Magnetic resonance imaging-directed transrectal ultrasonography-guided biopsies in patients at risk of prostate cancer


Jonathan A. Coleman, National Cancer Institute, 10 Center Drive, CRC Rm 2–5940, MSC 1210, Bethesda, MD 20892, USA.



To evaluate whether using endorectal-coil magnetic resonance imaging (erMRI) before transrectal ultrasonography (TRUS)-guided biopsies of the prostate increases the yield of cancer in a high-risk population, and in a subset analysis to compare the yield with high-field (3 T) vs lower field (1.5 T) MRI.


Between March 2003 and November 2005, 26 consecutive patients had erMRI before their TRUS biopsy of the prostate (median age 62 years, range 32–76). The median prostate-specific antigen (PSA) level was 8.40  (2.1–85.9) ng/mL. All patients had at least one previous negative prostate biopsy (median 3, range 1–12). Twenty-three patients had at least one risk factor for prostate cancer (family history, high PSA velocity, low percentage of free PSA, prostatic intraepithelial neoplasia or atypical small acinar proliferation on previous biopsy). MRI studies consisted of T2-weighted and dynamic contrast-enhanced (DCE) imaging studies.


There was a close correlation between T2-weighted and DCE images (85% agreement, P < 0.001). Neither T2-weighted nor DCE imaging correlated with a higher yield for cancer on final biopsy (T2, positive predictive value, PPV, 20%, negative PV, NPV, 14%, P = 0.21; DCE, PPV 21%, NPV 15%, P = 0.26). Combining the two methods did not improve the cancer yield. There was no significant difference in the probability of cancer based on 1.5 T or 3 T imaging (17% vs 16%, P = 0.88).


Although erMRI before TRUS-guided biopsies tended to give higher cancer yields the difference was not statistically significant. Reasons for this might include suboptimal localisation of the MRI findings and the biopsy location. Better methods for fusing MRI and TRUS images are presently being developed at our institution to allow more accurate targeting.