Exploring DNA Methylation in Tumour-Adjacent Normal Prostate Tissue and Evaluating

its Role as a Biomarker for PCa Detection

by

Carmelle Fatima Cuizon

A thesis submitted in conformity with the requirements

for the degree of Master of Science

Department of Laboratory Medicine and Pathobiology

University of Toronto

© Copyright by Carmelle Fatima Cuizon 2017

 

   

ii  

Exploring DNA Methylation in Tumour-Adjacent Normal Prostate Tissue and Evaluating

its Role as a Biomarker for PCa Detection

Carmelle Fatima Cuizon

Master of Science

Department of Laboratory Medicine and Pathobiology

University of Toronto

2017

Abstract

DNA methylation alterations as a result of field cancerization may be used as diagnostic

biomarkers to improve Prostate Cancer (PCa) detection. To identify diagnostic biomarkers, we

evaluated DNA methylation of KRTAP27-1, APC, HOXD3, CRIP3, KLK6 and RASSF1A in

tumour, tumour-adjacent normal (TAN) and benign prostate tissue. DNA methylation of APC,

RASSF1A and GSTP1 was detected in TAN tissue up to 8 mm from the tumour. KRTAP27-1

was discovered to be hypermethylated in TAN tissue using Global DNA methylation profiling.

Overall methylation concordance of all genes between matched radical prostatectomy (RP)

tumour and RP TAN or biopsy (Bx) TAN tissue was 81% and 66%, respectively.

Hypermethylation of RASSF1A, KRTAP27-1 and HOXD3 in combination were most

significantly associated with PCa (Benign vs. RP TAN: p = 5.87 x 10-10, Benign vs. Bx TAN

cases: p = 4.44 x 10-8). My project demonstrates the benefit of using field cancerization

biomarkers for PCa diagnosis.

 

   

iii  

Acknowledgements

Firstly, I would like to thank my supervisor, Dr. Bharati Bapat for your continued support and

consistent guidance. Your enthusiasm and your reassurance gave me strength to push through

any challenges. You have been both a mother and a mentor towards me. Thank you helping me

grow and succeed in your lab.

Secondly, I would thank my better half, Sam Kohn, for your unwavering love and patience.

You have made it possible for me to strive further, work harder, stay resilient and be a better

human being. My journey as a Master’s Student has been enriched because of you.

Thank you to my parents and family for your love, strong faith and support. You keep me

motivated to work hard and stay focused on pursuing my dreams. Thank you for believing in

me and encouraging me never to give up.

Thank you to my committee members Dr. Theo van der Kwast and Dr. Susan Done for

encouraging me to think beyond the boundaries of my project. Your curiosity and advice

towards my research has enabled me to think more critically and explore different perspectives.

Thank you Ekaterina, Andrea, Julia, Nicole, Fang, Shivani, Madonna, Renu and Richard for

your willingness to provide me with advice and assistance whenever necessary. You truly

made my experience working in the Bapat Lab memorable and worthwhile. Thank you for

teaching me everything that I know about working in the lab. I thoroughly enjoyed

collaborating with you smart and wonderful people. A special thank you to Nicole for your

guidance and getting started on this project. This work would not have been possible without

your significant effort and contribution to this project.

Finally, I would like to thank Prostate Cancer Canada, Ontario Student Opportunity Trust

Funds, Lunenfeld-Tanenbaum Research Institute and the Laboratory Medicine and Pathobiology

Department (University of Toronto) for their financial support.

 

   

iv  

Contributions

 Dr.  Theo  van  der  Kwast,  Dr.  Farshid  Siadat  and  Dr.  Swati  Satturwar  from  the  University  

Health  Network  (Toronto,  Canada)  performed  histological  review  of  tumour,  TAN  and/or  

benign  prostate  tissue  from  radical  prostatectomy  and/or  cystoprostatectomy  cases.    Dr.  

Shuba  Bellur  from  the  University  Health  Network  performed  histological  review  of  TAN  

and  tumour  prostate  tissue  from  biopsy  cases.      

 

Chapter  2  

Dr.  Nicole  White-­‐Al  Habeeb  performed  DNA  extraction  and  bisulfite  modification  of  benign  

and  TAN  prostate  tissue  cases  analyzed  for  methylation  using  the  Infinium  

HumanMethylation450k  Bead  Chip  array.    Dr.  Pingzhao  Hu  (PH)  at  the  University  of  

Manitoba  (Winnipeg,  Canada)  performed  normalization  of  Infinium  

HumanMethylation450k  Bead  Chip  array  data  and  differential  methylation  analysis  to  

identify  differentially  methylated  regions.    PH  also  provided  statistical  advice  on  how  to  

identify  candidate  genes.    Carmelle  Fatima  Cuizon  performed  DNA  extraction,  bisulfite  

conversion  and  DNA  methylation  analysis  for  the  remaining  data  included  in  Chapter  2.      

 

Chapter  3  

A  subset  of  TAN  and  tumour  DNA  cases  from  radical  prostatectomy  tissue  were  extracted,  

bisulfite  converted  and  analyzed  for  the  methylation  of  APC,  RASSF1A,  CRIP3  and  HOXD3  

by  Dr.  Ken  Kron  (former  PhD  student)  and  LiYang  Liu  (former  MSc  student).    DNA  

extraction,  bisulfite  conversion  and  methylation  analysis  of  TAN  and  tumour  DNA  cases  

from  radical  prostatectomy  tissue  were  analyzed  for  the  methylation  of  KLK6  by  Dr.  

Ekaterina  Olkhov-­‐Mitsel  (former  PhD  student).    Julia  Garcia  performed  methylation  

analysis  on  a  subset  of  Bx  TAN  cases.    Carmelle  Fatima  Cuizon  performed  DNA  extraction,  

bisulfite  conversion  and  DNA  methylation  analysis  for  the  remaining  data  included  in  

Chapter  3.    

 

 

   

v  

Table of Contents

Abstract ......................................................................................................................................... ii  

Acknowledgements ...................................................................................................................... iii  

Contributions ............................................................................................................................... iv  

Table of Contents ......................................................................................................................... v  

List of Tables ............................................................................................................................. viii  

List of Figures .............................................................................................................................. ix  

Abbreviations (in alphabetical order) ........................................................................................ x  

Chapter 1 Introduction ................................................................................................................ 1  1.1: The Prostate Gland  .................................................................................................................................  2  1.2: Characteristics of Prostate Cancer  ......................................................................................................  4  

1.2.1 Epidemiology of Prostate Cancer  ....................................................................................................................  4  1.2.2 Common Prostate Cancer Risk Factors  .........................................................................................................  4  1.2.3 Heterogeneity and Multifocality of Prostate Cancer  .................................................................................  5  1.2.4 Classifying Prostate Cancer  ...............................................................................................................................  6  

1.3: Prostate Cancer Screening and Diagnosis  ........................................................................................  12  1.3.1 Digital Rectal Examinations  ...........................................................................................................................  12  1.3.2 PSA Testing  .........................................................................................................................................................  12  1.3.3 Needle Biopsy Examinations for Clinical Diagnosis of PCa  ..............................................................  14  

1.4: Diagnostic Biomarkers for Prostate Cancer  ....................................................................................  18  1.4.1: What are Biomarkers?  .....................................................................................................................................  18  1.4.2 Genetic Diagnostic Biomarkers  ....................................................................................................................  18  1.4.3 Epigenetic Diagnostic Biomarkers  ...............................................................................................................  19  1.4.4 Proteomic Diagnostic Biomarkers  ...............................................................................................................  21  1.4.5 Diagnostic Biomarker Panels  .........................................................................................................................  21  1.4.6 Limitations of the Current Diagnostic Biomarkers  ................................................................................  23  

1.5: Field Cancerization  ...............................................................................................................................  23  1.5.1 What is Field Cancerization?  .........................................................................................................................  23  1.5.2 Epithelial & Stromal Changes in Field Cancerization  ...........................................................................  24  1.5.3 DNA Methylation in Field Cancerization  ..................................................................................................  26  1.5.4 Confirm MDx  ......................................................................................................................................................  26  

 

   

vi  

1.6 Hypothesis, Major Goal and Specific Aims  .......................................................................................  28  1.6.1: Hypothesis  ...........................................................................................................................................................  28  1.6.2: Major Goal  ..........................................................................................................................................................  28  1.6.3: Specific Aims  .....................................................................................................................................................  28  

Chapter 2: Discovery and Validation of Novel Differentially Methylated Genes in Tumour-

Adjacent Normal Prostate Tissue ............................................................................................. 29  2.1 Introduction  ..............................................................................................................................................  30  

2.1.1 Global DNA Methylation Profiling to Identify Diagnostic Biomarkers  .........................................  30  2.1.2 Characterizing the Extent of Field Cancerization  ...................................................................................  31  

2.2 Materials and Methods  ..........................................................................................................................  32  2.2.1 Patient Cohort  .....................................................................................................................................................  32  2.2.2 DNA Extraction and Bisulfite Conversion  ................................................................................................  32  2.2.3 Global DNA Methylation Profiling Using Illumina Infinium HumanMethylation450 Bead

Chip Array  .......................................................................................................................................................................  33  2.2.4 Identifying Candidate Gene-Associated Probes  ......................................................................................  34  2.2.5 Quantitative Methylation-Specific PCR  .....................................................................................................  34  2.2.6 Statistical Analysis  ............................................................................................................................................  37  

2.3 Results  ........................................................................................................................................................  37  2.3.1 Determining the Extent of Field Cancerization affecting DNA methylation in TAN tissues  ..  37  2.3.2 Characterization of the Global DNA methylation profile in TAN Prostate Tissue  .....................  38  2.3.3 Identification of Candidate DNA Methylation Biomarkers for PCa Detection  ............................  39  2.3.4 Technical Validation of Candidate Biomarkers using MethyLight Technology  ..........................  42  2.3.5 Independent validation of KRTAP27-1 using MethyLight Technology  .........................................  44  

2.4 Discussion  ..................................................................................................................................................  44  

Chapter 3: Evaluating the Methylation of Novel and Known genes in Normal, Tumour-

Adjacent Normal and Tumour Prostate Tissue ...................................................................... 49  3.1 Introduction  ..............................................................................................................................................  50  

3.1.1 Selecting Potential Diagnostic Biomarkers for PCa Detection  ..........................................................  50  3.1.2 Differential Methylation in Tumour and TAN Prostate Tissue  ..........................................................  52  

3.2 Materials and Methods  ..........................................................................................................................  53  3.2.1 Patient Cohort  .....................................................................................................................................................  53  3.2.2 DNA Extraction, Bisulfite Modification and Quantitative, Methylation-Specific PCR  ............  53  3.2.3 Statistical Analysis  ............................................................................................................................................  54  

3.3 Results  ........................................................................................................................................................  55  

 

   

vii  

3.3.1 Establishing the Quantitative Concordance of Methylation Between Matched Tumour and

TAN Prostate Tissue.  ...................................................................................................................................................  55  *Out of the 82 Bx cases extracted for DNA, 45 Bx cases had sufficient quality to run on our genes.

However, only up to 38 cases had matched RP Tumour and Bx TAN Methylation data for analysis.

 .............................................................................................................................................................................................  56  3.3.2 Qualitative analysis of methylation patterns between matched tumour and TAN prostate tissue

 .............................................................................................................................................................................................  56  3.3.3 Analyzing the detection ability of candidate gene methylation using Bx TAN prostate tissue

 .............................................................................................................................................................................................  58  3.3.4 Analyzing the detection ability of candidate gene methylation using benign CP and TAN

prostate tissue  .................................................................................................................................................................  59  3.4 Discussion  ..................................................................................................................................................  65  

Chapter 4: Summary of Main Findings and Future Directions ............................................ 70  4.1 Summary of Main Findings  .....................................................................................................................  71  4.2 Future Directions  .......................................................................................................................................  73  

References ................................................................................................................................... 74  

 

 

   

viii  

List of Tables

Chapter 1:

Table 1.1: New Prostate Cancer Grading System

Table 1.2: TNM staging according to the American Joint Committee on Cancer (AJCC) and the

International Union Against Cancer (UICC) as of 2010.

Chapter 2:

Table 2.1: Primer and Probe sequences used for methylation analysis of genes using Methylight

qPCR technology.

Chapter 3:

Table 3.1: Concordance of methylation data between matched tumour and TAN prostate tissue.

Table 3.2: Sensitivity analysis of paired gene biomarker panels in Bx TAN cases from PCa

patients only

Table 3.3: Individual gene methylation analysis between benign CP and TAN cases

Table 3.4: Paired and Multi-gene analysis of gene methylation in benign CP vs. RP TAN cases

Table 3.5: Paired and Multi-gene analysis of gene methylation in benign CP vs. Bx TAN cases

 

   

ix  

List of Figures

Chapter 1

Figure 1.1: The Prostate

Figure 1.2: Gleason Scoring System Schematic

Figure 1.3: Histological representation of cribriform glands (Gleason pattern 4)

Figure 1.4: Needle Biopsy Examinations

Figure 1.5. Field Cancerization in the Prostate

Figure 1.6. Epithelial and stromal cell interactions in Field Cancerization

Figure 1.7. ConfirmMDx Biopsy Examination Schematic

Chapter 2

Figure 2.1: Methylight: Quantitative, methylation-specific qPCR assay

Figure 2.2: Measuring the Extent of Field Cancerization.

Figure 2.3: Global DNA Methylation Profiling of TAN vs. Benign CP Prostate Tissue.

Figure 2.4: Schematic diagrams that represent the location of candidate CpG sites or DMR in

relation to associated gene

Figure 2.5: Validation of our genes of interest

Chapter 3

Figure 3.1: Biomarker Investigation Model

Figure 3.2: Correlation of methylation of genes of interest between RP Tumour vs. Bx Tumour,

RP or Bx TAN

Figure 3.3: ROC curves that reflect the performance of individual or multi-gene biomarker

panels involving benign CP and RP TAN cases

Figure 3.4: ROC curves that reflect the performance of individual or multi-gene biomarker

panels involving benign CP and Bx TAN cases

   

 

   

x  

Abbreviations (in alphabetical order)

4Kscore Four Kalikrein Score

AJCC American Joint Committee on Cancer

AMACR Alpha-methyl-CoA Racemase

APC Adenomatous Polyposis Coli

AUC Area Under the Curve

BCR Biochemical Recurrence

BPH Benign Prostatic Hyperplasia

BRCA2 Breast Cancer Type 2 Susceptibility Protein

Bx Biopsy

C5orf64 Chromosome 5 Open Reading Frame 64

CHEK2 Checkpoint Kinase 2

CNV Copy Number Variations

CP Cystoprostatectomy

CpG Cytosine base located adjacent to a guanine base

CRIP3 Cysteine Rich Protein 3

DMRs Differentially Methylated Regions

DRE Digital Rectal Examination

EPCA Early Prostate Cancer Antigen

ERG Transcriptional Regulator ERG

ETV1 ETS Variant 1

ETV4 ETS Variant 4

FDA Food and Drug Administration

FFPE Formalin-Fixed, Paraffin-Embedded

FOXD1 Forkhead Box D1

Fwer Family-wise error rate

GOLPH2 Golgi Membrane Protein 1

GS Gleason Score

GSTP1 Glutathione S-Transferase Pi 1

H & E Hematoxylin and Eosin

HOXB13 Homeobox B13

HOXD3 Homeobox D3

 

   

xi  

HSD3B2 Hydroxy-Delta-5-Steroid Dehydrogenase, 3 Beta- And Steroid Delta-

Isomerase 2

ISUP International Society of Urological Pathology

KLHDC7A Kelch Domain Containing 7A

KLK6 Kalikrein Related Peptidase 6

KRTAP3-3 Keratin-Associated Protein 3-3

KRTAP27-1 Keratin Associated Protein 27-1

LIMMA Linear Models for Microarray Analysis

miRNAs MicroRNA

Mi-PS Mi Prostate Score

MRI Magnetic Resonance Imaging

MYO15B Myosin XVB

NCCN National Comprehensive Cancer Network

NPV Negative Predictive Value

PCa Prostate Cancer

PCA3 Prostate Cancer Antigen 3

PCPT-CRC North American Prostate Cancer Prevention Trial Cancer Risk Calculator

PHI Prostate Health Index

PMR Percent Methylated Reference

PPV Positive Predictive Value

PSA Prostate-specific Antigen

PSADT PSA Doubling Time

PSAV PSA Velocity

RARB2 Retinoic Acid Receptor Beta

RASSF1A Ras Association Domain Family member 1

ROC Receiver Operating Characteristic

RP Radical Prostatectomy

SNV Single Nucleotide Variants

TAN Tumour-Adjacent Normal

TGFB Transforming Growth Factor Beta

TMPRSS2 Transmembrane Protease, Serine 2

TNM Tumour, Node, Metastasis

 

   

xii  

tPSA Total PSA

TRUS Transrectal Ultrasonography

UICC International Union Against Cancer

VTRNA2-1 Vault RNA 2-1

X2 Chi-Squared Test

 

 1  

Chapter 1 Introduction

 

 

   

2  1.1: The Prostate Gland

The prostate is a walnut-sized gland that mainly functions to nourish the sperm by releasing

prostatic fluid containing acid phosphates, prostaglandins and proteases during ejaculation1–3. It

surrounds the urethra below the bladder and is anterior to the rectum2. The prostate is organized

into an apex, a base, and anterior and posterior surfaces (Figure 1.1). Furthermore, it is

encapsulated by connective tissue that mostly consists of smooth muscle fibers and elastic

connective tissue2.

The prostate is made of fibromuscular stroma and glandular tissue2,4. Fibromuscular stroma is

located beneath the basal lamina that separates it from the glandular epithelial layer4. This

compartment mostly consists of smooth muscle cells, fibroblasts and an extracellular matrix

containing collagen fibers2,4. In contrast, the glandular tissue consists of epithelial tissue located

on the opposite side of the basal lamina and secretes factors such as prostate-specific antigen

(PSA) and phosphatases into the luminal space5. There are 3 zones of the prostate: the

peripheral, central, and transition zone (Figure 1.1) 1–3. The peripheral zone contains about 70%

of the glandular tissue within the prostate. In contrast, the central zone consists of 25% and the

transition zone consists of 5% of the glandular tissue. Prostate cancer (PCa) is most commonly

found within the peripheral zone, while Benign Prostatic Hyperplasia (BPH) is often found in

the transition zone2,3.

 

   

3  Figure 1.1 The Prostate. The prostate is divided into a) the central zone, b) anterior fibromuscular stroma, c) transition zone, d & f) peripheral zone and e) periurethral gland. Modified image adapted from De Marzo et al. Inflammation in prostate carcinogenesis. Nat. Rev. Cancer. 2007, 7(4):256-69.

       

   

     

 

   

4  1.2: Characteristics of Prostate Cancer

1.2.1 Epidemiology of Prostate Cancer

Prostate Cancer is the most common type of non-cutaneous cancer to affect men globally, with a

higher incidence of the disease observed in developed countries6. In Canada, it was expected

that 21,600 men would be diagnosed with the disease in 20167. Although men have a 1 in 8

chance of being diagnosed with PCa in their lifetime, only 1 in 27 men with the disease are

expected to die from PCa. This is due to the majority of PCa cases that are indolent or slow-

growing in nature7. Globally, Canada and the US have the highest incidence of PCa. However,

the incidence of PCa has been declining since 20016,7. This may be associated with the decrease

in PSA testing that had been commonly utilized since the early 1990s, but is currently not

recommended for PCa screening in both countries7–9. As changes to screening techniques and

novel screening technology emerge, this will likely affect the incidence of PCa worldwide.

1.2.2 Common Prostate Cancer Risk Factors

There are various established risk factors that may indicate the presence of PCa in men10–12.

Age is one of the strongest risk factors for having PCa. Men with increasing age are at a higher

risk of having PCa, with the incidence rising rapidly after the age of 7010. Furthermore, men

over the age of 70 have the highest risk of harbouring occult disease12,13. Ethnicity is another

established risk factor for PCa10–12. The incidence of PCa in patients with different ethnicities

varies depending on the geographical location6. Notably, African American men have been

associated with higher stage disease at diagnosis and have the highest incidence of PCa in the

US14. This may be associated with socioeconomic status, limited access to healthcare and late

PCa screening for this population15. Similarly, patients who have a family history of PCa are

also considered to have a higher risk of developing PCa10,12. This depends on the number of

relatives affected with the disease and the relationship of the patient to the affected relative.

Furthermore, known genes explain approximately 35% of the familial risk for PCa16. Germline

mutations in genes such as BRCA2 are often present in families with higher rates of breast and

ovarian cancer17. Although it is relatively rare (1-2% of early onset PCa patients10), men who

are BRCA2 mutation carriers also have a higher likelihood of developing PCa18,19. Other gene

mutations that have also been associated with an increased risk of PCa include HOXB13 and

 

   

5  CHEK2 genes20,21.

1.2.3 Heterogeneity and Multifocality of Prostate Cancer

 Prostate cancer is a highly heterogeneous disease in both disease presentation and clinical

outcome22. Not only can multiple tumours be found within the prostate at one time, but they can

also have distinct genetic, epigenetic and morphological profiles23–27. Some are categorized as

aggressive tumours that grow quickly and can result in metastatic disease and mortality from

PCa. However, the majority of PCa tumours are indolent and do not pose significant harm to

patients. This type of heterogeneity between tumours makes it difficult to predict course of

disease upon biopsy examination. Biopsy examination usually involves sampling of tissue from

multiple locations within the prostate28,29. Often, tumours captured upon biopsy examination

can underrepresent the predicted outcome of disease due to the presence of higher grade

tumours that were not captured in biopsy22. Therefore, the multifocality of PCa limits the

accuracy of pre-treatment diagnosis of PCa.

Different theories have been proposed to explain the development of multiple tumours within

the prostate. One theory suggests that multiple fields or areas within the prostate gain mutations

and other oncogenic properties that function synergistically to sustain the simultaneous growth

of independent tumour populations24. As the size of these lesions grow, they will often fuse to

create a single heterogeneous lesion. Another theory proposes that all tumours originate from a

single cell lineage and can gain the capacity to transform into distinct lesions. This could occur

when tumours acquire various carcinogenic properties that can alter their molecular profile and

potentially promote aggressive tumour development24. It is possible that both theories work

together to promote the multifocality of PCa.

Distinct genetic profiles have been reported in a single prostate further supporting intraprostatic

heterogeneity due to multiple tumours23. Boutros et al. (2015) observed extensive variation in

the amount of copy number variations (CNV), genetic rearrangements and single nucleotide

variants (SNVs) detected in separate prostate tumours from the same patient23. Similar genetic

differences have also been observed between cell populations that originate from a single

tumour lesion25. In a recent study performed by Cooper et al. (2015), distinct genetic variations

 

   

6  in the number and type of mutations and CNVs were observed that separated a tumour into

multiple cell lineages25.

The heterogeneity of PCa also underlies differences in disease progression that are difficult to

anticipate upon diagnosis. Commonly used screening methods and evidence of prostate cancer

in a biopsy cannot reliably predict the presence of aggressive PCa or course of disease. When

low-grade tumours are detected upon biopsy, patients are often placed under programs called

Active Surveillance to monitor the potential progression to aggressive disease30. It has been

reported that approximately one-third of patients diagnosed with low-grade PCa upon initial

biopsy progress (upgrade) to intermediate-grade disease during active surveillance30. Another

significant clinical challenge is determining which PCa patients will experience biochemical

recurrence (BCR) after RP (radical prostatectomy, removal of the prostate) or radiation therapy.

