First-in-Class Small Molecule to Inhibit CYP11A1 and Steroid Hormone Biosynthesis

This article is featured in Highlights of This Issue, p. 1745

Hormonal drugs that interfere with the binding of steroid hormones to their cognate receptors are effective in the treatment of most prostate and breast cancers, but acquired drug resistance is common (1, 2). Multiple resistance mechanisms may cause treatment failure. For example, prostate cancer progression may result from high sensitivity of cancer to steroid ligands due to an increase in the androgen receptor (AR) gene copy number (3–5), or from AR mutations that enhance the binding of steroid ligands with a lower binding affinity compared with testosterone or dihydrotestosterone (6). In preclinical models of castration-resistant prostate cancer (CRPC), extragonadal steroid hormones other than androgens may maintain cancer growth by activating the mutated or overexpressed AR (7, 8). The adrenal glands and prostate cancer itself may synthesize and convert pregnenolone, progesterone, dehydroepiandrosterone, and their derivatives into more active androgens that can bind and activate the AR (7). Therefore, inhibition of the synthesis of most, or even all, steroid hormones might be beneficial for some patients with hormone regulated cancer.

All steroid hormones are produced from a single precursor, cholesterol, via a complex series of enzymatic reactions catalysed by the cytochrome P450 (CYP) and hydroxysteroid dehydrogenase enzymes in the gonads, the adrenal gland cortex, the placenta, and to a much smaller extent in some peripheral tissues such as the blood and the brain (9, 10). Cleavage of the cholesterol side chain is the first step in steroid hormone production, catalysed by a single mitochondrial inner membrane enzyme CYP11A1. The reaction consists of three sequential steps and leads to the formation of pregnenolone, the precursor common to glucocorticoids, mineralocorticoids, and sex steroids (11, 12).

We hypothesized that specific inhibition of CYP11A1 using a small-molecule inhibitor might halt the production of all steroid hormones. Because glucocorticoids and mineralocorticoids are essential for life in the human, we speculated that a CYP11A1 inhibitor can be administered successfully provided that a physiological dose of corticosteroid replacement therapy is administered concomitantly.

Here, we describe an oral, non-steroidal CYP11A1 inhibitor, ODM-208, with potential to treat endocrine regulated cancers. To our knowledge, this is the first report of a selective CYP11A1 inhibitor. We assessed the efficacy of ODM-208 both in vitro and in vivo in animal models, and in patients with CRPC. The results suggest that the activity of CYP11A1 can be blocked with ODM-208 leading to reduction of the blood steroid hormone levels to undetectable concentrations, and that ODM-208 can be administered with a simple oral glucocorticoid and mineralocorticoid replacement therapy.

Animal studies were conducted in accordance with the European Union legislation (Directive 2010/63/EU) and Good Laboratory Practice guidelines. The rodent studies were approved by the National Animal Experiment Board of Finland (permissions ESAVI/5535/2012, ESAVI/229/2016, ESAVI/9118/2016), except for the rat autoradiography study, performed at Charles River Laboratories (Edinburgh Ltd.), and conducted under the UK Home Office project licence no. PPL 70/8781. The dog study design was approved by the ethics committee of the test facility, Charles River Laboratories (Saint-Germain-Nuelles, France), as per the standard document “Chien_Tox subchronique_2014avril14cea”. Charles River Laboratories Ltd test facilities are American Association for Accreditation of Laboratory Animal Care (AAALAC) accredited. The clinical trial (CYPIDES; ClinicalTrials.gov identifier NCT03436485) protocol was approved by the ethics committees and the supervising medicines agencies of the participating sites/countries. An institutional review board at each participating trial site approved the trial, and the patients signed informed consent prior to starting the trial procedures. The study was conducted following the principles of the Declaration of Helsinki and Good Clinical Practice.

NCI-H295R cells resemble adrenal cortical cells as they produce most of the adrenal steroid hormones (androgens, mineralocorticoids, and glucocorticoids; ref. 13). NCI-H295R cells (American Type Cell Culture, CRL-2128, RRID CVCL_0458) were authenticated with short-tandem repeat profiling. The cells were cultured in Dulbecco's modified eagle's medium F12 with phenol red, and supplemented with 1.25% Nu-serum, 1% insulin-transferrin-selenous acid (ITS) + Premix, 2 mmol/L Glutamax, 100 IU/mL penicillin, and 100 μg/mL streptomycin (the reagents used in the study are listed in Supplementary Table S1). For the experiments, the medium was changed to a corresponding steroid depleted assay medium consisting of the Dulbecco's modified eagle's medium F12 without phenol red and supplemented with 1.25% Nu-serum stripped in loco. The MDCK-MDR1 and MDCK-BCRP cell lines were obtained from NIH, Bethesda, MD and Exton, PA, respectively.

