Recently, there has been an increasing role of MRI to non-invasively detect the characteristics of prostatic lesions. Areas of suspicion can be targeted to improve the detection rate of clinically significant cancers over standard biopsy sampling techniques hence, fewer biopsies are preformed^1-5.

According to PI-RADS v2, T2-w sequence primarily determined the final score in TZ (Transitional zone) lesions with exception of one scenario in which a high DWI can upgrade and intermediate T2 score of three⁶. We explore the role DWI might play in downgrading the PI-RADS score, specifically, if a low or intermediate DWI score can be used to downgrade a higher T2 score.

The study was a retrospective review of TZ lesions identified on MRI and subsequently referred to MR-US fusion biopsy. Cases were excluded if they did not have a targeted biopsy, if the DWI acquisition was different than our standard (acquisition with b values of 50, 800 and 2000), and if the full study, including original target ROIs, could not be retrieved from our archives. A radiologist blinded to the biopsy results and the radiology report was presented with only the DWI/ADC images of the lesions. The radiologist assigned a DWI PI-RADS score and created a 3D ROI of the lesion. At a later time, the radiologist was presented with the T2-weighted turbo spin echo (T2 TSE) images to provide a T2 PI-RADS score. The PI-RADS scores for each sequence were recorded and were compared with the targeted biopsy. The final PIRADS v2 score was calculated. A downgraded PI-RADS score was also calculated by using a DWI score of 3 or less to override a higher T2 score (refer to Fig 1). We also compared targets with Gleason > 6 versus targets with Gleason ≤6 with respect to average ADC value, volume and largest dimension using Mann Whitney U test.

78 TZ lesion targets were identified. 55 of 78 TZ lesions met research criteria. 14% (8) had a summed Gleason score of 7, 86 % (47) had Gleason ≤6 (44 negative, 3 Gleason 6). There was no significant difference, between these 2 groups in average ADC value (p= 0.10), volume (p= 0.37), and maximum dimension (p=0.57). The distribution of the lesion among DWI PI-RADS score of 5, 4, 3, 2 and 1 were 42 %, 38.6%, 8.77%, 7.01%, and 0%. The percent of lesions with Gleason of ≥7 in each of those scores were 20.83% (5/24), 9.09% (2/22), 20% (1/5), and 0% (0/4). A total of 9 targets had an intermediate to low DWI score (defined as ≤3) four of which had a higher T2 score and were downgraded using our suggested downgrading scheme. All four downgraded lesions had negative Gleason.

In clinical practice TZ lesions are overcalled with low yield of targeted biopsy reflecting a high false positive rate. Lesions of low DWI PI-RADS score are less commonly called. If we use intermediate or low DWI score to downgrade the PI-RADS score, we can downgrade the score of few negative targets without downgrading any of the positive targets. Due to the small number of positive targets as well as the small number of lesion with low or intermediate DWI score these data lacked statistical significance.