Biochemical recurrence is defined as the rise of PSA levels after RP or therapy to indicate the

presence of prostate tumour cells within the patient. Studies have shown that BCR can range

from 20% to 60% of patients after RP or therapy31,32. Continued investigation of PCa disease

characteristics and heterogeneity will improve our ability to address the clinical challenges

associated with PCa prognosis.

1.2.4 Classifying Prostate Cancer  

1.2.4.1 The Gleason Grading System  The Gleason grading system was initially developed by Dr. Donald F. Gleason in 1966 to

maintain consistency of tumour classification between hospital centers33,34. It is currently one of

the strongest prognostic factors for PCa35,36. The Gleason grading system is used to classify

prostate tumours based on their histological morphology and stratify patients based on their risk

of having indolent vs. aggressive PCa (Figure 1.2). Prostate tumours are stratified into 5

different glandular patterns referred to as Gleason patterns35. Gleason patterns range from 1-5,

and represent morphologies of the prostate gland depending on their differentiation. Lower

Gleason patterns represent well-differentiated glands, more similar to normal prostate

morphology. Higher Gleason patterns represent poorly differentiated glands and indicate high

risk-PCa. The Gleason grading system is synonymous with the Gleason scoring system since a

Gleason score (GS) is assigned to patients to stratify them based on their risk of having low or

 

   

7  high grade PCa. A Gleason Score is assigned to a tumour based on the sum of two most

predominant glandular patterns within the lesion. Generally, tumours assigned with a GS of 6

or less are classified as low-risk, whereas tumours that are assigned a GS of 7 and a GS of 8 or

above are classified as intermediate and high risk, respectively36. Multiple tumours found within

the prostate (may) have different GS and reflect the multifocality of the disease22,27. This makes

it difficult to predict disease progression at an early stage. In recent years, the original Gleason

grading system has been modified to address clinical dilemmas associated with patient risk

stratification.

The most recent International Society of Urological Pathology (ISUP) Consensus Conference

was held in 2014 to discuss and propose changes to the existing Gleason grading system37. The

intention was to make clearer guidelines for assigning GS that would more accurately reflect the

disease prognosis of PCa patients in order to make proper treatment decisions. A significant

modification was made to categorize the GS into 5 Grade Groups (Table 1.1)37–39. This is based

on evidence that PCa patients with GS 7 tumours in which Gleason pattern 4 is predominant in

the area (GS 7 (4+3) have worse prognosis compared to PCa patients with GS 7 tumours in

which Gleason pattern 3 is predominant in the area (GS 7 (3+4))40,41. Furthermore, patients with

GS 8 tumours have better prognosis than GS 9-10 tumours42. As a result, the newly proposed

grade grouping system separates GS 7 (3+4), and GS 8 into their own categories (Table 1.1).

Other guidelines proposed by this new system combines large and small cribriform architectural

patterns under the Gleason pattern 4 category since both morphologies have shown a higher

likelihood of poor outcomes (Figure 1.3)43. In addition, GS 6 tumours are also categorized

under grade group 1 in an effort to clearly reflect its low risk prognosis37,39.

 

   

8  Figure 1.2. The Gleason Scoring System Schematic. Original (left, 1966 & 1967) vs. Modified (right, 2014). Modified image adapted from Epstein et al. The 2014 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma. Am J Surg Pathol 2016. 40 (2): 244-52.

Table 1.1. 2014 Gleason Grading System for Prostate Cancer Tumours. Each Grade Group corresponds to a Gleason score from the original Gleason scoring system.

Grade Group (2014) Gleason Score Histological Definition

1 3+3 = 6 Well-formed glands; glands are not fused

2 3+4 = 7 High composition of well-formed glands with lesser composition of poorly formed

or fused glands

3 4+3 = 7 High composition of poorly formed or

fused glands, lesser composition of well-formed glands

4 8 Poorly formed or fused glands with some areas lacking glands

5 9-10 Mainly composed of areas lacking glands, with a lesser component of

poorly formed or fused glands

 

   

9  Figure 1.3. Histological representation of cribriform glands (gleason pattern 4). Cribriform glands are indicated by the black arrows. Modified image adapted from Epstein et al. The 2014 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma. Am J Surg Pathol. 2016 Feb 40(2):244-52.

1.2.4.2 Clinical and Pathological Staging

The tumour, node and metastasis (TNM) staging system was developed in 1992 by the

American Joint Committee on Cancer (AJCC) and the International Union Against Cancer

(UICC) in order to assist with the stratification of low-, intermediate- or high risk PCa

patients44. It categorizes prostatic tumours based on their extension into surrounding or distant

areas throughout the body (Table 1.2)24,45. Both clinical and pathological TNM staging is

performed on PCa patients to describe their predicted disease prognosis45. Clinical staging is

performed prior to treatment and it reflects evidence of tumour spread revealed through the use

of imaging and other modalities (i.e. transrectal ultrasonography (TRUS), magnetic resonance

imaging (MRI)) or digital rectal examinations (DREs)45,46. In contrast, pathological staging is

often performed after radical prostatectomy, where the whole organ can be assessed

histologically for tumour extension within or outside of the prostate45. TNM staging consists of

a tumour component (T stage) that describes the spread of the tumour within the prostate or into

nearby structures (Table 1.2)44,45. It also consists of lymph node and distant metastasis

component (N and M stage, respectively) that describes whether a tumour has metastasized into

regional lymph nodes or distant organs (Table 1.2). Both clinical and pathological staging uses

the same scoring system. However, a “p” is placed in front of the stage category (i.e. pT2c vs.

T2c) to represent pathological staging.

 

   

10  

Generally, TNM staging is used to help predict disease outcomes and guide treatment

decisions24,45. However, clinical staging poorly predicts pathological stage assigned after RP.

This is mainly due to the multifocality of PCa. As a result, PCa is often under represented by

clinical stage when compared to their final pathologic stage24,27,45. Nomograms such as the

Partin tables and Kattan nomograms combine clinical and pathological parameters (i.e. clinical

stage, biopsy Gleason score, PSA) to develop a score that more accurately reflects the potential

risk of non-organ-confined PCa or pathological stage at RP47,48. These will be discussed in

more detail below. Classification systems that can accurately predict the course of disease for

PCa patients without the need for RP would be ideal for earlier and potentially more effective

treatment.

 

   

11  Table 1.2. TNM staging according to the American Joint Committee on Cancer (AJCC) and the International Union Against Cancer (UICC) as of 2016.

Stage Clinical Pathological

Tx Primary tumour cannot be assessed - T0 No evidence of primary tumour -

T1 (a-c)

Clinically inapparent tumour -­‐ T1a: Tumour found in < 5% of

resected tissue -­‐ T1b: Tumour found in > 5% of

resected tissue -­‐ T1c: Tumour upon needle biopsy

-

T2 (a-c)

Tumour palpable and confined to prostate -­‐ T2a: Tumour spread within < ½ of

one lobe -­‐ T2b: Tumour spread within > ½ of

one lobe, but not both lobes -­‐ T2c: Tumour spread to both lobes

Organ-confined tumour (No sub-classification)

T3 (a-b)

Extraprostatic tumour that does not invade adjacent structures -­‐ T3a: Extraprostatic extension -­‐ T3b: Tumour invades seminal

vesicles

Extraprostatic Extension -­‐ pT3a: Extraprostatic

extensions or invasion into bladder neck

-­‐ pT3b: Seminal vesicle invasion

T4

Invasion of adjacent structures (except seminal vesicles): external spincter, rectum, bladder, levator muscles and/or pelvic wall

Invasion of rectum, levator muscle or pelvic wall

Nx Regional Lymph nodes not assessed Regional lymph nodes not sampled

N0 No regional lymph node metastasis No positive regional lymph nodes

N1 Metastasis in regional lymph nodes Metastasis in regional lymph nodes

M0 No distant metastasis No distant metastasis

M1 (a-c)

Distant Metastasis -­‐ M1a: Non-regional lymph nodes -­‐ M1b: Bone metastasis -­‐ M1c: Other sites with or without

bone metastasis

Distant Metastasis -­‐ M1a: Non-regional lymph

nodes -­‐ M1b: Bone metastasis -­‐ M1c: Other sites with or

without bone metastasis

 

   

12  1.3: Prostate Cancer Screening and Diagnosis

1.3.1 Digital Rectal Examinations

 Digital Rectal Examinations performed by the urologists were the primary screening tests used

for PCa for many years prior to the 1980s49. During this procedure, the urologist will massage

the prostate through the rectum to determine if an abnormal mass can be felt to indicate the

presence of a tumour. DREs are still part of common clinical practice to supplement PSA testing

and contribute to clinical staging46. However, there is low sensitivity and considerable variation

between urologists using this method50. Therefore, DREs are often used in conjunction with

other tests to increase its sensitivity and accuracy of screening51,52.

1.3.2 PSA Testing

 Prostate Specific Antigen (PSA) is a glycoprotein that is released by prostatic epithelial cells to

reduce the thickness of semen and enhance sperm motility1,53,54. In PCa, the tissue barrier that

separates blood vessels from the glandular lumen is disrupted, causing leakage of PSA into the

blood. Therefore, elevated levels detected in the blood stream could indicate the presence of

PCa53. PSA was first discovered in 1970 by Richard J. Ablin and gained widespread clinical use

in the early 1990s to screen for patients that may be at risk of PCa55. The incidence of PCa rose

as a result and caused increased treatment of PCa through surgery or therapy. However, there is

much controversy over the continued use of PSA testing due to its limited specificity51,56,57.

Patients who have serum PSA blood levels between 4.0 – 10 ng/ml are often subjected to further

testing to confirm a diagnosis of PCa58. However, this range of PSA blood levels is also

referred to as the diagnostic grey zone since there are many non-PCa patients that have PSA

levels within this range49. For example, BPH and prostatitis are conditions that can enlarge the

prostate organ, causing PSA levels to rise in the blood stream1. These patients are often

subjected to unnecessary testing, with the added risk of experiencing complications and

infections. Conversely, a subset of PCa patients have low PSA levels (below 4.0 ng/ml),

making it difficult to identify such patients who may benefit from further treatment if their

serum PSA levels are below 4.0 ng/ml59. Finally, PSA testing cannot determine which patients

are likely to progress to advanced stage PCa60. Therefore, patients with indolent disease may

experience overtreatment in attempts to identify those with high grade PCa61. PSA derivatives

 

   

13  and isoforms have been recommended to supplement and improve the specificity of the

longstanding PSA test49,60.

1.3.2.1 PSA Isoforms and Derivatives

Serum or total PSA (tPSA) is a combination of complexed and free PSA1,54. Complexed PSA is

the most common isoform of PSA that is found in the blood. About 70% of PSA is bound to

protease inhibitors such as alpha1-antichymotrypsin62. However, 30% of PSA is detected in its

unbound form called free PSA. Free PSA is further broken down into three isoforms: pro-PSA,

BPH-associated PSA and intact PSA49. Measurement of free PSA has been shown to be useful

in the detection of PCa. In particular, [-2]proPSA, a type of proPSA cleaved between leucine 5

and 6, is found to be elevated in serum in PCa patients and its levels may be able to aid in

detection of PCa prior to initial or repeat biopsy1,54. Furthermore, the ratio of free-to-total PSA

is lower in men with PCa and improves PCa detection in men with tPSA levels between 4-10

ng/ml58,63. Other predictive models have been developed to measure PSA isoforms in

combination to improve the specificity of PCa detection.

PSA velocity (PSAV) and PSA Doubling time (PSADT) are derivatives of total PSA

measurements that consider the changes in PSA levels over time1,60. PSAV has been shown to

be a strong predictor of PCa and is often used to suggest biopsy testing in patients with serum

PSA levels within the diagnostic grey zone. National Comprehensive Cancer Network (NCCN)

recommends its use to suggest biopsy examinations when patients have a PSAV > 0.35

ng/ml/year64. In contrast, PSADT has been suggested as a marker to predict the need for active

therapy prior to treatment. Although it has not shown promise as a diagnostic marker, it may be

useful as a post treatment marker for predicting outcome and high-risk disease.

1.3.2.2 Age-adjusted PSA and PSA Density

As men age, the prostate will naturally grow in size65. This causes an increase in the baseline

levels of PSA, which can affect PSA measurements during PCa screening. Therefore, age-

adjusted PSA cut offs and PSA density measurements have been proposed to account for the

increased prostate volume when screening for PCa using PSA testing66,67. Age-adjusted PSA

cut offs have been shown to increase the specificity of PCa detection and reduce the number of

unnecessary biopsies68. However, many different cut offs have been proposed and further

 

   

14  validation of specific cut off values is necessary to confirm their clinical utility. Similarly, PSA

Density is calculated by dividing the tPSA by the volume of the prostate66. This measurement

takes the size of the prostate into consideration since naturally larger prostates will produce a

larger amount of PSA. This measurement was intended to distinguish between BPH and PCa as

the cause of elevated PSA levels. However, its use as a predictor of PCa at biopsy is not

recommended due to the difficulty in obtaining the volume measurements of the prostate69.

1.3.2.3 Nomograms  Although conventional PSA testing alone may have limited specificity, its clinical performance

may be improved when combined with other clinicopathological factors as part of a nomogram.

Nomograms are predictive tools that have been developed to assess the risk of PCa or high-

grade disease in patients47,48,70,71. Clinicians will often use nomograms to help them decide

whether the patient may benefit from a biopsy exam or a specific type of treatment. Usually,

nomograms combine clinical and pathological information of patients (i.e. family history, age,

positive or negative DRE, TRUS results, GS) to calculate a risk score that can either predict

their risk of detecting PCa upon biopsy or stratify patients based on their risk of having

aggressive disease after a confirmatory biopsy exam. For example, the Kattan nomogram is a

commonly used predictive model that uses preoperative PSA, GS upon biopsy and clinical stage

to predict the risk of biochemical recurrence48. In addition, the North American Prostate

Cancer Prevention Trial-based Cancer risk calculator (PCPT-CRC) was developed to assess the

likelihood of finding PCa at biopsy and their risk of high-grade disease14. Although nomograms

may improve the predictive accuracy over any one factor alone, many types have been proposed

and must be validated in multiple cohorts to assess their value for widespread clinical use.

1.3.3 Needle Biopsy Examinations for Clinical Diagnosis of PCa

1.3.3.1 Needle Biopsy Examination Schemes  Despite the emergence of novel screening tests that show promise in predicting PCa, a needle

biopsy examination is still required to confirm a diagnosis of PCa. Currently, the recommended

needle biopsy procedure involves the use of a TRUS-guided system to remove 10-12 cores from

the prostate (Figure 1.4, A & B)72–74. Although this is considered the gold standard for

diagnosing PCa, there is currently no standardized procedure that is followed when performing

 

   

15  the removal of cores73,74. Hodge and colleagues introduced needle biopsy examinations for PCa

detection in 1989, performing a TRUS-guided sextant biopsy procedure75. Since then, sextant

biopsies are no longer recommended due to the low cancer detection rate of high-grade PCa73.

A 12-core extended biopsy has been shown to increase the detection rate by 20-35% compared

to a sextant biopsy76–78.

In an effort to improve PCa detection upon biopsy, a saturation biopsy procedure has been

proposed, which involves the removal of 14 or more cores from the prostate77–79. The rationale

for this method is based on the thinking that more cores will lead to higher detection rates80.

However, there are conflicting results comparing the performance of saturation vs. a 12-core

biopsy procedure during initial and repeat biopsy77–79,81. Further investigation is necessary to

determine the optimal conditions to use either the extended or saturation biopsy schemes for

diagnosis.

1.3.3.2 Clinical Dilemmas associated with Needle Biopsy Examinations  Although TRUS-guided needle biopsies are recommended for the diagnosis for PCa, there are

many limitations to using this method. When low risk PCa is detected upon biopsy, there is no

way to determine if the patient may harbour aggressive PCa unless RP is performed. Many

patients with initial low-risk disease detected upon biopsy experience disease up grading due to

the under estimation of PCa at biopsy82. In addition, approximately 20-30% of negative biopsy

cases are false negative83,84. Therefore, many patients with a negative initial biopsy result will

undergo a repeat biopsy examination85. This can cause patients with indolent or no PCa to

experience overtreatment, while delaying diagnosis of patients with aggressive PCa.

Patients subjected to a needle biopsy examination are also more likely to suffer from

complications and infections associated with biopsy examinations86,87. Some of these

complications include hematospermia, hematuria, urinary retention and rectal bleeding.

Additionally, increasing antibiotic resistance is speculated to cause hospitalization of patients

due to infections87. Post-biopsy complications can increase anxiety levels in patients, especially

in those who experienced a negative biopsy result88. These clinical issues make it imperative to

find methods to optimize biopsy examinations in order to improve PCa detection and minimize

the need for repeat biopsy procedures.

 

   

16  Figure 1.4. Needle Biopsy Examinations. A) Diagram of the prostate. Dots represent areas that may be sampled using the 12-core systematic TRUS-guided biopsy technique. B) Transrectal ultrasound-guided biopsy. C) Transperineal ultrasound-guided biopsy. B&C: Modified images adapted from Trocazz et al. Medical Image Computing and Computer-Aided Medical Interventions Applied to Soft Tissues: Work in Progress in Urology. Proceedings of the IEEE. 2006. 94(9): 1665–1677.

   

A   B  

C  

 

   

17  1.3.3.3 Transrectal vs. Transperineal Needle Biopsy Examinations  TRUS-guided biopsies are performed using a biopsy gun attached to an ultrasound probe that is

directed into the prostate through the rectum (Figure 1.4 B)29. Patients are placed under local

anaesthesia to allow for the removal of 10-12 cores. Generally, this procedure can be performed

within the clinic in a short period of time. Similarly, transperineal prostate biopsy procedures

can also be performed, although this is a less commonly used method29. This involves the use

of a grid or guiding frame with perforations that guide the biopsy needle through the perineum

to obtain prostate tissue using transrectal ultrasonography as guidance (Figure 1.4 C). With this

method, approximately 20 cores or greater are removed. Although the transrectal approach is

more commonly used, there are benefits to performing transperineal biopsy procedures that may

improve the detection rate of PCa.

One of the main advantages of the transperineal approach is the ability to obtain prostate tissue

in the anterior and apical regions of the prostate29,89. These areas are often left unsampled

during a TRUS-guided biopsy and have a higher likelihood of harbouring cancer. Furthermore,

the needle is directed into the prostate parallel to the urethra. Therefore, there is less likelihood

of perforating the urethra, reducing complications such as hematuria29. Transperineal biopsy

approach can also be an alternative method to minimize the occurrence of infections86,90. Some

disadvantages associated with performing transperineal biopsy examinations are due to the

higher cost of equipment needed, the increased identification of indolent disease, and the

complexity of the procedure29. Since it is less commonly performed, there are fewer urologists

that are familiar with this method and it may take longer to implement this method more

regularly in the clinic. In addition, with the higher detection rate, more patients with indolent

PCa may be identified, leading to unnecessary biopsies and further testing. Although there is

evidence that the transperineal biopsy method may outperform cancer detection using TRUS-

guided biopsy, both methods have advantages and disadvantages and can still miss significant

cancer, leading to false negative results.

1.3.3.4 MRI-guided biopsies  MRI-guided biopsies have been shown to improve detection of PCa due to enhanced imaging of

the prostate91–93. Any suspicion of cancer observed from an MRI result is a strong predictor of

disease. MRI technology can be used alone or in combination with ultrasonography and other

 

   

18  modalities to perform needle biopsy examinations91–93. Multiparametric MRI modalities are

also available and can improve the sensitivity and specificity of detecting PCa94,95. However,

this method is more costly and requires experienced and well-trained urologists to interpret

images accurately96. Additionally, obese patients and the presence of hip prosthetics can lower

the quality of the image, making it difficult to see the organ. More investigation and trials are

necessary to determine when to use this technology during clinical diagnosis.

1.4: Diagnostic Biomarkers for Prostate Cancer

1.4.1: What are Biomarkers?

 A common goal of many PCa diagnostic tests is to accurately identify patients with PCa prior to

or upon biopsy while minimizing the false positive and false negative results. In addition to the

screening tests mentioned above, other ways to address this clinical need is with the use of

biomarkers associated with PCa. Biomarkers are characterized as factors that can be measured

to indicate normal or abnormal biology within a patient, or to indicate the response to

therapeutic interventions97. They can also be categorized as diagnostic, prognostic, predictive

and therapeutic biomarkers to reflect their intended clinical application. Genetic, epigenetic and

proteomic alterations have been investigated in PCa patients, which show strong potential as

diagnostic biomarkers for early detection of disease60,98,99. The most clinically relevant and

useful diagnostic biomarkers are ones that can be detected in biological specimens such as urine,

serum and biopsy tissue, which are obtained through non- or minimally-invasive procedures.

1.4.2 Genetic Diagnostic Biomarkers

1.4.2.1 Chromosomal Rearrangements  Genetic alterations in PCa tumours are often associated with the gain or loss of expression of a

gene100. More specifically, chromosomal rearrangements often occur to dysregulate the

expression of a gene. One of the most common chromosomal rearrangements observed in PCa

involves a deletion on chromosome 21 to create the TMPRSS2:ERG gene fusion101–103.

TMPRSS2 is an androgen-regulated gene that encodes Transmembraine Protease, Serine 2 and

is located upstream of the ERG gene, which is a part of the ETS gene family of transcription

factors. As a result of this gene fusion, ERG expression is increased and causes aberrant

expression of downstream target genes that promote cell motility and carcinogenesis101,103.

 

   

19  TMPRSS2:ERG gene fusion occurs in approximately 50% of PCa cases, mainly due to the

heterogeneity of the disease. TMPRSS2 has also been reported to fuse with other ETS gene

family members such as ETV1 and ETV4103.

TMPRSS2:ERG gene fusion is speculated to occur during early PCa development with evidence

that shows lower levels have been found in more aggressive disease104. Therefore,

TMPRSS2:ERG has been implicated as a diagnostic biomarker for early PCa diagnosis. Studies

have investigated the efficacy of the TMPRSS2:ERG gene fusion to detect PCa patients using

urinary sediments105. Although the specificity of this biomarker is high (93%), the sensitivity is

low (37%)105. Combining the TMPRSS2:ERG biomarker with other promising diagnostic

biomarkers could improve the overall detection rate.

1.4.2.2 Long non-coding RNAs  Long non-coding RNAs are defined by their size, which range from 200 bp up to 100 kb long106.

They are transcribed by RNA Polymerase II and are thought to be regulatory elements that do

not encode proteins. Prostate Cancer Antigen 3 (PCA3) is a long non-coding RNA molecule

that is exclusively expressed in the prostate by the gene DD3107. PCA3 in particular has no

known function and is over expressed in PCa, making it a promising diagnostic marker. In

2012, the Food and Drug Administration (FDA) approved the use of the PCA3 as a diagnostic

test called Progensa PCA3 for men over the age of 50 to aid in the decision to perform prostate

biopsy exams108. The PCA3 test involves the calculation of PCA3 score based on the levels of

PCA3 and PSA mRNA in post-DRE urine. PCA3 score has been reported to outperform the

predictive ability of tPSA to indicate PCa upon biopsy60,109. Patients with a PCA3 score > 25

are advised to undergo initial or repeat biopsy examinations to confirm the diagnosis of PCa109.

