SUO 2018: Pooled Results From Two Prospective Validation Studies of the EPI Test Demonstrates Consistent Performance to Predict High-grade Prostate Cancer at Initial Biopsy

Phoenix, Arizona ( As our knowledge of prostate cancer (PCa) continues to grow, we have increasingly come to accept the fact that there is a subset of PCa that has little long-term clinical implications and can be watched – clinically insignificant prostate cancer, often Gleason 6 PCa (Gleason Grade Group 1).

However, some Gleason Grade Group 2 cancers may also fit the bill. The rest of the PCa, as far as our current knowledge enables us to know, is clinically significant (csPCa) warranting treatment. Currently, our ability to discriminate is limited to prostate biopsy and histopathology, though genomic markers are helping to discriminate csPCa from insignificant disease. As of yet, though, no biomarkers have been approved to determine eligibility for active surveillance (AS).

One such biomarker that has been evaluated was the ExoDx Prostate(IntelliScore) (EPI) urine exosome assay.

Figure 1

Two prior studies have assessed this test in this setting – the original study (McKiernan, et. al., JAMA Oncol 2016, 2(7), 882-889) and the second (McKiernan,, Eur Urol., 2018, in press) validation cohort (N=519 and 503, respectively) representing 1022 subjects. In this abstract, the authors combine these two independent cohorts to assess outcome and cut-point performance EPI urine exosome assay vs. optimized standard of care models (SOCm) (i.e. prostate specific antigen [PSA], age, race, and family history) and the Prostate Cancer Prevention Trial-Risk Calculator 2.0 (PCPT-RC), two established clinical nomograms to predict outcomes. Specifically, their aim was to determine ability to distinguish between Gleason score (GS)≥7 (Grade group, GG2) from GS6 (GG1) PCa and benign disease on initial biopsy.

For both studies, (all 1022 patients), eligibility criteria included : >50-years, PSA 2-10 ng/mL, scheduled for initial prostate needle biopsy. Performance of all three markers was assessed using area under the receiver operating characteristic curve (AUC), negative predictive value (NPV), Positive predictive value (PPV), Sensitivity, and Specificity.

The pooled cohort include 1022 biopsy naïve patients - mean age 64 years, mean PSA 5.6 ng/mL, 16% African heritage, 71% Caucasian, 51% positive biopsy rate. Of those, 30% had ≥GS7(GG2) while 13% ≥GS4+3 (GG3). Full demographics and endpoint results below:

Table 2
Looking at the three markers, the EPI AUC=0.70 was superior to SOCm AUC=0.62, PSA AUC=0.56 and PCPT-RC AUC=0.62 (all p-values<0.001) for discriminating ≥GS7 (GG2) PCa from benign and GS6 (GG1). Using the previously validated cut-point of 15.6 (or alternative 20) would avoid 23% (or 34%) of all prostate biopsies and 30% (or 43%) of unnecessary biopsies, with an NPV of 90% for both cut-points and miss only 7.5% (or 12%) of ≥GG2, respectively. The AUC curves are below:

Figure 2

Based on this, the authors conclude that the use of EPI, a non-invasive urine test, can better discriminate high-grade (≥GS7, GG 2) from low-grade (GS6, GG 1) PCa and benign disease that current standard of care nomograms. The test improves identification of patients with higher grade disease, providing a tool for shared decision making and should reduce the total number of unnecessary biopsies.

Speaker: Michael Donovan, MD

Co-Author(s): J McKiernan; A Margolis; A Partin; B Carter; G Brown; P Torkler; M Noerholm; J Skog; N Shore; G Andriole; I Thompson and P Carroll

Institution(s): Columbia University Medical Center, NYC, NY, Icahn School of Medicine at Mt. Sinai, Urology Center of Englewood, Englewood, NJ, Johns Hopkins Hospital, Baltimore, MD, Delaware Valley Urology, Vorhees, NJ, Exosome Diagnostics GmbH, Martinsried, Germany, Exosome Diagnostics, Inc., Waltham, MA, Atlantic Urology Clinics, Myrtle Beach, SC, Washington University, St. Louis, MO, UT Health Science Center, San Antonio, TX, University of California at San Francisco, CA

Written by: Thenappan Chandrasekar, MD, Clinical Instructor, Thomas Jefferson University, twitter: @tchandra_uromd, @TjuUrology, at the 19th Annual Meeting of the Society of Urologic Oncology (SUO), November 28-30, 2018 – Phoenix, Arizona