NCI DCP News Note

Research from the Early Detection Research Network on New Methods to Detect Prostate Cancer

Prostate cancer is the most frequently diagnosed non-skin cancer in men in the United States. In 2010 there were 218,000 men diagnosed with prostate cancer. The prevalence of the diagnosis makes the disease a major health burden. While the majority of the diagnosed men will survive the disease, about 15% will die from it, a rate that is affected by over-diagnosis and the consequent over-treatment. The major objectives of the EDRN Prostate and other Urological Cancers Collaborative Group are to identify, develop and validate biomarkers for cancer risk and earlier detection, especially of aggressive forms of prostate cancer. In addition, the group is developing biomarkers to assist clinicians in decision making for questions such as who needs primary and secondary biopsies, which may help to reduce the number of unnecessary procedures.

The Prostate and other Urological Cancers Collaborative Group has made considerable progress in addressing the challenging questions of early detection of prostate cancer and especially for identifying clinically significant disease.

Major Accomplishments

• Discovery of cancer-specific gene fusions (TMPRSS2-ETV1; TMPRSS2:ERG) for the first time in solid epithelial tumors - in prostate cancer and associated with aggressiveness of disease
• Completed validation of %[-2]proPSA to improve PSA-based detection of prostate cancer; IVD pending Food and Drug Administration (FDA) review
• Completed validation of PCA3 as a urine-based test for early detection and diagnosis of prostate cancer
• Developed molecular assay and FISH for detection of T2S:ERG fusion in urine and tissue, respectively. Both assays are offered by Clinical Laboratory Improvement Amendment (CLIA) labs
• Constructed standard reference sets for rapid triage and validation of biomarkers for prostate cancer
• Discovered sarcosine as a metabolite biomarker in tissue and urine
• Developed a Prostate Cancer Risk Calculator (http://www.prostate-cancer-risk-calculator.com)

Prostate Cancer - The Plan

To accomplish the strategic goals the collaborative group proposed the following plan:

-Develop biomarkers which can discriminate between indolent and aggressive prostate cancers based on a variety of “omics” approaches (genomics, epigenomics, proteomics and metabolomics);

-Develop urine-based and blood-based assays for all promising prostate cancer biomarkers;

-Conduct rigorous clinical evaluation of promising biomarkers for early detection and of aggressive prostate based on recently discovered cancer-specific fusion transcripts such as TMPRSS2-ERG, TMPRSS2-ETV1, etc;

-Perform a meta-analysis on the performance of all validated promising prostate cancer biomarkers to select the best markers for earlier detection of clinically significant prostate cancers;

-Evaluate reactive cancer stromal markers in tissue and urine samples; and

-Combine biomarkers with imaging modalities for better and earlier detection of aggressive prostate cancer.

Ongoing Projects

Investigators in the Prostate and Other Urological Cancers Collaborative Group conduct new biomarker discovery, pre-validation and validation projects by pursuing the EDRN strategic goals. Ongoing projects include:

