Use of TMPRSS2-ERG gene rearrangement and quantitative ERG expression to predict clinical recurrence after radical prostatectomy.

Abstract

36 Background: The association of TMPRSS2-ERG fusions and ERG expression in prostate cancer (PC) with adverse clinical outcomes has been controversial, with mixed results in the literature. We conducted a study to test whether tumor-derived gene expression profiles, including the presence of TMPRSS2-ERG fusions and ERG gene expression, are associated with clinical recurrence (cR) after radical prostatectomy (RP). Methods: All patients with clinical stage T1/T2 prostate cancer treated with RP at CC from 1987 to 2004 were identified (n∼f2,600). A cohort sampling design was used to select 127 patients with cR and 374 patients without cR after RP. For each patient a primary Gleason pattern (GP) sample, secondary (or highest) GP sample, and an adjacent nontumor tissue sample were evaluated. Surgical Gleason Score (GS) and clinical data were centrally reviewed. RNA was extracted from 6 manually dissected 10 μ m formalin-fixed paraffin-embedded sections obtained from RP specimens and expression of TMPRSS2-ERGa, TMPRSS2-ERGb, ERG and reference genes were quantified using RT-PCR. Times to cR, PSA recurrence, and PC death were analyzed using Cox PH regression. Results: Blocks from 441 patients were evaluable. Median F/U was 5.8 years. Patients were mostly Caucasian (83%), clinical stage T1 (66%), had baseline PSA <10 ng/mL (82%), and had surgical Gleason score ≤7 (87%). 848 tumor samples and 410 non-tumor samples were assessed. TMPRSS2-ERGa and/or TMPRSS2-ERGb fusions were present in 51.8% of tumor samples and 7.5% of non-tumor samples. There was 89% concordance (95% CI: 86%, 92%) for TMPRSS2-ERG fusion status between the 2 tumor samples for each patient. High ERG expression was strongly associated with the presence of TMPRSS2-ERG fusions (p <0.01). We did not find an association between TMPRSS2-ERG a/b gene rearrangement or ERG expression with cR, PSA recurrence, PC death, or surgical GS (p > 0.2). Conclusions: This study was notable for the large number of cR events, use of a standardized quantitative assay, and rigorous central review of pathology and clinical data. We did not find an association of TMPRSS2-ERG gene rearrangements or ERG expression with aggressiveness of prostate cancer post RP. [Table: see text]