Increased frequency of CHD1 deletions in prostate cancers of African American men is associated with rapid disease progression without inducing homologous recombination deficiency

We analyzed genomic data derived from the prostate cancer of African and European American men in order to identify differences that may contribute to racial disparity of outcome and that could also define novel therapeutic strategies. In addition to analyzing patient derived next generation sequencing data, we performed FISH based confirmatory studies of Chromodomain helicase DNA-binding protein 1 (CHD1) loss on prostate cancer tissue microarrays. We created CRISPR edited, CHD1 deficient prostate cancer cell lines for genomic, drug sensitivity and functional homologous recombination (HR) activity analysis. We found that subclonal deletion of CHD1 is nearly three times as frequent in prostate tumors of African American men than in men of European ancestry and it associates with rapid disease progression. We further showed that CHD1 deletion is not associated with homologous recombination deficiency associated mutational signatures in prostate cancer. In prostate cancer cell line models CHD1 deletion did not induce HR deficiency as detected by RAD51 foci formation assay or mutational signatures, which was consistent with the moderate increase of olaparib sensitivity. CHD1 deficient prostate cancer cells, however, showed higher sensitivity to talazoparib. CHD1 loss may contribute to worse outcome of prostate cancer in African American men. A deeper understanding of the interaction between CHD1 loss and PARP inhibitor sensitivity will be needed to determine the optimal use of targeted agents such as talazoparib in the context of castration resistant prostate cancer.

Abstract 3805: Immunologic transcript and cell type evaluation of prostate tumors from a military cohort of African American and Caucasian American patients

Background: Health and racial disparities in prostate cancer place African American (AA) men at greater risk of developing and having a poorer outcome from the disease, especially at a younger age compared to Caucasian American (CA) men. This incidence is also reflected among active-duty service members, and patient biospecimens obtained from an equal access healthcare setting at the Walter Reed National Military Medical Center provide a valuable resource for the evaluation of cancer health disparities. The objectives of this study are to identify immunobiological differences influencing prostate cancer disparities and to elucidate the immune cell profiles of patient tumors associated with advanced disease.

Methods: Patients provided written consent to both biospecimen and clinical database collection under IRB-approved protocols. Fresh frozen tumor biopsy tissues were collected ex vivo, following radical prostatectomy. Total tumor RNA was amplified by PCR-based multiplexed target enrichment, and barcode-tagged transcripts were quantified using NanoString technology. Raw and relative abundances of immune cells were determined using published deconvolution algorithms. Differential expression of immune-related genes and cell type contrasts were evaluated for correlation with clinico-pathologic features.

Results: Genes regulating metabolism and innate immune responses were differentially expressed between AA and CA prostate tumors (AA n=26, CA n=25). Comparing high vs. low expression of each of these top genes, two were associated with biochemical recurrence (BCR)-free survival. Most immune cell subtypes did not differ significantly between AA and CA, but mast cells appeared to be enriched within AA tumors. When cell types were stratified by clinical and pathologic variables, we identified consistent trends in immune cell content that changed with increasing diagnostic age, PSA group, Grade Group, Gleason Sum, and with disease progression as defined by future development of BCR and/or metastasis.

Conclusions: Attention should be directed toward observed immunobiological differences based on race and other clinico-pathologic factors at the time of radical prostatectomy. Patient-centered studies mindful of existing health disparities will aid in diagnostic and therapeutic strategies that are inclusive of an increasingly diverse US and US military population.

Disclaimer: The contents of this publication are the sole responsibility of the authors and do not necessarily reflect the views, opinions or policies of the USUHS, HJF, the DoD or the Departments of the Army, Navy, or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.

IRB protocol: DBS.2019.032 (Ref Number 930187)

Citation Format: Cara C. Schafer, Jiji Jiang, Sally Elsamanoudi, Darryl Nousome, Denise Young, Yingjie Song, Isabell Sesterhenn, Gregory Chesnut, Shyh-Han Tan. Immunologic transcript and cell type evaluation of prostate tumors from a military cohort of African American and Caucasian American patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3805.

Race, tumor location, and disease progression among low-risk prostate cancer patients

Background

The relationship between race, prostate tumor location, and BCR‐free survival is inconclusive. This study examined the independent and joint roles of patient race and tumor location on biochemical recurrence‐free (BCR) survival.

Methods

A retrospective cohort study was conducted among men with newly diagnosed, biopsy‐confirmed, NCCN‐defined low risk CaP who underwent radical prostatectomy (RP) at the Walter Reed National Military Medical Center from 1996 to 2008. BCR‐free survival was modeled using Kaplan‐Meier estimation curves and multivariable Cox proportional hazards (PH) analyses.

Results

There were 539 eligible patients with low‐risk CaP (25% African American, AA; 75% Caucasian American, CA). Median age at CaP diagnosis and post‐RP follow‐up time was 59.2 and 8.1 years, respectively. Kaplan‐Meier analyses showed no significant association between race (P = .52) or predominant tumor location (P = .98) on BCR‐free survival. In Cox PH multivariable analysis, neither race (HR = 1.18; 95% CI = 0.68‐2.02; P = .56) nor predominant tumor location (HR = 1.13; 95% CI = 0.59‐2.15; P = .71) was an independent predictor of BCR‐free survival.

Conclusions

Neither race nor predominant tumor location was associated with adverse oncologic outcome.

Abstract 3165: Identification of candidate biomarkers for aggressive prostate cancer using targeted proteomics and FFPE tissue samples with outcomes data

Introduction: Although many (~40%) of the screen-detected prostate cancers are indolent (Gleason Score <= 6) and pose minimal risk for progression, advanced stage prostate cancer is a lethal disease with 5-year survival rates around 29%. The challenge is to identify biomarkers for early detection of aggressive disease, when the cancer is still organ confined. Such markers would be also used to better select patients with indolent and low risk cancers for active surveillance.

Material and Methods: To identify a panel of protein markers that could predict prostate cancer progression, we developed ultra-sensitive, high-pressure, high-resolution separations coupled with intelligent selection and multiplexing-selected reaction monitoring (PRISM-SRM) assays for 52 protein markers. Candidate protein markers were identified and selected from existing prostate cancer genomics data sets and validated lists of known prostate cancer drivers. The PRISM-SRM assays used heavy isotope-labeled synthetic peptides as internal standards for quantitative proteomics analysis. Study comprised of a prostate cancer patient cohort with organ confined primary tumors (N=338) presenting following post-surgery features: 53 (15.7%) metastatic progression, 124 (36.7%) biochemical recurrence (BCR), and 161 (47.6%) no progression after more than ten years of follow-up after radical prostatectomy. Index tumor region for each case was scraped from representative 10-m sections of formalin-fixed paraffin embedded (FFPE)-whole-mounted prostatectomy specimens and processed for PRISM-SRM analysis.

Results: Overall, PRISM-SRM analysis of the FFPE tissue samples enabled the detection of 42 (80.8%) out of 52 biomarker candidates; in comparison regular LC-SRM without the front-end chromatographic enrichment could detect only 21 (40.4%) of these candidates at the protein level. Kruskal-Wallis testing was used for statistical evaluation of the PRISM-SRM results and comparison of relative protein levels between the “no progression”, BCR and “metastatic progression” groups. Several prostate differentiation/androgen receptor signaling related proteins (FOLH1, PSA and NCOA) and tumor progression-related proteins (TGFB1, CCND1 and SPRC) had significantly different expression levels between the three groups, and showed initial promise in predicting progression to invasive cancer, BCR and metastasis.

Conclusion: Ultra-sensitive targeted proteomics can be used to select and verify performance of early prognostic markers based on the analysis of prostatectomy specimens. The top performing markers appeared able to predict progression from organ confined cancer to BCR and metastasis.

Citation Format: Yuqian Gao, Yi-Ting Wang, Hui Wang, Denise Young, Jennifer Cullen, Yingjie Song, Yongmei Chen, Athena Schepmoes, Gyorgy Petrovics, Thomas Fillmore, Tujin Shi, Wei-Jun Qian, Richard Smith, Sudhir Srivastava, Jacob Kagan, Albert Dobi, Inger Rosner, Karin Rodland, Isabell Sesterhenn, Shiv Srivastava, Tao Liu. Identification of candidate biomarkers for aggressive prostate cancer using targeted proteomics and FFPE tissue samples with outcomes data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3165.

