Skip to Content
Translational Research Program (TRP)
Contact CIP
Show menu
Search this site
Last Updated: 10/10/18

Prostate Cancer SPORE

Memorial Sloan Kettering Cancer Center

Principal Investigator:
Howard I. Scher, MD

Principal Investigator Contact Information

Howard I. Scher, MD
Chair, Urology Department
Memorial Sloan-Kettering Cancer Center
1275 York Avenue
New York, NY 10021
Tel: (212) 639-4322
Fax: (212) 422-1095


For American men, prostate cancer remains the most common cancer and the second leading cause of cancer deaths. The two most significant problems in the management and treatment of prostate cancer lie at opposite ends of the disease continuum. For localized disease, the objective is to avoid overdiagnosis and treatment of tumors that are unlikely to adversely affect the quality or duration of an individual patient’s life. For more advanced disease, the need is to develop new drugs and therapeutic strategies that build on the recent unparalleled advances in systemic therapy. Our program is focused on defining the mechanisms that contribute to disease progression and resistance to therapy, developing validated assays that can be used in a clinical setting to identify them in patient tumors, designing dedicated biomarker trials to validate them clinically, and demonstrating their utility to inform medical decision making.

The design of this SPORE is informed by the clinical states model, which separates the disease continuum into milestones, each representing a clinical context where key medical decisions are needed. Each research project is centered on the unmet needs of patients at particular milestones. A central tenet of the clinical states model is that prostate cancer is a dynamic disease in which the biologic determinants of tumor growth change as the disease progresses, due to the inherent genomic instability of the tumor itself, as well as selection pressure exerted by the specific treatments to which it has been exposed.

Guiding the choice of therapy to maximize patient benefit at each point requires a direct biologic characterization of both the tumor (through directed biopsy or blood-based analyses) and the patient (functional status, tumor microenvironment, immune factors, etc.). Recent discoveries enabled by state-of-the-art molecular profiling and bioinformatics, combined with a commitment by the research community to obtain human prostate cancers representing all clinical states for analysis, are leading rapidly to a biology-based taxonomy that will further refine medical decision making — personalized, precision medicine. Our SPORE is organized into 5 major research projects that are tightly integrated to achieve this objective.

Clinical states model showing unmet needs at each clinical state and the SPORE Research Projects designed to address them
Clinical states model showing unmet needs at each clinical state and the SPORE Research Projects designed to address them


Barry S. Taylor, PhD — Basic Science Co-Leader
David B. Solit, MD — Clinical Science Co-Leader

The era of genomically directed oncology has arrived, catalyzed by efforts to rationally design therapies targeting the molecular aberrations that promote individual tumor growth and progression. This has led to genomic sequencing of tumors from patients with progressive disease to guide the treatment of their cancer. At present the clinical utility of prospective tumor profiling is limited in many cancer types, perhaps most so in prostate cancer. Due in part to their profound clinical and mutational heterogeneity, we know very little about how to therapeutically target the molecular lesions that drive the emergence of the most aggressive metastatic castration-resistant prostate cancers (CRPC).

Recent work by our group showed that prospective clinical sequencing of active, advanced prostate cancers with linked clinical annotation can reveal both gene- and pathway-level genomic aberrations enriched in CRPC. We found that genetic lesions related to homologous recombination (HR) were far more common in castration-resistant cancers than in even advanced castration-sensitive metastases. These findings underscore the potential therapeutic utility of defining the lesions that drive CRPC, but without a clinical-translational approach that tests these hypotheses clinically, the gap in our understanding will only widen. We intend to overcome this urgent clinical challenge by first developing and comprehensively analyzing a one-of-kind cohort of approximately 2,000 sequenced advanced prostate cancers from patients under active care for whom we have detailed clinical annotation, as well as treatment-response and outcome data. A major goal of this effort will be to identify alterations associated with progression to castration-resistance, a strategy that will facilitate the testing of novel approaches to treat such tumors. We will then assess the extent to which genomic loss of heterozygosity and related hallmarks of HR deficiency are correlated with different underlying somatic or germline HR/DNA repair genotypes, specific clinical variables, disease-specific progression, and outcome. Finally, we will clinically test the hypothesis that men with CRPCs bearing the hallmark of HR dysfunction, including those with germline or somatic mutations in BRCA genes, derive clinical benefit from a class of drugs known as PARP inhibitors.


