Effects of proton pump inhibitors, esomeprazole and vonoprazan, on the disposition of proguanil, a CYP2C19 substrate, in healthy volunteers

Abstract

Aims

Vonoprazan, a new class of potassium‐competitive proton pump inhibitors has been found to attenuate the antiplatelet function of clopidogrel in a recent clinical study, despite weak in vitro activity against CYP2C19. To elucidate the mechanism of this interaction, the present study investigated the effects of esomeprazole and vonoprazan on the pharmacokinetics of proguanil, a CYP2C19 substrate.

Methods

Seven healthy male volunteers (CYP2C19 extensive metabolizers) received a single oral administration of 100 mg proguanil/250 mg atovaquone (control phase), oral esomeprazole (20 mg) for 5 days followed by proguanil/atovaquone (esomeprazole phase) and oral vonoprazan (20 mg) for 5 days followed by proguanil/atovaquone (vonoprazan phase). Concentrations of proguanil and its metabolite, cycloguanil, in plasma and urine in each phase were determined using liquid chromatography–tandem mass spectrometry.

Results

Coadministration with proton pump inhibitors resulted in increase and decrease in the area under the plasma concentration–time curve (AUC) of proguanil and cycloguanil, respectively, significantly reducing their AUC ratio (cycloguanil/proguanil) to 0.317‐fold (95% confidence interval [CI] 0.256–0.379) and 0.507‐fold (95% CI 0.409–0.605) in esomeprazole phase and vonoprazan phase, respectively. Esomeprazole and vonoprazan also significantly reduced the apparent formation clearance (cumulative amount of cycloguanil in urine divided by AUC of proguanil) to 0.324‐fold (95% CI 0.212–0.436) and 0.433‐fold (95% CI 0.355–0.511), respectively, without significant changes in renal clearance of proguanil and cycloguanil.

Conclusions

Although further studies are needed, both esomeprazole and vonoprazan potentially inhibit CYP2C19 at clinical doses, suggesting caution in the coadministration of these drugs with CYP2C19 substrates.

What is already known about this subject

• In vitro studies using human liver microsomes have shown that vonoprazan, a proton pump inhibitor, hardly inhibits cytochrome P450 2C19 (CYP2C19) at the clinically relevant concentration.

• However, it has been clinically reported that the antiplatelet effect of clopidogrel, which is metabolically activated by CYP2C19, is more strongly attenuated by vonoprazan than by another proton pump inhibitor, esomeprazole.

• Vonoprazan‐related drug–drug interactions have not been fully elucidated.

What this study adds

• Vonoprazan significantly inhibited proguanil metabolism in our clinical study, suggesting its potential inhibition against CYP2C19.

• There is a possibility of both esomeprazole and vonoprazan causing pharmacokinetic drug–drug interactions with CYP2C19 substrates at their therapeutic doses.

1. INTRODUCTION

Proton pump inhibitors (PPIs), having potent gastric acid secretion inhibitory activity, are used in the treatment of gastric/duodenal ulcers and reflux oesophagitis as well as Helicobacter pylori eradication. Five PPIs (omeprazole, lansoprazole, rabeprazole, esomeprazole and vonoprazan) are commercially available as prescription drugs in Japan.

Omeprazole perpetrates drug–drug interaction (DDI) with cytochrome P450 2C19 (CYP2C19) substrate drugs. CYP2C19 is involved in the metabolic activation of clopidogrel, citalopram and proguanil and its polymorphisms are associated with changes in the pharmacokinetics of drugs such as omeprazole, lansoprazole, pantoprazole, phenytoin, voriconazole, citalopram and carisoprodol.1 Omeprazole (a racemate) as well as its R‐isomer and S‐isomer (esomeprazole) all have in vitro CYP2C19 inhibitory effect, and this inhibitory effect is enhanced by pre‐incubation with human liver microsomes in the presence of NADPH, thus indicating time‐dependent inhibition (TDI).2, 3, 4 Various omeprazole metabolites are also found to inhibit CYP2C19.5

CYP2C19‐mediated pharmacokinetic DDI by omeprazole is clinically‐relevant. The antiplatelet effects of clopidogrel, which is metabolically activated by CYP2C19, are attenuated when coadministered with omeprazole, with an increased risk of re‐hospitalization for treatment of acute coronary syndrome.6, 7 Furthermore, repeated oral administration of omeprazole perpetrates DDI with CYP2C19 substrates such as citalopram,8 escitalopram9 and proguanil.10

