• Home
  • News
  • Calendar
  • About DF/HCC
  • Membership
  • Visitor Center





Project 1: Tumor and circulating markers as links between obesity and lethal prostate cancer

Project Co-Leaders:

Lorelei Mucci, ScD (HSPH)
Edward Giovannucci, MD, ScD (HSPH)

Neil Martin, MD, MPH (BWH)
Jennifer Rider, ScD (HSPH)
Massimo Loda, MD (DFCI)


Matthew Vander Heiden, MD, PhD
Philip Saylor, MD (MGH)

Obesity is a significant public health problem globally, and two-thirds of US men are overweight or obese. The obesity epidemic is of particular concern for prostate cancer patients, as obesity is strongly linked to worse prognoses. The study of obesity in the context of lethal prostate cancer is appealing since it is a potentially modifiable risk factor as well as a means of gaining mechanistic insight into drivers of lethal disease. Research has shown that high insulin and low adiponectin levels, which are altered in the obese state, are linked to lower cancer survival rates. In our preliminary data, tumors from patients who are overweight and develop lethal prostate cancer show enriched expression of metabolic-related genes including adipocytokine signaling. Obesity is further characterized by a low-level, chronic inflammation that may promote the evolution of a more aggressive tumor or propel a tumor to the metastatic state. In our preliminary data, obesity is linked to greater chronic inflammation and the presence of post-atrophic hyperplasia adjacent to prostate tumors. Patients with key tumor genomic features may be more vulnerable to the effects of obesity; we show that obesity is linked to worse cancer prognosis primarily in men with tumors harboring the TMPRSS2:ERG rearrangement.

The project will elucidate mechanisms underlying the link between obesity and lethal prostate cancer and identify patient subgroups more susceptible to the obesity milieu. We hypothesize that obesity may act through local tumor effects, which are “the seed” to increase a prostate cancer’s lethal potential as well as via systemic effects that fuel “the soil” for metastatic growth. We focus on metabolism and inflammation pathways—two domains with established links to obesity. Further, we include a tumor biomarker discovery aim to identify novel pathways that mediate the link between obesity and lethal disease. The Project is nested among patients in the Prostate Cancer DF/HCC SPORE biorepository and integrates anthropometric data, tumor (N=1200 patients) and circulating biomarkers (N=1000 patients), and uniquely focuses on lethal cancer as the primary endpoint. We will address the following aims:

  1. Examine the link between obesity, whole body, and tumor metabolism, and lethal prostate cancer
  2. Examine the link between obesity, inflammatory markers, and lethal prostate cancer. We will study obesity as an inflammatory condition, and hypothesize that obesity may act locally via increased peri-tumoral inflammation to stimulate local growth and drive metastatic potential, as well as via systemic inflammatory markers that may fuel cancer progression
  3. Using a discovery-based approach, identify tumor markers associated with obesity and lethal prostate cancer

Project 2: Gleason-based mRNA and metabolomic profiling to predict prostate cancer progression

Project Co-Leaders:
Massimo Loda, MD (DFCI)
Matthew Freedman, MD (DFCI)

Kathryn Penney, ScD (BWH)
Adam Kibel, MD (BWH)

Clinicians and researchers are currently unable to distinguish at diagnosis (with sufficient confidence) men with prostate cancer (CaP) who have indolent disease, from those who have aggressive disease. At present, the strongest predictor of lethal CaP is Gleason score. Utilizing gene expression array data, we previously identified a 157 gene mRNA signature that distinguished high from low Gleason score. This signature significantly improved prediction of lethal disease among men with clinically heterogeneous Gleason score 71. The signature may have clinical utility, but before applying it to patients it must be further refined and then tested in biopsy specimens to determine if the predictive accuracy is sufficient to influence treatment decisions at the time of diagnosis. This mRNA study also identified metabolic pathways differentially enriched in high and low grade disease, generating hypotheses for biological mechanisms that may underlie CaP differentiation and clinical progression. Since different types of biological data may add to the mRNA signature as well as provide different biological information, it is worthwhile to further investigate the metabolic pathways identified. We therefore propose to build upon our promising expression profiling findings with the following aims:

  1. Preparation and testing of Gleason signature of lethal disease for clinical use. To determine if the mRNA signature can be applied in the clinic, we will further validate it by testing its predictive accuracy in biopsy specimens at the time of diagnosis. We hypothesize that within Gleason score 7, the signature’s ability to predict lethal disease may improve if applied specifically to the grade 3 or grade 4 focus of the tumor.
  2. Metabolomics of Gleason grade in tumor as predictor of lethal disease. Our mRNA study identified metabolic pathways (pyrimidine, propanoate, and beta-alanine metabolism) differentially enriched in high and low grade tumors. Our preliminary data suggests these same pathways may be differentially enriched using metabolomic data; we hypothesize that metabolites may themselves be associated with Gleason score and lethal disease.
  3. Metabolomics of Gleason grade in serum as biomarker for upgrading. We hypothesize that metabolites in serum associated with Gleason grade may indicate the presence of higher-grade tumor not detected at biopsy and could serve as a biomarker for monitoring disease progression of active surveillance patients.

Project 3: Genomic determinants of resistance to primary androgen deprivation therapy and aggressive disease

Project Co-Leaders:
Levi Garraway, MD (DFCI) (HMS)
Mary-Ellen Taplin, MD (DFCI)

Prostate cancer (CaP) is the most common solid tumor and the second most common cause of cancer death in men in the United States. A significant proportion of patients with prostate cancer have aggressive tumors that require treatment. Despite their best efforts at disease control, some men experience tumor progression culminating in death from prostate cancer. Many other prostate cancers are indolent and do not result in cancer mortality, even without treatment. Although the field has witnessed important therapeutic advances in recent years, the identification of molecular or genetic features that distinguishes indolent from lethal disease has remained challenging.

