044PG Management of Prostate Cancer: A Case Based Approach with

Transcription

044PG Management of Prostate Cancer: A Case Based Approach with
044PG
Management of Prostate Cancer: A
Case Based Approach with
Emphasis on Integrating New
Molecular Diagnostics into Clinical
Practice
Sunday, May 18, 2014
3:30 PM – 6:30 PM
Faculty
Eric A. Klein, MD - Course Director
Andrew J. Stephenson, MD
Robert Dreicer, MD
Management of Prostate Cancer: A Case Based Approach with
Emphasis on Integrating New Molecular Diagnostics into Clinical
Practice
Faculty
Eric A. Klein, MD
Chairman, Glickman Urological and Kidney Institute
Cleveland Clinic
Robert Dreicer, MD
Chairman, Department of Solid Tumor Oncology
Taussig Cancer Institute
Cleveland Clinic
Program Agenda
3:30PM
Welcome and Introduction
3:35 – 5:15PM
Overview of genomics and their use via case presentations
5:15 – 6:30PM
Management of Advanced Disease with case presentations
Learning Objectives
Design appropriate screening strategies based on individual demographics, risk factors,
and PSA history and to incorporate new biomarkers into routine clinical practice
Distinguish and understand the use of new molecular and genomic based tests for
decisions on initial and rebiopsy, and choosing and following men on surveillance, and
deciding on adjuvant therapy after radical prostatectomy
Describe new therapeutic agents for the management of castrate resistant disease and
outline a coherent strategy for their use
Disclosures:
EAK: Research support from and consultant to Genomic Health, Metamark, and
GenomeDx Biosciences
RD: Consultant to Millenium, Janssen, Medivation, Bayer, Dendreon, and Roche
The Genomics of Prostate Cancer
Eric A. Klein, MD
Familial and Germline Genetic Influences
Epidemiologic and molecular evidence suggests that prostate cancer has as strong familial
component as demonstrated by epidemiologic studies and germline genetic analysis. The first reports
of familial clustering were published in the mid-20th century and suggested that the risk of developing
prostate cancer was higher in those with an affected first-degree relative (Woolf, 1960). Subsequent
case control and cohort studies have confirmed this observation (Eeles et al, 1997), and twin studies
demonstrate that the inherited component of prostate cancer risk is over 40%, substantially higher
than for other common cancers (Lichtenstein et al, 2000). Relative risk increases according to the
number of affected family members, their degree of relatedness, and the age at which they were
affected (Table 1) (Zeegers et al, 2003). About 15% of all prostate cancer is estimated be caused by
germline factors (Carter et al. 1992).
Table 1. Family history and risk of prostate cancer
Family history
Relative risk
95% Confidence Interval
None
1
Father affected
2.17
1.90-2.49
Brother affected
3.37
2.97-3.83
First-degree family member affected age
< 65 years at diagnosis
3.34
2.64-4.23
>2 first-degree relatives affected
5.08
3.31-7.79
Second-degree relative affected
1.68
1.07-2.64
Early linkage and segregation studies identified a number of candidate prostate cancer
susceptibility
genes
(HPC1/RNAseL,
HPC2/ELAC,
and
MSR1)
and
loci
(PCAP/1q42.2-43,
CAPB/1p36, and Xq27-28). Most subsequent studies have not replicated initial findings and the role of
these genes/regions is not fully established (Eeles et al, 2013), although a recent population based
study identified variant RNAseL alleles as 1 of 5 predictive of prostate cancer specific mortality (Lin et
al, 2011). More recently, genome-wide association studies (GWAS) have emerged as a new approach
to identify alleles associated with prostate cancer risk in an unbiased fashion, i.e., without prior
knowledge of their position or function. Using this technique > 70 prostate cancer susceptibility risk
alleles, many confirmed in multiple studies, have been identified on chromosomes 2, 3, 4, 5, 6, 7, 8,
10, 11, 12, 13, 17, 19, 22 and X (reviewed in Choudhury et al, 2012 and Eeles et al, 2013), which
account for 25 - 30% of germline-determined risk. Studies in African-American and Japanese
populations have identified additional risk alleles specific to these populations (Haiman et al, 2011 and
Takata et al, 2010).
Reported GWAS for prostate cancer have generally included only common
inherited variants (ie, a minor allele frequency of ~5%) and in total have captured only a small fraction
of the germline component of risk. As a consequence, the predictive value of most single alleles (rarely
> 1.5 times baseline risk) is too low to provide clinical utility as a way of identifying individual men at
risk for developing prostate cancer. One approach to this challenge is to combine multiple risk alleles
into a predictive model, as risk increases with the number of specific alleles carried. One such casecontrol study evaluated the ability of 5 loci (3 on 8q24 and 2 on 17q) to predict the likelihood of prostate
cancer in a population of 3161 men. The odds ratio for prostate cancer in men who carried 4 or 5
alleles was 4.47, and increased to 9.46 for those who carried all 5 alleles and had a positive family
history (Zheng et al, 2008). While this study demonstrates the power of risk information contained
within the germline, its clinical utility is limited by the fact that only a minority of the population (1.4%)
carried all 5 risk alleles, and that the model was unable to distinguish between the risk of low versus
high grade disease. In a follow-up study, adding additional alleles only marginally improved the
predictive value of the model (Sun et al, 2011).