Abstract 3129: Investigating the importance of prostate tumor location and symmetry in a racially diverse, military cohort

Introduction and objective

Prostate cancer (CaP) tumor location (TL) and symmetry (TS), may help forecast CaP outcomes in radical prostatectomy (RP) patients. The study aim was to examine associations between CaP TL and TS with disease pathology and progression, including biochemical recurrence (BCR)-free survival especially whether such associations differ across race.

Methods

This retrospective cohort study included 1140 patients enrolled at the Walter Reed National Military Medical Center who underwent RP between 1993 and 2008. TL and TS were recorded for index tumors (i.e, largest/highest grade) in whole-mounted RP specimen sections. TLs were divided into coronal versus axial planes. For coronal plane, TL was sub-divided into anterior (A), antero-lateral (AL), lateral (L), postero-lateral (PL), posterior (P) and diffuse (D) (includes A, AL, L, PL and P). The axial plane locations included base (B), mid (M) and apex (A). Kaplan Meier (KM) estimation curves were used to model BCR-free and DM-free survival.

Results

Mean age at diagnosis was 59.3 years, 27.1% were African-American (AA) and 76.7% were non-symmetric (NST). According to KM analysis patients with NST experienced significantly worse BCR-free survival (p=0.003) with no racial differences found (CA: p=0.02, AA: p=0.04). Tumor location (coronal plane) distribution by race was: A and AL tumors (CA: 9.6%, 7.5% vs. AA: 9.5%, 5.5% respectively), D tumors (CA: 6.9% vs. AA: 9.4%, respectively), L tumors (CA: 9.3% vs. AA: 6.5%, respectively), P and PL tumors (CA: 21.2%, 45.6% vs. AA: 22.7%, 46.3%, respectively) (p=0.36). In the axial plane exclusively A or B tumors are rare (CA: 3.1%, 1.3% vs. AA: 2.6%, 0%, respectively) while exclusively M tumors are more frequent (CA: 27.9% vs. AA: 23%, respectively). Tumors spanning two axial location categories MA and BM tumors were (CA: 27.6%, 15.6% vs. AA: 34%, 8.4%, respectively). Tumors spanning three axial location categories BMA were (CA: 21.8% vs. AA: 30.7%, respectively) (p<.0001). No racial differences in BCR-free survival were observed for D (CA: p<.0001, AA: p=.07) or BMA (CA p<.0001; AA p=0.0004) tumors. The share of tumors with nuclear grade III in D or BMA location categories was (10.5% and 7.2%, respectively; p<.0001) while the difference in NST and ST was (4.2% vs. 0.8%, p=.0002).

Conclusions

In this longitudinal, racially diverse cohort of RP patients, TL and TS were informative features in predicting BCR. However, this study did not find evidence of racial differences in these associations with the exception of axial tumor location. These findings have implications for treatment stratification at time of RP.

Citation Format: William Gesztes, Grant Williams, Jennifer Cullen, Allen Burke, Denise Young, Justin Mygatt, Yongmei Chen, Huai-Ching Kuo, Shiv Srivastava, Kevin Rice, Inger Rosner, Isabell Sesterhenn. Investigating the importance of prostate tumor location and symmetry in a racially diverse, military cohort [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3129.

ERG Tumor Type is Less Frequent in High Grade and High Stage Prostate Cancers of Chinese Men

Background: The incidence of prostatic adenocarcinoma has been rapidly increasing among Chinese men. This alarming trend prompted evaluations of early causal genomic alterations known to drive prostate tumorigenesis. Recurrent activation of the ETS-Related Gene (ERG) by genomic rearrangements is the most recognized early event in prostate cancer. Following the initial detection of ERG rearrangement at gene expression and genomic and levels, development of diagnostic quality antibodies against ERG oncoprotein have streamlined the rapid assessment of ERG frequencies world-wide. Unexpectedly, these studies revealed highest frequencies of ERG among Caucasian descents, lower frequencies among African Americans and even lower prevalence of ERG among Asian men.

Objective: To asses in a prospective study ERG frequencies, clinico-pathological and prognostic associations of ERG among prostate cancer patients of the Dalian region of Northeast China, by an established immunohistochemical procedure that have been used in studies world-wide.

Methods: Formalin fixed paraffin embedded specimens donated by patients (N=50) diagnosed with prostatic adenocarcinoma who underwent transurethral resection of the prostate (TURP) between 2007 and 2012 were evaluated for ERG by immunohistochemistry.

Results: Of the 50 cases, 13/50 (26.0%) tumors were positive for ERG. In all cases, normal prostatic epithelial were ERG negative. ERG was more frequently detected in the lower Gleason score (≤7) and low T-stage. Consistent with reports from Asian countries the results of our study shows lower overall frequencies of ERG positive tumors when compared to reports from Western countries.

Conclusion: The intriguing association of even lower ERG frequencies with high Gleason scores and higher T-stages provides impetus for current driver gene discoveries focused on the predominantly ERG negative prostate cancers of Asian men.

The relative contributions of the Genomic Prostate Score and comorbidity profile in predicting outcomes in surgically treated men with clinically low and intermediate-risk prostate cancer

e16607

Background: The 17-gene Oncotype Dx Genomic Prostate Score® (GPS™) assay has been validated as a predictor of aggressive prostate cancer (PCa) in men treated with radical prostatectomy (RP) for clinically low and intermediate risk PCa. This study examined the performance of the GPS assay as a predictor of adverse pathology (AP) and biochemical recurrence (BCR)-free survival (BRFS), after adjusting for the presence of major comorbidities commonly seen in older men. Methods: Additional analyses were performed using data from a prior clinical validation study of the GPS assay. GPS values (scaled 0-100), determined from diagnostic biopsy tissue, were categorized into 3 levels: lowest quartile (Q1), middle quartiles (combined Q2-Q3) versus the highest quartile (Q4). Major comorbidities included heart disease, stroke, COPD, and other cancers. The associations between GPS result and the presence of ≥1 major comorbidity and outcomes were examined. AP was defined as high-grade (Gleason Score ≥ 4+3) and/or pT3 tumor. BCR was defined as 2 successive PSA levels > 0.2 ng/mL. Logistic regression analysis was used to examine AP; Cox proportional hazards analysis was used to model BRFS. Results: Among 389 eligible men, median age at diagnosis and follow-up time was 62 and 5.6 years, respectively. The prevalence of ≥1 major comorbidity differed significantly across GPS category, rising from 10.2% to 19.7% to 25.5% for GPS Quartiles 1, 2-3, and 4, respectively (p = 0.0024). However, presence of ≥1 major comorbidity was not a significant predictor of AP or BCR in multivariable models with GPS, age, race, and NCCN risk stratum. Men whose GPS result was in the highest quartile had 4-fold higher odds of AP (OR: 4.1; 95% CI = 2.1-7.99, p < 0.0001) at RP and a 3.5 times higher risk of BCR (HR = 3.49; 95% CI = 1.59-7.64, p = 0.002) compared to men in the lowest GPS quartile (Table 1). Conclusions: A high GPS result remains the strongest predictor of BCR and AP in this cohort of low and intermediate risk PCa men, after adjusting for the presence of major comorbidities. Further work will explore the relationships between specific comorbidities and metabolic syndrome, with GPS testing and PCa outcomes.

Abstract A018: Distinct genomic alterations in prostate cancer of African American men

Background: In recent years remarkable ethnic/racial differences in the frequency of prostate cancer (CaP) driver gene alterations have emerged. Along the lines of ethnic differences in CaP genomes, oncogenic activation of ERG proto-oncogene and deletion of the PTEN tumor suppressor gene are most widely studied. These findings underscore distinct biology of prostate cancer in different ethnicities and races. We have reported the cumulative analyses of 435 patients (whole genome sequencing (WGS): 14, FISH evaluations: 101, SNP array: 320) comparing CaP genomes of African American (AA) and Caucasian American (CA) patients. An AA CaP genome-associated deletion was identified and mapped to the Limbic System-Associated Membrane Protein (LSAMP) gene locus. Further examination of the data indicated the AA CaP-genome associated deletion of the Chromodomain Helicase DNA Binding Protein 1(CHD1) gene. The goal of the current study is the comparative assessment of frequencies and prognostic associations of CaP genome alterations in the context of racial/ethnic differences.