Vonoprazan is a new category of PPI referred to as a potassium‐competitive acid blocker. Vonoprazan had not been considered to perpetrate CYP2C19‐mediated DDI provided that it is a weak CYP2C19 inhibitor.11 The 50% inhibitory concentration (IC₅₀) of vonoprazan against S‐mephenytoin 4′‐hydroxylation (a CYP2C19 probe reaction) was >30 μM and that pre‐incubation for 30 minutes in the presence of NADPH resulted in a decrease in IC₅₀ to 13 μM in human liver microsomes.12 However, the maximum plasma unbound concentration (approximately 0.01 μM), as calculated based on the maximum plasma concentration (C_max) of 23.32 ng/mL11 during repeated oral administration of the clinical dose (20 mg) of vonoprazan and the human plasma protein binding of 86.5% (vonoprazan 100 ng/mL),11 is far lower than the aforementioned IC₅₀.

It was recently reported that the antiplatelet effects of clopidogrel are attenuated by coadministration with either esomeprazole or vonoprazan. In that report, the antiplatelet effects of clopidogrel were more strongly attenuated by repeated oral administration of 10 mg vonoprazan than by repeated oral administration of 20 mg esomeprazole.13 However, it remains controversial whether the DDI between clopidogrel and vonoprazan was caused by pharmacokinetic interaction because of lack of its inhibitory effect on the formation of active metabolite of clopidogrel in vitro.14, 15

The present DDI study was conducted in healthy volunteers to clarify the risk of CYP2C19‐mediated DDIs associated with esomeprazole and vonoprazan. Proguanil was selected as the CYP2C19 substrate, and the plasma and urine concentrations of proguanil and the metabolite cycloguanil ¹⁰ were measured over time with and without coadministration of esomeprazole or vonoprazan.

As the plasma concentration of esomeprazole and vonoprazan reaches a steady state on day 3 of repeated oral administration,11, 16 proguanil was administered after 5 days of repeated administration of either drug in this study. In view of CYP2C19 turnover (elimination half‐life = 26 h),17 a 7‐day washout period was established as a sufficient period of time for recovery of enzyme activity.

2. METHODS

2.1. Materials

Nexium capsules were obtained from AstraZeneca K.K. (Osaka, Japan) and Takecab tablets from Takeda (Tokyo, Japan), while Maralon combination tablets were obtained from Glaxo Smith Kline (Middlesex, UK). Proguanil was purchased from the US Pharmacopeial Convention (MD, USA). Cycloguanil and proguanil‐d ₆ were purchased from Toronto Research Chemicals (North York, Canada). Esomeprazole was purchased from LKT Laboratories (MN, USA). Vonoprazan was purchased from ChemScene (NJ, USA). All other reagents were of analytical grade and were commercially obtained.

2.2. Subjects

From the panel of 12 healthy Japanese adult male volunteers who had been registered as CYP2C19 extensive metabolizers in P‐one clinic, and provided informed consent, 7 volunteers were enrolled in this study. A review of previous medical history and the results of medical examination, laboratory tests, vital signs (sitting blood pressure, pulse rate and axillary body temperature), 12‐lead electrocardiography, infection tests and urine drug tests confirmed that no subjects had any diseases requiring treatment. The exclusion criteria are summarized in Supporting Information (Appendix Text). The median (minimum‐maximum) age was 37 (20–42) years, median body height was 167.3 (163.2–182.9) cm, median body weight was 59.2 (51.3–76.7) kg and median body mass index was 21.2 (19.3–24.3) kg/m². No subjects withdrew prematurely from the study.

