Oseltamivir Dosing in Premature Infants

Abstract

We conducted a 2-sample pharmacokinetic study of oseltamivir in 12 premature infants. Oseltamivir 1 mg/kg/dose twice daily in infants <38 weeks postmenstrual age (n = 8) resulted in oseltamivir carboxylate exposure comparable to previously published pediatric data, which helps prospectively validate this regimen. Oseltamivir 3 mg/kg/dose once daily in premature infants >38 weeks postmenstrual age (born prematurely but chronologically past term, n = 4) resulted in similar oseltamivir and oseltamivir carboxylate exposure. Although these results suggest persistence of immature renal function in this subgroup, further pharmacokinetic/pharmacodynamic description is required to confirm the appropriateness of this regimen.

Oseltamivir is a neuraminidase inhibitor indicated for the treatment and prevention of influenza A and B infection. After oral administration, the prodrug oseltamivir is rapidly absorbed and converted predominantly by hepatic esterases to the active metabolite oseltamivir carboxylate, which is then cleared by the kidney. Pharmacokinetic studies have been conducted and dosing guidance is available in infants, young children, and adults [1, 2]. Pharmacokinetic properties of oseltamivir may be different in premature infants, due to the immaturity of various organ systems impacting absorption, conversion, and clearance [3]. Infants are at highest risk for severe complications of influenza infection [4]; thus, pharmacokinetic studies in premature infants are essential to define an appropriate regimen.

Historically, use of oseltamivir in infants has been limited due to safety concerns, based on reports of deaths and high cerebrospinal fluid concentrations of prodrug in 7-day-old rats exposed to a single, high dose [5]. Despite these concerns, retrospective studies have emerged in human infants describing its safe use [6–8]. Dosing of oseltamivir in these reports is highly variable, likely due to limited pharmacokinetic data. Recently, the pharmacokinetics of oseltamivir have been described in a large cohort of term infants and children <2 years of age (Collaborative Antiviral Study Group [CASG] 114) [1], resulting in a preliminary treatment dosing recommendation of 3 mg/kg/dose twice daily for infants less than 1 year of age. Available guidelines recommend 3 mg/kg/dose once daily for chemoprophylaxis in infants 3–11 months of age, with no specific guidance for chemoprophylaxis in infants <3 months of age [9]. Of note, once-daily dosing of oseltamivir in pediatric patients lacks prospective pharmacokinetic evaluation. Initial dosing guidance in premature infants is based on a small, single-center pharmacokinetic sampling study (CASG 119) [10]. Due to the inherent limitations of pharmacokinetic trials in premature infants, this dataset lacks robustness. A single sample was collected from each infant and utilized to develop a population area-under-the-curve (AUC) estimate. On the basis of these data, a regimen of 1 mg/kg/dose twice daily was recommended for premature infants <38 weeks’ gestational age (GA), but this regimen requires prospective validation.

After a confirmed influenza exposure in our neonatal intensive care unit (NICU), we rapidly implemented a sampling strategy to determine oseltamivir and oseltamivir carboxylate exposure in premature infants. We sought to improve on earlier pharmacokinetic estimates by obtaining 2 samples per infant in order to prospectively assess attainment of targeted oseltamivir carboxylate exposure from once and twice daily dosing regimens.

METHODS

A 12-week-old infant born at 28 weeks’ gestation and cared for in the NICU at St. Louis Children's Hospital developed symptoms of an upper respiratory tract infection. Fluorescent antibody stain performed on a nasopharyngeal swab confirmed the infant was positive for influenza virus type A. The infant was placed on droplet isolation and treated with oseltamivir 3 mg/kg/dose twice daily. Between the onset of symptoms and the infant being placed on isolation precautions, other infants were potentially exposed through shared contacts. After an infectious diseases consultation, the neonatologist elected to administer oseltamivir to 28 infants with shared caregivers, 22 born prematurely. Exposed infants <38 weeks postmenstrual age (PMA) received oseltamivir 1 mg/kg/dose twice daily, and those >38 weeks PMA (born prematurely but chronologically past term) and term infants received 3 mg/kg/dose once daily [9, 10]. All doses were administered enterally and with feedings.

The study protocol was drafted expeditiously to enable collection of steady-state plasma samples for measurement of oseltamivir and oseltamivir carboxylate concentrations. The study was approved by the institutional human research protection office at Washington University in St. Louis, and informed consent was obtained from parents of exposed infants. A study design involving collection of 2 whole blood samples (0.5 mL each) per infant was chosen to improve the robustness of the pharmacokinetic estimates while limiting phlebotomy losses. For each infant, 1 sample was obtained during the typical oseltamivir carboxylate formation phase (0–3 hours post dose) and 1 sample was obtained during the elimination phase (>3 hours post dose). Mass spectrometry was used to quantitate oseltamivir and oseltamivir carboxylate in plasma samples; lower limits of detection for the assay were 1 and 10 ng/mL, respectively [11].

