Genomic Differences in Prostate Cancer Stages Are Coming Into Focus

OncologyLive, Vol. 17/No. 18, Volume 17, Issue 18

Prostate cancer is slow-growing and readily curable in many cases, but the propensity for some tumors to develop into an aggressive, metastatic form that becomes resistant to androgen-targeting therapies continues to present a significant obstacle.

Prostate cancer is slow-growing and readily curable in many cases, but the propensity for some tumors to develop into an aggressive, metastatic form that becomes resistant to androgen-targeting therapies continues to present a significant obstacle.

The current methods for definitively identifying patients who are at risk of progression to metastatic disease are based on histological and clinical differences and the expression of specific biomarkers.

The Prostate Cancer Challenge

But these strategies are imperfect, and researchers are turning to genome sequencing to tease out molecular differences that might underlie the clinical heterogeneity in prostate cancer progression. They hope to identify new markers to guide diagnosis and treatment and new targets for drug development.Understanding what drives the variability in clinical behavior in prostate cancer and developing effective treatments for patients with metastatic disease are among the greatest challenges today.

Measurement of prostate-specific antigen (PSA) levels, a protein produced by the epithelial cells of the prostate gland, is the gold standard for determining which patients are at risk for developing aggressive disease and therefore require aggressive treatment. But the use of PSA to guide treatment remains controversial because of the tendency for overdiagnosis that often leads to unnecessary treatment and significant morbidity.

The treatment of advanced prostate cancer has been shaped by the knowledge that it is driven by the androgens testosterone and its more active metabolite, 5-dihydroxytestosterone (5-DHT). Reducing the circulating levels of androgens or blocking their cellular effects by inhibiting the activity of the androgen receptor (AR) has become the cornerstone of treatment via surgical or medical castration, the latter through the use of androgen-deprivation therapy (ADT).

Despite initial response, castration-resistant prostate cancer (CRPC) inevitably develops, in which the tumor progresses in spite of low circulating levels of androgens or inhibition of the AR signaling pathway. Insight into the mechanisms underlying the development of CRPC revealed, somewhat surprisingly, that the disease had not become androgen-independent as many suspected, but that the AR pathway is still highly activated in these cancers.

This knowledge has spurred the development of numerous second-generation AR-targeting drugs that are more potent and selective or have novel mechanisms of action, including the AR antagonist enzalutamide (Xtandi) and the CYP17A inhibitor abiraterone acetate (Zytiga). Yet patients still continue to develop resistance to these drugs and there is a pressing need for new therapeutic options.

The Primary Prostate Cancer Landscape


Genome-level studies are starting to provide a comprehensive picture of the molecular basis of prostate cancer at different stages. The first sequencing study was published in 2011 and since then several research entities have added their contributions to the field, including The Cancer Genome Atlas (TCGA) research network. The TCGA analyzed 333 primary prostate cancer samples and reported findings in 2015 (Figure 1).Studies have highlighted significant heterogeneity in the molecular makeup of prostate cancer and have also revealed that the types of genomic aberrations involved are quite unique. The gain or loss of whole chromosomes is relatively rare and the number of individual “spelling errors” in the DNA of key driver genes is also low compared with other solid tumors.

The TCGA study found a median mutation rate of just 0.94 mutations per megabase (range, 0.02-28 per megabase), which is comparable with cancers at the bottom end of the spectrum of mutational load, such as acute myeloid leukemia.

Instead, prostate cancers are a prime example of a class of complex DNA rearrangements called chromoplexy. This is where whole “paragraphs” of DNA break off and move to another part of the genome, creating numerous rearranged or deleted genes as the DNA is stuck back together in new configurations. Since chromoplexy can generate multiple disrupted genes, this can result in numerous oncogenic events occurring in a single cell cycle, giving the cancer cell a significant proliferative advantage.

