HSD3B1 genotype and outcomes in metastatic hormone-sensitive prostate cancer with androgen deprivation therapy and enzalutamide: ARCHES

Abstract

HSD3B1 encodes 3β-hydroxysteroid dehydrogenase-1, which converts adrenal dehydroepiandrosterone to 5α-dihydrotestosterone and is inherited in adrenal-permissive (AP) or adrenal-restrictive forms. The AP allele is linked to castration resistance, mainly in low-volume tumors. Here, we investigate the association of HSD3B1 alleles with outcomes in ARCHES, a multinational, double-blind, randomized, placebo-controlled phase 3 trial that demonstrated clinical benefit with enzalutamide plus androgen deprivation therapy (ADT) in men with metastatic hormone-sensitive prostate cancer (mHSPC) compared to those treated with placebo plus ADT. There are no significant differences between genotypes for clinical efficacy endpoints. Enzalutamide significantly improves radiographic progression-free survival and overall survival vs. placebo irrespective of HSD3B1 status. Men with the AP genotype have higher post-progression mortality and treatment-emergent adverse events, including hypertension, cardiovascular events, and gynecomastia, but a lower fracture rate. Overall, enzalutamide is beneficial in men with mHSPC independent of the HSD3B1 genotype. Inherited polymorphisms of HSD3B1 may account for differential toxicities.

Keywords: HSD3B1; androgens; enzalutamide; metabolism; metastatic hormone-sensitive prostate cancer.

Graphical abstract

Figure 1

Kaplan-Meier curves displaying clinical outcomes by HSD3B1 genotype associations (A) rPFS in the enzalutamide plus ADT arm between HSD3B1 genotypes. The data cutoff date for the rPFS curve was October 14, 2018, before crossover. HR comparisons were made for patients with AP vs. AR genotypes. (B) rPFS in the placebo plus ADT arm between HSD3B1 genotypes. The data cutoff date for the rPFS curve was October 14, 2018, before crossover. HR comparisons were made for patients with AP vs. AR genotypes. (C) OS in the enzalutamide plus ADT arm between HSD3B1 genotypes. The data cutoff date for OS was May 28, 2021, after crossover. HR comparisons were made for patients with AP vs. AR genotypes. (D) OS in the placebo plus ADT arm between HSD3B1 genotypes. The data cutoff date for OS was May 28, 2021, after crossover. HR comparisons were made for patients with AP vs. AR genotypes. (E) Time to PSA progression in the enzalutamide plus ADT arm between HSD3B1 genotypes. The data cutoff date for time to PSA progression was May 28, 2021, after crossover. HR comparisons were made for patients with AP vs. AR genotypes. (F) Time to PSA progression in the placebo plus ADT arm between HSD3B1 genotypes. The data cutoff date for time to PSA progression was May 28, 2021, after crossover. HR comparisons were made for patients with AP vs. AR genotypes. Analyses were conducted using Cox proportional hazards models that were stratified for prior docetaxel use and disease volume for HRs, 95% CIs, and nominal (unadjusted) p values. ADT, androgen deprivation therapy; AP, adrenal-permissive; AR, adrenal-restrictive; CI, confidence interval; HR, hazard ratio; NE, not estimable; OS, overall survival; PSA, prostate-specific antigen; rPFS, radiographic progression-free survival.

Figure 2

rPFS and OS based on HSD3B1 genotype by volume of disease and metastasis status (A) Kaplan-Meier curve of rPFS in patients with high-volume disease between the AP and AR HSD3B1 genotypes. The data cutoff date for the rPFS curve was October 14, 2018, before crossover. (B) Kaplan-Meier curve of rPFS in patients with low-volume disease between HSD3B1 genotypes. The data cutoff date for the rPFS curve was October 14, 2018, before crossover. (C) Kaplan-Meier curve of OS in patients with high-volume disease between HSD3B1 genotypes. The data cutoff date for the OS curve was May 28, 2021, after crossover was allowed. (D) Kaplan-Meier curve of OS in patients with low-volume disease between HSD3B1 genotypes. The data cutoff date for the OS curve was May 28, 2021, after crossover was allowed. (E) Forest plot of rPFS by volume of disease and metastasis status. (F) Forest plot of OS by volume of disease and metastasis status. Analyses were conducted using Cox proportional hazards models that were stratified for prior docetaxel use and disease volume for HRs, 95% CIs, and nominal (unadjusted) p values. AP, adrenal-permissive; AR, adrenal-restrictive; CI, confidence interval; E, number of events; ENZA, enzalutamide; HR, hazard ratio; M0, non-metastatic; M1, metastatic; N, number of patients; NE, not estimable; OS, overall survival; PBO, placebo; rPFS, radiographic progression-free survival.

Figure 3

Proportion of patients reaching a PSA reduction of ≥50% from baseline, a PSA reduction of ≥90% from baseline, and undetectable PSA responses Undetectable PSA rate was defined as the percentage of patients with detectable PSA (≥0.2 ng/mL) at baseline, which becomes undetectable (<0.2 ng/mL) during the study treatment. AP, adrenal-permissive; AR, adrenal-restrictive; ENZA, enzalutamide; PBO, placebo; PSA, prostate-specific antigen.

Figure 4

Tornado plots by genotype and treatment group (A) Most frequent TEAEs in the enzalutamide plus ADT arm. (B) Most frequent TEAEs in the placebo plus ADT arm. (C) TEAEs of special interest in the enzalutamide plus ADT arm. (D) TEAEs of special interest in the placebo plus ADT arm. ADT, androgen deprivation therapy; AP, adrenal-permissive; AR, adrenal-restrictive; TEAE, treatment-emergent adverse event.