Objectives: To determine the accuracy of the preoperative serum concentration of prostate-specific antigen (PSA) plus the Gleason pathology score of biopsy specimens in predicting the presence of disease in the pelvic lymph nodes in patients with prostate cancer.
Methods: The medical records of all patients treated for prostate cancer at eight medical centers from January 1988 to June 1993 were reviewed. There were 932 patients with newly diagnosed prostate cancer for whom all relevant data were available who had undergone pelvic lymphadenectomy with (n = 912) or without (n = 20) radical prostatectomy. The rate of false-negative predictions of metastases based on combined preoperative biopsy Gleason score and serum PSA concentration was analyzed. A multivariate logistic regression analysis was performed to assess the value of preoperative serum PSA and biopsy Gleason scores individually and in combination in predicting pelvic lymph node metastases.
Results: The false-negative rate of metastases was 0% for preoperative PSA concentrations < or = 6 ng/mL and biopsy Gleason scores < or = 5 (n = 142) and 1.0% for PSA concentrations < or = 10 ng/mL and Gleason scores < or = 6 (n = 388). The 95% upper confidence limit for the rate of false negativity at this PSA cut-off level was 2.0%. A combination of preoperative serum PSA levels and biopsy Gleason scores provided the best prediction for the false-negative rates.
Conclusions: For patients with newly diagnosed prostate cancer who have biopsy Gleason scores < or = 6 and preoperative PSA concentrations < or = 10 ng/mL (42% of our series), a staging pelvic lymphadenectomy appears to be unnecessary. The substantial cost associated with both cross-sectional imaging and staging lymphadenectomy may therefore be avoidable in this group of patients.