FormalPara Key Points

Fluconazole pharmacokinetics is extensively studied in the neonatal population but requires more extensive research in children and adolescents. Voriconazole pharmacokinetics is extensively studied in children and adolescents and could benefit from more information in the critically ill neonatal and pediatric population despite its limited clinical use in these populations.

Isavuconazole, posaconazole, and itraconazole pharmacokinetics are studied to a limited extend in pediatric populations. To our opinion, specifically isavuconazole and posaconazole pharmacokinetics need to be investigated, as these drugs are frequently used in the hemato-oncology setting.

For all triazole agents, there is very limited knowledge on pharmacokinetics in critically ill patients who are likely to have altered pharmacokinetics. In addition, information on the impact of dialysis, extracorporeal membrane oxygenation as well as renal or hepatic impairment is lacking in most cases and should warrant further exploration.

1 Introduction

Immunocompromised pediatric patients are at high risk for invasive fungal disease (IFD). Although advances have been made in the management of IFD, the incidence and mortality rates are still high whereas treatment options remain limited and challenging. Triazoles represent the most important class of antifungal drugs for the prophylaxis and treatment of IFD. Within this class, isavuconazole, itraconazole, posaconazole, and voriconazole are recommended for managing invasive aspergillosis [1] and fluconazole and voriconazole are recommended for managing invasive candidiasis [2, 3].

Understanding the pharmacokinetics (PK) of these triazoles in pediatric patients is crucial to provide the most beneficial treatment. While the PK of triazoles is extensively studied in adult populations, knowledge on the PK of triazoles in pediatric patients is limited. Pediatric dose recommendations of triazoles have either been adjusted several times in the past years (i.e., voriconazole) or have been reported in the literature to a limited extent (i.e., isavuconazole, itraconazole, and posaconazole). This review provides an overview of current knowledge on the PK of the triazoles fluconazole, itraconazole, voriconazole, posaconazole, and isavuconazole in pediatric populations and summarizes factors that have to be taken into account to select the optimal dose.

2 Search Methodology

Relevant articles that describe the PK of triazoles in pediatric patients were searched until 26 November, 2020 using the databases PubMed and Embase. A detailed description of the literature search strategy is given in the Electronic Supplementary Material. Conference abstracts and unpublished data from conference proceedings were not included in this review.

The order of appearance of each triazole in this article is in the order of appearances of market introduction. This emphasizes the need for more prompt action to investigate the PK for the newest released drugs and to learn from pitfalls from the past. After providing a general introduction on pharmacology for all triazoles, a general introduction of each triazole will be given including indications and dose recommendations from the current labels and guidelines. Next, triazole absorption, distribution, metabolism, and elimination characteristics in adults will be described followed by relevant details on pediatric PK for both non-compartmental analyses (NCA) and population PK analyses.

3 Mechanism of Action: Pharmacology

All triazoles block the conversion of lanosterol to ergosterol through inhibition of the enzyme lanosterol 14α-demethylase (cytochrome P450 [CYP] 51). The depletion of ergosterol and accumulation of its toxic sterol precursors weaken the cell membrane structure and lead to cell membrane dysfunction [4,5,6,7,8]. Next to their fungal pharmacological target, triazoles are substrates and/or inhibitors of the human equivalent CYP enzyme system [4,5,6,7,8]. An overview of the metabolic routes and enzyme affinities of triazoles is provided in Table 1.

Table 1 An overview of the metabolic routes and enzyme inhibition of triazoles

4 Fluconazole

The US Food and Drug Administration (FDA) approval of fluconazole in adult patients was received in 1990 and fluconazole is licensed in individual European member states since 1988 [4, 9]. Fluconazole formulations include a solution for intravenous infusion and capsules, tablets, syrup, and powder for suspension for oral administration [9]. Currently, fluconazole is approved in pediatric patients aged 0–17 years for the treatment of mucosal candidiasis, for invasive candidiasis and cryptococcal meningitis, for prophylaxis and treatment of Candida infections in immunocompromised patients, and for prophylaxis (of relapse) and treatment of cryptococcal meningitis in high-risk patients [9, 10]. The fluconazole dosing recommendations in the European and American labels, the European Society of Clinical Microbiology and Infectious Diseases (ESCMID), and the Infectious Diseases Society of America guidelines are given in Table 2. The recommendations in the labels are different from the international guidelines, but also differ slightly between these international guidelines. Consensus between labels and guidelines is necessary to provide good clinical practice.

Table 2 Fluconazole dose recommendation in European and American labels and international guidelines

Fluconazole is characterized by a bioavailability (F) of 90% in adults, which makes intravenous and different oral formulations interchangeable. Absorption of fluconazole is not affected by food intake. The volume of distribution (Vd) of fluconazole is approximately 0.7 L/kg [4]. Fluconazole shows good penetration in a variety of body fluids and tissues, such as cerebrospinal fluid, sputum, saliva, urine, and skin [11]. The affinity of fluconazole for plasma proteins is low (10–12%). Fluconazole is minimally metabolized (∼ 10%) and the route of elimination is primarily (~ 80%) unchanged via renal excretion. Mean clearance (CL) of fluconazole is around 0.0138 L/h/kg in adults [4].

