Introduction

An updated version of PIRADS (PIRADS v2) was introduced in August 2015 for MRI interpretation. While a similar scoring pattern ranging from 1-5 is used, PIRADS v2 employs a dominant sequence rule, where depending on the location of the target the overall score is the score from that particular sequence (DWI for PZ lesions, and T2 for TZ lesions). However, for lesions with an indeterminate score of 3, a secondary sequence is used (DCE for PZ targets, and DWI for TZ targets) to determine if the score should be upgraded to a 4 or left as a 3. Our objective was to validate whether the rate of detecting Gleason 7 or higher cancer was different between an upgraded PIRADS 4 lesion versus a true PIRADS 4.

Methods

A prospective cohort of 336 men underwent mp-MRI US fusion biopsy between August 2015 to August 2016 for evaluation of prostate cancer or active surveillance. Among these men, there were 462 targets that were biopsied. For this analysis we selected all targets that were scored as PIRADS 3 or 4, and differentiated between an upgraded PIRADS 4 and a &[Prime]true&[Prime] PIRADS 4 target. We report the detection rate of Gleason 7+ cancer between PIRADS 3 and 4 targets and use logistic regression to assess the association between PIRADS score and cancer detection.

Results

Among the 462 targets in our dataset, we had 333 targets that were either PIRADS 3 or 4. Among these targets, 166 were PIRADS 3, 49 were an upgraded PIRADS 4, and 118 were a true PIRADS 4. We had 190 targets that were located in the PZ, and 143 targets located in the TZ. For targets in the PZ, the detection of Gleason 7+ cancer was 8%, 27%, and 33% for PIRADS 3, upgraded PIRADS 4, and true PIRADS 4 targets, respectively (p <0.01). Compared to a PIRADS 3 in the PZ, the odds of Gleason 7+ cancer was 3.98(CI 1.41-11.23) for an upgraded PIRADS 4 and 5.42(CI 2.14-13.73) for a true PIRADS 4 target. We performed a similar logistic regression with upgraded PIRADS 4 as the reference group and saw no difference between a upgraded PIRADS 4 and true PIRADS 4 for the detection of Gleason 7+ cancer. We did not have enough significant cancer detected in the TZ to make any reliable comparisons in this zone.

Conclusions

Upgraded PIRADS 4 targets in the PZ have a higher rate of Gleason 7+ cancer compared to PIRADS 3 targets. We found no difference between an upgraded PIRADS 4 and a true PIRADS 4. Our findings validate the revised scoring system for PIRADS.

Funding

none