Coadministration of the prostaglandin F2α receptor antagonist preterm labour drug candidate OBE022 with magnesium sulfate, atosiban, nifedipine and betamethasone

Abstract

Aims

To investigate presence or absence of clinically relevant drug interactions (pharmacokinetic and safety/tolerability) of OBE022 with standard‐of‐care medicines for preterm labour, enabling coadministration and further clinical development.

Methods

Part A: open‐label, randomized, 3‐period crossover assessing coadministration of single doses of OBE022 (1100 mg) and MgSO₄. Part B: open‐label, single‐sequence crossover assessing the interactions following administration of OBE022 (1000 mg/day) at steady state coadministered with single doses of atosiban, nifedipine and betamethasone. Twenty‐five healthy nonpregnant women of reproductive age were enrolled (Part A: n = 12; Part B: n = 13).

Results

OBE022, alone or in combination with standard‐of‐care medications, was well tolerated. Headache and dizziness were the most frequently reported adverse events; dizziness occurred more often with the nifedipine/OBE022 combination. There were no clinically significant pharmacokinetic interactions when coadministered with MgSO₄. Co‐administration had no notable effect on atosiban exposure. Atosiban reduced exposure to OBE002 (peak concentration [C_max] 22%, area under the concentration–time curve [AUC] 19%). Coadministration with betamethasone slightly increased betamethasone exposure (C_max + 18%, AUC +27%) and OBE002 exposure (C_max + 35%, AUC +15%). These changes were not considered clinically significant. Coadministration with nifedipine slightly increased OBE002 exposure (C_max + 29%, AUC +24%) and markedly increased nifedipine exposure (C_max by 2‐fold and AUC by 2‐fold), which may be clinically significant.

Conclusions

The use of OBE022, a PGF2α antagonist prodrug, in combination with standard‐of‐care medicines may provide new treatment alternatives for preterm labour. All tested combinations were well tolerated. Nifedipine doses could potentially be reduced or staggered when coadministered with OBE022.

What is already known about this subject

• In uterine tissues, prostaglandins E2 and F2_α have been shown to exert cervical changes and elicit uterine contractility.

• Activation of the prostaglandin F receptor in the human myometrium by prostaglandin F2_α results in the elevation of intracellular calcium concentration, which, in turn, leads to contraction of the uterine smooth cell muscle.

• OBE022 is safe and well tolerated in healthy postmenopausal women.

What this study adds

• OBE022 can be safely coadministered with other currently used tocolytics and standard‐of‐care medicines in preterm labour patients.

• Nifedipine exposure increased significantly when coadministered with OBE022, leading to an increase in expected nifedipine side‐effects.

1. INTRODUCTION

1.1. Background to the target indication

Preterm birth, defined as birth before week 37 of gestation is a leading cause of infant mortality and morbidity.1 Estimates in 2010 showed that 35% of the 3.1 million deaths in neonates (infants aged <4 weeks) worldwide were due to preterm complications.2 Survivors of preterm birth can suffer lifelong neurodevelopmental and learning impairments.3

Tocolytics including atosiban (approved in Europe4, 5) and nifedipine are used as treatments to delay preterm labour by reducing uterine contractions. The problem with current tocolytics is they can be inefficacious and/or cause treatment‐limiting side effects.6 Other medications given to women who are prone to going into preterm labour include those to protect against preterm complications in the neonate. Clinical studies have demonstrated that magnesium sulfate (MgSO₄) protects against neurological morbidities7 and, in clinical practice, maternal corticosteroids such as betamethasone are coadministered with tocolytics to promote fetal lung maturation.8 The combined use of tocolytics and protective agents is recommended by obstetric clinical guidelines in the USA and Europe.8, 9

1.2. Background to the investigational medicinal product, OBE022

Prostaglandins and their G‐protein‐coupled receptor subtypes provide a potential target for therapeutic intervention in preterm labour.10, 11, 12, 13, 14, 15, 16 In particular, prostaglandin F2α (PGF2α) causes contraction of the myometrium through activation of prostaglandin F (FP) receptor by raising intracellular calcium concentrations17 and also upregulates the matrix metalloproteinase MMP1 (an enzyme that breaks down collagen in cervical fibroblasts leading to cervical ripening).18 Indomethacin, a prostaglandin synthesis inhibitor has been shown to be effective in delaying labour by at least 48 hours.19 However, it is nonselective and blocks signalling of many prostaglandin receptor subtypes, increasing the risk of severe side‐effects in newborn babies.20