We used a modified isocaproid aldehyde release assay (IARA) to measure sample CYP11A1 activity (Fig. 1A; ref. 14). NCI-H295R cells were seeded on poly-D-lysine 96-well plates in 100 μL of the culture medium at a density of 45,000 cells/well and incubated overnight at +37°C. ODM-208, dissolved in DMSO and diluted in a 10-point serial dilution range from 3 nmol/L to 6,000 nmol/L in the assay medium, was added, and after 30 minutes of incubation at the final DMSO concentration of 0.6%, ³H-labeled cholesterol and forskolin were added at the final concentrations of 3 nmol/L and 10 μmol/L, respectively. ³H-labeled cholesterol was used as the CYP11A1 substrate and forskolin to induce the cAMP/PKA signaling, an important second messenger pathway for trophic hormone-stimulated steroid biosynthesis and StAR expression (15). After 72 hours of incubation at +37°C and 5% CO₂, sample supernatants were collected, and the steroids were extracted from the reaction mixture by adding an equal volume of 1.25% aqueous dextran-coated charcoal solution with 2 mmol/L DL-dithiothreitol (DTT) to the samples. The samples were incubated for 20 minutes at room temperature and centrifuged at 1,076 × g for 10 minutes at +4°C. The product, isocaproic aldehyde, was measured in the supernatants by adding 2 volumes of scintillation liquid to each sample and measuring radioactivity (cpm) in the resulting mixture the next day using a Wallac MicroBeta TriLux instrument (1450 LSC & Luminescence Counter, PerkinElmer). Four independent experiments were performed in duplicates.

Inhibition of pregnenolone, progesterone, 11-deoxycortisol, cortisol, dehydroepiandrosterone [DHEA], and testosterone synthesis by ODM-208 and by the reference compounds ketoconazole, etomidate, and abiraterone was studied in NCI-H295R cells. Cells were seeded on poly-D-lysine 96-well plates in 100 μL of the assay medium at a density of 60,000 cells/well and incubated overnight at +37°C and 5% CO₂ before starting the assay. ODM-208 and the reference compounds, dissolved in DMSO and diluted in a 6-point serial dilution range from 0.1 to 10,000 nmol/L in the assay medium were added, and the cells were incubated for 48 hours at the final DMSO concentration of 0.5%. The steroid hormone concentrations were measured from the sample supernatants with liquid chromatography-mass spectrometry (LC/MS-MS) as described below. To test the recovery of pregnenolone production, NCI-H295R cells treated for 24 hours with a 4-point serial dilution range (from 50 to 1,000 nmol/L) of ODM-208 were washed twice with PBS and cultured in a fresh assay medium for 24 hours before sample pregnenolone concentration was measured.

Thermal profiling of ODM-208 with hrCYP11A1 was studied by measuring the melting temperature shift (ΔTm) with the nano differential scanning fluorimetry (nanoDSF) assay, performed at Proteros Biostructures GmbH (Martinsried, Germany). ODM-208 was dissolved in DMSO, and diluted with the assay buffer to the final DMSO concentration in the sample of 1%. hrCYP11A1 (2.5 μmol/L) was incubated with ODM-208 (10 μmol/L) in a buffer (50 mmol/L NaH₂PO4 at pH 6.5, 75 mmol/L NaCl, 1 mmol/L DTT, and 0.05% Tween-20) at room temperature for 30 minutes. The measurements were performed on a Prometheus NT.48 (NanoTemper Technologies) using High Sensitivity Capillaries (NanoTemper Technologies).

Inhibition of hrCYP11A1 enzyme activity was studied by measuring the pregnenolone formation from the substrate 22(R)-hydroxycholesterol. ODM-208 was dissolved in DMSO and diluted in a 11-point serial dilution range (0.2–10,000 nmol/L) in the incubation buffer with a final DMSO concentration of 0.5%. ODM-208 was preincubated with the hrCYP11A1 enzyme (100 nmol/L) and co-proteins adrenodoxin (500 nmol/L) and adrenodoxin reductase (30 nmol/L). Preincubations were performed either with or without NADPH (1 mmol/L) for 30 minutes at 37°C. The enzymatic reaction was initiated by adding the substrate, 22(R)-hydroxycholesterol (1 μmol/L) with a final ethanol concentration of 1%. The reaction mixture was incubated for 20 minutes at 37°C with orbital shaking (300 rpm). Samples were collected into ice-cold acetonitrile. LC/MS-MS (see below) was used to measure pregnenolone formation. The incubation buffer was 100 mmol/L phosphate buffer, pH 7.4, 5 mmol/L MgCl2, and 0.05% Tween20.

