Clinical outcomes and genomic analysis in patients with very high-risk clinically localized prostate cancer treated by radical prostatectomy.

Abstract

5067 Background: Very high risk prostate cancer is associated with poor response to local and systemic treatments, however few cases have been molecularly profiled. We studied patients with primary Gleason pattern 5 disease (PG5; Gleason scores 5+5 or 5+4) after undergoing radical prostatectomy (RP). Methods: Consecutive PG5 cases were retrospectively studied. Clinical-pathologic features, tumor somatic sequencing (UW Oncoplex), germline sequencing and PTEN, TP53 and ERG status by immunohistochemistry were correlated with biochemical relapse (BCR), metastasis-free survival (MFS), time to castration-resistance (TCR) and overall survival (OS). Kaplan–Meier curves were used in time-to-event data. Univariate and multivariable (MVA) Cox proportional-hazards models (95% confidence intervals (CI Clinical-pathological variables were selected for the multivariate analysis using stepwise method and were used to adjust the effect of each mutation. Results: Of 60 patients with available tissue and follow-up, 90% had non-organ confined disease, 58% had seminal vesicle invasion (SVI), 27% had lymph node involvement, and 15% had ductal or intraductal histologic features. Overall, 43% developed metastasis with a TCR of 12 months. On multivariable analysis of clinical-pathologic variables, only ductal/intraductal histology (HR = 4.43; 95% CI: 1.76-11.15; p = 0.002) and SVI (HR = 5.14; 95% CI: 1.83-14.47 p = 0.002) were associated with metastasis. Of patients with somatic sequencing data, 22% (11/49) had homologous recombination (HR) and 12% (6/49) had mismatch repair (MMR) gene alterations. In addition, 33% (16/49) had TP53 mutation and 51% (29/57) had PTEN loss. Only TP53 mutation and PTEN loss were associated with metastasis on univariable analysis and no genomic alterations remained significant in multivariable analyses. Conclusions: Patients with PG5 have poor outcomes and more than one third have mutations in DNA repair genes. The unusually high incidence of actionable HR and MMR alterations provides the rationale for the design of adjuvant clinical trials in this population. [Table: see text]