OBE022 is a potent, small molecule PGF2α receptor antagonist being developed to inhibit preterm labour.21 Ongoing clinical investigations are testing twice daily administrations of OBE022 during up to 7 days in preterm labour patients. This oral valine‐ester prodrug readily hydrolyses to its equally potent and highly selective PGF2α antagonist metabolite OBE002. In preclinical studies, human liver microsomes were incubated with cytochrome‐P450 (CYP) enzyme‐specific probe substrates alone or in the presence of OBE002. Liquid chromatography coupled with tandem mass spectroscopy (LC–MS/MS) analysis of the probe substrates showed that OBE002 did not inhibit common CYP pathways at clinically relevant concentrations (ObsEva, data on file). Similarly, monolayer cultures of human fresh hepatocytes were incubated with OBE002 and mRNA levels of CYP1A2, CYP2B6 and CYP3A4 were analysed using quantitative real‐time polymerase chain reaction to investigate OBE002 CYP induction potential. OBE002 did not induce CYP enzymes at clinically relevant concentrations (ObsEva, data on file). Incubations of OBE002 with individual cDNA expressed human CYP enzyme preparations (Bactosomes, Cypex Ltd., UK), which were analysed for OBE002 using LC–MS/MS, showed that OBE002 was metabolized by CYP3A4, CYP2C19 and CYP2D6 indicating several metabolism pathways.

In contrast to indomethacin, OBE022 and OBE002 have no fetal side effects related to prostaglandin synthesis inhibition.22 Combining OBE022 with other treatments may generate additive or synergistic effects on uterine contractions thereby, extending gestation periods.22

1.3. Background to the clinical trials programme

The early phase drug development programme consisted of 2 interdependent, adaptive trial protocols. The first protocol (Protocol 1) consisted of first‐in‐human (FIH) single ascending dose and multiple ascending dose parts in healthy nonpregnant female volunteers to assess OBE022's safety, tolerability and pharmacokinetics (PK).23 Cardiac safety assessments and an assessment of the effect of food on the PK were integrated into the multiple ascending dose part.23

The intention was to progress to preterm labour patients early in the drug development programme. It was important to first conduct trials in healthy nonpregnant female volunteers rather than preterm labour patients as it was important to determine the safety and tolerability of OBE022 when coadministered with various standard of care medications. Data from the FIH trial demonstrated that OBE022 would not expose them to an increased risk.21, 23

As preterm labour requires treatment with several medications owing to its complex aetiology and the risk posed to the fetus, OBE022 could not be tested in labouring women until it had been demonstrated that there was no significant interaction with other potential concurrent medical therapy. The aim of the drug–drug interaction (DDI) study (Protocol 2) was to investigate the presence or absence of clinically relevant drug interactions (PK and safety/tolerability) with standard‐of‐care medicines for preterm labour and was dependent on emerging data from Protocol 1.

2. MATERIALS AND METHODS

2.1. Study population

Healthy women aged 18–45 years with body mass index between 18.0 and 30.0 kg/m² were included. Subjects with current/recurrent disease or history of any condition precluding administration of MgSO₄, atosiban, nifedipine or betamethasone were not enrolled. Subjects with any clinically significant abnormality in laboratory blood or urine results, electrocardiograms or vital signs were also excluded. Subjects who had used any of the following within the stated time periods were also excluded: tobacco in any form (3 months), prescription or over‐the‐counter medication (2 weeks); hormonal/systemic contraception (3 months); products containing St John's Wort (3 weeks); any substance, including dietary, known to induce or inhibit CYP P450 enzymes (3 weeks); caffeine/xanthine/energy drinks or alcohol (48 hours).

2.2. Study design

This study was performed at Richmond Pharmacology London, UK in accordance with UK law, the Declaration of Helsinki, Good Clinical Practice guidelines, and European and US clinical drug‐interaction guidelines.24, 25 The protocol (EudraCT: 2016–001958‐18) was approved by a Health Research Authority Ethics Committee (South Central‐Berkshire B, UK)26 and the Medicines and Healthcare products Regulatory Authority (MHRA). The draft revised Food and Drug Administration guidelines have since been updated (i.e. following completion of this study) and the study is in compliance with the updated guidelines.25 All subjects voluntarily signed informed consent forms before enrolment.

