Predicting pathological stage of localized prostate cancer using volume weighted mean nuclear volume

With the use of prostate specific antigen (PSA) and transrectal ultrasound guided biopsy of the prostate, increasing numbers of clinically localized prostate cancers have recently been detected. Presently, approximately 60% of men newly diagnosed with prostate cancer are believed to have organ confined disease but at the time of surgery less than 50% are ultimately found to have organ confined disease on final pathological analysis. Prediction of pathological stage before surgery using various prognostic factors is needed to determine whether radical prostatectomy is indicated. We have reported that estimates of volume weighted mean nuclear volume can accurately predict the prognosis of clinically localized prostate cancer treated with androgen ablation. In this study we examine whether estimates of mean nuclear volume can predict the pathological stage of cases treated with radical prostatectomy.

A retrospective, prognostic study of 52 patients with clinically localized prostate cancer (21 cases T1c, 27 T2, 4 T3) diagnosed at Kobe City General Hospital between January 1996 and December 1999 and treated with radical prostatectomy was performed. Unbiased estimates of mean nuclear volume measured from transrectal biopsy specimens were compared with PSA at diagnosis, clinical stage, estimated tumor volume and Gleason score with regard to prediction of pathological stage.

Univariate analysis revealed that estimates of mean nuclear volume (p <0.0001), PSA at diagnosis (p = 0.0148) and estimated tumor volume (p = 0.0005) significantly correlated with pathological stage but Gleason score did not (p = 0.2011). In addition, multivariate logistic regression analysis demonstrated that estimates of mean nuclear volume (p = 0. 0073), PSA at diagnosis (p = 0.0277) and estimated tumor volume (p = 0.0197) were significantly independent predictors of pathological stage.

The results of our study suggest that combining PSA and estimated tumor volume with estimates of mean nuclear volume can significantly contribute to the prediction of pathological stage of prostate cancer. We recommend use of these 3 factors to predict pathological stage of prostate cancer before surgery.