Absence of ERG Expression Predicts Early Prostate Cancer Biochemical Recurrence When Combined with DNA Methylation Status of a Development-Associated Gene
Dominique Trudel, Ken Kron, Liyang Liu, John Trachtenberg, Neil Fleshner, Bharati Bapat, Theodorus H Van der Kwast. Princess Margaret Hospital, University Health Network, Toronto, Canada; Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Canada
Background: Recent work supports interplay between ERG expression secondary to TMPRSS2-ERG fusion and epigenetic changes. Through a genome-wide epigenetic screen, we discovered a development-associated gene (DAG) with high promoter methylation levels in prostate cancer. Our objective was to correlate DAG methylation status with ERG expression and biochemical recurrence-free survival (BRFS) in an independent prostate cancer cohort.
Design: Two-hundred-fifty-eight patients treated by radical prostatectomy between 1998 and 2001 were identified. Macrodissection of tumors was performed prior to quantitative DNA methylation analysis (MethyLight) for DAG promoter. Paraffin-embedded blocks were rearranged in tissue microarrays in order to identify ERG expression in tumors using immunohistochemistry as a surrogate for TMPRSS2-ERG fusion. Univariate BRFS was assessed using the Kaplan-Meier curve and log-rank tests. Cox proportional hazards regression analysis was used for multivariate BRFS.
Results: A total of 219 pT2 to pT4 cases were available for analysis. Median follow-up was 4.67 years. ERG expression was found in 51% of the cases. DAG methylation was significantly greater in tumor specimens compared to benign glands and was greater in ERG positive cases (p-values < 0.001). Univariate analysis showed no association between DAG methylation or ERG expression and BRFS. Patients with ERG negative/high DAG methylation tumors however had shorter BRFS. Furthermore, in a multivariate model that included clinicopathological data, absence of ERG expression and high methylation of DAG were both significant predictors of recurrence (p-values 0.03 and 0.04, respectively).
Conclusions: Although TMPRSS2-ERG fusion is identified in approximately 50% of prostate cancer, there is conflicting evidence regarding its prognostic value in the literature. We show that (1) ERG expression is associated with increased methylation of DAG promoter and (2) patients with tumors diverging from this profile, i.e. with high methylation of DAG promoter in the absence of ERG expression have shorter BRFS. These results suggest that TMPRSS2-ERG fusion is associated with methylation levels of specific genes, and also add prognostic value to TMPRSS2-ERG fusion status of prostate cancer.
Category: Special Category - Pan-genomic/Pan-proteomic approaches to Cancer
Tuesday, March 20, 2012 9:30 AM
Poster Session III # 288, Tuesday Morning