Prostate Cancer

A Panel of TMPRSS2:ERG Fusion Transcript Markers for

Urine-Based Prostate Cancer Detection with High Specificity

and Sensitivity

Phuong-Nam Nguyen, Philippe Violette, Sam Chan, Simon Tanguay, Wassim Kassouf,Armen Aprikian, Junjian Z. Chen *

Department of Surgery/Division of Urology, McGill University Health Center, Montreal, Quebec, H3G 1A4 Canada

E U R O P E A N U R O L O G Y 5 9 ( 2 0 1 1 ) 4 0 7 – 4 1 4

ava i lable at www.sciencedirect .com

journal homepage: www.europeanurology.com

Article info

Article history:Accepted November 17, 2010Published online ahead ofprint on November 26, 2010

Keywords:

TMPRSS2:ERG fusion

Prostate cancer

Urine test

ERG

Cancer diagnosis

Abstract

Background: The TMPRSS2:ERG fusion is both prevalent and unique to prostate

cancer (PCa) and has great potential for noninvasive diagnosis of PCa in bodily fluids.

Objectives: To evaluate the specificity and sensitivity of the TMPRSS2:ERG fusion in

urine from diverse clinical contexts and to explore potential clinical applications.

Design, setting, and participants: A total of 101 subjects were enrolled in 2008 from

urologic oncology clinics to form three study groups: 44 PCa free, 46 confirmed PCa,

and 11 negative prostate biopsies. The PCa-free group included females, healthy

young men, and post–radical prostatectomy (RP) patients. The confirmed PCa

group was composed of patients under active surveillance, scheduled for treat-

ment, or with metastatic disease.

Measurements: Urine was collected after attentive digital rectal exam (DRE) and

coded to blind group allocation for laboratory test. RNA from urine sediments was

analyzed using a panel of four TMPRSS2:ERG fusion markers with quantitative

polymerase chain reaction (qPCR).

Results and limitations: Our fusion markers demonstrated very high technical

specificity and sensitivity for detecting a single fusion-positive cancer cell (VCaP)

in the presence of at least 3000 cells in urine sediments. In clinical analysis, there

were no fusion-positive samples in the PCa-free group (0 of 44 samples), while there

were 16 of 46 (34.8%) fusion-positive samples in the confirmed PCa group. The fusion

incidence varied significantly among the three PCa subgroups. The clinical sensitivity

increased to 45.4% in cancer patients prior to treatments. The fusion markers were

detected in 2 of 11 (18.2%) biopsy-negative patients, suggesting potentially false

negative biopsies. This study is not prospective and is limited in sample sizes.

Conclusions: Our novel panel of TMPRSS2:ERG fusion markers provided a very

specific and sensitive tool for urine-based detection of PCa. Theses markers can

potentially be used to diagnose patients with PCa who have negative biopsies.

soc

. Department of Surgery/Division of Urology, Research Institute–McGillr, Room R2.133, 1650 Cedar Avenue, Montreal, Quebec, H3G 1A4 Canada.4601; Fax: +1 514 934 8261.chen@mcgill.ca (J.Z. Chen).

# 2010 European As

* Corresponding authorUniversity Health CenteTel. +1 514 934 1934x4E-mail address: junjian.

0302-2838/$ – see back matter # 2010 European Association of Urology. Publis

iation of Urology. Published by Elsevier B.V. All rights reserved.

hed by Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2010.11.026

E U R O P E A N U R O L O G Y 5 9 ( 2 0 1 1 ) 4 0 7 – 4 1 4408

1. Introduction

As a seminal discovery, many prostate tumors contain a

specific genetic change that involves the fusion of two genes

[1,2]. The fusion of an androgen-regulated gene, TMPRSS2, to

a nearby E twenty-six (ETS) transformation-specific tran-

scription factor gene, ERG, has been reported as the most

common genetic change in prostate cancer (PCa), found in

approximately 50% of surgical tumors [3–8]. The TMPRSS2

gene also becomes fused to other ETS genes in approximately

10% of cancer cases [1–3,9]. Such fusion events provide a

novel mechanism for androgen-regulated overexpression of

the ETS genes, with immediate implications in PCa progres-

sion [10,11]. ERG overexpression has recently been shown to

inhibit normal prostate differentiation by shutting down

androgen signaling [12] and to induce an embryonic stem

cell–like dedifferentiation program by turning on EZH2

expression [12–14].