The quantitative analysis of steroid hormones was done using LC/MS-MS at Orion, Finland. A Thermo Scientific TSQ Altis mass spectrometer or a Sciex 6500+ Triple quad mass spectrometer (Singapore) with an electrospray interface coupled with a Thermo Vanquish Horizon (Germering), and a Shimadzu Nexera XP ultrahigh-performance liquid chromatograph were used. The samples were thawed, and an internal standard solution containing the analytes (except 11-deoxycortisol) were added, and after mixing, the samples were evaporated to dryness under nitrogen flow at +40°C. The samples were next reconstituted into water:methanol 60:40 (v/v) or 0.1% acetic acid:methanol 70:30 (v/v) and mixed on a plate shaker. After centrifugation at 4,000 rpm for 10 minutes at +10°C, the supernatants were transferred onto a 96-well plate or autosampler bottles for the LC/MS-MS analysis. The standards and the quality control samples were processed similarly.

LC/MS-MS was used to measure ODM-208 and steroid hormones from dog and rodent samples after protein precipitation, and liquid–liquid extraction or solid liquid–liquid extraction. A triple quadrupole mass spectrometer with an electrospray interface coupled with an ultra-high performance liquid chromatography (Sciex, Shimadzu Nexera, or Waters Acquity) was used. The analyte was detected using selected reaction monitoring and quantified by internal standardization. For the ODM-208 analyses, the calibration ranges were from 2 ng/mL to 5,000 ng/mL or from 2 ng/mL to 10,000 ng/mL, and the lower level of quantitation (LLOQ) was 2 ng/mL. The LLOQ for the steroid hormones ranged from 0.05 to 0.25 ng/mL in the plasma and from 0.15 to 26.3 ng/g in tumor tissue and adrenal gland tissue. The fit-for-purpose method was performed in terms of specificity, sensitivity, calibration range, precision, and accuracy.

Patient serum steroid hormone concentrations were analysed centrally at Esoterix using mass spectrometry after liquid–liquid extraction or protein precipitation. The LLOQs of ranged from 0.1 and 3.0 ng/dL, except for dehydroepiandrosterone sulphate that had a LLOQ of 0.1 μg/dL.

Inhibition of drug metabolising CYP enzymes (CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, and CYP3A4) was studied using human liver microsomes as the enzyme source. Inhibition of each CYP enzyme was evaluated by incubating the model substrates and monitoring the respective CYP enzyme catalysed metabolite. Determination of the IC₅₀ values for each CYP enzyme was done for metabolite formation during incubation in the presence of ODM-208 compared with a solvent control (16). The study was conducted at Cyprotex Ltd (currently part of Evotec Ltd.).

A diversity profile study was conducted using a Eurofins broadened SafetyScreen44Panel (17). In addition to the targets in the Cerep's SafetyScreen44, additional targets that may be linked to adverse effects were included in our broad diversity profile. ODM-208 was tested in two replicates at the concentration of 10 μmol/L. In each experiment and if applicable, the respective reference compound was tested concurrently with ODM-208, and the data were compared with historical values determined at Eurofins. The experiment was accepted following the Eurofins validation standard operating procedure.

The efflux properties of ODM-208 (1 μmol/L) were studied in the Madin-Darby Canine Kidney (MDCK) cells transfected with the MDR1 gene (encodes the P-glycoprotein) or with ABCG2 (encodes the breast cancer resistance protein, BCRP; ref. 18). The cells were from National Institute of Health (Bethesda, MD). The study was conducted at Absorption Systems Ltd. (Pharmaron).

Intact male 9- to 11-week old Balb/cOlaHsd mice (Envigo) received 20 mg/kg ODM-208 suspended in the vehicle (0.5% Tween in 0.5% aqueous methyl cellulose) orally twice daily for seven days. The animals received 0.9% NaCl instead of drinking water during the ODM-208 treatment. Prednisone 0.6 mg/kg was given orally once daily with the morning dose of ODM-208 to a half of the mice. Blood samples were collected by a terminal cardiac puncture under anaesthesia on day 7 prior to the morning ODM-208 dose, and 0.5, 1, 2, 4, and 8 hours after the morning dose, and 0.5, 1, 2, 4, 7, and 16 hours after the afternoon ODM-208 dose (n = 3 mice per time point). Corticosterone, progesterone, testosterone, and ODM-208 were measured using LC/MS-MS.

