Prostate MRI evolution in clinical practice: Audit of tumour detection and staging versus prostatectomy with staged...

We conducted a retrospective audit to compare dominant nodule detection and local staging before and after the introduction of functional sequences and PI-RADS v2 reporting to MRI prostate scans in routine private practice.

A retrospective audit was performed of 245 patients in four separate groups undergoing robotic prostatectomy for prostate cancer by a single urologist between 2009 and 2017. The initial 100 consecutive patients had T2 imaging only. The next 43 patients had T2 and DWI. 52 subsequent patients had T2, DWI and DCE sequences (mpMRI). A final 50 consecutive patients had mpMRI using PI-RADS v2 reporting. Preoperative MRI reports were compared with prostatectomy histopathology to determine the sensitivity of MRI in detecting dominant tumour nodule and T3 extension.

The addition of DWI and DCE sequences improved sensitivity for detection of dominant tumour nodule, with a significant further increase using PI-RADS v2 reporting (38% for T2 vs. 62% for T2/DWI vs. 67% for mpMRI vs 91% for PI-RADS v2). The accuracy of detecting T3 disease was initially very low. The use of additional imaging techniques did not significantly influence this, but the use of a three category likelihood of extraprostatic extension in the PI-RADS v2 group had a significant increase in detection of T3 disease (sensitivity 27% vs. 23% vs. 38% vs 63%).

This audit tracks the significant improvements in MRI detection of prostate cancer dominant tumour nodule and T3 extension in patients undergoing prostatectomy with changing techniques and reporting standards in routine clinical practice.