4.1 Non-Compartmental Analysis of Fluconazole PK in Pediatric Patients

Six studies described NCA of fluconazole PK in pediatric patients [12,13,14,15,16,17]. One study was performed in neonates [12] and five studies were performed in infants and children [13,14,15,16,17]. A detailed overview of the dosing regimens and fluconazole pharmacokinetic results is given in Table 3. The neonatal study included 12 premature neonates aged < 24 h after birth receiving fluconazole intravenously in a dose of 6 mg/kg with a dose interval of 72 h [12]. The five studies in preterm and term infants and children included patients with hematological or non-hematological malignancies, congenital disease, neoplastic disease, human immunodeficiency virus (HIV), or patients with and without peritoneal dialysis (PD) after open heart surgery with an age range of 2 weeks to 16 years [13,14,15,16,17]. Doses of fluconazole were 2–8 mg/kg per day administered either intravenously or as an oral suspension [13,14,15,16,17].

Table 3 Non-compartmental analyses of fluconazole

Although three out of these six studies included fluconazole as an oral formulation, none of them described the relative or absolute F of fluconazole [13, 15, 16]. During the first 2 weeks after birth, the Vd of fluconazole in premature neonates almost doubled and CL increased more than two times [12]. After 2 weeks of life, the Vd of premature neonates was found to be higher compared with children [12, 14, 15, 17]. After this period, the Vd decreased [14, 15, 17] and comparable values to adults were reported in children aged ≥ 12 years. [4, 15] These data suggest that premature neonates aged ≥ 2 weeks need adequate loading doses compared to premature neonates straight after birth and that children aged < 12 years need adequate loading doses compared to older children and adults. The higher Vd of fluconazole in premature neonates vs children and adults might be explained by the characteristics of fluconazole and body composition of neonates. Fluconazole is a hydrophilic compound, and neonates tend to have a higher water: fat ratio and as such a higher Vd [18]. The increasing fluconazole CL observed in neonates during the first 2 weeks of life might be explained by the maturation of the kidney function during this period [19]. Clearance of fluconazole in premature neonates seemed to reach the same range as children 2 weeks after birth [14, 17] but was still higher compared with adults [4]. A higher maintenance dose or shorter dosing intervals might be needed in premature neonates, infants, and children compared with adults. Contrary to these studies, one study in premature infants aged < 3 months reported comparable CL to adults, after a single dose of fluconazole [15]. Three studies described exposure of fluconazole after different dosing regimens and found a dose-proportional increase in exposure [15,16,17]. In patients with PD, no statistical differences in Vd and CL were reported compared to non-PD children with mild renal dysfunction. However, the elimination half-life of fluconazole was significantly longer in PD patients. This points towards the need for a lower maintenance dose or a longer dosing interval in this pediatric PD population [14]. To our knowledge, no other disease variables, such as HIV, have been found to alter the exposure of fluconazole [15,16,17].

4.2 Population Pharmacokinetic Analysis of Fluconazole in Pediatric Patients

Nine population pharmacokinetic studies were conducted that included either neonatal patients [20, 21], a mixed patient population of neonates and infants [22,23,24,25,26,27], or children and adolescents aged 3 days to 15.9 years [28]. One of these studies pooled data from three previously reported studies [26]. A detailed overview of the dosing regimens and fluconazole pharmacokinetic results is given in Table 4. The following patient groups were included in these studies: preterm and term patients at risk for IFD, patients with suspected or documented oral or invasive Candida infections, patients supported with extracorporeal membrane oxygenation (ECMO), or immunocompromised hemato-oncology patients. Eight studies described fluconazole PK in a one-compartment model [20,21,22,23,24,25,26,27], of which two studies included first-order absorption in the pharmacokinetic model [20, 21]. One study described fluconazole data best with a two-compartment model and first-order absorption [28]. The pharmacokinetic models and tested covariates are summarized in Table 5.

Table 4 Population pharmacokinetic estimates of fluconazole
Table 5 Pharmacokinetic models of fluconazole

Overall, population pharmacokinetic studies showed that the relative F from 90.9 to 100% [20, 21, 28] in neonates, infants, and children was excellent, and was comparable to a F of >90% in adults [4]. The rate of oral bioavailability (Ka) was from 0.538 to 3.76 h-1 [20, 21, 28]. It is difficult to compare values of Vd and CL between fluconazole population pharmacokinetic studies directly, as a variety of covariates were included on Vd and CL. Allometrically scaled bodyweight with fixed [20, 21, 23] and/or estimated [20] exponents was added on either Vd [20, 21, 23] and/or CL [20, 21, 23]. Age (inversely related) [27], ECMO [25], a coefficient for ECMO [26] and/or linearly scaled bodyweight [26, 28] were included as covariates on Vd. Covariates as linearly scaled bodyweight [26], body surface area [28], serum creatinine [24, 25], and exponents for estimated glomerular filtration (estimated) [20], serum creatinine [21, 23, 26], postmenstrual age (PMA) as a function of gestational age (GA) and postnatal age (PNA), [21] gestational age at birth (BGA) [23] and/or PNA [23], were included on CL. Serum creatinine was inversely related to CL [21, 23,24,25,26]. In one study, it was not clear if postmenstrual age was included as a covariate on fluconazole CL in the final model [22]. Another study reported that bodyweight influenced fluconazole CL but did not report the covariate equation [22]. Three studies used a linear regression analysis to test covariates [24, 25, 28]. One study concluded that fluconazole CL in premature neonates was low at birth and doubled within the first month after birth, but did not report on changes in fluconazole Vd [23]. This conclusion is slightly different from a previous NCA report, which reported a more than two-fold increase in CL during the first 2 weeks of life. Another study included both ECMO and non-ECMO patients and reported a significantly higher Vd but similar CL in pediatric ECMO patients compared with non-ECMO patients [26]. This higher Vd is likely due to the hydrophilic nature of fluconazole and the large circulating volume of ECMO procedures [29]. These population pharmacokinetic results point toward the need for an adequate loading dose of fluconazole in pediatric ECMO patients.

