Abstract A28: Correlation of ERG expression with other genetic and expression abnormalities in prostate cancer

Abstract

Abstract Fusion between TMPRSS2 (transmembrane protease, serine 2) and ERG (v-ets erythroblastosis virus E26 oncogene like (avian)) (TMPRSS2:ERG) is the most commonly found genetic alteration in prostate cancer and is found in 35–50% of all cases. TMPRSS2 is induced by the androgen receptor (AR). Fusion with TMPRSS2 renders ERG androgen inducible, causing most of its overexpression in prostate cancer. Studies suggest that SPINK1 (serine peptidase inhibitor, Kazal type 1) expression is exclusive to TMPRSS2:ERG fusion negative tumors, PTEN (phosphatase and tensin homolog) and TP53 (tumor protein p53) deletions are enriched in fusion positive tumors, a small region in chromosome 3p14 (including three possible tumor suppressor genes, FOX1, RYBP and SHQ1) is specifically deleted in a subset of fusion positive tumors, and that AR expression is down regulated in fusion positive tumors. Correlation with EZH2 (enhancer of zeste homolog 2 (Drosophila)) and Ki-67 (antigen identified by monoclonal antibody Ki-67) expression has also been suggested. To confirm these correlations we used fluorescence in situ hybridization (FISH) and immunohistochemistry (IHC) techniques in our sample material, which consisted of 330 prostatectomy samples and 150 locally recurrent castration resistant prostate cancers (CRPC). About half of the prostatectomy (47%, 107/230) and CRPC (45%, 70/156) samples showed positive ERG staining and ERG positivity was significantly associated with longer progression-free survival (p=0.0001, Log rank test) compared to ERG negative cases. Positive SPINK1 expression was seen in 14% (34/243) of the prostatectomy samples and in 12% (17/138) of the CRPC tumors. SPINK1 was almost exclusively expressed in ERG negative tumors (in prostatectomies p<0.0001 and in CRPCs p=0.0132, Fisher's exact test). High AR staining was seen in 82% (206/252) of the prostatectomy samples and in 93% (127/136) of the CRPC tumors. In prostatectomy samples, AR expression was significantly associated with ERG positivity (p<0.0001, Fisher's exact test). Low PTEN expression was seen in 15% (42/282) of the prostatectomy samples and in 45% (55/122) of the CRPC tumors. In CRPC tumors, but not in prostatectomy samples, loss of PTEN expression was associated with ERG positivity (p=0.0035, Fisher's exact test). Low PTEN expression was significantly associated with shorter progression-free survival compared to PTEN positive tumors (p=0.0133, Log rank test). Deletion of TP53 was seen in 8% (19/237) of the prostatectomy samples and in 18% (25/140) of the CRPC tumors. Deletion of chromosome region 3p14 was seen in approximately 4% of both prostatectomy (8/229) and CRPC (5/118) samples. Neither of these deletions was associated with ERG positivity. Analysis of Ki-67 and EZH2 staining is ongoing. SPINK1 was almost exclusively expressed in ERG negative samples, and the association was statistically significant in both the prostatectomy samples and the locally recurrent CRPC samples. PTEN loss of expression was associated with ERG positivity in the CRPC samples, but not in the prostatectomy samples, indicating that this late event occurs preferentially in ERG positive cases. High AR expression was statistically significantly associated with ERG positivity. Since there shouldn't be any ERG expression from TMPRSS2:ERG fusion gene without AR, the results are in line with the known regulatory mechanisms of TMPRSS2:ERG. Contrary to previous reports, the specific deletions of 3p14 or TP53 were not enriched in the ERG positive samples. TP53 deletion was more frequent in the advanced disease. Citation Format: Katri A. Leinonen, Teuvo L.J. Tammela, Outi R. Saramäki, Tapio Visakorpi. Correlation of ERG expression with other genetic and expression abnormalities in prostate cancer [abstract]. In: Proceedings of the AACR Special Conference on Advances in Prostate Cancer Research; 2012 Feb 6-9; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2012;72(4 Suppl):Abstract nr A28.