First-in-man / proof of concept study with molidustat – a novel selective oral HIF-prolyl hydroxylase inhibitor for the treatment of renal anaemia

AIMS
Insufficient erythropoietin (EPO) synthesis is a relevant cause of renal anaemia in patients with chronic kidney disease. Molidustat, a selective hypoxia-inducible factor prolyl hydroxylase (HIF-PH) inhibitor, increases endogenous EPO levels dose-dependently in preclinical models. We examined the pharmacokinetics, safety, tolerability and the effect on EPO levels after single oral doses of molidustat in healthy male volunteers.This was a single centre, randomized, single-blind, placebo-controlled, group-comparison, dose-escalation study. Molidustat was administered at doses of 5, 12.5, 25, 37.5 or 50 mg as a polyethylene-glycol-based solution.In total, 45 volunteers received molidustat and 14 received placebo. Molidustat was rapidly absorbed, and mean maximum plasma concentration and area under the concentration–time curve increased dose-dependently. Mean terminal half-life was 4.64–10.4 h. A significant increase in endogenous EPO was observed following single oral doses of molidustat of12.5 mg and above. Geometric mean peak EPO levels were 14.8 IU/L (90% confidence interval: 13.0 – 16.9) for volunteers who received placebo and 39.8 IU/L (90% confidence interval: 29.4 – 53.8) for those who received molidustat 50 mg. The time course of EPO levels resembled the normal diurnal variation in EPO. Maximum EPO levels were observed approximately 12 h post-dose and returned to baseline after approximately 24–48 h. All doses of molidustat were well tolerated and there were no significant changes in vital signs and safety laboratory parameters.Oral administration of molidustat to healthy volunteers elicited a dose-dependent increase in endogenous EPO. These results support the ongoing development of molidustat as a potential new treatment for patients with renal anaemia.This trial was registered in the European Union Clinical Trials Register with the identifier: EudraCT 2010-018707-28.

Introduction
In the event of hypoxia or blood loss, erythropoietin (EPO) is released, mainly by the kidneyin adults, stimulating the production of red blood cells (1). For patients with chronic kidney disease (CKD), impaired kidney function may result in insufficient production of EPO to maintain normal haematocrit levels (2). Patients with CKD may have similar concentrations of circulating EPO to those of healthy individuals, but levels are low considering the presence of anaemia (3). Thus, the standard therapy for patients with anaemia associated with CKD is treatment with recombinant human EPO (rhEPO) and darbepoetin, a related erythropoiesis- stimulating agent (ESA). Although the introduction of these agents was a major advance, it was subsequently shown that they can increase the risk of major cardiovascular events (4, 5). This may reflect the development or worsening of hypertension in a considerable proportion of patients treated with ESAs, thought to be due, in part, to increased vascular resistance as a result of a haematocrit-induced increase in blood viscosity (6). There is a striking correlation between serious cardiovascular events (as well as mortality) and the cumulative epoietin dose(5) and it has been suggested that the rate of increase in haemoglobin (Hb), as well as Hb oscillations and overshoots are important in the development of cardiovascular events (4). Increased morbidity and mortality have been ascribed to the levels of rhEPO greatly exceeding the normal physiological levels of endogenous EPO (5). In a study in patients with CKD and type 2 diabetes mellitus, those in the lowest quartile of haematopoietic response to rhEPO (who subsequently received high doses of rhEPO) had an increased risk of death or cardiovascular events compared with patients in the upper quartiles of response (7).

In addition, rhEPO has been associated with increased risk of cancer and tumour progression (8, 9). A clinical phenotype associated with global HIF activation has been observed in patients with Chuvash polycythemia. Despite a lifelong activation of the HIF system an increase in the incidence of malignancies was not observed in such patients (10). Therefore, a differentapproach to treat EPO deficient anaemia might overcome these risks associated with rhEPO treatment.For healthy individuals, the change in atmospheric oxygen concentration that occurs on ascent from sea level to high altitude (2500 m) is normally well tolerated and not accompanied by symptoms associated with acute altitude sickness, such as headache, nausea, vomiting and lethargy (11). At the physiological level, the decreased oxygen concentration at altitude is compensated by a 1.5–2-fold increase in EPO production within the first 24 h (12), followed by an increase in reticulocyte levels within the next 2–3 days (13). In conditions of reduced oxygen, EPO gene transcription is activated by hypoxia-inducible factor (HIF) (14). In the absence of hypoxia, HIF is hydroxylated by specific oxygen-dependent dioxygenases, known as the HIF prolyl hydroxylases (HIF-PHs) or prolyl hydroxylase domain-containingproteins 1–3 (PHD 1–3) (15), thereby tagging it for degradation. Molidustat (BAY 85-3934)is a selective, orally bioavailable HIF-PH inhibitor, which behaves as a hypoxia mimetic both in vitro and in vivo, and thereby protects HIF from degradation (16). It is likely that the intermittent molidustat-mediated activation of the HIF system would have similar effects to intermittent and moderate high-altitude exposure.

