A multicentre prospective study evaluating the impact of proton‐pump inhibitors omeprazole and pantoprazole on voriconazole plasma concentrations

Abstract

Voriconazole is an antifungal metabolised by CYP2C19 enzyme, which can be inhibited by proton‐pump inhibitors (PPIs). A prospective observational study was carried out to determine the influence of PPIs on voriconazole pharmacokinetic. The 78 patients included were divided into 4 groups: omeprazole (n = 32), pantoprazole (n = 25), esomeprazole (n = 3) and no PPI (n = 18). Patients with no PPI had no significant difference in plasma voriconazole concentration when compared with those with PPI (2.63 ± 2.13 μg/mL [95% confidence interval {CI} 1.57–3.69] vs 2.11 ± 1.73 μg/mL [95%CI 1.67–2.55], P > .05). However, voriconazole plasma concentration was significantly lower in patients treated with pantoprazole vs those treated with omeprazole (1.44 ± 1.22 μg/mL [95%CI 0.94–1.94) vs 2.67 ± 1.88 μg/mL [95%CI 2.02–3.32], P = .013) suggesting a greater CYP2C19 enzyme inhibitory effect of omeprazole. This study demonstrates the greater CYP inhibition capacity of omeprazole and should be helpful for the choice of PPIs for patients treated with voriconazole.

What is already known about this subject

• Voriconazole is metabolized by CYP2C19 and proton‐pump inhibitors are enzymatic inhibitors of CYP2C19. Although the inhibitory effect of omeprazole has been demonstrated in in vivo studies, there is little information about the interaction of voriconazole and other proton‐pump inhibitors different to omeprazole.

What this study adds

• This is a multicentre prospective observational study that analysed the relationship between omeprazole and pantoprazole and voriconazole plasma concentrations in 78 patients treated with this antifungal. The results demonstrate the greater CYP inhibitory capacity of omeprazole vs pantoprazole regardless of CYP2C19 polymorphism as well as the lowest percentage of patients with subtherapeutic concentrations in the group of patients treated with omeprazole.

1. INTRODUCTION

Invasive fungal infections (IFI) are among the most feared complications of prolonged neutropenia patients with an increasing incidence in recent years. ¹ , ² Fungal infections caused by Candida spp. are the third cause of infection in critically ill patients while invasive aspergillosis (IA) is the most common life‐threatening invasive mould infection. ³

Voriconazole is an effective broad‐spectrum antifungal agent with potent activity against significant pathogens, including Aspergillus, fluconazole‐resistant Candida spp., Fusarium and Scedosporium. ⁴ It has nonlinear pharmacokinetics and narrow therapeutic range. Poor treatment outcome have been reported in patients with voriconazole trough plasma concentrations (Cmin) <1 mg/L ⁵ while Cmin >5.5 mg/L have been associated with adverse events such as vision disturbances, rash and hepatotoxicity. ⁵ , ⁶ , ⁷

Voriconazole have a significant interpatient variability depending on age, weight, liver function, CYP2C19 polymorphism and drug interactions. ⁸ One main source of the variability is hepatic metabolism. CYP2C19 contributes mainly in the hepatic conversion of voriconazole into the inactive metabolite voriconazole‐N‐oxide. Significant genetic polymorphism in the CYP2C19 gene encoding for CYP2C19 may result phenotypically in rapid or slow metabolism of voriconazole, ⁹ possibly resulting in approximately 30–50% variation of plasma concentrations. ¹⁰ Furthermore, cytochrome P450 is the main drug metabolising enzyme and is involved in the metabolism of many drugs. Therefore, voriconazole interacts with an exhaustive list of medications, many of which can significantly impact plasma concentrations.

Proton‐pump inhibitors (PPIs) such as omeprazole, pantoprazole or esomeprazole are drugs of particular interest widely used medication in the treatment of acid‐related gastrointestinal disorders. ¹¹ , ¹² Because these gastrointestinal problems are common in hospitalised patients, the likelihood of concomitant medication of PPIs and voriconazole is high in patients with IFI treated with voriconazole. PPIs undergo CYP‐mediated metabolism, mainly through CYP2C19, indicating the potential pharmacokinetic interaction with voriconazole.

Previous studies have demonstrated increased voriconazole exposure with omeprazole ¹³ due to an inhibition of CYP2C19; however, there have been few reports regarding the interaction of voriconazole with other commonly used PPIs, such as pantoprazole. In vitro studies have shown different CYP inhibitory ability according to the type of PPIs ¹² but the impact of this interaction in clinical practice remains unclear. Thus, the aim of this study is to analyse the relationship between different PPIs and voriconazole plasma concentrations to support the decision‐making regarding the selection of the best PPI in patients treated with voriconazole.

