Swedish Watchful Waiting Cohort Research Articles

Gene set analysis and reduction for a continuous phenotype: Identifying markers of birth weight variation based on embryonic stem cells and immunologic signatures

BackgroundGene set analysis is a popular approach to examine the association between a predefined gene set and a phenotype. Few methods have been developed for a continuous phenotype. However, often not all the genes within a significant gene set contribute to its significance. There is no gene set reduction method developed for continuous phenotype. We developed a computationally efficient analytical tool, called linear combination test for gene set reduction (LCT-GSR) to identify core subsets of gene sets associated with a continuous phenotype. Identifying the core subset enhances our understanding of the biological mechanism and reduces costs of disease risk assessment, diagnosis and treatment. ResultsWe evaluated the performance of our analytical tool by applying it to two real microarray studies. In the first application, we analyzed pathway expression measurements in newborns’ blood to discover core genes contributing to the variation in birth weight. On average, we were able to reduce the number of genes in the 33 significant gene sets of embryonic stem cell signatures by 84.3% resulting in 229 unique genes. Using immunologic signatures, on average we reduced the number of genes in the 210 significant gene sets by 89% leading to 1603 unique genes. There were 180 unique core genes overlapping across the two databases. In the second application, we analyzed pathway expression measurements in a cohort of lethal prostate cancer patients from Swedish Watchful Waiting cohort to identify main genes associated with tumor volume. On average, we were able to reduce the number of genes in the 17 gene sets by 90% resulting in 47 unique genes. ConclusionsWe conclude that LCT-GSR is a statistically sound analytical tool that can be used to extract core genes associated with a continuous phenotype. It can be applied to a wide range of studies in which dichotomizing the continuous phenotype is neither easy nor meaningful. Reduction to the most predictive genes is crucial in advancing our understanding of issues such as disease prevention, faster and more efficient diagnosis, intervention strategies and personalized medicine.

Abstract B33: Alterations in the binding of Meis proteins to HoxB13 or to DNA promote prostate cancer progression

Abstract Introduction: One of the first germline mutations shown to confer an increased risk of early-onset prostate cancer (PrCa) was recently found in the Meis-binding domain of Homeobox B13 (HoxB13). Neither the function of this mutation nor the role of Meis proteins in PrCa has been elucidated. There are three Meis genes, each with variously spliced isoforms, some of which lack a DNA binding domain. Meis proteins normally function as Hox gene cofactors, and regulate proliferation and cell fate. In normal and other disease contexts, Meis proteins with an abrogated homeodomain have been shown to act as a dominant negative and squelch normal Hox gene transcription. The androgen receptor (AR), the major oncogene in PrCa, has been shown to interact with both Meis and Hox proteins. We hypothesize that Meis proteins function as AR-regulated tumor suppressors in advanced prostate cancer, and isoforms lacking the homeodomain prevent HoxB13 DNA binding to promote uncontrolled cell proliferation. Methods: Meis1/2 expression stratified outcomes of men with Gleason 6/7 tumors in a retrospective bioinformatics analysis of microarray data obtained from the Swedish Watchful Waiting cohort. Meis1/2/3 mRNA expression in normal and cancerous prostate cell lines as well as adjacent normal stromal and epithelial patient-derive cell lines was compared by qRT-PCR. Cell number over time was determined by SRB assay for lines stably overexpressing full length or homeodomains-less Meis2 as compared to GFP in castration-sensitive LNCaP and castration-resistant CWR-22Rv1 PrCa cells. Meis2 expression was analyzed by qPCR in androgen-dependent LAPC4 cells. LAPC4 cells were incubated in distinct androgen conditions, including incubation with an AR agonist (R1881) or potent antagonist (enzalutamide). Publicly available microarray data from 19 patients were obtained for analysis from the Gene Expression Omnibus. Results: Through retrospective bioinformatics analysis, we found that men diagnosed with intermediate Gleason scores and low Meis expression had a worse prognosis than men with high Meis levels. qPCR data indicated lower Meis expression in PrCa cell lines as compared to normal prostate epithelial cells. Stromal patient-derived cell lines had significantly higher levels of Meis1 and 2 than their patient-matched epithelial counterparts, while Meis3 expression was consistent across all cell types. Proliferation assays indicated CWR-22Rv1 cells that overexpress full length Meis2 showed significantly lower cell number over five days than GFP- or abrogated homeodomain Meis-expressing cells. qPCR data indicates that PrCa cells exposed to R1881 had significantly lower Meis2 expression as compared to cells treated with enzalutamide. Analysis of publicly available microarrays showed that patients with metastatic prostate cancer were more likely to have lower Meis2 expression than patients with localized or benign disease. Conclusions: Our data indicate that full length Meis protein levels are consistently lower in cancer as compared to normal tissues, and are decreased to a greater extent in more aggressive cancers. Full length Meis proteins may act as tumor suppressors while the homeodomain-less isoforms may act as dominant negatives. Further, we demonstrate a regulatory interaction between Meis expression and AR signaling, providing evidence for a downstream role of Meis proteins in the most important pathway for PrCa progression. We are the first to investigate the functional differences of these isoforms in the context of PrCa. An understanding of how the functional domains of Meis and HoxB13 interact with each other, the AR and the genome are crucial steps towards evaluating the utility of Meis expression as a biomarker to stratify aggressive versus indolent PrCa patients, as well as aid our general understanding of prostate tumor initiation and progression. Citation Format: Hannah Brechka, Raj Bhanvadia, Donald Vander Griend. Alterations in the binding of Meis proteins to HoxB13 or to DNA promote prostate cancer progression. [abstract]. In: Proceedings of the AACR Special Conference: Developmental Biology and Cancer; Nov 30-Dec 3, 2015; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Res 2016;14(4_Suppl):Abstract nr B33.