Somatic Mutations

Chromoplexy has been observed in the majority of prostate tumors examined to date. Genes that have been found to be frequently deleted in the setting of chromoplexy include PTEN, TP53, and RB1, as well as NKX3-1 and CDKN1B. Alterations in these tumor suppressor genes are thought to be early events in prostate oncogenesis.Despite the low mutation rate, several recurrently mutated genes have been detected in primary prostate cancers. The most commonly mutated gene is SPOP, found in 6% to 13% of samples. The gene encodes a component of an E3 ubiquitin ligase complex that tags target proteins with ubiquitin to drive their removal from the cell by proteasomal degradation.

One of its target proteins is thought to be the death-associated protein 6 (DAXX), a transcriptional repressor of histones, histone acetylases, and other histone-associated proteins. Several studies, including the TCGA analysis, have suggested that SPOP-mutant prostate cancers represent a distinct class of tumors and that the SPOP mutations occur early in their development.

The TCGA identified 13 recurrently mutated genes; in addition to SPOP, several other genes previously linked to prostate cancer were among them, including TP53, and PTEN, as well as novel genes such as ZMYM3, an epigenetic regulatory protein not previously implicated in prostate cancer but frequently observed in various pediatric tumors.

ETS Fusions

Mutations in a number of genes involved in histone modification, including KDM6A, MLL2, and MLL3, as well as deletion of CHD1, all of which are involved in the regulation of chromatin states and in transcriptional control, have also been reported in several different studies.The most common oncogenic aberrations observed in prostate cancer overall are gene fusions involving ETS transcription family genes. Found in more than half of all prostate cancers, these mutations most commonly involve the ERG gene, which becomes fused to the constitutively activated promoter of an androgen-regulated gene. The ETS gene’s fusion partner is most often the TMPRSS2 gene, but more than 10 androgen-related genes have been shown to be involved in these oncogenic fusions in prostate cancer.

Like SPOP mutations, ETS fusions are thought to represent an early event in prostate carcinogenesis and to be maintained throughout prostate cancer progression. ETS fusion-positive tumors have been found to be notably distinct from those which do not have ETS fusions in a number of different ways, including in their gene expression signature, the types of copy number alterations they display, and the other genomic rearrangements observed.

Chromoplexy is more common in tumors with ETS fusions, as are PTEN deletions, but these fusions are mutually exclusive with SPOP mutations, suggesting the 2 types of alterations likely define distinct classes of prostate cancer, a hypothesis that was reinforced by the TCGA’s findings. As the most prevalent driver of prostate cancer, ETS fusions have been the focus of significant research seeking to determine any effect on patient prognosis, but thus far data from different studies have been conflicting.

Cell-Signaling Networks

Many of the recurrent genomic aberrations observed in prostate cancer serve to activate pathways involved in prostate cancer development. These include signaling networks associated with chromatin modification, cell cycle regulation, and androgen signaling (Figure 2).

Prostate Cancer Classes

Alterations in the phosphatidylinositol-3-kinase (PI3K) pathway occur in approximately 25% of all prostate tumors; these are predominantly PTEN deletions but other genes in this pathway are also affected. Mitogen-activated protein kinase (MAPK) pathway alterations are also common, although not the typical mutations in BRAF and RAS that are observed in other tumor types. Finally, inactivation of key DNA repair pathways was reported to occur in 19% of prostate cancer samples in the TCGA study.Given the significant heterogeneity among prostate cancers, the definition of subtypes could be potentially helpful to better diagnosis and treatment. The TCGA study identified 7 subtypes that were defined by specific gene fusions or mutations, and were able to place three-quarters of the analyzed tumor samples into 1 of these subtypes.

AR Signaling Dominates Metastatic Disease

The class-defining events were ERG fusions (in 46% of samples), ETV1 fusions (8%), ETV4 fusions (4%), and FLI1 fusions (1%), in addition to SPOP mutations (11%), FOXA1 mutations (3%) and IDH1 mutations (1%). Remarkably, however, this leaves a substantial number of prostate tumors that are driven by molecular events that are not yet clearly defined.Sequencing studies have now begun to highlight some of the genomic distinctions between primary and metastatic prostate cancer. Most striking is the difference in mutational load. One study revealed 4 times as many mutations per megabase on average in a cohort of 150 metastatic CRPC samples as would be expected in primary tumors, with up to 50 mutations per megabase in some metastatic tumor samples.