Methods: A prostate tumor tissue microarray (TMA) of benign and tumor foci (500+ cores) was generated by sampling 2-3 cores representing the range of cell morphology in matched cohort of 42 AA and 59 CA patients with up to 20 years of follow-up. Comparative evaluation of frequencies and prognostic associations of ERG oncoprotein by immunohistochemistry and the deletion of LSAMP, CHD1, and PTEN genes by FISH was performed. ERG frequencies were further assessed in index tumors of Chinese CaPs (N=100) and were compared to ERG frequencies in index tumors of patients from the United States equal-access military health-care system (N=336 AA and N=594 CA).

Results: Higher frequencies of recurrent deletions of CHD1 (29% AA vs. 10% CA p=0.017) and LSAMP (26% AA vs. 7% CA, p=0.006) were identified in AA CaP compared to CA CaPs. These deletions were associated with significantly higher risk of rapid biochemical recurrence. In contrast, PTEN deletions (15% AA vs. 63% CA) and the frequency of ERG positive CaPs (26% AA vs. 64% CA) were more frequent among CA CaPs. Comparative evaluation of ERG frequencies in CaPs (N=1,292) underscores highest ERG frequency among CA patients (49.3%) followed by AA (23.2%) and Chinese (22%) men.

Conclusions: In light of distinct biology of CaPs in ethnically/racially diverse CaP patient populations, there is a need for developing broadly applicable diagnostic, prognostic marker panels and therapeutic approaches. Higher frequency of CHD1 deletion in CaPs of AA patients may provide new therapeutic opportunities due to the sensitivity of CHD1 deletion harboring tumors to PARP inhibitors and platinum agents.

Acknowledgments: This research was supported by the CPDR-USU program HU0001-10-2-0002, the National Cancer Institute R01CA162383 grant, the EDRN/NCI ACN12011-001-0 grant, the Snyder Medical Foundation, the Otto Mønsteds Foundation, the Mazzone Foundation, the Breast Cancer Research Foundation, the Novo Nordisk Foundation, and grants by MTA-TKI643/2012, KTIA_NAP_13-2014-0021.

Citation Format: Albert Dobi, Gyorgy Petrovics, Hua Li, Denise Young, Shyh-Han Tan, Yongmei Chen, Qingyu Xiao, Yidi Sun, Hong Li, Yixue Li, Yuan Ji, Jun Hou, Wendy Wang, Jacob Kagan, Guo-Ping Zhao, Sudhir Srivastava, Reinhard Ebner, Inger L. Rosner, Jennifer Cullen, Matthew Freedman, Isabell Sesterhenn, Zoltan Szallasi, Shiv Srivastava. Distinct genomic alterations in prostate cancer of African American men [abstract]. In: Proceedings of the AACR Special Conference: Prostate Cancer: Advances in Basic, Translational, and Clinical Research; 2017 Dec 2-5; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(16 Suppl):Abstract nr A018.

Abstract 1230: Association of common germline variants with TMPRSS2-ERG gene fusion status in prostate cancer

Introduction and Objectives: Oncogenic activation of ERG resulting from prevalent gene fusions (predominantly as TMPRSS2-ERG) is a key driver event in prostate cancer (CaP) pathogenesis. Our laboratory and others have recently reported that major cancer driver genes, including ERG, show significant racial/ethnic differences in CaP. It is present in two thirds of CaP patients of European Ancestry including Caucasian Americans (CA) but is present at lower frequencies in African Americans (AA), Africans and Asians. Racial differences of CaP associated SNPs have also been extensively described. However, there is limited data on germline association with ERG fusion status. The goal of this study is to identify germline molecular determinants associating with ERG status of CaP. Methods: Blood derived genomic DNA samples were prepared from 270 AA men and 130 CA men treated by radical prostatectomy. ERG status was determined by immuno-histochemistry (IHC) for ERG protein expression. SNP genotyping was performed on the Illumina Golden Gate platform using Infinium Oncoarray SNP chip. Data analysis approaches included association analyses based on EMMAX and imputation analysis by IMPUTE2. SNP genotyping was performed using droplet digital polymerase chain reaction (ddPCR) approach Results: SNP genotyping analysis was performed in 321 patients with 478,299 SNPs. We identified SNPs associated with ERG status using EMMAX analysis. The SNPs most significantly (p &lt;10-5) associated with ERG fusion status included rs6698333, an intron variant of Kruppel-like factor 17 (KLF17) and two SNPs (rs1889877, rs3798999) in the intron of adhesion G protein-coupled receptor B3 (ADGRB3). 4 SNPs (rs10215144, rs3818136, rs9380660 and rs1792695) were found to be significantly (p &lt;10-5) associated with ERG positive phenotype under any tumor foci positive for the fusion. Fine-mapping of SNPs by genotype imputation analysis (IMPUTE2) using the 1000 Genomes reference dataset, found rs34349373 and rs2055272 to be significantly associated (p &lt;10-7). The 2 variants were found to be in strong linkage disequilibrium (LD) in both CA and AA populations with r2 of 1.0 and 0.91 respectively. Imputed SNP rs2055272 was further experimentally evaluated by Taqman based ddPCR SNP genotyping approach. Concordance between Taqman genotypes and imputed genotypes was found to be 98.04%. Association analysis of the SNPs with clinico-pathological features of CaP and functional annotation of the significant SNPs by in silico eQTL based analysis are being performed. Conclusions: This study identified SNPs associated with ERG status of CaP, a major driver oncogene in CaP. Although the biological significance as it relates to ERG status of CaP still needs to be determined, these SNPs, with independent validation, may help as markers in stratifying patients early (even before CaP is detected) for targeted prevention and treatment options.

Citation Format: Indu Kohaar, Yongmei Chen, Lakshmi Ravindranath, Denise Young, Amina Ali, Qiyuan Li, Albert Dobi, Inger L. Rosner, Isabell Sesterhenn, Jennifer Cullen, Matthew Freedman, Shiv Srivastava, Gyorgy Petrovics. Association of common germline variants with TMPRSS2-ERG gene fusion status in prostate cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1230.

Abstract 2580: A mosaic ERG oncoprotein staining pattern predicts upstaging and poor pathologic features in prostate cancer patients

Purpose: Building on previous work, this study examined the independent role of ERG staining pattern in the index prostate cancer (CaP) tumor, on probability of upstaging and biochemical recurrence (BCR) following radical prostatectomy (RP).

Methods: A retrospective cohort study of patients enrolled at the Walter Reed National Military Medical Center who underwent RP between 1993 and 2016 was conducted. ERG oncoprotein expression was assessed in CaP index tumors (i.e., largest and poorest grade) using immunohistochemistry (IHC) staining with 9FY monoclonal antibody of representative sections of formalin-fixed paraffin-embedded whole-mounted RP specimens. ERG status was classified as positive (ERG+), negative (ERG-) and a pattern of alternating +/- regions within the index tumor (ERG “Hybrid” or ERG-h). Kaplan Meier (KM) estimation curve analysis was used to examine BCR-free survival across ERG status. Multivariable (MV) logistic regression (LR) modeling was used to determine whether ERG status predicted upstaging (cT1-2 to pT3-4).

Results: Among 1,385 eligible patients, mean age at RP and follow up time were 58.8 and 7 years, respectively. Twenty-nine percent of patients self-classified as African-American (AA), and 18% experienced BCR. The ERG-h pattern was observed in 81 patients (5.8%). Pathologic stage T3a-T4a was observed in 51.3%, 32.2% and 30.5% of patients with ERG-h, ERG+ and ERG- tumors, respectively. In unadjusted KM analysis, 10-year BCR-free survival probabilities were 78%, 75%, and 82% for ERG-h, ERG-, and ERG+ patients, respectively (p=0.08). A trend of incrementally poorer BCR-free survival was observed for ERG-h, ERG- versus ERG positive patients. MV LR analysis revealed the greatest odds of upstaging for ERG-h vs. ERG- patients (OR=2.8, p=0.0008), controlling for age at RP (p&lt;.0001), biopsy Gleason score (p&lt;.0001), PSA at diagnosis (p=0.0054) and patient race (p=0.07).

Conclusions: In this racially diverse cohort of RP patients, a mosaic-like pattern of ERG staining predicted greater upstaging and poorer pathologic features. There may be a role for examination of this “hybrid” mosaic staining pattern in patient risk stratification during treatment decision making.