The performance of predictive models based on germline alleles and thus their clinical utility
may improve with the incorporation of rarer variants that confer higher risk. Several such variants, with
a minor allele frequencies
~1%,
have recently been described for prostate cancer. A recurrent
mutation in the coding region of the HOXB13 gene, which maps to an area of interest at 17q21-22
identified by GWAS, was present in 1.4% of
cases compared to only 0.1% of controls and was
significantly more common in men with early-onset, familial prostate cancer (3.1%) than in those with
late-onset, sporadic disease (0.6%) (Ewing et al, 2012). This mutation increases overall risk of disease
almost 5 times, and > 8 times in men under age 55 or with a family history (Witte et al, 2013). Several
studies have suggested a familial co-aggregation of prostate cancer with breast cancer (Goldgaret al,
1994, Thiessen, 1974, Tuliniuset al, 1992), and there is clear evidence that both BRCA1 and BRCA2
carriers are at increased risk of prostate cancer, especially for early onset disease. BRCA1 has been
estimated to increase risk by 1.8 – 3.5 fold and BRCA2 from 4.6 – 8.6 fold in men under 65 (reviewed in
Castro and Eeeles, 2012). BRCA-associated cancers, especially for BRCA2, are also more likely to
present with higher grade, locally advanced, and metastatic disease and have worse cancer-specific
and metastasis-free survival after prostatectomy (Castro et al, 2013).
The relative contribution of common and rare alleles to the overall germline risk of prostate
cancer is illustrated in Figure 1. One interesting observation from GWAS is that most of the variant
alleles that confer increased risk are found in non-coding regions of the genome (Choudhury et al, 2012
and Eeles et al, 2013), such that their underlying mechanism of action is not readily understood.
Another common germline variation, copy number variants, have only recently begun to be studied in
prostate cancer and their biologic and clinical relevance is as yet undetermined (reviewed in Barbieri et
al, 2012).
Using germline information to predict the risk of developing prostate cancer for individuals or on
a population basis has yet to be realized owing to the low penetrance of relevant alleles in the general
population, cost, lack of ability of most alleles (BRCA being a notable exception) to predict for disease
that is biologically significant, and lack of evidence that targeted prevention strategies or early
intervention will have a meaningful impact on outcome. This body of knowledge, however, sets the
stage for improved screening, prevention and intervention strategies as the biologic function of each
risk allele is understood.
Somatic Molecular Genetics
Prostate cancer is unique among solid tumors in that it exists in two forms: a histologic or
clinically occult form, which can be identified in approximately 30% of men > 50 years and 60-70% of
men > 80, and a clinically evident form, which affects approximately 1 in 6 U.S. men. Latent prostate
cancer is believed to have a similar prevalence worldwide and among all ethnicities, whereas the
incidence of clinical prostate cancer varies dramatically between and within different countries. For this
reason, an understanding of prostate cancer etiology must encompass the steps leading to both the
initiation of histologic cancer and progression to clinically evident disease. The exact molecular
relationship between latent and clinical cancers is not known, and it is likely that the progression from
the former to the latter is a biologic continuum with overlap in the associated molecular events.
Mutations, down-regulation by promoter methylation and other mechanisms, and protein modification
have all been implicated in progression of prostate cancer.
Substantial evidence exists that prostate cancer arises and progresses by core genetic
alterations that activate oncogenes and inactivate tumor suppressors. These changes result most
commonly from epigenetic and structural genomic changes, including amplification, deletion, somatic
copy number aberrations, and chromosomal rearrangements that result in gene fusions with novel
biologic properties. Unlike many metabolic diseases, the incidence of point and missense mutations
resulting in altered proteins are rare in prostate cancer, estimated to occur in only about 1% of primary
tumors (Taylor et al, 2010). As noted earlier, GWAS has shown that many germline mutations occur in
noncoding regions of the genome, highlighting the potential role of regulatory molecules such as
microRNA and long non-coding (lnc) RNA, suggesting an even deeper biologic complexity. A plethora
of studies using next generation sequencing, microarray data, and functional studies has led to an
emerging comprehensive understanding of the temporal genomic events that occur in prostate cancer
development and progression to the lethal phenotype of metastatic castrate resistant disease, and an
emerging molecular classification according to whether ETS family gene fusions are present or absent
(Barbieri et al, 2010, Barbieri and Tomlins, 2014, Figure 2). This section will highlight the most well
characterized genomic events in early stage prostate cancer; for a comprehensive overview the reader
is referred to a number of excellent primary sources (Barbieri and Tomlins, 2014, Taylor et al, 2010;
Frank et al, 2013, Presner and Chinnaiyan, 2011; Jerómino et al, 2011 )
Epigenetic Changes
Epigenetic events affect gene expression without altering the actual sequence of DNA. Known
mechanisms include DNA hyper- and hypo-methylation, chromatin remodeling, and microRNA (miRNA)
and long-noncoding RNA (lncRNA) regulation.