4.3 Physiologically Based PK of Fluconazole

Two studies have obtained interesting pharmacokinetic information with physiologically based pharmacokinetic models and assessed fluconazole dosing by predicting either cerebrospinal fluid exposure or the influence of ECMO [30, 31]. Data from plasma samples of 166 infants (< 750 g) with a median PNA of 21 days (range 3–93 days) and cerebrospinal fluid samples of 22 infants with a median PNA of 28 days (range 24–33 days) showed fluconazole exposure in the central nervous system, with a central nervous system-to-plasma ratio of ~ 1 [30]. In the second study, the edema disease state of ECMO patients was added to the model and the authors suggested that edema contributes to lower fluconazole exposure [31]

4.4 Summary of Findings and Recommendations

Pharmacokinetic data of fluconazole in neonates and infants are abundant, and pharmacokinetic data of fluconazole in children and adolescents are scarce. Research topics should include the F of all different oral fluconazole formulations and full pharmacokinetic investigations in children and adolescents. Special patient populations such as critically ill pediatric patients with renal impairment or other renal replacement therapy and solid organ transplant recipients should be further investigated. Additionally, the influence of the disease state of patients, such as excess fluid retention, on fluconazole PK might be interesting to further explore.

The relative F of fluconazole in pediatric patients is comparable to the F described in adults, which suggests that different formulations of fluconazole are interchangeable in pediatric patients. Most of these studies included the suspension as oral formulation, data on F of other oral formulations are very limited in pediatric patients.

Non-compartmental analyses report a higher Vd in preterm neonates compared with children and adults. These results suggest that adequate loading doses are needed. In preterm neonates, the fluconazole CL increases during the first 2 weeks after birth. The CL after 2 weeks of birth is comparable to CL in children but higher as compared to CL in adults. These results imply that higher maintenance doses or shorter dosing intervals are needed in preterm neonates and children. Non-compartmental analyses in pediatric PD patients report a significantly increased elimination half-life for fluconazole and these data suggest a lower maintenance dose or a longer dosing interval in this pediatric population.

Population PK studies report that allometrically scaled bodyweight and ECMO are significant covariates on Vd.. As a consequence, pediatric patients receiving ECMO might need higher loading doses. Allometrically scaled bodyweight, serum creatinine (inversely related), and either PMA (as a function of GA and PNA), or GA and PNA are significant covariates on CL. Dose adjustments based on serum creatinine, GA, and PNA might be taken into account to optimize fluconazole use. A standardized method to report both allometric scaling and maturation would be useful to compare pharmacokinetic results from different studies and populations.

Dose recommendations for fluconazole are inconsistent between the labels and the ESCMID and Infectious Diseases Society of America guidelines. As outlined previously by others [22], agreement between labels and international guidelines is necessary for clinical practice. Currently, there is no possibility to translate expert consensus from guidelines to an updated product information sheet. A reference in the summary of product characteristics to relevant guidelines would be an option to cover this. However, the legal background to make it possible for authorities and the pharmaceutical industry to request and update their product information will be tremendously challenging.

5 Itraconazole

Itraconazole was approved for adult patients in 1992 by the FDA [6] and itraconazole has been licensed in individual European member states. The oral capsules and oral solution are widely available in contrast to the intravenous formulations [32]. Itraconazole is not approved in pediatric patients aged < 18 years [6, 33]. However, the pediatric ESCMID-ECMM guideline for invasive aspergillosis and the pediatric ESCMID guideline for invasive candidiasis recommend a dose of 2.5 mg/kg twice daily of the oral solution for the purpose of mold and yeast active prophylaxis in children aged 2–18 years [1, 2]. For treatment of a proven or probable invasive aspergillosis, itraconazole is recommended in a loading dose of 5 mg/kg twice daily of the oral solution on day 1, followed by 2.5 mg/kg twice daily in patients aged 2–18 years [1].

In adults, itraconazole has a variable F with an absolute oral F of the oral solution of 55% [6]. The F of the oral solution is ~ 30% higher compared with the oral capsules [34]. Because of the variable F between formulations, these are not interchangeable. Food intake and pH fluctuation influence the itraconazole uptake, therefore the oral capsules are advised to be administered in a fed state and the oral solution in a fasted state [35]. The Vd of itraconazole is > 700 L [6]. Itraconazole penetrates into a variety of body tissues, including the lung, kidney, liver, bone, stomach, spleen, muscle, keratinous tissue, and skin but does not penetrate well into the cerebrospinal fluid [36,37,38]. Itraconazole has an active metabolite hydroxy-itraconazole with comparable in vitro activity to the parent compound. Both itraconazole (99.8%) and hydroxy-itraconazole (99.6%) are highly bound to plasma proteins. Itraconazole is mainly metabolized via CYP3A4 (Table 1). Renal elimination of both itraconazole and hydroxy-itraconazole is < 1%. The inactive metabolites of itraconazole are excreted in the urine (35%) and feces (54%). Mean CL of itraconazole in adults is 16.68 L/h [6].