1.4.3 Epigenetic Diagnostic Biomarkers

1.4.3.1 DNA Methylation Biomarkers  

DNA methylation is an epigenetic modification that occurs when a methyl group is added to the

5’ carbon of a cytosine base located adjacent to a guanine base (CpG)110,111. Regions rich in

CpG dinucleotides (>50%) are called CpG Islands and are often linked to the promoter region of

genes. It has been reported that changes to the DNA methylation patterns are associated with

 

   

20  the development of various types of cancer112,113. Generally, there is a global loss of

methylation accompanied by gene-specific enrichment of methylation within the promoter

regions of genes, including tumour suppressor genes and oncogenes, respectively. In PCa in

particular, DNA hypermethylation has been extensively studied and hypermethylation of key

genetic players, including GSTP1, APC and RASSF1A, have been reported by many groups114–

117. These genes have been investigated individually and in panels for their ability to serve as

biomarkers for both PCa diagnosis and prognosis 118–122.

The identification of aberrantly methylated genes as potential PCa biomarkers has gained much

interest due to the stability of DNA, which can easily be detected in urine, serum and biopsy

specimens113. GSTP1 is the most commonly hypermethylated gene in PCa tumours and is

found in approximately 90% of all cases123, making it a promising biomarker for PCa detection.

Similarly, promoter hypermethylation of APC, RASSF1A and RARB2 methylation has been

extensively studied and occur frequently in PCa116,124. Multiple hypermethylated genes have

been combined to improve the diagnostic ability of any individual methylation biomarker to

detect PCa121,125,126. Furthermore, genome-wide methylation studies have also been conducted

to identify novel differentially methylated genes outside of promoter regions that may also

improve PCa diagnosis127–129. Due to many methylation biomarkers that have been proposed,

further validation is necessary before implementing them into clinical practice.

1.4.3.2 MicroRNA  MicroRNAs (miRNAs) are small non-coding RNA molecules that are responsible for post-

translational, epigenetic regulation of gene expression130. Typically, they are 18-25 nucleotides

in length and when exported into the cytoplasm, will bind to mRNA transcripts in a

complementary fashion to facilitate their repression. A single miRNA molecule can regulate the

expression of over 200 transcripts130. Therefore, the dysregulation of a single miRNA may have

a significant impact on cancer development.

In PCa, miRNA expression is found to be altered in PCa and has been studied across multiple

biological specimens, such as urine, tissue and serum131. Some of the most commonly

dysregulated miRNAs that have been reported in PCa include let-7a, miR-21, miR-99a, miR-

141, mir-145, miR-200c, miR-221 and miR-375. Many of these miRNAs are observed to be

 

   

21  dysregulated across multiple specimen types. For example, it has been reported that let-7a and

miR-21 is overexpressed and miR-145 is downregulated in both tissue and blood samples from

PCa patients compared to healthy control patients132,133. This characteristic makes them ideal

biomarkers for early PCa detection. Furthermore, multiple miRNAs have also been combined

into biomarker panels to improve the detection ability of individual miRNAs134–136. Although

the area of miRNA biology is relative new, miRNAs serve as a promising source of biological

information that can be used as evidence for PCa diagnosis.

1.4.4 Proteomic Diagnostic Biomarkers

 Protein expression can be altered in PCa through multiple mechanisms such as aberrant miRNA

expression, chromosomal rearrangements, DNA methylation and histone modifications.

Aberrantly expressed proteins have been discovered/proposed as diagnostic biomarkers for early

PCa detection137–139. As mentioned previously, PSA encodes a glycoprotein that is

overexpressed in PCa1,53,54. Although it is not PCa specific, is the most commonly used

biomarker for PCa screening, despite recommendations against its use. GOLPH2 is a golgi

membrane antigen protein that is also observed to be up regulated in approximately 90% of PCa

patients138,140. In addition to its high sensitivity, it can be assayed in urine, making it a promising

protein biomarker for PCa detection141. Other overexpressed proteins in PCa include Alpha-

methyl-CoA racemase (AMACR) and Early Prostate Cancer Antigen (EPCA), all of which

have also been proposed as potential diagnostic biomarkers for PCa detection137–139.

1.4.5 Diagnostic Biomarker Panels

 Many of the PCa biomarkers that were mentioned previously have been studied in combination

to enhance the detection ability of individual biomarkers. When biomarkers are combined, their

diagnostic power is added and may minimize potential false positive and false negative results.

Recently, novel biomarker panels have emerged that show strong diagnostic potential and

outperforms the conventional PSA screening test60. In addition to their intended use for PCa

diagnosis, some biomarker panels can also provide prognostic information prior to performing

any invasive procedures. Moving forward, combining biomarkers with the strongest diagnostic

potential will have the most success in clinical practice to ensure patients are given an early and

accurate diagnosis.

 

   

22  1.4.5.1 Prostate Health Index  The Prostate Health Index (PHI) is a diagnostic test that is used to determine the probability of

detecting PCa upon biopsy in patients with serum PSA levels within the diagnostic grey zone

and a negative DRE60,142,143. The PHI was approved by the FDA in 2012 and calculates a score

based on the levels of tPSA, % fPSA, and -2proPSA. In comparison to its individual PSA

measurements, it has a stronger predictive ability to distinguish between benign vs. malignant

prostatic conditions in men aged 50 or older, since men with PCa are more likely to have a

higher PHI score142,143.

1.4.5.2 4Kscore  Currently, there are no reliable diagnostic tests that can distinguish between low and high-risk

PCa during initial screening. However, the Four-Kalikrein Score (4Kscore) has been developed

to determine the risk of having aggressive PCa prior to having initial or repeat biopsy

examinations144,145. The 4Kscore is calculated based on an algorithm that combines tPSA,

fPSA, intact PSA and human kallikrein 2 levels in serum to create a probability of score of 0-

100%. It will also consider clinical information of the patients, such as age, history of previous

biopsy and positive or negative DRE. The 4Kscore has not been approved by the FDA, but

there is evidence that it has strong potential as a pre-treatment prognostic marker to distinguish

between aggressive vs. indolent PCa60.

1.4.5.3 Mi-Prostate Score  As mentioned previously, the TMPRSS2:ERG gene fusion occurs in approximately 50% of PCa

patients101. Although it has a high specificity, it is limited by its low sensitivity105. However,

when combined with the detection of PCA3, the sensitivity is improved from 62% for PCA3

alone to 73% when combined with TMPRSS2:ERG105. A diagnostic test called the Mi-Prostate

Score (Mi-PS) measures the levels of TMPRSS2:ERG and PCA3 in urine and serum PSA to

create a score that predicts PCa within patients146. Mi-PS is a commercially available test that is

offered at the University of Michigan Health System and is mainly used for early PCa detection

that can aid in determining the need for subsequent biopsy. A recent study reported that the test

has a sensitivity of 80% and a specificity of 90%147.

 

   

23  1.4.6 Limitations of the Current Diagnostic Biomarkers

 Many of the biomarkers mentioned above have been proposed to improve the detection of

PCa98. Ideally, biomarkers that are stable and can easily be detected in biofluids (i.e. urine,

serum, whole blood) are favourable since they can be obtained from patients without performing

invasive procedures. However, many different biomarkers have been proposed and require

extensive validation in large multi-national cohorts before being implemented for widespread

clinical use. Therefore, the current gold standard for PCa diagnosis is a histopathological

examination of a needle biopsy, which is also limited in its ability to detect PCa due to sampling

bias72–74. If PCa is not detected on biopsy, there are currently no reliable diagnostic tests that

can confirm whether patients who received an initial negative biopsy result are positive for the

PCa. Despite the emergence of many promising diagnostic biomarkers for PCa, markers that

can specifically identify patients with PCa upon negative biopsy are warranted.

1.5: Field Cancerization

1.5.1 What is Field Cancerization?

 It has been proposed that the prostatic tissue microenvironment upon which a tumour lesion

grows may develop molecular, genetic and epigenetic aberrations before any histological

evidence of cancer is observed (Figure 1.5)148–151. This process is called field cancerization.

Various molecular changes and environmental carcinogens can interact in the prostate to create

a field of disturbed tissue and stroma149–151. With the right combination of events and cellular

states, the microenvironment can be sufficiently primed to cause the accumulation of other

mutations and alterations within this area24,148,149. This can activate oncogenic pathways and

ultimately promote PCa development. These changes often precede any histological changes

that indicate the presence of disease. In particular, genetic and epigenetic changes have been

observed in tumour-adjacent normal (TAN) prostate tissue as a result of field cancerization150–

152. These have been proposed as potential diagnostic biomarkers, especially in patients with an

initial negative biopsy result120,153. Ideally, genetic and epigenetic biomarkers identified in

negative biopsy tissue can improve PCa screening by allowing for early detection and

minimizing the number of repeat biopsy examinations in patients who are cancer-free.

 

   

24  Figure 1.5. Field Cancerization in the Prostate. Tumours located within the prostate organ are surrounded by histologically normal-appearing tissues affected by field cancerization.

1.5.2 Epithelial & Stromal Changes in Field Cancerization

 The prostate is composed of two main compartments – epithelial tissue and the stroma154,4. In

the context of PCa, the transformation of both compartments may contribute to the development

of tumour lesions. Prostatic epithelial cells will often undergo genetic and epigenetic

transformations that enhance the activity of oncogenic pathways. Stromal cells can also acquire

molecular alterations that promote tumourigenesis. In tumour-associated stroma, there is an

increase in the proportion of fibroblasts and myofibroblasts and a decrease of smooth muscle

cells observed154,4. This is commonly described as Reactive Stroma, since it mimics the

molecular composition of stromal compartments in wounds148,154-5. Normally, stromal and

epithelial cells interact to regulate normal prostate development and differentiation. Therefore,

it is possible that alterations in both compartments can modify their interactions and contributes

to the development of field cancerization (Figure 1.6).

Interactions between prostate stroma and epithelial tissue occur via receptor signalling. For

instance, stromal cells will release growth factors (i.e. fibroblast growth factor, epidermal

growth factor) that are bound by their corresponding receptor located on normal prostatic

epithelial cells to promote normal cell growth5,154. However, the changes that occur in each

compartment can alter their interactions to promote field cancerization and tumour development.

For example, stromal cells release TGFB to inhibit the growth of epithelial cells5. However, if

 

   

25  epithelial cells often lose their ability to express TGFB receptors, they become insensitive to

TGFB. This increases the propensity of cells to grow. With increased levels of TGFB in stromal

cells, this can induce the differentiation of fibroblasts into myofibroblasts, leading to synthesis

of Reactive Stroma that can improve cell motility 5. As a result, this area of tissue has gained the

characteristics that are optimal for tumour spread. In combination with other molecular

aberrations, field cancerization will continue to develop in this area and further promote

tumourigenesis (Figure 1.6).

Figure 1.6. Epithelial and stromal cell interactions in Field Cancerization. When epithelial or stromal cells are exposed to stimuli that alter their genetic, epigenetic or molecular profile, their interactions can cause tumour development. This describes Field Cancerization. Black arrows represent interactions that promote carcinogenesis.

   

 

   

26  1.5.3 DNA Methylation in Field Cancerization

Global loss and gene-specific gain in DNA methylation is a key epigenetic feature of cancer112.

In PCa, DNA hypermethylation has been extensively studied and often occurs in promoter CpG

islands, as mentioned previously114. Interestingly, many hypermethylated genes observed in

PCa tumours have also been observed in the TAN tissue 127,155–157. This supports the idea of

field cancerization where cells within a given tissue microenvironment acquire alterations that

predispose the development of tumours. Therefore, DNA methylation is considered an early

event in carcinogenesis and represents a potential biomarker source that may indicate the

presence of PCa at an early stage.

Differentially methylated genes can easily be interrogated in many specimens, such as plasma,

serum, urine, and tissue158. Furthermore, the hypermethylation of tumour suppressor genes

RASSF1A, APC and GSTP1 that are frequently found in tumours have also been observed in

adjacent normal tissue151,155,157. These characteristics make DNA methylation biomarkers in

TAN tissue ideal for PCa diagnosis, especially if they can be detected in negative biopsy tissue.

It is likely that negative biopsy tissues are removed from TAN prostate tissue. Therefore, it is

possible that differentially methylated genes associated with PCa can indicate the presence of

disease if they can be detected in negative biopsy tissue. This will prevent the need to repeat

biopsy examinations and identify patients who may need more aggressive treatment in a timely

manner.

1.5.4 Confirm MDx

Confirm MDx is a commercially available assay that uses methylation-specific qPCR

technology to detect DNA promoter hypermethylation of GSTP1, APC and RASSF1A in

negative biopsy tissue159–162. This test uses the concept of field cancerization to identify whether

a patient must undergo a repeat biopsy examination (Figure 1.7). If a patient has methylation of

any one of the three genes in any of their negative biopsy cores, they are advised to undergo

repeat biopsy examinations. Due to its high negative predictive value (NPV) (89%), there is a

high likelihood that the ConfirmMDx assay can ensure all patients with no methylation of

GSTP1, APC and RASSF1A are cancer free to avoid unnecessary repeat biopsy procedures.

 

   

27  However, the positive predictive value (PPV), sensitivity and specificity of this test is 29%, 68%

and 64%, respectively159,160. Therefore, approximately 30% of patients with a positive test

result may actually be negative for PCa.

Although ConfirmMDx may lower the number of unnecessary biopsy procedures, there are a

significant proportion of cancer-free patients that may have a positive test result and undergo

unnecessary biopsy procedures. Based on the performance of this test, GSTP1, APC and

RASSF1A may not be exclusively PCa-associated markers and can indicate the presence of

other conditions within the prostate. The addition of more DNA methylation biomarkers to this

panel may improve the performance of this test and prevent false positive results that lead to

overtreatment.

Figure 1.7. ConfirmMDx Biopsy Examination Schematic. A) Map of the prostate that reveals areas where biopsy cores are often taken, missing the tumour focus. B) Map of the prostate reveals biopsy cores that overlap areas around the tumour that are affected by field cancerization. Modified image adapted from mdxhealth.com.

 

A. B.

 

   

28  1.6 Hypothesis, Major Goal and Specific Aims

1.6.1: Hypothesis

 I hypothesize that differentially methylated genes identified in TAN tissue may be used as field

cancerization biomarkers to improve PCa detection and/or prognostication.

1.6.2: Major Goal

 My major goal is to identify promising DNA methylation biomarkers that can be used for PCa

diagnosis in negative biopsy tissues by evaluating the DNA methylation pattern of known and

novel genes in TAN prostate tissues. Newly identified biomarkers may be used in combination

with existing biomarkers to improve PCa and minimize the overtreatment of patients. This will

reveal the importance of field cancerization for clinical assessment of PCa.

1.6.3: Specific Aims

 1. To characterize the DNA methylation profile in TAN prostate tissue in order to identify and

validate novel differentially methylated genes and determine the extent of field cancerization

surrounding prostate tumours.

2. To evaluate the methylation of a selected panel of known and novel genes in matched

tumour, TAN and benign prostate tissue to determine their ability to detect PCa patients.

 

 29  

         

Chapter 2: Discovery and Validation of

Novel Differentially Methylated Genes in

Tumour-Adjacent Normal Prostate Tissue

 

   

30  2.1 Introduction

When patients who are suspected to have PCa undergo a needle biopsy examination,

approximately 25-30% of patients may receive a false negative result83,84. Therefore, patients

with an initial negative biopsy result will often undergo a repeat biopsy examination. This will

cause many patients who do not have PCa to be subjected to unnecessary procedures.

Biomarkers that can be detected in negative biopsy tissue to indicate the presence of PCa are

necessary to minimize false negative results and prevent the overtreatment of patients who do

not have PCa.

As mentioned previously, field cancerization is a phenomenon that causes genetic, epigenetic

and proteomic changes to occur in histologically normal appearing tissue surrounding a

tumour150,151. A common epigenetic alteration that occurs in PCa is DNA methylation, which

has been reported in TAN tissue as a result of field cancerization121,126,127,155,157,163. Negative

biopsy cores from PCa patients are likely to be removed from TAN areas and harbour DNA

methylation changes indicative of carcinogenesis. Therefore, differentially methylated genes in

TAN tissue may be used as biomarkers to detect PCa in negative biopsy tissue. In this chapter, I

characterized the DNA methylation profile in TAN tissue to identify differentially methylated

genes that can potentially be used as biomarkers for PCa detection in negative biopsy tissue.

2.1.1 Global DNA Methylation Profiling to Identify Diagnostic Biomarkers

The commercially available ConfirmMDx assay partially address the dilemma associated with

identifying PCa patients who have received an initial negative biopsy result159,160. With a high

NPV, this assay can make strong assumptions that patients without the methylation of APC,

GSTP1 or RASSF1A in their negative biopsy cores are negative for PCa. However,

approximately 30% of patients with a positive assay result are also negative for PCa and may

still undergo unnecessary biopsy procedures. Therefore, more robust biomarkers that have a

higher sensitivity and specificity are necessary to ensure only PCa patients undergo further

treatment.

 

   

31  Ideal biomarkers for PCa detection in negative biopsy tissue are those that have a higher

methylation signal in TAN compared to benign prostate tissue from non-PCa patients. As

mentioned previously, DNA hypermethylation in PCa is often found in promoter regions that

are rich in CpG sites163. However, it is possible that other genomic regions outside of promoter

regions may also exhibit robust differential methylation as a result of field cancerization.

Therefore, we performed global DNA methylation profiling using the Illumina

HumanMethylation450 BeadChip array164. This array can detect the methylation pattern of over

480,000 CpG sites using a small amount of DNA per sample. CpGs interrogated by this array

are associated with approximately 96% of RefSeq genes and 99% of all CpG islands within the

genome. The benefit of using this array is that it investigates the methylation of CpG sites

located outside of promoter regions, which are less studied in the context of PCa164. By using

this platform to analyze the methylation patterns in TAN vs. benign prostate tissue, we have a

greater likelihood of discovering robust differentially methylated regions or CpG sites across the

genome that may be used as diagnostic biomarkers to detect PCa patients using negative biopsy

tissue.

2.1.2 Characterizing the Extent of Field Cancerization

It is likely that patients with PCa will have negative biopsy tissue taken from TAN areas

affected by field cancerization121,126,127. However, the ability to detect DNA methylation

alterations associated with PCa in negative biopsy tissue depends on how far field cancerization

occurs around the tumour itself and where the biopsy core was removed in relation to this area.

Therefore, characterizing the extent of field cancerization in TAN prostate tissue is necessary to

determine the target area surrounding a tumour upon which DNA methylation alterations can be

detected. Conversely, it will also assist in estimation of the distance where a tumour focus lies

in relation to where the biopsy core was removed. To do this, we evaluated the methylation

patterns of APC, GSTP1 and RASSF1A at increasing distances away from the tumour. APC,

GSTP1 and RASSF1A promoter hypermethylation has been identified in both tumour and TAN

prostate tissue and used as a part of the ConfirmMDx panel of biomarkers161. Therefore, we

analyzed the methylation of these genes in TAN tissues at varying distances from a tumour

focus to indicate the extent of field cancerization.

 

   

32  2.2 Materials and Methods

2.2.1 Patient Cohort

Prostate tissue was obtained from twenty-four patients to perform global DNA methylation

profiling. TAN and benign cystoprostatectomy (CP) tissues from radical prostatectomy (RP) or

benign prostate tissues from CP cases were microdissected for DNA methylation analysis. Eight

patients underwent CP and had no histological evidence of PCa within the prostate. The

remaining 16 patients underwent a RP due to histological evidence of PCa upon biopsy. Among

these, eight patients had low-grade cancer (< GS 6 tumours), and 8 patients had high-grade

cancer: 4 with GS 8 tumours and 4 with GS 9 tumours. To validate the methylation patterns of

our candidate genes in benign CP and TAN prostate tissue, an independent cohort of 69 patients

with benign CP tissue (n = 11) or low- (n = 26) or intermediate- (n = 32) grade prostate tumours

were recruited. This independent cohort represents Cohort 2. An additional cohort of 18 PCa

patients were recruited to determine the distance that altered DNA methylation patterns can be

detected in TAN tissue surrounding a tumour. All patients were recruited from the University

Health Network, Toronto, Canada following Research Ethics Board review and approval.

2.2.2 DNA Extraction and Bisulfite Conversion

Radical prostatectomy and cystoprostatectomy tissue sections from each patient were obtained

from formalin-fixed, paraffin-embedded (FFPE) blocks. One representative block per case was

obtained and cut into 7-10 µm serial sections with up to 5 sections per case. The first slide was

stained with hematoxylin and eosin (H&E) for each case and TAN areas were marked by Dr. T.

van der Kwast. Similarly, H&E slides were prepared for each cystoprostatectomy case and

benign CP areas were marked by Dr. F. Siadat. H&E slides were superimposed over the

unstained serial sections of tissue to mark the TAN or benign CP areas for microdissection.

Extraction was performed using either the QIAmp® DNA Mini Kit (Qiagen, Mississauga, On,

Canada) using a modified protocol165 or the ReliaPrepTM FFPE gDNA Miniprep System

(Promega, Madison, WI, USA). DNA was quantified using Nanodrop 8000 Spectrophotometer

(Thermo Scientific, Wilmington, USA). 100 ng or greater of total DNA was bisulfite-converted

 

   

33  using EZ DNA Methylation Kit (Zymo Research, Irvine, CA, USA) using the protocol

provided.

2.2.3 Global DNA Methylation Profiling Using Illumina Infinium HumanMethylation450 Bead

Chip Array

500-700 ng of bisulfite-converted DNA from each case was sent to The Center for Applied

Genomics at The Hospital for Sick Children Research Institute by Dr. N. White-Al Habeeb (past

Post Doctoral Fellow in the lab) to perform genome-wide methylation profiling using the

Illumina Infinium HumanMethylation450 BeadChip Array. This platform allows for the

simultaneous methylation analysis of over 480,500 CpG using minimal DNA input164. A single

BeadChip can accommodate up to 12 samples for methylation quantification. The quantification

of each CpG site involves the use of two chemical assays: Infinium I and II164. The Infinium I

assay involves the binding of either an unmethylated bead-type or methylated bead-type probe

that will discriminate between methylated or unmethylated CpG sites. The infinium II assay

uses beads that will bind to a region of interest and allow for single-base extension using biotin-

labelled dNTPs. dNTPs will then emit a signal to indicate whether a specific CpG site is

methylated. Fluorescent channels will detect the methylation status of each CpG site

interrogated and quantify the methylation signal based on its intensity. Beta (β) values are used

to quantify the level of methylation within a CpG site and range from 0-1166. β values closer to

0 represent low or no methylation, while β values closer to 1 represent high methylation.

The Illumina GenomeStudio software was used to perform signal normalization and background

subtraction to process the data for further analysis164. Functional normalization and quantile

normalization were both applied to the microarray data separately to remove unwanted

variation, control for batch effects, and remove probes that contribute to background

noise164,167,168. Beta (β) values were then assigned to each probe that corresponds to a distinct

CpG site to quantify their methylation status.

 

   

34  2.2.4 Identifying Candidate Gene-Associated Probes

Differential methylation analysis was performed on the microarray data to identify potential

differentially methylated CpG sites and regions for further validation. Out of the 24 samples

analyzed using the Illumina HumanMethylation450 BeadChip array, only 19 samples could be

used for differential methylation analysis due to poor DNA quality. Linear Models for

Microarray Analysis (LIMMA) and methyAnalysis were applied separately to identify

significant differentially methylated CpG sites between groups after quantile normalization was

performed169,170. In both analyses, β values were log transformed into M values to quantify the

methylation value of each CpG site (M value = log2 (β/1- β)). Bayesian t-tests were used to

compare the methylation of all interrogated CpG sites between TAN and benign CP samples,

and TAN samples stratified by low and high-risk cases. Differentially methylated CpG sites

with an unadjusted (unadj.) p < 0.05 were considered significant. Similarly, Bump hunting

(performed by Dr. P. Hu, University of Manitoba) was also applied to identify significant

differentially methylated regions (DMRs) after functional normalization was performed171.