Citation Format: William Gesztes, Shahnoza Dusmatova, Justin Mygatt, Jennifer Cullen, Denise Yong, Huai-Ching Kuo, Yongmei Chen, Lauren Hurwitz, Inger Rosner, Shiv Srivastava, Isabell Sesterhenn. A mosaic ERG oncoprotein staining pattern predicts upstaging and poor pathologic features in prostate cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2580.

Abstract 1290: Association of common genetic variants with TMPRSS2 ERG fusion status in prostate cancer

Introduction and Objectives: Oncogenic activation of ERG resulting from prevalent gene fusions is present in two thirds of prostate cancer (CaP) patients of European Ancestry including Caucasian Americans (CA). Our laboratory and others have recently reported that major cancer driver genes, including ERG, show significant racial/ethnic differences in CaP with lower frequencies in African Americans (AA), Africans and Asians. Racial differences of CaP associated SNPs have also been extensively described. However, there is limited data on germline association with ERG fusion status. The goal of this study is to identify germline molecular determinants associating with ERG status of CaP.

Methods: Blood derived genomic DNA samples were prepared from 270 AA men and 129 CA men treated by radical prostatectomy at Walter Reed National Military Medical Center (WRNMMC). ERG status was determined in whole mounted prostate specimens by immuno-histochemistry (IHC) for ERG protein expression as a surrogate for the TMPRSS2-ERG fusion. Blinded blood samples were genotyped for SNPs on the Illumina Golden Gate platform using Infinium Oncoarray, a 500K genome wide BeadChip kit from Illumina. Data analysis approaches included association analyses based on logistic regression, Principal Component Analysis (PCA) and Efficient Mixed-Model Association eXpedited (EMMAX) analysis. Genotype imputation analysis is being performed by IMPUTE2 program.

Results: After applying rigorous sample and SNP QC steps on the datasets, SNP genotyping analysis was performed in 321 patients with 478,299 SNPs. Logistic regression, principal component analysis by EIGENSTRAT and a variance component approach, EMMAX analysis were performed to account for population structure. By EMMAX we identified SNPs associated with ERG status. The SNPs most significantly (&lt;10-5) associated with ERG fusion status included rs6698333, an intron variant of Kruppel-like factor 17 (KLF17) and two SNPs (rs1889877, rs3798999) in the intron of adhesion G protein-coupled receptor B3 (ADGRB3). Fine-mapping of SNPs is underway by genotype imputation analysis (IMPUTE2) using the 1000 Genomes reference dataset, followed by independent validation.

Conclusions: This study identified SNPs differentially associated with ERG status of CaP, a major driver oncogene in CaP. Although the biological significance as it relates to ERG status of CaP still needs to be determined, these SNPs, with independent validation, may help as markers in stratifying patients early (even before CaP is detected) for targeted prevention and treatment options.

Citation Format: Indu Kohaar, Lakshmi Ravindranath, Denise Young, Amina Ali, Qiyuan Li, Albert Dobi, David McLeod, Inger L. Rosner, Isabell Sesterhenn, Matthew Freedman, Shiv Srivastava, Gyorgy Petrovics. Association of common genetic variants with TMPRSS2 ERG fusion status in prostate cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1290. doi:10.1158/1538-7445.AM2017-1290

Abstract 2830: Secreted protein acidic and rich in cysteine antigen and autoantibodies in sera of prostate cancer patients: Potential use in diagnosis /prognosis

Objectives: Improvements in blood based biomarkers for detecting clinically significant prostate cancer (CaP) or for distinguishing between indolent and aggressive CaP are critical in enhancing the management of this most common non-skin cancer of men in the US. These biomarkers include tumor associated antigens (TAAs) and autoantibodies (AAbs) against TAAs in patient sera. SPARC/Osteonectin is highly expressed in metastatic cancers (glioblastoma, melanoma, breast cancer and prostate cancer) and promotes bone metastasis and epithelial-mesemchymal transition. Our comparative transcriptome analyses of well/moderately differentiated CaP with poorly differentiated CaP, defined SPARC as central node in the network of genes with common regulatory elements (NKXH_NKXH_HOX) associating with poorly differentiated CaP. Further, we reported associations of high levels of SPARC mRNA or protein with Gleason 8-10 or poorly differentiated primary tumors and association with metastatic progression. The aims of this study are: i) Quantify SPARC antigen in the sera CaP patients; ii) Are AAbs against SPARC present in the serum of CaP patients?; iii) Is there a correlation between SPARC AAb level in patient sera and disease status? iii) Does SPARC AAb level vary according to distinct ethnic groups?

Methods: Sera from CaP patients and healthy controls were evaluated for SPARC antigen using a commercial enzyme-linked immunosorbent assay (ELISA) kit. AAbs against SPARC were determined by ELISA utilizing recombinant full-length human SPARC protein as a substrate. For evaluation we used Caucasian American (CA, n=117) and African American (AA, n=111) CaP patients comprising Gleason 6-10, and healthy controls (CA, n=52; AA, n=45).

Results: SPARC antigen levels in the sera showed a difference in the overall case versus control groups (p=0.0016) and this trend was observed only in AA (p=0.0015) in comparison to CA patients (p=0.1709). SPARC AAbs were detected in the sera and the values were significant in the overall case versus control (p&lt; .0001) and also in CA (p&lt;.0001) and AA (p&lt;.0001) CaP patients with levels significantly lower in patients compared to controls. AAb reactivity for CaP patients was similar between CA and AA groups.

Conclusions: This study demonstrated the presence of AAbs against SPARC in CaP patient serum for the first time. Of note, highly significant differences were noted between CaP patient (low) and controls (high) sera, across different ethnic groups. These data suggest further evaluation of SPARC AAbs as a promising serum biomarker for CaP.

Source of Funding: This research was supported by the Center for Prostate Disease Research, Uniformed Services University Grant HU0001-10-2-0002.

Citation Format: Anshu Rastogi, Andy Martinez, Amina Ali, Shyh-Han Tan, Jennifer Cullen, Yongmei Chen, Gyorgy Petrovics, Albert Dobi, Lakshmi Ravindranath, Denise Young, Isabell Sesterhenn, Jacob Kagan, Sudhir Srivastava, David McLeod, Inger Rosner, Shiv Srivastava, Alagarsamy Srinivasan. Secreted protein acidic and rich in cysteine antigen and autoantibodies in sera of prostate cancer patients: Potential use in diagnosis /prognosis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2830. doi:10.1158/1538-7445.AM2017-2830

History of PSA screening on prostate cancer aggressiveness

5066

Background: In 2012, PSA screening for prostate cancer (CaP) detection was given a “Grade D” recommendation for all men by the USPSTF. Recent U.S. studies report declines in PSA screening with concomitant increases in advanced CaP at diagnosis. This study examined the association between PSA screening history and CaP aggressiveness in a racially diverse, military cohort with equal health care access. Methods: This retrospective cohort study evaluated CaP patients undergoing radical prostatectomy (RP) from 1994-2015 at Walter Reed National Military Medical Center. Whole-mounted prostatectomy specimens were classified using 2014 ISUP Gleason grading system. Excluding the diagnostic PSA, screening history was categorized as: ≥ 6 PSA’s prior to CaP diagnosis (uppermost quartile), 1-5 (lower 3 quartiles), vs. no screening history. Multivariable logistic regression (MLR) was used to examine NCCN risk stratum (intermediate-high vs. low) and Gleason upgrade (GU) from biopsy to RP. Multivariable Cox proportional hazards (Cox PH) analyses were used to model time to biochemical recurrence (BCR). Multivariable models controlled for age at RP, race, family history and obesity (BMI > 30 vs. ≤ 30 kg/m2). The GU and BCR models also controlled for NCCN risk classification. Results: There were 1,772 eligible patients with a median follow-up and age at RP of 7.0 and 59.8 years, respectively. Prior to CaP diagnosis, 42% and 19% of men had 1-5 and ≥ 6 PSA’s screenings, respectively. MLR showed greater odds of intermediate or high vs. low risk disease for PSA screening history of none vs. 1-5 (OR = 1.33, CI = 1.03-1.70, p = 0.028) but not for none vs. ≥ 6 (p = 0.44). MLR showed increased odds of GU for none vs. ≥ 6 (OR = 1.81, CI = 1.23-2.7, p < 0.001). Multivariable Cox PH models showed incrementally poorer BCR-free survival as screening history decreased (HRNone vs. ≥6 = 2.27, CI = 1.54-3.33, p < 0.001; HRNone vs. 1-5= 1.49, CI = 1.15-1.92, p = 0.002). Conclusions: In this RP cohort, higher risk stratum, increased GU, and poorer BCR-free survival were associated with no PSA screening history. BCR-free survival was incrementally worsened by less PSA screening. A complete absence of PSA screening may lead to more aggressive disease at presentation and poorer clinical outcomes.

