An analysis of risk factors for biochemical progression in patients with seminal vesicle invasion: validation of Kattan's nomogram in a pathological subgroup.

Abstract

To assess the ability of a recently published nomogram to predict failure after radical retropubic prostatectomy (RRP) in a cohort of patients with seminal vesicle invasion (SVI) of the resected specimen, widely regarded as a very poor prognostic factor in patients with prostate cancer. Men with SVI after RRP were analysed retrospectively; patients with positive lymph nodes, adjuvant radiotherapy or hormone therapy were excluded. Age, race, baseline prostate specific antigen (PSA) level, clinical stage, Gleason score, margin status, perineural invasion, capsular invasion, laterality and route of invasion were recorded. Biochemical recurrence was defined as one PSA measurement of 0.4 ng/mL. Biochemical disease-free survival probability was plotted using the Kaplan-Meier method. Kattan's nomogram was applied to each patient and a receiver-operator characteristic (ROC) curve produced to assess the test's reliability. In all, 67 patients with SVI were identified (mean age 64.6 years, sd 5.9; median follow-up 30.5 months). Biochemical disease-free survival was significantly better for men with negative margins, unilateral SVI and perineural invasion. No other significant factors were detected. The median (95% confidence limit) area under the ROC curve was 0.739 (0.604-0.847). There were significant prognostic differences for three of the factors assessed. Although components of the Kattan nomogram were not significant in this group its predictive value approached that found originally, suggesting that it can be applied to patients with SVI in helping to direct adjuvant therapy and guide postoperative counselling.