Regions of the genome that include highly correlated, CpG methylation levels were compared

between groups. DMRs with a family-wise error rate (fwer) < 0.25 were considered significant.

2.2.5 Quantitative Methylation-Specific PCR

The quantitative, methylation-specific PCR technique called Methylight was used to validate the

methylation patterns of our candidate biomarkers and analyze the methylation patterns of known

biomarkers to measure the extent of field cancerization (Figure 2.1)172–174. Primers and probes

were designed for our CpG sites or regions of interest associated with our candidate genes

KRTAP3-3, KLHDC7A, KRTAP27-1 and VTRNA2-1 (Table 2.1). Each tube containing

lyophilised primers and probes were suspended in 1x TE buffer (pH 8.0) to a final concentration

of 100 mM. Oligomix was created using 3 ul probe solution, 9 ul forward primer solution, 9 ul

reverse primer solution and 79 ul H20 for 100 reactions.

Methylight was performed on the QuantStudio 6 Flex Real-Time PCR system (ThermoFisher

Scientific, Waltham, MA, USA). Standard curves were created using positive control super-

methylated DNA at a 1:4 dilution factor. ALU-C4 is used as an input control to normalize the

 

   

35  amount of DNA added175. Percent Methylated Reference (PMR) values were calculated to

quantify the methylation of our genes within each DNA sample165,176. PMR values were

calculated as follows: [(Gene of interest/ALU fluorescence quantity ratio for sample

DNA)/(Gene of interest/ALU fluorescence quantity ratio for super-methylated DNA)] x 100%.

Each DNA sample was analyzed in duplicate.

Figure 2.1 Methylight: Quantitative, methylation-specific qPCR assay. Double-stranded DNA is represented by the horizontal red lines. DNA located within the orange box represents unmethylated DNA. DNA located within the green box represents methylated DNA. Methylated DNA is bound by primers and probes to allow for the amplification of our gene of interest. Taq Polymerase cleaves the fluorescent probe that will emit a fluorescent signal. DNA methylation level of our gene is quantified based on the amount of amplification detected by the amount of fluorescence emitted. Image adapted from Dr. E. Olkhov-Mitsel.

 

   

36  Table 2.1 Primer and Probe sequences used for methylation analysis of genes using Methylight qPCR technology.

Primer/Probe Sequences

KLHDC7A Forward: ATT TGT TTT AAT GAG AGA TTT TGT ATT ACG Probe: CAC AAC AAA ACC TTA AAC CCG CAC TCT CAA Reverse: AAT TAA ACA ACT AAC CCA AAA AAA CG

KRTAP27-1 Forward: TGA AGA ATT TTT TGA TTG GTA CG Probe: CTA CTT ACT CCA ATT CCA AAC CCT ACG AAA A Reverse: ATA TAC AAA ATA ACT ACT TCC CCG

HSD3B2 Forward: AAT CTA AAA AAA CCA AAA AAA CGC Probe: TAT CGG GGA AGA GTT ATT TTT TAT CGG TTT T Reverse: TTT TGA GTT AGT TTG TTT AAA GTC G

C5orf64 Forward: TTT TAG ATT GGT TAA TAG GAG ATC G Probe: GGT CGT GGT CGT TGT ATT TTG TTC GAG A Reverse: ACC CCT ATT CCA AAA ACC G

KRTAP3-3 Forward: GAAAGATTTGGTTTTAGTATTAGTACG Probe: AAC CAC AAA CCC ACC TTA TAC TCC CTC TT Reverse: AAC CAA ACT AAA CGA AAA TAA TTC G

VTRNA2-1 Forward: TGACGATGTGGAGAGGGAC Probe: AGAGGTTTGCGTTATGCGGTTTCGTTCGTTT Reverse: CAC GTT AAC AAA ACG CCT AAC G

GSTP1 Forward: GTC GGC GTC GTG ATT TAG TAT TG Probe: AAA CCT CGC GAC CTC CGA ACC TTA TAA AA Reverse: AAA CTA CGA CGA CGA AAC TCC AA

APC Forward: GAA CCA AAA CGC TCC CCA T Probe: CCC GTC GAA AAC CCG CCG ATT A Reverse: TTA TAT GTC GGT TAC GTG CGT TTA TAT

RASSF1A Forward: ATT GAG TTG CGG GAG TTG GT Probe: CCC TTC CCA ACG CGC CCA Reverse: ACA CGC TCC AAC CGA ATA CG

HOXD3 Forward: TTC GGA ATA GTT AGA GAT TAA AGT GC Probe: ACA AAA CGT TCC CGA CGC TTC TAA AA Reverse: CCA AAA TAA TAA AAC AAA ACG AAA CGA C

KLK6 Forward: AGG AAG TTA TTG ATG TAATCG TTT TTT CG Probe: ACT CCG ACC TCA ACC TCT CTT CCG ACG AAC A Reverse: AAA ACA ATC GAACTT TAT CCG CC

 

   

37  2.2.6 Statistical Analysis

Final PMR values for each gene of interest were determined by calculating the mean PMR of

two replicates. If two replicates had PMR values that were more than 10 % apart from each

other, a 3rd replicate was run and the mean PMR of all three replicates was calculated. Mann-

Whitney U test was performed to compare methylation patterns between two groups. Kruskal-

Wallis Test was performed to compare methylation patterns between 3 or more groups. Paired

analysis was performed using Wilcoxon Ranked Sum test for all matched samples. Statistical

analysis was performed using SPSS software (v. 21, Chicago, Illinois, USA).

2.3 Results

2.3.1 Determining the Extent of Field Cancerization affecting DNA methylation in TAN tissues

Negative biopsy cores removed from the prostate are likely to come from TAN tissues

surrounding a prostate tumour. However, it is unclear how far DNA methylation changes in

TAN tissues can be detected from the tumour focus itself. Therefore, we analyzed the

methylation patterns of known genes at varying distances away from the tumour using

Methylight technology. APC, GSTP1 and RASSF1A methylation has previously been reported

in TAN prostate tissue and is currently used in combination to create The ConfirmMDx

biomarker panel155,157,159. Therefore, we selected these genes to analyze the extent of field

cancerization due to the higher likelihood of detecting their methylation in TAN prostate tissue.

APC, GSTP1 and RASSF1A methylation was analyzed in 24 RP tumour and matched TAN

tissue at 0-4 mm and 4-8mm away from the tumour. The methylation of all three genes was

higher in tumour compared to TAN prostate tissue up to 8 mm away (Figure 2.2, p < 0.05).

Interestingly, a significant decrease in the methylation was observed for GSTP1 in TAN tissue

4-8 mm away from the tumour (Figure 2.2, p = 0.008). In contrast, the methylation levels for

APC and RASSF1A were consistent in TAN tissue up to 8 mm away from the tumour (Figure

2.2, p > 0.05).

 

   

38  Figure 2.2. Measuring the Extent of Field Cancerization. Methylation comparison of APC, GSTP1 and RASSF1A between tumour and TAN tissue at 0-4 mm or 4-8 mm away from tumour (n = 18). PMR values represent the level of methylation for each gene.

2.3.2 Characterization of the Global DNA methylation profile in TAN Prostate Tissue

To explore the genome-wide methylation alterations as a result of field cancerization, we

performed global DNA methylation profiling on TAN and benign CP prostate tissue using

Illumina HumanMethylation450 BeadChip Array. Due to low sample quality, only 19 out of 24

samples were included in the analysis. methyAnalysis identified 1,152 differentially methylated

CpG sites (unadj. p < 0.05, Fold Change (FC) > 2) (Figure 2.3 A). Similarly, LIMMA analysis

identified 3,472 differentially methylated CpG sites (unadj. p < 0.05, FC > 2) (Figure 2.3 B).

Unsupervised hierarchical clustering revealed a distinct methylation pattern between TAN and

benign CP cases for these identified differentially methylated CpG sites. To narrow down our

search for candidate DNA methylation biomarkers, we applied more stringent criteria (unadj. p

< 0.01, FC > 4 and Δβ > 0.2) and selected differentially methylated CpG sites identified by both

LIMMA and methyAnalysis. In total, 934 differentially methylated CpGs were selected.

Unsupervised hierarchical clustering also revealed a distinction between the methylation pattern

of all TAN vs. benign CP cases (Figure 2.3 C).

 

   

39  2.3.3 Identification of Candidate DNA Methylation Biomarkers for PCa Detection

 Out of the 934 differentially methylated CpG sites identified by LIMMA and methyAnalysis, 5

CpG sites were selected based on their robust differential methylation pattern observed (p <

0.01, FC > 4 and Δβ > 0.2). Each CpG site is associated with a specific gene: KRTAP27-1,

KRTAP3-3, HSD3B2, KLHDC7A and C5orf64 (Figure 2.4). LIMMA and methyAnalysis can

only identify differentially methylated CpG sites. To identify DMRs, Bump hunting analysis

was applied to our microarray data. After comparing methylation between benign CP vs. TAN

cases, we decided to focus on the comparison between benign CP vs. all TAN cases or benign

CP vs. TAN GS6 cases. Ideally, this would allow us to identify candidate biomarkers that may

be associated with early PCa development. As a result, 3 DMRs were identified (Figure 2.4,

fwer < 0.25). Two DMRs were identified in the comparison between benign CP vs. all TAN

cases (MYO15B, fwer = 0.224 and uncharacterized DMR 100 kb away from FOXD1, fwer =

0.192, Figure 2.4) and 1 DMR was identified between benign CP vs. TAN GS 6 cases

(VTRNA2-1, fwer = 0.022, Figure 2.4).

 

   

40  Figure 2.3. Global DNA Methylation Profiling of TAN vs. Benign CP Prostate Tissue. A) Top differentially methylated genes from methyAnalysis (un-adjusted p<0.01, FC > 2). B) Top differentially methylated genes from LIMMA Analysis (un-adjusted p<0.01, FC > 2). C) Top differentially methylated genes from both LIMMA and methyAnalysis (unadj. p < 0.01, FC > 4 and Δβ > 0.2) (Benign CP: n = 6, TAN GS6: n = 7, TAN GS89: n = 6).

A.

B.  

A N

C.

 

 

Benign CP

TAN

Legend  

 

   

41  Figure 2.4. Schematic diagrams that represent the location of candidate CpG sites or DMR in relation to associated gene. Black boxes represents amplified region for MethyLight analysis. Green boxes represent nearby CpG islands.

cg05809586

KRTAP27-­‐1

 

KRTAP27-­‐1  Location  of  Gene:  chr21:31,709,331-­‐31,710,012  Candidate  CpG  site:  cg05809586  (chr17:  31,709,691)    Number  of  Exons:  1  Amplified  region  for  Methylight:    chr17:  31,709,630-­‐31,709,751    

VTRNA2-­‐1

 

DMR

 VTRNA2-­‐1  Location  of  Gene:  chr5:  134,416,187-­‐134,416,286  DMR  includes  the  following  CpG  sites:  cg18678645,  cg06536614,  cg26328633,  cg25340688,  cg2686946,  cg00124993,  cg08745956,  cg16615357  (chr5:  135,416,331-­‐135,416,613)  Number  of  Exons:  1  Amplified  region  for  Methylight:    chr5:1  35,416,328-­‐135,416,426    

KLHDC7A

cg23688827

 

KLHDC7A  Location  of  Gene:  chr1:  18,807,424-­‐18,812,480  Candidate  CpG  Site:  cg23688827  (chr1:  18,767,776)    Number  of  Exons:  1  Amplified  region  for  Methylight:    chr1:  18,767,716-­‐18,767,837    

cg17951978

     

HSD3B2

HSD3B2  Location  of  Gene:  chr1:  119,957,270-­‐119-­‐965,662  Candidate  CpG  Site:  cg17951978  (ch1:  119,960,008)  Number  of  Exons:  3  Amplified  region  for  Methylight:    chr1:  119959947-­‐119960067    

 

cg13634242

     

C5orf64

 

C5orf64  Location  of  Gene:  chr5:  60,920,228-­‐60-­‐923,894  Candidate  CpG  Site:  cg13634242  (chr5:  60,921,891)  Number  of  Exons:  3  Amplified  region  for  Methylight:    chr5:  60,921,831-­‐60,921,952    

KRTAP3-­‐3

cg13183651

 

KRTAP3-­‐3  Location  of  Gene:  chr17:39,149,682-­‐39,150,385  Candidate  CpG  Site:  cg13183651  (chr17:39,151,857)  Number  of  Exons:  1  Amplified  region  for  Methylight:    chr17:39,151,797-­‐39,151,918    

     

         

 

FOXD1  gene  DMR Uncharacterized  DMR  

Nearest  gene:  FOXD1  (chr5:72,742,085-­‐72,744,352)  DMR  includes  the  following  CpG  Sites:  cg11176481,  cg01558931  (chr5:72,676,039-­‐72,676,121)    

 

   

42  

2.3.4 Technical Validation of Candidate Biomarkers using MethyLight Technology

Technical validation of our genes of interest was performed on TAN (n = 11) and benign CP (n

= 6) samples from the array using MethyLight technology. Primers and probes were designed

to interrogate the methylation patterns of KRTAP27-1, KRTAP3-3, HSD3B2, KLHDC7A,

C5orf64 and VTRNA2-1 within the CpG site or region of interest. Unfortunately, the

primer/probe sequences for C5orf64 and HSD3B2 were unable to amplify our region of interest

and could not be used for qPCR methylation analysis. This may be due to non-specific binding

or inability to bind due to unmethylated CpG sites within the primer or probe sequence.

Comparative analysis was performed between benign CP and all TAN cases for KRTAP27-1,

KRTAP3-3, VTRNA2-1 and KLHDC7A (Figure 2.5 A&B). Methylation of KRTAP27-1 and

KLHDC7A was significantly elevated in TAN cases (Figure 2.5 A, p = 0.028 for KRTAP27-1,

p = 0.02 for KLHDC7A). After stratifying TAN cases by low and high-risk disease, KRTAP27-

1 and KLHDC7A methylation in TAN GS 6 (n=6) was significantly increased compared to

benign CP cases (Figure 2.5 B). In contrast, methylation of KRTAP3-3 and VTRNA2-1 was

similar between benign CP and TAN cases (Figure 2.5 A&B, p > 0.05).

MYO15B  (26  Exons)  

 

 DMR MYO15B  Location  of  gene:  FOXD1  (chr17:73,584,139-­‐73,609,552)  DMR  includes  the  following  CpG  Sites:  cg00204465,  cg12184886,  cg06390536,  cg22287064  (chr17:  73,583,945-­‐73,584,070)  Number  of  Exons:  26    

 

   

43  Figure 2.5. Validation of our genes of interest. A & B) Technical validation of our candidate CpG sites or DMRs (Benign CP n=6, TAN GS6 n = 6, TAN GS89 = 5). C & D) Independent validation of KRTAP27-1 and methylation analysis of APC, GSTP1, HOXD3 and RASSF1A. PMR values represent the level of methylation for each gene. (APC, GSTP1, RASSF1A, HOXD3: Benign CP n < 17, TAN n < 66; KRTAP27-1: Benign CP n = 12, TAN GS6 n = 25, TAN GS7 n = 34).

A B  

C D  

 

   

44  2.3.5 Independent validation of KRTAP27-1 using MethyLight Technology

In addition to these novel genes, we decided to measure the methylation patterns of

known/established gene methylation markers HOXD3, APC, GSTP1 and RASSF1A in all

samples analyzed by the array and from Cohort 2 that consists of TAN (up to 66 samples) and

benign CP (up to 17 samples). Except for HOXD3, methylation of these known genes in TAN

tissue has been reported in the literature155,157,165. As mentioned previously, APC, GSTP1 and

RASSF1A constitute the gene panel in the ConfirmMDx assay. We observed a significant

increase in the methylation of RASSF1A in TAN cases (n = 64) compared to benign cases (n =

16) (Figure 2.5 C, p = 0.002). When TAN cases were stratified by GS, HOXD3 and RASSF1A

methylation was significantly elevated in TAN GS 6 cases (Figure 2.5 D, n = 30, p = 0.007 and

n=31, p = 0.004, respectively), while RASSF1A also showed significantly higher methylation in

TAN GS 7 cases (Figure 2.5 D, n = 33, p = 0.007). An increasing trend of APC methylation in

TAN GS 6 cases (n = 34) was also observed. However, the difference was not statistically

significant (Figure 2.5 D, p = 0.066). Surprisingly, there was a significant decrease in GSTP1

methylation in TAN GS 6 compared to benign CP cases (Figure 2.5 D, p = 0.04).

Independent validation of KRTAP27-1 was performed on Cohort 2, which included benign CP

cases (n = 12) and TAN cases (n = 59) from patients with low- (GS 6, n = 25) and intermediate-

(GS 7, n= 34) grade PCa that were not analyzed on the array. Comparative analysis revealed no

difference between methylation patterns of KRTAP27-1 in benign CP vs. TAN cases in this

cohort (Figure 2.5 C&D, p > 0.05). When stratified based on Gleason score, the difference is not

statistically significant.

2.4 Discussion

In this chapter, we established that TAN prostate tissue at a distance of up to 8 mm away from

the tumour focus harbours DNA methylation changes that can be detected using methylation-

specific qPCR. APC and RASSF1A displayed similar methylation profiles between TAN

prostate tissue 0-4 and 4-8 mm away from a tumour. In contrast, GSTP1 methylation

significantly decreased at increasing distances away from the tumour. Most importantly, these

results revealed that APC, RASSF1A and GSTP1 are methylated in both tumour and TAN

 

   

45  tissues, despite the significant difference in methylation levels. These findings suggest that

negative biopsy tissue taken within this distance from a tumour focus may harbour differentially

methylated genes that can potentially be used for clinical diagnosis of PCa, even when the

histologic diagnosis does not show evidence for cancer.

Due to the limited size of the RP specimens examined, we were unable to obtain enough cases

with TAN areas greater than 8 mm from a tumour. Beyond this distance, it is possible that the

presence of other tumours can influence then methylation signals for our genes of interest.

Furthermore, our TAN cases were micro dissected from 2-dimensional RP specimens. Other

tumours could also be present above or below the interrogated region in 3-dimensional space,

influencing the methylation patterns we observed in the TAN tissues. Although we were able to

detect DNA methylation alterations up to 8 mm away from a tumour, this may be caused by

field cancerization surrounding more than one tumour.

We also observed a distinct methylation signature that separates TAN and benign CP prostate

tissue. This provides evidence of field cancerization altering DNA methylation patterns in TAN

prostate tissue. Three different analytical approaches were used to identify 8 differentially

methylated CpG sites and regions to ensure that the most robust candidate biomarkers would be

identified. A total of 5 CpG sites and 3 DMRs were identified based on stringent criteria

mentioned above. Each CpG site of interest was associated with one gene. Both KRTAP3-3

and KRTAP27-1 encode Keratin-associated proteins that are involved in maintaining the

structural integrity of hair cells177. HSD3B2 (hydroxyl-delta-5-steroid dehydrogenase) encodes

a protein that converts cholesterol to progesterone178,179. The function of both KLHDC7A

(Kelch-domain-containing protein 7A family member) and C5orf64 has not been characterized.

Three DMRs were identified using Bump hunting analysis. However, due to the inability to

design reliable primers and probes for qPCR analysis of MYO15B and an uncharacterized

DMR, we performed technical validation of VTRNA2-1 only. VTRNA2-1 is a non-coding

RNA that functions as a tumour suppressor and represses cellular Protein Kinase RNA and

indirectly promotes cell death180–182. A region containing 8 CpG sites located in the promoter

region of VTRNA2-1 was significantly hypermethylated in TAN GS6 cases. The function of

MYO15B and the uncharacterized DMR are unknown. Understanding the functional purpose of

genes differentially methylated in TAN tissues could suggest how their methylation pattern may

 

   

46  be affecting their expression and contributing to PCa development. However, genes that are

differentially methylated in TAN tissues that can be used to identify PCa patients are most

relevant, regardless of their biological function.

Our inability to validate VTRNA2-1 and KRTAP3-3 methylation using methylation-specific

qPCR could be due to a number of reasons. Firstly, the primer and probe design often requires

binding to additional CpG sites near our candidate CpG to allow for optimal amplification of the

region. Consequently, the methylation levels of these additional CpG sites may have affected

the overall methylation patterns we observed between groups using MethyLight. Secondly,

technical validation was performed on a relatively small number of array cases (n = 19). As a

result, the variability in methylation levels between samples greatly influenced the results of our

analysis. A larger cohort of samples is necessary to minimize the effects of variability and

accurately compare the methylation patterns between benign CP cases and TAN cases.

Although the methylation of KLHDC7A in TAN cases was significantly higher compared to

benign CP cases, the overall methylation levels for both groups were very high (PMR > 100%).

This may be due to incomplete methylation of super-methylated DNA (positive control),

causing highly methylated genes to have PMR values over 100%173. If used as a biomarker, it

may be difficult to identify patients with PCa if non-cancer patients also exhibit high

methylation of this CpG site. Therefore, only the independent validation of KRTAP27-1 was

performed on Cohort 2 using MethyLight technology. Since we are interested in investigating

its potential as a biomarker for early and accurate PCa diagnosis, we only analyzed benign,

TAN GS6 and TAN GS7 cases in our independent cohort. Methylation patterns of known genes

(GSTP1, RASSF1A, HOXD3 and APC) were also analyzed in both the array cohort and Cohort

2. Since methylation of these genes has previously been reported in TAN or matched normal

prostate tissues155,157,165, detecting similar methylation patterns in our cohorts can confirm the

validity of our biomarker validation method. Our independent validation of KRTAP27-1

methylation revealed comparable methylation levels between benign CP and TAN cases taken

together and stratified by GS. This may be due to the difference in the proportion of TAN cases

studied in this cohort. In the array and the technical validation, TAN tissue was taken from

patients with GS 6, GS 8 and 9 tumours. However, Cohort 2 consisted of TAN tissue from

patients with GS 6 and GS 7 tumours. The differences in the two cohorts may have contributed

 

   

47  to discordant results observed in our study. Furthermore, we observed high KRTAP27-1

methylation in the benign CP cases. Interestingly, GSTP1 methylation was also significantly

elevated in benign CP cases compared to TAN GS 6 cases. This was also reported in a study

performed by Kwabi-Abdo et al. (2007)183. Therefore, we speculate that the elevated

methylation pattern of GSTP1 and KRTAP27-1 in benign CP cases may be due to the abnormal

pathology of each patient within the benign CP group.

Cases included in the benign CP cohort were microdissected from cystoprostatectomy

specimens obtained from patients who have high-grade urothelial carcinoma, although the

prostate tissue remains histologically normal. Since organs surrounding the prostate have

evidence of carcinoma, it is possible that the benign CP prostate tissue analyzed may also be

affected by field cancerization, altering the DNA methylation profile within benign prostate

tissue. For instance, none of the CpG sites interrogated on the array were significantly

differentially methylated according to LIMMA and methyAnalysis when using adjusted p-value

criteria (corrected for multiple comparisons) to filter candidate genes. Therefore, we were

limited to the use of unadjusted p-values to identify differentially methylated CpG sites or

regions. It is possible that the methylation patterns of these benign CP cases may not reflect the

true methylation profile of healthy prostate tissue. As a result, the extent of differential

methylation observed between groups is subtle and revealed higher-than-expected methylation

of KRTAP27-1 and GSTP1 in the benign CP cohort. Unfortunately, it is difficult to obtain

normal prostate tissue from healthy subjects and remains a limitation of this project.