Male noncastrated, 8- to 10-week-old rats (RccHan:WIST, RRID RGD_5508396, Envigo; n = 5 per treatment) were dosed orally once daily for 14 days either with a vehicle (0.5% aqueous methyl cellulose) or ODM-208 at the dose of 75 mg/kg. All animals received 0.9% NaCl instead of drinking water from day -3 onwards. After a 14-day dosing period, the animals were sacrified and selected tissues were collected for histopathologic processing.

Male noncastrated, 12- to 17-month-old Beagle dogs (Marshall Bioresources) were dosed orally once daily for 28 days with a vehicle (aqueous 0.5% methyl cellulose), the vehicle + replacement therapy (prednisone 2.5 mg/day and fludrocortisone 0.1–0.15 mg/day), ODM-208 alone at the dose of 2 mg/kg, or ODM-208 at the dose of 5 or 10 mg/kg with the replacement therapy (n = 3–5 dogs per treatment), followed by a 4-week recovery period after stopping of ODM-208 administration. Body weight was recorded twice weekly. Blood samples were collected 24 hours after the day 28 dosing, and corticosterone, progesterone, testosterone, and ODM-208 were measured using LC/MS-MS at Admescope. Plasma triiodothyronine (T3), thyroxine (T4), and thyroid stimulating hormone (TSH) concentrations were measured at BioVetim, VetAgro Sup, T3 and T4 using a radioimmune assay and TSH using the Luminex method. The animals were necropsied on day 29 or after the 4-week recovery period and the selected organs were excised and weighed.

In the 4-week dog study, 40 different tissues were sampled for review by a professional toxicologic pathologist. The adrenals and the testes were sampled and reviewed in the 2-week rat study. The fixative was 10% neutral formalin (most tissues) or modified Davidson's fluid (testes). From rats, one adrenal was also frozen fixed. Formalin-fixed tissues were embedded in paraffin, and 4 μmol/L tissue sections were cut and stained with hematoxylin–eosin. Frozen samples were cryosectioned for Oil Red O fat staining.

Quantitative whole-body autoradiography (QWBA) was performed as Wang and colleagues (19) with minor modifications. A 25 mg/kg dose of [¹⁴C]-ODM-208 (equivalent to a radioactive dose of 4 MBq/kg) was administered orally in saline to intact male Crl:WI(Han) rats (RRID RGD_2308816, Charles River Laboratories), and blood was collected by a cardiac puncture under deep anaesthesia 0.25, 1, 8, 12, and 24 hours postdosing. After sacrification, the carcasses were deep frozen (−70°C) in hexane/dry ice and embedded in a 2% (w/v) aqueous carboxymethylcellulose block. Sagittal 30-μm sections were cut at −20°C, freeze-dried, and exposed to phosphor imaging plates for 7 days. The plates were analyzed with radioluminography using a Typhoon FLA7000 scanner (GE Healthcare) and an AIDA image analysis software (version 4.06.034, Raytest Isotopenmeβgerate GmbH). Radioactivity concentrations were expressed as μg equivalents of ODM-208 per mL of tissue (μg-eq/mL) assuming 1 gram of tissue is equivalent to 1 mL. Each phosphor screen was exposed to external standards that were used to determine the upper and lower limits of quantification.

We studied the efficacy of ODM-208 on cancer growth in a castration-resistant prostate cancer VCaP xenograft mouse model (20). VCaP cells harbor the TMPRSS2-ERG fusion gene, express high levels of AR, and secrete prostate-specific antigen (PSA) (21). Adrenalectomy leads to significant growth inhibition of VCaP xenografts (22). VCaP cells (ATCC-CRL-2876, RRID CVCL_2235), authenticated with a short-tandem repeat profiling, were cultured and used to generate a xenograft as described previously (23). In brief, Hsd:Athymic Nude-Foxn1^nu male mice (7 to 8 weeks of age; from Envigo, n = 8 per group) were used. Two million VCaP cells in 150 μL of the RPMI-1640 medium and Matrigel in a 1:2 ratio were injected subcutaneously. Mice were castrated under isoflurane anaesthesia when the average tumor volume reached 200 mm³. Oral treatments with the vehicle (0.5% Tween in aq. methyl cellulose 0.5%), prednisone 0.6 mg/kg/day, or ODM-208 30 mg/kg twice daily with prednisone 0.6 mg/kg/day were started when the tumors were regrowing (∼2 weeks after castration) and continued for 32 days. Tumors and adrenals were collected 4 hours after the last dose for analyses of progesterone, testosterone, and corticosterone using LC/MS-MS.

