Development of a second-generation antiandrogen for treatment of advanced prostate cancer

Abstract

Metastatic prostate cancer is treated with drugs that antagonize androgen action, but most patients progress to a more aggressive form of the disease called castration-resistant prostate cancer, driven by elevated expression of the androgen receptor. Here we characterize the diarylthiohydantoins RD162 and MDV3100, two compounds optimized from a screen for nonsteroidal antiandrogens that retain activity in the setting of increased androgen receptor expression. Both compounds bind to the androgen receptor with greater relative affinity than the clinically used antiandrogen bicalutamide, reduce the efficiency of its nuclear translocation, and impair both DNA binding to androgen response elements and recruitment of coactivators. RD162 and MDV3100 are orally available and induce tumor regression in mouse models of castration-resistant human prostate cancer. Of the first 30 patients treated with MDV3100 in a Phase I/II clinical trial, 13 of 30 (43%) showed sustained declines (by >50%) in serum concentrations of prostate-specific antigen, a biomarker of prostate cancer. These compounds thus appear to be promising candidates for treatment of advanced prostate cancer.

Fig. 1

Effect of RD162 and MDV3100 in human prostate cancer cells in vitro. (A) Chemical structures of the parent arylthiohydantoin scaffold compound RU59063 and the novel AR antagonists RD162 and MDV3100. (B) Representative competition binding curve showing inhibition of ¹⁸F-FDHT equilibrium binding to AR by FDHT, RD162, MDV3100 and bicalutamide (Bic) in LNCaP/AR cells. The IC₅₀ values from this experiment were: 5.1 nM (FDHT), 36 nM (MDV3100), 50 nM (RD162) and 159 nM (Bic) (error bars are reported as standard deviation of triplicate measurements). The inset shows the mean IC₅₀ values (± SEM) from five replicate experiments. (C) qRT-PCR analysis of the AR-dependent genes PSA and TMPRSS2 in LNCaP/AR cells cultured in androgen-depleted media with 5% charcoal-stripped serum (CSS). Cells were treated for 8 hours with or without 1nM of the synthetic androgen R1881 combined with DMSO (Veh), bicalutamide (Bic, 1 and 10 μM), RD162 (1 and 10 μM) and MDV3100 (1 and 10 μM) (normalized to actin mRNA, Mean ± SD, n=3). (D) Effect of bicalutamide, RD162 or MDV3100 on cell proliferation. VCaP cells were treated with the indicated antiandrogen and concentration (dashed line = 1 μM, solid line = 10 μM) in media containing fetal bovine serum (FBS). Vehicle-treated cells were in media containing either FBS or CSS. The viable cell fraction was determined by CellTiter-GLO (n=3, error is SEM). (E) Effect of bicalutamide, RD162 or MDV3100 on cleavage of poly-ADP-ribose polymerase (PARP). VCaP cells were treated for 1 or 3 days with 10 μM antiandrogen in media containing FBS. Vehicle-treated cells were in media containing either FBS or CSS. Cells treated with 10 μM etoposide for 1 day served as a positive control for apoptosis. Whole cell lysates were analyzed by Western blot.

Fig. 2

Activity of RD162 in mice. (A) Pharmacokinetic analysis of RD162 in male mice (n=3 per timepoint) dosed by oral gavage as a slurry of 20 mg/kg in 0.5% hydroxy-methyl-propyl-cellulose. (B) Pharmacodynamics of AR antagonism in castrate male mice after 5 days of treatment with daily oral RD162 (10 mg/kg) or vehicle control (n=3 mice per treatment group). Antagonism was measured by luciferase imaging of LNCaP/AR xenograft tumors expressing a luciferase reporter construct driven by the promoter of the probasin gene which is androgen regulated. Light emission at day 5 (2 hours after final dose) was quantified from a region of interest (ROI) drawn over the tumors and normalized for each animal to light emission at day 0. Luciferase activity was quantified at day 5 (normalized for each animal to day 0 luciferase activity) and expressed as fold-change as indicated for the individual mouse. Means ± SD: 2.5 ± 1.9 (Vehicle), 0.4 ± 0.2 (RD162). (C) Effects of RD162 and bicalutamide in a xenograft model of CRPC. Castrate male mice bearing LNCaP/AR tumors > 100 mm³ in size were treated by oral daily gavage with vehicle, bicalutamide (10 mg/kg) or RD162 (10 mg/kg). The percent change in volume for each tumor (12 tumors per treatment group) after 28 days is shown as a waterfall plot (y-axis, left). Plasma levels of bicalutamide and RD162 for each mouse were measured 20 hours after the final dose on day 28 (filled circles above the waterfall plot; y-axis, right). (D) Time to progression in mice continually treated as in (C). The percent change in volume for each tumor (12 tumors per treatment group) was assessed weekly. Events (tumor volume > 50% of baseline) were plotted on a Kaplan-Meier curve. Differences between all three groups were statistically significant: Bic vs R162 (p=0.04), Bic vs Veh (P<0.0001), RD162 vs Veh (P<0.0001) by log-rank (Mantel-Cox) test.

Fig. 3

RD162 and MDV3100 impair AR nuclear translocation, DNA binding and coactivator peptide recruitment. (A) Chromatin immunoprecipitation analysis of AR in LNCaP cells. The cells were cultured in androgen-depleted media with 5% CSS and treated for 8 hours with or without 1nM R1881 combined with DMSO (Veh), bicalutamide (Bic, 1 and 10 μM), RD162 (1 and 10 μM) and MDV3100 (1 and 10 μM). Real-time PCR quantification of immunoprecipitated PSA enhancer and TMPRSS2 enhancer is shown (% input mean ± SD, n=3). (B) Representative confocal microscopic images (scale bar = 10 μm) of LNCaP cells transfected with AR-EYFP in androgen-depleted media with 5% CSS and treated with DMSO, 1nM R1881, 10 μM bicalutamide, 10 μM RD162 or 10 μM MDV3100. The ratio of nuclear to cytoplasmic fluorescence intensity of individual cells were calculated (n=3, mean ± SEM). (C) Activation of an androgen-regulated reporter gene by VP16-AR. Cos-7 cells were cotransfected with ARE(4x)-luciferase plasmid and either wild-type AR or VP16-AR fusion protein. Cells were treated for 24 hours with or without 1nM R1881 combined with DMSO (Veh), bicalutamide (Bic, 1 and 10 μM), RD162 (1 and 10 μM) and MDV3100 (1 and 10 μM) for 24 hours. A luciferase assay was conducted using cell lysates and relative light units shown (n=3, mean ± SEM). (D) In vitro fluorescence resonance energy transfer (FRET) analysis of the interaction between AR LBD and FxxLF coactivator peptide. Increasing concentrations of DHT, bicalutamide, RD162 or MDV3100 were incubated with purified AR LBD, terbium-labeled anti-AR antibody and fluorescein-labeled AR FxxLF coactivator peptide (4 replicates). FRET between terbium and fluorescein, indicative of binding of FxxLF peptide to AR-LBD, was measured by ratio of fluorescence emission at 525 nm (fluorescein emission) to 488 nm (terbium emission) after excitation at 322 nm (terbium excitation). Normalized ratio is plotted (mean of 2 experiments with SEM).

Fig. 4

PSA response data in the first 30 patients receiving MDV3100 in Phase I/II trial. Thirty men with CRPC were treated with oral daily MDV3100 at doses of 30 mg/day (n=3) or 60 mg/day (n=27). The percent change in serum PSA for each patient after 12 weeks is shown as a waterfall plot. Patient 30 came off study prior to 12 weeks due to disease progression and is therefore not shown.