Intracellular location of BRCA2 protein expression and prostate cancer progression in the Swedish Watchful Waiting Cohort.

Prostate cancer patients with inherited BRCA2 mutations have a survival disadvantage. However, it is unknown whether progression is associated with BRCA2 protein expression in diagnostic prostate cancer tissue, among men without inherited mutations. We conducted a nested case-control study within the Swedish Watchful Waiting cohort. The case group included all 71 patients who died from prostate cancer within 5 years from diagnosis and controls were all patients (n = 165) who lived at least 7 years after diagnosis. Tissue microarrays were stained using antibodies for C- and N-terminal domains of the BRCA2 protein. Location (nuclear, cytoplasmic and membranous) and magnitude (intensity and percentage) of expression were assessed. Logistic regression models produced odds ratios (OR) and 95% confidence intervals (CI) adjusted for age, year of diagnosis and Gleason score. Positive BRCA2 staining at the cell membrane was associated with reduced risk of death within 5 years (N-terminal: OR = 0.47, 95% CI = 0.21-1.04, P = 0.06; C-terminal: OR = 0.41, 95% CI = 0.18-0.91, P = 0.03) and low Gleason scores (P = 0.006). Positive cytoplasmic C-terminal staining was associated with higher Gleason scores and increased lethality (OR = 3.61, 95% CI = 1.61-8.07, P = 0.002). BRCA2 protein expression at the cell membrane and lack of C-terminal expression in the cytoplasm were associated with a reduced risk of rapidly fatal prostate cancer. BRCA2 protein expression in prostate cancer tissue may have independent prognostic value. The potential biological significance of BRCA2 expression at the cell membrane warrants further investigation.

