Voriconazole is a second-generation triazole antifungal agent with enhanced in vitro activity against various fungi. It is considered to be a first line primary therapy for invasive aspergillosis, and is equivalent to a regimen of amphotericin B followed by fluconazole for the treatment of candidaemia. However, since clinical studies with voriconazole have included only adults (>11–18 years old), experience in paediatric patients has been limited to case reports or series of cases. Accordingly, the optimal dose of voriconazole in children has not been fully established, but it is believed to be greater than adult dosing. The purpose of this paper is to report the cases of five

METHODS

Five children treated in different hospitals in the United Kingdom were studied. All patients received voriconazole as a loading dosage (range, 5.2–6.0 mg/kg intravenously in two dosages) and 3.4–6 mg/kg intravenously twice daily thereafter.

Voriconazole plasma levels were determined at the Regional Mycology Laboratory, Salford. Before 2002, samples were processed using a bioassay. In recent years, voriconazole has been measured by liquid chromatography–tandem mass spectrometry. This validated method is sensitive and faster than the bioassay or high-pressure liquid chromatagraphy-UV methods.1 Targeted trough voriconazole levels were 0.25–6.0 μg/ml.2 The lower limit of quantification was 0.1 μg/ml. Concentrations below that limit were recorded as zero.

RESULTS

Table 1 summarises the clinical findings from the five paediatric patients studied (three patients with leukaemia and two patients with burns). Voriconazole levels and dose adjustments for these patients are presented in fig 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Voriconazole levels for individual patients included in the study. Voriconazole trough levels are represented by squares. Stars represent peak levels and circles random levels. The boxes’ height indicate voriconazole total daily dosage, the first box showing the loading dose. Frequency of voriconazole administration is also indicated.

DISCUSSION

Voriconazole is known to have variable pharmacokinetics. In adults, steady-state plasma levels after 3–6 mg/kg twice daily intravaneous infusion range from 3–6 μg/ml. Steady-state concentrations are achieved after 5–6 days (shorter if a loading dose is given). After 200 mg twice daily orally in adults, steady-state plasma concentrations generally range from 2–3 μg/ml.3 However, in a study of voriconazole for the treatment of invasive aspergillosis, patients on standard doses demonstrated levels in plasma ranging from <0.1 μg/ml to as high as 9.7 μg/ml.2 Studies in adults revealed nonlinear pharmacokinetics, which are thought to be related to saturation of metabolism, resulting in a ∼threefold increase in the area under the curve (AUC_T) following a 33% increase in dosage.3 Thus, there is substantial inter-subject variability in the plasma concentrations achieved, especially in patients suffering from drug–drug interactions. CYP2C19, the major isozyme implicated in the metabolism of voriconazole, exhibits greater genetic polymorphism. As a result of a point mutation in the respective gene, some people are poor metabolisers, while some others are rapid metabolisers. Up to 7% of whites and ∼20% of non-Indian Asians have a deficiency in the expression of this enzyme. The observed interpatient variability in voriconazole levels is ∼100-fold,2 and we are not aware of any other anti-infective agent with a similarly profile.

Unlike the findings in adults, studies have revealed that voriconazole undergoes linear pharmacokinetics in children. Paediatric patients have a higher capacity for elimination of voriconazole per kg of body weight than do adult healthy volunteers, and dosages of 4 mg/kg may be required in children to achieve exposures consistent with those in adults following dosages of 3 mg/kg.4 Preliminary evidence suggests that the mean AUC_T associated with a dose of 8 mg/kg taken intravenously in children approaches that seen with 4 mg/kg in adults.5 In Europe, recommended dosage of intravenous voriconazole in 2–11-year-old patients is 7 mg/kg every 12 h, and no loading dose is recommended. Recommended oral maintenance dosage for these paediatric patients is 200 mg every 12 h. Because of the assumed limited gastrointestinal transit time, oral suspension should be preferred to the tablet formulation. There are no formal recommendations so far for the use of voriconazole in children in the United States.6

