Zanamivir Pharmacokinetics and Pulmonary Penetration into Epithelial Lining Fluid following Intravenous or Oral Inhaled Administration to Healthy Adult Subjects

Abstract

Zanamivir serum and pulmonary pharmacokinetics were characterized following intravenous (i.v.) or oral inhaled administration. I.v. zanamivir was given as intermittent doses of 100 mg, 200 mg, and 600 mg every 12 h (q12h) for two doses or as a continuous infusion (6-mg loading dose followed by 3 mg/h for 12 h). Oral inhaled zanamivir (two 5-mg inhalations q12h for two doses) was evaluated as well. Each zanamivir regimen was administered to six healthy subjects with serial pharmacokinetic sampling. In addition, a single bronchoalveolar lavage (BAL) fluid sample was collected at various time points and used to calculate epithelial lining fluid (ELF) drug concentrations for each subject. For intermittent i.v. administration of 100 mg, 200 mg, and 600 mg zanamivir, the median zanamivir concentrations in ELF collected 12 h after dosing were 74, 146, and 419 ng/ml, respectively, each higher than the historic mean 50% inhibitory concentrations for the neuraminidases of wild-type strains of influenza A and B viruses. Median ELF/serum zanamivir concentration ratios ranged from 55 to 79% for intermittent i.v. administration (when sampled 12 h after the last dose) and 43 to 45% for continuous infusion (when sampled 6 to 12 h after the start of the infusion). For oral inhaled zanamivir, the median zanamivir concentrations in ELF were 891 ng/ml for the first BAL fluid collection and 326 ng/ml for subsequent BAL fluid collections (when sampled 12 h after the last dose); corresponding serum drug concentrations were undetectable. This study demonstrates zanamivir's penetration into the human pulmonary compartment and supports the doses selected for the continuing development of i.v. zanamivir in clinical studies of influenza.

INTRODUCTION

Zanamivir (Relenza) for oral inhalation is approved for the treatment and prophylaxis of uncomplicated acute illness due to influenza A and B virus infections in adult and pediatric patients (7). Zanamivir inhibits viral neuraminidase, affecting the release of viral particles. Mean zanamivir neuraminidase 50% inhibitory concentrations (IC₅₀s) of 0.76, 1.82, and 2.28 nM for subtype A/H1N1 (N1), A/H3N2 (N2), and B neuraminidases, respectively, were reported for more than 1,000 clinical influenza virus isolates collected from various parts of the world between 1999 and 2002 as part of the Neuraminidase Inhibitor Susceptibility Network (12). Prompt use of neuraminidase inhibitors has been recommended for the treatment of patients with influenza (18); however, administration of anti-influenza drugs via conventional routes (tablets, suspension, or oral inhalation) may not be practical or appropriate for hospitalized patients, as a result of impaired absorption (of oral agents) from the gut or inability to administer these drugs to patients requiring mechanical ventilation. Furthermore, recent reports of resistance to oseltamivir further highlight the critical need for additional treatment options (4, 5, 6, 13, 17).

Although not registered, an aqueous formulation of zanamivir suitable for intravenous (i.v.) administration was used during the clinical development program for oral inhaled zanamivir. The pharmacokinetics of i.v. zanamivir have been studied after single escalating doses ranging from 1 to 600 mg and repeated doses of 600 mg twice daily for 5 days (3). The pharmacokinetics of zanamivir were linear, and there was no evidence of a change in disposition after repeated twice-daily administration. Approximately 90% of the zanamivir administered was excreted unchanged in the urine, with first-order elimination, a half-life of approximately 2 h, and a volume of distribution that was similar to that of extracellular water. An experimental influenza virus challenge study was conducted in which healthy male subjects received 600 mg of zanamivir i.v. twice daily for 5 days, commencing 4 h prior to inoculation with influenza A/Texas/91 (H1N1) virus (2). I.v. zanamivir had a significant prophylactic effect, as demonstrated by a low seroconversion rate (14% in the zanamivir group versus 100% in the placebo group) and a lack of isolation of virus in culture (0% in the zanamivir group versus 100% in the placebo group); in addition, i.v. zanamivir at this dose was well tolerated by the study subjects. I.v. zanamivir is currently in late-stage clinical development.