2.3. Study protocol

We conducted an open‐label, single‐arm study with a total of 3 phases. Control phase (1^st phase): One Malarone tablet (atovaquone 250 mg and proguanil 100 mg combination) was orally administered. Esomeprazole phase (2^nd phase): One Nexium capsule (esomeprazole 20 mg) per day was orally administered for 5 consecutive days, and one Malarone tablet (atovaquone 250 mg and proguanil 100 mg combination) was orally administered simultaneously on day 5 only. Vonoprazan phase (3^rd phase): one Takecab tablet (vonoprazan 20 mg) per day was orally administered for 5 consecutive days, and one Malarone tablet (atovaquone 250 mg and proguanil 100 mg combination) was orally administered simultaneously on day 5 only. A 250‐mg dose of atovaquone was unlikely to affect the pharmacokinetics of esomeprazole, vonoprazan, proguanil and cycloguanil (see below). All drugs were orally administered with 250 mL of water in fasting condition, and it was confirmed that the medication had been taken. In each phase, subjects were hospitalized the day before study medication, and were discharged 2 days after the last study medication dose. There was a 7‐day washout period between each phase. The dose of each drug used in the study was the standard dose used in Japan (Figure 1).

Figure 1.

Time schedule of the present clinical study. p.o., per os

This study was conducted in accordance with the Ethical Guidelines for Medical and Health Research Involving Human Subjects (partially revised on February 28, 2017) and the Declaration of Helsinki (revised in 2013). The study was conducted with the approval of the Ethics Committee of Kameda Medical Center (Approval No. 17–066), the Ethics Committee of P‐One Clinic, Keikokai Medical Corporation (approval dated August 29, 2017), and the Ethics Committee for Scientific Research Involving Human Subjects of the Graduate School of Pharmaceutical Sciences, the University of Tokyo (Approval No. 29–08). The study was registered in the University Hospital Medical Information Network Clinical Trials Registry at http://www.umin.ac.jp/ctr/index.htm (UMIN 000029539).

To ensure study quality, the study was managed by a medical doctor who had completed Good Clinical Practice Education and Training (TransCelerate, e‐learning program) and monitored by a clinical research coordinator from P‐One Clinic.

2.4. Sample collection

Blood samples were taken by direct venepuncture (sodium heparin anticoagulant) before dosing and at 1, 2, 3, 4, 6, 8, 10, 24 and 48 hours after dosing. Blood samples were centrifuged to isolate plasma, which was stored at −70°C until analysis. Urine samples were collected at 0–8, 8–24, and 24–48 hours after dosing.

2.5. Quantification of the drug concentration by liquid chromatography–tandem mass spectrometry

The plasma concentration of proguanil, cycloguanil, atovaquone, esomeprazole and vonoprazan and the urinary concentration of proguanil and cycloguanil were determined as follows: Urine specimens were diluted 1,000‐fold with Milli‐Q water. The plasma and diluted urine specimens (20 μL) were mixed with 80 μL of internal standard solution (2.6 ng/mL proguanil‐d ₆ in acetonitrile). Mixed solutions were centrifuged at 20,000 × g for 10 minutes, and the supernatant was subjected to liquid chromatography–tandem mass spectrometry analysis. The liquid chromatography system used was a Prominence UFLC (Shimadzu, Kyoto, Japan) equipped with a PC HILIC column (3.0 μm, 2.0 × 150 mm; Shiseido, Tokyo, Japan) with an isocratic mobile phase: 20% acetonitrile/80% water with 10 mM ammonium acetate and 0.1% formic acid (A); 95% acetonitrile/5% water with 10 mM ammonium acetate and 0.1% formic acid (B) (A: B = 30: 70) at a flow rate of 0.4 mL/min. Mass spectrometric data were acquired using QTRAP 5500 (AB SCIEX, Foster City, CA, USA) equipped with an electrospray ionization source. The multiple reaction monitoring precursor/product ion transitions are summarized in Supporting Information (Appendix Table S1). The linearity of the calibration curves and the lower limit of quantification of drugs are shown in Supporting Information (Appendix Table S2).

2.6. Pharmacokinetic analysis

The maximum plasma drug concentration (C_max) and the time to reach C_max (T_max) were obtained directly from the data. The area under the plasma concentration–time curve from 0 to 48 hours (AUC_0–48) was calculated by the trapezoidal rule. AUC from 0 hours to infinity (AUC_0–∞) was calculated as the sum of AUC_0–48 and AUC from 48 hours to infinity (AUC_48–∞). AUC_48–∞ was calculated as the drug concentration at 48 hours divided by the slope of the log‐linear regression of the terminal elimination phase. The renal clearance (CL_r) was calculated by division of the cumulative amount of drug in urine collected for 48 hours by AUC_0–48. The apparent systemic clearance (CL/F) was calculated by division of the dose by AUC_0–∞. The apparent distribution volume (Vd/F) was calculated by division of the dose by C_max. The metabolic ratio was calculated by division of AUC_0–∞ of cycloguanil by AUC_0–∞ of proguanil. The apparent formation clearance was calculated by division of the cumulative amount of cycloguanil in urine up to 48 hours by AUC_0–48 of proguanil.

