Quantification of extraprostatic extension in prostate cancer: different parameters correlated to biochemical recurrence after radical prostatectomy

Abstract

Different methods to substage extraprostatic extension (EPE) were correlated with biochemical recurrence (BCR) after radical prostatectomy (RP). A total of 157 consecutive RP specimens with EPE were completely embedded. Twenty-three patients with adjuvant therapy or detectable postoperative PSA levels were excluded, leaving 134 patients for BCR analysis. Data were analysed using Kaplan-Meier survival and Cox regression analyses. In univariate analysis, maximal radial distance (RD) was associated with BCR as continuous (P = 0.006) and dichotomous (P = 0.002) parameters. In multivariate analysis, independent predictors of BCR were preoperative prostate-specific antigen (PSA) (P = 0.006), Gleason score (P = 0.001), positive surgical margins (P = 0.005), maximal RD dichotomized at 0.6 mm [= one high-power field (HPF)]; hazard ratio (HR) 3.4; 95% confidence interval (CI) 1.48-7.85; P = 0.004), total RD (P = 0.009) and EPE quantification according to Epstein (P = 0.002) and to Wheeler (P = 0.004). The 5-year risk of BCR was 20% (95% CI 0.65-0.94) in patients with RD ≤ 0.6 mm and 47% (95% CI: 0.41-0.65) with RD > 0.6 mm. The restriction of focal EPE in no more than two slides (Epstein and Wheeler) gave no better results. Maximal RD dichotomized at one HPF is an objective method to subdivide EPE and a strong, independent predictor for BCR after RP. Its use is recommended for substaging pT3a in future TNM classifications.