Abstract 4346: Prognostic gene signature for intermediate risk prostate cancer

Abstract The over treatment of prostate cancer patients is a significant concern, as recent clinical trials suggest that many patients are over treated which can lead to significant patient morbidity. Although the Gleason score is a powerful predictor of lethal or indolent disease, a significant proportion (∼40%) of men present with early stage Gleason score (GS) 7 tumors, for whom prognosis is variable. The goal of this study is to develop and optimize a robust prognostic gene signature that can be utilized on formalin fixed paraffin embedded (FFPE) core biopsy tumor material to better classify patients with intermediate risk, GS 7 tumors into good and poor outcome groups. Three gene signatures were derived from publicly available gene expression profiles of the Swedish Watchful Waiting cohort. The Genomic Grade Index consisted of the top 25 molecular signatures discriminating between high (8, 9 & 10) and low (≤ 6) GS tumors. The Lethal Gene Score consisted of the top 25 molecular signatures discriminating between lethal and indolent disease within GS 7 tumors only. A network-based gene signature consisted of 88 genes which accurately stratified GS 7 patients into high risk and low risk groups, resembling the survival curves of high GS and low GS patients. The prognostic capacity of the combined gene signature was tested in silico on the gene expression profiles of the Mayo cohort. Results demonstrated the gene signature's highly robust capacity for differentiating low risk and high risk patients within GS 7 patients. The NanoString nCounter System will be used to quantify mRNA from prostate FFPE blocks to assess the expression of the 138 prognostic genes. 156 archived prostate tumor blocks will be collected from intermediate risk, GS 7 patients enrolled in the 2005 PR5 prostate trial, which also collected 12 years of clinical follow-up information. Results will be correlated with biochemical (PSA) failure rates and overall survival. In short, our findings provide proof-of-principle that through the use of gene signatures it is possible to separate prostate cancer patients of intermediate risk into good and poor outcome groups. Furthermore, they also identify multiple gene candidates whose expression could likely be formulated into a clinically applicable assay, the implementation of which could serve to stratify prostate cancer patients with tumors of intermediate risk into more accurate high and low risk groups. Citation Format: Brian Li, Robin Hallet, Ying Wu, Greg Pong, John Hassell, Sebastien Hotte, Mark Levine, Himansu Lukka, Anita Bane. Prognostic gene signature for intermediate risk prostate cancer. [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 4346. doi:10.1158/1538-7445.AM2015-4346

Discovery and validation of a 30-gene expression signature to identify prostate cancer patients who are candidates for active surveillance.

10 Background: Considerable attention has been devoted to the search for biomarkers of aggressive prostate cancer. While many promising markers have been proposed, it is often unclear whether their ability to risk classify is adequate to reduce overtreatment of indolent cancers. We present and validate a gene expression signature that is highly specific for indolent disease and that, when combined with Gleason, improves upon the prognostic power of Gleason alone. Methods: A 30-gene signature of indolent disease was derived from a case-control sample of men (n=254) from the Health Professionals Follow-Up Study (HPFS) who were followed prospectively from cancer diagnosis for a median of 13.7 years. Cases were defined as men with prostate cancer who experienced a metastatic event or died of cancer during follow-up, while indolent controls survived at least 8 years without metastases. Whole-transcriptome gene expression was quantified from archival formalin-fixed, paraffin-embedded (FFPE) tumor tissue specimens acquired at prostatectomy. Genes were selected by a novel analytic strategy that maximizes a partial area under the curve (pAUC) to accurately identify indolent tumors. We validated the signature in two independent cohorts: the Physicians’ Health Study (PHS; n=150) and a Swedish Watchful Waiting cohort (WW; n=253) with respective median follow-up times of 14.4 and 9.0 years. Results: When compared to a model with Gleason alone, application of the signature to both validation data sets significantly improved prognostic accuracy as measured by pAUC (p=0.003 in PHS and p<0.001 in WW). Performance was particularly strong among men diagnosed with Gleason 7, where unit standard deviation increases in the signature score were associated with odds ratios of indolence of 5.24 (95% CI 2.21-15.75; p<0.001) and 2.37 (95% CI 1.45-4.19; p=0.001) in the PHS and WW cohorts, respectively. Conclusions: We present a signature of indolent prostate cancer that adds prognostic information beyond Gleason and has the potential to guide men to active surveillance and avoid overtreatment. Validation across both treated and untreated cohorts supports the discovery of a robust signature.