Meanwhile, diverse TMPRSS2:ERG fusion subtypes have

been uncovered, ranging from chromosomal rearrangements

to fusion transcripts [6,8,15–17]. Such fusion subtypes not

only allow stratification of clinically aggressive forms of

cancer [6,15] but also provide redundant and cancer-specific

transcript markers for potentially noninvasive cancer detec-

tion in bodily fluids. Indeed, a common TMPRSS2:ERG fusion

isoform is shown to be detectable in the urine of men with PCa

and has been coupled with other molecular markers in urine-

based cancer detection [18–20]. However, critical evaluations

remain scarce on the specificity, sensitivity, and clinical utility

of one or multiple fusion markers in urine-based detection. In

this study, we aim to evaluate the specificity and sensitivity—

from both technical and clinical perspectives—of multiple

TMPRSS2:ERG fusion isoforms in diverse clinical contexts and

to explore their potential clinical applications.

2. Materials and methods

2.1. Human subjects

Human subjects were enrolled through an institutional review board–

approved protocol and written informed consent at the McGill

Table 1 – The baseline clinical and pathologic characteristics of distin

Clinical parametersa PCa-free group

Female Young men Post-RP Acti

Samples, no. 18 14 12

Age, yr 67 (34–79) 27.5 (19–37) 61 (48–82)

Presample PSA – – 0.01

(0.01–0.03)

Gleason score – – 6 (6–8)

Metastasis No No No

Treatment

(1-yr follow-up)

– – RP (12)

PCa = prostate cancer; RP = radical prostatectomy; PSA = prostate-specific antigena Values expressed as median, with ranges in parentheses.b Treatments received after urine collection.

University Health Center to form three distinctive study groups:

PCa-free subjects (n = 44), confirmed PCa patients (n = 46), and patients

with negative biopsies (n = 11). The PCa-free group was designed as a

negative control to test the specificity of fusion markers. It included

females (n = 18), healthy males <37 yr of age (n = 14), and post–radical

prostatectomy (RP) patients with a prostate-specific antigen (PSA) of

0 (n = 12). The confirmed PCa group was designed for comparative

analysis of fusion markers among different cancer cohorts. It was

composed of patients under active surveillance (n = 21), those

scheduled for treatment (n = 11), and those with metastasis (n = 14).

The active surveillance cohort consisted of patients with biopsy-

proven PCa who were followed clinically for signs of cancer progression

but were not receiving active therapy. The pretreatment cohort

consisted of patients who had biopsy-proven PCa and were scheduled

for either RP or radiation therapy (RT). The metastatic cohort consisted

of patients with biopsy-proven PCa and evidence of metastasis. This

late definition was independent of therapies (eg, androgen-deprivation

therapy [ADT] or RT) but excluded patients who received RP. Finally,

urine samples from patients with negative biopsies were used to

identify potentially false-negative biopsies. The baseline clinical

characteristics of each cohort and 1-yr follow-up of treatments are

provided in Table 1.

2.2. Urine collection and prostate cancer cell lines

The first voided urine after attentive digital rectal examination (DRE)

was collected from male subjects. Female and post-RP subjects were

exempted from DRE because of the absence of a prostate, and 10–40 ml

of urine was collected from each subject in a sterile collection cup

containing RNA/DNA preservatives (Sierra Diagnostics). Urine sediments

were collected by low-speed centrifugation at 4 8C and resuspended in

TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) for immediate RNA

extraction or stored at�80 8C until use. The PCa cell lines, VCaP and NCI-

H660, were gifts from Dr. van Bokhoven at the University of Colorado

Denver Health Sciences Center.