Statistical analyses were performed with the SAS Enterprise Guide version 7.13 (SAS software version 9.4; SAS Institute, Cary, NC). The dog steroid hormone levels and the tumor volumes in the mouse xenograft study were analysed with repeated measures analysis of variance (ANOVA) with pairwise comparisons. The values before treatment administration were considered the baseline values. The relative tissue weights of dogs and the steroids in the xenograft study were analysed with a one-way ANOVA model with pairwise comparisons. Two-sided P values <0.05 were considered statistically significant.

All steroid values were log-transformed and the values below the LLOQ were replaced as a value of LLOQ/2. If >50% of the treatment group values were below the LLOQ limit, all treatment group values were excluded from the analysis.

The data generated in this study are available within the article and its Supplementary Data files.

The crystal structure of human CYP11A1 in a complex with cholesterol (Protein Data Bank ID 3N9Y) was used to screen a library of compounds for activity. A glide ligand-receptor docking software (Schrödinger) was used to dock commercially available compounds to the CYP11A1 cholesterol-binding pocket, and IARA was used as a high throughput screening method to measure CYP11A1 inhibition achieved with the compounds in NCI-H295R cells.

To find compounds with high CYP11A1 selectivity, we first excluded compounds with potential heme-binding moieties from the screening library, especially aromatic nitrogens that are hydrogen bond acceptors. The selected compounds resembled the binding of cholesterol, the CYP11A1 natural substrate (Fig. 1B). We called the hits from this approach “substrate-like” binders, while conventional unselective inhibitors, etomidate or ketoconazole, were referred to as “heme binders” (24). The most potent screening hit, Compound 1, exhibited high selectivity and inhibited CYP11A1 providing a starting point for a systematic structure-activity relationship optimization, which led to the candidate compound, ODM-208 (Fig. 1C). Detailed experimental procedure for the preparation and characterization of ODM-208 is disclosed (25, example 185).

We assume that the enzymatic oxidation of ODM-208 occurs in a similar fashion as cholesterol is oxidized during the cholesterol side-chain cleavage. The isoindoline carbon 5 is supposed to have a similar orientation as cholesterol carbon 22 toward the heme of CYP11A1 (Fig. 1B). In situ formed metabolites may also contribute to the observed CYP11A1 inhibition.

ODM-208 inhibited the biosynthesis of pregnenolone at a low nanomolar concentration (IC₅₀, 15 nmol/L), and resulting from this, the biosynthesis of all measured downstream steroid hormones was also inhibited (Fig. 2A). ODM-208 was superior in steroid hormone biosynthesis inhibition compared with the non-selective CYP inhibitors ketoconazole and etomidate (26) (Supplementary Table S2). Pregnenolone production recovered within 24 hours after ODM-208 washout (Fig. 2B). Unlike ODM-208, abiraterone that blocks androgen production by inhibiting CYP17A1 induced accumulation of pregnenolone and progesterone (Fig. 2A).

When the binding of ODM-208 to the human recombinant CYP11A1 protein (hrCYP11A1) was assessed with a nanoDFS assay, ODM-208 bound to CYP11A1 showed thermal stabilization with a melting temperature shift (ΔTm) of 5.4°C, indicating direct binding of ODM-208 to hrCYP11A1 (Fig. 2C). We studied next the ability of ODM-208 to inhibit hrCYP11A1 activity by measuring pregnenolone formation from its substrate, 22(R)-hydroxycholesterol, when ODM-208 and hrCYP11A1 were preincubated with co-proteins adrenodoxin and adrenodoxin reductase with or without NADPH before initiating the enzymatic reaction by adding 22(R)-hydroxycholesterol. The IC₅₀ for pregnenolone production inhibition was 108 nmol/L with NADPH and 2,167 nmol/L without NADPH, suggesting that ODM-208 is a time-dependent inhibitor of hrCYP11A1 (Fig. 2D). The time dependency observed is likely explained by the formation of inhibitory metabolites in situ.

ODM-208 inhibited CYP11A1 >1,000 times more potently compared to the drug metabolising CYP enzymes studied (CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, and CYP3A4). ODM-208 did not inhibit these enzymes directly or in a time-dependent manner even at the maximum ODM-208 concentration studied (100 μmol/L). At the concentration of 10 μmol/L ODM-208 showed no activity in vitro on a panel of 125 pharmacologic targets that included G-protein–coupled receptors, nuclear receptors, ion channel proteins, and enzymes resulting in at least a 100-fold margin in inhibition compared with CYP11A1 (Supplementary Table S3).

