AUA 2018: Using Multi-parametric Magnetic Resonance Imaging and Targeted Biopsy to Rule Out Seminal Vesicle Invasion in Prostate Cancer

San Francisco, CA USA ( Patients that have seminal vesicle involvement at the time of radical prostatectomy (pT3b), represent a subset of patients with the highest risk of biochemical recurrence and development of subsequent metastasis for patients undergoing radical prostatectomy. In addition to Gleason ≥8 disease, pT3b is one of the strongest predictors for poor outcomes, across many published series. However, historically, identifying seminal vesicle invasion prior to radical prostatectomy has been a diagnostic challenge.

Multiparametric MRI (mpMRI) and mpMRI-guided targeted biopsies have been shown to improve detection of high-grade cancer in the prostate, however this technology has not been directly assessed for detecting disease in the seminal vesicles. Samuel Gold, MD and colleagues from the NCI presented results of their study using mpMRI and targeted biopsy to rule out seminal vesicle invasion for men with prostate cancer in an attempt to better risk stratify patients prior to surgery.

The authors used a single institutional setting to identify appropriate high-risk patients with prostate cancer. Cases with pathologic investigation of the seminal vesicles, from targeted biopsies or robotic prostatectomy, were identified. Imaging features including mpMRI evidence of seminal vesicle invasion, extraprostatic extension, or lesions in the prostatic base were examined for associations with pathologic seminal vesicle invasion. Targeted and systematic biopsy of the prostate and seminal vesicles were analyzed to identify other risk factors for seminal vesicle invasion. Uni- and multivariate regression analyses determined relative strengths of association of these features with seminal vesicle invasion before and after biopsy.

There were 583 patients from 2007-2017 that were identified as being suspicious for seminal vesicle invasion. Among these patients, 63 (11%) were confirmed positive for pathologic seminal vesicle invasion. Targeted biopsy of seminal vesicles was performed in 42 patients and in those who went on to robotic prostatectomy, targeted biopsy was 100% accurate (n=7). Sensitivity and specificity of mpMRI evidence for seminal vesicle invasion was 62% and 96%, respectively. When no prostate base lesion was identified on mpMRI, the negative predictive value (NPV) for pathologic seminal vesicle invasion was 99%. When targeted and systematic biopsy of the prostate base were benign, NPV was 99%. Significant features associated with seminal vesicle invasion included:

• mpMRI evidence of seminal vesicle invasion pre-biopsy (OR 17.8, p<0.001) and post-biopsy (OR 5.29, p=0.001)
• mpMRI evidence of extraprostatic extension on pre-biopsy (OR 2.95, p=0.007) and post-biopsy (OR 3.15, p=0.01)
• mpMRI evidence of prostate cancer at the base on pre-biopsy (OR 6.28, p=0.001)
• Biopsy Gleason group 3 (OR 7.16, p=0.04)
• Biopsy Gleason group 4 (OR 11.28, p=0.04)
• Biopsy Gleason group 5 (OR 42.11, p<0.001)

This is a provocative study demonstrating feasibility of using mpMRI to evaluate and target biopsy the seminal vesicle. Important clinical implications include preoperatively knowing that seminal vesicle invasion is present, and thus making sure of a wide excision at the prostate base/seminal vesicles. A limitation of the model is the likely overfitting of the statistical model with almost certain variable collinearity. From their results, the authors concluded that high-grade prostate cancer biopsy from the prostate base and MRI evidence of seminal vesicle invasion or extraprostatic extension are significant features associated with increased risk of seminal vesicle invasion. Furthermore, absence of base lesions or benign base biopsy are very strong negative predictors. In fact, the authors argue that even if the SVs are unable to be examined on MRI, inclusion of these features still yields strong predictive value at pre- and post-biopsy stages.

Presented by: Samuel Gold, Urologic Oncology Branch, National Cancer Institute, Bethesda, MD

Co-Authors: Sherif Mehralivand, Jonathan Bloom, Bethesda, MD, Stephanie Harmon, Frederick, MD, Graham Hale, Kareem Rayn, Vik Sabarwal, Shawn Marhamati, Marcin Czarniecki, Clayton Smith, Vladimir Valera Romero, Maria Merino, Baris Turkbey, Peter Pinto, Bethesda, MD

Written by: Zachary Klaassen, MD, Urologic Oncology Fellow, University of Toronto, Princess Margaret Cancer Centre, Twitter: @zklaassen_md, at the 2018 AUA Annual Meeting - May 18 - 21, 2018 – San Francisco, CA USA