2.3. Preparation of whole transcriptome amplification

complementary DNA libraries

Total RNA was extracted from urine sediment and cancer cell lines using

the TRIzol Reagent. A whole transcriptome amplification (WTA) step was

used to generate a complementary DNA (cDNA) library from a limited

amount of RNA from each sample using a TransPlex WTA kit (Sigma-

Aldrich, St. Louis, MO, USA) according to the manufacturer’s instructions

[18].

ctive clinical cohorts

Confirmed PCa group Negative biopsy

ve surveillance Pretreatment Metastatic

21 11 14 11

73 (55–82) 68 (53–85) 77 (73–88) 67 (51–81)

5.29

(1.55–33.09)

10.54

(5.32–22.7)

13.29

(0.01–399.04)

4.57

(0.56–9.24)

6 (6–7) 6 (6–9) 9 (7–10) –

No No Yes (14) No

– RP (6)b

ADT (3)

RT (3)

ADT(13)

RT (2)

–

; ADT = androgen-deprivation therapy; RT = radiation therapy.

Table 2 – Primer sequences of TMPRSS2:ERG fusion markers and additional molecular markers

Gene Primer Sequence Amplicon (bp) Annealing Tma (8C)

TM-e1:ER-e4; isoform I TM-e1:ER-e4F

ERG-e4R(1)

CTGGAGCGCGGCAGGAA

GTAGGCACACTCAAACAACGACTGG

65 70

TM-e2:ER-e4; isoform II TM-e2:ER-e4F

ERG-e4R(2)

GATGGCTTTGAACTCAGAAGC

TCCGTAGGCACACTCAAACAAC

70 64

TM-e1:ER-e2; isoform III TM-e1:ER-e2F

ERG-e2R(1)

TGGAGCGCGGCAGGTTATT

TTGTCTTGCTTTTGGTCAACACG

70 70

TM-e1:ER-e5; isoform IV TM-e1:ER-e5F

ERG-e5R(1)

GGAGCGCGGCAGGAACT

GTTCATCCCAACGGTGTCTGG

85 70

Isoform I and II flanking TM-e1F

ERG-e4R(3)

GGAGCGCCGCCTGGAG

GTCTTAGCCAGGTGTGGCGTTC

98 or 169 70.5

Isoform III flanking TM-e1F

ERG-e2R(2)

GGAGCGCCGCCTGGAG

TTGCCCTTGGTTCTGCCATC

120 70

ERG(5-6) ERG5-6F

ERG5-6R

CGCAGAGTTATCGTGCCAGCAGAT

CCATATTCTTTCACCGCCCACTCC

86 67

ERG(6-7) ERG6-7F

ERG6-7R

AGCTACAACGCCGACATCC

GAAGTCAAATGTGGAAGAGGAGTC

71 67

GAPDH GAPDH(3)f

GAPDH(3)r

AAGGTCGGAGTCAACGGATTT

ACCAGAGTTAAAAGCAGCCCTG

66 69

qPCR = quantitative polymerase chain reaction.a The qPCR program consisted of initial denaturing at 95 8C for 1.5 min, followed by 50 cycles of a two-step reaction at 95 8C for 15 s and 64–70 8C (varying for

each marker) for 30 s. The qPCR was performed using the MyiQ real-time PCR system (Bio-Rad).

E U R O P E A N U R O L O G Y 5 9 ( 2 0 1 1 ) 4 0 7 – 4 1 4 409

2.4. Detection of TMPRSS2:ERG fusion and other markers

Real-time quantitative polymerase chain reaction (qPCR) was used to

detect a panel of TMPRSS2:ERG fusion markers. Additional markers were

also quantified, including two ERG markers targeting exons 5-6 and 6-7

and a housekeeping gene GADPH. The primer sequences are listed in

Table 2. For qPCR, 9 ng of cDNA were amplified in a 20-ml reaction

containing 1X SYBR Green Supermix (Bio-Rad, Hercules, CA, USA) and

300 nM each of forward and reverse primers using a two-step program

(see Table 2). A melt-curve analysis step was enabled at the end of the

amplification, and the threshold cycle (Ct) was calculated using the MyiQ

software (Bio-Rad). The relative expression of each target gene was

normalized to the housekeeping gene GAPDH using the comparative

Ct method (Applied Biosystems User Bulletin 2).