5.1 Non-Compartmental Analysis of Itraconazole PK in Pediatric Patients

To our knowledge, there are no NCA reports of itraconazole PK described in neonates. Six studies performed NCA of itraconazole in infants, children, and adolescents aged 0.5–17 years at risk of mucosal fungal infection or IFD. A detailed overview of the dosing regimens and itraconazole pharmacokinetic results is given in Table 6. Patients with hematological and non-hematological malignancies, liver transplantation, respiratory tract infections, HIV, cystic fibrosis (CF), other infections/diseases, or undergoing hematopoietic stem cell transplantation (HSCT) were included in these studies. Itraconazole was administered in different oral and intravenous dosing regimens for prophylaxis and/or treatment. Dosages of itraconazole were from 2.5 to 5 mg/kg once or twice daily, with or without a loading dose of 5 mg/kg twice daily [39,40,41,42,43,44].

Table 6 Non-compartmental analyses of itraconazole

In five studies, itraconazole was administered as an oral solution [40,41,42,43,44], of which one study also included the intravenous formulation but the authors did not report the F of itraconazole [40]. Three studies stratified pharmacokinetic results of itraconazole by age [39, 42, 43]. A single dose of 2.5 mg/kg or multiple dosing regimens of 5 mg/kg once daily or 2.5 mg/kg twice daily have been investigated in patients aged 0.5–2 years, 2–5 years, and/or > 5 years [39, 42, 43]. Exposures differ widely between groups and studies. Both CL and Vd appear to change strongly within these groups. Interestingly, administration of a 2.5-mg/kg twice-daily regimen resulted in much higher itraconazole and hydroxy-itraconazole exposures compared with a 5-mg/kg once-daily regimen of itraconazole [42,43,44]. This is possibly owing to saturable absorption. One study in patients undergoing HSCT reported a considerably higher exposure compared with other studies, which is most likely explained by including a loading dose for itraconazole (5 mg/kg twice daily on day 1, followed by 5 mg/kg once daily) and pharmacokinetic sampling after the third administered dose [40]. Special pediatric populations, such as patients with HIV, showed comparable exposures of itraconazole and hydroxy-itraconazole to other populations, while patients with CF showed a considerably lower exposure after 2.5 mg/kg of itraconazole twice daily compared with other pediatric populations [41, 44]. Higher dosages than 2.5 mg/kg twice daily might be needed in pediatric patients with CF.

5.2 Population Pharmacokinetic Analysis of Itraconazole in Pediatric Patients

Two population pharmacokinetic studies in pediatric patients have been published [39, 45]. A detailed description of the dosing regimens and itraconazole pharmacokinetic results is given in Table 7. The pharmacokinetic models and covariates tested are summarized in Table 8.

Table 7 Population pharmacokinetic estimates of itraconazole
Table 8 Pharmacokinetic models of itraconazole

In 33 patients at risk for IFD aged 0.5–17 years, itraconazole was given intravenously as a single 2.5-mg/kg dose. Underlying diseases included CF, malignancies with febrile neutropenia, respiratory tract infections, or other diseases/infections. A three-compartment model best fitted the data for itraconazole. All parameter estimates were scaled to a total body weight of 30 kg [39], but the covariate equations were not reported.

In 49 patients with CF and undergoing bone marrow transplantation aged 0.4–30 years, including five adult patients, a median itraconazole dose of 5.4 mg/kg was given orally as capsules or solution. The vast majority of patients received itraconazole in a once-daily regimen. A one-compartment model was used with delayed absorption and included both itraconazole and hydroxy-itraconazole. The Ka for the solution and capsules was 0.96 h-1 and 0.09 h-1, respectively. The relative F of capsules was 0.55 compared to the solution. Clearance and Vd of itraconazole were allometrically scaled to a total body weight of 70 kg [45]. Values of exponents used for allometric scaling were not reported.

5.3 Summary of Findings and Recommendations

Pharmacokinetic studies of itraconazole are limited in pediatric patient populations and are lacking in neonates. Future research should focus on retrieving pharmacokinetic data in these patient populations and should address the F of the different itraconazole formulations.

The itraconazole oral solution is the preferred formulation, as the relative F was 45% higher compared with itraconazole capsules. Given the unknown absolute F and the difference in F of the oral formulations, dosing of itraconazole and switching between formulations should be accompanied by therapeutic drug monitoring. Furthermore, a twice-daily itraconazole regimen instead of a once-daily regimen is suggested to optimize itraconazole exposure.

Non-compartmental analyses suggest a great extent of variability across different age groups, attributable to both CL and Vd. Differences in studies preclude final conclusions and warrant further investigation. Pediatric patients with CF might need a higher itraconazole dose as a considerably lower exposure is reported compared with patients without CF.

Population pharmacokinetic studies included allometrically scaled bodyweight on itraconazole pharmacokinetic parameters. As itraconazole and hydroxy-itraconazole are highly bound to plasma protein, the unbound drug concentrations of itraconazole and hydroxy-itraconazole could be interesting variables for future research specifically in the critically ill population. Research in critically ill populations might be of interest in resource-poor countries where posaconazole and voriconazole may not be available.

Itraconazole is not approved for patients aged < 18 years in the labels, but international guidelines provide a dose recommendation for patients aged ≥ 2 years for both prophylaxis and treatment. Agreement between labels and guidelines is important for clinical practice and needs to be established.

