivo effects, it was shown that when mice with human neuroblastoma xenografts were treated with G-CSF for twenty-one days, the mice had increased tumor weights and increased bone marrow metastasis compared to untreated mice.68 Similarly, Morales-Arias et al. treated mice with Ewing sarcoma xenografts with G-CSF or PBS for nineteen consecutive days.65 They observed that the mice treated with G-CSF had significantly bigger tumors averaging about 1218 mm3 compared to mice treated with PBS which had tumors averaging about 577 mm3.65 Thus, there is a possibility that G-CSF can stimulate tumor growth in patients as well. I hypothesize that G-CSF administration promotes tumor growth, which can be measured by increased levels of tumor cfDNA in the plasma of pediatric solid tumor patients. 40  3.1.2 Methodology 3.1.2.1 Consent and ethics Written informed consent from the participants and/or their guardians was obtained by BC Children’s Hospital BioBank (BCCHB) staff.  This study was approved by the UBC Children’s and Women’s Health Centre of BC Research Ethics Board. After consent was obtained, the study participants were assigned unique BCCHB identification codes, as well as a study specific number. The latter is used in this thesis.  3.1.2.2 Blood collection Plasma samples from fourteen children with solid tumors who were assigned unique identification codes were obtained from the BCCHB along with their associated clinical information. For each child, plasma samples were collected at two time points: before G-CSF and after G-CSF administration. Within six hours of collection, BCCHB staff performed one spin of the blood tubes at 1500g for 10 minutes at 22°C and stored the plasma at -80°C. Then, on the day of cfDNA extraction, the matching plasma aliquots from each child were thawed out, and a second spin using the Eppendorf microcentrifuge 5424 R at 16000g for 10 minutes at 22°C was performed. The plasma was then aspirated out and placed in a new microtube. 3.1.2.3 Isolating cfDNA from plasma cfDNA was extracted from 1000 µL of plasma from each patient, except for patients 2, 7 and 8 for whom volumes of 800 µL, 900 µL and 500 µL were used respectively. For the latter three patients, PBS was added to the plasma samples to make a total volume of 1000 µL. cfDNA extraction was performed using the QIAamp® kit from Qiagen by following the protocol on pages 22-25 of the kit handbook.38 Refer to section 2.1.2.3 of this thesis for a brief summary. Elution of cfDNA was performed by eluting twice with the same 60 µL eluate. 41  3.1.2.4 Real-time PCR Total cfDNA was quantified by qPCR using the 7500 Fast Real-Time PCR instrument and software v.2.3 from Applied Biosystems® (Massachusetts, USA). The thermal cycling condition that was followed was 95°C for 3 minutes followed by 45 cycles of 95°C for 5 seconds and 60°C for 30 seconds. Each reaction contained 12.5 µL of the PrimeTime® Gene Expression Master Mix from Integrated DNA Technologies (California, USA), 1.5 µL of the forward and reverse primers (300 nmol/L concentration each), 1.25 µL of the TaqMan probe for the RNA polymerase II gene (POLR2) at a concentration of 200 nmol/L, 7 µL of nuclease free water and 1.2 µL of cfDNA. The primer pair sequences were obtained from Mussolin et al.39 and are found in Appendix A. A standard curve was prepared and used to determine the starting cfDNA concentrations of the samples, using a 5-fold serial dilution of the PowerQuant™ human male gDNA standard from Promega that was first diluted to a concentration of 7576 GE/µL using the PowerQuant™ dilution buffer. For each qPCR run, the standard and unknown samples were quantified in triplicate. Also, a no template control which consisted of nuclease-free water and an internal control gDNA were included in every run. 3.1.2.5 Bisulfite treatment of cfDNA samples Extracted total cfDNA was treated with sodium bisulfite using the Epitect Bisulfite kit from Qiagen  by following pages 26-30 from the kit handbook.61 This protocol was optimized for treating short and fragmented nucleic acids such as cfDNA. Briefly, 20 µL of cfDNA samples were added to 200 µL PCR tubes along with 85 µL of bisulfite mixture dissolved in water and 35 µL of DNA protect buffer. The mixtures were then vortexed and placed in the Veriti 96 well thermal cycler from Applied Biosystems® by following the thermal cycling conditions in 2.3.2.4 of this thesis. On the following day, purification of the bisulfite treated cfDNA samples was 42  performed. For this step, 310 µL of binding buffer with 10 µg/ml of carrier RNA and 250 µL of 100% ethanol were mixed with the bisulfite treated cfDNA before loading to the Epitect spin columns. After the cfDNA samples have bound to the column, a wash step using 500 µL of wash buffer was performed. Then, 500 µL of desulfonation buffer was loaded to the spin column and incubated at 22°C. After 15 minutes of incubation, the cfDNA samples were washed twice with the wash buffer. In order to remove excess ethanol, the columns were placed in a 56°C heat block. Finally, the bound bisulfite treated cfDNA were eluted twice from the spin columns using the same 60 µL of elution buffer by centrifugation at 15000g for 1 minute. The eluted bisulfite treated cfDNA samples were subsequently stored at -20°C until MS-qPCR could be performed.  3.1.2.6 Methylation specific real-time PCR Bisulfite treated cfDNA samples were quantified by MS-qPCR. The same methodology found in section 2.3.2.5 of this thesis was carried out with a few differences. First, instead of bisulfite treated gDNA from cells, bisulfite treated cfDNA from plasma was quantified by MS-qPCR. Second, the unknown and standard samples were quantified in triplicate.  3.1.2.6.1 Controls for MS-qPCR A negative control, a positive control and a no template control were included in every run of the MS-qPCR assay which consisted of WBC gDNA extracted from tonsil tissue obtained from pediatric non-oncology patients post tonsillectomy, gDNA from the human neuroblastoma cell line called IMR-32 was received as a gift from Dr. G. Reid (British Columbia Children’s Hospital Research Institute) and nuclease free water respectively. The positive and negative control gDNA were extracted using the automated Maxwell® RSC instrument and its associated Blood DNA kit from Promega by following the protocol in the TM419 technical manual.60  Refer to section 2.3.2.2 for a brief summary. Subsequently, nucleic acid concentrations were 43  determined using fluorometric quantification as per section 2.3.2.3. Then, bisulfite treatment of 2 µg of gDNA solutions were performed using the Epitect Bisulfite kit from Qiagen by following the protocols on pages 16-20 of the kit handbook.61 Refer to section 2.3.2.4 for a brief summary.  3.1.2.6.2 Standard curves for MS-qPCR Two standard curves were prepared and used to determine the starting cfDNA concentrations of the patient samples. One standard curve was prepared using a 4-fold serial dilution of the 455 GE/µL of bisulfite treated WBC gDNA extracted from tonsil tissue from healthy children. The second standard curve was prepared using a 3-fold serial dilution of the 303 GE/µL of bisulfite treated gDNA extracted from the IMR-32 cell line. The nucleic acid concentrations of the bisulfite treated WBC gDNA and the IMR-32 gDNA were determined spectrophotometrically using the NanoDrop instrument from Thermo Fisher Scientific (Massachusetts, USA). The NanoDrop instrument was set to measure RNA due to the fragmented condition of the bisulfite treated gDNA solutions. For this method of quantification, 1 µL of nuclease free water was placed on the instrument and measured as the blank sample. Next, 1 µL of the bisulfite treated WBC gDNA or the IMR-32 gDNA were placed on the instrument and read. 3.1.2.7 Statistical analysis The same statistical analysis as found in section 2.1.2.6 was performed except that the data was expressed as the average ± standard deviation of the three qPCR technical replicates since the experiment was only performed once. Also, graphs were created using GraphPad Prism software from GraphPad Prism Incorporation (La Jolla, California, USA). 44  3.1.3 Results The study cohort for the pre and post G-CSF comparison portion of this study was composed of children with solid tumors who were receiving treatment at BC Children’s Hospital. Patient diagnosis was determined by clinical pathologists using tumor biopsy samples. Metastasis was assessed by one or more of the following procedures: ultrasound, magnetic resonance imaging (MRI), computerized tomography (CT) scan, combined positron emission tomography and computed tomography (PET/CT) scan and/or bone marrow biopsy. Plasma samples before and after G-CSF treatment were obtained from 14 children from the BCCHB. Patient characteristics are listed in Table 3-1.The mean age of the patients at diagnosis was 5.4 years old (range: 0.4 - 14.2 years old) and the male to female ratio was 7:7.   Table 3-1 Patient characteristics for the pre and post G-CSF study Patient No. Diagnosis Sex Age Days between pre and post G-CSF sample collection 1 rhabdomyosarcoma f 5.7 14 2 rhabdomyosarcoma m 13.2 27 3 neuroblastoma m 2.9 11 4 neuroblastoma f 3.7 21 5 neuroblastoma f 8.7 17 6 Burkitt’s lymphoma m 14.2 22 7 neuroblastoma m 4.8 20 8 neuroblastoma m 3.8 18 9 neuroblastoma f 2.3 7 10 glioblastoma f 0.9 18 11 rhabdomyosarcoma f 2.9 11 12 choroid plexus carcinoma f 3.8 32 13 rhabdoid tumor m 0.4 14 14 medulloblastoma m 8.3 28   45  The number of days between pre G-CSF and post G-CSF ranged from 7 days to 32 days. Total cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients are illustrated in Figure 3-1 with the raw data found in Table 3-2. Total cfDNA represents both the normal cfDNA and tumor cfDNA molecules. Table 3-3 summarizes the number of days in between relevant clinical events and sample collections. An increase in total cfDNA levels after G-CSF treatment was observed in half of the children studied: patient #1, 6, 8, 10, 11, 13 and 14. The median total cfDNA levels before G-CSF treatment was 8740 GE/ml of plasma (range: 731 to 110136 GE/ml of plasma) which was higher than the median total cfDNA levels after G-CSF treatment of 5610 GE/ml of plasma (range: 822 to 98754 GE/ml of plasma).  