Prognostic value of ERG oncoprotein in prostate cancer recurrence and cause-specific mortality

BACKGROUND ETS-related gene (ERG) protein is present in 40-70% of prostate cancer and is correlated with TMPRSS2-ERG gene rearrangements. This study evaluated ERG expression at radical prostatectomy to determine whether it was predictive of earlier relapse or prostate cancer-specific mortality (PCSM). METHODS One hundred patients who underwent radical prostatectomy at Virginia Mason in Seattle between 1991 and 1997 were identified. Recurrence was confirmed by tissue diagnosis or radiographic signs. PCSM was confirmed by death certificates. Thirty-three patients with metastases or PCSM were matched to patients without recurrence at a 1:2 ratio. Paraffin embedded tissue was stained with two anti-ERG monoclonal antibodies, EPR3864 and 9FY. Nuclear expression intensity was evaluated as present/absent, on a 4-point relative intensity scale, and as a composite score (0-300). RESULTS Mean follow-up was 10.26 years. The two antibodies were highly correlated (P < 0.0001). Patients with higher ERG expression intensity and composite scores were significantly more likely to develop biochemical relapse, metastases, and PCSM. Kaplan-Meier survival curve analysis for the composite score of ERG expression revealed a significant association between higher ERG expression (EPR3864) and shorter PCa-specific survival (P = 0.047). CONCLUSIONS While the presence of ERG expression at the time of surgery was not predictive of earlier relapse or PCSM, the relative intensity and composite score for ERG expression was prognostic for the development of biochemical relapse, metastases, and PCSM. Quantitative ERG scoring may be useful to identify patients who would benefit from adjuvant treatment or closer follow-up, allowing more accurate individual patient treatment plans.