2.7. Statistical analyses

Differences in the pharmacokinetic parameters were compared by Dunnett's t‐test using SPSS software version 24 (IBM Japan, Tokyo). A value of p < 0.05 was considered statistically significant.

2.8. 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,18 and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18.19, 20

3. RESULTS

3.1. Adverse reactions during the study

No adverse events were reported in any subjects during the control phase or esomeprazole phase. Nonserious, moderate constipation was reported in only 1 subject in the vonoprazan phase. No treatment was required for the event, and the resolution was confirmed at post‐study examination. It was also concluded that the drug taken in the study was not causally related to the constipation reported during the vonoprazan phase.

3.2. Pharmacokinetics of proguanil and cycloguanil without administration of esomeprazole/vonoprazan

The plasma concentration–time profiles of proguanil and cycloguanil during the control phase are shown in Figure 2A, the cumulative amounts of urinary excretion of proguanil and cycloguanil are shown in Figure 3A, and the pharmacokinetic parameters are shown in Table 1. The T_max of cycloguanil was 3.4 hours longer than that of proguanil, and C_max and AUC_0–∞ were 31% and 16% lower, respectively, compared with those of proguanil. Both compounds were nearly completely eliminated from blood 48 hours postdose. Almost the same amount of proguanil and cycloguanil were excreted in urine up to 48 hours.

Figure 2.

The plasma concentration–time profiles for proguanil (closed circles and solid lines) and cycloguanil (open circles and dotted lines) in healthy subjects. A, Single oral administration of 100 mg proguanil/250 mg atovaquone (control phase); B, oral administration of esomeprazole (20 mg/day) for 5 days followed by 100 mg proguanil/250 mg atovaquone (esomeprazole phase); C, oral administration of vonoprazan (20 mg/day) for 5 days followed by 100 mg proguanil/250 mg atovaquone (vonoprazan phase). Inlet shows the semi‐log plots of the profiles. n = 7, mean + standard deviation

Figure 3.

The cumulative urinary excretion of proguanil (closed circles and solid lines) and cycloguanil (open circles and dotted lines) in healthy subjects. A, Single oral administration of 100 mg proguanil/250 mg atovaquone (control phase); B, oral administration of esomeprazole (20 mg/day) for 5 days followed by 100 mg proguanil/250 mg atovaquone (esomeprazole phase); C, oral administration of vonoprazan (20 mg/day) for 5 days followed by 100 mg proguanil/250 mg atovaquone (vonoprazan phase). n = 7, mean + standard deviation

Table 1.

Pharmacokinetic parameters for proguanil and cycloguanil

	Control phase	Esomeprazolephase	Vonoprazanphase
Proguanil
Cmax (ng/mL)	128 ± 75	132 ± 34	133 ± 31
Tmax (h)	2.86 ± 1.77	2.86 ± 0.90	2.71 ± 0.49
AUC0–48 (ng h/mL)	1481 ± 444	2351 ± 688*	1998 ± 415
AUC0–∞ (ng h/mL)	1611 ± 459	2662 ± 762**	2221 ± 439
Relative change (vs control phase)a		1.67 (1.40–1.94)	1.42 (1.17–1.67)
CL/F (L/h)	66.3 ± 17.2	40.3 ± 11.6**	46.7 ± 10.3*
Vd/F (L)	920 ± 285	796 ± 182	785 ± 181
Urinary excretion (mg)b	18.7 ± 3.1	33.8 ± 6.4**	20.5 ± 9.4
Relative change (vs control phase)a		1.83 (1.54–2.12)	1.07 (0.70–1.43)
CLr (L/h)	13.1 ± 2.4	7.2 ± 6.9	5.9 ± 6.9
Cycloguanil
Cmax (ng/mL)	87.5 ± 41.0	27.4 ± 15.3**	47.8 ± 20.3*
Tmax (h)	6.29 ± 0.76	5.71 ± 0.76	6.57 ± 1.51
AUC0–48 (ng h/mL)	1245 ± 539	569 ± 281*	825 ± 400
AUC0–∞ (ng h/mL)	1299 ± 549	677 ± 313*	916 ± 459
Relative change (vs control phase)a		0.522 (0.442–0.602)	0.716 (0.551–0.881)
Urinary excretion (mg)b	18.0 ± 6.3	8.7 ± 2.7**	10.6 ± 3.7*
Relative change (vs control phase)a		0.504 (0.367–0.641)	0.593 (0.484–0.702)
CLr (L/h)	15.8 ± 4.8	17.5 ± 6.5	14.3 ± 4.9
Metabolic ratioc	0.837 ± 0.373	0.261 ± 0.115***	0.410 ± 0.168**
Relative change (vs control phase)a		0.317 (0.256–0.379)	0.507 (0.409–0.605)
Apparent formation clearance (L/h)d	12.6 ± 3.5	3.98 ± 1.73***	5.36 ± 1.72***
Relative change (vs control phase)a		0.324 (0.212–0.436)	0.433 (0.355–0.511)