Normal cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients are illustrated in Figure 3-2 with the raw data found in Table 3-2. Normal cfDNA represents the cfDNA with the unmethylated RASSF1a promoter. The median normal cfDNA levels before G-CSF treatment was 1936 GE/ml of plasma (range: 336 to 23845 GE/ml of plasma) which was comparable to the median normal cfDNA levels after G-CSF treatment of 1962 GE/ml of plasma (range: 199 to 59880 GE/ml of plasma). Out of the fourteen patients studied, tumor cfDNA was detected in nine patients using the RASSF1a method. Tumor cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients are illustrated in Figure 3-3 with the raw data found in Table 3-2. Tumor cfDNA represents the cfDNA with the methylated RASSF1a promoter. The median tumor cfDNA levels before G-CSF treatment was 6778 GE/ml of plasma (range: 0 to 52355 GE/ml of plasma) which although higher than the median tumor cfDNA levels after G-CSF treatment of 599 GE/ml of plasma (range: 0 to 13277 GE/ml of plasma) was not statistically different (p=0.3438; Wilcoxon signed rank test). Overall, it was observed that G-CSF is not associated 46  with higher tumor cfDNA post-treatment. In the nine patients who had detectable tumor cfDNA, a decrease in tumor cfDNA after G-CSF administration was observed in patient # 2,4,7,8 and 9 while patient # 1,3,11 and 13 had constant tumor cfDNA levels after G-CSF treatment. There were two outstanding differences between these two patient populations. The patients with constant tumor cfDNA levels after G-CSF treatment had lower average tumor cfDNA levels (pre G-CSF: 52 versus 18190 GE/ml of plasma; post G-CSF: 71 versus 4370 GE/ml of plasma) and had received, on average, less dosage of chemotherapy (mean 0.75 days versus 5 days of treatment) in between sample collections than the patients with decreased tumor cfDNA post G-CSF treatment. 47  p r e p o s t p r e p o s t p r e p o s t p r e p o s t p r e p o s t p r e p o s t p r e p o s t1 0 01 0 11 0 21 0 31 0 41 0 51 0 6S a m p le sTotal cfDNA concentration (GE/ml of plasma)P a tie n t #  3P a tie n t #  4P a tie n t #  5P a tie n t #  7P a tie n t #  8P a tie n t #  9P a tie n t #  1P a tie n t #  2P a tie n t #  1 1P a tie n t #  6P a tie n t #  1 3P a tie n t #  1 0P a tie n t #  1 2P a tie n t #  1 4G -C S FN e u ro b la s to m a R h ab d o m yo sa rc o m a B u rk it t 'sLym phom aR h a b d o id  tu m o r G lio b la s to m a C h o ro id  P le x u sC a rc in o m aM e d u llo b la s to m a Figure 3-1 Total cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients Total cfDNA from the plasma of pediatric solid tumor patients before and after G-CSF treatment were quantified via qPCR using the RNA polymerase II (POLR2) gene without prior PCR amplification. The figure shows the mean total cfDNA concentrations from three technical replicates along with its associated standard deviation. Total cfDNA represents the combined cfDNA from normal and tumor cells.  48  Table 3-2 Raw data for before and after G-CSF treatment comparison of cfDNA levels Patient #  Diagnosis  Days b/w samples  Total cfDNA (GE/ml of plasma)  Normal cfDNA (GE/ml of plasma)  Tumor cfDNA (GE/ml of plasma)  BEFORE  G-CSF  AFTER  G-CSF  BEFORE  G-CSF  AFTER  G-CSF  BEFORE  G-CSF  AFTER  G-CSF  1 rhabdomyosarcoma 14 4781 ± 290 5977 ± 189 2098 ± 407 2214 ± 196 26 ± 5  9 ± 173 2 rhabdomyosarcoma 27 26900 ± 977 11234 ± 815 1898 ± 185 1092 ± 147 15689 ± 872 4831 ± 301 3 neuroblastoma 11 5840 ± 1574 2419 ± 279 1400 ± 243 615 ± 108 20 3 4 neuroblastoma 21 15438 ± 1982 5803 ± 318 1711 ± 298 949 ± 309 8834 ± 286 1471 ± 198 5 neuroblastoma 17 11640 ± 355 822 ± 209 3594 ± 593 199 ± 116 BLD BLD 6 Burkitt's lymphoma 22 1228 ± 201 3524 ± 484 864 ± 117 1798 ± 352 BLD BLD 7 neuroblastoma 20 61378 ± 2814 5417 ± 280 17466 ± 2236 2125 ± 642 7295 ± 119 599 ± 186 8 neuroblastoma 17 14594 ± 2016 18432 ± 2205 1923 ± 425 4861 ± 410 6778 ± 359 1671 ± 183 9 neuroblastoma 7 110136 ± 11339 30868 ± 1294 6728 ± 424 6358 ± 574 52355 ± 3727 13277 ± 1156 10 glioblastoma 19 3562 ± 297 4750 ± 496 1506 ± 187 2653 ± 67 BLD BLD 11 rhabdomyosarcoma 11 5679 ± 561 13868 ± 1228 1949 ± 134 4761 ± 190 48 57 ± 73 12 choroid plexus carcinoma 32 3655 ± 1117 1474 ± 288 2125 ± 678 517 ± 221 BLD BLD 13 rhabdoid tumor 14 53154 ± 1443 98754 ± 1882 23845 ± 2262 59880 ± 1868 115  ± 88 215 ± 78 14 medulloblastoma 28 731 ± 115 1811 ± 152 336 ± 44 918 ± 248 BLD BLD Median (range) 18(7-32) 8740(731-110136) 5610(822-98754) 1936 (336-23845) 1962(199-59880) 6778(0-52355) 599 (0-13277)  BLD: below the lower limit of detection  49  Table 3-3 Days between important clinical events and sample collection Patient # Diagnosis Disease outcome # of days b/w Total # of days of treatment in b/w sample collection Total # of days of G-CSF treatment in b/w sample collection last G-CSF treatment & sample 1 diagnosis & sample 1 diagnosis & sample 2 sample 1 & 2 diagnosis & death diagnosis & remission last treatment and sample 1 last treatment & sample 2 1 rhabdomyosarcoma deceased 14 46 60 14 255 N/A 0 14 1 7 2 rhabdomyosarcoma deceased 21 160 187 27 252 N/A 1 7 4 12 3 neuroblastoma under treatment 17 103 114 11 N/A N/A 1 12 0 11 4 neuroblastoma deceased 15 602 623 21 717 N/A 0 0 5 7 5 neuroblastoma in remission 20 243 260 17 N/A 460 27 14 4 6 6 Burkitt's lymphoma in remission NPG 2 24 22 N/A 54 NPT 0 10 7 7 neuroblastoma under treatment NPG 17 37 20 N/A N/A 0 6 5 9 8 neuroblastoma under treatment NPG 5 23 18 N/A N/A NPT 10 6 10 9 neuroblastoma under treatment NPG 2 9 7 N/A N/A NPT 2 5 2 10 glioblastoma in remission 17 120 138 18 N/A 378 49 10 2 10 11 rhabdomyosarcoma under treatment NPG 7 18 11 N/A N/A 1 5 1 1 12 choroid plexus carcinoma in remission NPG 44 76 32 N/A 230 0 0 20 16 13 rhabdoid tumor deceased 6 45 59 14 104 N/A 2 0 1 14 14 Medulloblastoma in remission 15 200 228 28 N/A 261 15 0 3 14 Average  (range) 16 (6 - 21) 114 (2 - 602) 133 (9 - 623) 19 (7 - 32) 332 (104 - 717) 277 (54 - 460) 9 (0 - 27) 6 (0 - 14) 5 (0 - 20) 9 (2 - 16)  N/A: Not applicable NPT: No previous treatment NPG: No previous G-CSF treatment50   Figure 3-2 Normal cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients cfDNA from the plasma of pediatric solid tumor patients before and after G-CSF treatment were treated with sodium bisulfite and the normal cfDNA levels were determined via methylation specific quantitative PCR (MS-qPCR) using the unmethylated Ras associated domain family member 1 (RASSF1a) gene. The figure shows the mean normal cfDNA concentrations from three technical replicates along with its associated standard deviation. Normal cfDNA represents the cfDNA derived from normal cells.  51  p r e p o s t p r e p o s t p r e p o s t p r e p o s t p r e p o s t p r e p o s t p r e p o s t1 0 01 0 11 0 21 0 31 0 41 0 5S a m p le sTumor cfDNA concentration (GE/ml of plasma)G -C S FN e u ro b la s to m a R h ab d o m yo sa rc o m a R h a b d o id  tu m o rB u rk it t 'sLym phom aP a tie n t #  3P a tie n t #  4P a tie n t #  5P a tie n t #  7P a tie n t #  8P a tie n t #  9P a tie n t #  1P a tie n t #  2P a tie n t #  1 1P a tie n t #  6P a tie n t #  1 3P a tie n t #  1 0P a tie n t #  1 2P a tie n t #  1 4G lio b la s to m a C h o ro id  P le x u sC a rc in o m aM e d u llo b la s to m a Figure 3-3 Tumor cfDNA concentrations before and after G-CSF treatment in pediatric solid tumor patients cfDNA from the plasma of pediatric solid tumor patients before and after G-CSF treatment were treated with sodium bisulfite and the tumor cfDNA levels were determined via methylation specific quantitative PCR (MS-qPCR) using the methylated Ras associated domain family member 1 (RASSF1a) gene. The figure shows the mean tumor cfDNA concentrations from three technical replicates along with its associated standard deviation. Tumor cfDNA represents cfDNA derived from tumor cells.  52  Table 3-4 outlines the results for the pre and post G-CSF tumor cfDNA inter-assay coefficient of variation (CV) determination experiment of our study. The patients were divided into three broad categories: high tumor cfDNA levels, low tumor cfDNA levels and no detectable tumor cfDNA. The inter-run CV between six separate experiments for three patients with high tumor cfDNA levels was 15% and the inter-assay CV for the two patients with low tumor cfDNA levels was 137%. The tumor cfDNA levels for the two patients with no detectable tumor cfDNA consistently were below the lower limit of detection.53  Table 3-4 Pre and post G-CSF tumor cfDNA inter-assay coefficient of variation (CV) determinations  High tumor cfDNA Low tumor cfDNA No detectable tumor cfDNA Patient  2 4 9 1 3 12 14 Pre or Post G-CSF Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post Experiment No. 1 15689 4831 8834 1471 52355 13277 26 9 20 3 BLD BLD BLD BLD 2 19089 5234 10455 1642 52225 17694 BLD BLD 61 48 BLD BLD BLD BLD 3 18830 5418 9035 947 48065 15516 BLD 33 15 BLD BLD BLD BLD BLD 4 18463 4641 9673 1624 40098 14823 31 15 BLD BLD BLD BLD BLD BLD 5 20715 5659 12522 2005 38660 18358 27 20 12 BLD BLD BLD BLD BLD 6 21186 5783 10675 2217 35407 15660 BLD BLD 12 BLD BLD BLD BLD BLD Mean 18995 5261 10199 1651 44469 15888 14 13 20 8 BLD BLD BLD BLD SD 1950 453 1355 442 7352 1871 16 13 21 19 BLD BLD BLD BLD Inter assay CV (%) 10 9 13 27 17 12 110 98 107 231 BLD BLD BLD BLD Mean inter assay CV (%) 15 137 BLD  BLD: Below the lower limit of detection 54  Outlines of what treatments or procedures each patient received around the time of sample collections are found in Figure 3-4 until Figure 3-17. Unfortunately due to logistics of patient sample collection, a substantial number of patients (patient # 1, 2, 3, 4, 5, 10, 13 and 14) had received prior G-CSF treatment as part of their chemotherapy cycles before the pre G-CSF blood sample was collected (range 6 days to 21 days). 55     Figure 3-4 Timeline of clinical treatment for patient 1 A timeline summarizing important treatment procedures around the time of the sample collections for patient 1. There were 14 days in between the pre and post G-CSF sample collections in which the patient received 1 day of chemotherapy and 7 days of G-CSF treatment. The patient also received G-CSF 14 days before the pre G-CSF sample was collected. The patient has since succumbed to the disease.  