CYP11A1 is a highly conserved enzyme in mammals with a high (∼98%) amino acid sequence homology between the human and the primates, and a high (≥71%) sequence homology also between the mouse, the rat, and the dog (27)_. To study the effect of ODM-208 on steroid hormone production in vivo, we administered ODM-208 20 mg/kg twice daily to adult noncastrated male Balb/c mice for 7 days. Plasma corticosterone, progesterone, and testosterone concentrations decreased rapidly from the baseline level achieving the maximum reduction (>95%) 4 hours after the first dose of ODM-208 regardless of whether 0.6 mg/kg prednisone was administered orally concomitantly with ODM-208 or not (Fig. 3A). Mineralocorticoid replacement therapy was not added, because adrenalectomized rodents can be maintained by offering physiologic saline instead of drinking water (28). The plasma steroid hormone levels were restored within 8 hours from ODM-208 dosing indicating rapid functional recovery in mice.

When ODM-208 was administered to adult noncastrated Beagle dogs once daily orally for 4 weeks either alone (2 mg/kg/day) or concomitantly with prednisone and fludrocortisone replacement therapy with higher dosages (5 or 10 mg/kg/day), a significant and dose-dependent decrease in the plasma cortisol and testosterone concentrations was observed (Fig. 3B). The lowest steroid hormone plasma levels, 1% to 3% of the baseline concentrations, were measured 4 hours after ODM-208 dosing. The replacement therapy could be tapered down within 3 days after stopping ODM-208 administration suggesting rapid functional recovery of the adrenal glands.

Rats tolerated a relatively a high dose (75 mg/kg) of ODM-208 administered twice daily without corticosteroid replacement therapy, whereas in the dog concomitant administration of corticosteroids with ODM-208 improved tolerability and permitted administration of ODM-208 at the dose level of 10 mg/kg/day for 4 weeks. The higher ODM-208 dose levels led to signs of dehydration in dogs, likely due to the inhibitory effect of ODM-208 on the adrenal gland function, unless corticosteroid replacement with prednisone 2.5 mg/dog/day and fludrocortisone 0.1 to 0.15 mg/dog/day was given. Dog body weight changes are provided in Supplementary Fig. S1.

In postmortem examinations the adrenal glands of both rats and dogs were enlarged and showed hypertrophy in all cortical zones, and vacuolation in the zona fasciculata and reticularis. The vacuoles contained lipids based on Oil Red O-staining (Fig. 4A). In both species, hypertrophy was present in the pituitary gland and the thyroid gland. Hypertrophy was associated with an increased serum thyroid-stimulating hormone (TSH) concentration (the dog data provided in Supplementary Table S4). In dogs, we found hypertrophy and vacuolation of the steroid producing Leydig cells in the testes, and acinar atrophy in the prostate and decreased weight (Fig. 4B). Dogs treated with ODM-208 5 or 10 mg/kg/day and concomitant corticosteroid replacement therapy, and the dogs treated with 2 mg ODM-208/kg/day without replacement therapy had similar findings in the histopathologic examination of the target organs. Prostate atrophy recovered fully within the first 4 weeks after ODM-208 discontinuation, whereas histologic alterations in the adrenals, the thyroid and pituitary glands, and the testes reversed only partially within this time period. No changes were observed in the other tissues examined.

To assess penetration of ODM-208 to the brain and the possible side effects resulting from neurosteroid inhibition (29, 30), we studied the distribution of [¹⁴C]-ODM-208 with quantitative whole-body autoradiography. After administration of [¹⁴C]-ODM-208 to male rats the highest radioactivity was measured in the adrenal cortex, whereas only little radioactivity was detected in the brain, suggesting low penetration of ODM-208 across the blood–brain barrier (Fig. 4C). In vitro studies with the MDCK-MDR1 and MDCK-BCRP cell lines suggested that ODM-208 is a substrate for the P-glycoprotein (MDR1) and the breast cancer resistance protein (BCRP; Supplementary Table S5), which may restrict the compound entry into the central nervous system (31).

In the VCaP model, ODM-208 administered orally 30 mg/kg twice daily with or without prednisone (0.6 mg/kg/day) inhibited the growth of the xenograft tumors (Fig. 5A). ODM-208 treatment decreased markedly progesterone, corticosterone, and testosterone concentrations both in the adrenal gland and in the tumor tissue (Fig. 5B). ODM-208 treatment did not influence the body weight of the mice (Supplementary Fig. S2; ref. 32).

ODM-208 was administered at the dose of 50 or 75 mg twice daily orally to the first six patients accrued to the multicenter phase I/II clinical trial (CYPIDES, ClinicalTrials.gov identifier NCT03436485). The patients who had progressive metastatic CRPC had previously been treated with at least one line of novel AR-targeted therapy (abiraterone or enzalutamide) and at least one line of taxane chemotherapy, and had serum testosterone concentration <50 ng/dL. Oral dexamethasone (1 mg/day) and fludrocortisone (0.05–0.1 mg/day) were administered concomitantly with ODM-208 as glucocorticoid and mineralocorticoid replacement therapy, respectively, and the patients continued androgen deprivation therapy.