6 Voriconazole

Voriconazole was both European Medicines Agency and FDA approved in 2002 for adult patients and has been available as oral tablets, oral suspension, and powder for concentrate for solution [5, 46]. The current approved indications for both adult and pediatric patients aged ≥ 2 years are treatment of invasive aspergillosis, candidemia in patients without neutropenia, esophageal candidiasis, infections caused by Scedosporium and Fusarium species [5, 46], fluconazole-resistant invasive Candida infections, and prophylaxis of IFD in high-risk allogenic HSCT [46]. The labels, the pediatric ESCMID-ECMM guideline for invasive aspergillosis, and the pediatric ESCMID invasive candidiasis guideline provide dose recommendations for pediatric patients aged ≥ 2 years. For prophylaxis and treatment of both invasive aspergillosis and candidiasis, a loading dose of 9 mg/kg twice daily on day 1, followed by 8 mg/kg twice daily intravenously or 9 mg/kg (maximum 350 mg) twice daily for the oral formulations in pediatric patients aged 2–11 years or aged 12–14 years (< 50 kg) is recommended. A loading dose of 6 mg/kg twice daily on day 1, followed by 4 mg/kg twice daily intravenously or 200 mg twice daily for the oral formulations is recommended in pediatric patients aged 12–14 years (≥ 50 kg) or aged ≥ 14–15 years [1, 2, 5, 46].

In adults, voriconazole is characterized by a F of 96% for both tablets and suspension [5], which makes it possible to switch between the two available formulations. As food intake can reduce voriconazole absorption, both oral formulations are advised to be administered in a fasted state [5, 47]. The Vd of voriconazole is around 4.6 L/kg. [5] The distribution of voriconazole is suggested to be extensive into different body tissues, including the cerebrospinal fluid [48] and aqueous and vitreous parts of the eye [49]. Voriconazole is bound to plasma proteins for around 58% [5]. Voriconazole is characterized by nonlinear pharmacokinetics in adult patients. The main CYP450 enzyme involved in the metabolism of voriconazole is CYP2C19 with also CYP2C9 and CYP3A4 playing a less prominent role (Table 1). Elimination via renal excretion accounts for only 2% in its unchanged form [5, 46].

6.1 Non-Compartmental Analysis of Voriconazole PK in Pediatric Patients

There are no NCA of voriconazole PK available in neonates and infants. Five NCA are available in pediatric patients aged 2–17 years. A detailed overview of the dosing regimens and voriconazole pharmacokinetic results is given in Table 9. Patients with hematological and non-hematological malignancies and patients undergoing BMT or HSCT were included in these studies. Voriconazole was administered either orally or in a combined intravenous to oral regimen. The oral voriconazole dose was from 4 to 9 mg/kg (maximum 350 mg) twice daily or was fixed at 200 or 300 mg twice daily. The intravenous voriconazole dose was from 4 to 8 mg/kg twice daily, either with or without a loading dose of 6 to 9 mg/kg twice daily [50,51,52,53,54].

Table 9 Non-compartmental analyses of voriconazole

Overall, only one study reported the F of voriconazole from 43.6 to 90.0% [52]. This F in pediatric patients was lower compared with the F of 96% seen in adults [5]. In the other studies, a lower F was hypothesized, as lower exposures were reported after oral administration compared with exposures after intravenous administration [50, 51, 54]. Unlike observations in adults where food intake reduces voriconazole absorption [5, 46], it remains unclear if the influence of food intake attributes to the variable F of voriconazole in pediatric patients. The reported lower F and subsequent lower exposure after oral administration imply that there is no bioequivalence between intravenous and oral formulations of voriconazole in pediatric patients. Two studies stratified pharmacokinetic results of voriconazole by age [52, 54]. One of these studies reported an overall comparable exposure of voriconazole in the group aged 2–5 years and aged 6–11 years after administration of 4, 6, or 8 mg/kg of voriconazole in a twice-daily intravenous to oral regimen. This study also reported a ~ 2.5 times increased exposure after increasing voriconazole from 4 to 8 mg/kg, suggesting non-linear PK in these pediatric patients over a dose range of 4–8 mg/kg [52]. The other study administered voriconazole according to the current labels and guidelines. For a detailed description of the dosing strategies, see Table 9. This study reported that patients aged 12–14 years (< 50 kg) had a higher exposure compared with patients aged 2–11 years and that patients aged 12–14 years (≥ 50 kg) had a lower exposure compared with patients aged < 15 years (< 50 kg) [54]. The sample sizes in the different age groups were small and the authors mentioned that the CYP2C19 genotype in their Asian population might also have played a role in the differences in voriconazole PK [54]. Two studies showed an overall higher exposure of voriconazole compared with the other studies [53, 54]. This higher exposure might be explained by the higher dosing regimens used.

6.2 Population Pharmacokinetic Analysis of Voriconazole in Pediatric Patients

There are no population pharmacokinetic analyses of voriconazole available in neonates. One study included infants, but did not describe the pharmacokinetic results for this population separately [55]. In total, nine studies were performed in pediatric patients aged 0.8–21 years [55,56,57,58,59,60,61,62,63], of which two studies pooled data of three earlier published studies [57, 62] and included data of healthy adult patients [57]. A detailed overview of the dosing regimens and voriconazole pharmacokinetic results is given in Table 10. These studies included immunocompromised patients with hematological or non-hematological diseases, immunodeficiency or autoimmune diseases, liver transplantation, CF, other infections/diseases or undergoing HSCT or BMT [55,56,57,58,59,60,61,62,63]. Voriconazole was administered either intravenously [55, 61, 63], orally [55], or in a combined intravenous to oral regimen [56,57,58,59,60, 62]. All studies reported PK of voriconazole in a two-compartment model [55,56,57,58,59,60,61,62] and one study included also one compartment for the metabolite of voriconazole [63]. The models included delayed absorption [55, 57, 59] and first-order absorption [55,56,57,58,59,60, 62] and either linear [61], nonlinear [55, 56, 58, 60, 62], or mixed linear and nonlinear elimination [57, 59]. In one study, voriconazole elimination was included as linear CL but in addition also as non-linear CL to its metabolite [63]. Two other studies included both concentration- and time-dependent voriconazole elimination [57, 59]. The PK models and covariates tested are summarized in Table 11.