56     Figure 3-5 Timeline of clinical treatment for patient 2 A timeline summarizing important treatment procedures around the time of the sample collections for patient 2. There were 27 days in between the pre and post G-CSF sample collections in which the patient received 4 days of radiation treatment and 12 days of G-CSF treatment. The patient also received G-CSF 21 days before the pre G-CSF sample was collected. The patient has since succumbed to the disease.  57     Figure 3-6 Timeline of clinical treatment for patient 3 A timeline summarizing important treatment procedures around the time of the sample collections for patient 3. There were 11 days in between the pre and post G-CSF sample collections in which the patient received 11 days of G-CSF treatment. The patient also received G-CSF 17 days before the pre G-CSF sample was collected. The patient has residual disease and is currently undergoing treatment.    58     Figure 3-7 Timeline of clinical treatment for patient 4 A timeline summarizing important treatment procedures around the time of the sample collections for patient 4. There were 21 days in between the pre and post G-CSF sample collections in which the patient received 5 days of chemotherapy and 7 days of G-CSF treatment. The patient also received G-CSF 15 days before the pre G-CSF sample was collected. The patient has since succumbed to the disease.  59     Figure 3-8 Timeline of clinical treatment for patient 5 A timeline summarizing important treatment procedures around the time of the sample collections for patient 5. There were 17 days in between the pre and post G-CSF sample collections in which the patient received 4 days of chemotherapy and 6 days of G-CSF treatment. The patient also received G-CSF 20 days before the pre G-CSF sample was collected. The patient has achieved complete remission.   60     Figure 3-9 Timeline of clinical treatment for patient 6 A timeline summarizing important treatment procedures around the time of the sample collections for patient 6. There were 22 days in between the pre and post G-CSF sample collections in which the patient received 10 days of chemotherapy and 7 days of G-CSF treatment. The patient also received no prior G-CSF treatment before the pre G-CSF sample was collected. The patient has achieved complete remission.  61     Figure 3-10 Timeline of clinical treatment for patient 7 A timeline summarizing important treatment procedures around the time of the sample collections for patient 7. There were 20 days in between the pre and post G-CSF sample collections in which the patient received 5 days of chemotherapy and 9 days of G-CSF treatment. The patient also received no prior G-CSF treatment before the pre G-CSF sample was collected. The patient has residual disease and is currently undergoing treatment.  62     Figure 3-11 Timeline of clinical treatment for patient 8 A timeline summarizing important treatment procedures around the time of the sample collections for patient 8. There were 18 days in between the pre and post G-CSF sample collections in which the patient received 6 days of chemotherapy and 10 days of G-CSF treatment. The patient also received no prior G-CSF treatment before the pre G-CSF sample was collected. The patient has residual disease and is currently undergoing treatment.  63    Figure 3-12 Timeline of clinical treatment for patient 9 A timeline summarizing important treatment procedures around the time of the sample collections for patient 9. There were 7 days in between the pre and post G-CSF sample collections in which the patient received 5 days of chemotherapy and 2 days of G-CSF treatment. The patient also received no prior G-CSF treatment before the pre G-CSF sample was collected. The patient has residual disease and is currently undergoing treatment.  64     Figure 3-13 Timeline of clinical treatment for patient 10 A timeline summarizing important treatment procedures around the time of the sample collections for patient 10. There were 18 days in between the pre and post G-CSF sample collections in which the patient received 2 days of chemotherapy and 10 days of G-CSF treatment. The patient also received G-CSF treatment 17 days before the pre G-CSF sample was collected. The patient has achieved complete remission.  65     Figure 3-14 Timeline of clinical treatment for patient 11 A timeline summarizing important treatment procedures around the time of the sample collections for patient 11. There were 11 days in between the pre and post G-CSF sample collections in which the patient received 1 day of chemotherapy and 1 day of G-CSF treatment. The patient also received no prior G-CSF treatment before the pre G-CSF sample was collected. The patient has residual disease and is currently undergoing treatment.  66     Figure 3-15 Timeline of clinical treatment for patient 12 A timeline summarizing important treatment procedures around the time of the sample collections for patient 12. There were 32 days in between the pre and post G-CSF sample collections in which the patient received 20 days of chemotherapy and 16 days of G-CSF treatment. The patient also received no prior G-CSF treatment before the pre G-CSF sample was collected. The patient has achieved complete remission.  67     Figure 3-16 Timeline of clinical treatment for patient 13 A timeline summarizing important treatment procedures around the time of the sample collections for patient 13. There were 14 days in between the pre and post G-CSF sample collections in which the patient received 1 day of chemotherapy and 14 days of G-CSF treatment. The patient also received G-CSF treatment 6 days before the pre G-CSF sample was collected. The patient has succumbed to the disease.  68     Figure 3-17 Timeline of clinical treatment for patient 14 A timeline summarizing important treatment procedures around the time of the sample collections for patient 14. There were 28 days in between the pre and post G-CSF sample collections in which the patient received 3 days of chemotherapy and 14 days of G-CSF treatment. The patient also received G-CSF treatment 15 days before the pre G-CSF sample was collected. The patient has achieved complete remission.   69  3.1.4 Discussion To our knowledge, this is the first study investigating the effect of G-CSF treatment in tumor growth by non-invasively measuring total and tumor cfDNA levels, in pediatric solid tumor patients.  Our results are in agreement with reports in the literature that some children with cancer have high amounts of total cfDNA that can vary from 7000 to 17000 GE/ml of plasma.14,21,39 Our findings also demonstrate that it is possible to detect tumor cfDNA in some children with solid tumors using the MS-qPCR technique targeting promoter methylation of the RASSF1a gene. This assay is quite robust in that it gives reproducible results for patients with no detectable tumor cfDNA levels and for patients with high tumor cfDNA levels. On the other hand, for patients with low detectable tumor cfDNA levels, the average inter-assay CV was high. This is most likely because the tumor cfDNA levels are at the lower end of the detection range of the assay, hence subject to greater variability. Thus, in the future, in order to accurately quantify low tumor cfDNA levels, a more sensitive assay could be used such as digital droplet PCR (ddPCR) which is 10x more sensitive than our assay.69 Out of the fourteen children analyzed, tumor cfDNA was detectable in nine children. The presence of tumor cfDNA was suggestive of a negative outcome in this pilot study as all children with detectable tumor cfDNA either succumbed to their disease or has residual disease and is undergoing treatment. On the other hand, all five children without tumor cfDNA have achieved complete remission and are still alive. This findings were in line with Muller et al.’s observation that promoter methylation of cfDNA through the RASSF1a gene or the adenomatous polyposis coli (APC) gene was a strong predictor of negative outcome in eighty-six breast cancer patients.70 It should be noted that we did not test diagnostic tumor samples for promoter methylation of the RASSF1a gene; therefore, 70  we cannot ascertain that the five children without detectable tumor cfDNA truly do not have tumor cfDNA. Tumor specific primers or testing the diagnostic material for promoter methylation of the RASSF1a gene would be contributory for further studies. Of note, all three children with a brain tumor diagnosis had no detectable tumor cfDNA. Previous studies have shown that tumor cfDNA can be detected in brain tumor patients, though in low frequencies (<10 to 55%).18,24,71–73 However, most brain tumor studies were done on adults. One study has successfully reported detection of tumor cfDNA in a child with a brain tumor. Chakravadhanula and colleagues analyzed tumor cfDNA with deletions of the SWI/SNF-related matrix-associated actin-dependent regulator of chromatin subfamily B member 1 (SMARCB1) gene from a single child diagnosed with atypical teratoid rhabdoid tumor.20 This patient, however, had no evidence of metastasis as per MRI analysis.20 It is possible that the three brain tumor patients in our study had very low quantities of tumor cfDNA in their blood which were not detectable by the MS-qPCR assay. Alternatively, it is also possible that these patients do not have promoter methylation of the RASSF1a gene. The frequency of promoter methylation of the RASSF1a gene in brain tumor patients reported in literature varies from 34% to 79%.47,52,74,75 In this case, testing promoter methylation of the RASSF1a gene in diagnostic tumor tissue samples should be explored in future studies. If by doing so reveals that the patients do not have promoter methylation of the RASSF1a gene, then, patient-specific primers should be designed against specific mutations found in the tumor tissue.  