2. MATERIALS AND METHODS

2.1. Study design and population

A multicentre prospective observational study was conducted from July 2016 to September 2018 at 5 Spanish hospitals. Patients were included in the study if they had received systemic voriconazole for antifungal prophylaxis or for antifungal treatment. Patients received voriconazole doses according to clinical practice. Patients were excluded if aged <18 years and/or if the length of voriconazole treatment was <4 days.

Patient collected data included demographic information (age, sex, weigh and race), voriconazole treatment information (indication, dose, weight‐adjusted dose, route of administration, duration) and information regarding concomitant medication (strong and moderate inducers or inhibitors of CYP2C19, CYP3A4 or CYP2C9). The use of PPIs including the type of PPI, dose and route of administration and the CYP2C19 polymorphisms of each participant was also registered.

This study was approved by the Local Ethics Committee and written informed consent was obtained from each participant.

2.2. Measurement of voriconazole trough plasma concentration

Blood samples were drawn at steady state just prior to next dose of voriconazole (Cmin) and centrifuged at 13 800 × g at 4°C for 15 min. Plasma was kept frozen at –80° C until analysis. Voriconazole concentration was determined using a high‐performance liquid chromatography/photodiode array detection (Agilent 1260 series HPLC system; Agilent Technologies, USA, equipped with Diode Array Detector HS), a solvent delivery quaternary pump system, maximum pressure 400 bar and an autosampler with thermostat. The analytical method was fully validated according to the European Medicines Agency guidelines for bioanalytical method validation and met all required quality criteria.

2.3. CYP2C19 genotyping and CYP2C19 assignment

Blood samples were collected in EDTA‐Vacutainer collection tubes and were kept frozen at – 80° C until analysis. Genomic DNA from blood was isolated using a DNA Blood Mini Kit (Wizard Genomic DNA Purification kit, Promega). Genotyping was performed on the QuantStudio 12 K Flex System platform (Applied Biosystem) and polymorphisms were analysed by using real‐time polymerase chain reaction with TaqMan probes in the QuantStudio 12 K Flex (Thermofisher). According to nomenclature by the Clinical Pharmacogenetics Implementation Consortium, ¹⁴ patients with the *1/*17 genotype were classified as rapid metabolizers, and those with the *17/*17 genotype were classified as ultrarapid metabolizers. Patients with the 1 copy of a *2 or *3 allele (e.g. *1/*2,*1/*3, *2/*17) were assigned the intermediate metabolizers phenotype, and carriers of 2 copies (e.g. *2/*2) were assigned the poor metabolizer phenotype. The extensive metabolizer phenotype was assigned by default to patients without a *2, *3, or *17 allele.

2.4. Statistical methods

The demographic details were expressed as median and range. The ratios of voriconazole concentrations and doses were expressed as mean and standard deviation. Categorical variables are reported as frequencies and percentages.

Differences in voriconazole Cmin were analysed between patients without PPI, and patients who had received pantoprazole or omeprazole using the ANOVA and the Mann–Whitney U test. Patients treated with esomeprazole were excluded from the analysis because it was a very small group (n = 3). Simple and multiple logistic regression analyses were carried out to estimate the differences in voriconazole trough plasma concentrations with the PPIs after adjusting for other covariates such age, weigh, sex, route of voriconazole administration, concomitant administration of glucocorticoids and CYP2C19 phenotype. All statistical tests were 2‐tailed and a P‐value <.05 was deemed significant. The analysis was performed with SPSS (version 19.0).

3. RESULTS

3.1. Patient characteristics

A total of 78 patients were included in the study and divided in groups (patients without PPI, omeprazole, pantoprazole). Demographic information, CYP2C19 polymorphism and information regarding voriconazole treatment are included in Table 1.

TABLE 1.