DNA hypermethylation generally causes gene silencing and is the most well characterized
epigenetic alteration in prostate cancer, affecting > 50 genes across a diverse number of basic cellular
processes including
hormone response (ERαA, ERβ, & RARβ), signal transduction (EDNRB &
SFRP1), cell cycle control (CyclinD2, & 14-3-3σ), DNA repair (GSTpi, GPX3 & GSTM1), inflammatory
response genes (PTGS/COX2), tumor suppressor genes (APC, RASSF1α, DKK3, p16INK4α, Ecadherin, & p57WAF1), tumor invasiveness (CD44) and apoptosis (Li et al, 2005 and Jerómino et al,
2011). DNA hypomethylation, which usually affects areas of the genome distinct form hypermethlyated
regions, causes activation of oncogenes and leads to genetic instability and has been reported for
genes associated with tumor progression (CAGE, HPSE, & PLAU) (Li et al, 2005).
Promoter
methylation of some genes is influenced by diet and age, and is frequently seen in high grade PIN and
morphologically normal prostate tissue, suggesting that these events are drivers early in the
development of prostate cancer (Henrique et al, 2006). Both hypo- and hypermethylation define a field
cancerization effect in normal prostate tissue, as revealed by methylation microarray analysis of tumor
–associated and non-tumor associated normal prostatic tissue (Yang et al, 2013). Clinical studies have
shown that quantitative methylation analysis of the GSTP1, APC, PTGS2, RASSF1α, MDR1, p16, and
MGMT genes can improve sensitivity and specificity for the diagnosis of cancer (Dobosy, et al. 2007).
These observations have clinical utility, as demonstrated by a study that showed that the methylation
status of GSTP1, APC and RASSF1a on prostate needle biopsy can be used to predict the likelihood
of cancer on subsequent biopsy with a negative predictive value of 90% (Stewart el, 2013).
Chromatin remodeling and histone post-translational modifications are also important epigenetic
mechanisms of gene deregulation in prostate cancer. A number of histone-modifying enzymes have
been reported to be altered, the best characterized of which is the histone methyltransferase polycomb
protein EZH2. EZH2 overexpression is correlated with promoter hypermethylation leading to gene
silencing and is associated with higher proliferation rates and disease recurrence (van Leenders et al,
2007). Other histone modifiers, including histone deacetylators (HDACs), are upregulated in prostate
cancer and are targets for both prevention and therapy using agents that can inhibit or reverse their
effects. Histone acetylation also appears important in regulating AR function (Jerómino et al, 2011).
Newly discovered forms of noncoding RNA species, including miRNA and lncRNA, affect posttranscriptional gene expression. miRNA are typically 18–25 nucleotides long and act by binding to and
thereby silencing
2009).
messenger RNAs (mRNA) that have complementary sequences (Garzon et al,
lncRNA are species of > 200 nucleotides that regulate gene expression by a variety of
mechanisms. While numerous miRNAs have been demonstrated to affect the cell cycle, intracellular
signaling, DNA repair, and adhesion/migration in prostate cancer, their main effects seem to be on
suppression of apoptosis and AR regulation (Catto et al, 2011). lncRNA are emerging as molecules
with fundamental biologic and clinical importance in prostate cancer: PCA3 is a lncRNA that can be
detected in urine after a DRE and has clinical utility both in cancer detection and deciding on the need
for subsequent biopsy after an initial negative biopsy (Marks et al, 2007); the expression levels of the
lncRNA SChLAP1 has been shown to associated with metastasis and prostate-cancer specific mortality
after radical prostatectomy (Presner et al, 2013); and a previously unknown lncRNA called PRNCR1
(prostate cancer non-coding RNA 1) has been isolated from the “gene desert” region of 8q24 (the
germline susceptibility locus most repeatedly identified in GWAS, is overexpressed in PIN and cancer,
and causes ligand-independent activation of the AR (Chung et al, 2011 and Yang et al 2013).
There is a complex interplay between the described epigenetic mechanisms in prostate cancer.
For example, several miRNAs are known to be regulated through promoter methylation or
hypomethylation and some miRNAs control the expression of histone modifying enzymes. At another
level of complexity, both miRNAs and EZH2 independently interact with the ETS-genes fusion axis
(Jerómino et al, 2011).
Androgen Receptor
As discussed earlier, germline polymorphisms of the AR are linked epidemiologically to prostate cancer
risk. The role of AR is well established in the progression of castrate resistant prostate cancer, which is
characterized by AR point mutations and amplification, alternative splice mechanisms, and ligand
promiscuity which make it exquisitely sensitive to low levels of intratumoral androgen and/or
constitutively active (Scher and Sawyers. 2005). While these lesions are absent in early stage disease,
dysregulation of the AR signaling axis may occur earlier in disease progression, involving activating
mutations in FOX1A and amplification of
NCOA2 that increase androgen dependent proliferation
(Barbieri and Tomlins, 2014). The PI3K/Akt pathway has reciprocal interactions with AR, such that
inhibition of one activates the other to maintain tumor viability and suggesting that blocking both
pathways simultaneously may be needed for therapeutic efficacy (Carver et al, 2011). Finally, whole
genome analysis has demonstrated that rearrangement break points are more common near AR
binding sites,
suggesting AR-mediated transcription brings together distant genomic loci and
predisposes to genomic rearrangements (Berger et al, 2011). For example, androgen signaling
promotes co-recruitment of AR and topoisomerase II beta (TOP2B) to sites of TMPRSS2-ERG genomic
breakpoints, triggering DNA double strand breaks and resulting in de novo production of TMPRSS2ERG fusion transcripts (Haffner et al, 2010). These observations suggest that AR-mediated
transcriptional activity acts as an early driver of genomic rearrangements in prostate cancer, and
reinforces AR-mediated transcription as a critical signaling pathway in both primary and advanced
disease (Barbieri and Tomlins, 2014).