2.5. Data analysis

The Mann-Whitney U test was used to compare the log-transformed

relative expression of molecular markers from clinical cohorts. This

nonparametric test was also used to assess statistical significance in the

clinical cohorts stratified by fusion status for ERG markers. The x2 test

was used to establish an association between confirmed PCa subgroups

and the presence of TMPRSS2:ERG fusion markers; it was also used to

assess an association between the ERG expression level and fusion status

in PCa patients or between ERG overexpression and ADT in fusion-

negative PCa patients. All statistical analyses were run on GraphPad

Prism v.4 software (GraphPad, La Jolla, CA, USA). Two-sided p values

<0.05 were considered statistically significant.

3. Results

3.1. Validation of a new panel of TMPRSS2:ERG fusion

markers for urine detection

A panel of isoform-specific fusion probes was designed

based on fusion junction sequences and validated for the

common fusion isoform I and three additional reported

isoforms (II, III, and IV; Table 2). The technical specificity of

the new probe for fusion isoform I was confirmed by the

expected melt-curve and fragment size of amplified

products in a fusion-positive cell line (VCaP) and by the

absence of unspecific amplification in fusion-negative urine

samples within a 50-cycle qPCR reaction (Fig. 1A and B). The

technical sensitivity of the fusion marker was evaluated in a

titration experiment that mixed various amounts of VCaP

RNA with fusion-negative urine RNA for a total of 100 ng

prior to WTA amplification. We demonstrated that as little

as 0.01% fusion-positive cancer RNA (ie, 10 pg) could be

effectively detected for fusion isoform I using this approach

(Fig. 1C).

3.2. Urine-based detection of TMPRSS2:ERG fusion markers

in diverse clinical cohorts

The WTA cDNA templates from 101 urine samples were

analyzed using the new fusion markers in a blind fashion

and without prior selection with urine PSA (Fig. 2). No

fusion isoforms were detected in 44 PCa-free cases,

representing a very high clinical specificity. In contrast,

16 urine samples were detected with at least one fusion

marker in 46 confirmed PCa cases, representing a clinical

sensitivity of 34.8%. However, the distribution of fusion-

positive cases varied significantly between cancer cohorts,

with 47.6% in the active surveillance cohort and 14.3% in the

metastatic cohort. All metastatic patients except one had

received ADT at the time of urine collection. Coincidentally,

the sole hormone-naı̈ve case was detected with multiple

fusion isoforms. Thus, the low fusion incidence in the

metastatic cohort could be associated with ADT treatment.

When the negative-biopsy cohort was screened for fusion

markers, 2 of 11 (18.2%) patients were found to be fusion

positive, one of whom was subsequently diagnosed with

PCa on repeat biopsy. In contrast, of the 18 fusion-positive

samples, 17 were detected with the common isoform I, and

7 contained two or more fusion isoforms (Table 3). All

fusion-positive samples were confirmed independently

Fig. 1 – Validation of the TMPRSS2:ERG fusion marker with a prostatecancer cell line and urine samples. (A) A melt-curve specific to the fusionisoform I amplified from a fusion-positive cell line (VCaP) using the MyiQreal-time polymerase chain reaction (PCR) system. (B) Gel electrophoresisof PCR products generated by specific (upper panel) versus unspecific(lower panel) markers for fusion isoform I. A new marker designed in thecurrent study generated an expected PCR fragment only in the VCaP cellline but not in fusion-negative urine samples (upper panel); a less-specificmarker generated the expected PCR fragment not only in the VCaP cell linebut also in fusion-negative urine samples after 35 cycles (lower panel). (C)Detection of titrated VCaP RNA (fusion isoform I) in fusion-negative urineRNA. The log value of relative amplification was normalized to the signal of100% VCaP RNA. The 0.0% amplification was defined by the lack of specificamplification in a 50-cycle reaction.Ct = threshold cycle; WTA = whole transcriptome amplification.

E U R O P E A N U R O L O G Y 5 9 ( 2 0 1 1 ) 4 0 7 – 4 1 4410

using probes flanking the fusion markers, which also ruled

out potential cross-contamination from fusion PCR prod-

ucts. However, the fusion status was not associated with

Gleason score or cancer progression in the current study,

probably because of the limited number of samples and

short follow-up time.