In the FIH study, OBE022 was shown to be rapidly hydrolysed to OBE002, which was present in the blood at noticeably higher levels than OBE022: The maximal plasma concentration (C_max) of OBE002 generally exceeded that of OBE022 by more than 100‐fold with up to 1200‐fold higher area under the plasma concentration–time curve (AUC) values. Based on OBE022's drug interaction profile (see Introduction, Background to the investigational medicinal product, OBE022), the index drug design was not considered appropriate. Furthermore, OBE022's target indication was very specific with a relatively limited number of medications likely to be used concomitantly. It was considered more relevant to assess potential interactions with the medications most likely to be used during preterm labour and hence a concomitant drug study was designed. The concomitant medications used with OBE022 in this study were MgSO₄, atosiban, nifedipine or betamethasone—all standard of care medicines that could be used with OBE022 for women who are at risk of preterm delivery.

Magnesium is not metabolized but instead excreted solely by the kidney. OBE022 was not expected to affect renal clearance therefore no foreseeable metabolic interaction between MgSO₄ and OBE022 was anticipated. Atosiban has been shown in in vitro testing to be neither a substrate for nor inhibitor of the cytochrome P450 system. Nifedipine, a dihydropyridine, is metabolized by CYP3A4 and the inactive metabolites are excreted in the urine and faeces via biliary excretion. Betamethasone inhibits and is metabolized by CYP3A4. As OBE002 was shown to not inhibit the cytochrome P450 system and potentially have multiple metabolic pathways, no significant drug–drug interaction was expected with atosiban, nifedipine or betamethasone.

2.3. Design of Parts A and Part B of the study and OBE022 dose selection

OBE022 was administered as an oral solution in both study parts and doses were anticipated to produce exposures below the protocol‐defined PK exposure limit.

2.3.1. Part A

Part A was conducted as an open‐label, randomized, 3‐period crossover study, consisting of 3 treatment periods (Figure 1). Twelve healthy premenopausal women were included in 1 cohort and were randomized to receive either OBE022, MgSO₄ or OBE022 coadministered with MgSO₄.

Figure 1.

Study design Part A

The OBE022 dose was selected to ensure that the anticipated mean exposures (C_max and AUC from administration to 24 hours [AUC_0–24]) would not exceed those previously explored in the FIH study (i.e. Protocol 1) with acceptable safety and tolerability, i.e. dosing regimens at which no study specific criteria stopping dose progression and/or escalations were met.23 Based on the safety, tolerability and PK data from Protocol 1,23 the safety review committee selected single doses of 1100 mg of OBE022.

Subjects were screened 55 days prior to entering the study on Day −1. For each treatment period, volunteers were admitted on Day −1 and discharged on Day 3. All subjects attended an outpatient visit on Days 4 and 5 and a follow‐up visit ±1 day after treatment period 3. All subjects fasted for 10 h predose and 4 h postdose. Treatments were administered on day 1 of each treatment period. Treatment periods 1 and 2 were run as a randomized cross‐over between OBE022 and MgSO₄ (Figure 1). Venous blood samples were collected for PK analysis predose and 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 4, 6, 8, 12, 16, 20, 24, 48, 72 and 96 h postdose.

2.3.2. Part B

Part B was conducted as an open‐label, single‐sequence crossover study (Figure 2).

Figure 2.

Study design Part B

Doses of OBE022 for Part B were also selected from Protocol 1,23 the highest multiple dose (1000 mg) was selected to be administered from Days 4–12.

Twelve subjects in Part B were administered atosiban, nifedipine, betamethasone and OBE022 sequentially. Once OBE022 had reached steady state (Day 9), OBE022 was then coadministered with atosiban, nifedipine or betamethasone.

Subjects were screened up to 55 days prior to admission on to the study on Day −1 and remained hospitalized for 14 days Subjects were dismissed from the unit on Day 14 and attended the unit for a follow‐up visit on Day 21.

2.4. Dose selection and route of administration of standard‐of‐care medications

Doses selected and the route of administration of standard‐of‐care medications were in‐line with those used in clinical practice for MgSO₄ 27 and in the Royal College of Obstetricians and Gynaecologists' guidelines for betamethasone and nifedipine28: MgSO₄ was administered intravenously as loading dose of 4 g over 30 minutes followed by a maintenance dose of 1 g/h for 11.5 hours. Nifedipine was administered as a 20‐mg oral dose and betamethasone as a 12‐mg intramuscular injection. Atosiban was administered as an infusion using the high‐dose part of the standard clinical regimen. A 6.75‐mg dose was administered as a 0.9‐ml intravenous bolus injection given over 1 minute followed by a 54‐mg dose administered as a 3 hours intravenous loading infusion at 24 ml/h (300 μg/min). This infusion duration was considered sufficient for the purpose of testing potential interactions with OBE022. All standard of care medications were administered as single doses without and then with OBE022.