Gene fusions
Gene fusions resulting from chromosomal translocations are the most common genetic alteration in
human cancers (Futreal et al, 2004). These were previously thought to be an oncogenic mechanism
exclusively limited to hematologic malignancies and sarcomas, as exemplified by the BCR-ABL1 fusion
protein in chronic myeloid leukemia. In 2005, recurrent genomic rearrangements in prostate cancer
were identified, resulting in the fusion of the 5’ untranslated end of TMPRSS2 (an androgen responsive,
prostate specific, transmembrane serine protease gene) to members of the ETS family of oncogenic
transcription factors (Tomlins, et al. 2005). Since then other important gene fusions involving the RAF1
kinase family and SPINK1 have been described, highlighting the fundamental importance of this
genetic mechanism in the genesis of prostate cancers (Rubin et al, 2011 and Figures 3 and 4). These
fusions, and other gross chromosomal rearrangements, occur by a process termed chromoplexy,
where in translocations and deletions arise in an interdependent manner, and disrupts multiple cancer
genes in a coordinated fashion (Baca et al, 2013).
ETS family gene fusions
The most common fusion identified in localized prostate cancer involves TMPRSS2 or other promoters
(SLC45A3, HERPUD1, or NDRG) fused to ERG (ETS-related gene,) in 50 - 60% of patients (KumarSinha et al, 2008 and Rubin et al, 2011). Gene fusions involving other members of the ETS family,
most commonly ETV1 (5-10%), ELK4 (5%) , ETV4 (2%) and ETV5 (2%) also occur. Both TMPRSS2
and SLC45A3 are androgen responsive such that fusion of either of these genes to a growth-promoting
gene of the normally androgen indifferent ETS family brings a powerful signal for cellular growth under
androgen control. These fusions are not observed in benign prostate tissue or PIA, but are present in
prostate stem cells, high grade PIN and early stage, low grade prostate cancer suggesting this is an
early and seminal event in prostate tumorigenesis that may drive the transition from PIN to cancer
(Polson et al, 2103 and Rubin et al, 2011). Recent data from animal models suggests that defects in
the PTEN/PI3K/Akt pathway in the presence of TRMPSS2:ERG fusions drive early tumor progression,
the former stimulating proliferation and the latter cell migration that together may result in a more
aggressive phenotype (Carver et al, 2009). However, there is mixed data on whether the presence of
TMPRSS2:ERG fusions affect prognosis (reviewed in Rubin et al, 2011), such that tumor
aggressiveness may not be determined by the fusion alone but by the presence of the fusion and which
other specific genetic defects are present in an individual tumor.
The high specificity of TPMRSS2:ERG fusions for cancer makes it an attractive target for clinical
use. Fusion transcripts can be detected in the urine and clinical evidence suggests that when combined
with an assay for PCA3 its use can improve the detection of cancer in screened populations over PSA
alone (Tomlins et al, 2011). Some data suggests that quantification of TPMRSS2:ERG fusion in urine
can predict both tumor volume and tumor aggressiveness, perhaps making it useful for selecting
appropriate candidates for active surveillance (Lin et al, 2013). It has been observed that not all tumor
foci within a prostate harbor ETS fusions, so that a positive urine for TRMPSS2:ERG fusion in the face
of a negative biopsy would suggest that cancer was missed due to sampling error, and that additional
evaluation with MRI or repeat biopsy is indicated. The presence of gene fusions that occur only in
cancer also makes them targets for novel therapies (Figure 3).
Other gene fusions
As noted, prostate cancers also are rarely observed to contain fusions involving SPINK1 and RAF
kinases. SPINK1 fusions occur in about 10-15% of cancers, exclusively in ETS fusion negative tumors,
and in cell lines seem to drive tumor invasion (Tomlins et al, 2008). Fusions involving RAF kinases are
even more rare and also define another ETS-fusion negative phenotype that is associated with
aggressive cancers (Palanisamy et al, 2010). Both examples likely represent alternative growth
pathways for ETS-fusion negative tumors and may represent distinct phenotypes (Figure 2).
NKX3.1
NKX3.1 is an androgen-regulated and prostate-specific gene belonging to the homeobox gene
family that protects against DNA damage and promotes DNA repair. Decreased expression of this gene
by mutation, promoter methylation, or post-transcriptional events leads to epithelial DNA damage and
increased rates of proliferation. Loss of NKX3.1 function is seen in areas of bacterial-induced prostatitis
in a mouse model Khalili et al., 2010) and in human PIA, PIN and most prostate cancers and is likely an
early event in prostate tumorigenesis (Bethel et al, 2006 and Bowen et , 2013).
Phosphoinositide 3-kinase (PI3K) pathway
PI3K is one of the most frequently dysregualted signaling pathways in human cancer and plays an
important role in both early and late stage prostate cancer, with alterations occurring in 25 -70% of
tumors (Barbieri et al, 2013). The pathway may be activated by several mechanisms and results in
alterations in proliferation, cell survival, and invasion. Loss of function mutations in PTEN and PHLPP1,
and amplification and gain of function mutations in PIK3CA are the commonest mechanisms of PI3K
activation in prostate cancer. PTEN deletions occur in about 40% of primary tumors, are a central
mechanism of tumor progression, and are associated with the risk of advanced disease and poor
prognosis (Frank et al, 2013).
