Oncologic outcomes of organ-confined Gleason grade group 4-5 prostate cancer after radical prostatectomy

BACKGROUND: Organ-confined prostate cancer (CaP) at radical prostatectomy (RP) is associated with good long-term outcomes. However, information for aggressive Gleason organ-confined CaP is scant. To investigate the impact of Gleason grade group (GG) 4-5 on long-term oncologic outcomes after RP.

METHODS: Within a high-volume center database we identified patients who harbored organ-confined CaP (pT2) at RP between 1992 and 2017. Only patients with negative surgical margins, without lymph node invasion and without neo- and/or adjuvant androgen deprivation therapy and/or adjuvant radiotherapy were included. Patients with GG1 were excluded. Kaplan-Meier analyses and Cox regression models tested the effect of GG4 and GG5 on biochemical recurrence-free (BFS), metastasis-free (MFS), overall survival (OS) and cancer-specific mortality (CSM) free survival.

RESULTS AND LIMITATIONS: Of 10,855 identified pT2 patients, 0.1% (n=81) and 0.1% (n=114) harbored GG4 and GG5, respectively. At 10-years after RP, BFS, MFS, OS and CSM-free rates were 80.3 vs. 68.6 vs. 55.4% (P<0.001), 96.7 vs. 89.9. vs. 83.4% (P<0.001), 93.2 vs. 78.3 vs. 72.6% (P<0.001) and 99.3 vs. 98.0 vs. 82.7% (P<0.001) for GG2 and GG3 vs. GG4 vs. GG5, respectively. In multivariable Cox regression models, GG5 represented an independent predictor for biochemical recurrence (Hazard ratio [HR] 3.00, P<0.001), metastasis (HR 5.01, P<0.001), death (HR 2.72, P<0.01) and cancer-specific death (HR 30.1, P<0.001). Conversely, GG4 represented an independent predictor for death (HR 2.10, P=0.04) and cancer-specific death (HR 6.09, P=0.01) but not for biochemical recurrence and metastasis.

CONCLUSION: GG4/5 in organ-confined CaP is rare. But its associated with worse oncologic outcomes after RP, namely biochemical recurrence, metastasis, death and cancer-specific death. Patients with organ-confined GG4/5 and negative margins should be closely followed and may be candidates for risk stratification by genomic markers.