OBJECTIVETo validate the 2007 Partin tables externally, which are based on the population of United States, using a cohort of Chinese prostate cancer patients.

METHODSAll of the patients enrolled and underwent radical prostatectomy between January 2006 and February 2010 were reviewed. The cases without preoperative hormone therapy and pelvic lymph node involvement according to radiologic tests were used for the external validation of the 2007 Partin tables. A comparative analysis of the clinical and pathological parameters of this Chinese cohort and Partin tables cohort was performed. Values of areas under the receiver operating characteristic (ROC) curve were used to assess predictive accuracy for the Chinese cohort.

RESULTSThe mean age of the whole cohort was 67 years. The serum prostate specific antigen level, Gleason score and clinical stage of this cohort were higher than the Partin tables cohort. The pathological outcomes analysis revealed that the rates of organ confined disease, capsular penetration, seminal vesicle involvement and lymph node involvement were 62.3%, 16.7%, 12.3% and 8.8%, respectively. The area under the ROC curve (AUC) for organ confined disease, capsular penetration, seminal vesicle involvement and lymph node involvement were 0.735, 0.653, 0.601 and 0.845.

CONCLUSIONSThe Partin tables discriminate well for Chinese patients at risk for positive lymph node. The discrimination of organ confined disease is also acceptable and the discrimination of capsular penetration and seminal vesicle involvement is more limited.