3.3. Associations of ERG overexpression in urine with

TMPRSS2:ERG fusion and metastasis

Two ERG markers targeting exons 5-6 and 6-7 were

detectable in all seven clinical cohorts (Fig. 3A). When

confirmed PCa cases were stratified by the TMPRSS2:ERG

fusion status, a 6- to 15-fold increase in urine ERG

expression was observed among fusion-positive versus

fusion-negative samples ( p < 0.01; Fig. 3B). When the

median ERG expression in fusion-positive cancer cases was

used as a cutoff, 11 of 16 fusion-positive samples were

found to overexpress one or both ERG markers in urine.

Thus, the ERG overexpression in urine was strongly

associated with the fusion status in cancer samples

( p < 0.001; x2 test; Fig. 3C). In contrast, 5 of 30 fusion-

negative cancer samples exhibited ERG overexpression;

four of the overexpressed samples were distributed in 12

fusion-negative metastatic cases ( p < 0.05; x2 test; Fig. 3D).

4. Discussion

The TMPRSS2:ERG fusion is both prevalent and unique to PCa

[21] and hence has great potential for noninvasive diagnosis

and prognosis of PCa. The main challenges of urine-based

fusion detection are technical demands for detecting rare

fusion markers and the need to clarify the specificity of

fusion markers in complex urine specimens. In the current

study, we developed a new panel of TMPRSS2:ERG fusion

markers for a urine-based qPCR test and evaluated their

specificity and sensitivity using fusion-positive cancer cell

lines (VCaP and NCI-H660) and in diverse clinical contexts.

We demonstrate that no fusion markers are detectable in

fusion-negative templates in the qPCR test, while as little as

10 pg of VCaP RNA could be effectively detected in the

presence of 100 ng of fusion-negative urine RNA. This

translates into superior technical specificity and sensitivity

for detecting a single VCaP cancer cell in the presence of at

least 3000 cells in urine sediments.

The clinical specificity and sensitivity of the fusion

markers were evaluated in diverse clinical subjects. We

demonstrate in a blind analysis that no fusion markers are

detectable in 44 subjects from the PCa-free group. The very

high specificity in both technical and clinical terms suggests

that the detectable presence of one or more fusion isoforms

may be sufficient to detect a positive result in bodily fluids.

This ‘‘nonthreshold’’ feature distinguishes the TMPRSS2:ERG

fusion marker from other quantitative markers that require

arbitrary cutoff values [22]. In contrast, one or more fusion

markers are detected in 16 of 46 (34.8%) confirmed PCa

patients. Considering a suppressing effect of ADT on fusion

incidence, the actual clinical sensitivity of fusion markers

will increase to 45.4% in cancer patients prior to any

treatment. This sensitivity is not only higher than previ-

ously reported values [18,19] but is achieved without

prescreening urine specimens. It is useful to point out that

PCa cell lines and urine specimens have limitations in the

[()TD$FIG]

Fig. 2 – Urine-based detection of TMPRSS2:ERG fusion markers in diverse clinical cohorts. The prostate cancer (PCa)–free group consisted of female, youngmen, and post–radical prostatectomy cohorts, while the confirmed PCa group consisted of active surveillance, pretreatment, and metastatic cohorts. Thebar with vertical lines indicates the number of cases in each cohort; the bar with horizontal lines indicates the number of cases detected with at least oneof the fusion markers. The fusion detection rate is indicated above the fusion-positive bar. Values were compared for statistical significance as indicatedby solid lines between clinical groups or subgroups.PCa = prostate cancer.* p < 0.05.**<0.01.***<0.001.

E U R O P E A N U R O L O G Y 5 9 ( 2 0 1 1 ) 4 0 7 – 4 1 4 411

development of new markers and that the usability of the

panel of TMPRSS2:ERG fusion markers could be further

strengthened by validating fusion status in both urine and

surgical tissues of the same patient in future prospective

studies. With exceptional specificity and 35–45% sensitivity,

the urine-based test is used to test the feasibility of detecting

potentially false-negative biopsy cases. As a significant

clinical issue, a typical patient population undergoing a

prostate biopsy based on serum PSA and DRE has a 65–70%

biopsy-negative rate, among which about 20% of biopsy-

negative subjects are diagnosed with PCa in repeat biopsies

Table 3 – The TMPRSS:ERG fusion isoform and combination detected icancer cell lines

Samples, no. Age, yr Gleason score Fusion I

(TM-e1:ER-e

(Ct)a

10 70.2 (55–82) 6 (6–7) 34.0 (31.4–38

3 75 (66–82) 7 (6–9) 32.3 (31.4–33

1 77 7 33.6

1 85 7 35.4

1 80 6 NQ

1 73 6 32.3

1 57 6 27.8

NCI-H660b – – 27.6

VCaPb – – 17.8

Ct = threshold cycle; NQ = not quantifiable; PCa = prostate cancer.a Cutoff Ct value is 50 cycles.b Fusion-positive PCa cell lines.

