Feasibility for active surveillance in biopsy Gleason 3 + 4 prostate cancer: an Australian radical prostatectomy cohort.

OBJECTIVE To examine the feasibility of active surveillance for low volume Gleason sum (GS) 3 + 4 disease compared to GS 3 + 3 disease. PATIENTS AND METHODS Retrospective review of 929 patients, with biopsy proven GS 3 + 3 and 3 + 4 PCa, undergoing upfront radical prostatectomy (RP) was performed. Suitability for AS was adapted from protocols by Royal Marsden Hospital, University of Toronto, and PRIAS by allowing Gleason 3 + 4 disease. The outcomes assessed were adverse pathology at RP (upgrading ≥GS 4 + 3 and/or upstaging ≥pT3) and biochemical recurrence (BCR) after RP. RESULTS Adverse pathology at RP was compared between GS 3 + 3 vs 3 + 4 groups. When selecting patients using Royal Marsden (n = 714) or University of Toronto (n = 699) protocols, there was statistically significantly more adverse pathology at RP in GS 3 + 4 group (21% vs 31%, P = 0.0028 and 19% vs 33%, P=<0.001 respectively). Using the more stringent PRIAS protocol (n = 198), there was no statistical significant difference in groups. There was no difference in BCR survival between biopsy GS 3 + 3 and 3 + 4 groups, regardless of which AS protocol assessed. Pre-operative PSA and clinical staging were the predictors for BCR. CONCLUSION Presence of Gleason 3 + 4 at biopsy, when compared to 3 + 3, increases the risk of adverse pathology being present at radical prostatectomy for less stringent selection criteria. When considering AS, a stricter protocol such as PRIAS, limiting PSA density and number of positive cores to ≤2, appears to decrease the risk of adverse pathology. No differences in BCR were seen between biopsy 3 + 3 and 3 + 4 disease, regardless of AS selection criteria.