The prognostic significance of combined ERG and androgen receptor expression in patients with prostate cancer managed by androgen deprivation therapy

ERG and androgen receptor (AR) are known to function cooperatively in prostate cancer (PCa) progression. However, the prognostic value of combined ERG and AR expression and potential pathways are not well characterized. We assessed ERG and AR protein expression by immunohistochemistry in a cohort of 312 men with PCa diagnosed by transurethral resection of the prostate (TURP). Patients were divided into those with no prior hormonal treatment (designated as PCa/AdvPCa) vs. those with castrate-resistant PCa (CRPC) undergoing channel TURP to relieve obstructive symptoms. The expression status was correlated with various clinical-pathological parameters. The Swedish watchful-waiting cohort was used for validation and characterization of potential gene signatures associated with ERG and AR. Patients with combined ERG-positive/AR high expression profile demonstrated higher rates of PCa-specific mortality (PCSM) compared with patients with ERG-negative/AR low in patients with no prior treatment (n = 90, P = 0.032), but this was attenuated in the overall cohort which included the CRPC subgroup (n = 125, P = 0.096). The prognostic significance to PCSM was validated in the Swedish watchful waiting cohort in univariate (HR: 3.3; 95% CI: 1.9–5.6, P = 4.25E−5) and multivariate analysis (HR: 2; 95% CI: 0.97–4.1, P = 0.057), which included Gleason score. ERG/AR overexpression status characterized 152 genes signatures including WNT, PI3K/AKT and chemokine signaling pathways known to be deregulated in PCa. In conclusion, combined ERG/AR overexpression signifies a class of patients at highest-risk of PCSM with specific key genetic alteration likely responsible for disease progression. The prognostic value of combined ERG/AR overexpression and its associated genes should be further investigated as potential prognostic and therapeutic targets in prostate cancer progression.

MP41-06 SPOP MUTATIONS IN PROSTATE CANCER ACROSS DEMOGRAPHICALLY DIVERSE PATIENT COHORTS.

BACKGROUND: Recurrent mutations in the Speckle-Type POZ Protein (SPOP) gene occur in up to 15% of prostate cancers. However, the frequency and features of cancers with these mutations across different populations is unknown. OBJECTIVE: To investigate SPOP mutations across diverse cohorts and validate a series of assays employing high-resolution melting (HRM) analysis and Sanger sequencing for mutational analysis of formalin-fixed paraffin-embedded material. DESIGN, SETTING, AND PARTICIPANTS: 720 prostate cancer samples from six international cohorts spanning Caucasian, African American, and Asian patients, including both prostate-specific antigen–screened and unscreened populations, were screened for their SPOP mutation status. Status of SPOP was correlated to molecular features (ERG rearrangement, PTEN deletion, and CHD1 deletion) as well as clinical and pathologic features. RESULTS AND LIMITATIONS: Overall frequency of SPOP mutations was 8.1% (4.6% to 14.4%), SPOP mutation was inversely associated with ERG rearrangement (P < .01), and SPOP mutant (SPOPmut) cancers had higher rates of CHD1 deletions (P < .01). There were no significant differences in biochemical recurrence in SPOPmut cancers. Limitations of this study include missing mutational data due to sample quality and lack of power to identify a difference in clinical outcomes. CONCLUSION: SPOP is mutated in 4.6% to 14.4% of patients with prostate cancer across different ethnic and demographic backgrounds. There was no significant association between SPOP mutations with ethnicity, clinical, or pathologic parameters. Mutual exclusivity of SPOP mutation with ERG rearrangement as well as a high association with CHD1 deletion reinforces SPOP mutation as defining a distinct molecular subclass of prostate cancer. Neoplasia (2014) 16, 14–20 Introduction Prostate cancer is a significant health burden, with 241,740 new diagnoses and 28,170 deaths in the United States in 2012 [1]. The inability to distinguish indolent from aggressive disease is a challenge [2]. Identification of driver lesions for specific subsets of prostate cancer could ultimately lead to the development of biomarkers to improve prognostic ability and risk stratification. Recent advances have uncovered multiple recurrent alterations in prostate cancer. The TMPRSS2-ERG fusion has been observed in nearly 50% of prostate cancers [3,4]. We recently reported somatic mutations in the Speckle-Type POZ Protein (SPOP) gene in 6% to 15% of prostate cancers [5]. SPOPmutations define a distinct subclass of prostate cancer: SPOP mutations and ETS rearrangements are mutually exclusive; SPOP mutant (SPOPmut) prostate tumors generally lack lesions in the phosphatidylinositide 3-kinase (PI3K) pathway, and they are also independent of mutations in the tumor suppressor gene TP53 [5–7]. Although the recurrent nature of SPOP mutations is clear, less is known about the frequency of SPOP mutations across different ethnicities and screening practices, the associations with clinical and pathologic characteristics, and the effects on patient outcomes. This study represents the largest multi-institutional study to date investigating these associations with SPOPmutations across different cohorts. In addition, detection of SPOP mutations in older well-annotated archival prostate cancer samples represents a technical challenge. Formalin fixation of tissue followed by paraffin embedding (FFPE) is widely used; however, analysis of nucleic acid from FFPE material is difficult due to cross-linking between nucleic acid and proteins [8]. The tissue heterogeneity in prostate cancer samples can dilute the signal of tumor-associated mutations with benign wild-type contamination and make the detection of point mutations difficult. In an effort to overcome these challenges, we developed a series of assays employing high-resolution melting (HRM) analysis, which relies on alterations in the melting curve of mutated nucleic acids, and Sanger sequencing [9–11]. We optimized the HRM assay as a high sensitivity pre-screening tool, followed by Sanger sequencing for specific confirmation of mutations. Finally, we employed a nextgeneration sequencing approach on a small subset of samples to determine if massively parallel sequencing could rescue samples considered assay failures using the HRM-Sanger methods. Materials and Methods Patient Populations Table 1 lists the clinical, pathologic, and survival data according to each cohort. In total, prostate cancer samples from 996 patients [radical prostatectomy (RP), transurethral resection of the prostate (TURP), or metastatic biopsies] were examined. Of the 996 patients, 720 samples had DNA of sufficient quality and were screened for their SPOP status. Cohorts included patients from Memorial SloanKettering Cancer Center (MSKCC), cohort from Kyungpook National University School of Medicine, Korea (Korean), African American cohort from New York-Presbyterian Hospital (AA), Weill Cornell Medical College (WCMC) cohorts, and University Hospital of Zurich (USZ), as well as the Swedish watchful waiting cohort (SWWC). A detailed description about each individual cohort can be found in the Supplementary Materials and Methods. Samples were categorized into SPOP wild type (SPOPwt) and SPOPmut and analyzed for correlation Neoplasia Vol. 16, No. 1, 2014 SPOP Mutations in Prostate Cancer Blattner et al. 15

