From the Desk of the Editor: Testing in Localized Prostate Cancer

In an era when all men got radiation or surgery, there was no need for risk stratification. However, as newer options became introduced, risk stratification became essential. Towards that effort, PSA, tumor grade, and stage have been the backbone of prostate cancer risk stratification for over 20 years. However, in an era of active surveillance for low-risk disease and multi-modal therapy for high-risk disease, it is clear that they are often not good enough. 

For example, is it safe to put a man with very low volume grade group 2 disease, PSA of 5, and T1C disease on active surveillance (i.e. potentially indolent disease)? If the same patients choose treatment with radiation, does he need hormonal therapy (i.e. potentially lethal disease). In other words, the same patient could get active surveillance, local therapy, or multimodal therapy.  How do we decide who gets which treatments?

Into this morass of risk stratification, steps molecular markers. There are now many markers available – either using tissue, blood, or urine. Given that a comprehensive review of all markers for all applications would be beyond a single review, Ross et al, focused on tissue-based molecular markers. They succinctly reviewed the various markers available including more traditional often non-commercial tests often done in-house such as Ki67 and PTEN to the commercially available send-out tests include ProMark, Prolaris, Ocotype Dx Prostate, and Decipher. As the different tests were designed using different populations and for different end-points it is tough to do a direct comparison. However, the point of how the tests were developed is that the various tests were optimized for different outcomes. This review nicely summarizes the extant literature and puts the tests in context and provides some guidelines for when to use which test and why.

In general, the currently available molecular markers all have prognostic impact in that they can separate high-risk disease from low-risk disease.  The next step, which is just around the corner, is predictive markers. These markers will not just say who is high vs. low-risk, but dictate which treatment should be given. Androgen receptor splices variants (i.e. AR-V7) that predict non-response to further androgen therapy much be such a marker. Similarly, DNA damage defects may predict response to PARP inhibitors. Future studies predicting response to surgery vs. radiation for example, may finally help us end the classic argument and identify the best candidates for either surgery or radiation. We are at a golden time in markers and the exciting part is that the best is yet to come.


Written by: Stephen Freedland, MD, Editor-in-Chief, Prostate Cancer and Prostatic Diseases, Cedars-Sinai Medical Center, Los Angeles, CA