CaP is considered a hormone-dependent tumor: malignant prostate epithelial cells express the androgen receptor (AR) and require persistent androgen exposure for survival. The clinical evolution of an aggressive, localized tumor may result in metastases and/or progression to castration resistant prostate cancer (CRPC), a state wherein androgen deprivation therapy (ADT) has been implemented but tumor drug resistance has emerged. Metastatic CRPC is generally lethal. Although some mechanisms of resistance have been identified, the genetic basis for the development of CRPC remains incompletely characterized. Thus, two of the most important and challenging unanswered questions for clinicians who treat CaP (and for patients with CaP) are: 1) the determinants of indolent versus aggressive disease; and 2) a comprehensive understanding of mechanisms of resistance to ADT.

The application of systematic genomic and functional approaches to fundamental questions in cancer has reached a unique point of convergence. For the first time, it is feasible to garner knowledge of the complete spectrum of recurrent genome alterations in CaP and deploy this knowledge in a manner that may inform both diagnosis and response to existing and emerging therapeutics. The overarching goal of this project is to apply both “complete” and targeted massively parallel sequencing to characterize the genetic underpinnings of clinically aggressive and CRPC. In parallel, we will apply systematic functional screens to discover genes whose overexpression mediates resistance to ADT. Integration of this functional data with knowledge of recurrent CaP alterations may provide a framework for improved diagnosis and, in the future, durable therapeutic approaches for CaP.

Project 3 is divided into the following specific aims:

  1. Whole genome sequencing of CaP that shows de novo or acquired resistance to ADT
  2. Targeted genomic profiling of CaP cohorts to identify somatic determinants of clinical aggressiveness and progression to CRPC
  3. Characterization of resistance to ADT by systematic ORF screening

Project 4: Targeting Mechanisms Driving AR Activity in Advanced CRPC

Project Co-Leaders:
Steven Balk, MD (BIDMC)
Myles Brown, MD (DFCI)

Xin Yuan, MD, DSc

Our focus during the last funding period was on molecular mechanisms driving androgen receptor (AR) activity in castration resistant prostate cancer (CRPC) and on the specific functions of AR in CRPC. Based on our work, and others’, it is now clear that increased intratumoral androgen synthesis contributes to the reactivation of AR transcriptional activity in CRPC. This mechanism is targeted by CYP17A1 inhibitors, including the recently FDA approved drug abiraterone. Unfortunately, most patients who respond to abiraterone relapse within 1-2 years, and current data indicate that AR remains active in these recurrent tumors. Additional mechanisms that enhance AR activity and drive tumor progression in CRPC are almost certainly also contributing to abiraterone resistance/relapse, and to resistance to novel AR antagonists such as enzalutimide. Moreover, the marked decrease in androgen levels in patients treated with abiraterone likely generates selective pressure for the emergence of further resistance mechanisms.

Recent dramatic improvements in DNA sequencing methods and in methods for RNA and DNA amplification from small tissue samples now make it possible to much more comprehensively analyze patient samples and identify changes including mutations, gene amplifications, gene fusions, and alternative splicing that would have been missed in previous microarray based studies. Moreover, while TMPRSS2:ERG fusion and PTEN are examples of frequent genomic alterations in prostate cancer (CaP), it is becoming clear that multiple additional genomic and epigenomic events may be contributing to CaP in small subsets of patients, and that many of these may have therapeutic implications. This may be particularly true in advanced CRPC, and we hypothesize that a diversity of genomic alterations may be contributing to AR activation and tumor progression in CRPC and abiraterone resistant CRPC. Our objective in Aim 1 is to identify mechanisms mediating abiraterone resistant CRPC using both a hypothesis driven approach based on mechanisms discovered in xenograft models, and through an unbiased comprehensive genomic and transcriptional analysis of CRPC, including abiraterone resistant/recurrent tumors.

During the last funding period we also demonstrated that the AR transcriptional program can be markedly altered in CRPC. In particular, we found that AR in CRPC cells, but not androgen dependent cells, is recruited to and stimulates the expression of genes including CDK1 and UBE2C that drive cells through mitosis. Previous studies have shown that EZH2, a SET domain histone methyltransferase known to play a role in gene silencing through H3K27 trimethylation, is up-regulated in CRPC. In recent studies, we found that EZH2 can be recruited to the cis-regulatory elements of a subset of genes targeted by AR in CRPC, and can form a complex with AR. Significantly, EZH2 directly up-regulates these AR targets in CRPC cells and stimulates their growth, but does not have these effects in androgen-dependent CaP cells. This EZH2 function is dependent on SET domain methyltransferase activity, but is not dependent on H3K27 trimethylation, suggesting additional substrates. Finally, this AR coactivation function of EZH2 is enhanced by AKT mediated phosphorylation of a specific serine in EZH2.

These findings, in conjunction with our data showing that AR is also phosphorylated and activated by CDK1, indicate that a positive feedback loop involving AR, EZH2, AKT and CDK1 may be driving AR activity and proliferation in a subset of CRPC.

Specific aims for Project 4 are to:

  1. Identify mechanisms mediating AR activation and tumor progression in castration resistant and abiraterone resistant prostate cancers
  2. Develop pre-clinical therapeutic models and biomarkers to facilitate clinical translation of EZH2 as a therapeutic target in CRPC
  3. Assess the efficacy of AKT and CDK1 inhibitors in combination with abiraterone and develop a clinical trial targeting the AR-EZH2 axis in CRPC