Overall, this proof-of-concept study did not show an increase in tumor cfDNA levels with G-CSF treatment in children with different types of solid tumors. However, the true effect of G-CSF treatment cannot be conclusively determined as the samples used were not obtained in a systematic manner and are subject to confounders. For instance, samples were taken during 71  different treatment cycles; some patients received G-CSF treatment during previous cycles; there are different lengths of time between collection of pre and post G-CSF treatments samples, and patients received different types of treatment including radiation, chemotherapy and surgery. In actuality, five out of the nine patients with detectable tumor cfDNA showed decreased tumor cfDNA post G-CSF treatment. The average number of days in between the pre and post G-CSF plasma collection is 19 days (range: 7 days to 32 days). Therefore, much could have happened between the two collections. Upon close examination of the clinical data, these patients received, on average, five days of chemotherapy or radiation treatment (range: 4 days to 6 days) in between sample collections. The observed decrease in tumor cfDNA levels post G-CSF treatment was likely due to the treatment received in between sample collections. Interestingly, the remaining four children with detectable tumor cfDNA had constant tumor cfDNA levels before and after G-CSF treatment and had only, on average, 0.75 days of chemotherapy treatment in between sample collections. Moreover, these four children had tumor cfDNA levels at the low end of our assay’s detection range. Therefore, it is possible that any true changes in the tumor cfDNA levels in these patients are masked by the large variation at the detection limit of our assay. Another possible confounding factor to the tumor cfDNA analysis in this study is that eight patients received G-CSF treatment on average sixteen days (range: 6 days to 21 days) before the pre G-CSF blood sample collection was performed. The previous G-CSF treatment could potentially have residual effect on the pre G-CSF blood samples. Therefore, in future studies, it is critical that a standardized sample collection plan be put into place, to minimize possible confounding factors with regards to tumor cfDNA levels.  In addition, we speculate that the cfDNA levels (normal and tumor) as measured in this study likely underrepresent the actual values due to the harsh conditions of the sodium bisulfite 72  treatment which requires extremely high temperatures and low pH conditions, which can cause degradation and/or loss of cfDNA.61 Another caveat to this study is that the QIAamp® kit used to isolate cfDNA has cellular gDNA contamination which may falsely increase the total and normal cfDNA levels. As the number of circulating solid tumor cells is minimal to non-existent, tumor cfDNA values should not be affected by contaminating tumor gDNA and are more reflective of the true value.  3.1.5 Conclusion This study had too many confounding factors to conclusively answer the question whether G-CSF treatment increases tumor cfDNA levels in children with solid tumors. However, the presence of cfDNA either before or after G-CSF treatment is suggestive of a poor prognosis.    3.2 Comparison of tumor cfDNA levels in the plasma and tumor gDNA levels in the stem cell product 3.2.1 Introduction From 1991 to 2014, 469 autologous PBSC collection procedures have been performed at British Columbia Children’s Hospital with an average of 20 procedures per year.76 PBSC collections and the subsequent transplant of the harvested stem cells may be performed on children with solid tumors as part of their treatment protocol to rescue the bone marrow post myeloablative therapy. An outline of the stem cell collection and transplant process is shown in Figure 3-18. First, patients are administered mobilizing agents such as G-CSF which allows the release of CD34+ hematopoietic stem cells from the bone marrow into circulation. An apheresis machine performs continuous centrifugation of the patient’s blood. This allows different components of the blood to be separated out and placed in collection bags. The plasma product is 73  stored at -80°C while the stem cell product is stored at -180°C. Afterwards, the patient receives high dose chemotherapy to destroy all the cells in the bone marrow. The patient is then transplanted with their previously collected stem cells which will migrate to the bone marrow and re-populate there.   Figure 3-18 Peripheral blood stem cell collection and transplant flowchart A flowchart of the procedure for peripheral blood stem cell collection and subsequent transplant. First, the patient is injected with mobilizing agents such as G-CSF which stimulates the bone marrow to produce more stem cells and to mobilize them into the circulation. The patient’s blood is separated into different components using an apheresis machine, collected and stored. The patient then undergoes myeloablative chemotherapy to destroy all the cells in the bone marrow. Subsequently, the patient undergoes transplant of the previously collected stem cells which will migrate into the bone marrow to re-populate there. 74  Theoretically, harvesting and transplanting back bone marrow products from a patient with a solid tumor could result in reintroducing tumor cells. The latter could potentially be present in the bone marrow product, especially from patients with bone marrow involvement of their disease. Currently, there is no agreement in regards to the frequency of tumor cell contamination in the stem cell products, and whether this contributes to recurrence of the disease. Van Wezel et al., only found evidence of tumor cell contamination in six PBSC grafts and three CD34+ purged grafts of one hundred and four neuroblastoma patients.77  In addition, they found that there was no association between graft positivity and adverse outcome.77 On the other hand, Yaniv and colleagues detected the presence of the Ewing Sarcoma RNA binding protein 1 (EWS) and friend leukemia integration 1 transcription factor (FLI1) translocation in the PBSC grafts, an indication of tumor cell contamination, in all eleven of their Ewing Sarcoma patients.78 Out of the eleven patients, two are deceased, seven relapsed and two are in remission post stem cell transplant.78 Therefore, for our study, the presence of tumor cells in the stem cell products from various pediatric solid tumors were assessed using MS-qPCR. I hypothesize that after G-CSF stimulation, the level of genomic tumor DNA in the stem cell product is associated with the level of tumor cfDNA in the plasma product. I also hypothesize that there is a relationship between cfDNA levels and bone marrow disease status. 3.2.2 Methodology 3.2.2.1 Blood collection Matching plasma and stem cell products from nine children with solid tumors who were assigned unique identifications were obtained from the BCCHB along with their associated clinical information. On the day of stem cell harvest, the patient’s blood was directed into the Spectra Optia® instrument from Terumo BCT (Colorado, USA). The machine was set up to 75  perform a continuous centrifugation of the patient’s blood in order to separate out the blood components such as plasma and stem cells. These components are then collected in separate bags with ACDA as the anticoagulant and frozen. On the day of cfDNA extraction, the plasma samples from each child were thawed out and a second spin using the Eppendorf microcentrifuge 5424 R at 16000g for 10 minutes at 22°C was performed. The plasma was then aspirated out and placed in a new microtube.  3.2.2.2 Isolating cfDNA from the plasma product cfDNA was extracted from 5000 µL of plasma from each patient. cfDNA extraction was performed using the QIAamp® kit from Qiagen by following the protocol on pages 26-29 of the kit handbook which was optimized for larger volume samples.38 Briefly, 500 µL of Pro-K and 4000 µL of lysis buffer containing 1.0 µg of carrier RNA were added to 5000 µL of plasma and was incubated at 60°C for half an hour. After incubation, 9000 µL of binding buffer was added to the mixture. As per Figure 2-3, the mixture was then loaded onto the spin column and pulled through the silica membrane of the spin column using vacuum pressure. The bound cfDNA on the spin columns were then washed twice with wash buffers and then 750 µL of 100% ethanol to remove residual impurities. Then, to allow evaporation of excess ethanol, the spin columns were dried thoroughly on a 56°C heat block for 10 minutes. Finally, the bound cfDNA were eluted by adding 60 µL of the same elution buffer twice to the spin columns, incubating for 3 minutes and spinning at 21130g for 1 minute.  The eluted cfDNA samples were stored at -20°C until they were quantified by qPCR. 3.2.2.3 Isolating gDNA from the stem cell product A vial of stem cell product from each of the nine patients was thawed out and diluted to a total of 1x106 WBCs. The diluted samples were then spun at 300g for 5 minutes at 22°C. The 76  supernatant was removed and the cell pellet was re-suspended in 500 µL of nuclease free water. Then, gDNA extraction was performed using the automated Maxwell® RSC instrument and its associated Blood DNA kit from Promega by following the protocol in the TM419 technical manual.60  Refer to section 2.3.2.2 for a brief summary.  3.2.2.4 Real-time PCR Total cfDNA from plasma and total gDNA from the stem cells were quantified by qPCR using the 7500 Fast Real-Time PCR instrument and software v.2.3 from Applied Biosystems® (Massachusetts, USA). Refer to section 3.1.2.4 for a brief summary. 3.2.2.5 Bisulfite treatment of cfDNA from plasma and gDNA from stem cells Extracted cfDNA from plasma was treated with sodium bisulfite using the Epitect Bisulfite kit from Qiagen by following pages 26-30 from the Epitect Bisulfite handbook.61 Refer to section 3.1.2.5 for a summary. In addition, extracted gDNA from stem cells was treated with sodium bisulfite using the same kit by following 16-20 of the kit handbook.61 A summary of the protocol is found in section 2.3.2.4 except with the following three differences. First of all, 20 µL of the gDNA solution from the stem cells was used for extraction and so no nuclease free water was added. Second, during the purification step, the lysis buffer used contained 10 µg/ml of carrier RNA in order to keep the protocol consistent with the bisulfite treatment of the cfDNA samples from the plasma. Thirdly, bound bisulfite treated gDNA was eluted twice from the spin columns using the same 60 µL of elution buffer. 