2.4.1. Timing of administration

All IMPs were administered in the morning. OBE022 was given as an oral solution to subjects and was timed to start simultaneously either with the start of atosiban or MgSO₄ infusions or with the betamethasone injection. Nifedipine was administered within the same minute of OBE022 being administered.

2.5. Safety assessments

Adverse events (AEs) were recorded from screening until final visit and coded according to appropriate guidance.29, 30 Clinical laboratory parameters, 12‐lead electrocardiogram and vital signs were assessed. Additionally, in Part A, to manage the risk of hypermagnesaemia, neurological examinations were conducted by checking the subjects’ patellar reflexes and by assessing subjects using the Glasgow Coma Scale. MgSO₄ infusion safety monitoring was also performed by checking the subjects’ respiration rates.

2.6. PK analysis

Potential effects of MgSO₄, atosiban, nifedipine and betamethasone on OBE022/OBE002 PK were examined, and vice versa.

Details of all assays performed to calculate plasma concentrations are presented in the Supplementary Information. Noncompartmental analysis was used to estimate PK parameters and PK modelling to assess time‐dependent PK. Estimated parameters calculated for all drugs were: C_max and minimal plasma concentration (C_min); time C_max occurred (t_max), termination elimination rate constant (λz); terminal elimination half‐life (t_½); mean residence time (MRT); AUC from administration to last measured time‐point (AUC_0–last); AUC from time 0 extrapolated to infinity (AUC_0–∞); AUC_0–24h; apparent total plasma clearance (CL) and volume of distribution (Vdz). Values of AUC_0–24h, AUC_0–last and AUC_0–∞ were calculated using the linear/log trapezoidal method, applying the linear trapezoidal rule up to C_max and the log trapezoidal rule for the remainder of the curve. Other parameters were calculated using standard equations.

2.7. Statistical analysis

The sample size in this study was not established based on statistical considerations, a group size of 12 subjects was deemed to be appropriate to characterize the drug interaction potential of OBE022, atosiban, nifedipine, MgSO₄ and betamethasone.

The plasma PK profile of OBE022 and OBE002 were evaluated in Parts A and B. The plasma PK profile of Mg was evaluated in Part A, and the individual plasma PK profiles of atosiban, nifedipine and betamethasone were evaluated in Part B. Plasma concentrations of OBE022 and OBE002 were measured by a validated LC–MS/MS assay and analysed using noncompartmental methods to obtain estimates of the following PK parameters:

$${\mathrm{C}}_{\max },{\mathrm{C}}_{\min },{\mathrm{t}}_{\max },\mathrm{\lambda z},{\mathrm{t}}_{1/2},{\mathrm{AUC}}_{0-\text{last}},{\mathrm{AUC}}_{0–24\mathrm{h}},{\mathrm{AUC}}_{0-\infty ,}\mathrm{MRT},\mathrm{CL}\text{and}\mathrm{Vdz}.$$

Plasma concentrations of Mg, atosiban, nifedipine and betamethasone were measured by a validated assay and analysed using noncompartmental methods to obtain estimates of the following PK parameters:

$${\mathrm{C}}_{\max },{\mathrm{C}}_{\min },{\mathrm{t}}_{\max },\mathrm{\lambda z},{\mathrm{t}}_{1/2},{\mathrm{AUC}}_{0-\text{last}},{\mathrm{AUC}}_{0–24\mathrm{h}},{\mathrm{AUC}}_{0-\infty },\mathrm{MRT},\mathrm{CL}\text{and}\mathrm{Vdz}.$$

The safety set included subjects that received at least 1 dose of study drug. The PK set included subjects in the safety set with sufficient blood samples for at least 1 PK variable.

PK interaction assessments were performed on logarithmically transformed AUC_0–24, AUC_0–∞ and C_max values using analysis of variance with fixed effects for sequence (Part A only), period, treatment and a random term for subject within sequence. The 90% confidence intervals (CI) for the geometric mean ratio of AUC_0–24, AUC_0–∞ and C_max were constructed, comparing treatments administered alone versus combination treatments.