In animal models, oral administration of molidustat increased expression of the HIF-sensitive target gene EPO and dose-dependently raised endogenous EPO levels, thereby stimulating erythropoiesis (16). Furthermore, in preclinical models, molidustat was effective in raising haematocrit levels while maintaining endogenous EPO levels close to the normal physiological range which, unlike treatment with rhEPO, resulted in the normalization of hypertensive blood pressure in a model of CKD (16).The purpose of this first-in-man study was to investigate the pharmacokinetics (PK), safety and tolerability of single oral doses of molidustat in healthy volunteers. In addition to these aspects of the study, endogenous EPO levels and other blood parameters were measured toprovide proof-of-concept that the stabilization of HIF by inhibiting HIF-PHs could be an approach to the treatment of patients with renal anaemia.This first-in-man study was a single-centre, randomized, single-blind, placebo-controlled, parallel-group-comparison, single-dose-escalation study conducted in healthy male volunteers. Single doses of molidustat or placebo were administered orally in the fasting state to five groups of 12 volunteers; nine individuals per group received molidustat and three received placebo, administered as a solution in a polyethylene glycol based diluent (comprising macrogol, polysorbate 20 and levomenthol). Molidustat was administered at doses of 5, 12.5, 25, 37.5 or 50 mg as a solution in a polyethylene glycol based diluent (comprising macrogol, polysorbate 20 and levomenthol). The synthesis of molidustat, 2-[6- (morpholin-4-yl)pyrimidin-4-yl]-4-(1H-1,2,3-triazol-1-yl)-1,2-dihydro-3H-pyrazol-3-one, has been described before (17). Molidustat was manufactured by Bayer AG in accordance with Good Manufacturing Practice guidelines.The volunteers were admitted to the study ward 25 h before study drug administration and underwent a medical examination. Drug administration began on the morning of the following day at approximately 8:00 am and volunteers were discharged from the ward after an observation period of 6 days.

A final examination was performed approximately 7 days after discharge.The study (EudraCT 2010-018707-28) was approved by the Ethics Committee of North- Rhine Medical Council, Düsseldorf, Germany, and by the competent authority (Bundesinstitut für Arzneimittel und Medizinprodukte, Bonn, Germany). The study was conducted in accordance with the guidelines on Good Clinical Practice, and with the Declaration of Helsinki and the German Medical Product Act. All volunteers gave written, informed consent to participate in the study.Healthy, white male volunteers between the ages of 18 and 45 years with a body mass index (BMI) between 18 kg/m² and 29.9 kg/m² were eligible for enrolment. Major exclusion criteria included: participation in another clinical trial or donation of blood within the preceding3 months; a history of drug abuse; daily consumption of more than 20 g of alcohol or more than 25 cigarettes; a history of diseases that may affect the absorption, distribution, metabolism, elimination or effect of the study drug; systolic blood pressure less than100 mmHg or greater than 145 mmHg, diastolic blood pressure less than 50 mmHg or greater than 95 mmHg, and heart rate less than 45 beats per minute or greater than 95 beats per minute. During the study, use of concomitant medications, and grapefruit juice were excluded as standard practice in early clinical studies. Smoking was restricted on the profile day and thereafter was limited to 10 or fewer cigarettes per day.PharmacokineticsTo determine PK parameters, blood samples were collected pre-dose, and 0.25, 0.50, 0.75, 1, 1.25, 1.5, 2, 3, 4, 5, 6, 8, 10, 12, 16, 24 and 48 h after drug administration. Urine wascollected at 0–4, 4–8, 8–12, 12–24 and 24–48 h after administration. The quantitative analysis of molidustat concentrations in plasma and urine was performed at Bayerlaboratories using a fully validated high-performance liquid chromatography/mass spectrometry method. The calibration range of this assay was from 0.100 µg/L (lower limit of quantitation) to 100 µg/L.

