Can multiparametric MR imaging help identify prostate cancer patients who are candidates for active surveillance? "Beyond the Abstract," by Peter Choyke, MD and Baris Turkbey, MD

BERKELEY, CA (UroToday.com) - Many men harbor low-volume and low-grade prostate cancers that exhibit very low lethality. Increasingly, such men are advised to undergo active surveillance (AS) rather than active treatment (AT). The decision to undergo AS is typically based on PSA values, PSA density, and percent involvement of low-grade Gleason score on random biopsies. The popular clinical criteria used to determine suitability for AS, D’Amico system, Epstein criteria and CAPRA scoring system provide slightly different approaches to combining these data and therefore provide different guidance with regard to the decision to utilize AS. These clinical criteria produce two kinds of errors: Errors that direct patients to AT when they would have been better served by AS, or errors that direct the patient to AS when AT would have been more appropriate. One could argue that the latter error (which amounts to undertreatment) could be lethal and therefore, should be more strongly avoided, but there are also problems with overtreatment. Naturally, there is a tradeoff between the two types of errors depending on which error one chooses to minimize. However, regardless of the criteria used, and which errors are deemed more acceptable, a disturbingly high percentage of men, at least a third, fall off AS over time and receive AT. In part this is due to rising PSA values, increasing tumor grade on repeat biopsies, and/or fear of missing the “window of opportunity.”

In our series of 133 patients who were validated at prostatectomy, D’Amico and CAPRA correctly classified 93% of AS candidates but falsely directed 33% and 45%, respectively, to AS who should have been AT candidates. In contrast, the Epstein classification correctly classified only 64% of AS candidates correctly but falsely directed only 9% to AS who should have been AT candidates. Thus, it appears one must choose the criteria that yields the best results for the error one is most trying to avoid.

In the same group of patients, however, MRI correctly classified 93% of AS candidates but had only an 8% rate of patients falsely directed to AS. Thus, MRI does not appear to suffer from the same tradeoffs that affect the clinical criteria. It is important to point out that the MRI criteria do not include biopsy data and were solely based on the size of the dominant lesion as visualized on MRI. Guided biopsy is likely only to improve the results of MRI.

One limitation of this study is that all patients underwent radical prostatectomy, and so there is a verification bias towards patients with clinically significant disease. This was unavoidable as the only widely accepted “standard of truth” is histopathology of the entire prostate gland. However, the general population (on whom the decision to perform AS or AT is actually made) is composed of both less significant and more significant disease than the typical surgical population. How would we predict this bias to affect the results? For the clinical criteria, the risk of identifying inconsequential tumor is high in the general population and one would predict a worsening of the performance of these criteria. In contrast, because MRI is insensitive to small tumors and MR guided biopsy specifically increases the rate of detection of higher-grade disease, the addition of MRI-guided biopsy (typically in the form of MR-TRUS fusion biopsy at our institution) should further improve the performance of MRI. Moreover, the data in the supplement suggests combining MR information with PSA information further strengths the performance of MRI.

It must be acknowledged that this data was generated during the tail end of the “PSA era.” Since the United States Preventive Health Care Task Force recommendations on PSA were announced in 2012 , screening has plummeted in the community. With the possible abandonment of routine PSA screening, the nature of detected prostate cancers is likely to migrate back toward higher volume and higher grade. No doubt, the clinical criteria will improve in accuracy as a consequence of the shift to more advanced disease, and the number of men qualifying for AS will decline. However, in the near term, as we live in a transition period where PSA will be obtained in a haphazard manner and not as a routine, the data generated in our paper point strongly to an advantage to MRI for properly assigning patients to AS. Combined with its advantages of monitoring patients on AS, in lieu of repeated biopsies and the ability to re-sample the same lesion accurately over time, MRI becomes a powerful tool for selecting patients for AS and maintaining them on this course. To be responsive to cost containment, the endorectal coil MR may have to evolve to a coil-less study without intravenous contrast, but we believe that MR technology improvements will allow this to happen without loss of the diagnostic value of MR. Indeed, we believe in a strong future for prostate MRI.

Written by:
Peter Choyke, MD and Baris Turkbey, MD as part of Beyond the Abstract on UroToday.com. This initiative offers a method of publishing for the professional urology community. Authors are given an opportunity to expand on the circumstances, limitations etc... of their research by referencing the published abstract.

Molecular Imaging Program, National Cancer Institute, Building 10, Room B3B69F, Bethesda, MD 20892

Prostate cancer: Can multiparametric MR imaging help identify patients who are candidates for active surveillance? - Abstract

More Information about Beyond the Abstract