Preoperative Prediction of Neurovascular Bundle Involvement of Localized Prostate Cancer by Combined T2 and Diffusionweighted Imaging of Magnetic Resonance Imaging, Number of Positive Biopsy Cores, and Gleason Score

摘要

Because recovery of erectile function and avoidance of positive surgical margins are important but competingoutcomes with prostate cancer therapy, the decision to preserve or resect a neurovascular bundle (NVB) duringlaparoscopic radical prostatectomy (LRP) should be firmly based on information concerning the presence andlocation of extracapsular extension. In the current retrospective study, the propriety of actual decisions wasassessed using preoperative magnetic resonance imaging (MRI), combining T2-weighted imaging (T2WI) withdiffusion-weighted imaging (DWI), the apparent diffusion coefficient (ADC), numbers of positive biopsy cores,tumor volume and the Gleason score. MRI before prostate biopsy was performed in 35 patients who underwentLRP for clinically localized prostate cancer. A single radiologist retrospectively assessed whether the tumorlocalization, capsular penetration, seminal vesicle invasion, NVB involvement, and MRI findings correlated withthe postoperative histological results. With the postoperative specimens, 83 lesions demonstrated a Gleason scoreof 6 or more. Using T2WI with and without DWI and ADC, 39 and 27 of 54 lesions were correctly identified,respectively, the difference being significant. For cancers in the transitional zone, using a threshold Gleason scoreof 3 or greater, sensitivity was also significantly higher for T2+DWI+ADC than for T2WI alone. Of 35 patients,using all available clinical information (biopsy results including Gleason score, tumor location, percentage ofpositive biopsy cores, and the percentage of tumor-involved core tissue), we found that the preoperative andpostoperative staging were concordant in 25 cases. There is no universal consensus for nerve-sparing LRP;therefore, we performed an additional analysis using simplified clinically defined selection criteria (PSA level>15ng/mL, cT2, less than two positive biopsy scores in the unilateral lobe and less than 30% tumor volume, anda Gleason score of 6). Using this criteria, we selected 12 of 35 patients, and the detection rate of NVB involvementby MRI combined T2WI + DWI + ADC maps was 100% in their 30 lesions, and therefore we consider it safe toperform nerve-sparing LRP using our criteria. Our findings suggest that NVB can be safely preserved in patientswith low-grade tumors using simplified clinically defined selection criteria to determine margin involvement.