Absence of PK interaction was concluded if the 90% CI of the ratio μ_combined/μ_single was fully contained within the acceptance range for AUC and C_max (80%, 125%). Differences in t_max were explored using Hodges–Lehmann estimates and corresponding 90% CI according to the Tukey approximate method. The Wilcoxon signed rank test evaluated effect boundaries.

2.8. Assessment of drug interaction results

In line with the Food and Drug Administration guidance,25 an assessment of the clinical significance of any observed interaction was made. If the statistical analysis confirmed the absence of any PK interaction, no further interpretation of clinical significance was necessary. When a PK interaction was present, we used an approach that focussed on hazard identification, in view of the following factors:

1. No PK/pharmacodynamic curves, safety margins or no‐effect boundaries are available for any of the standard of care medications.

2. The study population for this trial was not pregnant and thus the PK data of all study drugs had to be considered with caution.

Therefore, we applied the following strategy to interpret clinical significance of PK interactions:

• 1Assessment of the impact on safety:

Decreased exposure: no impact.

1. Increased exposure: potential increase of undesired effects. This was considered of no impact if (i) no change in the safety profile was established during the study, (ii) estimates were close to bioequivalence limits and/or (iii) exposures have been shown to be safe (e.g. prescribing information, previous reports or publications).

• 2Assessment of the impact on efficacy:

Decreased exposure: potential lack of efficacy. This was considered of no impact if (i) estimates were close to bioequivalence limits and/or (ii) the decrease could be considered minor given the clinical treatment scenario.

1. Increased exposure: no impact.

2.9. Nomenclature of targets and ligands

Key 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,31 and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18.32, 33

3. RESULTS

3.1. Subject disposition

A total of 83 women were screened, of whom 36 (43.3%) passed and were suitable for admission. Reasons for screening failure included body mass index (4 subjects), unsuitable past medical history (6), laboratory blood results outside of normal range (6) and withdrawal of consent (13).

Overall, 25 subjects were randomized as shown in Figure 3. The demographic details of the subject population are shown in Table S8. For Part A; 12 subjects were included in Period 1; and, due to withdrawals 10 subjects continued to Period 2 and 9 subjects continued on to Period 3 to receive MgSO₄ coadministered with OBE022; all 9 completed. For Part B, a total of 13 subjects received treatment. Initially, 12 subjects were allocated to receive sequentially administered atosiban, nifedipine, betamethasone and OBE022 alone, then each drug coadministered with OBE022. One subject was withdrawn on Day 1 after receiving atosiban; the subject was replaced and 12 subjects completed Part B.

Figure 3.

Consort flow diagram

3.2. Safety

OBE022, alone or in combination with standard‐of‐care medications, was well tolerated. The safety and tolerability of MgSO₄, atosiban, nifedpine and betamethasone was in accordance with their reference safety information.

No severe or serious AEs occurred in Part A or Part B and all AEs either ‘recovered’ (76 AEs) or were ‘recovering’ (one AE).

In Part A, no AEs were considered related to OBE022 on its own, five AEs were considered related to the combination treatment of OBE022 and MgSO₄ (Table S1).

In Part B, six AEs were considered related to OBE022 on its own, 26 AEs were considered related to the combination treatments of OBE022 with atosiban (7 AEs) or nifedipine (19 AEs; Table S2).

The most frequently reported AEs were headache (22 events in both study parts) and dizziness (11 events in Part B only). The majority of dizziness was reported on coadministration of nifedipine and OBE022 (7 AEs) and nifedipine alone (2 AEs).

There were no clinically significant changes in laboratory safety tests, vital signs or electrocardiogram morphology time intervals and for Part A, no abnormal neurological findings.

3.3. PK results

The prodrug OBE022 was readily transformed into its stable, pharmacologically‐active metabolite OBE002: mean OBE022 peak concentrations and AUC_0–∞ were 133–246 and 771–1163 times lower than those of OBE002, respectively. Thus, this manuscript will only present general background on PK and the drug interactions evaluated using the OBE002 PK parameters as this is the clinically relevant moiety.

For interactions outside the bioequivalence range, the clinical significance was defined as follows.

Increases in exposures: Safety profile (reference safety information where available) and observed safety and tolerability during the study were considered to determine clinical significance.

Decreases in exposures: This is only relevant for OBE002 (i.e. not applicable to the other drugs). Exposures of OBE002 were still markedly above the pharmacological effect exposure range. All decreases were considered not clinically significant.

In Part B, OBE022 dosing began on Day 4. OBE002 approached steady state by Day 7, thus PK parameters could be reliably estimated when OBE022 was coadministered with other drugs.

