We retrospectively analysed a series of 1950 consecutive patients treated with radical prostatectomy (RP) for clinically localized prostate cancer between 1996 and 2005 at our institution. The patients without residual tumour on RP specimen were defined as pT0 patients. The group of pT0 patients was compared with a control group of 295 patients operated consecutively during the same period.

Overall, 11 (0.5%) patients were classified as pT0 on pathologic examination of the RP specimen. There was no pT0 tumour in the control group. Among the pT0 patients, five characteristics were particularly frequent: T1c clinical stage (90.9%), prostate-specific antigen (PSA) ≤15ng/ml (90.9%), one positive biopsy core only (81.8%), biopsy Gleason score <7 (100%), and prostate weight ≥60g (100%). All these characteristics were present in 8 of the 11 (72.7%) pT0 patients, while they were present in only 12 of the 295 (4.1%) controls. These parameters, when combined together, had a sensitivity of 72%, a specificity of 96%, and an accuracy of 99% for the prediction of pT0 stage. With a mean follow-up of 30 months after RP, no pT0 patient had clinical or biologic evidence of prostate cancer.

In our experience, the rate of pT0 tumours after RP is 0.5%. The combination of clinical stage, preoperative PSA, number of positive biopsy cores, Gleason score, and prostate weight could help to predict pT0 stage after RP.

Aurélien Descazeaud^a, Marc Zerbib^a, Thierry Flam^a, Annick Vieillefond^b, Bernard Debré^a, Michaël Peyromaure<sup>a

a</sup> Department of Urology, Cochin Hospital, Paris, France
^b Department of Pathology, Cochin Hospital, Paris, France

Accepted 5 June 2006 published online 23 June 2006.