Data shown as mean ± standard deviation. The significance of differences was assessed by Dunnett's t‐test

^* p < 0.05.

^** p < 0.01.

^*** p < 0.001 vs control.

^a Data shown as mean (95% confidence interval).

^b Cumulative amount of compound excreted up to 48 hours in the urine.

^c AUC_0–∞ of cycloguanil/AUC_0–∞ of proguanil.

^d Cumulative amount of cycloguanil excreted up to 48 hours in the urine/AUC_0–48 of proguanil.

C_max, maximum plasma concentration; T_max, time to reach C_max; AUC, area under the plasma concentration–time curve; CL/F, apparent systemic clearance; Vd/F, apparent distribution volume; CL_r, renal clearance.

3.3. Effect of esomeprazole on the pharmacokinetics of proguanil and cycloguanil

Figure 4A shows the plasma trough concentrations up to the 4^th dose and the plasma concentration–time profile after the 5^th dose during repeated esomeprazole administration. Esomeprazole was nearly completely eliminated from blood 24 hours post‐dose, revealing no accumulation with repeated administration, which is consistent with a previous report of reaching a steady state on day 3 of its repeated oral administration.16

Figure 4.

The plasma concentration–time profiles for esomeprazole and vonoprazan in healthy subjects. A, Oral administration of esomeprazole (20 mg/day) for 5 days (esomeprazole phase); B, oral administration of vonoprazan (20 mg/day) for 5 days (vonoprazan phase). Inlet shows the semi‐log plots of the profiles. n = 7, mean + standard deviation

The plasma concentration–time profiles of proguanil and cycloguanil following simultaneous administration of an atovaquone/proguanil combination at the 5^th dose during repeated esomeprazole administration (esomeprazole phase) are shown in Figure 2B, the cumulative amount of urinary excretion are shown in Figure 3B, and the pharmacokinetic parameters are shown in Table 1. Esomeprazole administration resulted in a significant increase (1.67‐fold) in the AUC_0–∞ of proguanil and a significant decrease (0.522‐fold) in that of cycloguanil. Correspondingly, a significant increase (1.83‐fold) and a significant decrease (0.504‐fold), respectively, were observed in the cumulative amount of urinary excretion up to 48 hours. The proguanil metabolic ratio (AUC_0–∞ of cycloguanil/AUC_0–∞ of proguanil) and the apparent formation clearance (cumulative amount of cycloguanil in urine/AUC_0–48 of proguanil) were significantly lower (0.317‐fold and 0.324‐fold, respectively) in the esomeprazole phase compared with the control phase. There were no significant changes in the CL_r of proguanil and cycloguanil, but that of proguanil was lower than in the control phase.