Neal X. Rosen, MD, PhD — Basic Science Co-Leader
Brett S. Carver, MD — Basic Science Co-Leader
Dana E. Rathkopf, MD — Clinical Science Leader

The androgen receptor (AR) signaling pathway is the key driver of human prostate cancer at all points in the illness. Prostate cancers are almost always sensitive to blockade of the AR (eg, androgen deprivation therapy), and tumors that develop resistance to these therapies are often sensitive to second-generation inhibitors of the pathway. However, the majority of men with advanced prostate cancer ultimately succumb to the disease.

Activation of the PI3K signaling pathway is the second most prevalent oncogenic event in prostate cancer, occurring in the majority of advanced castration-resistant tumors. One regulator of PI3K signaling, the PTEN tumor suppressor gene, is lost or inactivated in 40% of these tumors, and genetic events that activate PI3K are also found in a significant number of patients. Recent genomic profiling studies of metastatic castration-resistant prostate cancer have reported activating alterations in the subunits of the PI3K complex (PIK3CA, PIK3CB, PIK3R1) in approximately 10% to 20% of cases.

Our previous work revealed that inhibition of AR signaling activates PI3K signaling and inhibitors of PI3K signaling activate AR signaling. Hence, it is likely that this process reduces the benefit of therapy with single-agent AR or PI3K inhibitors. In support of this hypothesis, we have shown that combined inhibition of both pathways has profound therapeutic effects in preclinical models of advanced prostate cancer. The major goal of this project is to optimize such therapies, tailoring the therapeutic strategy to specific tumor genotypes. We will do this by determining the biologic consequences of different mechanisms of PI3K activation in prostate cancer, determining the best means of pharmacologically inhibiting PI3K and AR signaling in prostate tumors with PTEN loss and/or PI3K activation, and optimizing the dose and schedule of these combinations in novel organoid and genetically engineered mouse models of prostate cancer.


Charles L. Sawyers, MD — Basic Science Co-Leader
Howard I. Scher, MD — Clinical Science Co-Leader

Recent genomics studies have shown that men with metastatic castration-resistant prostate cancer (CRPC) have an unexpectedly high frequency of alterations in the tumor suppressor genes TP53 and RB. In laboratory studies, we have found that these genetic alterations lead tumor cells to change their identity, a phenomenon known as lineage plasticity, and thereby become resistant to anticancer therapies that target the androgen receptor (AR), such as abiraterone and enzalutamide. Parallel studies revealed that a blood sample from an individual patient with CRPC may have circulating tumor cells (CTCs) with markedly different morphological features. We have found that higher CTC heterogeneity is also correlated with worse treatment response to AR-targeted drugs (abiraterone and enzalutamide) but not to taxanes (docetaxel, cabazitaxel).

This project will investigate how the lineage plasticity induced by TP53/RB alterations is connected to CTC heterogeneity and resistance to AR-targeted therapies. In our first aim, we will use preclinical model systems (cell lines, xenografts, and organoids) to define the molecular changes associated with lineage plasticity and determine if these changes are reversible. We will track the emergence of tumor heterogeneity in these preclinical models and examine how AR-targeted drugs and taxanes impact lineage plasticity and tumor heterogeneity. We will also explore drugs with the potential to prevent or reverse resistance to AR-targeted drugs. In our second aim, we will analyze existing data from patients with metastatic CRPC treated at MSK, including data on CTCs isolated from blood and genomic data from biopsy samples. We will explore the relationship between CTC heterogeneity and TP53/RB alterations in the patients’ genomic data. We will also examine alterations in other genes such as PTEN, BRCA1/2, and ATM. We will then test whether there is a relationship between CTC heterogeneity and response to AR-targeted therapies and taxanes. Overall, this project is designed to enable development of biomarkers to predict which patients are unlikely to respond to AR-targeted drugs.


