In order to assess mechanisms predictive of response or resistance to intense neoadjuvant therapy, we performed automated immunohistochemistry on targeted biopsies from patients with intermediate or high risk prostate cancer. All biopsy were stained with H&E or antibodies against: p53, PTEN, AR, PSA, GR, Ki67, SYP and PIN4-cocktail (p63+CK5+K18+AMACR). We further performed automated immunohistochemistry on whole mount radical prostatectomy tissues containing residual tumor. RP tissues were stained with H&E or antibodies against: PTEN, AR, AR-V7, PSA, PSMA, GR, Ki67, SYP and PIN4-cocktail (p63+CK5+K18+AMACR), with some slides stained for other markers of residual tumor, to include NKX3.1 and CAM5.2. Slides were scanned at 20x magnification. For each patient, response to intense neoadjuvant therapy is known. As biopsies were targeted to MRI-visible lesions, lesion-level response to therapy is also known. For each biopsy, gene expression profile by RNA-seq and exome sequencing were performed on tumor tissue laser capture microdissected from serial sections. Up to 4 foci were dissected per biopsy. Whole-genome sequencing of germline DNA from each patient was also performed.

Exome and gene expression data corresponding to pre-treatment biopsy tissues has been deposited in dbGaP and GEO. Genome, exome and expression data from post-treatment radical prostatectomy tissues is underway. As slides were stained using an automated slide stainer, this is a rich resource for performing data mining for assessing histogenomic correlates or signatures of treatment response or resistance. Our goal is also to eventually deposit the pre- and post-treatment mpMRI data from this cohort that matches the tissue, but those data are not ready for deposit and thus cannot be deposited at this time.