• Researchers at the University of Texas Health Science Center San Antonio Clinical Validation Center (CVC) are conducting a validation study on gene polymorphisms, which are associated with critical pathways, such as enzymes involved in androgen metabolism, DNA repair and hereditary susceptibility genes. If validated, such markers will be added to the prostate cancer risk calculator (http://www.prostate-cancer-risk-calculator.com). EDRN investigators discovered that ~55% of all prostate cancers harbor genetic rearrangements such as fusion genes (e.g., TMPRSS2-ERG, SLC45A3-ETV1). Such cancer-specific markers provide an important opportunity to develop a non-invasive, highly specific diagnostic test for the early detection of prostate cancer.
• At the Beth Israel Deaconess Medical Center (BIDMC) CVC, researchers are comparing the accuracy of detection of TMPRSS2-ERG fusion in post-digital rectal exam (DRE) urine with the detection of the fusion from biopsy. The prevalence of TMPRSS2-ERG rearrangements will be also examined in a community-based study with emphasis on African Americans. Finally, a correlation between this genetic rearrangement and the development of aggressive cancer will be examined in a longitudinal study based on active surveillance of 300 men at BIDMC.
• PCA3 is a non-coding prostate-specific RNA that was reported to be frequently overexpressed in prostate tumor cells. EDRN investigators at the University of Michigan Biomarker Developmental Laboratory (BDL) are collaborating with Gen-Probe, Inc. on a multi-institutional validation study testing the utility of a PCA3 urine test to assist clinicians in decision making for biopsy or repeat biopsy. Interim analysis of data from this study shows great promise. In addition, a quantitative PCA3 urine test demonstrated potential as an adjunct to current methods for prostate cancer diagnosis.
• Epigenetic modification of DNA (particularly within the 5’ promoter region and the first exon) is a common alteration in cancer-related genes and is often associated with complete or partial repression of transcription. This is one of the mechanisms resulting in inactivation of tumor suppressor genes of cancers. Scientists at the Johns Hopkins University BDL developed a panel of promoter methylated genes which could serve as biomarkers of prostate cancer and high-grade prostatic intraepithelial neoplasia (HGPIN). The team developed an assay based on the percentage of methylated alleles and determined that these values are higher for the genes APC and RARß2 in HGPIN and carcinoma as compared to normal prostate tissue
• Using gene expression microarrays and selected reaction monitoring-mass spectrometry (SRM-MS) investigators at the Fred Hutchinson Cancer Research Center BDL identified stromal-based biomarkers. For example, CD90 (Thy-1) is a cell surface protein frequently over-expressed in prostate cancer stroma. Furthermore, increased levels of CD90 peptide fragments have been detected in all urine samples of prostate cancer patients tested so far. Another stomal marker, CD10, is frequently absent in the majority of prostate tumors. Gleason 4 and CD10+ cancer cells are associated with poor outcomes. The laboratory is also developing tests to detect AGRA2, an overexpressed protein in tumor tissue
• Validation of tissue expression of candidate biomarkers such as CCL3, CCL4, AGRA2, and others is being carried out using the intelligent prostate cancer tissue microarray TMA by the UCLA BRL. The University of California, Irvine BDL developed a classifier composed of 114 differentially expressed genes in reactive stromal cells adjacent to the cancer epithelial cells. The classifier was recently tested on 364 samples and showed promising results (sensitivity=98% and specificity= 88%). These results indicate that the prostate cancer microenvironment exhibits reproducible changes useful for categorizing the presence of tumor in patients when a prostate biopsy is derived from near the tumor site. The laboratory plans to further refine this classifier and test it on a larger cohort.
• Investigators at another BDL at Johns Hopkins University are applying a mass spectrometry-based analytical method to measure different glycosylated forms of glycoproteins from complex biological samples by coupling a glycopeptide extraction strategy for specific glycosylation with SRM-MS. Using this approach, they have recently demonstrated that the relative abundance of glycosylated PSA isoforms did not correlate with total PSA protein levels. Interestingly, a sialylated PSA was differentially distributed between tumor and non-tumor tissues. These findings suggest that glycosylated isoforms of glycoproteins can be quantitatively analyzed and have the potential to serve as cancer biomarkers. The laboratory is currently pursuing new candidate biomarkers based on their differential glycosylation in prostate cancer specimens. Their partnering Biomarker Reference Laboratory (BRL) at Johns Hopkins University intends to validate 15 promising prostate glycoproteins that were previously identified as candidate biomarkers. They are developing ELISA and glycan-linked immunosorbent assays (GLISA) for each of the candidate biomarkers to be followed up by construction of multiplex assays for analytical and clinical validation. In addition, the BRL will develop and validate multivariate predictive models which will be applied to identify aggressive prostate cancer.
• By combining the latest ultrasound technology with tumor angiogenesis-targeted microbubbles, the BDL at Stanford University is improving the imaging of prostate cancer. Microbubbles are gaseous bubbles encased by lipid shells functionalized with antibodies specific for the target tumor-associated angiogenesis marker VEGFR2. When introduced into the bloodstream, VEGFR2-targeted microbubbles bind to the endothelial cells of the tumor neovasculature to provide enhanced contrast during ultrasound. Incorporated into transrectal ultrasound, this approach will increase the accuracy of detection during the screening process. With the emergence of new prostate cancer cell surface associated markers this approach may enable targeted screening for earlier detection of prostate cancer.

• Team Projects

  Background and Clinical Goals: Given the high prevalence of non-aggressive prostate cancer, the team projects focus on the early detection of cancers that are likely to progress. Investigators developed highly interactive experiments that will eventually validate biomarkers in Phases II and III in urine and in tissue samples. The 3-year projects were developed around two major independent themes: pre-validation/validation of urine-based biomarkers for early detection of clinically significant prostate cancer, and for assisting clinicians in decision making; and molecular classification and diagnosis of clinically significant prostate cancer based on tissue-discovered markers.

  Project 1: Evaluation of Urine PCA3 and TMPRSS2-ERG

  Study: The project will determine whether multiplex combination of urinary measurement of TMPRSS2-ERG fusion and PCA3, together with serum PSA and percent-proPSA can improve the early detection of histopathologically aggressive prostate cancer.

