Perineural invasion and TRUS findings are complementary in predicting prostate cancer biology.

INTRODUCTION: Clinical variables with more accuracy to predict biologically insignificant prostate cancer are needed. We evaluated the combination of transrectal ultrasound-guided biopsy of the prostate (TRUSBx) pathologic and radiologic findings in their ability to predict the biologic potential of each prostate cancer. MATERIALS AND METHODS: A total of 1043 consecutive patients who underwent TRUSBx were reviewed. Using pathologic criteria, patients with prostate cancer (n = 529) and those treated with radical prostatectomy (RP) (n = 147) were grouped as: "insignificant" (Gleason score ≤ 6, prostate-specific antigen (PSA) density ≤ 0.15 ng/ml, tumor in ≤ 50% of any single core, and < 33% positive cores) and "significant" prostate cancer. TRUSBx imaging and pathology results were compared with the RP specimen to identify factors predictive of "insignificant" prostate cancer. RESULTS: TRUSBx pathology results demonstrated perineural invasion in 36.4% of "significant" versus 5.4% of "insignificant" prostate cancers (p < 0.01) and pathologic invasion of periprostatic tissue in 7% of significant versus 0% of insignificant prostate cancers (p < 0.01). TRUS findings concerning for neoplasia were associated with significant tumors (p < 0.01). Multivariable analysis demonstrated perineural invasion in the biopsy specimen (p = 0.03), PSA density (p = 0.02) and maximum tumor volume of any core (p = 0.02) were independently predictive of a significant prostate cancer. CONCLUSIONS: TRUS findings concerning for measurable tumor and perineural invasion in TRUSBx specimens appear to be complementary to Epstein's pathologic criteria and should be considered to aid in the determination whether a prostate cancer is organ-confined and more likely to be biologically insignificant.