Charles L. Sawyers, MD — Basic Science Co-Leader
Wassim Abida, MD, PhD — Clinical Science Co-Leader

The androgen receptor (AR)-directed agents enzalutamide and abiraterone acetate are standard first-line therapies for the treatment of patients with metastatic castration-resistant prostate cancer (CRPC). These agents are generally well tolerated and effective at delaying disease progression, prolonging survival, and improving quality of life, but both drugs have limited durations of effectiveness. Patients who experience progression on first-line AR-targeted therapies typically respond poorly to subsequent AR-targeted agents, and there is currently no predictive biomarker that aids in the selection of a second-line therapy. Our group and others have identified upregulation of the glucocorticoid receptor (GR) as one mechanism of resistance to enzalutamide and abiraterone acetate. Our findings suggest that 1) GR may be a biomarker that can predict whether a patient will respond to enzalutamide or abiraterone and 2) blocking GR upregulation in patients with cancer resistant to enzalutamide or abiraterone could result in tumor responses.

In preliminary studies, we found that a form of GR called GRaC can promote resistance to enzalutamide and is detectable in some patient-derived tumor cultures. It is necessary to characterize GRaC and other forms of GR in order to develop GR as a biomarker and a therapeutic target. Using a laboratory model of prostate cancer that is resistant to enzalutamide, we found that GR expression is epigenetically modulated, and that treatment with an inhibitor of the BET family of proteins suppresses GR expression and restores sensitivity to enzalutamide. Expanding on these findings, we will leverage our group’s laboratory and clinical expertise to 1) investigate the role of GR variants in resistance to enzalutamide, 2) define the broader effects of BET inhibition across a range of preclinical models of prostate cancer, and 3) investigate the activity of a novel BET inhibitor in men with metastatic CRPC whose disease has progressed during treatment with enzalutamide or abiraterone acetate.


Hans G. Lilja, MD, PhD — Basic Science Co-Leader
Andrew J. Vickers, DPhil — Applied Science Co-Leader

One of the primary clinical challenges in prostate cancer is the management of localized disease detected via cancer screening. Recent years have seen a dramatic increase in conservative management of low-risk disease using a strategy known as active surveillance. A corollary has been increased attention to the algorithms used for categorizing patients according to risk. Gleason grade, a measure of a cancer’s likelihood to spread as determined by microscopic evaluation of tissue samples, is strongly predictive of oncologic outcome. However, some men with low Gleason grade on biopsy do in fact harbor high-risk disease, and many men with higher Gleason grade would not in fact experience symptoms of prostate cancer during the natural course of their life even if their cancer was left untreated. We have shown that a statistical model based on a panel of kallikrein markers in blood — total PSA, free PSA, intact PSA, and human kallikrein 2 (hK2) — strongly predicts the risk of high-grade (Gleason score of 7 or higher) cancer on biopsy. The panel has been validated in multiple studies involving close to 15,000 men. We have also demonstrated that the panel is highly predictive of the long-term risk of distant metastasis in men followed for many years without screening.

In this project, we will evaluate whether the clinical role of the panel could expand from decisions about the need for a diagnostic biopsy to those concerning immediate treatment decision-making and active surveillance protocols. We will first use data from five independent cohorts, three from Europe and two from the US, to determine whether the panel of four kallikrein markers can predict either unfavorable pathology at radical prostatectomy or posttreatment oncologic outcomes such as metastases. If so, the panel — data for which is often available from the diagnostic biopsy — could be used to help make decisions about treatment versus conservative management. Our second aim is to determine whether the panel of markers can predict the result of biopsies administered while a patient is being treated with active surveillance. If so, using the panel to avoid biopsy in men at low risk of progression would decrease the number of biopsies required by active surveillance protocols, reducing morbidity and potentially increasing the acceptability of active surveillance as a management strategy. We also aim to determine whether the panel could be improved by either adding a novel marker, microseminoprotein-Β, or by assessing the presence of specific genetic variants (known as single-nucleotide polymorphisms or SNPs) related to the proteins included in the panel.