Two approaches were taken to analyze the concentration-time data. First, the measured oseltamivir and oseltamivir carboxylate concentrations were pooled to determine the median population steady-state concentration (C_ss). Second, a population pharmacokinetic model was generated (2-compartment model for oseltamivir and 1-compartment model for oseltamivir carboxylate) based on previously collected concentration-time data from CASG 114 and 119 [2, 10]. Absorption, metabolite formation, and clearance processes were assumed to be linear. Modeling was conducted using ADAPT 5 computational modeling platform (Biomedical Simulations Resource, Los Angeles, CA) [12]. The average oseltamivir carboxylate concentration over the dosing interval (C_ss) was calculated as the modeled area-under-the-curve (AUC_τ) divided by the dosing interval. Both approaches were undertaken to facilitate direct comparison with previously reported results.

RESULTS

Sixteen premature infants were enrolled. One preterm infant discontinued therapy due to increased emesis, and 2 samples had undetectable drug concentrations. Three preterm infants had only 1 sample collected after a dose. The median GA at birth and chronological age at sample collection of the remaining subjects was 28.5 weeks (range, 24–37 weeks) and 7.5 weeks (range, 1–29 weeks), respectively. The median serum creatinine was 0.3 mg/dL (range, 0.2–0.6 mg/dL). None of the exposed infants receiving oseltamivir developed influenza infection.

Eight study infants were <38 weeks PMA at enrollment and, therefore, received 1 mg/kg/dose twice daily. The median number of doses administered prior to sampling was 8 (range, 6–9). Figure 1 illustrates the raw concentration-time data from infants in the current study receiving 1 mg/kg/dose twice daily compared with those from all cohorts in the CASG 114 study (n = 78). The median oseltamivir concentration was 6.9 ng/mL (range, <1–39.6 ng/mL). The median oseltamivir carboxylate concentration was 318 ng/mL (range, 61–543 ng/mL). The population pharmacokinetic model identified a median oseltamivir carboxylate AUC₁₂ of 3846 ng × h/mL (range, 2257–6866 ng × h/mL) and a C_ss of 320 ng/mL (range, 188–572 ng/mL).

Figure 1.

Left panel, Measured oseltamivir phosphate concentrations from pediatric patients in CASG 114 (circles; dashed lines 95% CI) and from premature infants <38 weeks PMA in the current study (triangles). Right panel, Measured oseltamivir carboxylate concentrations from pediatric patients in CASG 114 (circles, dashed lines 95% CI) and from premature infants <38 weeks PMA in the current study (triangles). The x-axis is represented in collection time windows, as this was the sample collection design in CASG 114.

Abbreviations: CASG, Collaborative Antiviral Study Group; CI, confidence interval; PMA, postmenstrual age.

Four study infants were >38 weeks PMA (GA 28–37 weeks at birth) and received 3 mg/kg/dose once daily. The median number of doses administered prior to pharmacokinetic sampling was 4 (3–5). Figure 2 illustrates the raw concentration-time data from infants in the current study receiving 3 mg/kg/dose once daily. The median oseltamivir concentration for these subjects was 3.6 ng/mL (range, <1–132 ng/mL). The median oseltamivir carboxylate concentration was 347 ng/mL (range, 43–513 ng/mL). The population pharmacokinetic model identified a median oseltamivir carboxylate AUC₂₄ of 8005 ng × h/mL (range, 5718–10 891 ng × h/mL) and a C_ss of 334 ng/mL (range, 238–454 ng/mL).

Figure 2.

Left panel, Measured oseltamivir phosphate concentrations from premature infants >38 weeks PMA in the current study (triangles). Right panel, Measured oseltamivir carboxylate concentrations from premature infants >38 weeks PMA in the current study (triangles).

Abbreviation: PMA, postmenstrual age.

DISCUSSION

Our study provides unique oseltamivir and oseltamivir carboxylate pharmacokinetic data in premature infants. Although our sample size is small, these results help prospectively validate the previous dosing recommendation of 1 mg/kg/dose twice daily for the treatment of influenza in premature infants <38 weeks PMA. Our data also describe the use of 3 mg/kg/dose once daily in critically ill infants born prematurely but chronologically past term (>38 weeks PMA), although this population clearly requires further pharmacokinetic evaluation.