  Project 2: Establishing Community-Based Normal Distributions of Urinary PCA3 and TMPRSS2-ERG

  Study: The frequency of detection of post-DRE urinary PCA3 and TMPRSS2-ERG fusion, and percent-free PSA and percent-proPSA in serum of a community-based sample of men undergoing prostate cancer screening will be determined by this project. The goal is to examine whether frequency differs between African-American men as compared to men of other racial backgrounds, and between Hispanic-American men as compared to men of non-Hispanic ethnicity. The broad representation of many EDRN sites in this project facilitates a strong accrual of racial and ethnic minorities, thus improving the generalizability of the anticipated outcome from this work.

  Project 3: Measuring Multiple Cancer Secreted Proteins

  Study: Selected Reaction Monitoring (SRM) quantitative mass spectrometry (MS) will be used to identify and quantify proteins in urine. The study will discern aggressive from indolent prostate cancer by measuring multiple proteins secreted by prostate cancer cells that are shed in voided urine.

  Project 4: Tissue Microarrays for Biomarker Validation

  Study: Identification and testing of available tissue microarrays (TMAs) is essential for evaluation and validation of diagnostic and prognostic biomarkers. The immediate goal is to identify TMAs that are available through participating sites. A list of available TMAs is being developed which should contain statistically powered clinically defined end points, such as organ confined versus cancers with extracapsular extension; or cancers with lymph node involvement or with distant metastasis; hormone refractory cancers; neuroendocrine carcinomas. In addition, they should be accompanied by clinical data such as survival. A “Google-Document” will be developed that will allow all participants to monitor the current status of available and potentially available TMAs.

  Project 5: Upgrading Gleason Scores on Radical Prostatectomies

  Study: This project will determine the clinical parameters associated with upgrading Gleason scores on radical prostatectomies. The current EDRN biopsy collection protocol records significant clinical and pathology data. A subset of these individuals has undergone prostatectomy at their respective participating institutions. At the University of Michigan and Weill Cornell Medical College sites alone, there will be approximately 500 cases where patients have entered an EDRN protocol for the collection of urine and blood samples prior to surgery by the end of the second year. The major goal of this project is to ensure that all clinical parameters that can be used in conjunction with future biomarkers are assessed. This project will also include mandatory review of current literature and available data for modeling.

  Project 6: Molecular Sub-Classification of ETS and Non-ETS Rearrangements

  Study: This project addresses the molecular sub-classification of ETS and non-ETS rearrangements in prostate cancer. Significant data is developing to suggest that similar to the classification of acute myeloid leukemia (AML), prostate cancer can be molecularly classified to subcategories by specific genetic alterations including gene fusions and mutations. This project will try to determine the frequency of each category. This process is similar to developing a classification for AML even prior to known treatments. An immediate goal will be to identify individuals with targetable mutations. This study will use assays pioneered by EDRN investigators, which have been used in over 100 studies throughout the world and some of which are planned to be commercially available (e.g., the TMPRSS2-ERG in situ assay licensed and distributed by Ventana/Roche).

  Project 7: Prostate Cancer Phenotypes and Clinical Outcomes

  Study: This project is focused on prostate cancer phenotypes and clinical outcomes. Two markers, AGR2 and CD10, can group patients into three survival or disease risk categories: 1) AGR2hi/CD10lo for high survival/low risk; 2) AGR2lo/CD10hi for low survival/high risk; and 3) AGR2hi/CD10hi and AGR2lo/CD10lo for intermediate survival/risk. Additional informative sub-grouping will be provided by markers associated with Gleason pattern 4 vs. pattern 3 cells, and CD10+ vs. CD10– cancer cell types. Further reactive stroma markers will be added based on a recently developed classification.

  Project 8: Vascular Tissue Biomarkers for Molecular Imaging

  Study: This project centers on the identification of cancer related vascular tissue biomarkers for molecular imaging. This project directly addresses a need to develop molecular biomarkers that can identify tumor endothelium. Validated biomarkers will then be handed off for further development by the Stanford BDL group. These biomarkers will enhance the group’s ability to monitor cancers over time and seek associations with biomarkers developed by other team projects of this collaborative group. A specific example might be a particular vascular pattern seen in ERG-rearranged prostate cancers as compared to non-rearranged cancers given the important role of ERG in vascular development.

  ----

  Other EDRN News: Urine Test Added to PSA May Indicate Increased Risk for Prostate Cancer

  Researchers from NCI’s Early Detection Research Network have published findings on a urine test that may be used to better determine which men with elevated PSA readings are at highest risk for invasive prostate cancer.  Read more at the University of Michigan website at:

  http://www.uofmhealth.org/news/cancer-prostate-0803

  The paper, titled, Urine TMPRSS2:ERG Fusion Transcript Stratifies Prostate Cancer Risk in Men with Elevated Serum PSA, was published in the August 3, 2011 issue of Science Translational Medicine .