Higher Weight-Based Doses Are Required to Achieve and Maintain Therapeutic Voriconazole Serum Trough Concentrations in Children

Voriconazole, an extended-spectrum triazole antifungal with activity against a variety of clinically important yeasts and molds, is approved by the US Food and Drug Administration for use in adults and children aged ≥2 years.^1,2  Therapeutic drug monitoring (TDM) is recommended, with target serum trough concentrations variable among resources. Targets greater than 1 to 2 mcg/mL are generally recommended to optimize efficacy, while a maximum trough of 6 mcg/mL is typically recommended to limit toxicity.^3–6  Pharmacokinetic data have demonstrated that children require weight-based dosing proportionately larger than adults to achieve therapeutic serum trough concentrations, largely owing to non-linear saturable pharmacokinetics in children, as well as age-related changes in clearance mechanisms.^7–13  Recommendations for initial starting doses in children are published, but no formal guidelines exist to guide timing and frequency of TDM or dose modifications needed to achieve and maintain therapeutic drug concentrations.^11,14  We performed a quality improvement project to identify the initial voriconazole dose used in different age groups, the proportion of patients achieving target serum concentrations with the initial dose, and subsequent TDM and dose modifications needed to maximize the proportion of children achieving and maintaining therapeutic voriconazole serum concentrations. We hypothesized that currently recommended doses do not achieve target drug concentrations in most children and multiple TDM-guided dose modifications are necessary to achieve and maintain therapeutic voriconazole concentrations.

We retrospectively reviewed records of patients who were treated with voriconazole at Children's Wisconsin from November 1, 2012, to August 1, 2019. We included patients aged <18 years who received voriconazole for at least 48 hours. Patients prescribed voriconazole for prophylaxis were excluded. Patients were identified by using the electronic medical record. Institutional voriconazole clinical practice guidelines changed over time, based on local data, though generally followed the American Academy of Pediatrics (AAP) dose recommendation of 18 mg/kg/day divided twice daily for 2 doses (load) then 16 mg/kg/day divided twice daily (maintenance) for children aged <12 years, and 12 mg/kg/day divided twice daily for 2 doses (load) then 8 mg/kg/day divided twice daily (maintenance) for patients aged ≥12 years up to standard adult maximum doses.^11,15  Most modifications in the guideline were related to monitoring parameters. Though our guideline did not specify whether doses should be given intravenously (IV) or enterally, guidance was included to consider IV therapy in children aged <12 years owing to pharmacokinetic variability in this population. Directions on the medication administration record within the electronic health record recommend to hold tube feeds and/or give 1 hour before or after a meal. Voriconazole serum trough concentrations were drawn at steady state after initiation of therapy or a change in dose, every 1 to 2 weeks after achieving the target concentration, or at any time toxicities were suspected. Individual patient dosing (including decisions regarding receipt of a loading dose and dose modifications) and monitoring regimens were determined by the treating physician and a pharmacist with expertise in pharmacokinetics.¹⁵ 

Demographic data, voriconazole doses, TDM values and timing, and patient-reported drug toxicities were collected. Invasive fungal infection category (proven, probable, possible, and other) was classified by using definitions from the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group.¹⁶  Voriconazole serum trough concentrations drawn ≥2.5 days following drug initiation or dosage change were accepted as “steady state” measurements. We allowed this timeframe because voriconazole serum concentrations are a “send out” test at our institution, and this allowed for flexibility of scheduling around weekends and holidays. Steady-state voriconazole serum trough concentrations of 1 to 6 mcg/mL were designated “therapeutic” (within target range). The dose associated with each TDM measurement was defined as the majority of doses received prior to that TDM. For patients with more than 1 TDM measurement, this would be the majority of doses administered after the preceding TDM measurement. Patient-reported drug toxicities were defined as visual disturbances, audio or visual hallucinations, or skin photosensitivity.