Future studies could be performed using pyrosequencing or bisulfite sequencing strategies for

technical and independent validation of novel genes identified using the Illumina 450K

methylation microarray184–186. Both of these methods can interrogate and measure the

methylation pattern of a specific CpG site and consequently, provide more accurate validation

results. Also, the use of healthy prostate tissue from healthy male donors may serve as a better

control group and provide a more accurate comparison between cancer and non-cancer

associated normal tissue. Unfortunately, these specimens may be more difficult to come by.

Furthermore, our CpG site of interest associated with KRTAP27-1 was identified based on

significant hypermethylation observed in a cohort of TAN cases surrounding low (GS6) and

high (GS89) grade tumours. Independent validation using a larger series with TAN cases

 

   

48  surrounding low (GS6) and high (GS89) risk tumours and normal prostate tissue from healthy

patients may confirm the association between PCa and KRTAP27-1 hypermethylation at our

CpG site of interest.

 

 49  

                             

Chapter 3: Evaluating the Methylation of

Novel and Known genes in Normal, Tumour-

Adjacent Normal and Tumour Prostate

Tissue

 

 

   

50  3.1 Introduction

In the previous chapter, we identified and validated the hypermethylation of a CpG site within

an exon of the KRTAP27-1 gene. Although independent validation revealed similar

methylation patterns between TAN and benign CP cases, hypermethylation of this gene may be

useful as a diagnostic marker if used in combination with other genes. Many groups have

developed diagnostic tests that combine multiple biomarkers to improve the detection of PCa143–

146. For example, the 4K score measures the levels of tPSA, fPSA, intact PSA and human

kallikrein 2 and calculates a score that reflects the probability of detecting high-risk PCa upon

initial or repeat biopsy144,145. Similarly, Mi-PS test combines TMPRSS2:ERG, PCA3 and PSA

transcript levels, while the PHI test combines tPSA, %fPSA and -2proPSA to evaluate the risk

of having PCa based on strong evidence that these biomarkers perform optimally as a panel for

PCa detection143,146. Unfortunately, none of these diagnostic tests can predict the presence of

PCa in patients who received an initial negative biopsy result. Therefore, we evaluate the

methylation of our genes individually and in combination to determine their efficacy as potential

biomarkers for PCa detection in negative biopsy tissue.

3.1.1 Selecting Potential Diagnostic Biomarkers for PCa Detection

Genes reported to be hypermethylated in PCa that can discriminate between PCa and healthy

patients are the most promising diagnostic biomarkers for PCa detection. APC is a tumour

suppressor gene that functions as a regulator of the Wnt signalling pathway through degradation

of beta-catenin124. Similarly, RASSF1A is also a tumour suppressor gene that regulates cell

cycle progression by controlling the transition from G1 to S phase116. Although both genes are

known to be hypermethylated in PCa tumours, they are currently used as field cancerization

biomarkers to discriminate between PCa and non-PCa patients in negative biopsy tissues161.

This suggests that other differentially methylated genes in prostate tumours may also be useful

as field cancerization biomarkers for PCa detection. In this chapter, we analyze the methylation

of KRTAP27-1, HOXD3, KLK6 and CRIP3 in addition to APC and RASSF1A to evaluate their

efficacy individually and in combination as field cancerization biomarkers to improve PCa

diagnosis in negative biopsy tissue.

 

   

51  KRTAP27-1 was the most promising biomarker identified in our discovery analysis between

TAN and benign CP tissues. Using two separate techniques, we were able to verify the

differential methylation results between the two tissues. Therefore, we decided to further

examine the methylation of KRTAP27-1 in combination with other potential diagnostic

biomarkers. HOXD3 and KLK6 were also selected for further investigation as field

cancerization biomarkers due to previous evidence of their ability to distinguish between benign

and PCa patients165,187. HOXD3 is a transcription factor that is a part of the Homeobox (HOX)

gene family188. Previous studies in our lab revealed that HOXD3 is significantly

hypermethylated in tumour vs. matched normal prostate tissue and may have prognostic value in

identifying patients with BCR165. Additional studies in our lab also revealed that KLK6

methylation is elevated in tumours compared to matched normal prostate tissue187. KLK6 is a

serine protease that is part of the Kalikrein family of genes. Interestingly, a positive association

between tumour and matched normal methylation was also observed, which suggests that KLK6

methylation may be associated with field cancerization. Although differential methylation was

observed between tumour and TAN prostate tissue, these genes have not been studied in the

context of field cancerization and may have potential roles as biomarkers for PCa detection in

negative biopsy tissue.

Our lab previously discovered hypermethylation of CRIP3 in prostate tumours and observed its

progressive association with increasing stages of PCa. The biological function of CRIP3 is

currently unknown. Recently, our lab investigated CRIP3 methylation in post-DRE urinary

sediments from low risk (GS 6) PCa patients monitored by Active Surveillance189. CRIP3

methylation in combination with methylation of APC, GSTP1 and HOXD8 in urinary sediments

of these patients shows promise as a biomarker classifier panel to predict reclassification of PCa

patients for increased risk of cancer progression. This is the first publication that revealed the

significance of CRIP3 methylation as a PCa biomarker. However, no studies have been

conducted to evaluate CRIP3 methylation in the context of field cancerization. Therefore, we

were interested in investigating CRIP3 methylation in tumour, TAN and benign prostate tissue

to determine if it has diagnostic value for PCa detection in combination with other genes of

interest.

 

   

52  Figure 3.1. Biomarker Investigation Model. RP or Bx cases investigated were analyzed for the concordance of methylation with RP Tumour data. All Bx and RP cases analyzed are matched, which is reflected by the red diagonal arrows.

3.1.2 Differential Methylation in Tumour and TAN Prostate Tissue

 Tumour and TAN prostate tissues have been reported to share similar DNA methylation patterns

compared to tumour and normal prostate tissue from healthy patients163. Two studies have

revealed that many of the genes that were differentially methylated in tumour tissues were also

found in TAN tissues, but absent in normal prostate tissue127,163. This provides strong evidence

that field cancerization that promotes tumour development also alters the methylation profile in

TAN tissue in a similar fashion. Genes found to be differentially methylated in both tumour and

TAN tissues would be ideal biomarkers for PCa detection in cancer-negative biopsy tissue. In

this chapter, we follow our Biomarker Investigation Model (Figure 3.1) to evaluate the

concordance of methylation between matched tumour and TAN tissue for our genes of interest.

This will reveal their feasibility as DNA methylation biomarkers for PCa detection. To identify

which genes will have the best performance in detecting PCa patients with initial negative

biopsies, additional methylation analysis was performed between TAN and benign CP prostate

tissue for our genes of interest.

 

   

53  3.2 Materials and Methods

3.2.1 Patient Cohort

Matched RP and Biopsy (Bx) prostate tissue obtained from up to 152 patients per gene (range:

28-152) were analyzed for the methylation of our candidate genes. Tumour and TAN prostate

tissue from RP and Bx specimens were microdissected for DNA methylation analysis. The

concordance of methylation was analyzed between the following groups: matched RP tumour

and TAN prostate tissue, matched RP and Bx tumour prostate tissue, and matched RP tumour

and Bx TAN prostate tissue (Figure 3.1). Methylation was also analyzed between patients with

either RP or Bx TAN and benign CP prostate tissue to develop a cut off that ensured 100%

specificity. Methylation of selected genes was compared between three groups: up to 72 RP

TAN, 40 Bx TAN cases and 17 benign CP cases were compared. All patients were recruited

from the University Health Network (Toronto, Canada) with REB approval. Due to the limited

availability and/or quality of DNA, not all genes were analyzed for the same number of cases.

3.2.2 DNA Extraction, Bisulfite Modification and Quantitative, Methylation-Specific PCR

Benign CP, RP and Bx specimens from patients were obtained from FFPE blocks and cut into 5-

10 um serial sections for microdissection and DNA extraction. Samples within the benign CP

tissue cohort are the same cases evaluated in Chapter 2. RP TAN cohort is comprised of a

combination of all RP TAN cases analyzed in Chapter 2 and additional RP cases evaluated for

TAN methylation by previous members of the lab. H&E slides were prepared for each block

and were marked for normal, TAN and/or tumour tissue by Dr. T. van der Kwast, Dr. F. Siadat

and/or Dr. S. Bellur. Marked H&E slides were superimposed over the corresponding unstained

sections of benign CP, RP or Bx tissue to outline areas for DNA extraction. Method of

extraction and bisulfite modification is described in the Materials and Methods section of

Chapter 2. To measure the methylation of our genes of interest, we performed Methylight

technology. Assay protocol is described in Chapter 2. Primer and probe sequences for APC,

RASSF1A, HOXD3, CRIP3, KLK6 and KRTAP27-1 are listed in Table 2.1. Percent

Methylated Reference (PMR) values were calculated to quantify the methylation of our genes of

interest within each DNA sample165,173,176. PMR values were calculated as follows: [(Gene of

 

   

54  interest/ALU fluorescence quantity ratio for sample DNA)/(Gene of interest/ALU fluorescence

quantity ratio for super methylated DNA)] x 100%. Each DNA sample was analyzed in either

duplicate or triplicate.

3.2.3 Statistical Analysis

To examine the feasibility of our genes as field cancerization biomarkers for PCa detection,

concordance of methylation between matched TAN and tumour cases was evaluated. A gene

with a PMR > 0 was considered methylated, while a gene with a PMR = 0 was considered

unmethylated. Quantitative methylation concordance of an individual gene or a combination of

genes is described as the percentage of matched samples concordant in groups being compared.

If both matched samples are methylated, they have concordant methylation. If one sample is

methylated while the other sample is unmethylated, they have discordant methylation.

Sensitivity of our genes to detect PCa cases was calculated using only matched Bx TAN cases

from strictly PCa patients based on the number of true positive cases (concordant methylation)

and false negative cases (discordant methylation). Pearson correlation analysis was also

performed to determine the association of methylation between groups.

To evaluate the performance of our genes to accurately detect PCa patients, we analyzed the

methylation patterns between benign CP and RP or Bx TAN prostate tissue. Receiver Operating

Characteristic (ROC) analysis was conducted to determine the cut off value for each gene that

ensured 100% specificity. Cases with a PMR above the cut off were considered to be

hypermethylated. Multi-gene analysis models were performed to identify combinations of genes

that can improve the predictive ability over any one gene. When a case is above the cut off

value of at least one gene included in the multi-gene panel, they were considered to be

hypermethylated. Some cases included in a paired or multi-gene panel may have methylation

analyzed for only some of the genes. These cases were included only if 1 or more genes were

hypermethylated. However, if none of the genes had hypermethylation, they were excluded

from the analysis. Chi-squared (X2) analysis was performed to determine the association of gene

hypermethylation with PCa for any gene individually, or genes in paired or multi-gene panels.

Area under the curve (AUC), sensitivity, specificity, Positive Predictive Value (PPV) and NPV

 

   

55  were also evaluated for all individual or multi-gene models. Statistical methods have also been

performed previously126.

3.3 Results

3.3.1 Establishing the Quantitative Concordance of Methylation Between Matched Tumour and

TAN Prostate Tissue.

Matched RP tumour and either RP or Bx TAN prostate tissue were analyzed for the methylation

of APC, RASSF1A, HOXD3, CRIP3, KLK6 and KRTAP27-1 (Figure 3.1). Approximately 82%

of all patients had methylation detected in both RP tumour and RP TAN cases for 1 or more

genes (Table 3.1). KLK6 had the highest concordance of methylation (94.7%), while APC had

the lowest concordance of methylation (60%). A lower overall concordance of methylation was

observed between RP tumour and Bx TAN tissues (66%, Table 3.1). RASSF1A had the highest

concordance of methylation between both tissue types (88%), while APC had the least

concordance of methylation between tissue types (42%). As proof of principle, we also

analyzed the concordance of methylation between matched RP and Bx tumour tissue for APC

and HOXD3 (Table 3.1). As a result, we observed that 93% of patients had methylation

concordance of APC and HOXD3 between matched RP and Bx tumour prostate tissue. More

specifically, 100% of all patients had HOXD3 methylation in both RP and Bx tumour tissue,

while only 86% of patients had APC methylation in both tissues.

 

   

56  Table 3.1. Concordance of methylation data between matched tumour and TAN prostate tissue.

*Out of the 82 Bx cases extracted for DNA, 45 Bx cases had sufficient quality to run on our genes. However, only up to 38 cases had matched RP Tumour and Bx TAN Methylation data for analysis.

 

3.3.2 Qualitative analysis of methylation patterns between matched tumour and TAN prostate

tissue

The association between tumour and TAN prostate tissue methylation from RP or Bx tissue was

evaluated to determine the qualitative concordance of methylation between matched samples

from PCa patients. Due to the limited amount of DNA, each gene may have been run for a

different number of matched cases. Interestingly, KRTAP27-1 and APC methylation in RP

tumour and RP TAN tissue is positively correlated (n=46, p=6.99x10-6, r=0.598 for KRTAP27-

1, n=78, p=0.002, r=0.339 for APC) (Figure 3.2 A, C). This positive correlation was also

confirmed between RP tumour and Bx TAN tissue for the same genes (n=38, p=0.001, r= 0.529

for KRTAP27-1, n=28, p=0.002, r=0.561 for APC) (Figure 3.2 B, D). Notably, methylation of

HOXD3 in tumour prostate tissue from RP and Bx also revealed a strong, positive correlation

(Figure 3.2 E, n=14, p=0.007, r=0.688), while CRIP3 methylation in RP tumour and Bx TAN

tissue revealed a weak, but positive correlation (Figure 3.2 F, n=34, p=0.038, r=0.358).

Gene

RP Tumour vs. RP TAN Methylation

RP Tumour vs. Bx Tumour Methylation

RP Tumour vs. Bx TAN Methylation

n Concordance (%) n Concordance (%) n Concordance (%)

APC 78 60.3 14 85.7 28 46.4 RASSF1A 138 83.9 - - 34 88.2 HOXD3 131 81.7 14 100% 28 60.7 CRIP3 130 76.2 - - 34 52.9 KLK6 95 94.7 - - 33 72.7

KRTAP27-1 46 87 - - 38 68.4 Overall

Concordance 662 81.8 28 92.9 195 65.6

 

   

57  Figure 3.2. Correlation of methylation of genes of interest between RP Tumour vs. Bx Tumour, RP or Bx TAN. A) KRTAP27-1, n = 46. B) KRTAP27-1, n = 38. C) APC, n = 78. D) APC, n = 28. E) HOXD3, n = 14. F) CRIP3, n = 34.

   

     

A   B  

C   D  

E   F  

 

   

58  3.3.3 Analyzing the detection ability of candidate gene methylation using Bx TAN prostate

tissue

To evaluate the predictive ability of our genes of interest to indicate the presence of PCa, we

compared the sensitivity of APC, RASSF1A, CRIP3, HOXD3, KLK6 and KRTAP27-1

methylation in Bx TAN tissue from PCa patients (Table 3.2). Bx TAN cases that have a PMR >

0 for any gene indicated a true positive result. Bx TAN cases with PMR = 0 methylation for

any gene indicated a false positive result. Multi-gene sensitivity was calculated based on the

true positive and false negative result of 2 or more genes in combination. If one or more of the

genes had a PMR > 0, the case was given a true positive result. Up to 40 Bx TAN cases were

analyzed for any one gene. We observed that any gene paired with RASSF1A had stronger

predictive ability to detect PCa than any one gene alone (Table 3.2, average paired sensitivity =

92%). Similarly, KLK6, KRTAP27-1 and CRIP3 also revealed a high sensitivity when paired

with another gene (Table 3.2, Average paired sensitivity = 90%, 82% and 80%, respectively).

When CRIP3, KLK6, KRTAP27-1 and RASSF1A are combined as a panel, 100% of all patients

with PCa are predicted to have PCa.

Table 3.2. Sensitivity analysis of paired gene biomarker panels in Bx TAN cases from PCa patients only

Gene APC RASSF1A HOXD3 CRIP3 KLK6 KRTAP27-1

Average Sensitivity

(%)

APC N - 27 28 27 28 26

77.9 Sensitivity (%) - 85.2 75 70.4 85.7 73.1

RASSF1A N 27 - 27 34 31 30

93.7 Sensitivity (%) 85.2 - 92.6 97.1 96.7 96.7

HOXD3 N 28 27 - 27 28 26

85.3 Sensitivity (%) 75 92.6 - 77.8 92.8 88.4

CRIP3 N 27 34 27 - 31 30

83.8 Sensitivity (%) 70.4 97.1 77.8 - 90.3 83.3

KLK6 N 28 31 28 31 - 31

90.5 Sensitivity (%) 85.7 96.7 92.8 90.3 - 87.1

KRTAP27-1 N 26 30 26 30 31 -

85.7 Sensitivity (%) 73.1 96.7 88.4 83.3 87.1 -

 

   

59  3.3.4 Analyzing the detection ability of candidate gene methylation using benign CP and TAN

prostate tissue

Methylation of our genes of interest was analyzed between benign CP (up to n = 16) and RP (up

to n = 152) or Bx (up to n = 40) TAN cases to examine their efficacy as biomarkers for PCa

detection. Genes were analyzed individually and in combination in order to determine which

genes are most successful in identifying PCa patients and minimizing false negative results.

Using ROC curve analysis, we selected a PMR cut off value for all genes that ensured 100%

specificity.

Table 3.3. Individual gene methylation analysis between benign CP and TAN cases

Gene Tissue Type CRIP3 HOXD3 APC RASSF1A KRTAP27-1 KLK6

Total NN (n) RP TAN 16 16 15 16 17 - Bx TAN 16 16 15 16 17 15

Total PCa Cases (n)

RP TAN 145 144 95 152 72 - Bx TAN 34 28 28 34 40 44

True Positives (n)

RP TAN 56 57 30 66 10 - Bx TAN 11 5 7 21 4 5

False Negatives (n)

RP TAN 89 87 65 86 62 - Bx TAN 23 23 21 13 36 39

Specificity (%)

RP TAN 100 100 100 100 100 - Bx TAN 100 100 100 100 100 100

Sensitivity (%)

RP TAN 38.6 39.6 31.6 43.4 13.9 - Bx TAN 32.4 17.9 25.0 61.8 10.0 11.4

NPV (%) RP TAN 15.2 15.5 18.8 15.7 20.5 -

Bx TAN 41.0 41.0 41.7 55.2 32.1 27.8

PPV (%) RP TAN 100 100 100 100 100 -

Bx TAN 100 100 100 100 100 100

P Value (X2) RP TAN 0.002 0.002 0.011 0.001 0.113 - Bx TAN 0.01 0.073 0.034 3.66E-05 0.176 0.172

AUC RP TAN 0.71 0.65 0.59 0.70 0.53 - Bx TAN 0.55 0.52 0.20 0.62 0.48 0.41

Initially, we compared the methylation patterns between benign CP and RP TAN cases. We

observed that CRIP3 and RASSF1A were most successful in discriminating between PCa and

non-PCa patients as an individual biomarker (AUC = 0.71 and 0.70, sensitivity = 39% and 43%,

NPV = 15% and 15% respectively, Table 3.3, Figure 3.3 A). When we combined all genes into

 

   

60  paired biomarker panels, more PCa patients were accurately detected based on a higher

sensitivity of paired vs. individual gene biomarkers (Table 3.4, Figure 3.3 B). Out of the 10

paired biomarker panels, RASSF1A with KRTAP27-1 and HOXD3 with APC hypermethylation

were most significantly associated with PCa (p = 8.44 x 10-8 and 1.38 x 10-7, respectively).

Since these 2 paired gene panels showed the most promise as predictive biomarkers (RASSF1A

with KRTAP27-1: AUC = 0.78, sensitivity = 67%, NPV = 29%; HOXD3 with APC: AUC =

0.79, sensitivity = 66%, NPV = 28%), we combined them to create a multi-gene biomarker

panel. As a result, RASSF1A, KRTAP27-1, HOXD3 and APC in combination improved the

PCa detection rate while minimizing false negative results (AUC = 0.90, sensitivity = 0.79,

NPV = 0.52, p = 1.81 x 10-8, Table 3.4, Figure 3.3 C).

Table 3.4. Paired and Multi-gene analysis of gene methylation in benign CP vs. RP TAN cases

RP TAN vs. Benign CP

Gene combinations

Sensitivity (%)

Specificity (%)

NPV (%)

PPV (%)

P value (X2)

Benign CP

Cases (n)

PCa Cases

(n)

True Positives

(n)

False Negatives

(n) AUC

HOXD3, RASSF1A and KRTAP27-1

79.5 100 36.8 100 5.87E-10 14 117 93 24 0.90

KRTAP27-1, RASSF1A,

HOXD3 and APC

79.4 100 51.9 100 1.81E-08 14 63 50 13 0.90

RASSF1A and KRTA27-1 67.3 100 29.4 100 8.44E-08 15 110 74 36 0.78

CRIP3, KRTAP27-1 and

RASSF1A 70.9 100 33.3 100 1.29E-07 15 103 73 30 0.85

HOXD3 and APC 66.1 100 27.8 100 1.38E-07 15 115 76 39 0.79

CRIP3 and KRTAP27-1 65.3 100 30.0 100 2.76E-07 15 101 66 35 0.74

APC and RASSF1A 63.4 100 23.8 100 3.89E-07 15 131 83 48 0.84

CRIP3 and RASSF1A 62.3 100 21.4 100 5.16E-07 15 146 91 55 0.82

CRIP3 and APC 63.2 100 25.9 100 5.18E-07 15 117 74 43 0.79 HOXD3 and KRTAP27-1 61.9 100 27.3 100 1.17E-06 15 105 65 40 0.71

CRIP3 and HOXD3 60.1 100 21.9 100 1.43E-06 16 143 86 57 0.78

HOXD3 and RASSF1A 60.0 100 20.5 100 1.49E-06 15 145 87 58 0.77

APC and KRTAP27-1 52.0 100 28.0 100 5.95E-05 14 75 39 36 0.76

 

   

61   Figure 3.3. ROC curves that reflect the performance of individual or multi-gene biomarker panels involving benign CP and RP TAN cases. A) Individual gene biomarkers. Number of cases analyzed per gene, AUC and sensitivity statistics reported in Table 3.3. B) Paired gene biomarker panels. Number of cases analyzed, AUC and sensitivity statistics reported in Table 3.4. C) Multi-gene biomarker panels. Number of cases analyzed per gene, AUC and sensitivity statistics reported in Table 3.4.

Likewise, we also compared the methylation patterns between benign CP and Bx TAN cases.

CRIP3 and RASSF1A methylation also had the strongest predictive ability as individual

biomarkers based on AUC (AUC = 0.55 and 0.62, sensitivity = 32% and 62%, NPV = 41% and

55% respectively, Table 3.3, Figure 3.4 A). When genes were paired, any gene combined with

RASSF1A revealed a highly significant association with PCa (Range p = 2.08 x 10-6 to 4.32 x

10-5) compared to other paired gene models. Furthermore, RASSF1A paired with either

KRTAP27-1, CRIP3 or HOXD3 had the greatest performance in identifying PCa patients (AUC

= 0.82, 0.86 and 0.84, sensitivity = 72%, 74% and 70%, NPV = 64%, 63% and 63%,

respectively, Table 3.5, Figure 3.4 B). Combining all four genes into a multi-gene biomarker

A   B  

C  

 

   

62  panel improved the detection of PCa cases compared to the other gene models (AUC = 0.89,

sensitivity = 79% and NPV = 68%, p = 3.82 x 10-7, Table 3.5, Figure 3.4 C).