SPOP mutations
Mutations in SPOP, which encodes a subunit of a ubiquitin ligase, are the most common
point
mutations in primary prostate cancer, with a frequency of 6% to15% (Barbieri and Tomlins, 2014).
Tumors with SPOP mutations have several unique molecular characteristics. They do not occur in ETS
fusion positive tumors or in those with p53 abnormalities, usually lack defects in the PI3K pathway, and
typically contain deletions in the CHD1 gene and at 6q21. CHD1 encodes a DNA helicase binding
protein that regulates transcription epigenetically by chromatin remodeling, and CHD1 negative tumors
have an increased frequency of chromosomal rearrangements. Like RAF kinase and SPINK1
associated tumors, SPOP and CHD1 mutations may define a distinct molecular subtype of prostate
cancer (Figure 2).
p53
The well known tumor suppressor TP53 activates the transcription of genes involved in cell cycle arrest,
DNA repair, and apoptosis, and its dysregulation results in improved cell survival, genomic instability
and proliferation. About 25–30% of clinically localized cancers have lesions in in this gene. Whole
genome analysis suggest that in some cases disruption of p53 occurs early in tumorigeneis, following
deletion of NKX3-1 or FOXP1 and fusion of TMPRSS2 and ERG (Baca et al, 2013).
An Integrated Model of Prostate Cancer Tumorigenesis
A comprehensive, integrated model of the genetic and environmental events that underlies the
genesis and progression of prostate cancer from germline susceptibility to castrate resistant metastatic
disease is now emerging (Figure 2). Early events in genetically susceptible men include environmental
insults such as diet and infection that result in inflammatory insults to DNA integrity in prostate
epithelium. Early genetic events that fuel the progression of precursor lesions to early cancers include
NKX3.1 deletion and ETS fusion or alternatively, mutations in SPOP and FOXA1 in ETS-negative
tumors. Mutations in classical tumor suppressors such as p53 follow, leading to inactivation of the
PI3K/PTEN/Akt pathway and disease progression, culminating in the multi-faceted dysregulation of AR
function and signaling that leads to lethal disease. While many gaps in understanding this process still
exist, we are on the threshold of having a detailed molecular map with temporal sequencing that will
allow advances in the most important clinical challenges facing the field - improved identification of
those at risk of disease development who will be the best candidates for chemoprevention; improved
identification of those with indolent tumors who can avoid or delay initial therapy; biologic measures of
disease progression that identify those who need therapy; and targeted molecular therapy for those
with progressive disease.
SUGGESTED READING
Barbieri CE, Tomlins SA: The prostate cancer genome: Perspectives and potential. Urol Oncol.
2014;32:53
Eeles R, Goh C, Castro E, Bancroft E, Guy M, Olama AA, Easton D, Kote-Jarai Z: The genetic
epidemiology of prostate cancer and its clinical implications. Nat Rev Urol. 2013 Dec 3. doi:
10.1038/nrurol.2013.266. [Epub ahead of print]
Li LC, Carroll PR, Dahiya R. Epigenetic changes in prostate cancer: Implication for diagnosis and
treatment. J Natl Cancer Inst 2005; 97: 103-15.
Tomlins SA, Rhodes DR, Perner S, Dhanasekaran SM, Mehra R, Sun XW, et al Recurrent fusion of
TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 2005; 310: 644-8.
Key Points
Familial and Germline Genetic Influences
•
Both genetics and environment are important in the origin and evolution of prostate cancer.
•
Genome-Wide Association Studies (GWAS) have identified multiple chromosomal loci and
specific variant alleles in germline DNA that confer risk of getting prostate cancer.
•
For commonly inherited variants, the predictive value is rarely > 1.5 times baseline risk, which
is too low to provide clinical utility as a way of identifying individual men at risk for developing
prostate cancer. Models that include multiple risk loci or less common alleles that confer greater
risk will be necessary to accomplish individual risk prediction.
•
HOXB13 and BRCA are two genes that substantially increase individual risk. BRCA-related
tumors present with more aggressive clinical features.
Somatic Molecular Genetics
•
The primary androgen of the prostate is dihydrotestosterone (DHT), formed by the action of
5α-reductase on testosterone. Functional type-2 5α-reductase is a prerequisite for normal
development of the prostate and external genitalia in males and insufficient exposure of the
prostate to DHT appears to protect against the development of prostate cancer.
•
Prostatic stem cells are precursors with multilineage differentiation potential that give rise to
all 4 cell types of prostatic epithelium. Stem cells can repopulate damaged or post-therapy
depleted cancer epithelial cells, and may give rise to prostate cancer directly.
•
Prostate cancer arises and progresses by core genetic alterations that activate oncogenes
and inactivate tumor suppressors. These changes result most commonly from epigenetic
and structural genomic changes, including amplification, deletion, somatic copy number
aberrations, and chromosomal rearrangements that result in gene fusions with novel
biologic properties.
•
Epigenetic regulation of gene expression by promoter methylation, hypomethylation
chromatin remodeling is important in prostate cancer development and progression.
•
MicroRNA (miRNA) and long noncodingRNA (lncRNA) are also important epigenetic
mechanisms of modulating tumor growth and progression.
•
The androgen receptor plays a central role in prostate cancer development and progression.