Given the relative unpredictability of voriconazole levels, measurement of plasma concentrations to identify extreme levels is warranted. In a previous study, patients with invasive aspergillosis who failed to demonstrate random levels >0.25 μg/ml had a higher probability of treatment failure.2 These findings were also confirmed in a recent study, which showed a significant relationship between disease progression and drug concentration (p<0.025).7 A positive clinical response was observed in 100% (10/10) of patients with random voriconazole concentrations of above 2.05 μg/ml, whereas the disease progressed (and patients died) in eight out of 18 patients with concentrations below that level.8 It may be that these are useful threshold trough concentrations. However, more data are required on this point, particularly in challenging situations such as cerebral infections. In addition, patients with plasma concentrations >6 μg/ml may be at risk for concentration-related toxicities,2 including neurological adverse events.7

Only a few studies have reported the use of voriconazole levels to monitor therapy in children.9^–11 In one of the studies,9 the introduction of phenobarbitone promoted a decrease in voriconazole peak concentration by ∼50%. The premature patient described by Muldrew et al also required elevated voriconazole dosages (6 mg/kg every 8 h) owing to concomitant therapy with phenobarbitone.10 In the report by Destino and colleagues, dosages as high as 13.4 mg/kg every 12 h were required to obtain trough concentrations of 0.2 μg/ml.11 Similarly, Walsh et al.12 showed that median plasma concentrations in children receiving dosages of ⩾4 mg/kg per 12 h (1.6 μg/ml) were lower than those of adult volunteers receiving 4 and 5 mg/kg per 12 h (5.7 μg/ml and 7.4 μg/ml, respectively). Thus clearance of voriconazole is greater in children than in adults. Eiden et al reported that torsades de pointes occurred in a 14-year-old girl with high plasma trough levels (7 μg/ml).13 The patient was also being treatment with omeprazole, which is known to increase voriconazole C_max and AUC_T by 15% and 41%, respectively. Torsades de pointes had previously been described in a 15-year-old patient with normal voriconazole levels, suggesting that this rare adverse effect does not solely depend on voriconazole drug concentrations.14 In both reported cases, genotyping of CYP2C19 (and 2C9) revealed no mutations, indicating that these patients were standard metabolisers.

In this case series we demonstrate how unpredictable voriconazole levels can be in children. For instance, patient 1 had levels frequently below the therapeutic target despite the use of voriconazole dosages as high as 8.1 mg/kg three times a day (a dosage much higher than that recommended for children). Patient 1 was also receiving vincristine — apparently, voriconazole levels are not affected by the use of vinca alkaloids, although vincristine may be increased to toxic concentrations.6 Patients 2 and 5 received lower than currently recommended dosages of voriconazole — curiously, a marked increase in voriconazole trough levels was observed for these two patients after a small change in dosage. The occurrence of repeated low levels in patients 3 shows how much drug is required in some children to reach detectable concentrations. Burn injury may lead to accelerated azole clearance as demonstrated for fluconazole.15 It remains to be elucidated whether higher voriconazole dosage is required for burn patients. Increased voriconazole clearance probably affected patient 4, also a patient with burns, who had low plasma levels despite the use of the recommended voriconazole dosage (7 mg/kg twice daily). This situation was probably aggravated by a drug interaction with phenobarbitone, which also accelerates clearance. Although stopped 1 week before voriconazole was started, this long-acting barbiturate is known to be a potent CYP450 inducer. Renal failure may have prolonged phenobarbitone’s half-life in this patient. As could have been anticipated, voriconazole trough concentrations markedly increased over the following days despite no change in dosage, which is probably related to phenobarbitone induction wearing off, with a reduction in metabolism of voriconazole. Voriconazole has been shown to inhibit midazolam metabolism in vitro, apparently with no effect on voriconazole levels.6

In conclusion, while the optimal dose of voriconazole in children is unknown, monitoring plasma levels may be helpful to at least ascertain a minimal level of exposure, especially in patients concomitantly receiving other drugs that may interfere with voriconazole metabolism. Although trough levels are preferred, random/peak levels can also provide useful information if levels are extremes. Unfortunately, the limited number of patients included in this series does not allow us to define target voriconazole concentrations for children. Furthermore, it should be noted that no study has defined voriconazole exposure or levels that can predict clinical success or failure in children. As discussed in this article, to achieve measurable plasma concentrations similar to those in adults, paediatric patients may require higher doses of voriconazole or the administration of the drug at shorter intervals. Owing to the huge variability observed in patients’ levels, the use of fixed doses of voriconazole is probably incorrect for both adults and children. Guidelines about dose modification for patients with very low or high plasma concentrations also need to be developed and validated.