Risk Stratification for Positive Lymph Nodes in Prostate Cancer

Abstract

To evaluate the risk of positive lymph nodes using preoperative clinical parameters. We reviewed our prospectively collected database for all patients who received RRP and PLND between January 1993 and November 2005 as primary therapy for prostate cancer. We excluded patients who had hormonal ablation or radiation therapy prior to surgery and patients with missing PSA, clinical stage, or biopsy Gleason score. We evaluated risk for nodal disease using the following definitions: low risk: PSA <or=10 ng/mL, clinical stage <or=T(2a), and Gleason score <or=6; intermediate risk: PSA 10 to 20 ng/mL, clinical stage >or=T(2b), or Gleason score of 7; and high risk: PSA >or=20 ng/mL, or clinical stage >or=T(2c), or Gleason score >or=8. Logistic regression was used to determine the association between the risk groups and pathologic lymph node involvement, and a receiver operating characteristics (ROC) curve was constructed to evaluate the performance of the stratification scheme in detecting nodal disease. A total of 760 patients with 43 (5.7%) patients with node-positive disease were available for analysis. Risk classification was significantly associated with positive nodes (P<0.001), even after controlling for year of surgery and age. The area under the ROC curve was 0.77 (95% CI: 0.69, 0.83). Omitting PLND in the low-risk group would have spared 368 (49.2%) of the entire cohort with a false-negative rate of 5/369 (1.3%) for the low-risk group, and 5/760 (0.7%) for the entire cohort. Sensitivity was 88.4%, and negative predictive value was 98.7%. Patients can be risk stratified for node-positive disease and potentially excluded from lymphadenectomy with minimal risk.