3.4. Effect of vonoprazan on the pharmacokinetics of proguanil and cycloguanil

Figure 4B shows the plasma trough concentrations up to the 4^th dose and the plasma concentration–time profile after the 5^th dose during repeated vonoprazan administration. Vonoprazan showed no accumulation with repeated administration, which was consistent with a previous report of reaching a steady state on day 3 of its repeated oral administration.11

The plasma concentration–time profiles of proguanil and cycloguanil following simultaneous administration of the atovaquone/proguanil combination at the 5^th dose during repeated vonoprazan administration (vonoprazan phase) are shown in Figure 2C, the cumulative amounts of urinary excretion are shown in Figure 3C, and the pharmacokinetic parameters are shown in Table 1. Vonoprazan administration was associated with an increase (1.42‐fold) in the AUC_0–∞ of proguanil and a decrease (0.716‐fold) in that of cycloguanil, although the differences were not significant. The cumulative amount of urinary excretion of proguanil up to 48 hours was about the same as in the control phase (1.07‐fold), while that of cycloguanil was significantly decreased (0.593‐fold). The proguanil metabolic ratio (AUC_0–∞ of cycloguanil/AUC_0–∞ of proguanil) and the apparent formation clearance (cumulative amount of cycloguanil in urine/AUC_0–48 of proguanil) were significantly lower (0.507‐fold and 0.433‐fold, respectively) in the vonoprazan phase when compared with the control phase. There were no significant changes in the CL_r of proguanil and cycloguanil, but that of proguanil was lower than in the control phase.

4. DISCUSSION

In this study, the coadministration of esomeprazole or vonoprazan was confirmed to change the pharmacokinetics of proguanil, used as a CYP2C19 substrate, in healthy volunteers.

Esomeprazole is the S‐isomer of omeprazole and is known to be associated with the risk of CYP2C19‐mediated DDIs.21 The proguanil metabolic ratio (AUC_0–∞ of cycloguanil/AUC_0–∞ of proguanil) and the apparent formation clearance (cumulative amount of cycloguanil in urine/AUC_0–48 of proguanil) were significantly lower in the esomeprazole phase than in the control phase. A significant increase in the urinary excretion of proguanil and a significant decrease in that of cycloguanil were also observed (Table 1). These results suggest that esomeprazole inhibits the metabolism of proguanil to cycloguanil. Repeated oral administration of the racemate omeprazole (40 mg/day, 7 days) reportedly resulted in a 1.49‐fold increase in proguanil AUC and a 0.53‐fold decrease in cycloguanil AUC.22 In the present study, oral administration of esomeprazole 20 mg day for 5 consecutive days resulted in a 1.67‐fold increase in proguanil AUC_0–∞ and a 0.522‐fold decrease in cycloguanil AUC_0–∞ (Table 1). It has been reported that CYP2C19 inhibitory effect by esomeprazole is stronger than the inhibition by omeprazole (the ratio of the maximum inactivation rate constant [k_inact] and the concentration causing half‐maximal inactivation [K_I] for esomeprazole was 2.1‐fold higher than that for omeprazole4), thus suggesting that the changes in the AUC of proguanil and cycloguanil in the esomeprazole phase observed in the present study are consistent with the results of the previous study on the administration of omeprazole 40 mg/day^.22

Although weaker than during esomeprazole coadministration, the proguanil metabolic ratio (AUC_0–∞ of cycloguanil/AUC_0–∞ of proguanil) in the vonoprazan phase was lower than in the control phase (esomeprazole coadministration: 0.317‐fold; vonoprazan coadministration: 0.507‐fold). Although also weaker than during esomeprazole coadministration, the cumulative amount of urinary excretion of proguanil increased (esomeprazole coadministration: 1.83‐fold; vonoprazan coadministration: 1.07‐fold) and that of cycloguanil decreased (esomeprazole coadministration: 0.504‐fold; vonoprazan coadministration: 0.593‐fold), thus suggesting that vonoprazan also inhibited the metabolism of proguanil to cycloguanil (Table 1).

The AUC_0–∞ of cycloguanil in vonoprazan phase tended to be lower than in control phase, but the difference was not statistically significant, which might be attributed in part to the large inter‐individual variability (Table 1, Figure 5). There was a significant negative correlation (p < 0.01) between the AUC_0–∞ and CL_r of cycloguanil (Figure 6), suggesting that the inter‐individual variability in the renal excretion of cycloguanil was partly responsible for the interindividual variability in its AUC_0–∞.

Figure 5.

The individual values of AUC_0–∞ of proguanil A, AUC_0–∞ of cycloguanil B, metabolic ratio (AUC_0–∞ of cycloguanil/AUC_0–∞ of proguanil; C, and apparent formation clearance (cumulative amount of cycloguanil in urine/AUC_0–48 of proguanil; d) in 3 phases. n = 7

Figure 6.