•
Gene fusions, especially those involving androgen sensitive promoters like TMRPSS2 and
the ETS family of oncogenic transcription factors, are fundamental drivers of prostate cancer
initiation and progression.
•
Somatic mutations in a variety of genes with diverse biological functions have been
implicated in prostate cancer development and progression
•
Mutations, amplification, and ligand promiscuity of the AR are important determinants of
progressive castrate-resistant prostate cancer
Figure 1
Figure 2
Figure 3
Figure 4
3/2/2014
P1 2014
• 73 year old previously robustly healthy man
• Not seen a doc in 35 year
• Develops bilateral LE edema, presents to primary
care
• DVT’s ruled out, PSA 116
• Urologic evaluation lead to Trus bx high volume
Gleason 4 + 4
• Normal cbc and chemisties
• Bone scan: DJD only
1
3/2/2014
Evolving Initial Management of
Metastatic Disease
• US Intergroup Study E 3805
• Asked the simple question of earlier
integration of docetaxel
• Patients randomized to ADT plus/minus 6
cycles of docetaxel ( without prednisone)
ECOG 3805 (CHAARTED)
Press Release 12/9/13



790 men with metastatic disease randomized
to CAB with or without 6 cycles of docetaxel
Planned interim analysis: 69% receiving
ADT/chemorx were alive after 3 years
compared to 52.5
Greatest impact (63.4 percent vs. 43.9 %) in
men with high volume disease: visceral mets
and/or four or more bone lesions
P2 2014
• 55 year old presents with progressive back
pain and weight loss, PSA 114
• 3 years prior, iPSA 14, Gleason 4 + 5 on bx
• EBRT 72 Gy + 2 years CAB
• On this presentation, testosterone 466
• Imaging ordered
– CT abd/pelvis bone findings no nodes
2
3/2/2014
P2 2014
• ADT initiated
• Initial pain response, PSA nadir 23, 7 months
later new pain, PSA 78 testosterone 15 ng/dl
• Hgb 9.8, 15 lb weight loss
• CT abd/pelvis no nodes/visceral disease
3
3/2/2014
P2 2014
You Recommend
A. Discontinue LHRH, start docetaxel
B. Docetaxel
C. Discontinue LHRH, start
abiraterone/prednisone
D. Abiraterone/prednisone
E. Enzalutamide
F. Radium 223
P2 2014
• Abiraterone 1000 mg/day plus prednisone 5
mg bid started
• One week later, back pain resolved
• Six weeks later PSA 21 (nadirs at 5.6)
• Tolerated rx with only minor LE edema
• 7 months following initiation of Abi/pred, PSA
54, patient without disease-related symptoms
P2 2014
You Recommend
A.
B.
C.
D.
E.
F.
Discontinue Abiraterone, start docetaxel
Discontinue Abiraterone, start Enzalutamide
Add Enzalutamide to Abiraterone
Add docetaxel to Abiraterone
Continue Abiraterone
Something else
4
3/2/2014
Steroid Synthesis
Cholesterol
Low-dose steroid replacement decreases ACTH
and minimizes mineralocorticoid-related toxicity
Desmolase
Pregnenolone
Progesterone
Deoxycorticosterone
Corticosterone
11-Deoxycortisol
Cortisol
Aldosterone
X
CYP17
17α-hydroxylase
17α-OHpregnenolone
17α –OHprogesterone
X
ACTH
↓
CYP17
C17,20-lyase
5α-reductase
Androstenedione
DHEA
Testosterone
DHT
CYP19: aromatase
Estradiol
Attard G, et al. J. Clin. Oncol. 26: 4563–4571, 2008
Attard G, et al. J. Clin. Oncol. 26: 4563–4571, 2008
COU-AA-301 Study Design
Patients
• 1195 patients with
progressive, mCRPC
• Failed 1 or
2 chemotherapy
regimens, one of which
contained docetaxel
R
A
N
D
O
M
I
Z
E
D
2:1
Efficacy endpoints (ITT)
Abiraterone1000 mg daily
Prednisone 5 mg BID
N=797
Primary end point:
• OS (25% improvement; HR
0.8)
Secondary end points (ITT):
Placebo daily
Prednisone 5 mg BID
n=398
• TTPP
• rPFS
• PSA response
• Phase 3, multinational, multicenter, randomized, double-blind,
placebo-controlled study (147 sites in 13 countries; USA, Europe,
Australia, Canada)
• Stratification according to:
–
–
–
–
ECOG performance status (0-1 vs. 2)
Worst pain over previous 24 hours (BPI short form; 0-3 [absent] vs. 4-10 [present])
Prior chemotherapy (1 vs. 2)
Type of progression (PSA only vs. radiographic progression with or without PSA progression)
Clinicaltrials.gov identifier: NCT00638690.