Table 3.5. Paired and Multi-gene analysis of gene methylation in benign CP vs. Bx TAN cases.

As mentioned above, we identified a 4-gene panel consisting of RASSF1A, CRIP3, KLK6 and

KRTAP27-1 that detected all PCa patients based on Bx TAN methylation (PMR > 0) observed

for at least one gene. We also analyzed the predictive ability of these genes using the cohorts

that contain both benign CP and TAN cases. Since KLK6 was not analyzed for the same set of

RP TAN cases, it was excluded from analysis between benign CP and RP TAN cases. We

Bx TAN vs. Benign CP

Gene combinations

Sensitivity (%)

Specificity (%)

NPV (%)

PPV (%)

P value (X2)

Benign CP

Cases (n)

PCa Cases

(n)

True Positives

(n)

False Negatives

(n) AUC

RASSF1A, HOXD3 and KRTAP27-1

86.2 100 78.9 100 4.44E-08 15 29 25 4 0.93

CRIP3, KRTAP27-1,

RASSF1A and KLK6

81.8 100 70.0 100 2.12E-07 14 33 27 6 0.91

RASSF1A, CRIP3,

KRAP27-1 and HOXD3

78.8 100 68.2 100 3.82E-07 15 33 26 7 0.89

RASSF1A and CRIP3 73.5 100 62.5 100 2.08E-

06 15 34 25 9 0.86

KRTAP27-1 and

RASSF1A 71.9 100 64.0 100 2.62E-

06 16 32 23 9 0.82

RASSF1A and HOXD3 70.0 100 62.5 100 9.12E-

06 15 30 21 9 0.84

RASSF1A and APC 68.8 100 60.0 100 1.07E-

05 15 32 22 10 0.82

RASSF1A and KLK6 64.7 100 53.8 100 4.32E-

05 14 34 22 12 0.82

APC and CRIP3 48.3 100 50.0 100 1.00E-

03 15 29 14 15 0.7

CRIP3 and KLK6 47.1 100 45.5 100 1.00E-

03 15 34 16 18 0.74

HOXD3 and CRIP3 43.3 100 48.5 100 2.00E-

03 16 30 13 17 0.68

KRTAP27-1 and CRIP3 37.5 100 44.4 100 5.00E-

03 16 32 12 20 0.64

APC and HOXD3 39.3 100 46.9 100 5.00E-

03 15 28 11 17 0.7

APC and KLK6 39.3 100 45.2 100 6.00E-

03 14 28 11 17 0.7

KRTAP27-1 and HOXD3 32.1 100 45.7 100 1.10E-

02 16 28 9 19 0.64

HOXD3 and KLK6 32.1 100 44.1 100 1.40E-

02 15 28 9 19 0.68

KRTAP27-1 and APC 31.0 100 42.9 100 1.60E-

02 15 29 9 20 0.62

KRTAP27-1 & KLK6 22.5 100 32.6 100 4.50E-

02 15 40 9 31 0.64

 

   

63  observed that the multi-gene panel consisting of RASSF1A, CRIP3, KRTAP27-1 and KLK6

(for benign CP vs. Bx TAN cases only) performed optimally in both analysis involving either

Bx or RP TAN cases. AUC, sensitivity and NPV were 0.85, 71% and 33%, respectively for the

analysis involving benign CP vs. RP TAN cases (Table 3.4, Figure 3.3 C) and 0.91, 82% and

70% respectively, for the analysis involving benign CP vs. Bx TAN cases (Table 3.5, Figure 3.4

D). Additionally, we observed that RASSF1A, HOXD3, and KRTAP27-1 were identified as

top candidate biomarkers that can best distinguish between benign and TAN cases when

combined in paired or multi-gene panels (Table 3.4 and 3.5). When combined into a three-gene

panel, their ability to identify PCa patients was comparable to or improved the performance of

other multi-gene models that were investigated (Benign CP vs. RP TAN cases: AUC = 0.90,

sensitivity = 79%, NPV = 37% (Table 3.4, Figure 3.3 C); Benign CP vs. Bx TAN cases: AUC =

0.93, sensitivity = 86%, NPV = 79% (Table 3.5, Figure 3.4 E).

 

   

64  Figure 3.4. ROC curves that reflect the performance of individual or multi-gene biomarker panels involving benign CP and Bx TAN cases. A) Individual gene biomarkers. Number of cases, AUC and sensitivity statistics reported in Table 3.2. B) Paired gene biomarker panels. Number of cases analyzed per gene, AUC and sensitivity statistics reported in Table 3.5. C-E) Multi-gene biomarker panels. Number of cases analyzed per gene, AUC and sensitivity statistics reported in Table 3.5.

A   B  

C   D  

E  

 

   

65  3.4 Discussion

Previous studies have reported that the DNA methylation profiles of prostate tumours are more

similar to DNA methylation profiles of matched TAN prostate tissue compared to normal

prostate tissue from non-cancer patients due to field cancerization127,163. To confirm this

finding, we determined the methylation concordance of APC, RASSF1A, KRTAP27-1,

HOXD3, CRIP3 and KLK6 in tumour and TAN RP tissue. This allowed us to evaluate their

feasibility as field cancerization biomarkers to indicate the presence of PCa using negative

biopsy samples. We observed that out of all patients who had tumour methylation of any gene,

81% had methylation of the same gene detected in TAN tissue. The high concordance of

methylation between RP TAN and tumour prostate tissue confirms that both tissue types share a

similar methylation profile. This reveals the potential role of our selected genes as field

cancerization biomarkers to indicate the presence of a nearby tumour when methylation is

detected in TAN tissue from negative biopsy cores.

It has been well established that APC, RASSF1A, HOXD3 and KLK6 are methylated in prostate

tumours116,124,165,187. Therefore, it is likely that a high concordance of methylation of our genes

would be observed between tumour tissues from RP and matched pre-operative Bx specimens

indicating a cancer diagnosis. As proof of principle, we examined the methylation of APC and

HOXD3 in matched RP and Bx tumours and observed that 92% of our patients had concordant

methylation between tissues. This confirms that methylation of our genes is likely to be found

in tumour tissue from either RP or Bx specimens. We decided not to analyze the methylation of

other genes of interest due to limited tumour DNA sample availability. Revealing the

concordance of methylation for all our genes between matched RP and Bx tumour tissue would

not provide us supportive evidence to further test these genes for their potential as diagnostic

biomarker in matched negative biopsy tissue.

When the methylation between RP tumours and Bx TAN cases was examined for our genes of

interest, we observed an overall concordance of 66%, with concordance of methylation varying

for each individual gene (Range: 46.4-88.2%). Compared to RP tumour and TAN prostate

tissue, we observed a lower overall concordance of methylation between RP tumours and Bx

TAN cases that could be explained by the multifocality of PCa. It was previously reported that

 

   

66  50.2% of patients with low risk PCa (GS 6) upgraded to intermediate or high-risk disease upon

RP82. This often occurs since many tumours exist within the prostate simultaneously and

tumours found upon Bx are not necessarily obtained from the tumour foci within the prostate

that has the highest GS. As a result, Bx TAN tissue that we examined could represent the

methylation profile of different tumour foci than what was investigated on the RP specimen.

Furthermore, the methylation of certain genes, such as APC, occurs at a lower frequency in

TAN tissues compared to other genes, such as RASSF1A157,159. This was confirmed in our

analysis, where approximately 42% of cases with RP tumour methylation also had Bx TAN

methylation of APC, compared to the 88% concordance of methylation observed for RASSF1A.

Our results suggest that using a combination gene methylation panel in TAN tissue from Bx

specimens may be more successful in detecting PCa.

We also investigated whether an association exists between prostate tumour and TAN

methylation of RP and Bx tissues. Interestingly, both APC and KRTAP27-1 methylation in RP

tumours was positively correlated with TAN methylation from both RP and Bx specimens.

Similarly, a positive correlation was observed for CRIP3 methylation between RP tumour and

Bx TAN tissue. This suggests that increased tumour methylation of APC, KRTAP27-1 and

CRIP3 may also indicate increased levels of methylation in TAN tissue. Previous studies have

reported that increased methylation of genes such as HOXD3 and APC are associated with high-

grade PCa and have been implicated as predictive biomarkers for aggressive disease165,176. If

TAN and tumour methylation is qualitatively associated in genes implicated as prognostic

biomarkers, they may be used to detect aggressive PCa using negative biopsy tissue.

KRTAP27-1 and APC show the most promise as prognostic biomarkers since a positive

association was observed between tumour and TAN prostate tissue. Furthermore, we also

demonstrated that higher KRTAP27-1 TAN methylation was significantly associated with a

higher GS using the array cases (Chapter 2, Figure 2.5 A & B), while an increasing trend in

APC methylation was observed in TAN GS6 compared to benign CP cases (Chapter 2, Figure

2.5 D). Although we were unable to reproduce these results for KRTAP27-1 in our independent

validation due to the TAN cohort composition (TAN GS 6 and TAN GS 7 cases), further

analysis comparing KRTAP27-1 and APC TAN methylation between high and low risk PCa

patients may reveal the potential role of these genes as predictive biomarkers for PCa

aggressiveness.

 

   

67  

To determine the efficacy of multiple genes in combination to detect PCa, we compared the

sensitivity of 2 or more genes based on the presence or absence of methylation in Bx TAN

tissue from PCa patients only. This revealed which genes can potentially be used as a

biomarker panel to identify PCa in patients when found to be methylated in negative biopsy

tissue. We observed that RASSF1A, KLK6, KRTAP27-1 and CRIP3 had the highest overall

sensitivity when paired with any other gene (average paired sensitivity > 80%). Combined into

a 4-gene biomarker panel, methylation of at least 1 gene in Bx TAN tissue was observed in all

PCa patients (sensitivity = 100%). These results suggest that RASSF1A, KLK6, KRTAP27-1

and CRIP3 used in combination may successfully detect PCa in negative biopsy tissue affected

by field cancerization. However, biomarker panels should also reveal high specificity, PPV and

NPV to provide strong evidence so that they can be used successfully in a clinical setting.

When we analyzed the methylation of our genes between TAN and benign CP prostate tissue,

we selected a cut off value that ensured 100% specificity. This allowed us to investigate the

ability of individual or multi-gene panels to identify PCa after ensuring no false positive results

(100% PPV). The ConfirmMDx assay can confidently rule out patients with no PCa if they do

not receive a positive methylation result. However, the assay has a poor PPV (29%), causing

potential false positive results. By selecting cut offs that ensured 100% specificity for each

gene, we could identify a biomarker panel that can supplement the ConfirmMDx assay and

minimize the amount of non-PCa patients from experiencing overtreatment.

Overall, we observed that genes combined into paired or multi-gene biomarker panels had

greater predictive ability to detect PCa cases than any one gene. Between benign CP and RP

TAN prostate tissue, the AUC, sensitivity and NPV improved when multiple genes were

combined. Multi-gene biomarker panels consisting of HOXD3, RASSF1A and KRTAP27-1 or

HOXD3, RASSF1A, APC and KRTAP27-1 revealed that their methylation was most

significantly associated with PCa and had the highest AUC, sensitivity, and NPV compared to

any other gene models. In our sensitivity analysis involving only Bx TAN cases from PCa

patients, RASSF1A, KLK6, CRIP3 and KRTAP27-1 were able to identify all PCa when at least

1 gene was observed to have a PMR > 0 for all patients (sensitivity = 100%). Therefore, we

also evaluated the performance of these genes in a cohort with both benign CP and RP TAN

 

   

68  prostate tissue methylation. Since KLK6 RP TAN cases are not the same cases evaluated for

RASSF1A, CRIP3 and KRTAP27-1, KLK6 was excluded from the analysis. RASSF1A, CRIP3

and KRTAP27-1 biomarker panel also revealed strong predictive ability that revealed an AUC

of 0.85 with a sensitivity and NPV of 71% and 33%, respectively. Although it wasn’t the best

performing biomarker panel based on AUC, methylation of these genes combined was very

significantly associated with PCa.

The same type of analysis was performed between benign CP and Bx TAN prostate tissue.

Unfortunately, we did not have access to an additional benign patient cohort and decided to use

the same cohort that was compared to the RP TAN cases. Similarly, we observed that paired or

multi-gene biomarker panels outperformed the predictive ability of individual genes to detect

PCa patients while minimizing false negative results. Notably, RASSF1A combined with any

other gene in a paired model performed better any other paired gene biomarker panels. This

confirmed previous studies that revealed RASSF1A methylation was a highly sensitive

biomarker116,157,159. However, when RASSF1A was combined with KLK6, CRIP3 and

KRTAP27-1 into a multi-gene panel, this improved the predictive ability of RASSF1A alone or

paired, revealing an AUC of 91% with a sensitivity and NPV of 82% and 70%, respectively.

Based on the results from both analyses, the paired biomarker panels with the strongest PCa

detection ability included RASSF1A, KRTAP27-1 and HOXD3. When combined into a three-

gene panel, we observed that the methylation of these genes were most significantly associated

with PCa in both comparisons. Furthermore, they had very high AUC, sensitivity and NPV.

Since this was observed in both comparisons, this combination of genes shows the most promise

as a diagnostic biomarker panel to detect PCa patients.

The main limitation of this project is the absence of an additional negative control cohort of

healthy patients with no PCa to analyze alongside Bx TAN methylation. In previous studies,

cohorts are often separated into a training and test set to evaluate the efficacy of a cut off value

in separating two groups accurately153,190,191. Although we had two separate TAN cohorts (Bx

and RP), we only had 1 cohort of benign CP cases that was used in both analyses. Therefore, we

were unable to identify an optimal methylation cut off value for a specific gene or gene panel

that could most accurately distinguish between PCa and non-PCa patients. Currently, we are in

 

   

69  the process of obtaining negative biopsy tissue from patients with at least 2 consecutive biopsy

examinations and no evidence of PCa. This will allow us to create separate training and test

cohorts to perform more comprehensive statistical analysis.

As mentioned in Chapter 2, the benign CP prostate tissue was taken from patients with high-

grade urothelial carcinoma and may possess methylation patterns that indicate field

cancerization. For example, we observed high methylation values for KRTAP27-1 and GSTP1

in benign CP tissue that may be associated with field cancerization. Since the same cohort of

benign CP cases were used in our analysis with RP and Bx TAN tissue in this chapter, we

selected high methylation cut off values for certain genes to ensure 100% specificity. This may

have influenced the number of false negative results detected since many PCa cases with a

lower PMR value fell below the cut off for specific genes. Obtaining negative prostate biopsy

tissue from patients with no urothelial carcinoma may allow us to identify methylation cut off

values that can more accurately discriminate between non-PCa and PCa patients and improve

the NPV and sensitivity of our biomarker panels under investigation.

Other limitations involved the quality and quantity of Bx tissues that were available for

methylation analysis. Microdissection and DNA extraction from FFPE Bx prostate tissue

yielded very limited amounts of DNA, which made it difficult to analyze all 6 genes per Bx

sample. This caused us to exclude GSTP1 in our Chapter 3 analysis since not enough DNA was

available. Furthermore, different genes have a different number of TAN Bx cases analyzed. By

having the same number of cases analyzed per gene or gene panel, a more direct comparison of

their performance can be made since the number of cases analyzed affects our calculations for

NPV and sensitivity. Furthermore, only a few of our genes could be multiplexed using

Methylight technology. Therefore, more DNA was required to analyze the methylation of each

gene on separate assays. To overcome this limitation, redesigning primers and probes for our

genes of interest may allow multiple genes to be multiplexed optimally. However, this would

require a shifting of the amplified region and could create incomparable results to our previous

analysis.

 

 70  

Chapter 4: Summary of Main Findings and

Future Directions

 

   

71  4.1 Summary of Main Findings

Although needle biopsy examinations are the gold standard procedure for PCa diagnosis, it is

prone to sampling bias, causing approximately 25-30% of PCa patients to receive false negative

results on their first biopsy76–78. The commercially available ConfirmMDx assay was designed

to address this dilemma. However, it is also limited by its sensitivity (68%), specificity (64%)

and PPV (29%), which may lead to 30% false positive results159. Consequently, many non-PCa

patients may undergo repeat biopsy examinations. DNA methylation alterations in TAN tissue

are promising biomarkers to identify PCa patients with an initial negative biopsy result since

biopsy tissue from these patients are likely removed from TAN areas affected by field

cancerization. Therefore, the goal of my thesis project was to investigate DNA methylation

patterns in TAN tissue and identify differentially methylated genes that can improve PCa

detection in patients who received an initial negative biopsy test.

We observed that APC, RASSF1A and GSTP1 methylation in TAN prostate tissue could be

detected up to 8 mm away as a result of field cancerization. This suggests that negative biopsy

cores taken up to 8 mm away from a tumour may harbour differential methylation patterns that

can indicate the presence of PCa. We also observed a distinct methylation signature between

TAN and benign CP prostate tissue, further confirming the influence of field cancerization to

alter the DNA methylation pattern in TAN prostate tissue. Five significantly differentially

methylated CpG sites and 3 DMRs were identified from our methylation array results and were

selected for further validation as potential candidate diagnostic biomarkers.

We were able to verify the differential methylation patterns of KRTAP27-1 and KLHDC7A

observed in the methylation array using Methylight technology. Methylight results revealed that

both genes had significantly elevated TAN methylation compared to benign CP. When TAN

cases were stratified by GS, KLHDC7A methylation was significantly higher in both TAN GS 6

and TAN GS 8 and 9 when combined compared to benign CP cases. Similarly, KRTAP27-1

methylation in TAN GS6 cases was significantly elevated compared to benign CP cases. Due to

the high methylation of KLHDC7A in both TAN and benign CP cases, KRTAP27-1 was the

only gene selected for independent validation. Interestingly, independent validation of

KRTAP27-1 methylation was not concordant with the results from the array and technical

validation. Despite the discordance, we pursued further investigation of KRTAP27-1

 

   

72  methylation to examine its potential as a biomarker for PCa diagnosis in combination with other

genes.

In addition to KRTAP27-1, we evaluated APC, RASSF1A, HOXD3, CRIP3 and KLK6

methylation levels to determine the concordance of methylation between matched tumour and

TAN prostate tissue and examine their diagnostic ability as PCa biomarkers. Overall

concordance of methylation between RP tumour and RP TAN, Bx tumour and Bx TAN tissue

was approximately 81.8%, 92.9% and 65.6%, respectively. KLK6 had the highest concordance

of methylation between RP Tumour and TAN cases, while HOXD3 and RASSF1A had highest

methylation concordance between RP tumour and Bx tumour and RP tumour and Bx TAN

cases, respectively. Interestingly, KRTAP27-1 and APC revealed a significant positive

correlation between RP tumour and RP TAN methylation and RP tumour and Bx TAN

methylation. When RASSF1A, CRIP3, KLK6 and KRTAP27-1 were combined, at least one of

the genes was methylated per case to identify all PCa patients in Bx TAN tissue.

To further examine the predictive ability of our candidate biomarkers, methylation of

RASSF1A, CRIP3, KLK6, KRTAP27-1, APC and HOXD3 was also analyzed in a cohort

containing benign CP and TAN prostate tissue from RP or Bx specimens. CRIP3 and

RASSF1A individually outperformed the predictive ability of other genes to detect PCa patients

when analyzed between benign CP and either RP or Bx TAN tissue. However, when genes are

combined into paired or multi-gene biomarker panels, the sensitivity, AUC and NPV increased,

improving the overall detection rate. A biomarker panel containing RASSF1A, CRIP3, KLK6

and KRTAP27-1 revealed strong predictive ability when hypermethylation of these genes were

evaluated in combination. This confirmed our previous results that revealed strong sensitivity of

these genes in combination to detect PCa patients using Bx TAN tissue only. Furthermore,

hypermethylation of HOXD3, RASSF1A and KRTAP27-1 in combination was most

significantly associated with PCa in both analyses. This multi-gene biomarker panel revealed a

high AUC, sensitivity and NPV that was superior to other biomarker panels. In conclusion, this

project revealed the clinical potential of using field cancerization biomarkers for PCa detection

using negative biopsy tissue.

 

   

73  4.2 Future Directions

Moving forward, we intend to evaluate the efficacy of our biomarkers for both PCa detection

and prognosis using an additional cohort of negative biopsy tissue from non-cancer patients who

had at least two consecutive negative biopsies and no histological evidence of PCa. By

including these cases, additional analysis can be performed to identify methylation cut off

values that can detect and stratify PCa patients based on their risk of having aggressive disease.

Methylation cut off values for each gene will be determined using a training cohort and will be

applied to a test cohort to compare the performance of our biomarkers individually and in

combination (paired or multi-gene biomarker panel) for PCa detection and risk stratification.

The most promising PCa diagnostic and prognostic biomarker panel identified using the method

mentioned above will also be compared to existing screening (i.e. PSA, the ConfirmMDx panel

of biomarkers (APC, RASSF1A and GSTP1), PCA3 score, PHI, 4Kscore) and prognostic (i.e.

OncotypeDx, Prostarix test) tests used prior to RP. Furthermore, we can determine the additive

value of combining our biomarker panel with these existing clinical tests to see if this may

improve PCa detection and patient risk stratification while minimizing false-positive and false-

negative results.

To further investigate the efficacy of our biomarker panel separately or in combination with

other screening tests, we will assess their performance in a separate cohort of patients

(validation cohort) that consist of PCa cases and benign controls. Prostate tissue examined for

gene methylation will be obtained from biopsy cores, which closely reflects clinical practice for

PCa diagnosis and early prognosis. Initial negative biopsy cores will be obtained from PCa

cases who had an initial false-negative followed by a positive biopsy result. Similarly, true-

negative biopsy cores will be obtained from benign controls who had at least 2 consecutive

negative biopsy results and no evidence of PCa. Comparative analysis will be performed to

determine if our biomarker panel can improve PCa detection and prognosis as a stand-alone test

or in combination with other clinical tests. This analysis will confirm the benefit of using field

cancerization biomarkers for clinical PCa detection and early prognosis in negative biopsy

tissue. Long-term goal of these studies will be to translate these results into clinical practice.

 

 74  

References

1. Salman JW, Schoots IG, Carlsson S V., Jenster G, Roobol MJ. Prostate Specific Antigen

as a Tumor Marker in Prostate Cancer  : Biochemical and Clinical Aspects. Adv Exp Med

Biol. 2015;867:93-114. doi:10.1007/978-94-017-7215-0.

2. Bhavsar A, Verma S, Bhavsar A, Verma S. Anatomic Imaging of the Prostate. Biomed

Res Int. 2014;2014:1-9. doi:10.1155/2014/728539.

3. Shen M, Abate-Shen C. Molecular genetics of prostate cancer: new prospects for old

challenges. Genes Dev. 2010;(212):1967-2000. doi:10.1101/gad.1965810.GENES.

4. Barron DA, Rowley DR. The reactive stroma microenvironment and prostate cancer

progression. 2012:187-204. doi:10.1530/ERC-12-0085.

5. Berry PA, Maitland NJ, Collins AT. Androgen receptor signalling in prostate  : Effects of

stromal factors on normal and cancer stem cells. 2008;288:30-37.

doi:10.1016/j.mce.2008.02.024.

6. Center MM, Jemal A, Lortet-Tieulent J, et al. International variation in prostate cancer

incidence and mortality rates. Eur Urol. 2012;61(6):1079-1092.

doi:10.1016/j.eururo.2012.02.054.

7. Canadian Cancer Society. Canadian Cancer Statistics Special topic: HPV-associated

cancers. Public Heal Agency Canada. 2016.