3.2.2.6 Methylation specific real-time PCR Bisulfite treated cfDNA from plasma and bisulfite treated gDNA from stem cells were quantified by MS-qPCR. The same methodology as found in section 2.3.2.5 of this thesis was carried out except for a few differences. First, the unknown and standard samples were 77  quantified in triplicate. Second, a negative control, a positive control and a no template control were included in every run which consisted of WBC gDNA extracted from tonsil tissue from a healthy child, gDNA from the IMR-32 cell line and nuclease free water, respectively. 3.2.2.6.1 Controls and standard curves for MS-qPCR The same preparation for the controls and the standard curves as found in sections 3.1.2.6.1 and 3.1.2.6.2 were performed except that one standard curve was prepared using a 3.5-fold serial dilution of the 5318 GE/µL or 1515 GE/µL of bisulfite treated WBC gDNA extracted from tonsil tissue from healthy children. The second standard curve was prepared using a 3-fold serial dilution of the 303 or 102 GE/µL of bisulfite treated gDNA extracted from the IMR-32 cell line.  3.2.3 Results The study cohort of this study was composed of children with solid tumors who underwent stem cell collections at BC Children’s Hospital. Patient diagnosis was determined by clinical pathologists using tumor biopsy samples. Metastasis was assessed by one or more of the following procedures: ultrasound, MRI, CT scan, combined PET/CT scan and/or bone marrow biopsy. Matching plasma and stem cell products on the day of the stem cell harvests were obtained from nine children from BCCHB. A list of the patient characteristics appears in Table 3-5. The mean age of the patients at diagnosis was 6 years old (range: 0.4 to 15.8 years old) and the male to female ratio was 6:3. Moreover, six out of the nine patients were also included in the previous part of this study.   78  Table 3-5 Patient characteristics for the association study Patient No. Diagnosis Sex Age 3 neuroblastoma m 2.9 5 neuroblastoma f 8.7 7 neuroblastoma m 4.8 8 neuroblastoma m 3.8 10 glioblastoma f 0.9 13 rhabdoid tumor m 0.4 15 Hodgkin’s lymphoma f 15.8 16 Hodgkin’s lymphoma m 15.6 17 choroid plexus carcinoma m 1.5  As seen in Table 3-6, four out of the nine children studied had evidence of bone marrow involvement of their disease while the status of the remaining five children were unknown because they did not undergo a bone marrow biopsy procedure. Very high total cfDNA and normal cfDNA concentrations were observed in plasma products on the day of stem cell harvests in two patients (Table 3-6). For instance, patient #13 had a total cfDNA concentration of 98247 ± 11822 GE/ml of plasma and a normal cfDNA concentration of 76617 ± 3833 GE/ml of plasma. Patient #17 had a total cfDNA concentration of 117526 ± 6129 GE/ml of plasma and a normal cfDNA concentration of 45409 ± 2467 GE/ml of plasma. A semi-quantitative analysis of possible gDNA contamination reveals that patient #13 and patient#17 samples have gDNA contamination that are between 545 – 1136 GE/ml of plasma and 1136 – 2272 GE/ml of plasma respectively (see Figure 3-19). Tumor cfDNA was detected in six out of the nine children’s plasma products on the day of stem cell harvest. The average detectable tumor cfDNA levels detected was 147 ± 171 GE/ml of plasma (range: 18 to 466 GE/ml of plasma). No tumor gDNA was detected in any of the stem cell products regardless of the presence of tumor cfDNA in the plasma. As of September 2017, four patients are in complete remission (two out of the four 79  patients had detectable tumor cfDNA on the day of stem cell harvest), three patients have residual disease and are undergoing treatment (all three patients had detectable tumor cfDNA on the day of stem cell harvest) and two patients are deceased (one out of the two patients had detectable tumor cfDNA during stem cell harvest).     80   Table 3-6 Total DNA and normal DNA concentrations from plasma and stem cell products from pediatric solid tumor patients  Patient# Diagnosis Bone marrow involvement Survival Total DNA (GE/ml of plasma or GE/106 of cells) Normal DNA (GE/ml of plasma or GE/106 of cells) Plasma Stem cells Plasma Stem cells 3 neuroblastoma extensive on treatment 9807 ± 902 410342 ± 28542 6148 ± 367 741867 ± 38915 5 neuroblastoma yes on treatment 2754 ± 245 334069 ± 12771 928 ± 109 454374 ± 27916 7 neuroblastoma yes on treatment 23742 ± 1431 743510 ± 40664 6590 ± 290 537976 ± 10375 8 neuroblastoma extensive on treatment 20122 ± 583 394467 ± 21918 2749 ± 173 255054 ± 14513 10 glioblastoma not assessed in remission 6149 ± 49 517483 ± 23197 3773 ± 164 758277 ± 1273 13 rhabdoid tumor not assessed deceased 98247 ± 11822 382570 ± 38518 76617 ± 3833 569912 ± 36065 15 Hodgkin lymphoma not assessed in remission 38222 ± 2250 420489 ± 44624 13367 ± 288 318668 ± 24965 16 Hodgkin lymphoma not assessed in remission 21131 ± 2167 748563 ± 30597 7629 ± 251 255118 ± 12800 17 choroid plexus carcinoma not assessed deceased 117526 ± 6129 219345 ± 16011 45409 ± 2467 258344 ± 9791      81                  Figure 3-19 Semi-quantitative gDNA co-isolation check for patient 13 and 17’s cfDNA sample from plasma products collected during stem cell harvest Isolated cfDNA from the plasma products of patient #13 and 17 which were collected on the day of stem cell harvest as well as serially diluted white blood cells (WBC) gDNA from healthy individuals were amplified via PCR using the toll-like receptor 3 (TLR3) gene. The PCR products were visualized on 1% agarose gel and run at 100V for 50 minutes. The 944bp fragment represents the presence of high molecular weight gDNA.82  Table 3-7 Tumor cfDNA yields in plasma and stem cell products Patient # Diagnosis  Bone marrow involvement of disease Tumor DNA Yield Survival Plasma (GE/ml of plasma) Stem cells (GE/106 of cells) 3 neuroblastoma extensive 62 ± 18 BLD residual disease & on treatment 5 neuroblastoma yes 27 ± 7 BLD in remission 7 neuroblastoma yes 100 ± 14 BLD residual disease & on treatment 8 neuroblastoma extensive 466 ± 28 BLD residual disease & on treatment 10 glioblastoma not assessed BLD BLD in remission  13 rhabdoid tumor not assessed 206 ± 15 BLD deceased 15 Hodgkin lymphoma not assessed 18 ± 15 BLD in remission  16 Hodgkin lymphoma not assessed BLD BLD in remission 17 choroid plexus carcinoma not assessed BLD BLD deceased  BLD: below the lower limit of detection83   3.2.4 Discussion High dose chemotherapy and subsequent bone marrow rescue by autologous stem cell transplant is an effective treatment option for children with solid tumors especially for those who have metastatic disease or have relapsed.79 Numerous studies have found evidence of tumor cell contamination of harvested stem cell products with frequencies ranging from 8% up to 100%.77,78,80–84 These findings are in discordance with our study’s findings where we did not find any evidence of tumor gDNA contamination from any of the nine pediatric solid tumor patients in our study. These published studies relied on detection of tumor mRNA using reverse transcriptase PCR (RT-PCR) which has a lower limit of detection of one tumor cell in 106 normal cells.81,83,84 Therefore, their assays are 100x more sensitive than our MS-qPCR assay which may explain the difference in observations in our study and previous studies.   Overall, we did not find any relationship between the tumor gDNA levels in the stem cell product with the tumor cfDNA levels in the plasma product. However, we were able to detect tumor cfDNA in the plasma product collected on the day of stem cell harvest in six patients and its presence was suggestive of negative outcome in which four out of the six patients have either passed away or have residual disease. This finding is in disagreement with previous studies where the researchers found no association with stem cell positivity for tumor contamination and negative outcome77,80,84 However, compared to our study cohort of nine, these studies have large sample sizes ranging from thirty-eight to eighty-eight patients.77,80,84 Therefore, a larger cohort of patients would be needed in order to validate our study’s findings. In our study we also wanted to investigate whether bone marrow involvement of the disease plays a role in tumor contamination of the bone marrow product. Though our 84  quantification assay might not have been sensitive enough to detect tumor contamination of the stem cell product, it was able to detect tumor cfDNA contamination of the plasma product. Our results show that all four patients with known bone marrow involvement of the disease have detectable tumor cfDNA levels at the time of the stem cell harvest. To our knowledge, there has been no other studies reported that analyzed tumor cfDNA contamination of the leukopharesis products. Thus, this finding should be validated in a larger cohort of patients.  