5
3/2/2014
COU-AA-301: Abiraterone Acetate Improves Overall
Survival in mCRPC
HR = 0.646 (0.54-0.77) P< 0.0001
100
Abiraterone acetate:
14.8 months (95%CI: 14.1, 15.4)
Survival (%)
80
60
40
Placebo:
10.9 months (95%CI: 10.2, 12.0)
20
2 Prior Chemo OS:
14.0 mos AA vs 10.3 mos placebo
1 Prior Chemo OS
15.4 mos AA vs 11.5 mos placebo
0
0
100
300
200
500
400
600
700
Days from Randomization
de Bono J et al: N Engl J Med 364:19952005, 2011
Adverse Events of Special Interest
All
Grades
Grade
3
Grade
4
Placebo
Plus Prednisone
(n=394)
All
Grade Grade
Grades
3
4
241 (31)
16 (2)
2 (<1)
88 (22)
4 (1)
135 (17)
27 (3)
3 (<1)
33 (8)
3 (1)
0
Cardiac disorders
106 (13)
26 (3)
7 (1)
42 (11)
7 (2)
2 (<1)
LFT abnormalities
82 (10)
25 (3)
2 (<1)
32 (8)
10 (3) 2 (<1)
Hypertension
77 (10)
10 (1)
0
31 (8)
1 (<1)
Abiraterone Acetate
Plus Prednisone (n=791)
Adverse Event, no.
patients (%)
Fluid retention and
edema
Hypokalemia
0
0
• Adverse events associated with elevated mineralocorticoid levels, cardiac events, and LFT
abnormalities were deemed of special interest
• These events were more common in the abiraterone acetate group (55% vs 43%; P<0.001) but
were largely mitigated by the use of low-dose prednisone
deBonoJS et al. N Engl J Med. 2011;364(21):1995-2005.
Overall Study Design of COU-AA-302
Patients
• Progressive chemonaïve mCRPC patients
(Planned N = 1088)
• Asymptomatic or
mildly symptomatic
R
A
N
D
O
M
I
Z
E
D
Efficacy end points
AA 1000 mg daily
Prednisone 5 mg BID
(Actual n = 546)
1:1
Placebo daily
Prednisone 5 mg BID
(Actual n = 542)
Co-Primary:
• rPFS by central review
• OS
Secondary:
• Time to opiate use (cancerrelated pain)
• Time to initiation of
chemotherapy
• Time to ECOG-PS
deterioration
• TTPP
•
Phase 3 multicenter, randomized, double-blind, placebo-controlled study conducted at 151 sites in 12 countries; USA,
Europe, Australia, Canada
•
Stratification by ECOG performance status 0 vs 1
6
3/2/2014
N Engl J Med 2013;368:138-48.
Ryan CJ, et al. N Engl J
Med 2013;368:138-48
MDV3100/Enzalutamide
1.
MDV3100 is an oral investigational drug
rationally designed as a new hormonal agent
to target androgen receptor (AR) signaling, a
key driver of prostate cancer growth.
2.
MDV3100 is the first in a
new class of Androgen Receptor Signaling
Inhibitors that affects
multiple steps in the
androgen receptor
signaling pathway.
Reduces the
efficiency of its
nuclear translocation
and impairs both
DNA binding to
androgen response
elements and
recruitment of
coactivators
T
T
1
Inhibits Binding of
Androgens to AR
M DV 3 1 0 0
AR
Cell cytoplasm
2
Inhibits Nuclear
Translocation of AR
Cell nucleus
3
AR
Inhibits Association
Of AR with DNA
Scher H, et al. J Clin Oncol 30, 2012 (suppl 5; abstr LBA1)
7
3/2/2014
Phase III Trial of MDV-3100 vs Placebo in
CRPC (AFFIRM)
R
A
N
D
O
M
I
Z
E
Progressive prostate cancer
1-2 prior chemotherapy regimens, 1
must have contained docetaxel
No prior abiraterone or ketoconazole
MDV-3100 160 mg PO QD
Placebo PO QD
2:1
Primary endpoints: Overall survival
Secondary endpoints: PFS and pain control
Available at: http://www.clinicaltrials.gov/ct2/show/NCT00974311
Accessed April 18, 2010.
MDV3100 Prolonged Survival by a Median of
4.8 Months in the Phase 3 AFFIRM Trial
HR = 0.631 (0.529, 0.752) P < 0.0001
37% Reduction in Risk of Death
100
90
MDV3100: 18.4 months
(95% CI: 17.3, NYR)
Survival (%)
80
70
60
50
40
30
Placebo: 13.6 months
(95% CI: 11.3, 15.8)
20
10
0
MDV3100
800
775
701
627
400
211
72
7
0
Placebo
399
376
317
263
167
81
33
3
0
Scher H, et al. 2012 Genitourinary Cancers Symposium
Adverse Events of Interest
All Grades
Grade ≥ 3 Events
MDV3100
(n = 800)
Placebo
(n = 399)
MDV3100
(n = 800)
Placebo
(n = 399)
Fatigue
33.6%
29.1%
6.3%
7.3%
Cardiac Disorders
6.1%
7.5%
0.9%
2.0%
0.3%
0.5%
0.3%
0.5%
LFT Abnormalities*
1.0%
1.5%
0.4%
0.8%
Seizure
0.6%
0.0%
0.6%
0.0%
Myocardial Infarction
*Includes terms hyperbilirubinaemia, AST increased, ALT increased, LFT abnormal,
transaminases increased, and blood bilirubin increased.
