Purpose: Cross-resistance renders multiple lines of androgen receptor (AR) signaling inhibitors increasingly futile in metastatic castration-resistant prostate cancer (mCRPC). We sought to determine acquired genomic contributors to cross-resistance. Experimental Design: We collected 458 serial plasma cell-free DNA samples at baseline and progression timepoints from 202 patients with mCRPC receiving sequential AR signaling inhibitors (abiraterone and enzalutamide) in a randomized phase II clinical trial (NCT02125357). We utilized deep targeted and whole-exome sequencing to compare baseline and posttreatment somatic genomic profiles in circulating tumor DNA (ctDNA). Results: Patient ctDNA abundance was correlated across plasma collections and independently prognostic for sequential therapy response and overall survival. Most driver alterations in established prostate cancer genes were consistently detected in ctDNA over time. However, shifts in somatic populations after treatment were identified in 53% of patients, particularly after strong treatment responses. Treatment-associated changes converged upon the AR gene, with an average 50% increase in AR copy number, changes in AR mutation frequencies, and a 2.5-fold increase in the proportion of patients carrying AR ligand binding domain truncating rearrangements. Conclusions: Our data show that the dominant AR genotype continues to evolve during sequential lines of AR inhibition and drives acquired resistance in patients with mCRPC.

Purpose: Clusterin is a cytoprotective chaperone protein that promotes cell survival and confers broad-spectrum treatment resistance. OGX-011 is a 2 ¶-methoxyethyl^modified phosphorothioate antisense oligonucleotide that is complementary to clusterin mRNA, has a prolonged tissue half life, enhances drug efficacy in xenograft models, and reduces clusterin expression in humans with a biologically effective dose of 640 mg.The objective of this study was to determine a recommended phase II dose of OGX-011in combination with docetaxel. Experimental Design: Patients with cancers known from the literature to express clusterin were eligible. OGX-011was given by 2-h i.v. infusion starting at 40 mg weekly after loading doses on days 1, 3, and 5. Docetaxel was given i.v. 30 mg/m 2 weekly for 5 of 6 weeks (schedule A) or 75 mg/m 2 every 3 weeks (schedule B). All patients had serial samples of peripheral blood mononuclear cells and serum assessed for clusterin expression. Results: Forty patients were enrolled to eight cohorts. OGX-011could be given at the full biologically effective single-agent dose of 640 mg with both docetaxel schedules. Toxic effects were primarily myelosuppression, fatigue, hair loss, gastrointestinal effects (expected docetaxel effects), as well as dose-related chills and fever (expected OGX-011effects). OGX-011AUC and C max increased proportionally with no apparent effect on docetaxel pharmacokinetics. At the end of cycle 1, serum clusterin showed mean decreases of 34% and 38% (range, 15-99%) at the 640-mg dose levels. Conclusions: OGX-011 can be given at a biologically effective dose with standard doses of docetaxel. Phase II trials of combined OGX-011 and chemotherapy are ongoing in patients with prostate, breast, and lung cancers.