Tumor Volume, Tumor Percentage Involvement, or Prostate Volume: Which Is Predictive of Prostate-specific Antigen Recurrence?
Received 17 April 2009; accepted 25 June 2009. published online 12 October 2009.
Objectives
To compare the effects of tumor volume (TV), tumor percentage involvement (TPI), and prostate volume (PV) on prostate-specific antigen (PSA) recurrence (PSAR) after radical prostatectomy (RP).
Methods
A cohort of 3528 patients receiving RP between 1988 and 2008 was retrieved from the Duke Prostate Center. Patients were stratified by TV (< 3, 3-6, > 6 cm3), TPI (< 10%, 10%-20%, > 20%), and PV (< 35, 35-45, > 45 cm3) and their effects on PSAR evaluated using Kaplan-Meier analysis. Clinicopathologic variables included in univariate analysis were age at surgery, race, year of surgery, PSA, pathologic Gleason score, pathologic tumor stage, margin status, extracapsular extension, and seminal vesicle invasion. The effects of TV, TPI, and PV (as continuous and categorical variables) on PSAR were compared using Cox analysis.
Results
TPI, TV, and PV were predictive of PSAR (P <.05) in Kaplan-Meier analysis. In multivariate analysis as continuous variables, TPI and PV were predictive of PSAR (hazard ratio [HR] = 1.16 and HR = 0.65, P <.05). As categorical variables, TPI > 20% and PV 10-35 cm3 were predictive of PSAR (HR = 1.45 and OR = 1.25, P <.05). TV was not predictive of PSAR in either analysis. Pathologic Gleason score ≥ 7, PSA, positive margins, seminal vesicle invasion, and tumor stage T3/T4 were found to be predictors of PSAR (P <.05).
Conclusions
TV, TPI, and PV were predictive of PSAR in univariate analysis, but in multivariate analysis, only TPI and PV were predictive of PSAR. TPI and PV should be considered when evaluating, assessing, and counseling patients regarding PSAR risk.
Division of Urologic Surgery and Duke Prostate Center, Department of Surgery, Duke University Medical Center, Durham, North Carolina
Reprint requests: Judd W. Moul, M.D., Division of Urologic Surgery and Duke Prostate Center, Duke University Medical Center, PO Box 3707, Durham, NC 27710
This study was supported by NIH Grant TL1 RR 024126.
ABSTRACT