Seasonal variation in expression of markers in the vitamin D pathway in prostate tissue

Recent studies suggest variation in genes along the vitamin D pathway, as well as vitamin D receptor (VDR) protein levels, may be associated with prostate cancer. As serum vitamin D levels vary by season, we sought to determine whether the expression of genes on the vitamin D pathway, assessed in prostate tumor tissue, do the same. Our study incorporates mRNA expression data from 362 men in the Swedish Watchful Waiting cohort, diagnosed between 1977 and 1999, and 106 men enrolled in the US Physicians' Health Study (PHS) diagnosed between 1983 and 2004. We also assayed for VDR protein expression among 832 men in the PHS and Health Professionals Follow-up Study cohorts. Season was characterized by date of initial tissue specimen collection categorically and by average monthly ultraviolet radiation levels. One-way analysis of variance was used to examine variation in the expression levels of six genes on the vitamin D pathway-VDR, GC, CYP27A1, CYP27B1, RXRα, CYP24A1-and VDR protein by season, adjusted for age at diagnosis and Gleason grade. Variation was also examined separately among lethal and nonlethal cases. Tumor expression levels of the six genes did not vary significantly by season of tissue collection. No consistent patterns emerged from subgroup analyses by lethal versus nonlethal cases. Unlike circulating levels of 25(OH) vitamin D, expression levels of genes on the vitamin D pathway and VDR protein did not vary overall by season of tissue collection. Epidemiological analyses of vitamin D gene expression may not be biased by seasonality.

