Login to Access Video or Poster Abstract: MP64-03
Sources of Funding: None.

Introduction

Advancements in prostate imaging have improved prostate cancer detection, localization and staging. Prediction of tumor capsular breaches on multiparametric MRI (mpMRI) may assist with surgical planning as well as improve prostate cancer prognostication. The objective of this study was to determine what tumor and MRI characteristics are suggestive of extracapsular extension (ECE) on pathologic prostate specimens.

Methods

High risk prostate cancer patients who underwent mpMRI and radical prostatectomy from January 2011 to April 2015 were retrospectively reviewed. Clinicopathological characteristics (PSA, PSA density, clinical stage, biopsy Gleason score and percent positive core), MRI features and pathology results were analyzed. Univariate and multivariate logistic regression analyses were performed to determine odds ratios predictive of ECE on pathology.

Results

Among 93 patients that underwent radical prostatectomy, 55 had ECE on final pathology. On univariate analysis, PSA (OR=1.1, 26.1 vs 10.9, p=0.03) was predictive of pathologic ECE while biopsy Gleason score (OR=1.2, p=0.62), percent positive cores (OR=2.6, p=0.27) and lowest ADC value on MRI (OR=1, p=0.27) were not. On multivariate analysis, gross ECE on MRI was the only factor associated with pathologic ECE (OR=1.9,p=0.03). PSA (p=0.07) and other MRI features were not suggestive of pathologic ECE. The sensitivity and specificity of MRI for identifying pathologic ECE was 52.6% and 65.8%, respectively; the associated positive and negative predictive values were 84.3% and 28%, respectively.

Conclusions

The presence of ECE on mpMRI was the most reliable predictor of pathological ECE in this high-risk cohort. mpMRI is a valuable adjunct in surgical planning and stage prognostication.

Funding

None.