Patients were categorized by age groups: <2 years, 2 years to <12 years, or ≥12 years. All data for patients <12 years were analyzed together owing to the low number of patients in the <2 years category. Continuous data are reported as median and IQR. We compared the dose associated with reaching therapeutic steady-state voriconazole serum trough concentration and the proportion of patients reaching the therapeutic target by age category, using Mann-Whitney U test and chi-square test. We used a generalized linear estimating equation assuming gamma distribution with an identity link function, patients as clusters and composite symmetry covariance structure to estimate the effects of patient age, administered dose, and dose formulation (IV or enteral) on the steady-state voriconazole serum trough concentration.¹⁷ 

A total of 105 patients received voriconazole during the study period (Figure S); 59 patients met inclusion criteria (Table 1). The median age (IQR) of all included patients was 10.4 (3.7–14.7) years; 33 (56%) were aged <12 years (7 of these were <2 years) and 26 (44%) were aged ≥12 years. Proven or probable infection was diagnosed in 25 of 59 (42%) patients. The median duration of therapy was 30 days (9–83 days). The number of serum voriconazole TDM measurements per patient was 6 (2–11) and 3 (1–7.5) among those aged <12 years and ≥12 years, respectively. Most patients (61%) initiated therapy with IV voriconazole and 37% received a loading dose. There were no differences in demographic or clinical characteristics (e.g., 30-day mortality, hospital length of stay) between age groups, except that children aged <12 years were more likely to start therapy with the IV formulation of voriconazole.

Forty-two patients had at least 1 steady-state voriconazole serum trough concentration measured; 21 were aged <12 years and 21 were aged ≥12 years. The median time to steady-state measurement was 5 days overall (Table 1). Half (50%) were prescribed a concomitant proton pump inhibitor. Of those with a steady-state measurement, 21 of 42 (50%) achieved the target concentration (1–6 mcg/mL) at the first steady-state measurement (regardless of age). Among those achieving the target concentration (n = 21) at the first steady-state measurement, the dose was 19.1 mg/kg/day (16.0–20.5) vs 11.8 mg/kg/day (8.9–12.0) in patients aged <12 years and ≥12 years, respectively (Table 2).

An additional 13 of 42 (31%) patients reached the target voriconazole serum trough concentration at some point after the first steady-state measurement. For those 13 patients, the dose was 25.8 mg/kg/day (24.4–33.8) and 14.6 mg/kg/day (13.2–16.1) in patients aged <12 years and ≥12 years, respectively (Table 3).

In total 34 of 42 (81%) patients had a steady-state voriconazole serum trough concentration value within the target range (either on initial TDM or subsequently). The dose required to first achieve a value in the target range showed great variability, especially in patients aged <12 years (Figure 1). The time to achieve a value within the target range was 8.5 days (5–32) for patients <12 years and 6.5 days (3–10.5) for patients ≥12 years (p = 0.1), consistent with reported half-life of the drug (3–8 days).^2,18–20  For the 13 patients who did not reach the target voriconazole serum trough concentration on the first steady-state measurement, 9 reached the target concentration by the second TDM evaluation, 3 by the third evaluation, and 1 on the fourth steady-state measurement.

For the 34 patients who ever achieved a therapeutic voriconazole concentration, there were 132 follow-up TDM measurements. Fifty-nine percent and 81% of those steady-state measurements were in the therapeutic range in patients aged <12 years and ≥12 years, respectively (Table 4).

The route of administration at first steady-state TDM was balanced overall, with 22 of 42 (52%) administered enteral voriconazole formulations (suspension or tablet) and 20 of 42 (48%) treated with the IV formulation. The dosage used prior to the first steady-state TDM measurement was higher for those receiving an enteral formulation compared with the IV formulation (Table 5). The plot of voriconazole serum concentration by administered dose showed a great overlap between enteral and IV formulations (Figure 2). In multivariable analysis of the dose used prior to the steady-state TDM, the effect of both administered dose (p = 0.013) and formulation (p = 0.016) was statistically significant for patients aged ≥12 years, but only dose (p = 0.016), not formulation (p = 0.115), was statistically significant for those aged <12 years (Table 6).

Eight of the 42 (19%) patients with at least 1 steady-state voriconazole serum trough measurement never achieved a concentration within the target range. Of those, 4 were subtherapeutic (<1 mcg/mL), 2 were supratherapeutic (>6 mcg/mL), and 2 had both sub- and supra-therapeutic measurements (but no measurements within the target range). Eighteen patients had a voriconazole serum trough concentration >6 mcg/mL at some point during therapy (including steady-state and non–steady-state measurements). Five patients had voriconazole discontinued owing to an elevated serum drug concentration (>6 mcg/ml); of these, 3 (60%) reported adverse effects (visual disturbances or hallucinations). Nine patients who never had an elevated voriconazole serum trough concentration reported a symptom potentially related to voriconazole toxicity.