While metastatic tumors displayed similar genomic alterations and subtype grouping as primary prostate tumors, the frequency with which individual genes and pathways are altered is significantly higher as the disease progresses.

For example, alterations in the PI3K pathway have been observed in nearly half of all metastatic samples, as opposed to 25% of primary cancers. These mutations include loss of PTEN, amplification and fusion of the PIK3CA gene, and activating mutations in AKT1.

Alterations in DNA repair pathways are also more frequent in metastatic prostate cancers. Most notably, metastatic samples appear to have a much higher frequency of BRCA2 mutations, both germline and somatic. Germline BRCA2 mutations were observed in more than 12% of patients with metastatic CRPC, compared with 3% in the TCGA study of primary prostate cancer. These observations have served as the rationale for investigating the use of poly(ADP)-ribose polymerase (PARP) inhibitors in prostate cancer, which target a DNA repair enzyme.

Patients with defects in DNA repair pathways, including mutations in the breast cancer susceptibility genes BRCA1 and BRCA2, are predicted to be highly sensitive to PARP inhibition. Their pre-existing faults in combination with the inhibition of PARP serves a double-whammy to their ability to repair DNA damage that can trigger cell death. Patients with mutations in other DNA repair enzymes, such as ATM, that confer a BRCAness phenotype are also thought to be sensitive to this treatment strategy.

Based on these hypotheses and the frequency of these mutations observed in metastatic prostate cancer, the authors of 1 study predicted that some 20% of patients might respond to PARP inhibition. Recently, the results of the phase II TOPARP-A study of olaparib in patients with metastatic prostate cancer were reported. In a subgroup analysis of 16 patients, the response rate was 88%, including all patients with BRCA2 mutations.

Several genes are selectively mutated or deleted in prostate cancer; most common is the TP53 gene, which could partly explain the increased genomic instability and mutational load. Most notable, however, are AR mutations. Although AR pathway alterations do occur in localized disease, alterations in the AR gene itself are almost exclusively observed in metastatic tumors.

Overall, aberrations in the AR pathway, including AR gene mutations, occurred in more than 60% of tumors. This reinforces the idea that alterations in the AR gene might occur under the selective pressure of treatment with ADT to help maintain AR signaling as a driver of CRPC.

The retinoblastoma protein (RB1) gene is another gene that is selectively altered in metastatic prostate cancer. It is rarely observed in localized disease, but its loss was detected in more than a fifth of metastatic samples. The pRb protein has been suggested to modulate AR signaling and may serve as a central inhibitor of progression to advanced stages of disease.


  1. Barbieri CE, Bangma CH, Bjartell A, et al. The mutational landscape of prostate cancer. Eur Urol. 2013;64(4):567-576.
  2. Barbieri CE, Tomlins SA. The prostate cancer genome: perspectives and potential. Urol Oncol. 2014;32(1):53.e15-e22. doi: 10.1016/j. urolonc.2013.08.025.
  3. Berger MF, Lawrence MS, Demichelis F, et al. The genomic complexity of primary human prostate cancer. Nature. 2011;470(7333):214-220.
  4. Cancer Genome Atlas Research Network. The molecular taxonomy of primary prostate cancer. Cell. 2015;163(4):1011-1025.
  5. Liu W, Xu J. Translation of genomics and epigenomics in prostate cancer: progress and promising directions. Asian J Androl. 2016;18(4):503-504.
  6. Mateo J, Carreira S, Sandhu S, et al. DNA-repair defects and olaparib in metastatic prostate cancer. N Engl J Med. 2015;373(18):1697-1708.
  7. Robinson D, Van Allen EM, Wu YM, et al. Integrative clinical genomics of advanced prostate cancer. Cell. 2015;161(5):1215-1228.