The effect of coadministration of OBE022 with the 4 standard‐of‐care medications on their respective PK parameters and the PK of OBE002, are shown in Table S3 for MgSO₄ and Tables S5 and S6 for atosiban, nifedipine and betamethasone. The interactions between MgSO₄ and of atosiban, nifedipine and betamethasone with OBE022 and OBE002 are shown in Tables S4 and S7, respectively.

3.4. Coadministration of MgSO4 and OBE022

Mean peak OBE002 plasma concentrations (C_max) were 23% lower when OBE022 was coadministered with MgSO₄ compared to administration of OBE022 alone, the mean t_max of OBE002 was around 3 hours when OBE022 was administered alone or with MgSO₄. There was also an insignificant 8.4% decrease in AUC_0–24 but no change in AUC_0–48 when OBE022 was administered with MgSO₄ compared to administration of OBE022 alone (Table 1).

Table 1.

Summary of the effect of coadministration on pharmacokinetic (PK) parameters*

		MgSO 4 (9–11 subjects, see Table S3)	Atosiban (12 subjects)	Nifedipine (12 subjects)	Betamethasone (12 subjects)
PK of standard of care medications**	C max (ng/mL or mM for Mg) [CV%]	no change	no change	↑ by 2×	↑ 18%
		2.178 [11.37] vs 2.064 [8.08]	674.500 [14.90] vs 701.250 [15.46]	132.683 [69.20] vs 264.117 [54.06]	116.633 [12.70] vs 137.333 [11.68]
	t max (h) [CV%]	no change	no change	no change	↓ from 2.5 to 1.5 h
		0.573 [29.06] vs 0.496 [3.03]	1.128 [59.39] vs 0.921 [97.36]	0.609 [27.62] vs 0.795 [98.03]	2.458 [29.36] vs 1.450 [25.39]
	AUC (h * ng/mL) [CV%]	no change	no change	↑ by 2×	↑ 27%
		98.240 [4.45] vs 97.031 [4.07]	1579.935 [16.74] vs 1696.559 [20.89]	239.400 [23.24] vs 554.900 [21.44]	997.200 [10.24] vs 1388.800 [10.18]
OBE002 PK**	C max (ng/mL or mM for Mg) [CV%]	↓ by 23%	↓ by 22%	↑ by 29%	↑ by 35%
		615.364 [89.47] vs 477.111 [49.51]	626.417 [38.31] vs 488.250 [69.65]	626.417 [38.31] vs 857.917 [54.91]	626.417 [38.31] vs 846.833 [49.82]
	t max (h) [CV%]	no change	↑ from 3.0 to 5.7 h	↓ from 3.0 to 2.2 h	↑ from 3.0 to 3.8 h
		2.961 [53.78] vs 3.200 [57.4]	3.006 [20.22] vs 5.670 [13.76]	3.006 [20.22] vs 2.236 [95.52]	3.006 [20.22] vs 3.807 [34.96]
	AUC 0–24 (h * ng/mL) [CV%]	↓ by 8.4%	↓ by 19%	↑ by 24%	↑ by 15%
		3882.465 [69.65] vs 3555.524 [51.94]	3874.100 [47.25] vs 3156.500 [61.54]	3874.100 [47.25] vs 4894.000 [54.54]	3874.100 [47.25] vs 4471.000 [52.78]
	AUC 0‐inf (h * ng/mL) [CV%]	no change
		5582.848 [61.53] vs 5551.384 [50.28]
Clinical interpretation		No interaction of clinical significance	No interaction of clinical significance	OBE002 changes minor and of no clinical significance; Nifedipine changes clinically significant and relevant, consider reducing dose of nifedipine if used in combination with OBE022	No interaction of clinical significance

^* All PK parameters are provided in Tables S3 to S7.

^** PK values shown as: PK parameter value of drug alone [CV%] vs PK parameter of drug in combination [CV%].

Administration of OBE022 induced no effects on the timing or extent of exposure for total Mg (Table 1, Figure 4). Mean C_max and t_max of total Mg concentrations were similar for MgSO₄ alone and MgSO₄ plus OBE022. All estimates of AUC were slightly lower (<1%) after coadministration of OBE022 and MgSO₄ (Table 1).

Figure 4.