Core A: Biospecimen Repository

Victor E. Reuter, MD — Core Director
Anuradha Gopalan, MD — Core Co-Director
Martin Fleisher, PhD — Core Co-Director

The Biospecimen Repository supports the basic translational research efforts of the SPORE by playing a central role in collecting, annotating, storing, distributing, and tracking prostate cancer tissue and blood biospecimens from patients enrolled in research protocols. These specimens are procured following the MSK Institutional Review Board guidelines and banked, with our direct input, by the Tissue Procurement Service and the Clinical Chemistry Biomarker Laboratory. Detailed biospecimen annotation, including documentation of pre-analytic processing variables, pathology findings, and patient clinical history information, are recorded in robust relational databases. The Biospecimen Repository provides SPORE investigators with expert histopathological evaluation of tumor samples both from patients enrolled in prostate cancer research protocols and from mouse models of the disease. It provides assistance in performing and interpreting immunohistochemical and in situ hybridization assays, in selecting tissue for microdissection and construction of arrays, and in collaborating with project leaders and the Biostatistics and Bioinformatics Cores. A central role of this core is also to characterize novel markers in a standardized fashion and establish staining protocols and performance characteristics across the prostate SPOREs, with the ultimate goal of credentialing markers for use in clinical trials.


Glenn Heller, PhD — Core Director
Andrew J. Vickers, D Phil — Core Co-Director

The role of the Biostatistics Core is to support investigators of the SPORE in Prostate Cancer in their research efforts, including laboratory experiments and the design and analysis of clinical trials. Core members will assist in the formulation of the experimental design and in the analysis and interpretation of the data at the conclusion of both preclinical and clinical studies. A core member will conduct a protocol review with the principal investigator during the trial design phase. Based on this review, we will provide a statistical section for the protocol, outlining major scientific objectives, population to be studied, primary and secondary endpoints, experimental design, a randomization procedure if necessary, analysis plans, and a targeted sample size justified in probabilistic terms. At the conclusion of the trial, data analyses will be performed to assess outcomes of the primary and secondary endpoints stated in the protocol. If current statistical methodology does not adequately address a research question in this SPORE, alternative methodologies will be explored.


Nicholas D. Socci, PhD — Core Director
Nikolaus Schultz, PhD — Core Co-Director

The Bioinformatics Core provides services that are vital to many of the projects within the SPORE in Prostate Cancer, including the development and implementation of statistical and computational techniques to process next-generation sequencing data from a variety of applications. The Core will apply standard processing algorithms and pipelines to a large number of samples and has the expertise to create and deploy custom methods and applications as needed by the SPORE researchers. The Core can provide SPORE projects with state-of-the-art-applications with very little development time or effort, greatly reducing the time and cost required for individual projects to develop their own analysis pipelines. The Bioinformatics Core will develop and provide a state-of-the-art genomic analysis pipeline to detect variants in targeted-DNA assays, the Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) test, and whole-exome sequencing (WES). This pipeline will detect both single nucleotide changes and small insertions and deletions for both somatic- and germline-variant calling scenarios. The core will also develop a state-of-the-art genomic analysis pipeline for DNA copy number analysis, which can measure both total copy number changes and allele-specific copy number, including loss of heterozygosity (LOH). Additionally, the Bioinformatics Core will facilitate the sharing of data generated in the SPORE research projects and enable collaborative, integrative analysis via the MSK cBioPortal by collecting, formatting, and importing data generated by these research projects into the cBioPortal.


Brett S. Carver, MD — Core Director
Anuradha Gopalan, MD — Core Co-Director

The overall purpose of the Animal Models Core is to facilitate the translational research goals and objectives of the SPORE in Prostate Cancer by providing investigators with assistance in the design and generation of transgenic and knockout/knockin mouse strains, a centralized repository for these mouse strains, and material of common interest, such as dissected tissues, DNA, RNA, or protein extracts from these mouse strains. The specific aims of the Animal Models Core are to 1) generate and provide genetically engineered mouse models and patient-derived xenograft models as needed by SPORE investigators; 2) provide small animal imaging, provide histopathologic analyses of tumor specimens, design and conduct preclinical trials, and serve as a repository and breeding service of mouse strains as requested by SPORE investigators; and 3) provide SPORE laboratories with dissected tissue specimens, DNA, RNA, and proteins from the various mouse models utilized by our investigators.