We also noticed that two patients with extremely high total cfDNA levels at the time of stem cell harvest have passed away shortly after harvest suggesting that very high levels of total cfDNA are a clinical indicator of poor prognosis. The high total cfDNA levels for patient #13 on the day of stem cell harvest is likely the real level because a similar level was observed for the post G-CSF plasma samples (compare Table 3-2 and Table 3-6). Likewise for patient #17, a high total cfDNA level (110169 ± 7770 GE/ml of plasma) was observed when another aliquot of the plasma on the day of the stem cell harvest was quantified. Hence, the high total cfDNA levels observed in these two patients are real values. Our finding is in line with Mussolin et al.’s observation that high total cfDNA levels are significantly associated with negative outcome.39 However, since our study only has two patients with very high cfDNA levels and it appears that cellular gDNA is present, further studies that involve filtering of the plasma before cfDNA extraction need to be done to confirm the relevance of high total cfDNA levels. One hypothesis is that the high amounts of total cfDNA levels are activating the body’s immune response because cfDNA is a known immune activator.85 Generally, an activated immune response is beneficial because it is capable of protecting the individual from harmful factors such as foreign invaders, harmful substances and cancerous cells. Though, if the body’s immune system is activated for long periods of time or is consistently at heightened levels, it can be harmful to 85  normal cells and tissues as well. The complete blood count (CBC) levels of the two patients on the day of the stem cell harvest do not reveal abnormally elevated WBC counts. However, a CBC is not a sensitive or accurate measure of the immune status of an individual since it does not give functional information of the body’s immune activation.86 Alternatively, a flow cytometry analysis of activation markers in the WBCs’ surface or analysis of the cytokines released by the WBCs would be more informative.86  3.2.5 Conclusion Tumor cfDNA can be detected in the plasma of some children with pediatric cancer at the time of stem cell harvest but its presence does not reflect tumor contamination of the stem cell product. 86  Chapter 4: Overall conclusions 4.1 Summary of research findings This thesis contains preliminary findings from a proof-of-principle research project. In chapter 2, I optimized three critical parts of the research protocol namely the blood sample collection, the total cfDNA isolation, and the tumor cfDNA quantification components of the protocol. We presented data to show that the type of anticoagulant used during the blood collection procedure does not significantly affect the total cfDNA levels. Our results demonstrate that the QIAamp® Circulating Nucleic acid kit which gives statistically similar yields of total cfDNA to Promega’s Maxwell® RSC ccfDNA plasma kit, is contaminated with higher amounts of cellular gDNA. Meanwhile, one tumor cell in the presence of 10000 normal cells is detectable in our MS-qPCR assay which relies on the detection of the promoter methylation of the RASSF1a gene.  In chapter 3, we demonstrated that we can detect tumor cfDNA in the plasma of some children with solid tumors, but not tumor gDNA in the matching stem cell products. Notably, the presence of tumor cfDNA, including at the time of the stem cell harvest procedure, is suggestive of a negative clinical outcome. Overall, four of the seven children (57%) in whom we did not detect any tumor cfDNA from, had a brain tumor diagnosis. This could mean that the tumor cfDNA is present at such low quantities that we could not detect it using our assay, or that promoter methylation of the RASSF1a gene is not present in these children. Alternatively, it is possible that the blood brain barrier is effective at preventing the passage of tumor cfDNA from the brain into the blood. At this time, this study was not able to find any evidence to show that the administration of G-CSF in pediatric solid tumor patients results in increased tumor cell growth. However, the lack of evidence may be confounded by various factors such as exposure 87  to cancer treatment, prior treatment with G-CSF and non-standardized timing of sample collections.   4.2 Strengths This is a pilot study in which we show that tumor cfDNA can be detected in children with solid tumors. G-CSF is used in many pediatric solid tumor patients to either treat neutropenia or to harvest stem cells for autologous transplantation. As some animal studies have shown that G-CSF treatment may cause tumor growth and metastasis, this is the first study looking at the changes in tumor levels pre and post G-CSF treatment using tumor cfDNA as a measurement in human samples. This research project was conducted using a sample of convenience, namely, biospecimens obtained from the BCCHB. All the samples obtained have associated demographic data as well as key clinical information including tumor type and grade, outlines of treatment regimens, and clinical outcome. To minimize the risk of bias, cfDNA samples were quantified in a randomized and blinded fashion. Lastly, a pilot work was performed to determine where variation could come from in our multi-step research protocol and quantify both intra-assay and inter-assay variations. This allowed us to have more confidence in the interpretation of our results.    4.3 Research significance This is a proof-of-principle research project where we showed that tumor cfDNA can be detected in pediatric patients with various solid tumors. G-CSF is widely used clinically to treat chemotherapy-induced neutropenia or to harvest hematopoietic stem cells for autologous transplantation in children with solid tumors. The prior knowledge that mouse studies have 88  shown that G-CSF treatment causes tumor growth and metastasis is concerning. This is the first research study that investigated the changes in tumor levels before and after G-CSF treatment using tumor cfDNA as a measure. Insight into this is important because it can affect the widespread use of G-CSF as a drug and also provide a non-invasive way to assess tumor burden. The latter contributes to ease of sample collection which in turn can increase sample availability, reduce hospital costs, decrease the length of stay of the patients in the hospital and, most importantly of all, improve the quality of life of the patients.    4.4 Limitations  One of the main limitations of this study is the limited number of sample size both to our optimization part of the study and our patient cohort. The latter is due to the rarity of solid tumors in the pediatric population. According to one epidemiology study in Canada over the sixteen year period between 1985 to 2001, there were only 14,601 reported cases of children with a solid tumor diagnosis.87 An additional disadvantage of pediatric studies is the relative small volume of samples readily available. As such, only 1000 µL or less of plasma samples was available for the pre versus post G-CSF comparison portion of this study.  Since this was an observational study, the patients were receiving treatment throughout our sample collection. The treatment regimens given could potentially confound this study’s findings. Another confounding factor in our study is the fact that some patients received prior G-CSF treatment less than one month before the pre G-CSF sample collection was performed. Therefore, for these patients, it is possible that the earlier dosage of G-CSF could have residual effect on the pre G-CSF plasma sample. A more systematic sample collection at defined time points during treatment should be carried out for future studies.  89   Chapter 2 of this thesis revealed that the cfDNA isolation method chosen, the QIAamp® kit, co-isolates large amounts of cellular gDNA with the cfDNA of interest. Unfortunately, the presence of cellular gDNA contamination ultimately prevents us from making definitive conclusions using total cfDNA concentrations.   A technical limitation of this research study involves the quantification method which relies on the detection of promoter methylation of the RASSF1a gene. As chapter 2 demonstrates, one tumor cell is detectable in the presence of 10000 normal cells using this assay. For the samples that showed undetermined values, we cannot conclude that the patient has no tumor cfDNA because it is plausible that the tumor cfDNA is present but is beyond the detection range of our assay. Moreover, because we did not investigate the presence of promoter methylation of the RASSF1a gene in diagnostic tumor tissue samples, we cannot rule out the possibility that the patient actually does not have methylation of the RASSF1a gene.   4.5 Future research directions 4.5.1 Sample collection As discussed earlier, the isolation of cellular gDNA along with the total cfDNA of interest greatly restricts our data interpretation. To address this issue, in future studies, the plasma samples should be filtered prior to storage in order to completely remove any residual normal WBCs present. However, in the rare occasion that WBCs remain after filtering of the plasma, lysis of these WBCs can be prevented by initial collection of the whole blood in Streck cell-free DNA Blood Collection® Tubes. These specialized blood collection tubes contain an additive that stabilizes cells to prevent unwanted cell lysis that leads to an increase in normal 90  gDNA contamination even if the tubes are kept at room temperature for 5 days post collection.88,89 Regarding the timing of sample collections, it is critical that a more standardized approach with definitive time points be put into place. The day of diagnosis where the patient has not been exposed to any treatments nor has been administered G-CSF is the desired time to perform the pre G-CSF sample collection. However, in cases where a patient receives G-CSF treatment substantially after the initial date of diagnosis, a sample on the day of G-CSF treatment just prior first administration could also be obtained. Since G-CSF is administered daily until the absolute neutrophil counts are back to normal or until the stem cell harvest procedure is performed, the post G-CSF sample could be collected every day after G-CSF administration until one week after the last administered dose. This strategy allows for standardization of the sample collection which can reduce variation in the sample quality. Another crucial improvement to the study design would be to obtain diagnostic tumor tissue from each patient enrolled in the study. This would allow for prior testing of the patient’s promoter methylation status for the RASSF1a gene. In this way, if promoter methylation was observed in the tumor tissue but no tumor cfDNA is detected in the plasma samples via the MS-qPCR assay using the methylated RASSF1a primers, then we can rule out the possibility that the patient is among the individuals who does not have promoter methylation of this gene. On the other hand, if promoter methylation was not observed in the tumor tissue, then the DNA from the tumor tissue can be sequenced to identify unique mutations in which patient specific primers can be designed to target.  