Interaction of OBE002 with MgSO₄

For OBE002, linear mixed model calculations showed all interaction estimates to be close to 1 with 90%CI exceeding bioequivalence limits; and bioequivalence values (0.8–1.25) were obtained for coadministration of OBE022 with MgSO₄ (Table S4). While OBE002 C_max, AUC_0–24 and AUC_0‐t was not fully contained in the BE limits, there was no clinically significant interactions when OBE022 was coadministered with MgSO₄ and no effect on t_max (Table 1, Figure 4).

3.5. Coadministration of atosiban and OBE022

Atosiban reduced exposures to OBE002 after i.v. administration (22% for C_max and 19% for AUC_0–24). The t_max of OBE002 increased from 3.0 to 5.7 hours with atosiban coadministration. Changes in t_max, were particularly evident while the 3‐hour atosiban infusion was ongoing, and partially recovered once atosiban was eliminated from circulation (Table 1 and Figure 5).

Figure 5.

Interaction of OBE002 with atosiban

Mean C_max and AUC of atosiban were similar with and without OBE022. However, t_max of atosiban decreased from 1.5 to 0.25 h following coadministration of OBE022 with Atosiban. Given the atosiban concentration–time curves during the 3‐hour atosiban infusion, changes in t_max, were not of PK relevance (Table 1 and Figure 5).

OBE002 C_max and AUCs were below the 90% CI for bioequivalence. The 90% CIs around the point estimates of atosiban C_max and AUCs were fully included in the 0.8–1.25 bioequivalence range (Table S7). Hence there was no clinically significant interaction between OBE022 and atosiban.

3.6. Coadministration of nifedipine and OBE022

OBE022 coadministered with nifedipine markedly increased OBE002 exposures (C_max by 29% and AUC_0–24 by 24%), with values above the 90% CI for bioequivalence. Mean nifedipine exposures (both C_max and AUC) also increased 2‐fold (there were no changes in t_max for both OBE002 and nifedipine), hence bioequivalence for nifedipine was also not obtained (Table S7). The changes in plasma levels of OBE002 were minor and of no clinical significance; however, changes in nifedipine plasma levels were considered clinically significant and relevant (Table 1, Figure 6).

Figure 6.

Interaction of OBE002 with nifedipine

3.7. Coadministration of betamethasone and OBE022

OBE022 coadministered with betamethasone resulted in increased OBE002 exposures by 35% for C_max and 15% for AUC_0–24. The t_max for OBE002 increased slightly from 3.0 to 3.8 hours with concurrent decrease of t_max for betamethasone from 2.5 to 1.5 hours. OBE022 slightly increased betamethasone exposure (C_max + 18%, AUC +27%), with 90% CI of geometric mean ratios on or just outside the upper limit of the 0.8–1.25 bioequivalence interval (Table S7). These changes were not considered clinically significant (Table 1, Figure 7).

Figure 7.

Interaction of OBE002 with betamethasone

4. DISCUSSION

Prodrug OBE022, alone or in combination with standard‐of‐care medicines, was well tolerated. Additionally, standard‐of‐care medicines were well tolerated in accordance with their reference safety information. Two of the 3 withdrawals were due to AEs following MgSO₄ and the third was due to a problem with cannulation; none was attributable to OBE022. An increased occurrence of dizziness was observed in women treated with OBE022 and nifedipine; this was not unexpected as it is a known side effect of nifedipine.34

The PK evaluation revealed some interactions outside of regulatory bioequivalence limits between OBE022/OBE002 and MgSO₄, nifedipine, betamethasone and atosiban.

Regarding their clinical significance, the interactions with MgSO₄, atosiban and betamethasone were not considered clinically significant. The rationale for this interpretation is that these interactions resulted either in a decrease of exposure to OBE002, which is not relevant given the clinical dosing scenario, or resulted in exposure increases well within the core safety information or proven to be safe during this and previous studies. For example, for atosiban the impact on the efficacy from the OBE002 exposure decrease was considered minor given the clinical treatment scenario: atosiban is eliminated within minutes after treatment stop (Figure 5); OBE022 is administered for 7 days in a clinical setting.

The interaction with nifedipine was considered to be of potential clinical significance. When coadministered with nifedipine, there was a marked increase in nifedipine exposure (C_max by 2‐fold (132.683[69.20] vs 264.117[54.06]); AUC_0–24 by 2‐fold (239.400[23.24] vs 554.900[21.44])) which could have led to the observed increased incidences of nifedipine‐related adverse reactions.