Table 10 Population pharmacokinetic estimates of voriconazole
Table 11 Pharmacokinetic models of voriconazole

Seven studies in pediatric patients administered either an oral solution or tablets of voriconazole in which the F was from 44.6 to 85% [55,56,57,58,59,60, 62]. The F found in these studies was also lower compared with the F of 96% reported in adults [5]. Similar to findings in the NCA, it remains unclear if the influence of food was attributed to this difference. The Ka had a range of 0.43–1.53 h-1 [55,56,57,58,59,60, 62]. Allometrically scaled bodyweight with fixed exponents [56,57,58,59,60, 63] was added on either CL [57, 59, 63], Vd [57,58,59,60, 63], and/or maximum rate of enzyme activity [56,57,58,59,60, 63]. Two studies included patients aged < 2 years [55, 63], of which one study had sufficient information to include a maturation factor to the pharmacokinetic model [63]. Two other studies incorporated the CYP2C19 genotype [61, 62], alanine aminotransferase (ALT) [61, 62], and alkaline phosphatase on CL. In these studies, the CYP2C19 genotype in the combined group of heterozygous extensive/poor CYP2C19 metabolizers [61, 62], ALT [61, 62], and alkaline phosphatase [61] significantly decreased CL, but according to the authors these variables were not predictive for voriconazole CL [61, 62]. Other covariates included linearly scaled weight and age on CL and Vd [55].

6.3 Physiologically Based PK of Voriconazole

One physiologically based pharmacokinetic model was developed for voriconazole in children. The physiologically based pharmacokinetic-derived values from the initial oral model showed an overprediction for F, area under the curve (AUC), and maximum serum concentration in children, which decreased substantially after adding intestinal CL to the model. Intestinal first-pass metabolism might explain the lower bioavailability of voriconazole in children compared with adults [64].

6.4 Summary of Findings and Recommendations

The PK of voriconazole in neonates and infants and children aged < 2 years is lacking, and future studies should take these patient populations into account. Future research should further focus on the highly variable F, differences in F between the oral formulations, the linear or non-linear relationship of voriconazole elimination, and PK in critically ill pediatric patients.

None of the reports highlight the difference in F of the oral solution and tablets. In contrast to adults, it seems that there is no bioequivalency between oral and intravenous formulations in pediatric patients. It is unclear if the intake of food or gastric-emptying time is (partly) responsible for this variability and/or if the influence of intestinal first-pass metabolism might play a role. These questions need to be further explored. Switching from intravenous voriconazole to oral formulations cannot be done as straightforwardly as in adults but should be accompanied by therapeutic drug monitoring.

Noncompartmental analyses report that patients aged < 12 years seem to have a higher CL and Vd compared with patients aged ≥ 12 years and therefore the recommended loading dose and maintenance doses of voriconazole is higher in patients aged 2–11 years compared with those above 12 years. Some population pharmacokinetic studies reported that the CYP2C19 genotype and ALT values were significant covariates on voriconazole CL, but were not predictive for voriconazole CL. Although CYPC19 might be correlated with voriconazole CL, upfront dose adjustments in clinical practice are not yet advised in populations with a low prevalence of homozygous allele variations. Further research is needed to explain the differences of voriconazole PK in pediatric patients, to explore the influence of CYP2C19, and to reflect on the role of ALT as a surrogate marker for liver function. Additionally, other possible elimination routes (i.e., flavin-containing monooxygenase 3 [65]) might be interesting topics to explore.

7 Posaconazole

In 2005, posaconazole received European Medicines Agency marketing authorization and in 2006 FDA approval for adult patients [8, 66]. The currently available formulations include a concentrate for solution for infusion, an oral suspension, and gastro-resistant tablets [66]. The FDA approved posaconazole in pediatric patients aged > 13 years for prophylaxis and treatment of invasive aspergillosis and invasive candidiasis [8], but in Europe posaconazole is not approved in pediatric patients aged < 18 years [66]. Both the new solid oral tablet and the intravenous solution of posaconazole require a loading dose of double the maintenance dose, whereas this loading dose is not of value for the marketed oral suspension. In the pediatric ESCMID-ECMM guideline for invasive aspergillosis, the recommended dose for posaconazole prophylaxis for patients aged ≥ 13 years is 300 mg once daily of the gastro-resistant tablet or a dose of 200 mg three times daily of the marketed oral suspension. For salvage therapy of a proven/probable invasive aspergillosis for patients aged ≥ 13 years, 300 mg once daily of the gastro-resistant tablet or intravenous formulation or a dose of either 400 mg twice daily or 200 mg four times daily of the marketed oral suspension is recommended [1]. The posaconazole dosing in the setting of prophylaxis for invasive candidiasis is identical to the dosing regimen of the marketed oral suspension for prophylaxis of invasive aspergillosis [2]. All the above-mentioned guidelines recommend using the gastro-resistant tablet over the marketed oral solution because of the anticipated more favorable oral bioavailability of the gastro-resistant tablet.