91  4.5.2 Quantification Out of the total cfDNA measured, the fraction of tumor cfDNA present in a cancer patient can vary from as high as 100% to as low as 0.005%.39,90 In order to accurately quantify low tumor cfDNA levels, in the future, a more sensitive technique could be used such as ddPCR. In ddPCR, DNA molecules are divided into multiple smaller reactions so that some reactions will either contain none or few molecules of DNA.69 Then, the number of reactions with the target DNA molecule present is counted as positive and the number of reactions without the target DNA molecule is counted as negative.69 The Poisson statistics is used to calculate the starting concentration of DNA based on the observed distribution of the positive and negative reactions.69 Compared to the MS-qPCR technique used in this study, the ddPCR technique is able to detect one mutated DNA per 100000 normal DNA.69 Another advantage of ddPCR is it allows absolute quantification of DNA since a standard curve is not required.69     4.5.3 Total cfDNA Even though tumor cfDNA concentrations are important, total cfDNA measurements can also be informative. In chapter 3, we observed that patient #13 and #17 who had extremely high total cfDNA concentrations have since succumbed to their disease. This trend would not have been obvious by assessing tumor cfDNA levels. The possible association between extremely high total cfDNA levels and poor outcome should be further investigated. Unlike tumor cfDNA, total cfDNA can be easily quantified. For instance, total cfDNA levels can be quantified via ddPCR by targeting the human long interspersed nuclear elements 1 (LINE1) fragments.91 Since approximately 520000 LINE1 sequences are present in the human genome, their great abundance can lead to increased sensitivity in quantifying total cfDNA.92  92  References 1. Canadian Cancer Society’s Advisory Committee on Cancer Statistics. Canadian Cancer Statistics 2017. Toronto, ON: Canadian Cancer Society; 2017. Available at: http://www.cancer.ca/~/media/cancer.ca/CW/publications/Canadian Cancer Statistics/Canadian-Cancer-Statistics-2017-EN.pdf. (Accessed: 21st August 2017) 2. Chaddah, M., Willemse, L. & Singh, S. Cancers: Solid Tumor. Stem Cell Network (2013). Available at: http://oirm.ca/sites/default/files/disease-cancer-solid_tumour.pdf. (Accessed: 21st August 2017) 3. Yuan, S. & Wang, S. How do we mobilize and collect autologous peripheral blood stem cells? Transfusion 57, 13–23 (2017). 4. Cotter, S. E., McBride, S. M. & Yock, T. I. Proton Radiotherapy for Solid Tumors of Childhood. Technol. Cancer Res. Treat. 11, 267–278 (2012). 5. Jahr, S. et al. DNA fragments in the blood plasma of cancer patients: quantitations and evidence for their origin from apoptotic and necrotic cells. Cancer Res. 61, 1659–65 (2001). 6. Bronkhorst, A. J. et al. Characterization of the cell-free DNA released by cultured cancer cells. Biochim. Biophys. Acta - Mol. Cell Res. 1863, 157–165 (2016). 7. Pérez-Barrios, C. et al. Comparison of methods for circulating cell-free DNA isolation using blood from cancer patients: impact on biomarker testing. Transl. Lung Cancer Res. 5, 665–672 (2016). 8. Diehl, F. et al. Detection and quantification of mutations in the plasma of patients with colorectal tumors. Proc. Natl. Acad. Sci. U. S. A. 102, 16368–16373 (2005). 9. Raymond, C. K., Hernandez, J., Karr, R., Hill, K. & Li, M. Collection of cell-free DNA 93  for genomic analysis of solid tumors in a clinical laboratory setting. PLoS One 12, e0176241 (2017). 10. Lee, T.-H., Montalvo, L., Chrebtow, V. & Busch, M. P. Quantitation of genomic DNA in plasma and serum samples: higher concentrations of genomic DNA found in serum than in plasma. Transfusion 41, 276–282 (2001). 11. Rago, C. et al. Serial Assessment of Human Tumor Burdens in Mice by the Analysis of Circulating DNA. Cancer Res. 67, 9364–9370 (2007). 12. Krumbholz, M. et al. Genomic EWSR1 Fusion Sequence as Highly Sensitive and Dynamic Plasma Tumor Marker in Ewing Sarcoma. Clin. Cancer Res. 22, 4356–65 (2016). 13. Murtaza, M. et al. Multifocal clonal evolution characterized using circulating tumour DNA in a case of metastatic breast cancer. Nat. Commun. 6, 8760 (2015). 14. Kurihara, S. et al. Circulating free DNA as non-invasive diagnostic biomarker for childhood solid tumors. J. Pediatr. Surg. 50, 2094–2097 (2015). 15. Xu, S. et al. Circulating tumor DNA identified by targeted sequencing in advanced-stage non-small cell lung cancer patients. Cancer Lett. 370, 324–331 (2016). 16. Aung, K. L. et al. Analytical Validation of BRAF Mutation Testing from Circulating Free DNA Using the Amplification Refractory Mutation Testing System. J. Mol. Diagnostics 16, 343–349 (2014). 17. Valenzuela, M. et al. Assessing the Use of p16INK4a Promoter Gene Methylation in Serum for Detection of Bladder Cancer. Eur. Urol. 42, 622–630 (2002). 18. Majchrzak-Celińska, A. et al. Detection of MGMT, RASSF1A, p15INK4B, and p14ARF promoter methylation in circulating tumor-derived DNA of central nervous system cancer 94  patients. J. Appl. Genet. 54, 335–44 (2013). 19. Misawa, A. et al. RASSF1A hypermethylation in pretreatment serum DNA of neuroblastoma patients: a prognostic marker. Br. J. Cancer 100, 399–404 (2009). 20. Chakravadhanula, M. et al. Detection of an Atypical Teratoid Rhabdoid Brain Tumor Gene Deletion in Circulating Blood Using Next-Generation Sequencing. J. Child Neurol. 29, NP81-NP85 (2014). 21. Primerano, S. et al. Kinetics of Circulating Plasma Cell-Free DNA in Paediatric Classical Hodgkin Lymphoma. J. Cancer 7, 364–366 (2016). 22. Yagyu, S. et al. Preoperative analysis of 11q loss using circulating tumor-released DNA in serum: A novel diagnostic tool for therapy stratification of neuroblastoma. Cancer Lett. 309, 185–189 (2011). 23. Hsieh, C.-C., Hsu, H.-S., Chang, S.-C. & Chen, Y.-J. Circulating Cell-Free DNA Levels Could Predict Oncological Outcomes of Patients Undergoing Esophagectomy for Esophageal Squamous Cell Carcinoma. Int. J. Mol. Sci. 17, 2131 (2016). 24. Bettegowda, C. et al. Detection of circulating tumor DNA in early- and late-stage human malignancies. Sci. Transl. Med. 6, 224ra24 (2014). 25. Sutherland, D. R., Stewart, A. K. & Keating, A. The CD34 antigen: Molecular features and potential clinical applications. Stem Cells 11, 50–57 (1993). 26. Rizzo, A. Immune cell markers. Abcam 1 (2016). Available at: http://docs.abcam.com/pdf/immunology/immune-cell-markers-poster.pdf. (Accessed: 7th September 2017) 27. Guthridge, M. A. et al. Mechanism of Activation of the GM-CSF, IL-3, and IL-5 Family of Receptors. Stem Cells 16, 301–313 (1998). 95  28. Fischmeister, G. et al. G-CSF versus GM-CSF for stimulation of peripheral blood progenitor cells (PBPC) and leukocytes in healthy volunteers: comparison of efficacy and tolerability. Ann. Hematol. 78, 117–123 (1999). 29. Viegas, L. Mozobil - A new treatment option for collecting stem cells. Claimsecure (2012). Available at: https://www.claimsecure.com/content/pdfs/en-CA/DrugReviews/DrugReview_Vol11_Issue9_en.pdf. (Accessed: 24th August 2017) 30. Fricker, S. P. Physiology and Pharmacology of Plerixafor. Transfus. Med. Hemotherapy 40, 237–245 (2013). 31. Hatse, S., Princen, K., Bridger, G., De Clercq, E. & Schols, D. Chemokine receptor inhibition by AMD3100 is strictly confined to CXCR4. FEBS Lett. 527, 255–62 (2002). 32. Petit, I. et al. G-CSF induces stem cell mobilization by decreasing bone marrow SDF-1 and up-regulating CXCR4. Nat. Immunol. 3, 687–694 (2002). 33. Saba, F. et al. G-CSF induces up-regulation of CXCR4 expression in human hematopoietic stem cells by beta-adrenergic agonist. Hematology 20, 462–468 (2015). 34. Ayan, I., Kebudi, R., Dogan, S., Tokuç, G. & Görgün, Ö. Granulocyte Colony-Stimulating Factor in Neutropenic, Pediatric Solid Tumor Patients Following Chemotherapy. Pediatr. Hematol. Oncol. 13, 417–424 (1996). 35. Steward, W. P. et al. Granulocyte-macrophage colony stimulating factor (GM-CSF) after high-dose melphalan in patients with advanced colon cancer. Br. J. Cancer 61, 749–54 (1990). 36. Arora, M. et al. Randomized comparison of granulocyte colony-stimulating factor versus granulocyte-macrophage colony-stimulating factor plus intensive chemotherapy for peripheral blood stem cell mobilization and autologous transplantation in multiple 96  myeloma. Biol. Blood Marrow Transplant. 10, 395–404 (2004). 37. Corporation, P. Maxwell RSC ccfDNA Plasma Kit Technical Manual, TM454. 2016 1–9 Available at: https://www.promega.com/-/media/files/resources/protocols/technical-manuals/101/maxwell-rsc-ccfdna-plasma-kit-protocol.pdf. (Accessed: 14th August 2017) 38. Qiagen. QIAamp Circulating Nucleic Acid Handbook 3rd edition. 22–29 (2013). Available at: https://www.qiagen.com/ca/resources/resourcedetail?id=0c4b31ab-f4fb-425f-99bf-10ab9538c061&lang=en. (Accessed: 14th August 2017) 39. Mussolin, L. et al. Plasma cell-free DNA in paediatric lymphomas. J. Cancer 4, 323–9 (2013). 40. Jain, K. K. Textbook of personalized medicine. (Springer Science + Business Media, 2015). doi:10.1007/978-1-4939-2553-7 41. Devonshire, A. S. et al. Towards standardisation of cell-free DNA measurement in plasma: controls for extraction efficiency, fragment size bias and quantification. Anal. Bioanal. Chem. 406, 6499–6512 (2014). 42. Sorber, L. et al. A Comparison of Cell-Free DNA Isolation Kits: Isolation and Quantification of Cell-Free DNA in Plasma. J. Mol. Diagnostics 19, 162–168 (2017). 43. Lam, N. Y. L., Rainer, T. H., Chiu, R. W. K. & Lo, Y. M. D. EDTA is a better anticoagulant than heparin or citrate for delayed blood processing for plasma DNA analysis. Clin. Chem. 50, 256–7 (2004). 44. Gillio-Meina, C., Cepinskas, G., Cecchini, E. L. & Fraser, D. D. Translational research in pediatrics II: blood collection, processing, shipping, and storage. Pediatrics 131, 754–66 (2013). 45. Adamson, P. C., Houghton, P. J., Perilongo, G. & Pritchard-Jones, K. Drug discovery in 97  paediatric oncology: roadblocks to progress. Nat. Rev. Clin. Oncol. 11, 732–9 (2014). 46. Djos, A., Martinsson, T., Kogner, P. & Carén, H. The RASSF gene family members RASSF5, RASSF6 and RASSF7 show frequent DNA methylation in neuroblastoma. Mol. Cancer 11, 40 (2012). 