8. Tasian GE, Cooperberg MR, Cowan JE, et al. Prostate specific antigen screening for

prostate cancer: Knowledge of, attitudes towards, and utilization among primary care

physicians Gregory. Urol Oncol Semin Orig Investig. 2012;30(2):155-160.

doi:10.1016/j.urolonc.2009.12.019.

9. Connor Gorber S, Singh H, Pottle K, Jaramillo A, Tonelli M. Canadian Task Force on

Preventive Health Care. 2015;184(14):1575-1581. http://canadiantaskforce.ca/.

10. Attard G, Parker C, Eeles RA, et al. Prostate cancer. Lancet. 2016;387(10013):70-82.

doi:10.1016/S0140-6736(14)61947-4.

11. Cuzick J, Thorat MA, Andriole G, et al. Prevention and early detection of prostate cancer.

Lancet Oncol. 2014;15(11):e484-e492. doi:10.1016/S1470-2045(14)70211-6.

12. Glass AS, Cary KC, Cooperberg MR. Risk-based prostate cancer screening: Who and

how? Curr Urol Rep. 2013;14(3):192-198. doi:10.1007/s11934-013-0319-8.

 

   

75  13. Bechis SK, Carroll PR, Cooperberg MR. Impact of age at diagnosis on prostate cancer

treatment and survival. J Clin Oncol. 2011;29(2):235-241.

doi:10.1200/JCO.2010.30.2075.

14. Thompson IM, Ankerst DP, Chi C, et al. Assessing Prostate Cancer Risk  : Results from

the Prostate Cancer Prevention Trial. 2006;78229:529-534. doi:10.1093/jnci/djj131.

15. Dale W, Vijayakumar S, Lawlor EF, Merrell K. Prostate cancer, race, and socioeconomic

status: Inadequate adjustment for social factors in assessing racial differences. Prostate.

1996;29(5):271-281. doi:10.1002/(SICI)1097-0045(199611)29:5<271::AID-

PROS1>3.0.CO;2-D.

16. Lichtenstein P, Holm N V., Verkasalo PK, et al. Environmental and Heritable Factors in

the Causation of Cancer — Analyses of Cohorts of Twins from Sweden, Denmark, and

Finland. N Engl J Med. 2000;343(2):78-85. doi:10.1056/NEJM200007133430201.

17. Rebbeck TR, Mitra N, Wan F, et al. Association of Type and Location of BRCA1 and

BRCA2 Mutations With Risk of Breast and Ovarian Cancer. Jama. 2015;313(13):1347.

doi:10.1001/jama.2014.5985.

18. Kote-Jarai Z, Leongamornlert D, Saunders E, et al. BRCA2 is a moderate penetrance

gene contributing to young-onset prostate cancer: implications for genetic testing in

prostate cancer patients. Br J Cancer. 2011;105(8):1230-1234. doi:10.1038/bjc.2011.383.

19. Castro E, Goh C, Olmos D, et al. Germline BRCA mutations are associated with higher

risk of nodal involvement, distant metastasis, and poor survival outcomes in prostate

cancer. J Clin Oncol. 2013;31(14):1748-1757. doi:10.1200/JCO.2012.43.1882.

20. Karlsson R, Aly M, Clements M, et al. A Population-based Assessment of Germline

HOXB13 G84E Mutation and Prostate Cancer Risk. Eur Urol. 2014;65(1):169-176.

doi:10.1016/j.eururo.2012.07.027.

21. Goh CL, Eeles RA. Germline genetic variants associated with prostate cancer and

potential relevance to clinical practice. Recent Results Cancer Res. 2014;202:9-26.

doi:10.1007/978-3-642-45195-9_2.

22. Arora R, Koch MO, Eble JN, Ulbright TM, Li L, Cheng L. Heterogeneity of Gleason

grade in multifocal adenocarcinoma of the prostate. Cancer. 2004;100(11):2362-2366.

doi:10.1002/cncr.20243.

23. Boutros PC, Fraser M, Harding NJ, et al. Spatial genomic heterogeneity within localized,

multifocal prostate cancer. Nat Genet. 2015;47(May):1-14. doi:10.1038/ng.3315.

 

   

76  24. Andreoiu M, Cheng L. Multifocal prostate cancer: biologic, prognostic, and therapeutic

implications. Hum Pathol. 2010;41(6):781-793. doi:10.1016/j.humpath.2010.02.011.

25. Cooper CS, Eeles R, Wedge DC, et al. Analysis of the genetic phylogeny of multifocal

prostate cancer identifies multiple independent clonal expansions in neoplastic and

morphologically normal prostate tissue. Nat Genet. 2015;47(4):367-372.

doi:10.1038/ng.3221.

26. Arora R, Koch MO, Eble JN, Ulbright TM, Li L, Cheng L. Heterogeneity of Gleason

grade in multifocal adenocarcinoma of the prostate. Cancer. 2004;100(11):2362-2366.

doi:10.1002/cncr.20243.

27. Ruijter ET, Van De Kaa CA, Schalken JA, Debruyne FM, Ruiter DJ. Histological Grade

Heterogeneity in Multifocal Prostate Cancer. Biological and Clinical Implications. J

Pathol. 1996;180(3):295-299. doi:10.1002/(SICI)1096-9896(199611)180:3<295::AID-

PATH663>3.0.CO;2-W.

28. Jones TA, Radtke JP, Hadaschik B, Marks LS. Optimizing safety and accuracy of

prostate biopsy. Curr Opin Urol. 2016;26(5):472-480.

doi:10.1097/MOU.0000000000000310.

29. Chang DTS, Challacombe B, Lawrentschuk N. Transperineal biopsy of the prostate — is

this the future  ? Nat Publ Gr. 2013;10(12):690-702. doi:10.1038/nrurol.2013.195.

30. Klotz L. Active Surveillance for Prostate Cancer: Overview and Update. Curr Treat

Options Oncol. 2013;14(1):97-108. doi:10.1007/s11864-012-0221-5.

31. Grimm P, Billiet I, Bostwick D, et al. Comparative analysis of prostate-specific antigen

free survival outcomes for patients with low, intermediate and high risk prostate cancer

treatment by radical therapy. Results from the Prostate Cancer Results Study Group. BJU

Int. 2012;109 Suppl(s1):22-29. doi:10.1111/j.1464-410X.2011.10827.x.

32. Kollmeier MA, Zelefsky MJ. How to select the optimal therapy for early-stage prostate

cancer. Crit Rev Oncol Hematol. 2012;84(SUPPL. 1):6-15.

doi:10.1016/j.critrevonc.2012.12.002.

33. Gleason DF, Mellinger GT, Arduino LJ, et al. Prediction of Prognosis for Prostatic

Adenocarcinoma by Combined Histological Grading and Clinical Staging. J Urol.

1974;197(2):S134-S139. doi:10.1016/j.juro.2016.10.099.

34. Gleason DF. Classification of prostatic carcinomas. Cancer Chemother reports.

1966;50(3):125-128. http://www.ncbi.nlm.nih.gov/pubmed/5948714. Accessed February

 

   

77  13, 2017.

35. Montironi R, Mazzucchelli R, Lopez-Beltran A, Scarpelli M, Cheng L. The Gleason

grading system: Where are we now? Diagnostic Histopathol. 2011;17(10):419-427.

doi:10.1016/j.mpdhp.2011.06.008.

36. Lopez-Beltran A, Mikuz G, Luque RJ, Mazzucchelli R, Montironi R. Current practice of

Gleason grading of prostate carcinoma. Virchows Arch. 2006;448(2):111-118.

doi:10.1007/s00428-005-0102-4.

37. Shen MM, Abate-Shen C, Kregel S, et al. The 2014 International Society of Urological

Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma:

Definition of Grading Patterns and Proposal for a New Grading System. Vol 8.;

2013:244-252. doi:10.1097/PAS.0000000000000530.

38. Van Der Kwast T, Bubendorf L, Mazerolles C, et al. Guidelines on processing and

reporting of prostate biopsies: The 2013 update of the pathology committee of the

European Randomized Study of Screening for Prostate Cancer (ERSPC). Virchows Arch.

2013;463(3):367-377. doi:10.1007/s00428-013-1466-5.

39. Matoso A, Epstein JI. Grading of Prostate Cancer: Past, Present, and Future. Curr Urol

Rep. 2016;17(3):25. doi:10.1007/s11934-016-0576-4.

40. Amin A, Partin A, Epstein JI. Gleason Score 7 Prostate Cancer on Needle Biopsy:

Relation of Primary Pattern 3 or 4 to Pathological Stage and Progression After Radical

Prostatectomy. J Urol. 2011;186(4):1286-1290. doi:10.1016/j.juro.2011.05.075.

41. Pierorazio PM, Walsh PC, Partin AW, Epstein JI. Prognostic Gleason grade grouping:

data based on the modified Gleason scoring system. BJU Int. 2013;111(5):753-760.

doi:10.1111/j.1464-410X.2012.11611.x.

42. Epstein JI, Zelefsky MJ, Sjoberg DD, et al. A Contemporary Prostate Cancer Grading

System: A Validated Alternative to the Gleason Score. Eur Urol. 2016;69(3):428-435.

doi:10.1016/j.eururo.2015.06.046.

43. Kweldam CF, Wildhagen MF, Steyerberg EW, Bangma CH, van der Kwast TH, van

Leenders GJ. Cribriform growth is highly predictive for postoperative metastasis and

disease-specific death in Gleason score 7 prostate cancer. Mod Pathol. 2014:1-8.

doi:10.1038/modpathol.2014.116.

44. Amin MB, Greene FL, Edge SB, et al. The Eighth Edition AJCC Cancer Staging Manual:

Continuing to build a bridge from a population-based to a more “personalized” approach

 

   

78  to cancer staging. CA Cancer J Clin. January 2017. doi:10.3322/caac.21388.

45. Cheng L, Montironi R, Bostwick DG, Lopez-Beltran A, Berney DM. Staging of prostate

cancer. Histopathology. 2012;60(1):87-117. doi:10.1111/j.1365-2559.2011.04025.x.

46. Oon SF, Pennington SR, Fitzpatrick JM, Watson RWG. Biomarker research in prostate

cancer-towards utility, not futility. Nat Rev Urol. 2011;8(3):131-138.

doi:10.1038/nrurol.2011.11.

47. Eifler JB, Feng Z, Lin BM, et al. An updated prostate cancer staging nomogram (Partin

tables) based on cases from 2006 to 2011. BJU Int. 2013;111(1):22-29.

doi:10.1111/j.1464-410X.2012.11324.x.

48. Kattan MW, Eastham JA, Stapleton AM, Wheeler TM, Scardino PT. A preoperative

nomogram for disease recurrence following radical prostatectomy for prostate cancer. J

Natl Cancer Inst. 1998;90(10):766-771. doi:10.1093/jnci/90.10.766.

49. Morgan T, Palapattu G, Wei J. Screening for Prostate Cancer—Beyond Total PSA,

Utilization of Novel Biomarkers. Curr Urol Rep. 2015;16(9). doi:10.1007/s11934-015-

0537-3.

50. Koulikov D, Mamber A, Fridmans A, Abu Arafeh W, Shenfeld OZ. Why I Cannot Find

the Prostate? Behind the Subjectivity of Rectal Exam. ISRN Urol. 2012;2012:1-4.

doi:10.5402/2012/456821.

51. Thompson IM, Ankerst DP, Chi C, et al. Assessing prostate cancer risk: Results from the

Prostate Cancer Prevention Trial. J Natl Cancer Inst. 2006;98(8):529-534.

doi:10.1093/jnci/djj131.

52. Roobol MJ, Van Vugt HA, Loeb S, et al. Prediction of prostate cancer risk: The role of

prostate volume and digital rectal examination in the ERSPC risk calculators. Eur Urol.

2012;61(3):577-583. doi:10.1016/j.eururo.2011.11.012.

53. Stamey TA, Caldwell M, McNeal J, Nolley R, Downs J. the Prostate Specific Antigen

Era in the United States Is Over for Prostate Cancer: What Happened in the Last 20

Years? J Urol. 2004;172(4):1297-1301. doi:10.1097/01.ju.0000139993.51181.5d.

54. Balk SP. Biology of Prostate-Specific Antigen. J Clin Oncol. 2003;21(2):383-391.

doi:10.1200/JCO.2003.02.083.

55. Rao AR, Motiwala HG, Karim OMAA. The discovery of prostate-specific antigen. BJU

Int. 2008;101(1):5-10. doi:10.1111/j.1464-410X.2007.07138.x.

56. Moyer VA. Screening for Prostate Cancer: U.S. Preventive Services Task Force

 

   

79  Recommendation Statement. Ann Intern Med. 2012;157(2).

57. Andriole GL, Crawford ED, Grubb RL, et al. Prostate cancer screening in the randomized

Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial: mortality results after 13

years of follow-up. J Natl Cancer Inst. 2012;104(2):125-132. doi:10.1093/jnci/djr500.

58. Roddam AW, Duffy MJ, Hamdy FC, et al. Use of prostate-specific antigen (PSA)

isoforms for the detection of prostate cancer in men with a PSA level of 2-10 ng/ml:

Systematic review and meta-analysis. Eur Urol. 2005;48(3):386-398.

doi:10.1016/j.eururo.2005.04.015.

59. Mitchell P, Med M, Wriedt C, et al. Prevalence of prostate cancer among men with a

prostate-specific antigen level < or =4.0 ng per milliliter. Sci York. 2009:557-568.

doi:10.1056/NEJMoa1214609.

60. Falzarano SM, Ferro M, Bollito E, Klein E a, Carrieri G, Magi-Galluzzi C. Novel

biomarkers and genomic tests in prostate cancer: a critical analysis. Minerva Urol Nefrol.

2015;67(3):211-231. http://www.ncbi.nlm.nih.gov/pubmed/26054411.

61. Draisma G, Boer R, Otto SJ, et al. Lead times and overdetection due to prostate-specific

antigen screening: estimates from the European Randomized Study of Screening for

Prostate Cancer. J Natl Cancer Inst. 2003;95(12):868-878.

http://www.ncbi.nlm.nih.gov/pubmed/12813170. Accessed February 17, 2017.

62. Lilja H, Christensson A, Dahlén U, et al. Prostate-specific antigen in serum occurs

predominantly in complex with α1-antichymotrypsin. Clin Chem. 1991;37(9):1618-1625.

63. Lee R, Localio AR, Armstrong K, Malkowicz SB, Schwartz JS. A meta-analysis of the

performance characteristics of the free prostate-specific antigen test. Urology.

2006;67(4):762-768. doi:10.1016/j.urology.2005.10.052.

64. Carroll PR, Parsons JK, Andriole G, et al. Prostate Cancer Early Detection, Version

2.2015. J Natl Compr Canc Netw. 2015;13(12):1534-1561.

65. Oesterling JE, Cooner WH, Jacobsen SJ, Guess HA, Lieber MM. Influence of patient age

on the serum PSA concentration. An important clinical observation. Urol Clin North Am.

1993;20(4):671-680. http://www.ncbi.nlm.nih.gov/pubmed/7505975. Accessed February

17, 2017.

66. Greene KL, Albertsen PC, Babaian RJ, et al. Prostate Specific Antigen Best Practice

Statement: 2009 Update. J Urol. 2009;182(5):2232-2241. doi:10.1016/j.juro.2009.07.093.

67. Djavan BOB, Remzi M, Zlotta AR, et al. Complexed prostate-specific antigen,

 

   

80  complexed prostate-specific antigen density of total and transition zone, complexed/total

prostate-specific antigen ratio, free-to-total prostate-specific antigen ratio, density of total

and transition zone prostate-sp. Urology. 2002;4295(2):0-5.

68. Reed A, Ankerst DP, Pollock BH, Thompson IM, Parekh DJ. Current Age and Race

Adjusted Prostate Specific Antigen Threshold Values Delay Diagnosis of High Grade

Prostate Cancer. J Urol. 2007;178(5):1929-1932. doi:10.1016/j.juro.2007.07.045.

69. Loeb S, Han M, Roehl KA, Antenor JOANN V, Catalona WJ. Accuracy of Prostate

Weight Estimation By Digital Rectal Examination Versus Transrectal Ultrasonography. J

Urol. 2005;173(January):63-65. doi:10.1097/01.ju.0000145883.01068.5f.

70. Caras RJ, Sterbis JR. Prostate Cancer Nomograms  : A Review of Their Use in Cancer

Detection and Treatment. Curr Urol Rep. 2014;15(3):1-6. doi:10.1007/s11934-013-0391-

0.

71. Chun FKH, Karakiewicz PI, Briganti A, et al. Prostate Cancer Nomograms: An Update.

Eur Urol. 2006;50(5):914-926. doi:10.1016/j.eururo.2006.07.042.

72. Heidenreich A, Bastian PJ, Bellmunt J, et al. EAU guidelines on prostate cancer. Part 1:

Screening, diagnosis, and local treatment with curative intent - Update 2013. Eur Urol.

2014;65(1):124-137. doi:10.1016/j.eururo.2013.09.046.

73. Stock C, Hruza M, Cresswell J, Rassweiler JJ. Transrectal Ultrasound-Guided Biopsy of

the Prostate  : Development of the Procedure , Current Clinical Practice , Processing

Biopsy Cores. 2008;22(6):1321-1329. doi:10.1089/end.2008.0068.

74. Patel AR, Jones JS. Optimal biopsy strategies for the diagnosis and staging of prostate

cancer. Curr Opin Urol. 2009;19(3):232-237. doi:10.1097/MOU.0b013e328329a33e.

75. Hodge KK, McNeal JE, Terris MK, Stamey TA. Random systematic versus directed

ultrasound guided transrectal core biopsies of the prostate. J Urol. 1989;142(1):71-4-5.

http://www.ncbi.nlm.nih.gov/pubmed/2659827. Accessed February 17, 2017.

76. Babaian RJ, Toi A, Kamoi K, et al. A comparative analysis of sextant and an extended

11-core multisite directed biopsy strategy. J Urol. 2000;163(1):152-157.

doi:10.1016/S0022-5347(05)67993-1.

77. Guichard G, Larré S, Gallina A, et al. Extended 21-Sample Needle Biopsy Protocol for

Diagnosis of Prostate Cancer in 1000 Consecutive Patients. Eur Urol. 2007;52(2):430-

435. doi:10.1016/j.eururo.2007.02.062.

78. Ploussard G, Nicolaiew N, Marchand C, et al. Prospective evaluation of an extended 21-

 

   

81  core biopsy scheme as initial prostate cancer diagnostic strategy. Eur Urol.

2014;65(1):154-161. doi:10.1016/j.eururo.2012.05.049.

79. Jones JS, Patel A, Schoenfield L, Rabets JC, Zippe CD, Magi-Galluzzi C. Saturation

technique does not improve cancer detection as an initial prostate biopsy strategy. J Urol.

2006;175(2):485-488. doi:10.1016/S0022-5347(05)00211-9.

80. Levy DA, Jones JS. Management of rising prostate-specific antigen after a negative

biopsy. Curr Urol Rep. 2011;12(3):197-202. doi:10.1007/s11934-011-0177-1.

81. Jiang X, Zhu S, Feng G, et al. Is an initial saturation prostate biopsy scheme better than

an extended scheme for detection of prostate cancer? A systematic review and meta-

analysis. Eur Urol. 2013;63(6):1031-1039. doi:10.1016/j.eururo.2013.01.035.

82. Shaw GL, Thomas BC, Dawson SN, et al. Identification of pathologically insignificant

prostate cancer is not accurate in unscreened men. Br J Cancer. 2014;110(10):2405-2411.

doi:10.1038/bjc.2014.192.

83. Serefoglu EC, Altinova S, Ugras NS, Akincioglu E, Asil E, Balbay MD. How reliable is

12-core prostate biopsy procedure in the detection of prostate cancer  ? 2013;7:293-298.

84. Roehl KA, Antenor J ann V., Catalona WJ. Serial Biopsy Results in Prostate Cancer

Screening Study. J Urol. 2002;167(6):2435-2439. doi:10.1016/S0022-5347(05)64999-3.

85. Welch HG, Fisher ES, Gottlieb DJ, Barry MJ. Detection of prostate cancer via biopsy in

the medicare-SEER population during the PSA era. J Natl Cancer Inst.

2007;99(18):1395-1400. doi:10.1093/jnci/djm119.

86. Bennett HY, Roberts MJ, Doi SAR, Gardiner RA. The global burden of major infectious

complications following prostate biopsy. Epidemiol Infect. 2015;(2016):1-8.

doi:10.1017/S0950268815002885.

87. Loeb S, Carter HB, Berndt SI, Ricker W, Schaeffer EM. Complications after prostate

biopsy: Data from SEER-Medicare. J Urol. 2011;186(5):1830-1834.

doi:10.1016/j.juro.2011.06.057.

88. Wade J, Rosario DJ, Macefield RC, et al. Psychological impact of prostate biopsy:

Physical symptoms, anxiety, and depression. J Clin Oncol. 2013;31(33):4235-4241.

doi:10.1200/JCO.2012.45.4801.

89. Shinohara K, Nguyen H, Masic S. Management of an increasing prostate-specific antigen

level after negative prostate biopsy. Urol Clin North Am. 2014;41(2):327-338.

doi:10.1016/j.ucl.2014.01.010.

 

   

82  90. Grummet JP, Weerakoon M, Huang S, et al. Sepsis and “superbugs”: Should we favour

the transperineal over the transrectal approach for prostate biopsy? BJU Int.

2014;114(3):384-388. doi:10.1111/bju.12536.

91. Panebianco V, Barchetti F, Manenti G, Aversa T, Catalano C, Simonetti G. MR imaging

� guided prostate biopsy  : technical features and preliminary results. Radiol Med.

2015:571-578. doi:10.1007/s11547-014-0490-0.

92. Filson CP, Natarajan S, Margolis DJA, et al. Prostate cancer detection with magnetic

resonance-ultrasound fusion biopsy: The role of systematic and targeted biopsies. Cancer.

2016;122(6):884-892. doi:10.1002/cncr.29874.

93. Bjurlin MA, Mendhiratta N, Wysock JS, Taneja SS. Multiparametric MRI and targeted

prostate biopsy: Improvements in cancer detection, localization, and risk assessment.

Cent Eur J Urol. 2016;69(1):9-18. doi:10.5173/ceju.2016.734.

94. Puech P, Potiron E, Lemaitre L, et al. Dynamic Contrast-enhanced-magnetic Resonance

Imaging Evaluation of Intraprostatic Prostate Cancer: Correlation with Radical

Prostatectomy Specimens. Urology. 2009;74(5):1094-1099.

doi:10.1016/j.urology.2009.04.102.

95. Delongchamps NB, Rouanne M, Flam T, et al. Multiparametric magnetic resonance

imaging for the detection and localization of prostate cancer: Combination of T2-

weighted, dynamic contrast-enhanced and diffusion-weighted imaging. BJU Int.

2011;107(9):1411-1418. doi:10.1111/j.1464-410X.2010.09808.x.

96. Sosnowski R, Zagrodzka M, Borkowski T. The limitations of multiparametric magnetic

resonance imaging also must be borne in mind. Cent Eur J Urol. 2016;69(1):22-23.

doi:10.5173/ceju.2016.e113.

97. Biomarkers Definitions Working Group. Biomarkers and surrogate endpoints: preferred

definitions and conceptual framework. Clin Pharmacol Ther. 2001;69(3):89-95.

doi:10.1067/mcp.2001.113989.

98. Verma M, Patel P, Verma M. Biomarkers in prostate cancer epidemiology. Cancers

(Basel). 2011;3(4):3773-3798. doi:10.3390/cancers3043773.

