PROSTATE, sa.15, ss.1783-1789, 2015 (SCI-Expanded)
BACKGROUND. We aimed to evaluate the relationship between perineural invasion (PNI) and bone metastasis in prostate cancer (PCa). METHODS. We retrospectively reviewed the data of 633 PCas who had whole-body bone scan (WBBS) between 2008 and 2014. We recorded the age, clinical T-stage, total PSA (tPSA) prior to biopsy, Gleason sum (GS), and PNI in transrectal ultrasound guided biopsy (TRUS-Bx) and digital rectal examination findings. Bone metastases were assessed with WBBS and magnetic resonance image if WBBS was suspicious. We divided the patients into two groups according to NCCN criteria: (Group 1) bone scan not indicated, (Group 2) bone scan indicated. RESULTS. There were 262 patients in Group 1 and 371 in 2. There is not significant relationship between PNI and bone metastasis in Group 1. However, there is very limited number of metastatic patients (n = 12) in this group. There is a strong relationship between PNI and bone metastasis in Group 2 (P = 0.001). Sensitivity, specificity and positive predictive value of PNI for bone metastasis were 72.4%, 81.7%, and 77.7%, respectively. In this group, tPSA, GS, positive DRE, and PNI were significant covariates for prediction of bone metastasis in univariate and multivariate analysis (except age). The most powerful predictor was PNI, and it increased the risk of bone metastasis 11-fold. CONCLUSIONS. PNI in the TRUS-Bx specimens is the most powerful predictive histopathological feature for bone metastasis, by increasing the risk of bone metastasis 11-fold in NCCN bone scan indicated patients (Group 2). (C) 2015 Wiley Periodicals, Inc.