[23–26]. Surprisingly, two fusion-positive cases were identi-

fied in 11 biopsy-negative patients, with one of the positive

cases confirmed as PCa in a subsequent biopsy in the current

study. This result is not only supported by the detection of

TMPRSS2:ERG fusion markers in the urine of biopsy-negative

patients in a previous study [19] but represents one of

the first successful attempts to validate a false-negative

biopsy patient using urine-based fusion detection. Thus,

the detection of the TMPRSS2:ERG fusion markers in biopsy-

negative patients may represent a distinctive molecular

subgroup, with significant clinical implications in the

n 18 fusion-positive urine samples and 2 fusion-positive prostate

Fusion II Fusion III Fusion IV

4) (TM-e2:ER-e4) (TM-e1:ER-e2) (TM-e1:ER-e5)

(Ct) (Ct) (Ct)

.8) NQ NQ NQ

.7) 32.9 (31.7–34.0) NQ NQ

NQ 36.5 NQ

33.7 34.1 NQ

NQ 32.5 NQ

NQ NQ 42.0

NQ 29.3 26.9

26.1 NQ NQ

NQ NQ 26.1

Fig. 3 – Associations of ERG overexpression in urine with TMPRSS2:ERG fusion and metastasis. (A) The prevalence of two ERG markers targeting exons 5-6 and 6-7 in urine from diverse clinical cohorts. The detectionrate in each cohort was calculated as the percentage of positive samples for each marker. (B) The expression levels of two ERG markers in urine were stratified by fusion status in 46 cases in the confirmed prostatecancer (PCa) group. The relative expression was based on a DD threshold cycle method and transformed into log2 values. (C) ERG overexpression in urine was stratified with fusion status in the confirmed PCagroup. ERG overexpression was defined as a relative expression that was greater than the median expression of either exon 5-6 or 6-7 in fusion-positive cancer cases. (D) ERG overexpression in urine was stratifiedwith androgen-deprivation therapy in fusion-negative cancer cases.RP = radical prostatectomy; ADT = androgen-deprivation therapy.

EU

RO

PE

AN

UR

OL

OG

Y5

9(

20

11

)4

07

–4

14

41

2

E U R O P E A N U R O L O G Y 5 9 ( 2 0 1 1 ) 4 0 7 – 4 1 4 413

management of biopsy-negative patients—a topic that we are

actively investigating in an ongoing prospective study.

The biologic implication of TMPRSS2:ERG fusion is

upregulation of oncogenic ERG expression in PCa cells

[27,28]. We demonstrate that the fusion-positive status is

strongly associated with ERG overexpression in the urine of

cancer patients prior to any treatment (Fig. 3B and C).

Interestingly, ADT significantly reduces TMPRSS2:ERG fu-

sion incidence in the urine of metastatic patients but has

little effect on ERG overexpression.

Although limited in sample size, the frequent ERG

overexpression in TMPRSS2:ERG fusion-negative metastatic

patients can be explained by the possible existence of

additional fusion events or because of altered signal

pathways associated with hormone-refractory cancers.

Regardless of the basis, ERG overexpression may be

essential to both androgen-dependent primary cancer

and hormone-refractory cancer [12]. Thus, ERG overexpres-

sion in urine may serve as a useful surrogate for potential

fusion events associated with PCa [29].