Patient characteristic of 4 groups. Data were expressed as median and range in quantitative variables and frequencies and percentages in categorical variables

Variable		All (n = 78)	No PPI (n = 18)	OME (n = 32)	PAN (n = 25)
Age (y)		68 (19–93)	72 (19–93)	74 (32–91)	62(34–80)
Sex (M/F)		43/35	10/8	16/16	14/11
Weigh (kg)		68 (39–115)	72 (39–115)	71 (56–97)	64 (45–104)
CYP2C19 phenotype (n, %)	PM	1 (1.3%)	0 (0%)	1 (3.1%)	0 (0%)
	IM	20 (25.6%)	3 (16.7%)	9 (28.1%)	6 (24%)
	EM	34 (43.6%)	11(61.1%)	11 (34.4%)	12(48%)
	RM	21 (26.9%)	4 (22.2%)	10 (31.2%)	6 (24%)
	UM	2 (2.6%)	0 (0%)	1 (3.1%)	1 (4%)
Co‐medication (n, %)
Glucocorticoids		43 (55.1%)	8 (44.4%)	19 (59.4%)	14 (56.0%)
Enzyme inhibitor		17 (21.8%)	2 (11.1%)	8 (25.0%)	6 (24.0%)
Enzyme inducer		3 (3.8%)	1(5.5%)	0 (0%)	2(8.0%)
Frequency of voriconazole oral therapy (n, %)		35 (44.8%)	10 (55.5%)	12 (37.5%)	11 (44.0%)

PPI, proton‐pump inhibitor; OME, omeprazole; PAN, pantoprazole; PM, poor metabolizers; IM, intermediate metabolizers; EM, extensive metabolizers; RM, rapid metabolizers; UM, ultrarapid metabolizers.

3.2. Influence of PPIs on voriconazole trough plasma concentrations

In the total of the 78 patients included in the study, the mean voriconazole trough level was 2.22 ± 1.78 μg/mL with no significant differences between voriconazole trough oral and intravenous treatment (2.29 ± 1.99 vs 2.31 ± 1.92 μg/mL, respectively; P > .05). A considerable interindividual variability in voriconazole trough concentrations at steady state, which ranged from 0.06 to 14.47 μg/mL, was observed.

The Mann–Whitney U test for the comparisons of the trough concentrations of pairs groups yielded significant difference between the omeprazole and the pantoprazole group. Patients treated simultaneously with voriconazole and pantoprazole (n = 32) had significantly lower concentrations of voriconazole than those treated with omeprazole (n = 25; 1.44 ± 1.22 vs 2.67 ± 1.88 μg/mL, P = .013, Figure 1). This suggests a greater CYP2C19 enzyme inhibitory effect of omeprazole compared to pantoprazole. However, there were no significant differences in voriconazole trough level patients with and without PPIs (2.63 ± 2.13 vs 2.11 ± 1.73 μg/mL, P > .05).

FIGURE 1.

Box‐and‐whisker plots of voriconazole trough concentration in the 3 groups. The line inside the box, the lower and upper box ends, and the lower and upper whiskers represented the median value, the 25th and 75th percentiles, and the 5th and 95th percentiles of voriconazole trough concentration. Filled circles were outliers. Comparisons of the voriconazole trough concentration between the OME group and PAN groups were performed by the Mann–Whitney U test. PPI, proton‐pump inhibitor; OME, omeprazole; PAN, pantoprazole

Of the 78 global patients included in the study, 42 had Cmin out of therapeutic range (1.0–5.5 μg/mL), with the majority under the lower therapeutic limit: 46.1% were <1 μg/mL, 9.1% were >5.5 μg/mL and 44.8% were within the target range. When we compare the percentage subtherapeutic patients according to the PPI used, we observe a higher percentage of patients with subtherapeutic concentration in the pantoprazole group compared with the omeprazole group (44 vs 22%, Figure 2). Therefore, patients treated with voriconazole and pantoprazole had lower voriconazole trough concentrations and are more likely to suffer subtherapeutic voriconazole concentration than patients treated concomitantly with omeprazole.

FIGURE 2.

Effect of concomitant medication of PPIs on the percentage of patients obtaining therapeutic, subtherapeutic or supratherapeutic voriconazole trough concentration. PPI, proton‐pump inhibitor; OME, omeprazole; PAN, pantoprazole

Finally, with the exception of age, the association between PPIs and trough concentration was not affected by other covariates (sex, body weight, CYP2C19 polymorphism, other drug interactions, and route of administration of voriconazole; Table S1). Regarding the age, this study shows that increasing patient age was found to be a significant predictor of increased voriconazole concentrations. This finding is consistent with previous studies. ¹⁵

4. DISCUSSION

Several factors have been shown to affect plasma voriconazole concentrations, such as drug–drug interactions and polymorphisms of the gene encoding the CYP2C19 enzyme. ⁸ , ¹⁶ , ¹⁷ , ¹⁸ In patients with IFIs, the prescription of PPIs is common to prevent or treat acid‐related gastrointestinal disorders ¹⁹ and PPIs undergo CYP2C19‐dependent metabolism making these drugs competitive inhibitors of voriconazole. ²⁰ Although there are pharmacokinetic studies evaluating the effect of PPIs on voriconazole plasma concentrations, the net effect of PPIs on voriconazole pharmacokinetics as well as the differences according to the type of PPI has not been delineated.