The mean inter-assay accuracy in calibrators ranged between 98.9% and 102 % and the corresponding precision was ≤ 7.36 %. Quality control samples in the concentration range from 0.300 µg/L to 500 µg/L in plasma were determined with an accuracy of 98.7 % to 103 % and a precision of ≤ 8.82 %. Quality control and test samples with a concentration > 100 µg/L were diluted by a factor 10. A quality sample with a concentration of 800 µg/L was subsequently prepared and analysed using the same dilution factor. Regarding specificity, chromatograms of 6 lithium heparin blank plasma samples from different donors did not reveal any relevant peak interfering with the peak of molidustat.On the basis of plasma-concentration-versus-time data, PK parameters were calculated for molidustat. The PK parameters were calculated using the model-independent (compartment- free) method in WinNonlin (version 4.1, Pharsight Corporation, St Louis, MO, USA) with an automation extension (developed by Bayer AG). For the non-compartmental analysis, the apparent terminal rate constant was calculated from the slope of a log-linear regression of the unweighted data considering the last concentration-time points > LLOQ. The PK parameters determined were: maximum plasma concentration (Cmax), area under the concentration–time curve (AUC), Cmax normalized by dose per kg bodyweight (Cmax,norm), AUC normalized by dose per kg bodyweight (AUCnorm), terminal half-life (t½), time to maximum concentration (tmax), apparent oral clearance (CL/F), amount of drug excreted via urine (Aeur) and renal clearance (CLR).PharmacodynamicsFor the assessment of the pharmacodynamic (PD) effects of molidustat, the following parameters were measured: endogenous EPO serum levels, reticulocyte counts, Hb andhaematocrit levels (packed cell volume).

EPO concentrations were measured by a solid- phase, two-site, one-cycle chemiluminescent enzyme immunometric assay (IMMULITE 1000mEPO, Siemens Healthcare Diagnostics).The PK/PD relationship was depicted calculated on the basis of molidustat- and EPO-plasma concentrations (geometric mean per time point and dose step).Clinical safety and tolerabilityClinical safety and tolerability were assessed by physical examination, monitoring of vital signs (blood pressure and heart rate), 12-lead electrocardiography and laboratory safety tests (blood and urine analysis). Adverse events were identified by participant self-reporting and by the investigators asking relevant non-leading questions. Laboratory safety tests were evaluated before and 24, 48, 96, 144 and 168 h after drug intake.Geometric means and percent coefficients of variation were calculated for AUC, AUCnorm, Cmax, Cmax,norm, t½, CL/F and CLR; arithmetic means and standard deviations (SD) were calculated for Aeur, age and BMI; and medians and ranges were calculated for tmax. For statistical analysis data, subjects treated with placebo were pooled across dose steps. The statistical comparison of maximum EPO data in the active treatment and placebo groups was carried out using a Bayesian approach with non-informative prior. The Bayesian approach facilitates the implementation on intuitive decision rule and interpretation of data in terms of probability statements (18). However, because no prior information was included in thisanalysis, sample sizes and precision of estimation of effects are similar as for classical (frequentistic) statistical approaches, e.g. the presented credible intervals would match the confidence intervals.

EPO data were assumed to be log-normally distributed, and an increase in EPO was considered statistically significant if the ratio of the potential underlying geometric mean of EPO Cmax for molidustat to the geometric mean of EPO Cmax for placebo was more than 1.0, with a posterior probability of more than 90%. The proof-of-concept for the study was considered to be achieved if there was an increase in the ratio of the potential underlying geometric means of EPO Cmax for molidustat versus placebo that exceeded 1.5 (comparable with a stay at 2500 m above sea level), with a posterior probability of 90% or more. Assuming that an increase in EPO Cmax by a factor of 2.0 relative to placebo could be observed in the study, minimum sample size needed per dose group in order to be confident of seeing an effect was nine for volunteers treated with molidustat and three for those who received placebo. Statistical analysis was performed using the SAS software package (version 9.1; SAS Institute Inc., Cary, NC, USA). The concentration–effect relationship between molidustat and EPO was calculated using SigmaPlot® version 12.5.Nomenclature of targets and ligandsKey protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY,(19) and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 (15, 20, 21).