The F of posaconazole is only reported for adult patients receiving the gastro-resistant tablets and is around 54% [8]. As the F of the marketed oral suspension is not available in the public domain, bioequivalence between the formulations cannot be assured. Both the marketed oral suspension and gastro-resistant tablets show saturable absorption, but for the gastro-resistant tablets this was only seen for daily doses above 800 mg of posaconazole [67, 68]. Absorption of the marketed posaconazole suspension is significantly influenced by food intake and administration in a fed state is advised [69]. The gastro-resistant tablets are less prone to food effects [66], but a fed state can still increase the absorption by ∼ 1.5 times [70]. The tablet cannot be broken because of the gastro-resistant coating, which makes it difficult to administer these tablets to patients who are unable to swallow. The mean apparent Vd (Vd/F) of posaconazole is 287 L for the gastro-resistant tablet and the Vd/F is around 1774 L for the marketed oral suspension [8]. Posaconazole penetrates into a variety of tissues, including the lung, heart, kidney, and liver, but penetrates poorly into brain tissue [71] and cerebrospinal fluid [72]. Posaconazole is bound to plasma proteins for > 98% [8]. In contrast to the other azoles, posaconazole is metabolized via uridine diphosphate glucuronosyltransferase enzymes, and particularly uridine diphosphate glucuronosyltransferase 1A4 (Table 1) [73]. About 77% of radioactive-labeled posaconazole was retrieved in the feces of which 66% was the parent compound. The formed metabolites that were excreted in the urine and feces accounted for about 17% of the radioactive-labeled posaconazole [8, 66]. Mean CL is 7.3 L/h [8].

7.1 Non-Compartmental Analysis of Posaconazole PK in Pediatric Patients

Currently, there are no NCA studies of posaconazole PK performed in neonates. A detailed overview of the dosing regimens and posaconazole PK results is given in Table 12. Three NCA were performed in immunocompromised patients aged 3 months to < 18 years. [74,75,76] Patients with hematological and non-hematological malignancies or undergoing HSCT were included in these studies. In two studies, posaconazole was only administered as the marketed oral suspension. The relative F of posaconazole was not determined in these studies [74, 75]. In the other study, posaconazole was administered as a not yet marketed new formulation, a powder for oral suspension (PFS), as well as an intravenous solution [76]. The first NCA investigated posaconazole orally as the marketed suspension at 6 or 9 mg/kg in a two or three times-daily regimen in three different age groups [74]. The second study used the marketed oral posaconazole suspension as 120 mg/m2 based on body surface area (BSA) [75]. In the third study, posaconazole was investigated as either an intravenous solution or as the new oral PFS at 3.5 mg/kg, 4.5 mg/kg, or 6 mg/kg in a twice-daily regimen on day 1, followed by the same dose in a once-daily regimen in two different age groups [76].

Table 12 Non-compartmental analyses of posaconazole

Increasing the daily dose from 6 to 9 mg/kg or increasing the dosing frequency of the marketed suspension from two times daily to three times daily did not increase the exposure of posaconazole. This suggests saturable absorption in pediatric patients, which is also seen in adults. The authors suggested that children aged >7 years showed higher exposures compared with patients aged 2–7 years [74], implying that higher dosages are needed in younger patients to achieve a comparable exposure to older patients. A dosing regimen based on BSA resulted in a comparable mean exposure as children aged 7–17 years on a 6-mg/kg twice-daily regimen [75]. However, data based on BSA were not available for different age groups and exposure in the youngest patients is therefore not exactly known with this approach. Administering posaconazole intravenously or as a PFS in a once-daily regimen (with a loading dose on day 1) resulted in higher exposures compared with the exposures after a twice-daily regimen of the marketed oral suspension in the previously described report [74, 76]. Similarly to this earlier report, posaconazole exposure was lower in younger patients compared with older patients in all dosing groups [74, 76]. Furthermore, the exposure after oral PFS administration was lower compared with intravenously administered posaconazole. As suggested by the authors, there seems to be no bioequivalence between the intravenous and new PFS formulations in pediatric patients [76].

7.2 Population Pharmacokinetic Analysis of Posaconazole in Pediatric Patients

Currently, there are no population pharmacokinetic studies of posaconazole performed in neonates. One population pharmacokinetic model was published in 117 immunocompromised infants, children, and adolescents aged 0.5–18 years. A detailed overview of the dosing regimens and posaconazole pharmacokinetic results is shown in Table 13. Posaconazole was administered as the marketed suspension in the vast majority of these patients, with a mean daily dose of 13.11 mg/kg [77]. A one-compartment model fitted the data best. An overview of the pharmacokinetic model and covariates tested is given in Table 10. Allometrically scaled bodyweight was added on CL and Vd and covariates such as diarrhea and concomitant use of proton pump inhibitors decreased posaconazole bioavailability only after administration of the marketed suspension [77]. The pharmacokinetic models and covariates tested are summarized in Table 14.

Table 13 Population pharmacokinetic estimates of posaconazole
Table 14 Pharmacokinetic models of posaconazole

The relative Ka of the marketed suspension and tablets was 0.197 h-1 and 0.588 h-1, respectively. The relative F of the marketed suspension and tablets was not described. A decrease of 33% in the relative F of the marketed suspension was seen in patients with diarrhea and a 42% decrease in patients using proton pump inhibitors. As only the oral marketed formulations were used, Vd/F and apparent CL were determined. Allometrically scaled bodyweight normalized to 70 kg was added as covariate on posaconazole Vd/F and apparent CL [77].