The present data suggest that 1 mg/kg/dose twice daily for infants <38 weeks PMA results in similar maximum oseltamivir concentrations to premature infants from CASG 119, but lower than term infants and young children receiving 3 mg/kg/dose twice daily (Figure 1) and adults receiving 75 mg twice daily (adult oseltamivir maximum concentration 65 ± 26 ng/mL) [1, 2]. Oseltamivir carboxylate exposure in the present cohort was similar to previously described pediatric cohorts. In the absence of appropriately designed concentration-response studies in adults, the initial target of pediatric dose-finding trials has been AUC exposure similar to established adult dosing. Adults receiving 75 mg twice daily achieve an average oseltamivir carboxylate AUC₁₂ of 2719 ng × h/mL [2]. In CASG 114, the target AUC₁₂ was 3800 ng × h/mL, and a median dose of 3 mg/kg/dose twice daily administered to patients ≤2 years of age produced an oseltamivir carboxylate AUC₁₂ of 4326 ng × h/mL [1]. In CASG 119, a median dose of 1.73 mg/kg/dose twice daily administered to infants <38 weeks GA produced a modeled oseltamivir carboxylate AUC₁₂ of 9250 ng × h/mL [10]. This was the basis for the lower dosage utilized in premature infants in our cohort. Utilizing this dose, the modeled oseltamivir carboxylate AUC₁₂ for premature infants in the current study was 3846 ng × h/mL. This dose produced a modeled C_ss of 320 ng/mL and a median of all raw concentrations of 318 ng/mL, which is similar to the target AUC₁₂ and C_ss observed in CASG 114. These parameters suggest that a dose of 1 mg/kg/dose twice daily is appropriate in premature infants <38 weeks PMA.

To our knowledge, this report represents the first pharmacokinetic description of premature infants >38 weeks PMA receiving 3 mg/kg/dose once daily. Daily oseltamivir dosing lower than that in CASG 114 (6 mg/kg/day) carries theoretical risks in young children. In pediatric trials, a dose of 2 mg/kg/day resulted in lower oseltamivir carboxylate exposure in children 1–2 years of age (AUC₂₄ = 1850 ng × h/mL) compared to older children (AUC₂₄ = 2410 ng × h/mL in children 3–5 years of age) [13]. Utilization of 4 mg/kg/day in young children was also associated with more frequent emergence of neuraminidase mutations conferring oseltamivir resistance [14]. The modeled data in this small cohort suggests that 3 mg/kg/dose once daily produces an oseltamivir carboxylate AUC₂₄ (8005 ng × h/mL) similar to twice the AUC₁₂ from CASG 114, although the dose utilized was 50% lower. Our findings must be interpreted with caution because the sample size is small and efforts to limit phlebotomy precluded sampling across the entire 24-hour dosing interval. Additionally, these results do not suggest that the efficacy of a once-daily regimen will be the same as twice daily. Importantly, there are no pharmacodynamic data to suggest which exposure variable (AUC, C_ss, C_max, etc.) is related to efficacy, and some parameters will be different between once- and twice-daily dosing. In fact, our results imply babies born prematurely but chronologically past term demonstrate persistence of immature renal function and optimal dosing may be a twice-daily regimen similar to infants <38 weeks PMA. As only 1 study infant was >40 weeks PMA, our preliminary results may be limited to premature infants below this maturation. More robust pharmacokinetic and pharmacodynamic studies are necessary to confirm the appropriateness of once-daily dosing of oseltamivir in this population.

Finally, the present data confirm the ability of premature infants to convert oseltamivir to oseltamivir carboxylate. Theoretical concerns exist about accumulation of parent drug in patients <1 month of age, given recent developmental data indicating minimal hepatic esterase activity during the first 31 days of life [15]. We did not observe undesired parent drug accumulation in our cohort, though it included 4 premature infants <1 month old and 11 infants ≤40 weeks PMA. This finding supports the presence of hepatic esterase activity in these young infants, although it may be limited compared to more mature subjects. Interestingly, the oseltamivir carboxylate panel for the current study (Figure 1) shows an approximately 1-hour metabolite formation lag phase, which is consistent with the youngest infants in CASG 114. This lag phase may reflect administration with feedings, which is manifest in the lower overall oseltamivir exposure. Collectively, these results agree with previously reported cohorts of both term and premature infants [1, 10].

Our dataset has clear limitations. Only 2 samples were obtained per infant. This is an improvement on previous data, which obtained only 1 sample per infant, although neither design is ideal for accurately estimating pharmacokinetic parameters. Unfortunately, pharmacokinetic studies in infants will likely continue to be limited by concerns over excessive invasive sampling and phlebotomy losses. Our cohort was also limited to the size of the exposed group and to those infants whose parents consented to study participation.

Although results of the current study support the appropriateness of oseltamivir 1 mg/kg/dose twice daily for premature infants ≤38 weeks PMA, further description of the pharmacokinetic properties of oseltamivir and oseltamivir carboxylate in premature infants is necessary. This aim will likely be achieved through collaboration between multiple centers with standing protocols for oseltamivir dosing and sample collection in the event of influenza exposure.