Expression Of ERG Protein In Prostate Adenocarcinoma And Its Clinicopathologic And Prognostic Significances

Objective:Prostate cancer is one of the most common malignancies that affect menin the world. Transmembrane protease serine2-ETS related gene (TMPRSS2-ERG)fusion has recently been found to be the most frequent gene rearrangement in prostatecancer, which is associated with expression of ERG protein. Recently, animmunohistochemical staining method using an anti-ERG antibody was shown to havea strong correlation with ERG protein expression. This prospective study was designedto investigate the clinicopathological and prognostic significances of ERG proteinexpression in Chinese prostate adenocarcinoma patients by ERG immunohistochemicalstaining.Materials and Methods:Paraffin-embedded prostate tissue samples were takenfrom a total of50patients diagnosed with prostate adenocarcinoma who underwenttransurethral resection of the prostate(TURP) between2007and2012. The mean age ofthe patients was69years (median76；range55to84years).The Gleason score andprimary Gleason grade were evaluated using pathologic records and slide review.Age,prostate volume,T-stage and preoperative levels of prostate-specific antigen (PSA)were collected from patient medical records. Immunohistochemistry was performed todetect the expression of ERG protein in prostate cancer tissue samples. ERG proteinexpression were microscopically evaluated by two board-certified pathologists withexpertise in urological pathology on the initial slides to verify the histologicaldiagnosis.All cases were followed up，with average follow-up time of42months(median35months; range9to65months).The correlation between ERG proteinexpression and clinicopathologic features and patients’prognoses were statisticallyanalyzed.Result:1.Of the50cases, only13/50(26.0%) showed positive ERG immunohistochemical staining,with5/13(38.46%) of patients showing low ERG intensit y versus8/13(61.54%) showing high ERG intensity. In all these specimens, normal prostatic epithelial were negative.2.ERG protein expression was more frequently detected in the subgroup with alower primary Gleason grade (Grade≤3) than a higher primary Gleason grade (Grade4or5) in the subgroup with(39.3%vs9.1%；p=0.017). Additionally, the frequency ofERG protein expression was significantly higher in the subgroup with the lowerGleason score(≤7) than in the subgroup with the higher Gleason score(8-10)(47.4%vs12.9%)(p=0.008). Moreover,we observed comparable rates for thedistribution low and high ERG intensities with5/13(38.46%) of patients demonstratinglow ERG intensity versus8/13(61.54%) patients with high ERG intensity.Using thiscriterion,there was significantly inverse relation relationship between ERG intensitylevels and Gleason score with weaker intensity of ERG more frequently occurring inhigher Gleason score tumors and higher intensity ERG occurring in lower Gleasonscore tumors(p=0.003).3.There was statistical significance between ERG protein expression and tumorstage [(T1+T2)vs(T3+T4),p=0.016]. Other factors including age, prostate volume andpreoperative prostate-specific antigen（PSA）level were not significantly related with theERG protein expression.Although statistical significance was not achieved due to thesmall sample size, the younger cases (≤65years) showed higher incidence of ERGprotein expression than others.4.Patients with ERG protein expression had shorter free progression time tocastration resistant after endocrine therapy for prostate cancer compared to patient withno ERG protein expression(p=0.04).ERG protein expression was significantlyassociated with higher rate of cancer specific mortality(p=0.037).With this subgroup ofERG positive patients, patients with low ERG intensity showed no overall survivaladvantage compared to patients with low ERG intensity(p=0.07). Gleason score(p=0.011), T-stage (p=0.028) and ERG expression (p=0.015) all had independentpredictive value for time to cancer specific death in multivariate analysis.Conclusions:1.Using ERG immunohistochemical staining, we found the ERG-positive rate inChinese prostate adenocarcinoma cases to be26.0%,but not in begin glands.2.ERG expression is significantly higher in tumors with well-formed glands andlow T-stage and is associated with a lower Gleason score and a lower primary Gleasongrade.High intensity ERG more frequently occurred in lower Gleason score tumors.There was no significant association between ERG expression and age, prostatevolume and initial prostate-specific antigen level.3.Patients with ERG protein expression had shorter free progression time tocastration resistant after endocrine therapy for prostate cancer.ERG expression wassignificantly associated with higher rate of cancer specific mortality and a poorerprognosis.