Molecular classification of prostate cancer using curated expression signatures

High Gleason score is currently the best prognostic indicator for poor prognosis in prostate cancer. However, a significant number of patients with low Gleason scores develop aggressive disease as well. In an effort to understand molecular signatures associated with poor outcome in prostate cancer, we analyzed a microarray dataset characterizing 281 prostate cancers from a Swedish watchful-waiting cohort. Patients were classified on the basis of their mRNA microarray signature profiles indicating embryonic stem cell expression patterns (stemness), inactivation of the tumor suppressors p53 and PTEN, activation of several oncogenic pathways, and the TMPRSS2-ERG fusion. Unsupervised clustering identified a subset of tumors manifesting stem-like signatures together with p53 and PTEN inactivation, which had very poor survival outcome, a second group with intermediate survival outcome, characterized by the TMPRSS2-ERG fusion, and three groups with benign outcome. The stratification was validated on a second independent dataset of 150 tumor and metastatic samples from a clinical cohort at Memorial Sloan-Kettering Cancer Center. This classification is independent of Gleason score and therefore provides useful unique molecular profiles for prostate cancer prognosis, helping to predict poor outcome in patients with low or average Gleason scores.

A miRNA expression signature that separates between normal and malignant prostate tissues

BackgroundMicroRNAs (miRNAs) constitute a class of small non-coding RNAs that post-transcriptionally regulate genes involved in several key biological processes and thus are involved in various diseases, including cancer. In this study we aimed to identify a miRNA expression signature that could be used to separate between normal and malignant prostate tissues.ResultsNine miRNAs were found to be differentially expressed (p <0.00001). With the exception of two samples, this expression signature could be used to separate between the normal and malignant tissues. A cross-validation procedure confirmed the generality of this expression signature. We also identified 16 miRNAs that possibly could be used as a complement to current methods for grading of prostate tumor tissues.ConclusionsWe found an expression signature based on nine differentially expressed miRNAs that with high accuracy (85%) could classify the normal and malignant prostate tissues in patients from the Swedish Watchful Waiting cohort. The results show that there are significant differences in miRNA expression between normal and malignant prostate tissue, indicating that these small RNA molecules might be important in the biogenesis of prostate cancer and potentially useful for clinical diagnosis of the disease.

MRNA Expression Signature of Gleason Grade Predicts Lethal Prostate Cancer

Prostate-specific antigen screening has led to enormous overtreatment of prostate cancer because of the inability to distinguish potentially lethal disease at diagnosis. We reasoned that by identifying an mRNA signature of Gleason grade, the best predictor of prognosis, we could improve prediction of lethal disease among men with moderate Gleason 7 tumors, the most common grade, and the most indeterminate in terms of prognosis. Using the complementary DNA-mediated annealing, selection, extension, and ligation assay, we measured the mRNA expression of 6,100 genes in prostate tumor tissue in the Swedish Watchful Waiting cohort (n = 358) and Physicians' Health Study (PHS; n = 109). We developed an mRNA signature of Gleason grade comparing individuals with Gleason ≤ 6 to those with Gleason ≥ 8 tumors and applied the model among patients with Gleason 7 to discriminate lethal cases. We built a 157-gene signature using the Swedish data that predicted Gleason with low misclassification (area under the curve [AUC] = 0.91); when this signature was tested in the PHS, the discriminatory ability remained high (AUC = 0.94). In men with Gleason 7 tumors, who were excluded from the model building, the signature significantly improved the prediction of lethal disease beyond knowing whether the Gleason score was 4 + 3 or 3 + 4 (P = .006). Our expression signature and the genes identified may improve our understanding of the de-differentiation process of prostate tumors. Additionally, the signature may have clinical applications among men with Gleason 7, by further estimating their risk of lethal prostate cancer and thereby guiding therapy decisions to improve outcomes and reduce overtreatment.

Validation of suitable endogenous control genes for expression studies of miRNA in prostate cancer tissues

When performing quantitative polymerase chain reaction analysis, there is a need for correction of technical variation between experiments. This correction is most commonly performed by using endogenous control genes, which are stably expressed across samples, as reference genes for normal expression in a specific tissue. In microRNA (miRNA) studies, two types of control genes are commonly used: small nuclear RNAs and small nucleolar RNAs. In this study, six different endogenous control genes for miRNA studies were investigated in prostate tissue material from the Swedish Watchful Waiting cohort. The stability of the controls was investigated using two different software applications, NormFinder and BestKeeper. RNU24 was the most suitable endogenous control gene for miRNA studies in prostate tissue materials.