Influence of common comorbid conditions on a 17-gene assay for prediction of adverse pathology in clinically low-risk prostate cancer

91

Background: Well-validated molecular biomarkers can improve risk stratification in newly diagnosed PCa. It is important to demonstrate that biomarkers are not influenced by common clinical conditions. A 17-gene biopsy-based RT-PCR assay (Oncotype Genomic Prostate Score, GPS) has been validated as an independent predictor of adverse pathology and biochemical recurrence after surgery. We sought to explore the association between comorbid conditions and GPS. Methods: We examined the association between GPS and obesity (OB), diabetes mellitus (DM), hypertension (HTN), hypercholesterolemia (HC), coronary artery disease (CAD) and Low Testosterone (Low T) in a cohort of surgically treated men with clinically low-risk PCa. All men were seen and treated at Madigan Army Medical Center or Walter Reed National Military Medical Center. Diagnoses were abstracted from the electronic medical record. OB was defined as body mass index (BMI) > 30 kg/m2. Low T was defined as total T < 300 ng/mL. Student’s T test and ANOVA were used to assess differences in GPS between the groups of men with or without relevant diagnoses. Results: GPS results were available in 402 patients; 389 (97%) had BMI data. Pre-treatment serum testosterone (T) levels were available for 120 patients. There were no significant differences in mean GPS between men with or without OB, HTN, HCL, DM, and low T (p > 0.05). This relationship persisted when men with no vascular risk factors were compared to men with any vascular risk factor (DM/HTN/HCL) and/or with diagnosed coronary artery disease. There was a modest but statistically significant inverse correlation between GPS and continuous serum T (r = -0.18; p = 0.049). Conclusions: Comorbidities that are common in older men with prostate cancer do not appear to influence GPS results. Although Low T was not associated with GPS, there was a weakly negative association between GPS and continuous T. This may reflect the more aggressive PCa phenotype that has been reported in patients with progressively lower serum T.

Patient-specific meta-analysis (MA) of two validation studies to predict pathologic outcomes in prostate cancer (PCa) using a 17-gene genomic prostate score (GPS)

100

Background: We validated a biopsy-based assay as a predictor of PCa aggressiveness in two independent patient cohorts. Using GPS with NCCN risk group, the test provides an estimate of likelihood of favorable pathology (LFP) in low-intermediate risk PCa based on data from the first validation study. The second study re-confirmed the association between GPS and LFP. We combined information from both studies to provide more precise estimates of LFP, and examined predictive models using GPS with other clinical risk tools. Methods: Patient-specific MA provides precision-weighted predictions for individual patients using data from multiple studies (Crager and Tang, J. Appl. Stat. 2014). MA was performed on the two validation studies (732 patients total) using GPS (scale 0-100) with CAPRA score, NCCN risk group, or AUA/EAU risk group as predictors of LFP (Gleason score 3+3 or 3+4, organ-confined disease). Decision curves were calculated using the MA risk estimates. Results: MA provided more precise estimates of LFP with narrower confidence intervals (CIs) than either study alone (median width 24% narrower than the smaller of the two individual study CIs). GPS added significant predictive value for LFP to each of the 3 clinical classifiers. A model utilizing GPS and CAPRA provided the most risk discrimination. In decision-curve analysis, greater net benefit was shown when combining GPS with each clinical classifier compared with the clinical classifier alone. Over a wide range of threshold probabilities, incorporation of GPS would be expected to lead to fewer treatments of patients with favorable pathology without increasing the number of patients with adverse pathology left untreated. The proportion of men with MA-estimated LFP > 80% was 31% with GPS and CAPRA, and 23% with GPS and NCCN risk group; 11% were identified with LFP > 80% using NCCN risk group alone. Conclusions: Patient-specific MA of two independent studies provided more precise risk estimates reflecting the complete body of evidence. GPS adds predictive value to three clinical risk tools. The MA-estimated LFP identified more patients with an LFP > 80% than identified by clinical classifiers alone.

Abstract 4679: Silencing of PMEPA1, a NEDD4 E3 ubiquitin ligase binding protein, stabilizes androgen receptor and confers resistance to AR inhibitors

INTRODUCTION: The androgen receptor (AR) dysfunctions contribute to prostate cancer (CaP) development and progression. Earlier we defined the regulation of AR protein levels by a negative feed-back loop between AR and the androgen regulated PMEPA1, a NEDD4 E3 ubiquitin ligase binding protein. Our data also revealed decreased or absent PMEPA1 transcripts in 65% prostate tumors. This study focuses on the relationship of PMEPA1 protein with AR and its transcriptional targets (PSA) in human CaP, as well as the impact of PMEPA1 silencing on growth of PMEPA1-shRNA LNCaP cells derived tumors in naïve and castrated nude mice.

METHODS: Ten weeks old athymic Ncr-nu/nu mice were evaluated for the growth of tumor xenografts derived from PMPEA1-shRNA LNCaP (N = 20) and control-shRNA LNCaP cells (N = 20). Each mouse received 4×106 cells by subcutaneous injection to right flank side. The mice with tumors were castrated at 9th week post-injection. The tumor sizes were monitored twice per week until 16th week post-injection. The xenograft tissues were assayed by immunohistochemical (IHC) staining for PMEPA1, AR and PSA protein levels. The AR, PMEPA1 and PSA expressions in whole mount human prostate specimens were evaluated by IHC staining and QRT-PCR.

RESULTS: The PMEPA1 shRNA-LNCaP and control shRNA-LNCaP cells started to form subcutaneous tumors at 4 weeks after injections. However, the growth rate of PMEPA1 shRNA-LNCaP derived tumors was significantly faster than control derived tumors (P&lt;0.05). PMEPA1 shRNA-LNCaP tumors expressed higher levels of AR and PSA proteins as assessed by IHC. At 7 weeks after castration the tumor sizes increased by 69% in the control group in contrast to 304% in the PMEPA1-shRNA harboring xenografts (P&lt;0.05). In human CaP, decreased PMEPA1 mRNA expression significantly correlated with increased levels of AR transcription target PSA, as a surrogate for elevated AR. Focally increased PSA and AR protein levels were detected in a subset of low PMEPA1 expressing tumors.CONCLUSIONS: PMEPA1 silencing in CaP cells leads to gain of AR, increased tumor growth with potential to induce resistance to AR inhibition targeted therapy. Decreased PMEPA1 in human CaP correlated with increased AR activity. Thus, monitoring PMEPA1 expression in CaP cells offers new opportunities in therapeutic stratification of CaP. Further, restoring PMEPA1 in CaP cells may provide a new strategy for AR degradation focused therapy.

Citation Format: Hua Li, Elizabeth Umeda, Yingjie Song, Denise Young, Lakshmi Ravindranath, Ahmed Mohamed, Shashwat Sharad, Gyorgy Petrovics, David McLeod, Isabell Sesterhenn, Taduru Sreenath, Albert Dobi, Shiv Srivastava. Silencing of PMEPA1, a NEDD4 E3 ubiquitin ligase binding protein, stabilizes androgen receptor and confers resistance to AR inhibitors. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4679. doi:10.1158/1538-7445.AM2015-4679

Abstract 623: Evaluation of a multiplex biomarker assay for the detection of prostate cancer cells in urine

Background and Objectives: Over one million men undergo prostate biopsies annually in the US, a majority of them due to elevated serum PSA. More than half of the biopsies turn out to be negative for prostate cancer (CaP). The limitations of PSA test as well as the biopsy procedure have prompted to develop more precise CaP detection assays in urine- (e.g. PCA3, TMPRSS2-ERG) or blood (e.g. PHI, 4K). Here we describe the development and evaluation of a Urine CaP Marker Panel (UCMP) assay for sensitive and reproducible detection of CaP cells in post-digital rectal examination (post-DRE) urine.

