Abstract
Abstract Background: In patients who receive definitive therapy for locally advanced prostate cancer, biochemical recurrence is due to local recurrence and/or the presence of undetected metastatic disease at the time of initial diagnosis and treatment. Approximately 25-40% of patients who receive radiotherapy plus androgen deprivation therapy (ADT) with curative intent are expected to recur within 5-10 years. We hypothesized that applying multiparametric (mp) MRI, gene expression, and genomic analyses to prostate tumors prior to EBRT+ADT would identify biomarkers predictive of recurrence (NCT01834001). Methods: At baseline and 6 months after completion of radiotherapy, mpMRI was performed and lesions were manually contoured. MR/ultrasound-fusion biopsies acquired at baseline targeted to up to three distinct MR lesions per patient. 29 patients with intermediate and high risk localized prostate cancer treated with EBRT+ADT with sufficient baseline biopsy tissue were selected for this analysis (median follow-up: 91 months). Tumor tissues were macrodissected to obtain DNA and RNA from 60 distinct biopsy regions (1-4 per patient). DNA was used for generating whole-exome sequencing libraries (92.5 × median coverage); RNA was converted into cDNA and hybridized to Affymetrix Human Exon 1.0 microarrays. Secondary gene expression cohorts with clinical annotation were obtained from the University of Miami and Queen’s University Belfast. Gene expression signatures were processed using Ingenuity Pathway Analysis (IPA) and the Decipher GRID. Results: Five of the 29 patients experienced biochemical failure by the time of data analysis; four of these patients were identified to have distant metastases with no evidence of local recurrence at the time of biochemical failure. Baseline and posttreatment lesion volumes by MRI and posttreatment ADCmax (apparent diffusion coefficient) were positively associated with recurrence (p<0.05). The Decipher signature positively correlated with these three MR features. Comparison of gene expression using a linear mixed-effect model identified 1,120 differentially expressed transcripts. IPA nominated TGF-beta signaling as the topmost enriched upstream regulator (z=3.32). This result was validated in the Miami (z=2.54) and Belfast (z=3.67) cohorts. By network analysis, genes that independently tracked with the baseline and posttreatment volumes and posttreatment ADCmax MR features similarly converged to the TGF-beta pathway. Co-occurrence of biallelic TP53 alterations with 1- or 2-copy PTEN losses was observed exclusively in patients who recurred. Conclusions: Genomic, transcriptomic and radiomic analyses have nominated predictive biomarkers that identify the subset of patients with locally advanced prostate cancer destined to recur after definitive EBRT+ADT. Identification of molecular features predictive of failure suggests that aggressive pre-existing subclones harboring these alterations have metastatic potential. Emerging systemic therapies including TGF-beta ligand traps targeted to these tumors may improve overall outcomes. Citation Format: Adam G. Sowalsky, Anson T. Ku, Uma Shankavaram, Shana Y. Trostel, Houssein A. Sater, Stephanie A. Harmon, Nicole V. Carrabba, Yang Liu, Radka Stoyanova, Elai Davicioni, Alan Pollack, Baris Turkbey, Deborah E. Citrin. Radiogenomic profiling of prostate tumors prior to external beam radiotherapy (EBRT) converges on a transcriptomic signature of TGF-beta activity driving tumor recurrence [abstract]. In: Proceedings of the AACR Special Conference: Advances in Prostate Cancer Research; 2023 Mar 15-18; Denver, Colorado. Philadelphia (PA): AACR; Cancer Res 2023;83(11 Suppl):Abstract nr B049.
