Differential prognostic impact of different Gleason patterns in grade group 4 in radical prostatectomy specimens.

INTRODUCTION There are questions regarding whether grade group (GG) 4 prostate cancer (PC) is heterogeneous in terms of prognosis. We assessed prognostic differences in PC patients within GG 4 treated with radical prostatectomy (RP). MATERIAL AND METHODS Biochemical recurrence (BCR)-free, cancer-specific, and overall survival were analyzed in 787 PC patients with GG 4 based on RP pathology (Gleason score (GS) 3 + 5: 189, GS 4 + 4: 500, and GS 5 + 3: 98). Logistic regression analysis was performed to assess factors predictive of high-risk surgical pathological features. Cox regression models were used to evaluate potential prognostic factors of survival. RESULTS Within a median follow-up of 86 months, 378 patients (48.0%) experienced BCR and 96 patients (12.2%) died, 42 of whom (5.3%) died of PC. GS 5 + 3 was significantly associated with worse BCR-free and cancer-specific survival, as well as higher positive surgical margin, lymph node metastasis, extraprostatic extension, and non-organ-confined disease rates, than GS 3 + 5 and higher positive surgical margin, lymph node metastasis, extraprostatic extension, and non-organ-confined disease rates than GS 4 + 4 (P < 0.05). GS 4 + 4 was significantly associated with worse BCR-free survival and higher extraprostatic extension, and non-organ-confined disease rates than GS 3 + 5 (P < 0.05). Inclusion of the different Gleason patterns improved the discrimination of a model for prediction of all survival outcomes compared to standard prognosticators. CONCLUSIONS There is considerable heterogeneity within GG 4 in terms of oncological and surgical pathological outcomes. Primary and secondary Gleason patterns should be considered to stratify high-risk PC patients after RP.