Methods: The cellular content of the post-DRE urine, was captured on translucent membrane for direct evaluation by microscopy and immunocytochemistry. The membranes placed on Cytoclear microscope slides were used for characterization of prostate epithelial cell by PSA and Prostein or tumor cells by ERG and AMACR protein markers. Immunostained cells were analyzed for quantitative and qualitative features and correlated with biopsy positive and negative status for malignancy.

Results: Assay was optimized for single cell capture sensitivity and downstream evaluations by spiking known number of cells from established prostate cancer cell lines, LNCaP and VCaP, into cleared control urine. The cells captured from the post-DRE urine of subjects, obtained prior to biopsy procedure, were co-stained for ERG, AMACR, Prostein and PSA, rendering a whole cell based analysis and characterization. A feasibility cohort of 63 post-DRE urine specimens was assessed. Comparison of the UCMP results with the blinded biopsy results showed an assay sensitivity of 64% (16 of 25) and a specificity of 68.8% (22 of 32) for CaP detection by biopsy.

Conclusion: This pilot study assessing a minimally invasive prostate cancer detection assay with single cell sensitivity cell-capture and characterization from the post-DRE urine holds promise for further development of this novel assay platform.

Citation Format: Kristen Nickens, Amina Ali, Tatiana Scoggin, Shyh-han Tan, Lakshmi Ravindranath, David McLeod, David Tacha, Isabell Sesterhenn, Albert Dobi, Shiv Srivastava, Gyorgy Petrovics. Evaluation of a multiplex biomarker assay for the detection of prostate cancer cells in urine. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 623. doi:10.1158/1538-7445.AM2015-623

Abstract 5277: ERG-based stratification of prostate cancer highlights ethnicity associated biological differences

Introduction: The molecular basis of racial/ethnic differences in the detection and outcomes of the carcinoma of the prostate (CaP) needs to be better defined to improve biomarkers and identify appropriate therapeutic targets. ETS-related gene fusions in proto-oncogenes (ERG, ETV1) are among the most common early oncogenic alterations described in CaP. Studies have highlighted the racial/ethnic variation in ERG fusion and ERG oncoprotein expression frequencies showing: the highest frequencies in Caucasian (CA) men, the lowest frequencies in Asian men, and intermediate frequencies in African American (AA) men. These differences persisted when patients were carefully matched for age and clinico-pathologic features. Building on our and other's previous findings, the current study was conducted to evaluate ERG expression frequencies in a large cohort of AA and CA CaP patients.

Methods: Patients enrolled at the Walter Reed National Military Medical Center who consented to provide radical prostatectomy (RP) specimens to the CPDR Tissue Bank and Clinical Database were evaluated. The current study included 443 AA and 696 CA patients. ERG oncoprotein expression was detected by immunohistochemistry using the 9FY mouse monoclonal anti-ERG antibody and was correlated with clinico-pathologic features and CaP outcomes, including time to biochemical recurrence.&lt;br

Results: The result revealed lower frequencies of ERG in the index tumors of AA men as compared to CA men (28.4% vs. 51.5%, p &lt; 0.0001). Furthermore, AA men with higher grade CaP (Gleason 8-10) had even lower ERG frequency (16%), suggesting distinct molecular biologic features in aggressive CaP for AA men. A trend with age was observed for both racial groups, with younger CA showing higher ERG frequencies. At least one focus of ERG positive PIN was present in prostates with ERG positive tumors. In both CA and AA patients, a very low frequency (8%) of ERG expression was noted in anterior index tumors compared to posterior tumors.

Conclusions: This study lends increasing support to differences of the most common early genomic alteration (ERG) between AA and CA CaP patients. The study also underscores the importance of identifying new or improving existing biomarkers with consideration of patient race/ethnicity.

Citation Format: Albert Dobi, Michael Degon, James Farrell, Wagner Baptiste, Denise Young, Yongmei Chen, Gyorgy Petrovics, Jennifer Cullen, Jacob Kagan, Sudir Srivastava, Inger Rosner, David G. McLeod, Shiv Srivastava, Isabell Sesterhenn. ERG-based stratification of prostate cancer highlights ethnicity associated biological differences. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5277. doi:10.1158/1538-7445.AM2015-5277

The association of a 17-gene genomic prostate score with adverse surgical pathology and recurrence following surgery for localized prostate cancer (PCa): A comparison of African American and Caucasian patients

86

Background: Molecular assays can improve risk assessment for newly diagnosed PCa, but it is imperative to characterize assay performance in different racial groups, since tumor biology and clinical outcomes may vary. A racially diverse cohort of men (20% AA) with PCa in the Center for Prostate Disease Research multi-center national database was used to determine the association of GPS with outcomes in men treated with radical prostatectomy (RP) for localized PCa. Methods: Biopsy specimens from 431 men treated with RP for NCCN very low, low or intermediate risk PCa at 2 U.S. military medical centers were tested with a 17-gene RT-PCR assay to validate the association between GPS (scale 0-100) and 1) biochemical recurrence (BCR) following RP, and 2) adverse pathology (AP) at RP. BCR was defined as 2 successive PSA levels > 0.2 ng/mL. AP was defined as high-grade (primary Gleason pattern 4 or any pattern 5) and/or pT3 disease. Cox proportional hazards and logistic regression models were used. Results: GPS was obtained in 402 cases (93%), including 82 AA men. A broad range of GPS results was observed in both AA and CA men; GPS distributions were similar between AA (median GPS = 30.3; inter-quartile range (IQR): 23-38) and CA (median GPS = 30.3; IQR: 23-40); no correlation was observed between GPS and race (r = -0.04, p = 0.45). No differences in expression of individual genes or gene groups in the assay were observed between the two groups. In univariable analysis, PSA, biopsy GS and NCCN risk group were associated with BCR and AP, but race was not. The associations between GPS and clinical outcomes were similarly strong and statistically significant in both AA and CA men - BCR HR/20 GPS units = 3.0 (95% CI: 2.0-4.3) for CA vs. 3.5 (95% CI: 1.0-11.7) for AA; AP OR/20 units = 4.0 (95% CI: 2.6-6.6) for CA vs. 2.9 (95% CI: 1.2-7.6) for AA (p < 0.05 for all). Conclusions: In this cohort of patients treated in a health care system with equal access, clinical outcomes and the tumor biology measured by GPS were similar between AA and CA patients. GPS is a significant predictor of BCR and AP in men treated with RP for localized PCa in both AA and CA men.

Racial/ethnic variation in a common genetic alteration in prostate cancer

14

Background: The molecular basis of racial/ethnic differences in the detection and outcomes of the carcinoma of the prostate (CaP) needs to be better defined to improve biomarkers and identify appropriate therapeutic targets. ETS-related gene fusions in proto-oncogenes (ERG, ETV1) are among the most common early oncogenic alterations described in CaP. Studies have highlighted the racial/ethnic variation in ERG fusion and ERG oncoprotein expression frequencies showing: the highest frequencies in Caucasian (CA) men, the lowest frequencies in Asian men, and intermediate frequencies in African American (AA) men. These differences persisted when patients were carefully matched for age and clinico-pathologic features. Building on previous work, the current study was conducted to evaluate ERG expression frequencies in a large cohort of AA and CA CaP patients. Methods: Patients enrolled at the Walter Reed National Military Medical Center who consented to provide radical prostatectomy (RP) specimens to the CPDR Tissue Bank and Clinical Database were evaluated. The current study included 443 AA and 696 CA patients. ERG oncoprotein expression was detected by immunohistochemistry using the 9FY mouse monoclonal anti-ERG antibody and was correlated with clinico-pathologic features and CaP outcomes. Results: Lower frequencies of ERG in the index tumors of AA men as compared to CA men (28.4% vs. 51.5%, p < 0.0001) were observed. Furthermore, AA men with higher grade CaP (Gleason 8-10) had even lower ERG frequency (16%), suggesting distinct molecular biologic features in aggressive CaP for AA men. A trend with age was observed for both racial groups, with younger CA showing higher ERG frequencies. At least one focus of ERG positive PIN was present in prostates with ERG positive tumors. In both CA and AA patients, a very low frequency (8%) of ERG expression was noted in anterior index tumors compared to posterior tumors. Conclusions: This study lends increasing support to differences of the most common early genomic alteration (ERG) between AA and CA CaP patients. The study also underscores the importance of identifying new or improving existing biomarkers with consideration of patient race/ethnicity.