The mechanism by which both prodrug OBE022 and OBE002 are metabolized may provide an explanation for the observed drug interactions. While neither OBE022 nor OBE002 inhibit standard drug elimination pathways, both are subject to CYP‐mediated metabolism via the CYP3A4 enzyme highly expressed in the intestine and liver (ObsEva, data on file). Nifedipine and betamethasone are also metabolized by CYP3A4, and thus interaction between them and OBE022/OBE002 may be due to competitive substrate binding, as previously described for nifedipine and other CYP3A4 substrates.35

The marked interaction with nifedipine may be explained by substrate–substrate interaction with CYP3A4 when high concentrations of OBE022/OBE002 and nifedipine favour OBE022/OBE002 metabolism and inhibit nifedipine metabolism during intestinal absorption, but no mechanistic data confirming this hypothesis is currently available. Conversely, OBE022/OBE002 and betamethasone concentrations could interact with hepatic CYP3A4, impacting almost exclusively on compound excretion. Thus, the betamethasone/OBE022 interaction is unlikely to cause any clinical concern. The increased exposure to nifedipine was well tolerated under the conditions of this study and, to mitigate potentially adverse effects in preterm labour patients, the dose of nifedipine could be reduced. Alternatively, staggered oral dosing of OBE022 and nifedipine could limit intestinal CYP3A4 related interactions.

Coadministration of OBE022 and atosiban resulted in reduction in exposure to OBE002 and a significant delay in t_max was observed for OBE002 during the infusion period, indicating a concentration‐dependent interaction with atosiban. Although the resulting interaction was not considered to be relevant in a clinical preterm labour setting, a possible explanation could be the documented effect of atosiban on gastric emptying.36

It is accepted that the study design had the following limitations: The study was conducted in healthy nonpregnant women and the sample size was not based on a statistical power analysis. Physiological differences and statistical limitations should be considered when extrapolating our results to preterm labour patients. Additionally, in contrast to DDI studies with index drugs, results from a concomitant‐use study with a nonindex drug can be difficult to extrapolate to other drugs, however, in this study the concomitant medications given are those which will be administered in a clinical setting with OBE022 and therefore study of the DDI of these drugs is useful.

5. CONCLUSION

The use of OBE022, a PGF2α antagonist prodrug in combination with standard of care medicines and other tocolytic treatments may provide new treatment alternatives for preterm labour. Nifedipine doses could potentially be reduced and/or staggered dosing implemented when coadministered with OBE022. There were no clinically relevant PK interactions between prodrug OBE022 and MgSO₄, betamethasone or atosiban; however, nifedipine exposure increased notably. Coadministration of OBE022 with MgSO₄, betamethasone, atosiban and nifedipine raised no safety concerns and may provide new effective treatment alternatives for preterm labour.

The results from this study, as well as favourable tolerability and safety results from the FIH study23 using OBE022, enabled the initiation of a Phase 2a study with OBE022 in pregnant women with spontaneous preterm labour with a gestational age of 24^0/7–33^6/7 weeks (PROLONG, NCT03369262).

COMPETING INTERESTS

U.L., J.T. and S.C. are employees of Richmond Pharmacology Ltd. J.P.G., L.M. and O.P. are employees of ObsEva SA. OBE022 is being developed by ObsEva SA. ObsEva SA funded this clinical trial, which was performed by Richmond Pharmacology Ltd.

CONTRIBUTORS

All authors conceived and designed the experiments. S.C., U.L. and J.T. performed the experiments. All authors read and revised the article and approved the final version.

Supporting information

Data S1.

Supporting information

Data S2.

Table S1. Summary of adverse events reported by more than 1 subject – Part A

Table S2. Summary of adverse events reported by more than 1 subject – Part B

Table S3. Plasma pharmacokinetic parameters of OBE002 and MgSO₄

Table S4. Evaluation of the interaction between MgSO₄ and OBE022/OBE002

Table S5. Pharmacokinetic parameters for atosiban, nifedipine and betamethasone in plasma, with and without OBE022 steady state dosing

Table S6. Summary of pharmacokinetic parameters of OBE002 with and without coadministered drugs

Table S7. Evaluation of the interaction between OBE002 in steady state and noninvestigational drugs

Table S8. Demographic data – summary table (randomized set)

Pohl O, Marchand L, Gotteland J‐P, Coates S, Täubel J, Lorch U. Coadministration of the prostaglandin F2α receptor antagonist preterm labour drug candidate OBE022 with magnesium sulfate, atosiban, nifedipine and betamethasone. Br J Clin Pharmacol. 2019;85:1516–1527. 10.1111/bcp.13925