Molecular sampling of prostate cancer: a dilemma for predicting disease progression

BackgroundCurrent prostate cancer prognostic models are based on pre-treatment prostate specific antigen (PSA) levels, biopsy Gleason score, and clinical staging but in practice are inadequate to accurately predict disease progression. Hence, we sought to develop a molecular panel for prostate cancer progression by reasoning that molecular profiles might further improve current clinical models.MethodsWe analyzed a Swedish Watchful Waiting cohort with up to 30 years of clinical follow up using a novel method for gene expression profiling. This cDNA-mediated annealing, selection, ligation, and extension (DASL) method enabled the use of formalin-fixed paraffin-embedded transurethral resection of prostate (TURP) samples taken at the time of the initial diagnosis. We determined the expression profiles of 6100 genes for 281 men divided in two extreme groups: men who died of prostate cancer and men who survived more than 10 years without metastases (lethals and indolents, respectively). Several statistical and machine learning models using clinical and molecular features were evaluated for their ability to distinguish lethal from indolent cases.ResultsSurprisingly, none of the predictive models using molecular profiles significantly improved over models using clinical variables only. Additional computational analysis confirmed that molecular heterogeneity within both the lethal and indolent classes is widespread in prostate cancer as compared to other types of tumors.ConclusionsThe determination of the molecularly dominant tumor nodule may be limited by sampling at time of initial diagnosis, may not be present at time of initial diagnosis, or may occur as the disease progresses making the development of molecular biomarkers for prostate cancer progression challenging.

Estrogen-Dependent Signaling in a Molecularly Distinct Subclass of Aggressive Prostate Cancer

The majority of prostate cancers harbor gene fusions of the 5'-untranslated region of the androgen-regulated transmembrane protease serine 2 (TMPRSS2) promoter with erythroblast transformation-specific transcription factor family members. The common fusion between TMPRESS2 and v-ets erythroblastosis virus E26 oncogene homolog (avian) (ERG) is associated with a more aggressive clinical phenotype, implying the existence of a distinct subclass of prostate cancer defined by this fusion. We used complementary DNA-mediated annealing, selection, ligation, and extension to determine the expression profiles of 6144 transcriptionally informative genes in archived biopsy samples from 455 prostate cancer patients in the Swedish Watchful Waiting cohort (1987-1999) and the United States-based Physicians(') Health Study cohort (1983-2003). A gene expression signature for prostate cancers with the TMPRSS2-ERG fusion was determined using partitioning and classification models and used in computational functional analysis. Cell proliferation and TMPRSS2-ERG expression in androgen receptor-negative (NCI-H660) prostate cancer cells after treatment with vehicle or estrogenic compounds were assessed by viability assays and quantitative polymerase chain reaction, respectively. All statistical tests were two-sided. We identified an 87-gene expression signature that distinguishes TMPRSS2-ERG fusion prostate cancer as a discrete molecular entity (area under the curve = 0.80, 95% confidence interval [CI] = 0.792 to 0.81; P < .001). Computational analysis suggested that this fusion signature was associated with estrogen receptor (ER) signaling. Viability of NCI-H660 cells decreased after treatment with estrogen (viability normalized to day 0, estrogen vs vehicle at day 8, mean = 2.04 vs 3.40, difference = 1.36, 95% CI = 1.12 to 1.62) or ERbeta agonist (ERbeta agonist vs vehicle at day 8, mean = 1.86 vs 3.40, difference = 1.54, 95% CI = 1.39 to 1.69) but increased after ERalpha agonist treatment (ERalpha agonist vs vehicle at day 8, mean = 4.36 vs 3.40, difference = 0.96, 95% CI = 0.68 to 1.23). Similarly, expression of TMPRSS2-ERG decreased after ERbeta agonist treatment (fold change over internal control, ERbeta agonist vs vehicle at 24 hours, NCI-H660, mean = 0.57- vs 1.0-fold, difference = 0.43-fold, 95% CI = 0.29- to 0.57-fold) and increased after ERalpha agonist treatment (ERalpha agonist vs vehicle at 24 hours, mean = 5.63- vs 1.0-fold, difference = 4.63-fold, 95% CI = 4.34- to 4.92-fold). TMPRSS2-ERG fusion prostate cancer is a distinct molecular subclass. TMPRSS2-ERG expression is regulated by a novel ER-dependent mechanism.