The Specific Definition of High Risk Prostate Cancer Has Minimal Impact on Biochemical Relapse-Free Survival

Publisher Website
Google Scholar
Abstract
Accurate categorization of high risk prostate cancer cases remains elusive. Various schemes based on clincopathological criteria have been proposed to stratify cases by presumed recurrence risk. We determined whether survival outcomes are dependent on the specific definition. The study population included men who underwent radical prostatectomy from 1987 to 1995 (708) and 1996 to 2007 (3,351). Patients who received adjuvant therapy or had no postoperative prostate specific antigen were excluded from analysis. High risk patients were identified based on 6 commonly used definitions. Biochemical failure was defined as a prostate specific antigen of 0.4 ng/ml or greater and increasing or initiation of salvage therapy. Estimates of biochemical relapse-free survival were generated with the Kaplan-Meier method. Hazard ratios for disease recurrence were estimated using Cox proportional hazards analysis. High risk patients determined by the 6 definitions demonstrated a 2.7 to 5.3-fold increased hazard of biochemical relapse, and 5 and 10-year biochemical relapse-free survival rates were 36% to 58% and 25% to 43%, respectively. When stratified by date of treatment high risk patients from 1987 to 1995 generally had worse biochemical relapse-free survival compared to those treated after 1996. Within each era the variation in biochemical relapse-free survival among various high risk definitions was not substantial. Biochemical relapse-free survival after radical prostatectomy does not vary substantially based on the specific definition of high risk prostate cancer. There is a trend toward improved biochemical relapse-free survival in patients treated more recently, perhaps reflecting stage migration or changes in surgical technique. The data suggest that high risk prostate cancer may represent a relatively homogeneous population.

This publication has 14 references indexed in Scilit: