PSA-tested participants (N=281) of the European Randomized Study of Screening for Prostate Cancer (ERSPC) who underwent radical prostatectomy were analyzed. Besides clinical features, and serum-PSA, histopathologic features as determined in the diagnostic biopsy and matching radical prostatectomy specimen were related to patient outcome.

At a median follow-up of 7 yr, 39 (13.9%), 24 (8.5%), and 12 (4.3%) patients had PSA ≥0.1ng/ml, PSA ≥1.0ng/ml, and clinical relapse after radical prostatectomy, respectively. Using Cox proportional hazards, PSA level (p=0.002), length of tumour (p=0.040), and length of high-grade cancer (p=0.006) in the biopsy, but not Gleason score, were independent prognostic factors for biochemical relapse (PSA ≥0.1ng/ml) when assessed as continuous variables. In radical prostatectomies, the proportion of high-grade cancer (p<0.001) was most predictive of relapse (PSA ≥0.1ng/ml). For PSA ≥1.0ng/ml and clinical relapse, the amount of high-grade cancer, both in the biopsy specimen (p=0.016 and p=0.004, respectively) and radical prostatectomy specimen (p=0.002 and p=0.005, respectively), but not Gleason score, was an independent predictor.

In biopsy and radical prostatectomy specimens of surgically treated prostate cancer, the amount of high-grade cancer is superior to the Gleason grading system in predicting patient outcome. We propose that, in addition to the Gleason score, the amount of Gleason growth patterns 4/5 in the biopsy (whether absolute length or proportion) should be mentioned in the pathology report.

André N. Vis, Stijn Roemeling, Ries Kranse, Fritz H. Schröder, Theo H. van der Kwast

Department of Urology, Erasmus Medical Center, Rotterdam, The Netherlands

Comprehensive Cancer Center Rotterdam, Rotterdam, The Netherlands

Department of Pathology & Laboratory Medicine, Princess Margaret Hospital, Toronto, Ontario, Canada

Accepted 28 July 2006 published online 16 August 2006.