Comparison of magnetic resonance imaging to ultrasound for prostate sizing

To compare pelvic ultrasound (PUS) and transrectal ultrasound (TRUS) to magnetic resonance imaging (MRI) in the estimation of prostate size.

After IRB approval, we performed a single-center, retrospective study of 91 patients who had prostate sizing between August, 2013 and June, 2017. Correlation, reliability, and agreement between PUS, TRUS, and MRI were calculated through the Pearson coefficient, intraclass correlation coefficient, and Bland-Altman analysis, respectively. Data was stratified by prostate size, body mass index, and time between imaging acquisition.

A total of 91 patients underwent all three imaging methods. Median age was 64, median body mass index (BMI) was 27 kg/m2, and median PSA value prior to PUS was 7.1 ng/mL. Pearson coefficient for MRI versus TRUS and MRI versus PUS was 0.90 and 0.87, respectively. Intraclass correlation coefficient was 0.90 (0.87-0.93) comparing all three modalities. BA analysis for MRI versus TRUS and MRI versus PUS showed that for prostates ≤ 50 cc, greater than 79% of the data fell within limits of agreement. Percentages decreased with increasing prostate size to 46% and 41% for prostates > 50 cc and ≤ 80 cc and to 28% and 25% for prostates > 80 cc for MRI versus TRUS and MRI versus PUS, respectively.

MRI may be considered clinically interchangeable with TRUS and PUS for prostate sizing at prostate volumes ≤ 50 cc. For larger prostates and when minor changes in prostate size would drastically alter surgical management, cross-sectional imaging should be considered.