The relationship between AUC_0–∞ and CL_r of cycloguanil. The symbols represent the values for individual subjects (closed circles: control phase, open squares: esomeprazole phase, open triangles: vonoprazan phase, n = 7 each). The dotted line represents linear least‐squares regression line (n = 21)

In in vitro studies using human liver microsomes, vonoprazan weakly inhibited S‐mephenytoin 4′‐hydroxylation12 and had almost no effect on the conversion of clopidogrel to its active metabolite H4.14 However, the results of the present study suggest that vonoprazan has a significant inhibitory effect on CYP2C19 metabolism in vivo, and the previously reported13 attenuation of the effects of clopidogrel when coadministered with vonoprazan may therefore be based on CYP2C19 inhibition. Caution should be taken on combination use of vonoprazan and CYP2C19 substrates until absence of significant DDI with the combination use is guaranteed in clinical studies.

An in vitro study using human liver microsomes reported that vonoprazan showed TDI against CYP3A4 and CYP2B6. Although CYP3A4 is partly involved in the formation of cycloguanil from proguanil,10 Sakurai et al. excluded the possibility of CYP3A4‐medaited DDI by vonoprazan based on a clinical study of vonoprazan with clarithromycin.23 By contrast, vonoprazan, but not esomeprazole, attenuated the antiplatelet function of prasugrel, which is metabolically activated mainly by CYP2B6 and CYP3A4.13, 24 Vonoprazan might therefore have a DDI potential with CYP2B6 at its therapeutic doses by TDI or other mechanisms.

The pharmacokinetic DDI between vonoprazan and proguanil remains unaccountable based on preclinical data. One of the possibilities is the contribution of the metabolites of vonoprazan. Vonoprazan is metabolized by CYP, UDP‐glucuronosyl transferase and sulfotransferase to various metabolites,25 which could potentially inhibit CYP2C19 as in the case of esomeprazole. Alternatively, the inhibitory effect of vonoprazan may be substrate dependent, as substrate dependency has been reported for the potency of CYP2C19 inhibition estimated via in vitro studies using human liver microsomes.4, 26 Both proguanil and cycloguanil have been shown to be substrates of organic cation transporter (OCT) 1 (a transporter involved in hepatic uptake) and also transporters involved in renal excretion (OCT2 and multidrug and toxin extrusion protein [MATE] 1/MATE2‐K).27 Because hepatic uptake is the first step of hepatic elimination, inhibiting the hepatic uptake of proguanil could result in the decrease in cycloguanil production. Recently, Matthaei et al. reported that OCT1 allele showing reduced transport activity is associated with lower cycloguanil‐to‐proguanil ratio.28 Thus, this offers a possibility of OCT1 inhibition by vonoprazan as the underlying mechanism. However, the inhibitory effect of vonoprazan on OCT1‐mediated transport was found to be very weak with no effect at 1 μM (345 ng/mL; (Supporting Information Appendix Figure S1)), suggesting that the inhibition of OCT1‐mediated hepatic uptake by vonoprazan is not a mechanism of the observed DDI. In addition, PPIs other than vonoprazan also do not appear to inhibit OCT1 in vivo despite showing some inhibitory effects in vitro.29, 30 Absence of vonoprazan effect on CL_r of cycloguanil excludes the inhibition of OCT2 and MATE1/MATE2‐K by vonoprazan. Alternatively, the absorption of proguanil might have been affected by esomeprazole and vonoprazan through the increase in gastric pH, as has been suggested for the interaction between esomeprazole and sonidegib.31 Further studies are warranted to elucidate the mechanism of the interaction between vonoprazan and proguanil, which might have been difficult to detect based on conventional preclinical studies.