47. Harada, K. et al. Aberrant promoter methylation and silencing of the RASSF1A gene in pediatric tumors and cell lines. Oncogene 21, 4345–4349 (2002). 48. Avigad, S. et al. Aberrant methylation and reduced expression of RASSF1A in Ewing sarcoma. Pediatr. Blood Cancer 53, 1023–1028 (2009). 49. Honda, S. et al. RASSF1A methylation indicates a poor prognosis in hepatoblastoma patients. Pediatr. Surg. Int. 29, 1147–1152 (2013). 50. Wong, I. H. N., Chan, J., Wong, J. & Tam, P. K. H. Ubiquitous aberrant RASSF1A promoter methylation in childhood neoplasia. Clin. cancer Res. 10, 994–1002 (2004). 51. Murray, P. G. et al. Frequent epigenetic inactivation of the RASSF1A tumor suppressor gene in Hodgkin’s lymphoma. Oncogene 23, 1326–31 (2004). 52. Lusher, M. E. et al. Biallelic epigenetic inactivation of the RASSF1A tumor suppressor gene in medulloblastoma development. Cancer Res. 62, 5906–11 (2002). 53. Huynh, K. T. et al. Aberrant hypermethylation in primary tumours and sentinel lymph node metastases in paediatric patients with cutaneous melanoma. Br. J. Dermatol. 166, 1319–1326 (2012). 54. Lázcoz, P. et al. Frequent promoter hypermethylation of RASSF1A and CASP8 in neuroblastoma. BMC Cancer 6, 254 (2006). 55. Michalowski, M. B. et al. Methylation of tumor-suppressor genes in neuroblastoma: The RASSF1A gene is almost always methylated in primary tumors. Pediatr. Blood Cancer 98  50, 29–32 (2008). 56. Wagner, K. J. et al. Frequent RASSF1A tumour suppressor gene promoter methylation in Wilms’ tumour and colorectal cancer. Oncogene 21, 7277–7282 (2002). 57. Ohshima, J. et al. Methylation of the RASSF1A promoter is predictive of poor outcome among patients with Wilms tumor. Pediatr. Blood Cancer 59, 499–505 (2012). 58. Zhang, Y. et al. DNA methylation analysis on a droplet-in-oil PCR array. Lab Chip 9, 1059–1064 (2009). 59. Stutterheim, J. et al. Methylated RASSF1a is the first specific DNA marker for minimal residual disease testing in neuroblastoma. Clin. Cancer Res. 18, 808–14 (2012). 60. Corporation, P. Maxwell RSC Blood DNA Kit Technical Manual, TM419. 1–9 (2015). Available at: https://www.promega.com/-/media/files/resources/protocols/technical-manuals/101/maxwell-rsc-blood-dna-kit-protocol.pdf. (Accessed: 14th August 2017) 61. Qiagen. Epitect Bisulfite Handbook. 16–20, 26–30 (2014). Available at: https://www.qiagen.com/us/resources/resourcedetail?id=9dabebff-fd12-4183-a428-6984c8569fb9&lang=en. (Accessed: 14th August 2017) 62. Balgkouranidou, I. et al. Prognostic role of APC and RASSF1A promoter methylation status in cell free circulating DNA of operable gastric cancer patients. Mutat. Res. 778, 46–51 (2015). 63. Donninger, H., Vos, M. D. & Clark, G. J. The RASSF1A tumor suppressor. J. Cell Sci. 120, (2007). 64. Kawamura, K. et al. Comparison of the efficacy of peripheral blood stem cell mobilization using G-CSF alone from healthy donors and patients with hematologic malignancies. Transfus. Apher. Sci. 49, 334–340 (2013). 99  65. Morales-Arias, J. et al. Expression of granulocyte-colony-stimulating factor and its receptor in human Ewing sarcoma cells and patient tumor specimens. Cancer 110, 1568–1577 (2007). 66. Hsu, D. M. et al. G-CSF Receptor Positive Neuroblastoma Subpopulations Are Enriched in Chemotherapy-Resistant or Relapsed Tumors and Are Highly Tumorigenic. Cancer Res. 73, 4134–46 (2013). 67. Gay, A. N. et al. Granulocyte colony stimulating factor alters the phenotype of neuroblastoma cells: implications for disease-free survival of high-risk patients. J. Pediatr. Surg. 43, 837–842 (2008). 68. Agarwal, S. et al. G-CSF Promotes Neuroblastoma Tumorigenicity and Metastasis via STAT3-Dependent Cancer Stem Cell Activation. Cancer Res. 75, 2566–2579 (2015). 69. Hindson, B. J. et al. High-throughput droplet digital PCR system for absolute quantitation of DNA copy number. Anal. Chem. 83, 8604–10 (2011). 70. Müller, H. M. et al. DNA methylation in serum of breast cancer patients: an independent prognostic marker. Cancer Res. 63, 7641–5 (2003). 71. Schwaederle, M. et al. Detection rate of actionable mutations in diverse cancers using a biopsy-free (blood) circulating tumor cell DNA assay. Oncotarget 7, 9707–9717 (2016). 72. Lavon, I., Refael, M., Zelikovitch, B., Shalom, E. & Siegal, T. Serum DNA can define tumor-specific genetic and epigenetic markers in gliomas of various grades. Neuro. Oncol. 12, 173–180 (2010). 73. Salkeni, M. A. et al. Detection of EGFRvIII mutant DNA in the peripheral blood of brain tumor patients. J. Neurooncol. 115, 27–35 (2013). 74. Hesson, L. et al. Frequent epigenetic inactivation of RASSF1A and BLU genes located 100  within the critical 3p21.3 region in gliomas. Oncogene 23, 2408–2419 (2004). 75. Horiguchi, K. et al. Epigenetic inactivation of RASSF1A candidate tumor suppressor gene at 3p21.3 in brain tumors. Oncogene 22, 7862–7865 (2003). 76. Chen, P. BC Children’s Hospital Pediatric Oncology Hematology Network Newsletter, Apheresis Program. 1–2 (2014). Available at: http://www.bcchildrens.ca/Oncology-Site/Documents/POHNSpring2014_Apr7 final.pdf. (Accessed: 15th August 2017) 77. van Wezel, E. M. et al. Minimal residual disease detection in autologous stem cell grafts from patients with high risk neuroblastoma. Pediatr. Blood Cancer 62, 1368–1373 (2015). 78. Yaniv, I. et al. Tumor cells are present in stem cell harvests of Ewings sarcoma patients and their persistence following transplantation is associated with relapse. Pediatr. Blood Cancer 42, 404–9 (2004). 79. Fraser, C. J. et al. Autologous stem cell transplantation for high-risk Ewing’s sarcoma and other pediatric solid tumors. Bone Marrow Transplant. 37, 175–181 (2006). 80. Ahn, M. J. et al. Clinical Impacts of Tumor Cell Contamination of Hematopoietic Stem Cell Products in Metastatic Breast Cancer Patients undergoing Autologous Peripheral Blood Stem Cell Transplantation: Multicenter Trial. J. Korean Med. Sci. 16, 175 (2001). 81. Burgess, J. et al. Breast tumor contamination of PBSC harvests: tumor depletion by positive selection of CD34(+) cells. Cytotherapy 3, 285–94 (2001). 82. Siehl, J. M. et al. Quantitative real-time RT-PCR detects elevated Wilms tumor gene (WT1) expression in autologous blood stem cell preparations (PBSCs) from acute myeloid leukemia (AML) patients indicating contamination with leukemic blasts. Bone Marrow Transplant. 29, 379–381 (2002). 83. Burchill, S. A. et al. Minimal residual disease at the time of peripheral blood stem cell 101  harvest in patients with advanced neuroblastoma. Med. Pediatr. Oncol. 36, 213–219 (2001). 84. Vermeulen, J. et al. Incidence and prognostic value of tumour cells detected by RT–PCR in peripheral blood stem cell collections from patients with Ewing tumour. Br. J. Cancer 95, 1326–1333 (2006). 85. Zinkova, A., Brynychova, I., Svacina, A., Jirkovska, M. & Korabecna, M. Cell-free DNA from human plasma and serum differs in content of telomeric sequences and its ability to promote immune response. Sci. Rep. 7, 2591 (2017). 86. Kuethe, J. W. et al. Assessing the immune status of critically ill trauma patients by flow cytometry. Nurs. Res. 63, 426–34 (2014). 87. Mezei, G. et al. Socioeconomic Status and Childhood Solid Tumor and Lymphoma Incidence in Canada. Am. J. Epidemiol. 164, 170–175 (2006). 88. Kang, Q. et al. Comparative analysis of circulating tumor DNA stability In K3EDTA, Streck, and CellSave blood collection tubes. Clin. Biochem. 49, 1354–1360 (2016). 89. Medina Diaz, I. et al. Performance of Streck cfDNA Blood Collection Tubes for Liquid Biopsy Testing. PLoS One 11, e0166354 (2016). 90. Diehl, F. et al. Circulating mutant DNA to assess tumor dynamics. Nat. Med. 14, 985–990 (2008). 91. Sunami, E., Vu, A.-T., Nguyen, S. L., Giuliano, A. E. & Hoon, D. S. B. Quantification of LINE1 in Circulating DNA as a Molecular Biomarker of Breast Cancer. Ann. N. Y. Acad. Sci. 1137, 171–174 (2008). 92. Ostertag, E. M. & Kazazian, H. H. Biology of mammalian L1 retrotransposons. Annu. Rev. Genet. 35, 501–38 (2001).102  Appendices Appendix A  Primer and probe sequences for qPCR* Primers and probes Sequences (5’  3’) Size (bp)† GC content (%)† Melting Temperature (°C)† POLR2 Forward CCCAGGTGACATGGAATCTTG 21 52.4 63.8 POLR2 Reverse GCAGAGGCACGTTCAGGAA 19 57.9 65.0 POLR2 Probe FAM-AGCCTTGTGCAGTGGCAGCCACT-TAMRA 23 60.9 73.0 Unmethylated RASSF1a Forward TGTGTTTGTTAGTGTTTAAAGTTAGTGAAGTATG 34 29.4 65.7 Unmethylated RASSF1a Reverse ACACTCCAACCAAATACAACCCTT 24 41.7 65.4 Unmethylated RASSF1a Probe FAM-CACACCCAACAAATACCAACTCCCACAACT-TAMRA 30 46.7 70.4 Methylated RASSF1a Forward GCGTTGAAGTCGGGGTTC 18 61.1 63.4 Methylated RASSF1a Reverse AAACCCGTACTTCGCTAACTTTAAAC 26 38.5 64.5 Methylated RASSF1a Probe FAM-ACAAACGCGAACCGAACGAAACCA-TAMRA 24 50.0 69.7  † The primer and probe sequences were analyzed using the OligoAnalyzer 3.1 tool from the Integrated DNA Technologies website. The parameters used were 50 mmol/L of sodium ion, 4.4 mmol/L of magnesium ion and 1.2 mmol/L of deoxynucleotides.  *The primer and probe sequences for the POLR2 gene were obtained from Mussolin et al. while the primer and probe sequences for the methylated and unmethylated RASSF1a gene were obtained from Stutterheim et al.          103  Appendix B  Primers used for PCR§ Primers and probes Sequences (5’  3’) Amplicon Size (bp) COG5-Forward AAAATTAAGATGCCAACTAACAGG 94 COG5-Reverse CATGGAAGATGATGCACAAGATATA TLR3-Forward TGTCTCTGAGTAACAGCCAG 944 TLR3-Reverse AAGGATGTGGAGGTGAGACA    COG5: Conserved oligomeric Golgi complex subunit 5 TLR3: Toll-like receptor 3     § The primer sequences were obtained from Dr. R. Morin from Simon Fraser University      104   Appendix C  Standard curve for the semi-quantitative analysis of gDNA contamination   Serially diluted white blood cell (WBC) gDNA from healthy individuals were amplified via PCR using the toll-like receptor 3 (TLR3) primers.  The PCR products were visualized in 2% agarose gel and run at 100V for 60 minutes with 1 second exposure 105   Appendix D  Confirmation of promoter methylation status of the RASSF1a gene in the IMR32 cell line and WBCs  Percentage of promoter methylation of the RASSF1a gene (%) IMR-32 cells WBCs 100 0 100 0 99.4 0 99.9 0 100 0 99.9 0 Mean ± Standard deviation 99.9 ± 0.2 0 ± 0    

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0362865/manifest

Comment

Related Items