PurposeTo evaluate the ability of the new Gleason grade groups (GGGs) to stratify risk in prostate cancer patients with locally adverse pathologic features after radical prostatectomy (RP) thereby allowing more accurate assessment for planning eventual adjuvant therapy.Patients and methodsData on 172 patients with locally adverse pathologic features (including seminal vesicle invasion, extracapsular extension, or positive surgical margins) who had been treated with wait and see policy after RP were retrospectively analyzed for biochemical recurrence (BCR)-free survival. Kaplan–Meier survival analysis and Cox proportional hazard regression models were used to test the association between the GGGs and BCR. Finally, concordance indices of different grading classifications were calculated to evaluate the predictive accuracy for biochemical failure after RP.ResultsThe five-year BCR-free survival rates were 71.2, 66.9, 25.7, 17.4, and 8.3 % for GGG 1–5 assessed on surgical specimens (p < 0.001, log-rank test). In the two-way log-rank test, men with prostatectomy GGG 2 had a lower progression risk relative to GGG 3 (p = 0.001), though similar risk as GGG 1 (p = 0.105). In multivariate Cox regression analysis, specimen GGG ≥3 and early postoperative PSA ≥0.1 ng/ml were independent risk factors for biochemical failure (p < 0.001). In addition, GGGs had higher predictive accuracy compared with the alternate classification system (improvement in concordance index by 0.036–0.141).ConclusionsFor the appropriate patient, depending on age, physical condition, early postoperative PSA, patient desire, etc., could be a candidate for wait and see policy with specimen GGG 2 disease, so to distinguish this from GGG 3 may facilitate discussions at the point of treatment decision making.