Diagnostic criteria for ductal adenocarcinoma of the prostate: interobserver variability among 20 expert uropathologists

AIMS

Ductal adenocarcinoma of the prostate (DAC) is clinically important, because its behaviour may differ from that of acinar adenocarcinoma. Our aims were to investigate the interobserver variability of this diagnosis among experts in uropathology and to define diagnostic criteria.

METHODS AND RESULTS

Photomicrographs of 21 carcinomas with ductal features were distributed among 20 genitourinary pathologists from eight countries. DAC was diagnosed by 18 observers (mean 13.2 cases, range 6-19). In 11 (52%) cases, a 2/3 consensus was reached for a diagnosis of DAC, and in five (24%) there was consensus against. In DAC, the respondents reported papillary architecture (86%), stratification of nuclei (82%), high-grade nuclear features (54%), tall columnar epithelium (53%), elongated nuclei (52%), cribriform architecture (40%), and necrosis (7%). The most important diagnostic feature reported for DAC was papillary architecture (59%), whereas nuclear and cellular features were considered to be most important in only 2-11% of cases. The most common differential diagnoses were intraductal prostate cancer (52%), high-grade PIN (37%), and acinar adenocarcinoma (17%). The most common reason for not diagnosing DAC was lack of typical architecture (33%).

CONCLUSIONS

Papillary architecture was the most useful diagnostic feature of DAC, and nuclear and cellular features were considered to be less important.

Abstract 1206: Evaluation of androgen receptor function index (ARFI) in prostate cancer

Introduction: Initially prostate cancer is driven by the male hormone androgen through the androgen receptor (AR). However, in some cases androgen receptor is functionally altered at late stages of tumorigenesis. Early knowledge of androgen receptor dysfunctions may assist in patient stratification for emerging therapeutic strategies. Although AR function can be altered by numerous mechanisms, the net effect of these changes is reflected in defective transcription factor functions of the AR. To develop readouts for AR function in prostate cancer we have evaluated the expression of a functionally relevant panel of AR-regulated genes in a prostate cancer cell culture model and in whole-mounted prostate tumor tissues.

Methods: In the VCaP cell culture model of prostate cancer that endogenously express, AR, TMPRSS2-ERG fusion (ERG), KLK3 (PSA), NKX3.1, PMEPA1, ODC1 and AMD1 genes we have evaluated the dose and time kinetics of these AR regulated genes in response to androgen (R1881) treatment at protein and mRNA levels. Radical prostatectomy specimens representing favorable or poor prognostic cases were selected for the study. Each prostate was fixed, sectioned and immunohistochemical (IHC) staining on adjacent four-micron sections of the whole-mounted blocks were performed by using mouse monoclonal anti-AR, anti-ERG antibody (clone 9FY), rabbit polyclonal anti-PSA and anti-NKX3.1 antibodies. IHC intensities were scored and the final score was determined after multiplying the intensity score and percentage of positively stained area in the respective lesions.

Results: Towards defining the expression levels of AR regulated genes in prostate cancer cells we have evaluated and confirmed the androgen dose and time kinetic response of endogenously expressed KLK3(PSA), PMEPA1, NKX3.1, ODC1, AMD1 and TMPRSS2-ERG (ERG) genes in VCaP cell culture model. Assessment of AR dysfunction by the immunohistochemical evaluation of AR regulated genes in whole-mounted prostate cancer specimens showed reduced expression of AR regulated genes in a sub-set of prostate cancer cases with notable inter- and intratumoral heterogeneity.

Conclusions: Androgen regulation of examined genes was confirmed in VCaP prostate cancer cell culture model. Assessment of AR, ERG, NKX3.1 and PSA protein levels by IHC indicated reduced expression of AR-regulated genes in a subset of the cases. The analysis revealed inter- and intratumoral heterogeneity, as well as, challenges in the quantitative assessment of IHC. The observed readouts of AR dysfunction may provide a new tool for improved prognostic accuracy and patient stratification at early stages of prostate cancer treatment.

Citation Format: Albert Dobi, Denise Young, Wei Huang, Lakshmi Ravindranth, Shashwat Sharad, Hua Li, Gyorgy Petrovics, David G. McLeod, Isabell Sesterhenn, Shiv Srivastava. Evaluation of androgen receptor function index (ARFI) in prostate cancer. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 1206. doi:10.1158/1538-7445.AM2013-1206

Abstract 1920: Variation of ERG oncoprotein expression in prostate cancer patients with different ethnicities

Introduction and Objectives: ERG gene fusions leading to ERG overexpression represent a highly prevalent oncogenic alteration (50-70%) in prostate cancer (CaP) patients of Caucasian ethnicity. We and others noted lower ERG frequency in CaP of African Americans (AA) in comparison to Caucasian Americans (CA). There is no study of ERG alterations in CaP patients from India. This study focused on comparative analysis of the frequency of ERG oncoprotein expression in CaP patients with Asian (India, Japan), African (AA) and Caucasian (CA) ethnicity.

Materials and Methods: Coded archived FFPE tissue blocks from radical prostatectomy (RP) specimens of 51 patients from the Rajiv Gandhi Cancer Institute (RGCI), New Delhi, India, were analyzed for the presence of ERG oncoprotein by immunohistochemistry (IHC) using a highly specific ERG monoclonal antibody (ERG MAb, 9FY). From Japan, tissue microarray (TMA) derived from surgical specimens of 64 patients (160 cancer spots) with localized CaP was analyzed by the same anti-ERG MAb. A USA cohort of whole mount prostate specimens, minimizing sampling bias and allowing for assessment of ERG in multifocal tumor context, from 91 AA and 91 CA CaP patients (matched for age, Gleason score and pathologic stage) was analyzed by ERG IHC.

Results: Of thirty evaluable specimens from India, ERG positive tumors were present in 8 cases (27%). Among 64 patients from Japan, 16 patients (25%) showed at least one positive cancer spot for ERG expression in the TMA. No correlation between ERG status and Gleason score or PSA failure was found. Systematic evaluation of whole mount prostatectomy specimens revealed that the frequency of ERG positive prostate tumors was significantly greater among CA than AA men when assessed in all tumor foci (41.9% vs. 23.9%, p&lt;0.0001) or in index tumors (63.3% vs. 28.6%, p&lt;0.0001).

Conclusions: Frequency of ERG oncoprotein expression is much lower in CaP patients with Asian (25-27%) or African (28%) descent than in CaP patients with European descent (63%). This study underscores that typing of prostate tumors for ERG may enhance our understanding of biological differences between ethnic groups.

This research in part was supported by the National Cancer Institute R01CA162383 (S. S.) and USU-CPDR funds.

Citation Format: Gyorgy Petrovics, Sudhir Rawal, Molly Williams, Philip Rosen, David Pfister, Takahiro Inoue, Osamu Ogawa, Monica Colombo, Yongmei Chen, Denise Young, Albert Dobi, David G. McLeod, Isabell Sesterhenn, Shiv Srivastava. Variation of ERG oncoprotein expression in prostate cancer patients with different ethnicities. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 1920. doi:10.1158/1538-7445.AM2013-1920

Abstract 2614: Distinguished features of ERG oncoprotein expression among matched cohorts of African-American and Caucasian-American prostate cancer patients

