Genomic biomarkers of sensitivity to bipolar androgen therapy (BAT) in metastatic castration-resistant prostate cancer (mCRPC).

Abstract

5064 Background: In most cases, mCRPC remains driven by persistent androgen receptor (AR)-signaling. BAT has shown benefit in a subset of patients (pts) with mCRPC through multiple mechanisms of action, including modulation of AR activity, induction of DNA double-strand breaks and cell cycle arrest. There is a need for predictive biomarkers to identify patients most likely to benefit from BAT. Methods: This retrospective multicenter study included pts with mCRPC treated with BAT who had undergone tumor genomic profiling up to one month prior to BAT initiation. Somatic alterations were correlated with PSA50 response rates (≥50% decline in PSA from baseline), time to treatment failure (TTF) and overall survival (OS). PSA50 associations were assessed using Fisher’s exact test and logistic regression. Time-to-event outcomes were analyzed using Kaplan–Meier and Cox models. Independent predictors were evaluated using multivariable logistic regression and Cox models. Results: We identified 104 pts for inclusion. The median age was 72 years and 49.4% of pts presented with metastases at diagnosis. PSA50 was achieved in 39.4% overall. Median TTF and OS were 3.7 (95% CI: 3.0–4.9) and 28.5 (95% CI: 23.5–35.1) months, respectively. AR amplification, TP53 alterations and double-hit tumor suppressor alterations involving TP53 / RB1 / PTEN were all strongly associated with better PSA50 (Table). In multivariable analysis (MVA), AR amplification (OR 4.00, 95% CI 1.23–13.01; p = 0.021) and TP53 alterations (OR 5.18, 95% CI 1.71–15.71; p = 0.0037) remained independently associated with PSA50 In survival analyses, AR ligand-binding domain mutations that reduce testosterone binding affinity (low-affinity AR-LBD mutations: L702H, W742C, T878A/S) were linked to shorter TTF and remained independently associated in MVA (HR 2.19, 95% CI 1.15–4.18; p = 0.017), together with metastatic disease at diagnosis (HR 1.82, 95% CI 1.17–2.84; p = 0.008). Conclusions: Distinct genomic alterations are associated with sensitivity to BAT in mCRPC patients. AR amplification, TP53 alterations and TP53/RB1/PTEN double-hit status were all associated with increased likelihood of PSA response, while low-affinity AR-LBD mutations identify patients at risk for early treatment failure. These findings are in line with prior studies and support genomic profiling to refine patient selection for BAT. PSA50 TTF OS OR (95% CI) P value HR (95% CI) P value HR (95% CI) P value TP53 alterations 5.3 (1.9–15.4) 0.002 0.7 (0.4-1.1) 0.100 1.8 (1.0-3.3) 0.067 Double hit (TP53, RB1, PTEN) 4.4 (1.4–13.6) 0.013 0.8 (0.5-1.3) 0.313 1.7 (0.9-3.2) 0.115 AR amplification 4.6 (1.7–13.0) 0.003 0.7 (0.4-1.1) 0.110 0.9 (0.5-1.7) 0.809 Low affinity AR-LBD mutations 0.4 (0.1-1.5) 0.231 2.6 (1.4-4.6) 0.002 1.9 (0.9-3.89) 0.076 DDR alterations 0.6 (0.2-1.3) 0.210 1.3 (0.8-2.0) 0.223 1.6 (1.0-2.6) 0.057