5. Conclusions

We have developed a new panel of TMPRSS2:ERG fusion

markers for a urine-based qPCR test with exceptional

technical specificity and sensitivity. Using these new

markers, we have demonstrated a lack of fusion-positive

samples in PCa-free subjects and a clinical sensitivity of 45.4%

in PCa patients prior to treatments. We suggest that our

TMPRSS2:ERG fusion markers may serve as nonthreshold

markers in urine-based PCa detection and find direct

applications in identifying false-negative biopsy cases. It is

anticipated that the clinical application of the panel of

TMPRSS2:ERG fusion markers could be further improved by

simultaneous detection of all fusion isoforms and additional

fusion genes in a single test.

Author contributions: Junjian Z. Chen had full access to all the data in the

study and takes responsibility for the integrity of the data and the

accuracy of the data analysis.

Study concept and design: Chen, Aprikian, Violette.

Acquisition of data: Chen, Nguyen, Violette, Chan, Aprikian, Kassouf,

Tanguay.

Analysis and interpretation of data: Chen, Nguyen.

Drafting of the manuscript: Chen.

Critical revision of the manuscript for important intellectual content: Chen,

Aprikian, Violette, Kassouf, Tanguay.

Statistical analysis: Chen, Nguyen.

Obtaining funding: Chen, Aprikian.

Administrative, technical, or material support: Chen, Nguyen, Chan.

Supervision: Chen, Aprikian.

Other (specify): None.

Financial disclosures: I certify that all conflicts of interest, including

specific financial interests and relationships and affiliations relevant to the

subject matter or materials discussed in the manuscript (eg, employment/

affiliation, grants or funding, consultancies, honoraria, stock ownership or

options, expert testimony, royalties, or patents filed, received, or pending),

are the following: None.

Funding/Support and role of the sponsor: Canadian Institute of Health

Research (NGH99087) provided support to JZ Chen and A Aprikian, and

the Canada Foundation for Innovation (11623) provided support to JZ

Chen for data collection.

Acknowledgment statement: The authors acknowledge M Chevrette for

comments on an earlier version of the manuscript and D Ayele for

statistical assistance.

References

[1] Tomlins SA, Rhodes DR, Perner S, Dhanasekaran SM, Mehra R,

Sun XW. Recurrent fusion of TMPRSS2 and ETS transcription factor

genes in prostate cancer. Science 2005;310:644–8.

[2] Tomlins SA, Mehra R, Rhodes DR, Smith LR, Roulston D, Helgeson BE.

TMPRSS2:ETV4 gene fusions define a third molecular subtype of

prostate cancer. Cancer Res 2006;66:3396–400.

[3] Hermans KG, van Marion R, van Dekken H, Jenster G, van Weerden

WM, Trapman J. TMPRSS2:ERG fusion by translocation or intersti-

tial deletion is highly relevant in androgen-dependent prostate

cancer, but is bypassed in late-stage androgen receptor-negative

prostate cancer. Cancer Res 2006;66:10658–63.

[4] Perner S, Demichelis F, Beroukhim R, Schmidt FH, Mosquera JM,

Setlur S. TMPRSS2:ERG fusion-associated deletions provide insight

into the heterogeneity of prostate cancer. Cancer Res 2006;66:

8337–41.

[5] Iljin K, Wolf M, Edgren H, et al. TMPRSS2 fusions with oncogenic ETS

factors in prostate cancer involve unbalanced genomic rearrange-

ments and are associated with HDAC1 and epigenetic reprogram-

ming. Cancer Res 2006;66:10242–6.

[6] Wang J, Cai Y, Ren C, Ittmann M. Expression of variant TMPRSS2/

ERG fusion messenger RNAs is associated with aggressive prostate

cancer. Cancer Res 2006;66:8347–51.

[7] Yoshimoto M, Joshua AM, Chilton-Macneill S, Bayani J, Selvarajah S,

Evans AJ. Three-color FISH analysis of TMPRSS2/ERG fusions in

prostate cancer indicates that genomic microdeletion of chromo-

some 21 is associated with rearrangement. Neoplasia 2006;8:

465–9.

[8] Clark J, Merson S, Jhavar S, Flohr P, Edwards S, Foster CS. Diversity of

TMPRSS2-ERG fusion transcripts in the human prostate. Oncogene

2007;26:2667–73.

[9] Hegeson BE, Tomlins SA, Shah N, et al. Characterization of

TMPRSS2:ETV5 and SLC45A3:ETV5 gene fusion in prostate cancer.

