Back to Program


Accuracy of 1.5-Tesla using 32 channel High Definition Cardiac coil Magnetic Resonance Imaging for Local Staging of High Risk Prostate Cancer in Comparison to the Partin Tables.
Tony Luongo, MD, Hoon Ji, M.D., Paul Mathew, M.D., Gennaro Carpinito, M.D., Kent Yucel, M.D., Monika Pilichowska, M.D..
Tufts Medical Center, Boston, MA, USA.

Background: Predicting the final pathologic stage is essential for proper patient selection to determine who would be a suitable candidate and experience maximum benefit from radical prostatectomy; especially for men with high risk localized prostate cancer. Several predictive tables and nomograms have been created to help predict the pathologic stage post prostatectomy. The Partin tables represent one of the most widely used statistical predictive tools. We compare the accuracy in predicting the pathologic stage in these high risk patients who undergo radical prostatectomy between the Partin tables and 1.5-Tesla using 32 channel high definition cardiac coil, including dynamic high-resolution 3-D magnetic resonance imaging (MRI) of the prostate preformed prior to surgery. This is the first time this MRI technique has been studied on the accuracy for local staging of prostate cancer.
Materials: Men with high risk localized prostate cancer who chose to be managed with radical prostatectomy underwent 1.5-Tesla with 32 channel high definition cardiac coil prostate MRI to better assess the local stage of the disease. High risk localized prostate cancer is defined as clinical stage > T2c and/or PSA higher than 20 ng/mL, and/or biopsy Gleason grade of > 8, with a normal bone scan and abdominal and pelvis CT scan. The staging results of the 1.5-Tesla MRI technique were compared in terms of staging accuracy with whole-mount-section histopathologic analyses as the standard of reference. Probabilities for pathologic stages were estimated according to Partin tables. The prostate MRI technique is compared to the probability determined by the Partin tables using Spearman rank correlation.
Results: Since January 2009, 10 men were identified with high risk localized prostate cancer that underwent robotic assisted laparoscopic radical prostatectomy. Histopathological evaluation of these prostate specimens revealed organ confined disease (OC) in 6 (60%) and extracapsular extension (ECE) and/or seminal vesicle invasion (SVI) in 4 (40%). 1.5-Tesla MRI staging technique was accurate in all patients with organ confined disease and in three out of the four men with extracapsular extension and/or seminal vesicle invasion. Accuracy of local staging was 90% (9 of 10). Sensitivity was 80% (95% confidence interval [CI] 0.22 - 0.95) and specificity 100% (95% CI 0.83-1) for the detection of ECE and/or SVI. Findings of the MRI and the Partin tables showed a Spearman rho of 0.760 vs. 0.456 for OC and 0.744 vs. 0.396 for ECE and/or SVI respectively.
Conclusions: This pilot study revealed that 1.5T using 32 channel high definition cardiac coil prostate MRI technique is very accurate for the staging of high risk clinically localized prostate cancer and it is more accurate in predicting the final pathologic stage than the Partin tables.


Back to Program