Targeting the PI3K/AKT Pathway Overcomes Enzalutamide Resistance by Inhibiting Induction of the Glucocorticoid Receptor.

The PI3K-AKT pathway has pleiotropic effects and its inhibition has long been of interest in the management of prostate cancer, where a compensatory increase in PI3K signaling has been reported following androgen receptor (AR) blockade. Prostate cancer cells can also bypass AR blockade through induction of other hormone receptors, in particular the glucocorticoid receptor (GR). Here we demonstrate that AKT inhibition significantly decreases cell proliferation through both cytostatic and cytotoxic effects. The cytotoxic effect is enhanced by AR inhibition and is most pronounced in models that induce compensatory GR expression. AKT inhibition increases canonical AR activity and remodels the chromatin landscape, decreasing enhancer interaction at the GR gene (NR3C1) locus. Importantly, it blocks induction of GR expression and activity following AR blockade. This is confirmed in multiple in vivo models, where AKT inhibition of established xenografts leads to increased canonical AR activity, decreased GR expression, and marked antitumor activity. Overall, our results demonstrate that inhibition of the PI3K/AKT pathway can block GR activity and overcome GR-mediated resistance to AR-targeted therapy. Ipatasertib is currently in clinical development, and GR induction may be a biomarker to identify responsive patients or a responsive disease state.

Genomic Heterogeneity Within Individual Prostate Cancer Foci Impacts Predictive Biomarkers of Targeted Therapy.

BACKGROUND: Most lethal prostate cancers progress from relapse of aggressive primary disease. Recently, the most significant advances in survival benefit from systemic therapy have come from moving the administration of therapy to an earlier disease state. There is movement toward using biomarkers from the intraprostatic index lesion to guide early systemic therapy. OBJECTIVE: To determine the genomic heterogeneity, including the heterogeneity of predictive biomarkers, within the index focus of treatment-naïve prostate cancer. DESIGN, SETTING, AND PARTICIPANTS: Ten patients with treatment-naïve prostate cancer underwent prostatectomy. DNA was extracted from 70 spatially distinct regions of the 10 index foci. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Single nucleotide mutations, small indels, and copy number changes were identified. Intrafocal genomic heterogeneity and heterogeneity of alterations that predict response to therapy was determined. RESULTS AND LIMITATIONS: Exome sequencing and copy number estimates demonstrate branched evolution with >75% of point mutations being subclonal, including numerous pathways associated with castrate-resistant prostate cancer. Seven of 10 patients harbor alterations in one of five genes that predict response to targeted therapies with survival benefit in prostate cancer. Within biomarker-positive cases, 25% of intraprostatic regions are biomarker negative, with discordance between intraprostatic regions and lymph node metastases. CONCLUSIONS: Treatment-naïve, nonmetastatic prostate cancer has marked intrafocal heterogeneity. Numerous alterations in pathways associated with castration-resistant prostate cancer are present in subclonal populations, including biomarkers predictive of response to targeted therapy. PATIENT SUMMARY: Untreated patients' tumors have alterations that predict response to targeted therapies, but the presence of a biomarker is dependent on what region of the tumor was evaluated.