7.3 Summary of Findings and Recommendations

Pediatric pharmacokinetic data of posaconazole are very limited, and future research is particularly needed to explain the PK of posaconazole in infants, and to further resolve its PK in children and adolescents. Research topics should include the F of all the oral formulations and the PK in critically ill patients and patients with CF. Furthermore, the drug–drug interaction between posaconazole and CF transmembrane conductance regulator modulators might be an interesting research topic. In adults, the gastro-resistant tablets are the preferred formulation, but there are no pharmacokinetic data of this formulation available in pediatric patients. This oral tablet formulation urgently needs to be studied in children and adolescents to confirm that this is the most appropriate oral pharmaceutical formulation to be used. For patients who are unable to swallow tablets, the new PFS needs to be further explored. Other new child-friendly formulations allowing the administration of smaller dosages might be needed to further expand posaconazole treatment.

Although all studies administered posaconazole as an oral formulation, the absolute and/or relative F were not described and need to be explored in pediatric patients. Exposures after administration of the not yet marketed posaconazole PFS were lower compared with intravenous administration, and suggests that there is no bioequivalence between these two formulations. Given the unknown F of the marketed formulations and the non-bioequivalence between intravenous and PFS formulations, dosing of posaconazole and switching between formulations should be accompanied by therapeutic drug monitoring.

The majority of available pediatric NCA only administered the suspension of posaconazole as an oral formulation. These data confirm adult observations that the marketed suspension shows saturable absorption. The new posaconazole PFS that is not yet on the market shows higher exposures in a once-daily regimen compared with the twice-daily regimen of the current marketed posaconazole suspension. After administration of both oral and intravenous formulations, posaconazole exposure seems lower in younger patients and higher dosages might be needed to reach the same exposure as older patients.

The population PK study included allometrically scaled bodyweight on CL and Vd. Diarrhea and concomitant use of proton pump inhibitors were negatively associated with the relative F of the marketed posaconazole solution. Because of the high protein binding of posaconazole, it might be interesting to explore the influence of its unbound drug concentrations on posaconazole PK.

8 Isavuconazole

The relatively new triazole isavuconazole is not licensed for pediatric patients. The European Medicines Agency approved isavuconazole for adult patients in 2014 and the FDA approved isavuconazole in 2016 [7, 78]. Available formulations include an oral formulation as hard capsules and an intravenous formulation as powder for concentrate for solution. In adult patients, isavuconazole is indicated for the treatment of invasive aspergillosis. In addition, it is licensed for mucormycosis for patients who have a contraindication or intolerance for amphotericin B [7, 78]. Isavuconazole has not yet been approved for pediatric patients and the international guideline does not provide recommendations for dosing of isavuconazole in pediatric patients [1]. Dose finding trials have been completed or are ongoing, thus more information is expected soon.

Isavuconazole is given as a pro-drug isavuconazonium sulfate. The oral F of isavuconazonium sulfate is 98% in adults [7]. After a rapid and complete absorption, isavuconazonium sulfate is quickly and completely cleaved to isavuconazole [7]. Oral and intravenous formulations can be used interchangeably. Food intake or fluctuations in pH do not influence the absorption of isavuconazole [79]. Based mostly on animal research, isavuconazole widely distributes in different tissues, including the liver, lungs, eyes, kidneys, skin, bone, nasal mucosa, and brain [80]. Isavuconazole is bound to plasma proteins for >99% and is metabolized by CYP3A4/A5 and uridine diphosphate glucuronosyltransferase (Table 1) [7].

To our current knowledge, there is only one pediatric study of isavuconazole available in the public domain outside of conference abstracts and case reports. This retrospective study included 29 patients with a hematological malignancy aged 3–18 years. In six patients, an 8-point sample curve was obtained over 12 h. The demographics and dosing regimens are not reported for these six patients separately. The median AUC0–12h (range) in these six patients was 153.16 mg × h/L (86.31–169.45) [81]. Because of the small sample size and missing demographics and dosing information, it is difficult to draw any conclusions from these data.

8.1 Summary of Findings and Recommendations

Data on the PK of isavuconazole are urgently needed in pediatric patients including population pediatric PK data. Specifically for pediatric patients, information on F including information on dosing via a nasogastric tube are needed as well as information on bioequivalence after the intake of whole or opened capsules. As isavuconazole is highly protein bound, more research is needed on unbound drug concentrations in, for instance, the critically ill patient populations.

9 Conclusions

This review shows that the PK of fluconazole is extensively studied in the neonatal population and the PK of voriconazole is extensively studied in children and adolescents. Isavuconazole, itraconazole, and posaconazole are studied to a limited extent. Fluconazole data in children and adolescents are understated, while for other triazoles pharmacokinetic data in neonates and infants urgently need to be studied. Future studies should explore the PK of the newest triazole agents, understanding the F of the available formulations and learning more about interactions with food or administration over a nasogastric tube, the effect of CYP genotypes and other metabolic routes, the influence of other factors such as unbound drug concentrations for highly protein-bound agents, and the development and PK of new oral formulations that can easily be deployed in pediatric patients. In addition, information on the PK of triazoles in critically ill patient populations, the impact of dialysis, ECMO as well as renal or hepatic impairment is lacking in most cases and should warrant further exploration. Better understanding of the PK is necessary for optimal clinical care and remaining knowledge gaps will need to be clarified.