Results
Per dose step, 9 patients were planned to receive active treatment and 3 patients placebo. In total, 65 volunteers were enrolled. Of these, six were not eligible and 59 received study medication (molidustat [n = 45], placebo [n = 14]). The mean age (SD) of participants was32.3  9.0 years and the mean BMI was 24.5  2.2 kg/m2. Demographic parameters were not statistically different between the study groups. At baseline, the haemoglobin levels and PCV were similar between the treatment groups and ranged between 14.01 ± 0.99 g/dL and 14.90±0.95 g/dL for haemoglobin and between 40.11 ± 2.49 % and 43.48 ± 3.33% for PCV.All those who received molidustat were deemed to be valid for PK analysis. The plasma concentration vs. time profile of molidustat was characterized by an early peak followed by a multiphasic decline. Molidustat plasma concentrations increased with increasing dose (Figure 1). Molidustat was rapidly absorbed, with median tmax values ranging from 0.250 h (for the 5mg dose) to 0.750 h (for the 50 mg dose); mean Cmax and AUC increased dose-dependently (Table 1). Mean t½ ranged from 4.64 h to 10.4 h. Mean CLR ranged from 0.530 L/h (for the 37.5 mg dose) to 1.75 L/h1 (for the 5 mg dose) and was considerably lower than the total oral clearance. Only a small proportion of molidustat (between 1.42% and 3.58%) was excreted unchanged in the urine.Samples from one volunteer treated with molidustat 12.5 mg were invalid for analysis (impaired results due to suspected cross-reactivity with heterophilic antibodies), meaning that data from 58 participants were available for PD evaluation.The profiles of endogenous EPO serum levels over time after oral administration of molidustat and placebo are shown in Figure 2.

Treatment with molidustat resulted in a dose- dependent increase in endogenous EPO. The geometric mean EPO Cmax, observed for the molidustat 50 mg dose group was 39.8 IU/L ((90% confidence interval: 29.4 – 53.8) compared with an EPO Cmax of 14.8 IU/L (90% confidence interval: 13.0 – 16.9) for volunteers who received placebo. For those who received the highest dose of molidustat, the EPO Cmax increased by a mean of 3.6-fold (90% confidence interval: 3.1 – 4.2) over baseline. In comparison, a mean increase of 1.4-fold (90% confidence interval: 1.3 – 1.6) was observed for volunteers treated with placebo. The time course of EPO levels resembled the normal diurnal variation in EPO. For those who received molidustat, EPO Cmax was reached about 12 h post-dose (range, 4–16 h) and EPO values returned to baseline after approximately 24–48 h.A pairwise comparison of molidustat with placebo was performed using a Bayesian approach. The potential underlying increase in EPO Cmax within 24 h of administration for molidustat relative to placebo, i.e. the ratio of the potential underlying geometric mean of EPO Cmax between molidustat and placebo, is given for different doses of molidustat in Table 2. The increase in EPO Cmax for molidustat relative to placebo was dose-dependent,with a 1.12-fold rise (90% credible interval: 0.89–1.41) for molidustat 5 mg, up to a 2.70-fold rise (90% credible interval: 1.94–3.75) for molidustat 50 mg.

Doses of 12.5 mg and above yielded an increase in EPO Cmax by a factor greater than 1.0 with more than 90% probability, and were considered to be statistically significantly different from placebo. According to thepredefined proof-of-concept criteria, significant increases in EPO Cmax relative to placebo were observed for molidustat 50 mg (i.e. the increase in EPO Cmax relative to placebo by a factor of greater than 1.5 was observed with more than 90% probability). A comparison of EPO Cmax based on the relative changes from baseline between molidustat treatment and placebo produced similar results.A significant increase in reticulocyte levels compared to placebo was seen 72 h and 120 h after administration of molidustat at doses of 37.5 mg and 50 mg, respectively (evidence of an increase with > 90% posterior probability according to proof of concept criterion). No clinically relevant increases in haemoglobin or haematocrit were observed.Concentration–effect relationships between molidustat and EPOThe concentration–effect relationships between molidustat and endogenous EPO were calculated (geometric means within the first 24 h after drug intake) (Figure 3). A nearly linear increase in EPO was observed, which correlated with the molidustat concentration for the first four doses (5–37.5 mg). The 50 mg molidustat dose showed a more than proportional increase in EPO, especially between 4 to 12 h after administration.Of the 59 volunteers, 14 experienced 23 treatment-emergent adverse events (TEAEs), which are summarized in Table 3. Five TEAEs were observed after placebo and 18 after molidustat treatment. One TEAE was of moderate intensity (migraine), and all other TEAEs were of mild intensity. No serious adverse events were observed. The most common TEAEs were throat irritation in five volunteers and headache in three volunteers. The frequency and intensity of the TEAEs were unrelated to the dose of molidustat, and all had resolved by the end of the observation period. No clinically relevant changes in clinical laboratoryparameters, urinalysis, blood pressure or heart rate or electrocardiogram parameters were observed.