Serum testosterone concentrations decreased about 50% from the predose value within the first 9 hours after the first dose of ODM-208. The serum concentrations of the monitored steroid hormones were undetectable 4 weeks after starting of ODM-208 except for detectable androstenedione and testosterone concentrations in one patient who had a 5-day interruption in the dosing 2 weeks after the start of ODM-208 treatment (Fig. 6). Serum oestradiol and progesterone concentrations were undetectable already before starting of ODM-208 and remained undetectable during the treatment.

Inhibition of CYP11A1 with ODM-208 reduced substantially the blood steroid hormone concentrations both in hormonally intact and castrated animals investigated and in patients with CRPC. The reversibility of the CYP11A1 inhibition was studied in non-castrated mice and dogs, where the functional recovery of the adrenal glands and the testicles was rapid with the key plasma steroid hormone levels generally recovering within a few days in dogs after stopping the 4-week administration of ODM-208. In preclinical models, ODM-208 inhibited CYP11A1 selectively compared with the other CYP enzymes studied. ODM-208 was well tolerated in rats and dogs, the main observation being adrenal hypertrophy and accumulation of lipid vacuoles in the adrenocortical cells. Lipid vacuole formation is likely a consequence of the ODM-208 induced metabolic block in the steroid hormone biosynthesis (33).

ODM-208 suppressed serum testosterone concentration rapidly in patients with CRPC, and the other monitored steroid hormone serum concentrations were also efficiently suppressed. These findings suggest that ODM-208 inhibits effectively steroid biosynthesis, and that there are no major redundant metabolic pathways that might compensate for steroid hormone production once CYP11A1 is inhibited. Because of the strong steroid hormone biosynthesis inhibition, ODM-208 needs to be administered with corticosteroid replacement therapy, which could be carried out with simple oral dosing. Abiraterone, an inhibitor of the 17α‐hydroxylase and 17,20‐lyase enzymes, is generally administered with prednisone (34, 35).

Achieving of very low tissue levels of the key hormonal ligands, such as testosterone, dihydrotestosterone, and testosterone intermediates may lead to durable responses and survival benefits in the treatment of patients with prostate cancer (36, 37). Such observations led to the development of oral CYP17A1 inhibitors, including steroidal abiraterone and nonsteroidal orteronel (TAK-700) and seviteronel (INO-464). While abiraterone inhibits both reactions catalyzed by CYP17, the 17-hydroxylase reaction and the 17,20-lyase reaction, orteronel and seviteronel have higher specificity for the 17,20-lyase reaction over the 17-hydroxylase reaction compared with abiraterone (38, 39). Orteronel plus prednisone prolonged progression-free survival, but not overall survival, in a large, randomized trial compared with placebo plus prednisone in a patient population with progressive metastatic CRPC, which result led to stopping of orterenol development for metastatic CRPC (40). At present, abiraterone remains the only approved CYP17 inhibitor for prostate cancer. Unlike CYP17A1 inhibitors that inhibit androgen production, ODM-208 blocks all steroid hormone biosynthesis including progesterone and other steroid hormones upstream of CYP17A1.

The steroids upstream of CYP17A1 have a lower affinity for the wild-type AR than the high-affinity steroid hormones, such as dihydrotestosterone, but may still drive cancer progression when the high-affinity steroid hormones are depleted or inhibited with endocrine therapy. The importance of these upstream hormones as drivers is well known in the late-stage prostate cancer patient population with mutated AR (6, 41, 42). AR mutations in the ligand-binding domain (LBD) may change the conformation of the AR and allow lower affinity steroid hormones to bind to the AR and promote cancer growth (43). Increased plasma levels of progesterone have been measured in CRPC patients treated with abiraterone (44, 45). In addition to progesterone, endogenous glucocorticoids like cortisol and corticosterone may activate mutated AR (46). The lower affinity steroids may be converted into higher affinity AR binding steroids within cancer tissues (8, 47). Hypothetically, the lower affinity steroids might also work in concert to activate the AR. The efficacy and safety of ODM-208 in the treatment of patients with advanced prostate cancer will be the topic for a future report, but encouraging serum PSA responses were observed in the CYPIDES trial in a substantial proportion of the patients treated with ODM-208 despite extensive prior treatments including at least one line of novel hormonal therapy and a taxane (48).

