Preoperative urinary exosome detection of TMPRSS2:ERG mRNA closely correlates with tissue expression at radical prostatectomy

Back to Publications and Abstracts

Annual Meetings of the American Urological Association — Atlanta GA (May 2012)
Authors: Piruz Motamedinia, Anna Scott, Kendall Bate, Guillermo Salazar, Michael Lipsky Neda Sadeghi, Wayne Comper, James McKiernan, Leileata Russo

Introduction and Objectives

Exosomes (including microvesicles) are lipid bilayer vesicles that are released into biofluids such as urine and carry high integrity RNA from the parent cell. We have recently shown that urinary exosomes can be isolated and found to contain prostate specific mRNA transcripts. Here we examine the concordance of our detection of TMPRSS2:ERG (T:E) in urinary exosomes versus prostatectomy tissue and determine the rate of T:E detection in various prostate cancer related groups.


Following Columbia University IRB approval, random urine samples (without prostate massage) were collected from 163 men who were stratified into 4 groups: TRUS prostate biopsy negative (Bx Neg, n=39), TRUS biopsy positive (Bx Pos, n=47), post-radical prostatectomy (RP, n=37) and controls (males <35 yrs, n=40). Additionally, whole-mount paraffin embedded prostate sections were obtained from 11 patients who underwent RP and had pre-RP urine specimens available (7 T:E positive, 4 T:E negative in pre-RP urine). Urine samples were stored at 4°C and 0.8 µm filtration was used to remove whole cells and debris. Urinary exosomal RNA and tissue RNA were isolated using an in-house technique (Exosome Diagnostics, NY) and analyzed using RP-qPCR.


Mean serum PSA levels were similar in the Bx Neg and Bx Pos groups. All groups except the control (males <35 yrs) were age matched. T:E fusion events were found to occur in 68% of Bx Pos and 44% of Bx Neg patients. Only 5% of controls expressed T:E and no patient in the post RP group had positive T:E detection. Furthermore, T:E expression was lost in the 4 Bx Pos patients that underwent prostatectomy and were retested following surgery. Tissue analysis revealed positive T:E in 8 of 11 prostates. T:E expression concordance between urine and tissue occurred in 10 of 11 (91%) patients with a sensitivity and specificity of 89% and 100%, respectively.


The detection rate of the T:E fusion in the Bx Pos group using urinary exosomal RNA was consistent with previous reports using tissue analyses. Our study exhibited near perfect concordance between urinary exosomal mRNA with matched prostate tissue analysis. The detection of T:E expression in the urine of patients with a negative TRUS biopsy may be due to false negative biopsy results or assay detection of T:E expression in premalignant lesions. The unique stability and yield of urinary exosomal RNA will broaden the role of exosomes in future diagnostic testing and simplify sample handing without the variability and patient discomfort inherent to prostate massage.