Our study highlights the challenges of dosing voriconazole to achieve therapeutic drug concentrations in children. When we used voriconazole at doses approximating the AAP recommendations, only half of the patients achieved target serum concentrations with the initial regimen. However, after subsequent dose modifications based on TDM, we were able to eventually reach the target in over 80% of patients, though this required multiple steady-state voriconazole serum trough concentration measurements. Achieving therapeutic concentrations also required administration of doses nearly 75% larger than the AAP-recommended dose in patients aged <12 years, with almost 4-fold variability between minimum and maximum doses; and 50% larger in patients aged ≥12 years.^11,14  Doses required to reach target therapeutic concentrations for patients in this study were similar to those reported for children from the Netherlands.²¹ 

Similar to other studies, we found large variability in the voriconazole dose needed to reach target serum concentrations, most notably in patients aged <12 years.^7–12,21  The dramatic variability of dose needed to reach and maintain goal concentrations in pediatric patients has been attributed to non-linear saturable pharmacokinetics in children, age-related changes in the type and function of clearance mechanisms, CYP2C19 genotype, race, and sex.^{1,2,13,18,19,22} 

Administered formulation may also be a factor in achieving and maintaining a therapeutic voriconazole trough concentration, as oral bioavailability has been noted to be lower and more variable in children (40%–80%) than in adults (90%–100%).^7,10,23  Our findings that dosing formulation was not associated with voriconazole serum trough concentrations in patients aged <12 years highlights the unpredictable pharmacokinetics and oral bioavailability in this age group. Though this effect of dosage formulation is more predictable in patients aged ≥12 years than in those aged <12 years, patients in both age groups need repeated monitoring after a change in dosing formulation to ensure the voriconazole serum concentration remains in the target range. We support initial IV therapy for children aged <12 years owing to these differences.

Similar to findings in previous studies, we achieved the target voriconazole serum concentration in only 50% of patients, using initial doses that approximated AAP recommendations.^8,9  Another 30% achieved the target only with dose adjustments and additional TDM. Larger initial doses (exceeding AAP recommendations) may be necessary to achieve therapeutic concentrations at the first steady-state measurement. Though the risk exists for a supratherapeutic concentration with larger initial doses, we did not observe that in our population, nor did we identify a concerning number of patients needing to stop therapy owing to toxic drug concentrations or patient-reported adverse effects.

Currently, no formal guidelines exist to direct TDM timing or frequency, or subsequent dose adjustments, when treating children with voriconazole. In our study cohort the median time to achieve target voriconazole serum trough concentrations was about 1 week. For children with a serious invasive fungal infection, it may be prudent to prescribe a second antifungal agent until therapeutic voriconazole concentrations can be confirmed. Our internal guideline recommends increasing the total daily dose by 50% for subtherapeutic voriconazole serum trough concentrations.¹⁵  For supratherapeutic voriconazole serum trough concentrations, our guideline recommends holding 1 dose followed by a decrease in the total daily dose by 25%, with repeated TDM at steady state. Our patients required TDM approximately every 1 to 2 weeks to ensure target concentrations were either achieved or maintained.

Strengths of our study include longitudinal data collection over a 7-year period with real-world experience, using a variety of doses over time. Limitations include small sample size within age subcategories, paucity of use of loading doses, and use of both IV and enteral dosing within a treatment course, which may have affected drug concentrations. Additionally, voriconazole serum trough concentrations were accepted as “steady state” if drawn at ≥2.5 days after initiation or dose change, though most were drawn around day 5. This was allowed for scheduling conflicts, such as weekends or holidays, and reflects our real-world experience. However, this could have resulted in premature dose reductions or escalations if early TDM was acted on. Other limitations include lack of data related to hepatotoxicities, differences in drug metabolism (e.g., CYP2C19 polymorphisms), drug interactions, and patient comorbidities.

Reaching therapeutic voriconazole serum trough concentrations required doses larger than currently recommended, and multiple TDM measurements were required to achieve and maintain therapeutic serum concentrations.

AAP

American Academy of Pediatrics;

IV

intravenous;

TDM

therapeutic drug monitoring