Similar activation of the estrogen receptor (ER) cannot be excluded in some patients with advanced breast cancer (36). Safety and efficacy of ODM-208 remain to be evaluated in preclinical models of breast cancer and in breast cancer patients, but because ODM-208 inhibits the synthesis of estrogens that the key activating ligands of the estrogen receptor (ER), ODM-208 is likely to have efficacy also in the treatment of ER-positive breast cancer. CYP11A1 inhibitors could be of interest also in the treatment of adrenal cortical carcinomas that may secrete multiple steroid hormones causing symptoms difficult to control with the currently available therapies (49).

The study has some limitations. Several dozens of steroid hormones have been identified (50), and we measured the serum concentrations of only a few. The steroid hormone levels in the human tumor tissues were not evaluated. The safety and efficacy of ODM-208 remains to be evaluated in the ongoing phase I/II multicenter trial (NCT03436485) in heavily pretreated patients (Supplementary Table S6) with mCRPC. Another CYP11A1 inhibitor, ODM-209 (25 example 184), is also being evaluated in a multicenter phase I/II trial (NCT03878823) in the treatment of advanced CRPC and breast cancer. At the time of this reporting, a few dozens of patients have been treated with either ODM-208 or ODM-209 (Supplementary Fig. S3) in these trials.

We conclude that ODM-208, a selective first-in-class CYP11A1 inhibitor, inhibits effectively steroid hormone biosynthesis and reduces the blood concentrations of steroid hormones in mice, dogs, and men. The safety and efficacy of ODM-208 warrants clinical evaluation in hormone regulated cancers.

M. Karimaa reports personal fees from Orion Corporation Orion Pharma during the conduct of the study and personal fees from Orion Corporation Orion Pharma outside the submitted work. R. Riikonen reports other support from Orion Corporation Orion Pharma during the conduct of the study and other support from Orion Corporation Orion Pharma outside the submitted work. H. Kettunen reports other support from Orion Corporation Orion Pharma during the conduct of the study and other support from Orion Corporation Orion Pharma outside the submitted work. P. Taavitsainen reports other support from Orion Corporation Orion Pharma during the conduct of the study and other support from Orion Corporation Orion Pharma outside the submitted work. M. Chrusciel reports personal fees from Orion Corporation outside the submitted work. P. Rummakko reports a patent for EP3558981 issued. G. Wohlfahrt reports a patent for US10717726B2 pending. H. Joensuu reports personal fees from Orion Corporation Orion Pharma during the conduct of the study, personal fees from Deciphera Pharmaceuticals and Neutron Therapeutics, other support from Orion Corporation Orion Pharma, and other support from Sartar Therapeutics outside the submitted work. K. Fizazi reports personal fees from Orion Corporation Orion Pharma during the conduct of the study, personal fees from CureVac outside the submitted work, and participation in advisory boards and talks for Amgen, Astellas, Astrazeneca, Bayer, Clovis, Daiichi Sankyo, Janssen, MSD, Novartis/AAA, Pfizer, and Sanofi, where honoraria go to Gustave Roussy, the author’s institution. R. Oksala reports other support from Orion Corporation Orion Pharma during the conduct of the study and other support from Orion Corporation Orion Pharma outside the submitted work. No disclosures were reported by the other authors.

M. Karimaa: Data curation, validation, investigation, visualization, methodology, writing–original draft, writing–review and editing. R. Riikonen: Software, formal analysis, investigation, visualization, methodology, writing–original draft, writing–review and editing. H. Kettunen: Formal analysis, investigation, methodology, writing–original draft, writing–review and editing. P. Taavitsainen: Conceptualization, investigation, methodology, writing–original draft. M. Ramela: Data curation, software, validation, investigation, methodology, writing–original draft. M. Chrusciel: Data curation, validation, investigation, visualization, methodology, writing–original draft. S. Karlsson: Formal analysis, investigation, methodology, writing–original draft, writing–review and editing. P. Rummakko: Conceptualization, data curation, software, investigation, writing–original draft, writing–review and editing. O. Simola: Conceptualization, investigation, methodology, writing–review and editing. G. Wohlfahrt: Conceptualization, data curation, software, investigation, visualization, writing–review and editing. P. Hakulinen: Resources, data curation, methodology, writing–original draft. A. Vuorela: Conceptualization, investigation, methodology, writing–original draft. H. Joensuu: Conceptualization, resources, supervision, investigation, methodology, writing–original draft. T. Utriainen: Investigation, methodology, writing–original draft. K. Fizazi: Conceptualization, supervision, investigation, methodology. R. Oksala: Conceptualization, data curation, formal analysis, supervision, investigation, visualization, writing–original draft.

We thank the personnel of the R&D Orion Corporation Orion Pharma for skillful technical assistance and patients in the CYPIDES trial.

This study was funded by Orion Corporation, Espoo, Finland.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.