Patterns of failure after surgically radical resection for organ-confined prostate cancer

Purpose: Some patients from our radical prostatectomy (RPx) series with organ-confined (pT2) prostate cancer and negative surgical margins show a PSA relapse. Our study aims at further analyzing this intriguing cohort of patients that otherwise would have been considered to be cured. Patients and Methods: Since the introduction of PSA in the follow-up after RPx 475 pelvic lymph node dissections (PLND) with subsequent RPx were performed in our department from 1988 to 1997. 227 were classified as pT2, 34 (15%) exhibited positive surgical margins (17 apically, 16 laterally, 1 both) and 4 others were excluded due to a short follow-up (<1 month). From the remaining 189 (study cohort), 19 (10%) developed a biochemical progression, defined as a minimum of two consecutive PSA measurements greater than or equal to 0.1 ng/ml. Only in one of them a G3 tumor was present. Median follow-up was 19.7 months. Results: The Kaplan-Meier biochemical progression analysis showed that after 1 year, 2 years and 5 years, 95% (confidence interval [CI] 91-99%), 91% (Cl 86-96%), and 77% (Cl 55-89%) were free of progression, respectively. This means that roughly one fourth of pT2 tumors will become progressive despite negative surgical margins. These 19 patients were subdivided into 4 groups: No. 1: biopsy proven local recurrence (n=2); no. 2: suspected local recurrence defined as slowly rising PSA less than or equal to 2 ng/ml, but negative biopsies (n=12); no. 3: radiologic imaging proven distant metastasis (n=1); no. 4: suspected distant metastasis defined as rapidly rising PSA > 9 ng/ml without direct radiologic evidence (n=4). All patients from groups 3 + 4 had negative bone scans preoperatively and 4/5 had preoperative PSA values < 10 ng/ml. Conclusion: A pathological diagnosis of organ-confined prostata cancer (pT2) and a meticulous analysis of negative surgical margins does not exclude the occurrence of local relapses in 7% (14/189), and there is evidence for suspect haematogenic spread of PC cells in at least 2% (4/189).