Scher H, et al. 2012 Genitourinary Cancers
Symposium
8
3/2/2014
Phase III Trial of MDV-3100 vs Placebo in CRPC
(PREVAIL)
1717 men with
progressive
mCRPC
Asymptomatic/
mildly
symptomatic
Chemotherapy
naïve
Steroids
allowed by not
required
R
A
N
D
O
M
I
Z
E
D
MDV-3100 160 mg
PO QD
Placebo PO QD
Co-Primary
Endpoints
Overall Survival
Radiographic
Progression-Free
Survival rPFS
1:1
9
3/2/2014
ELM-PC5 Study Design
Patients with
mCRPC that
progressed postdocetaxel, and
PSA ≥ 2ng/mL
at screening
R
A
N
D
O
M
I
Z
E
D
Enrolled
N = 1099
Orteronel 400mg BID
Prednisone 5mg BID
n = 734
Endpoints
Primary:
• OS
Key Secondary:
• rPFS
• PSA response
• Pain response
Placebo BID
Prednisone 5mg BID
n = 365
2:1
Dreicer R, et al. GU Symposium Abst 7 2014
Primary Endpoint: Overall Survival
DLL1
Orteronel + prednisone
P = 0.18976
Prednisone
HR: 0.886 (0.739, 1.062)
Median: 17.0 mo vs 15.2 mo
Events: 330 vs 182

Median follow-up time: 10.6 months
Regional analysis of OS
DLL2
non-Europe/NA
(N = 397)
Orteronel + prednisone
Prednisone
Europe
(N = 586)
Orteronel + prednisone
Prednisone
OS
nonEurope/NA
Europe
North
America
P value
0.019
0.721
0.680
HR
0.709
1.048
0.889
(0.531, 0.946)
(0.810, 1.356)
(0.508, 1.557)
Median
(mo)
15.3
10.1
18.3
17.8
20.9
16.9
# of
events
117
77
178
86
35
19
North America
(N = 112)
Orteronel + prednisone
Prednisone
10
Slide 29
DLL1
K-M figure to be redrawn:
- delete bottom left corner data
- replace with clear/larger, bolded font data inset (P-value, HR, Median, Events)
- color-coding of orteronel (blue) vs placebo (green) # of subjects at risk
Dawn Lee, 1/2/2014
Slide 30
DLL2
K-M figures to be redrawn:
- delete bottom left corner data
- add new region titles
- color-coding of orteronel (blue) vs placebo (green) # of subjects at risk
- [For non-Europe/NA plot: Enlarge+ bold font for '265' and '132' under '# of subjects at risk' at time 0]
- [For Europe + NA plot: only bold font for 1st column #s' under '# of subjects at risk']
New data table for all regional data instead (P-value, HR, Median, Events)
Dawn Lee, 1/6/2014
3/2/2014
P3 2014
• 57 year old 4 years out from RRP for Gleason 4
+ 3, 1 SV +, I PSA 8.9
• Declined adjuvant radiotherapy
• PSA 6 months post op 0.35, when 14.2 started
on monotherapy with bicalutamide
• PSA nadired to 7, now 44, bone scan NED, he
is asyptomatic
• CT abd/pelvis
11
3/2/2014
P3 2014
• He is evaluated by both urology and medical
oncology, testosterone suppression
recommended
• He has no interest, returns 3 months later
with:
The genomic explosion over the last 10 years
is due to two major events: the completion of
the Human Genome Project in 2003 and the
development of high-throughput DNA
sequencing technologies. The sequencing
technologies have made it possible to
characterize complex genomic signatures in a
rapid and affordable manner. Numerous startups and existing biotechnology companies
have joined in what has become the “omics”
revolution that includes genomics,
proteomics, metabolomics, and
pharmacogenomics. Genomic testing is
becoming the cornerstone of personalized
medicine and pharmacogenomics are part of
prescribing several newer oncology
medications.
Gomella LG Can J Urol 21:7091 2014
12
3/2/2014
Progress in cancer genomics has raised hopes
of increased precision in the identification of
patients suitable for targeted therapies
tailored to their genotypes. Effective
implementation of such precision medicine
will need to take into account diversity
between and within tumours to mitigate
tumour evolution through space and time
Swanton C Lancet Oncol Feb 2014
Burrell RA, et al. Nature 501:338, 2013
P4 2014
• 75 year old active otherwise healthy man
• Presented with metastatic disease, CAB,
followed upon progression with
abiraterone/prednisone
• Initial PSA at presentation 1244, nadired to 35
with ADT
• Now increasing fatigue, very mild bone pain,
PSA 3566
P4 2014
• Hgb 10.2
• Abd/pelvis CT NED
• Bone scan
13
3/2/2014
Radium-223 Targets Bone Metastases
Radium-223
acts as a
calcium mimic
Naturally
targets new
bone growth in
and around
bone
metastases
Ca
Ra
Radium-223 is
excreted by the
small intestine
ALSYMPCA (ALpharadin in SYMptomatic Prostate
CAncer) Phase III Study Design
TREATMENT
PATIENTS
• Confirmed
symptomatic
CRPC
• ≥ 2 bone
metastases
• No known
visceral
metastases
• Postdocetaxel or
unfit for
docetaxel
6 injections at
4-week intervals
STRATIFICATION
• Total ALP:
< 220 U/L vs ≥ 220 U/L
• Bisphosphonate use:
Yes vs No
• Prior docetaxel:
Yes vs No
R
A
N
D
O
M
I
S
E
D
Radium-223 (50 kBq/kg)
+ Best standard of care
Placebo (saline)
+ Best standard of care
2:1
N = 922
14
3/2/2014
Parker C, et al. N Engl J Med 2013;369:213-23
15