Yu Chen, MD, PhD — Core Director

The study of castration-resistant prostate cancer (CRPC) has been limited by lack of in vitro models that represent the molecular and phenotypic diversity of the disease. We have optimized conditions for the generation and growth of 3D in vitro prostate organoids from benign human and mouse prostate cells. The resulting organoids can be engineered with specific genetic lesions using patient-derived biopsy specimens. The organoids maintain the histology and 3D architecture of the cancer tissue. We will collaborate with each research project to engineer and study organoids that harbor specific genetic alterations or molecular phenotypes, including those with mutations in the DNA repair pathway, mutations in the PI3K pathway, mutations in TP35 and RB1, and overexpression of glucocorticoid receptor variants. Working with the Animal Models Core, we will generate murine prostate organoids from genetically engineered mouse models harboring the specific genetic alterations of interest for the research projects. We will work with each project to develop protocols for in vitro drug treatment, in vivo xenograft studies, and genetic engineering of the organoid lines. We will also conduct research aimed at improving services, including further optimization of organoid acquisition and growth protocols, identification of clinical and molecular determinants of successful organoid growth, and development of protocols to engineer genetic lesions into benign human prostate organoids. We have generated 17 organoid lines over the past 2 years from CRPCs that harbor a number of genetic alterations not present in publicly available cell lines. Leveraging our SPORE’s commitment to collect, annotate, and sequence biopsy specimens from more than 2,000 patients with CRPC, we plan to establish 20 clinically and molecularly annotated CRPC organoid lines each year.


Howard I. Scher, MD — Core Director
Peter T. Scardino, MD — Core Co-Director
Charles L. Sawyers, MD — Core Co-Director
Philip W. Kantoff, MD — Core Co-Director

The purpose of the Administrative Core is to support the translational research objectives of our SPORE in Prostate Cancer by serving as the organizational hub, optimizing collaboration among SPORE investigators within and outside the MSKCC. The specific aims of the Administrative Core are to (1) provide coordination for all educational and scientific activities of the SPORE, and (2) provide centralized administrative support of day-to-day activities to all the components of the SPORE: research projects, cores, and developmental programs. The services centralized in the Administrative Core provide the following administrative and communications infrastructure that serve all components of the SPORE: financial management, editorial services, grants administration, meetings coordination, communication, data management system support.


Howard I. Scher, MD — Principal Investigator

Dr. Scher is Chief of the Genitourinary Oncology Service in the Department of Medicine at MSK, as well as incumbent of the D. Wayne Calloway Chair in Urologic Oncology and Leader of the Genitourinary Oncology Disease Management Team. He leads the Prostate Cancer Clinical Trials Consortium (PCCTC), a nationwide, multi-institutional research organization headquartered at MSK and funded by the US Department of Defense and the Prostate Cancer Foundation, whose members collaborate to design, implement, and complete phase I/II and phase II trials in prostate cancer, translating scientific discoveries to improved standards of care. The PCCTC has been the designated clinical trials mechanism for the SPOREs in Prostate Cancer. Dr. Scher developed the clinical states model of prostate cancer progression, which, in categorizing the clinical spectrum of prostate cancer from diagnosis to metastasis, provides a framework for patient management, clinical trials, and biomarker qualification. He led the international Prostate Cancer Clinical Trials Working Groups (PCWG2 and PCWG3) to standardize the methodology for the conduct of clinical trials in castration-resistant prostate cancer. The PCWG2’s recommendations are now standards. Under his leadership, the PCWG3 seeks to evolve the prostate cancer states framework from a taxonomy based solely on clinical criteria and treatment algorithms to one that includes molecularly based measures. PCWG3 addresses the need for more trials that are specific to the different clinical subtypes of prostate cancer and the integration of biomarkers and trial design. Dr. Scher led the phase I/II clinical trial of enzalutamide at MSK and was PI of the phase III trial. He is also leading the development of circulating tumor cells as measures of prognosis, as an efficacy-response biomarker of clinical benefit, and as source of tissue for the development of predictive biomarkers to optimize treatment selection.