99. Hatakeyama S, Yoneyama T, Tobisawa Y, Ohyama C. Recent progress and perspectives

on prostate cancer biomarkers. Int J Clin Oncol. 2016:1-8. doi:10.1007/s10147-016-

1049-y.

100. Barbieri CE, Bangma CH, Bjartell A, et al. The mutational landscape of prostate cancer.

 

   

83  Eur Urol. 2013;64(4):567-576. doi:10.1016/j.eururo.2013.05.029.

101. Tomlins S a, Rhodes DR, Perner S, et al. Recurrent Fusion of TMPRSS2 and. Science

(80- ). 2005;310(October):644-648. doi:10.1126/science.1117679.

102. Demichelis F, Fall K, Perner S, et al. TMPRSS2:ERG gene fusion associated with lethal

prostate cancer in a watchful waiting cohort. Oncogene. 2007;26(31):4596-4599.

doi:10.1038/sj.onc.1210630.

103. Winnes M, Lissbrant E, Damber JE, Stenman G. Molecular genetic analyses of the

TMPRSS2-ERG and TMPRSS2-ETV1 gene fusions in 50 cases of prostate cancer. Oncol

Rep. 2007;17(5):1033-1036.

104. Scheble VJ, Scharf G, Braun M, et al. ERG rearrangement in local recurrences compared

to distant metastases of castration-resistant prostate cancer. Virchows Arch.

2012;461(2):157-162. doi:10.1007/s00428-012-1270-7.

105. Hessels D, Smit FP, Verhaegh GW, Witjes JA, Cornel EB, Schalken JA. Detection of

TMPRSS2-ERG fusion transcripts and prostate cancer antigen 3 in urinary sediments

may improve diagnosis of prostate cancer. Clin Cancer Res. 2007;13(17):5103-5108.

doi:10.1158/1078-0432.CCR-07-0700.

106. Guttman M, Amit I, Garber M, et al. Chromatin signature reveals over a thousand highly

conserved large non-coding RNAs in mammals. Nature. 2009;458(7235):223-227.

doi:10.1038/nature07672.

107. Bussemakers MJG, Van Bokhoven A, Verhaegh GW, et al. DD3: A new prostate-specific

gene, highly overexpressed in prostate cancer. Cancer Res. 1999;59(23):5975-5979.

doi:10.1038/ncb2161.

108. Deras IL, Aubin SMJ, Blase A, et al. PCA3: A Molecular Urine Assay for Predicting

Prostate Biopsy Outcome. J Urol. 2008;179(4):1587-1592.

doi:10.1016/j.juro.2007.11.038.

109. Capoluongo E, Zambon CF, Basso D, et al. PCA3 score of 20 could improve prostate

cancer detection: Results obtained on 734 Italian individuals. Clin Chim Acta.

2014;429:46-50. doi:10.1016/j.cca.2013.10.022.

110. Jurkowska RZ, Jurkowski TP, Jeltsch A. Structure and Function of Mammalian DNA

Methyltransferases. ChemBioChem. 2011;12(2):206-222. doi:10.1002/cbic.201000195.

111. Deaton A, Bird A. CpG islands and the regulation of transcription. Genes Dev.

2011;25(10):1010-1022. doi:10.1101/gad.2037511.1010.

 

   

84  112. Sproul D, Meehan RR. Genomic insights into cancer-associated aberrant CpG island

hypermethylation. Brief Funct Genomics. 2013;12(3):174-190. doi:10.1093/bfgp/els063.

113. Mikeska T, Craig JM. DNA methylation biomarkers: Cancer and beyond. Genes (Basel).

2014;5(3):821-864. doi:10.3390/genes5030821.

114. Strand SH, Orntoft TF, Sorensen KD. Prognostic DNA Methylation Markers for Prostate

Cancer. Vol 15.; 2014. doi:10.3390/ijms150916544.

115. Steiner I, Jung K, Schatz P, et al. Gene promoter methylation and its potential relevance

in early prostate cancer diagnosis. Pathobiology. 2010;77(5):260-266.

doi:10.1159/000318017.

116. Liu L, Yoon J-H, Dammann R, Pfeifer GP. Frequent hypermethylation of the RASSF1A

gene in prostate cancer. Oncogene. 2002;21(44):6835-6840. doi:10.1038/sj.onc.1205814.

117. Jerónimo C, Bastian PJ, Bjartell A, et al. Epigenetics in prostate cancer: Biologic and

clinical relevance. Eur Urol. 2011;60(4):753-766. doi:10.1016/j.eururo.2011.06.035.

118. Blute ML, Damaschke N a, Jarrard DF. The epigenetics of prostate cancer diagnosis and

prognosis: update on clinical applications. Curr Opin Urol. 2015;25(1):83-88.

doi:10.1097/MOU.0000000000000132.

119. Rouprêt M, Hupertan V, Yates DR, et al. Molecular detection of localized prostate cancer

using quantitative methylation-specific PCR on urinary cells obtained following prostate

massage. Clin Cancer Res. 2007;13(6):1720-1725. doi:10.1158/1078-0432.CCR-06-

2467.

120. Trock BJ, Brotzman MJ, Mangold LA, et al. Evaluation of GSTP1 and APC methylation

as indicators for repeat biopsy in a high-risk cohort of men with negative initial prostate

biopsies. BJU Int. 2012;110(1):56-62. doi:10.1111/j.1464-410X.2011.10718.x.

121. Brikun I, Nusskern D, Gillen D, et al. A panel of DNA methylation markers reveals

extensive methylation in histologically benign prostate biopsy cores from cancer patients.

Biomark Res. 2014;2(1):25. doi:10.1186/s40364-014-0025-9.

122. Troyer DA, Lucia MS, De Bruïne AP, et al. Prostate cancer detected by methylated gene

markers in histopathologically cancer-negative tissues from men with subsequent positive

biopsies. Cancer Epidemiol Biomarkers Prev. 2009;18(10):2717-2722.

doi:10.1158/1055-9965.EPI-09-0068.

123. Van Neste L, Herman JG, Otto G, Bigley JW, Epstein JI, Van Criekinge W. The

Epigenetic promise for prostate cancer diagnosis. Prostate. 2012;72(11):1248-1261.

 

   

85  doi:10.1002/pros.22459.

124. Chen Y, Li J, Yu X, et al. APC gene hypermethylation and prostate cancer: a systematic

review and meta-analysis. Eur J Hum Genet. 2013;21(9):929-935.

doi:10.1038/ejhg.2012.281.

125. Vasiljević N, Wu K, Brentnall AR, et al. Absolute quantitation of DNA methylation of 28

candidate genes in prostate cancer using pyrosequencing. Dis Markers. 2011;30(4):151-

161. doi:10.3233/DMA-2011-0790.

126. Møller M, Strand SH, Mundbjerg K, et al. Heterogeneous patterns of DNA methylation-

based field effects in histologically normal prostate tissue from cancer patients. Sci Rep.

2017;7(January):40636. doi:10.1038/srep40636.

127. Yang B, Bhusari S, Kueck J, et al. Methylation profiling defines an extensive field defect

in histologically normal prostate tissues associated with prostate cancer. Neoplasia.

2013;15(4):399-408. doi:10.1593/neo.13280.

128. Devaney JM, Wang S, Furbert-Harris P, et al. Genome-wide differentially methylated

genes in prostate cancer tissues from African-American and Caucasian men. Epigenetics.

2015;10(4):319-328. doi:10.1080/15592294.2015.1022019.

129. Bhasin JM, Lee BH, Matkin L, et al. Methylome-wide Sequencing Detects DNA

Hypermethylation Distinguishing Indolent from Aggressive Prostate Cancer. Cell Rep.

2015:2135-2146. doi:10.1016/j.celrep.2015.10.078.

130. Malumbres M. MiRNAs and cancer: An epigenetics view. Mol Aspects Med.

2013;34(4):863-874. doi:10.1016/j.mam.2012.06.005.

131. Bertoli G, Cava C, Castiglioni I. MicroRNAs as biomarkers for diagnosis, Prognosis and

theranostics in prostate cancer. Int J Mol Sci. 2016;17(3). doi:10.3390/ijms17030421.

132. Wach S, Nolte E, Szczyrba J, et al. MicroRNA profiles of prostate carcinoma detected by

multiplatform microRNA screening. Int J Cancer. 2012;130(3):611-621.

doi:10.1002/ijc.26064.

133. Kelly BD, Miller N, Sweeney KJ, et al. A Circulating MicroRNA Signature as a

Biomarker for Prostate Cancer in a High Risk Group. J Clin Med. 2015;4(7):1369-1379.

doi:10.3390/jcm4071369.

134. Haldrup C, Kosaka N, Ochiya T, et al. Profiling of circulating microRNAs for prostate

cancer biomarker discovery. Drug Deliv Transl Res. 2014;4(1):19-30.

doi:10.1007/s13346-013-0169-4.

 

   

86  135. Moltzahn F, Olshen AB, Baehner L, et al. Microfluidic-based multiplex qRT-PCR

identifies diagnostic and prognostic microRNA signatures in the sera of prostate cancer

patients. Cancer Res. 2011;71(2):550-560. doi:10.1158/0008-5472.CAN-10-1229.

136. Chen Z-H, Zhang G-L, Li H-R, et al. A panel of five circulating microRNAs as potential

biomarkers for prostate cancer. Prostate. 2012;72(13):1443-1452.

doi:10.1002/pros.22495.

137. Alinezhad S, Väänänen R-M, Ochoa NT, et al. Global expression of AMACR transcripts

predicts risk for prostate cancer - a systematic comparison of AMACR protein and

mRNA expression in cancerous and noncancerous prostate. BMC Urol. 2016;16:10.

doi:10.1186/s12894-016-0128-8.

138. Kristiansen G, Fritzsche FR, Wassermann K, et al. GOLPH2 protein expression as a

novel tissue biomarker for prostate cancer: implications for tissue-based diagnostics. Br J

Cancer. 2008;99(6):939-948. doi:10.1038/sj.bjc.6604614.

139. Uetsuki H, Tsunemori H, Taoka R, Haba R, Ishikawa M, Kakehi Y. Expression of a

novel biomarker, EPCA, in adenocarcinomas and precancerous lesions in the prostate. J

Urol. 2005;174(2):514-518. doi:10.1097/01.ju.0000165154.41159.b1.

140. Wei S, Dunn TA, Isaacs WB, De Marzo AM, Luo J. GOLPH2 and MYO6: putative

prostate cancer markers localized to the Golgi apparatus. Prostate. 2008;68(13):1387-

1395. doi:10.1002/pros.20806.

141. Laxman B, Morris DS, Yu J, et al. A First-Generation Multiplex Biomarker Analysis of

Urine for the Early Detection of Prostate Cancer. Cancer Res. 2008;68(3):645-649.

doi:10.1158/0008-5472.CAN-07-3224.

142. Jansen FH, van Schaik RHN, Kurstjens J, et al. Prostate-Specific Antigen (PSA) Isoform

p2PSA in Combination with Total PSA and Free PSA Improves Diagnostic Accuracy in

Prostate Cancer Detection. Eur Urol. 2010;57(6):921-927.

doi:10.1016/j.eururo.2010.02.003.

143. Lazzeri M, Haese A, De La Taille A, et al. Serum isoform [-2]proPSA derivatives

significantly improve prediction of prostate cancer at initial biopsy in a total PSA range

of 2-10 ng/ml: A multicentric european study. Eur Urol. 2013;63(6):986-994.

doi:10.1016/j.eururo.2013.01.011.

144. Braun K, Sjoberg DD, Vickers AJ, Lilja H, Bjartell AS. A Four-kallikrein Panel Predicts

High-grade Cancer on Biopsy: Independent Validation in a Community Cohort. Eur

 

   

87  Urol. 2016;69(3):505-511. doi:10.1016/j.eururo.2015.04.028.

145. Vickers AJ, Cronin AM, Roobol MJ, et al. A four-kallikrein panel predicts prostate

cancer in men with recent screening: Data from the european randomized study of

screening for prostate cancer, Rotterdam. Clin Cancer Res. 2010;16(12):3232-3239.

doi:10.1158/1078-0432.CCR-10-0122.

146. Tomlins SA, Day JR, Lonigro RJ, et al. Platinum Priority – Prostate Cancer Urine

TMPRSS2:ERG Plus PCA3 for Individualized Prostate Cancer Risk Assessment.

2016;0(4):5-5. doi:10.1016/j.eururo.2015.04.039.

147. Salami SS, Schmidt F, Laxman B, et al. Combining urinary detection of TMPRSS2: ERG

and PCA3 with serum PSA to predict diagnosis of prostate cancer. Urol Oncol Semin

Orig Investig. 2013;31(5):566-571. doi:10.1016/j.urolonc.2011.04.001.

148. Ge L, Meng W, Zhou H, Bhowmick N. Could stroma contribute to field cancerization?

Med Hypotheses. 2010;75(1):26-31. doi:10.1016/j.mehy.2010.01.019.

149. Lochhead P, Chan AT, Nishihara R, et al. Etiologic field effect: reappraisal of the field

effect concept in cancer predisposition and progression. Mod Pathol. 2015;28(1):14-29.

doi:10.1038/modpathol.2014.81.

150. Nonn L, Ananthanarayanan V, Gann PH. Evidence for field cancerization of the prostate.

Prostate. 2009;69(13):1470-1479. doi:10.1002/pros.20983.

151. Trujillo KA, Jones AC, Griffith JK, Bisoffi M. Markers of field cancerization: proposed

clinical applications in prostate biopsies. Prostate Cancer. 2012;2012:302894.

doi:10.1155/2012/302894.

152. Halin S, Hammarsten P, Adamo H, Wikström P, Bergh A. Tumor indicating normal

tissue could be a new source of diagnostic and prognostic markers for prostate cancer.

Expert Opin Med Diagn. 2011;5(1):37-47. doi:10.1517/17530059.2011.540009.

153. Truong M, Yang B, Livermore A, et al. Using the epigenetic field defect to detect

prostate cancer in biopsy negative patients. J Urol. 2013;189(6):2335-2341.

doi:10.1016/j.juro.2012.11.074.

154. Krušlin B, Ulamec M, Tomas D, Tomas D. Prostate cancer stroma: an important factor in

cancer growth and progression. Bosn J Basic Med Sci. 2015;15(2):1-8.

doi:10.17305/bjbms.2015.449.

155. Richiardi L, Fiano V, Grasso C, et al. Methylation of APC and GSTP1 in Non-Neoplastic

Tissue Adjacent to Prostate Tumour and Mortality from Prostate Cancer. PLoS One.

 

   

88  2013;8(7):1-7. doi:10.1371/journal.pone.0068162.

156. Geybels MS, Zhao S, Wong CJ, et al. Epigenomic profiling of DNA methylation in

paired prostate cancer versus adjacent benign tissue. Prostate. 2015;75(16):1941-1950.

doi:10.1002/pros.23093.

157. Mehrotra, Jyoti, Varde,Shobha, Wang, Haiying, Chiu, Hsiling, Vargo, Janet, Gray,

Karen, Nagle, Raymond B., Neri, Jaclyn R., Mazumder A. Quantitative, Spatial

Resolution of the Epigenetic Field Effect in Prostate Cancer. Prostate. 2008;68(2):161-

170. doi:10.1002/pros.

158. How Kit A, Nielsen HM, Tost J. DNA methylation based biomarkers: Practical

considerations and applications. Biochimie. 2012;94(11):2314-2337.

doi:10.1016/j.biochi.2012.07.014.

159. Stewart GD, Van Neste L, Delvenne P, et al. Clinical utility of an epigenetic assay to

detect occult prostate cancer in histopathologically negative biopsies: Results of the

MATLOC study. J Urol. 2013;189(3):1110-1116. doi:10.1016/j.juro.2012.08.219.

160. Partin AW, Van Neste L, Klein EA, et al. Clinical validation of an epigenetic assay to

predict negative histopathological results in repeat prostate biopsies. J Urol.

2014;192(4):1081-1087. doi:10.1016/j.juro.2014.04.013.

161. Wojno KJ, Costa FJ, Cornell RJ, et al. Reduced rate of repeated prostate biopsies

observed in ConfirmMDx clinical utility field study. Am Heal Drug Benefits.

2014;7(3):129-134.

162. Van Neste L, Partin AW, Stewart GD, Epstein JI, Harrison DJ, Van Criekinge W. Risk

score predicts high-grade prostate cancer in DNA-methylation positive,

histopathologically negative biopsies. Prostate. 2016;76(12):1078-1087.

doi:10.1002/pros.23191.

163. Luo JH, Ding Y, Chen R, et al. Genome-wide methylation analysis of prostate tissues

reveals global methylation patterns of prostate cancer. Am J Pathol. 2013;182(6):2028-

2036. doi:10.1016/j.ajpath.2013.02.040.

164. Touleimat N, Tost J. Complete pipeline for Infinium(®) Human Methylation 450K

BeadChip data processing using subset quantile normalization for accurate DNA

methylation estimation. Epigenomics. 2012;4(3):325-341. doi:10.2217/epi.12.21.

165. Kron KJ, Liu L, Pethe V V, et al. DNA methylation of HOXD3 as a marker of prostate

cancer progression. Lab Invest. 2010;90(7):1060-1067. doi:10.1038/labinvest.2010.57.

 

   

89  166. Du P, Zhang X, Huang C-C, et al. Comparison of Beta-value and M-value methods for

quantifying methylation levels by microarray analysis. BMC Bioinformatics.

2010;11(1):587. doi:10.1186/1471-2105-11-587.

167. Fortin J-P, Labbe A, Lemire M, et al. Functional normalization of 450k methylation array

data improves replication in large cancer studies. Genome Biol. 2014;15(12):503.

doi:10.1186/s13059-014-0503-2.

168. Heiss JA, Brenner H. Between-array normalization for 450K data. Front Genet.

2015;5(FEB):1-7. doi:10.3389/fgene.2015.00092.

169. Smith GK. limma: Linear Models for Microarray Data. Bioinforma Comput Biol Solut

Using R Bioconductor. 2005;(2005):397-420. doi:citeulike-article-id:5722720.

170. Du P, Bourgon R. methyAnalysis  : an R package for DNA methylation data analysis and

visualization. 2016:1-11.

171. Jaffe AE, Murakami P, Lee H, et al. Bump hunting to identify differentially methylated

regions in epigenetic epidemiology studies. Int J Epidemiol. 2012;41(1):200-209.

doi:10.1093/ije/dyr238.

172. Eads CA, Danenberg KD, Kawakami K, et al. MethyLight: a high-throughput assay to

measure DNA methylation. Nucleic Acids Res. 2000;28(8):E32.

doi:10.1093/nar/28.8.e32.

173. Trinh BN, Long TI, Laird PW. DNA methylation analysis by MethyLight technology.

Methods. 2001;25(4):456-462. doi:10.1006/meth.2001.1268.

174. Olkhov-Mitsel E, Zdravic D, Kron K, van der Kwast T, Fleshner N, Bapat B. Novel

multiplex MethyLight protocol for detection of DNA methylation in patient tissues and

bodily fluids. Sci Rep. 2014;4:4432. doi:10.1038/srep04432.

175. Weisenberger DJ, Campan M, Long TI, et al. Analysis of repetitive element DNA

methylation by MethyLight. Nucleic Acids Res. 2005;33(21):6823-6836.

doi:10.1093/nar/gki987.

176. Liu L, Kron KJ, Pethe V V., et al. Association of tissue promoter methylation levels of

APC, TGFβ2, HOXD3 and RASSF1A with prostate cancer progression. Int J Cancer.

2011;129(10):2454-2462. doi:10.1002/ijc.25908.

177. Zhao Z, Liu G, Li X, et al. Characterization of the promoter regions of two sheep keratin-

associated protein genes for hair cortex-specific expression. PLoS One. 2016;11(4):1-17.

doi:10.1371/journal.pone.0153936.

 

   

90  178. Udhane S, Kempna P, Hofer G, Mullis PE, Flück CE. Differential Regulation of Human

3β-Hydroxysteroid Dehydrogenase Type 2 for Steroid Hormone Biosynthesis by

Starvation and Cyclic Amp Stimulation: Studies in the Human Adrenal NCI-H295R Cell

Model. PLoS One. 2013;8(7). doi:10.1371/journal.pone.0068691.

179. Jernberg E, Thysell E, Ylitalo EB, et al. Characterization of prostate cancer bone

metastases according to expression levels of steroidogenic enzymes and androgen

receptor splice variants. PLoS One. 2013;8(11). doi:10.1371/journal.pone.0077407.

180. Romanelli V, Nakabayashi K, Vizoso M, et al. Variable maternal methylation

overlapping the nc886/vtRNA2-1 locus is locked between hypermethylated repeats and is

frequently altered in cancer. Epigenetics. 2014;9(5):783-790. doi:10.4161/epi.28323.

181. Kong L, Hao Q, Wang Y, Zhou P, Zou B, Zhang Y. Regulation of p53 expression and

apoptosis by vault RNA2-1-5p in cervical cancer cells. Oncotarget. 2015;6(29):28371-

28388. doi:10.18632/oncotarget.4948.

182. Lee H-S, Lee K, Jang H-J, et al. Epigenetic silencing of the non-coding RNA nc886

provokes oncogenes during human esophageal tumorigenesis. Oncotarget.

2014;5(11):3472-3481. doi:10.18632/oncotarget.1927.

183. Kwabi-Addo B, Chung W, Shen L, et al. Age-related DNA methylation changes in

normal human prostate tissues. Clin Cancer Res. 2007;13(13):3796-3802.

doi:10.1158/1078-0432.CCR-07-0085.

184. Tost J, Gut IG. DNA methylation analysis by pyrosequencing. Nat Protoc.

2007;2(9):2265-2275. doi:10.1038/nprot.2007.314.

185. Weisenberger DJ, Trinh BN, Campan M, et al. DNA methylation analysis by digital

bisulfite genomic sequencing and digital MethyLight. Nucleic Acids Res.

2008;36(14):4689-4698. doi:10.1093/nar/gkn455.

186. Olkhov-Mitsel E, Bapat B. Strategies for discovery and validation of methylated and

hydroxymethylated DNA biomarkers. Cancer Med. 2012;1(2):237-260.

doi:10.1002/cam4.22.

187. Olkhov-Mitsel E, Van der Kwast T, Kron KJ, et al. Quantitative DNA methylation

analysis of genes coding for kallikrein-related peptidases 6 and 10 as biomarkers for

prostate cancer. Epigenetics. 2012;7(9):1037-1045. doi:10.4161/epi.21524.

188. Gehring WJ, Affolter M, Bürglin T. Homeodomain proteins. Annu Rev Biochem.

1994;63(1):487-526. doi:10.1146/annurev.bi.63.070194.002415.

 

   

91  189. Zhao F, Olkhov-Mitsel E, van der Kwast T, et al. Urinary DNA Methylation Biomarkers

for Noninvasive Prediction of Aggressive Disease in Patients with Prostate Cancer on

Active Surveillance. J Urol. 2017;197(2):335-341. doi:10.1016/j.juro.2016.08.081.

190. Strand SH, Switnicki M, Moller M, et al. RHCG and TCAF1 promoter hypermethylation

predicts biochemical recurrence in prostate cancer patients treated by radical

prostatectomy. 2016;8(4):5774-5788. doi:10.18632/oncotarget.14391.

191. Holmes EE, Goltz D, Sailer V, et al. PITX3 promoter methylation is a prognostic

biomarker for biochemical recurrence-free survival in prostate cancer patients after

radical prostatectomy. Clin Epigenetics. 2016;8:104. doi:10.1186/s13148-016-0270-x.