Role of Magnetic Resonance Imaging In Predicting Adverse Pathology Post-Radical Prostatectomy
Multiparametric magnetic resonance imaging (mpMRI) has become a valuable tool in diagnosing prostate cancer. However, it's utility in predicting clinical outcomes has not been well defined. We sought to evaluate the ability of MRI to predict adverse pathology.
Using our institutional radical prostatectomy (RP) database, we identified 206 patients with preoperative mpMRI. Patients who underwent previous radiotherapy and/or androgen deprivation therapy were excluded from analysis. Biopsy slides were reread by a dedicated genitourinary pathologist. Baseline clinical and pathological data included age, race, BMI, cT stage, biopsy Gleason Score and % of maximum core involvement, diffusion weighted imaging (DWI), T2w imaging and PI-RADS score. PI-RADS v2 was used. Adverse pathology was defined as Gleason Group>2, pathological stage ≥T3 and/or pN+ disease. Univariate association was analyzed by ANOVA, Kruskal-Wallis H Test/ Mann-Whitney U Test, and chi-square/Fisher&[prime]s exact tests. Multivariable logistic regression was used to model the log odds of adverse pathology with radiological features.
In univariate analysis, factors associated with adverse pathology included age (p=0.02), PSA (p=0.001), maximum core percentage at biopsy (p=0.015), clinical T stage (p=0.005), Gleason score (p<0.001), PI-RADS (p<0.001), T2 (p=0.003), DWI (p<0.001), mpMRI lesion size (p<0.001) and extracapsular extension (p=0.003). In a logistic regression model adjusting for age, race, and clinical stage, PSA (p=0.015), PI-RADS (p=0.030) and Gleason score (p<0.001) were associated with adverse pathology. Lesions with PI-RADS 5 versus 3 are more likely to have adverse pathology (OR: 8.1, 95% CI: 1.2, 57.5; p=0.04), but this was not true when compared with PI-RADS 4 (OR: 1.9, 95% CI: 0.8, 4.5, p = 0.12).</p>
PI-RADS score, PSA and biopsy Gleason score are independent predictors for adverse pathology features on prostatectomy specimens.