In the present study the association between the use omeprazole or pantoprazole and plasma concentration of voriconazole was analysed. Unlike other studies, ¹¹ , ¹³ , ¹⁵ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ no significant differences were found in our study between patients without PPIs and patients treated with some kind of PPI. Curiously, voriconazole plasma concentration was lower in the group of patients treated with pantoprazole compared with patients with no PPI, but it was not statistically significant. This may be due to the low number of patients without any PPI (n = 18). In addition, this study did not include patients treated with other PPIs such as rabeprazole or lansoprazole, which is a limitation of the study.

However, the results showed that omeprazole produced a significative greater increase on voriconazole levels compared with pantoprazole regardless of CYP2C19 polymorphisms. This observation is consistent with the results of a retrospective study in 33 patients that revealed a statistically significant association between increased voriconazole concentrations and esomeprazole use but not with rabeprazole or pantoprazole use. ²¹ However, these authors did not include omeprazole, 1 of the most widely used PPI. In contrast, all PPIs were included in the in vitro and in vivo study of Yan et al. ²⁴ These authors revealed an increase on voriconazole concentrations in patients treated with lansoprazole, omeprazole or esomeprazole whereas there was no significant association with pantoprazole use. Finally, a multicentre study of voriconazole pharmacokinetics in 201 patients suggests a reduced interaction between voriconazole and pantoprazole compared with other PPIs, including omeprazole, although the effect was not statistically significant and this study did not include genetic information about CYP2C19. ¹⁵ These findings corroborated in vitro studies that showed that pantoprazole is the weakest inhibitor of CYP2C19. ²⁰ , ²⁶ The lower enzymatic inhibitory effect of pantoprazole and the reduced incidence of drug interactions caused by pantoprazole could be because the initial metabolite of pantoprazole undergoes a relatively nonsaturable phase II sulfate conjugation. ¹¹ , ²⁷

Achieving voriconazole target concentrations remains vital for the treatment of invasive fungal infections. However, the nonlinear pharmacokinetic of voriconazole and the large variability in voriconazole pharmacokinetics made it difficult. Corroborating the results of previous studies, ¹⁵ our study shows a high percentage of patients with subtherapeutic concentrations. This percentage was higher in the group of patients treated concomitantly with voriconazol and pantoprazole instead of omeprazole.

In conclusion, the results of our study corroborate the previous results in terms of lower enzymatic inhibitory effect of pantoprazole compared to omeprazole in the metabolism of voriconazole. The information obtained is of special interest when selecting the best PPIs in the usual clinical practice although it must be completed with information about the inhibitory effect of other PPIs such rabeprazole or lansoprazole.

Considering that invasive fungal infections pose significant morbidity and are often life threatening to many high‐risk patients, the selection of omeprazole could be useful in patients with subtherapeutic voriconazole plasma concentrations. Similarly, patients with voriconazole concentrations in the therapeutic range could most benefit from the concomitant treatment with pantoprazole to avoid the risk of drug toxicity.

CONTRIBUTORS

S.B.D, A.F.F concebid de idea of the work. S.B.D wrote the main paper. O.M. and A.L.P developed the pharmacogenetics analysis and revised the manuscript. M.T.R.J, A.L.V, A.L.M, B.B.G, D.B.M contributed to the recruitment of patients and supervised the work. G.B.C and M.L.P.M.B performed the microbiological analysis and A.B.H conducted the statistical analysis. I.Z.F , A.C, A.F.F and M.J.L supervised the project.

COMPETING INTERESTS

There are no competing interests to declare.

Supporting information

TABLE S1 Multivariate linear mixed‐effects regression analyses investigating influence of covariates on voriconazole exposure (Cmin)

Blanco Dorado S, Maroñas Amigo O, Latorre‐Pellicer A, et al. A multicentre prospective study evaluating the impact of proton‐pump inhibitors omeprazole and pantoprazole on voriconazole plasma concentrations. Br J Clin Pharmacol. 2020;86:1661–1666. 10.1111/bcp.14267

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.