Impact of Therapy on Genomics and Transcriptomics in High-Risk Prostate Cancer Treated with Neoadjuvant Docetaxel and Androgen Deprivation Therapy

Abstract

Purpose: The combination of docetaxel chemotherapy and androgen deprivation therapy (ADT) has become a standard treatment for patients with metastatic prostate cancer. The recently accrued phase III CALGB 90203 trial was designed to investigate the clinical effectiveness of this treatment approach earlier in the disease. Specimens from this trial offer a unique opportunity to interrogate the acute molecular response to docetaxel and ADT and identify potential biomarkers.Experimental Design: We evaluated baseline clinical data, needle biopsies, and radical prostatectomy (RP) specimens from 52 (of 788) patients enrolled on CALGB 90203 at one high volume center. Pathology review, tumor and germline-targeted DNA sequencing (n = 72 genes), and expression profiling using NanoString platform (n = 163 genes) were performed to explore changes in critical prostate cancer pathways linked to aggression and resistance.Results: Three of 52 patients had only microfocal residual cancer at prostatectomy. The most common alterations included TMPRSS2-ERG fusion (n = 32), TP53 mutation or deletion (n = 11), PTEN deletion (n = 6), FOXA1 (n = 6), and SPOP (n = 4) mutation, with no significant enrichment in posttreated specimens. We did not observe AR amplification or mutations. The degree of AR signaling suppression varied among treated tumors and there was upregulation of both AR and AR-V7 expression as well as a subset of neuroendocrine and plasticity genes.Conclusions: These data support the feasibility of targeted and temporal genomic and transcriptome profiling of neoadjuvant-treated prostate cancer with limited formalin-fixed paraffin embedded tissue requirement. Characterization of the heterogeneity of treatment response and molecular outliers that arise posttreatment provides new insight into potential early markers of resistance. Clin Cancer Res; 23(22); 6802-11. ©2017 AACR.

©2017 American Association for Cancer Research.

Figures

Figure 1. Study overview

A) Schematic illustrating clinical trial design and the nature and number of tissue specimens available for downstream analysis. B) Example photomicrographs of H&E stained radical prostatectomy specimens from the treatment arm demonstrating differing degrees of treatment score. See also Supplementary Figure 1. C) Schematic showing the breakdown of genomic and transcriptomic profiling techniques applied to each patient in the cohort. The effect of treatment on serum prostate specific antigen (PSA) levels is also provided.

Figure 2. Somatic mutations detected in diagnostic biopsy cores and matched radical prostatectomy specimens

The bar plot shows minimum tumor cellularity predicted from targeted DNA sequencing of 72 prostate cancer driver genes in biopsy (upper) and radical prostatectomy (RP; lower). For some specimens, no somatic mutations were detected within the target regions meaning that tumor cellularity was unknown (indicated by lack of yellow or green bars). Mutations in frequently altered genes are provided in an oncoprint (also see Supplementary Figure 3). The bottom plot is a heatmap highlighting persistent AR expression in post-treated specimens (seven patients did not have expression data available and are indicated with filled gray boxes)

Figure 3. Differentially expressed genes between treated and untreated patient subgroups

Differentially expressed genes (n=57) were calculated based on a t test with corrected p-values

Figure 4. Differences in the expression of the androgen receptor (AR), AR-variant 7, and PSA between treated and untreated subgroups

A) AR and AR-V7 gene expression in untreated and treated specimens. B) Change of tissue PSA expression (KLK3) versus change in serum PSA levels in untreated and treated patients. R= Pearson’s correlation.