As proguanil alone is not commercially available in Japan, a combination formulation with atovaquone 250 mg was used in this study. It has been reported that atovaquone is hardly metabolized by CYP2C19 or CYP3A432 and does not inhibit CYP2C19.33 Meanwhile, it has been reported that atovaquone inhibits CYP3A4 (IC₅₀ = 4.70 μM),33 but that the C_max of unbound atovaquone (≤0.037 μM calculated from C_max 33 and plasma unbound fraction34) after the administration of 250 mg is far lower than the IC₅₀ value noted above, thus suggesting that atovaquone is unlikely to inhibit CYP3A4 in vivo. Indeed, it has been reported that the differences in the AUCs of atovaquone, proguanil and cycloguanil were within 10% when 1000 mg atovaquone and 400 mg proguanil were given alone or in combination in healthy volunteers.35 The plasma concentration–time profiles of proguanil and cycloguanil in the control phase (Figure 2A) were nearly the same as previously reported during the oral administration of proguanil alone,22 thus suggesting that atovaquone was unlikely to affect the pharmacokinetics of proguanil in this study. As the plasma esomeprazole concentration (esomeprazole phase) was also nearly the same as reported previously,16 it appeared unlikely that atovaquone and proguanil affect the pharmacokinetics of esomeprazole (Figure 4A). Meanwhile, plasma concentration of vonoprazan (which is metabolized by CYP3A425) was higher than previously reported12, 36 and the AUC of atovaquone was 1.90‐fold higher in vonoprazan phase than in control phase (Supporting Information (Figure S2, Appendix Table S3). Although the underlying mechanism is not known, the observed atovaquone concentrations were comparable to previously reported values in control and esomeprazole phases,33, 34 and even those in vonoprazan phase were not sufficient to inhibit CYP3A4. Furthermore, the plasma vonoprazan concentration 24 h after the fifth dose of vonoprazan (simultaneously administered with atovaquone/proguanil combination) was nearly the same as the plasma trough concentrations after the first to fourth doses, thus suggesting that atovaquone was unlikely to inhibit the metabolism of vonoprazan in this study (Figure 4B).

There are 2 limitations of this study. First, this study was conducted as 3 phases of fixed order, which could have resulted in some order effects. However, a simulation analysis using physiologically based pharmacokinetic model has shown that the 7‐day washout period is more than enough for the inhibitory effect of esomeprazole on CYP2C19 to be completely recovered, based on its elimination half‐life of 26 h, by the time of administration of proguanil in the vonoprazan phase (data not shown). Second, number of days of hospitalization for control phase (4 days) was shorter than that for esomeprazole phase and vonoprazan phase (8 days each), which could have contributed to the differences observed in the disposition of proguanil and cycloguanil. However, at least proguanil and cycloguanil were completely eliminated by the start of the esomeprazole phase (9 days after the first proguanil administration) as shown in Figure 2.

In conclusion, the results of this study suggest that clinical doses of esomeprazole and vonoprazan may both potentially inhibit CYP2C19 and result in pharmacokinetic drug interactions. The use of these drugs in combination with CYP2C19 substrates, such as clopidogrel, citalopram and proguanil,1 requires caution.

COMPETING INTERESTS

There are no competing interests to declare.

CONTRIBUTORS

R.F., T.K., H.K., K.I. wrote the article; R.F., T.K., H.K., K.I. designed the research; H.K., Y.T., K.F. performed the research; H.K., T.K., K.I. analysed the data. All authors revised the article critically for important intellectual content and approved the final version.

Supporting information

Table S1. Analytical conditions for test compounds

Table S2. Validation parameters of liquid chromatography–tandem mass spectrometry Figure S1. Inhibitory effect of vonoprazan against hOCT1‐mediated uptake of [¹⁴C]TEA and proguanil‐d₆

Figure S2. The plasma concentration–time profiles for atovaquone in healthy subjects. Closed circles and solid line: single oral administration of 100 mg proguanil/250 mg atovaquone (control phase), open squares and dotted line: oral administration of esomeprazole (20 mg/day) for 5 days followed by 100 mg proguanil/250 mg atovaquone (esomeprazole phase), open triangles and chain line: oral administration of vonoprazan (20 mg/day) for 5 days followed by 100 mg proguanil/250 mg atovaquone (vonoprazan phase). Inset shows the semi‐log plots of the profiles. n = 7, mean + standard deviation

Table S3. Pharmacokinetic parameters for atovaquone

Funakoshi R, Tomoda Y, Kudo T, Furihata K, Kusuhara H, Ito K. Effects of proton pump inhibitors, esomeprazole and vonoprazan, on the disposition of proguanil, a CYP2C19 substrate, in healthy volunteers. Br J Clin Pharmacol. 2019;85:1454–1463. 10.1111/bcp.13914