Cancer Res 2008;68:73–80.

[10] Kumar-Sinha C, Tomlins SA, Chinnaiyan AM. Recurrent gene fusions

in prostate cancer. Nat Rev Cancer 2008;8:497–511.

[11] Zong Y, Xin L, Goldstein AS, Lawson DA, Teitell MA, Witte ON. ETS

family transcription factors collaborate with alternative signaling

pathways to induce carcinoma from adult murine prostate cells.

Proc Natl Acad Sci U S A 2009;106:12465–70.

[12] Yu J, Yu J, Mani RS, et al. An integrated network of androgen

receptor, polycomb, and TMPRSS2-ERG gene fusions in prostate

cancer progression. Cancer Cell 2010;17:443–54.

[13] Kunderfranco P, Mello-Grand M, Cangemi R, et al. ETS transcription

factors control transcription of EZH2 and epigenetic silencing of the

tumor suppressor gene Nkx3.1 in prostate cancer. PLoS One 2010;5:

e10547.

[14] Rostad K, Mannelqvist M, Halvorsen OJ, et al. ERG upregulation and

related ETS transcription factors in prostate cancer. Int J Oncol 2007;

30:19–32.

[15] Attard G, Clark J, Ambroisine L, et al. Duplication of the fusion of

TMPRSS2 to ERG sequences identifies fatal human prostate cancer.

Oncogene 2008;27:253–63.

E U R O P E A N U R O L O G Y 5 9 ( 2 0 1 1 ) 4 0 7 – 4 1 4414

[16] Demichelis F, Fall K, Perner S, et al. TMPRSS2:ERG gene fusion

associated with lethal prostate cancer in a watchful waiting cohort.

Oncogene 2007;26:4596–9.

[17] Nam RK, Sugar L, Wang Z, et al. Expression of TMPRSS2 ERG gene

fusion in prostate cancer cells is an important prognostic factor for

cancer progression. Cancer Biol Ther 2007;6:40–5.

[18] Laxman B, Tomlins SA, Mehra R, et al. Noninvasive detection of

TMPRSS2:ERG fusion transcripts in the urine of men with prostate

cancer. Neoplasia 2006;8:885–8.

[19] 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;13:5103–8.

[20] 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;68:645–9.

[21] Scheble VJ, Braun M, Beroukhim R, et al. ERG rearrangement is

specific to prostate cancer and does not occur in any other common

tumor. Mod Pathol 2010;23:1061–7.

[22] Downes MR, Byrne JC, Pennington SR, Dunn MJ, Fitzpatrick JM,

Watson RW. Urinary markers for prostate cancer. BJU Int 2007;99:

263–8.

[23] Djavan B, Remzi M, Schulman CC, Marberger M, Zlotta AR. Repeat

prostate biopsy: who, how and when? A review. Eur Urol 2002;

42:93–103.

[24] Thompson IM, Pauler DK, Goodman PJ, et al. Prevalence of prostate

cancer among men with a prostate-specific antigen level �4.0 ng

per milliliter. N Engl J Med 2004;350:2239–46.

[25] Babaian RJ, Johnston DA, Naccarato W, Ayala A, Bhadkamkar VA,

Fritsche HA. The incidence of prostate cancer in a screening

population with a serum prostate specific antigen between

2.5 and 4.0 ng/ml: relation to biopsy strategy. J Urol 2001;165:

757–60.

[26] Schroder FH, van der Cruijsen-Koeter I, de Koning HJ, Vis AN,

Hoedemaeker RF, Kranse R. Prostate cancer detection at low pros-

tate specific antigen. J Urol 2000;163:806–12.

[27] Kumar-Sinha C, Tomlins SA, Chinnaiyan AM. Evidence of recurrent

gene fusions in common epithelial tumors. Trends Mol Med 2006;

12:529–36.

[28] Kumar-Sinha C, Tomlins SA, Chinnaiyan AM. Recurrent gene

fusions in prostate cancer. Nat Rev Cancer 2008;8:497–511.

[29] Rice KR, Chen Y, Ali A, et al. Evaluation of the ETS-related gene

mRNA in urine for the detection of prostate cancer. Clin Cancer Res

2010;16:1572–6.