Introduction and Objective: The ETS-related gene (ERG) proto-oncogene is frequently overexpressed in prostate cancer (CaP) as a result of a genomic rearrangement that places ERG under the control of the androgen-dependent TMPRSS2 promoter. The initial report of quantitative ERG mRNA expression in micro-dissected prostate tumor cells showed significantly higher ERG expression in CaP of Caucasian-Americans (CA) patients vs. African-American (AA) patients (Petrovics et al, Oncogene, 2005). Recent studies (Magi-Galluzzi et al, Prostate, 2011; Elliott et al, USCAP Mtg., 2011) have also shown higher frequency of ERG rearrangement or ERG oncoprotein expression in CaP of CA patients. We evaluate the frequency and pattern of the ERG oncoprotein expression in prostate tumors of matched CA and AA CaP patients to better understand the biological basis for differences in prostate cancer between the two populations. Methods: Ninety one AA and 91 CA CaP patients were matched for age, Gleason score and pathologic stage. All underwent radical prostatectomy between 1993 and 2010. Whole mount prostate specimens were used for the immunohistochemical detection of the ERG oncoprotein by a highly specific ERG monoclonal antibody (clone 9FY). Individual foci of tumors were reported as either positive or negative. ERG staining data was linked to clinico-pathologic data. Biochemical recurrence was defined as two serum PSA measurements of 0.2 ng/mL or higher at least 8 weeks from surgery. Results: A higher percentage of CA (61.8%) than AA (28.2%) CaP patients had ERG positive index tumors and had at least one focus of tumor positive for ERG. A higher percentage of overall tumor foci were also positive for ERG in CA than AA CaP patients. There was high concordance of ERG positive prostatic intra-epitheilal neoplasia (PIN) and ERG positive CaP in both AA and CA patients. Conclusions: ERG expression is more prevalent among CA than AA CaP patients in matched cohorts. Differences in the pattern of ERG expression in CaP and differing trends in biochemical recurrence between CA and AA patients with ERG (+) index tumors suggest a dominant clonal selection of ERG-positive tumors in CA patients. These differences in ERG expression have the potential to delineate biological distinctions of CaP in the two patient populations. Acknowledgement: This study was supported in part by USU-CPDR and NIH grant CA162383 to SS.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2614. doi:1538-7445.AM2012-2614

Comparison of ERG oncoprotein expression among matched cohorts of African American and Caucasian American patients with prostate cancer

52

Background: The ERG proto-oncogene is frequently overexpressed in Prostate Cancer (CaP) following genomic fusion, which places ERG under the control of the TMPRSS2 promoter. Recent studies show higher frequency of ERG rearrangement or ERG oncoprotein expression in CaP of Caucasian-American (CA) patients compared to African-American (AA) patients. The frequency of the ERG oncoprotein expression in prostate tumors of matched CA and AA CaP patients is evaluated to better understand the biological basis for differences in prostate cancer between the two populations.

Methods: Ninety-one AA CaP patients were matched for age, Gleason sum and pathologic stage to 91 CA CaP patients. All underwent radical prostatectomy between 1993 and 2010. Whole mount prostate specimens were used for immunohistochemical detection of the ERG oncoprotein by an ERG monoclonal antibody (clone 9FY). Individual foci of tumors were noted as either positive or negative. ERG staining data was linked to dinico-pathologic data.

Results: Fifty-nine of 91 CA patients (64.8%) and 41191 (45.0%) AA patients had at least one focus of tumor positive for ERG in a whole-mount prostate section (p = 0.0073). The index tumor was positive for ERG expression in 55/89 (61.8%) CA and 24/85 (28.2%) of AA patients (p <0.001). Seventy-seven of 183 (42.1%) individual foci were positive in CA versus 511196 in AA (26.0%; P <0.001). There was a trend towards increased biochemical recurrence in CA patients with ERG positive index tumors.

Conclusions: The prevalence of ERG oncoprotein expression is significantly higher among Caucasian- American patients. Differences in pattern of ERG positivity of index and non-index tumors between Caucasian-American and African-American prostate cancer patients and trends towards decreased biochemical recurrence-free survival suggest a more accelerated growth of ERG-positive tumors in Caucasian-American patients. Differences in ERG expression have the potential to delineate biological distinctions of CaP in these patient populations.

Abstract A27: Comprehensive analysis of the TMPRSS2-ERG translocation during prostate cancer progression

Introduction: Approximately half of the diagnosed prostate carcinomas (PC) are characterized by a chromosomal rearrangement fusing the androgen regulated gene TMPRSS2 to the oncogenic ETS transcription factor ERG. Aim of this study was to comprehensively analyze the impact of this translocation on the expression of the ERG gene in hormone-naïve (untreated) and castration-resistant prostate cancers and to define the influence of AR protein expression and genomic amplification in this context.

Methods: We constructed a tissue microarray (TMA) containing 915 tissue cores from 107 hormone-naïve PC and from 101 castration-resistant (CR) PC. In addition, we included 56 specimens from distant metastases. We analyzed the TMPRSS2-ERG translocation status by fluorescence in-situ hybridization and the expression profiles of ERG, AR and the proliferation marker Ki67 by immunohistochemistry.

Results: Nearly half of the analyzed PC tissue specimens (untreated: 38%, castration-resistant: 46%) harbored a TMPRSS2-ERG gene fusion. Untreated PC with positive translocation status showed increased tumor cell proliferation (p &lt; 0.05). As expected, TMPRSS2-ERG gene fusion was strongly associated with increased ERG protein expression in untreated, as well as in castration-resistant PC (both p &lt; 0.0001). However, we detected a subgroup (26%) of CR PCs with the gene fusion, but without detectable ERG protein expression. This subgroup showed significantly lower levels of AR protein expression and of the androgen regulated serum PSA (both p&gt;0.05).

Conclusions: This comprehensive study comparing the TMPRSS2-ERG gene and ERG protein expression status in a large cohort of hormone-naïve and CR PC, identifies a subgroup of translocated CR PCs without ERG protein expression. Our results suggest that this subgroup may represent CR PCs with a dispensed AR pathway. Further analyses may help to explain if these tumors represent a distinct subgroup of AR independent CR PCs.

Citation Format: Shyh-Han Tan, I. Sesterhenn, S. Srivastava, L. Bubendorf, C. Ruiz, M. Braun, V.J. Scheble, T. Zellweger, C.A. Rentsch, A. Bachmann, Albert Dobi, S. Perner. Comprehensive analysis of the TMPRSS2-ERG translocation during prostate cancer progression [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 A27.

Abstract A29: Distinguishing features of ERG oncoprotein expression among matched cohorts of African-American and Caucasian-American prostate cancer patients

Introduction and Objective: The ETS-related gene (ERG) proto-oncogene is frequently overexpressed in prostate cancer (CaP) as a result of a genomic rearrangement that places ERG under the control of the androgendependent TMPRSS2 promoter. The initial report of quantitative ERG mRNA expression in micro-dissected prostate tumor cells showed significantly higher ERG expression in CaP of Caucasian-Americans (CA) patients vs. African-American (AA) patients (Petrovics et al, Oncogene, 2005). Recent studies (Magi-Galluzzi et al, Prostate, 2011; Elliott et al, USCAP Mtg., 2011) have also shown higher frequency of ERG rearrangement or ERG oncoprotein expression in CaP of CA patients. We evaluate the frequency and pattern of the ERG oncoprotein expression in prostate tumors of matched CA and AA CaP patients to better understand the biological basis for differences in prostate cancer between the two populations.

Methods: Ninety one AA and 91 CA CaP patients were matched for age, Gleason score and pathologic stage. All underwent radical prostatectomy between 1993 and 2010. Whole mount prostate specimens were used for the immunohistochemical detection of the ERG oncoprotein by a highly specific ERG monoclonal antibody (clone 9FY). Individual foci of tumors were reported as either positive or negative. ERG staining data was linked to clinico-pathologic data. Biochemical recurrence was defined as two serum PSA measurements of 0.2 ng/mL or higher at least 8 weeks from surgery.

Results: A higher percentage of CA (61.8%) than AA (28.2%) CaP patients had ERG positive index tumors and had at least one focus of tumor positive for ERG. A higher percentage of overall tumor foci were also positive for ERG in CA than AA CaP patients. There was high concordance of ERG positive prostatic intra-epitheilal neoplasia (PIN) and ERG positive CaP in both AA and CA patients.

Conclusions: ERG expression is more prevalent among CA than AA CaP patients in matched cohorts. Differences in the pattern of ERG expression in CaP and differing trends in biochemical recurrence between CA and AA patients with ERG (+) index tumors suggest a dominant clonal selection of ERG-positive tumors in CA patients. These differences in ERG expression have the potential to delineate biological distinctions of CaP in the two patient populations.

Citation Format: Gyorgy Petrovics, Philip Rosen, David Pfister, Denise Young, Yongmei Chen, Albert Dobi, David McLeod, Shiv Srivastava, Isabell Sesterhenn. Distinguishing features of ERG oncoprotein expression among matched cohorts of African-American and Caucasian-American prostate cancer patients [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 A29.