Molecular Evolution of Early-Onset Prostate Cancer Identifies Molecular Risk Markers and Clinical Trajectories

Abstract

Highlights d Clock-like mutation process attributed to APOBEC3 mediates earliest mutations in PC d Identification of four molecular subgroups that stratifies intermediate-risk disease d Rearrangements at the ESRP1 locus associated with aggressive and proliferative cancer d Development of method to predict clinical trajectories of PC from DNA sequencing data

Random forest-based modelling to detect biomarkers for prostate cancer progression

“…139 We applied our model to two independent PCa cohorts for validation of the good prediction 140 rate. We were able to validate our results using the ICGC PCa cohort of early and late onset 141 prostate cancer (n= 222) (Gerhauser et al, 2018). The AUC for the sensitivity analysis was 142 77.1% and the error rate was 35.1% ( Figure 3C).…”

“…Recent high-resolution genome-wide studies have significantly improved our understanding 55 of chromosomal and genetic alterations associated with PCa development, such as the 56 androgen-driven formation of gene fusions between the transmembrane serine protease 57 TMPRSS2 and a member of the oncogenic ETS transcription factor family like ERG in about 58 50% of all PCa cases, and frequent loss of the tumour suppressor gene PTEN (Cancer 59 Genome Atlas Research, 2015;Fraser et al, 2017;Gerhauser et al, 2018;Taylor et al, 2010;60 Tomlins et al, 2005;Weischenfeldt et al, 2013). These events affect signalling pathways and 61 lead to alterations in gene expression programs that have been used for the development of 62 gene signatures (genomic classifiers) as biomarkers for the prediction of PCa prognosis.…”

“…117 Random forest model 118 We applied random forest-based modelling to rank the selected DMS according to their 119 discriminative power (for details see description in the Experimental section). In addition to 120 the DMS, the Purity-Adjusted Epigenetic Prostate Cancer Index (PEPCI) of tumor 121 aggressiveness (Gerhauser et al, 2018) was included in the model. Mean PEPCI was 122 significantly different between the two prognosis groups of the discovery cohort (t-test p-123 value=0.03).…”

Androgen-regulated transcription ofESRP2drives alternative splicing patterns in prostate cancer

“…ESRPs have previously been shown to have a dual role in carcinogenesis with both gain and loss associated with poor patient prognosis (36). ESRP1 has recently been shown to be amplified in an aggressive subgroup of early onset prostate cancer (54). ESRP1 expression is linked to poor survival and metastasis in lung cancer (64), and both ESRP1 and ESRP2 are upregulated in oral squamous cell carcinoma relative to normal epithelium (53).…”

“…Figure supplement 1C). While this manuscript was in preparation, another group used an alternative ESRP1 antibody to show upregulation of ESRP1 in 12,000 prostate cancer tissue microarray tumours(54).We next investigated the effects of ESRP1/2 expression on the biology of prostate cancer cells in vivo. Because of their normal endogenous expression profiles (Figures 1C and 1D), we selected PC3 and DU145 cells to study the effects of ESRP1/ESRP2 protein upregulation on prostate cancer cells.…”