Presentation Authors: Christian Pavlovich*, Baltimore, MD, Matthew Eric Hyndman, Calgary, Canada, Gregg Eure, Virginia Beach, VA, Sangeet Ghai, Toronto, Canada, Vincent Fradet, Quebec, Canada
Introduction: To compare first generation high-frequency 29 MHz transrectal micro-ultrasound (micro-US) with conventional low-frequency 7-12 MHz transrectal ultrasound (conv-US) for the detection of clinically-significant prostate cancer (csPCA).
Methods: 1,676 men indicated for prostate biopsy and without known prostate cancer were randomized 1:1 to micro-US or conv-US guided biopsy at 5 sites in North America from 2013-16. Exactly 12-cores were taken transrectally from each subject, with each core taken either systematically or from a target near the systematic position. The trial was paused after 1,113 subjects to train investigators on the new PRI-MUSTM (Prostate Risk Identification using Micro-Ultrasound) protocol for micro-US targeting, developed using data from the first portion of the trial. csPCA was defined as any Gleason Grade Group > 1 and/or any core with > 50% cancer.
Results: The per-patient detection rates of csPCA in the intent-to-treat (ITT) analysis did not differ between transrectal micro-US and conv-US arms (34.6% (290/837) vs. 36.6% (307/839), respectively, p=0.21). However, the per-protocol analysis (PP) detection rate of csPCA was significantly greater with micro-US (43.7% (125/286) vs. 36.6% (301/823), p=0.02). After PRI-MUS training, sensitivity improved to 63.4% from 24.7% for micro-US (p < 0.01), but at the cost of lower specificity (post-training for micro-US = 61.7%, conv-US = 89.5%), suggesting a learning curve to micro-US interpretation across all investigators. Unexpected prostate under-sampling was noted in the micro-US arm, where 65.8% of men had inadequate sampling of the prostatic apex, which explains the disparity in number of subjects between the ITT and PP groups.
Conclusions: First generation transrectal micro-US with PRI-MUS achieved greater sensitivity to detect significant prostate cancers than conventional TRUS. Despite limitations of the first-generation instrument, when clinicians were instructed in micro-US image interpretation and used proper systematic biopsy technique with the experimental probe, greater cancer detection rates were achieved with the same number of biopsy samples. This work also represents the first prospective validation of a protocol for real-time lesion identification and biopsy targeting using micro-US (PRI-MUS). Additional studies are underway to refine use and application of this novel imaging technology.
Source of Funding: Exact Imaging (Trial sponsor)