Discussion
In this phase I study, the safety, tolerability, and PK and PD effects of the novel HIF-PH inhibitor molidustat (BAY 85-3934) were investigated in young healthy male volunteers following oral administration of a single dose. Molidustat was rapidly absorbed and mean values of both AUC and Cmax increased dose-dependently. The mean t½ ranged from 4.64 h to10.4 h. The amount of molidustat excreted in the urine was low (less than 4%), so alteration of renal function is unlikely to have a clinically significant effect on the elimination rate. Molidustat is metabolized exclusively via N-glucuronidation to an inactive metabolite which is excreted via urine (17). Molidustat was well tolerated and the incidence of TEAEs was typical of that seen in other phase I studies (22, 23). The frequency and intensity of the TEAEs were unrelated to molidustat dose and all had resolved by the end of the observation period; however, the drug formulation was viscous and was reported to have a very bitter taste, and this was suspected to be the cause of the most common TEAE (throat irritation). A tablet formulation of molidustat has been developed for further studies.In our study, endogenous EPO levels increased dose-dependently after oral administration of molidustat. The predefined proof-of-concept criteria were met following administration of the 50 mg dose. Based on PK/PD modelling of the preclinical data, an effective human dose of approximately 30 mg was predicted for treatment of renal anaemia (16), and those who received molidustat at doses of 12.5 mg and above showed significant increases in EPO compared with those who received placebo. Although peak molidustat concentrations were observed between 0.250 h and 0.750 h after dosing, maximum EPO levels were reachedapproximately 12 h post dose. This is to be anticipated given the mechanism of action, and reflects the time required for EPO expression and transcription following HIF-PH inhibition by molidustat.

Endogenous EPO levels returned to baseline after approximately 24–48 h.Similar kinetics of EPO induction were observed after single oral doses of another HIF-PH inhibitor, FG-2216, with maximal EPO response approximately 12 h post-dose (24). Studies in healthy volunteers have shown a diurnal rhythm in the circulating levels of EPO (25). This diurnal pattern is also apparent with the molidustat-triggered EPO response, with highest values observed in the evening after administration of molidustat in the morning. For volunteers who received placebo, a small increase in EPO secretion was observed, which may reflect the normal changes during circadian rhythm. Additionally, these increases in EPO may be partly a response to blood loss from venipuncture. However, the total volume of blood collected was limited to 160 mL over a period of approximately five weeks to overcome major influence on EPO values.The increased endogenous EPO levels observed were in the normal physiological range seen in individuals ascending to high altitude (13), and treatment with molidustat therefore avoids the high peak levels of rhEPOs associated with current standard of care with ESAs. It has been suggested that the increased risk of cardiovascular events observed during rhEPO therapy may be related to the oscillations in Hb levels that result from intermittent dosing (4). Molidustat can be administered orally once daily, thus avoiding the rapid changes in Hb levels that may be seen with rhEPO therapy. Furthermore, the lower concentrations of endogeneous EPO reached after administration of molidustat compared with rhEPO treatment might alleviate the increased risk of cancer and tumour progression associated with treatment with rhEPO (8, 9, 26).In preclinical models, adaptation of the EPO response or accumulation of EPO was not observed when molidustat was administered on consecutive days (16). In contrast, rapid acclimatization of the EPO response is observed at high altitude, with plasma EPO concentrations declining after 2 days (13). Long-term studies will demonstrate the potential of molidustat to treat patients with renal anaemia.The results from this study show that oral single-dose administration of molidustat can elicit dose-dependent increase of endogenous EPO in healthy volunteers. A phase 2b BAY 85-3934 program has been designed to determine dosing recommendations for phase III studies (manuscript in preparation). Molidustat is a promising new option for the treatment of patients with renal anaemia and clinical studies are ongoing to investigate the effect of this novel therapy in patients with CKD.