The present study was designed to evaluate serum and pulmonary pharmacokinetics following i.v. and oral inhaled administration (at the approved treatment dose of 10 mg twice daily) of various regimens of zanamivir. Since i.v. zanamivir at 600 mg every 12 h (q12h) had already demonstrated activity in vivo (2), this regimen was included in the present study to document the intrapulmonary pharmacokinetics of i.v. zanamivir for a regimen with known activity. Although zanamivir concentrations in nasal fluid (NF) with this regimen exceeded the IC₅₀ for influenza virus neuraminidase by many fold (3), direct measures of pulmonary exposure following i.v. administration had not been determined. Given the magnitude of estimated zanamivir concentrations above the IC₅₀ in the pulmonary compartment for zanamivir at 600 mg i.v., several lower-dose, intermittent i.v. infusion regimens and continuous infusions were evaluated in this study as well. Continuous i.v. infusion of zanamivir was included to evaluate the time course of intrapulmonary penetration using a regimen that could maintain serum drug concentrations using a lower total daily dose (relative to intermittent administration). The continuous-infusion regimen used in this study was selected to deliver average serum drug concentrations similar to the trough concentration previously reported following intermittent i.v. administration of 600 mg of zanamivir (3). The oral inhaled product was evaluated to provide an estimate of pulmonary pharmacokinetic exposure to zanamivir following oral inhalation of the approved dose (7, 14).

This study utilized bronchoalveolar lavage (BAL) for the collection of samples to determine zanamivir concentrations in epithelial lining fluid (ELF). Given the invasive nature of this procedure and the variability of the measurement of ELF drug concentrations, each subject had BAL fluid sampling on one occasion during the study and six subjects were evaluated at each time point for BAL fluid sampling. NF samples were collected simultaneously to enable informal comparisons to ELF samples.

(This study was presented at the 47th Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, IL, 17 to 20 September 2007.)

MATERIALS AND METHODS

Volunteers.

Healthy adult subjects were enrolled and evaluated between 15 June 2006 and 14 August 2006. Subjects were 18 to 55 years of age, were nonsmokers, had body mass indexes of 19 to 30 kg/m², and had a calculated creatinine clearance greater than or equal to 70 ml/min. Written informed consent was obtained from each subject prior to the performance of any study-specific procedures. During screening, physical examinations and clinical laboratory tests were performed and medical histories were recorded. The study protocol and informed consent were reviewed and approved by an institutional review board.

Study design.

This study was an open-label, nonrandomized, single period, multiple-dose pharmacokinetic study to evaluate serum, nasal, and pulmonary pharmacokinetics following the administration of i.v. or oral inhaled zanamivir (protocol NAI106784). Healthy subjects received one of the following regimens: intermittent i.v. infusion of zanamivir at 100 mg, 200 mg, or 600 mg q12h for two doses; continuous i.v. zanamivir (as a 6-mg loading dose over 2 min, followed by administration at a rate of 3 mg/h for 12 h for a total dose of 42 mg); or oral inhaled zanamivir at 10 mg (as two 5-mg inhalations via Rotadisk/Diskhaler) q12h for two doses. The intermittent zanamivir infusions were each administered over 30 min. For continuous infusion, there were three cohorts in which BAL fluid sampling occurred at 2, 6, or 12 h after the start of the infusion.

Eligible subjects were confined the evening prior to dosing and pharmacokinetic evaluation. For the intermittent i.v. infusion regimens, serum samples were collected and at 0.5 (end of the infusion), 0.75, 1, 1.25, 1.5, 2, 3, 4, 6, 8, 12, and 24 h after the start of the infusion of the second dose. For the oral inhaled zanamivir regimen, serum samples were collected predose and at 5, 15, 30, and 45 min and 1, 1.5, 2, 3, 4, 6, 8, 12, and 24 h after the second dose. For the continuous i.v. infusions, serum samples were collected predose and at 2, 6, 12 (end of the infusion), 12.5, 12.75, 13, 13.25, 13.5, 14, 15, 16, 18, 20, and 24 h after the start of the infusion.

BAL fluid and nasal wash samples were collected 12 h after the second dose of zanamivir for the intermittent regimens and at 2, 6, or 12 h (depending on the cohort) after the start of the continuous-infusion regimens.

For bronchoscopy, a fiber optic bronchoscope (Olympus P-20 and P-20D; Olympus America Inc., Melville, NY) was inserted into a subsegment of the right middle lobe of the lung. The following medications were administered to facilitate bronchoscopy: codeine (30 mg, oral), nebulized atropine-lidocaine (2 ml atropine at 0.4 mg/ml with 2 ml 4% lidocaine), and 1 to 2 mg midazolam (i.v.) with 25 to 75 μg fentanyl (i.v.). During bronchoscopy, 2% viscous lidocaine was applied to the distal end of the bronchoscope, 8 ml of 4% lidocaine was administered to the vocal chords, and approximately 10 ml of 1% lidocaine was administered to the lower airway. Following standardized bronchoscopy (one for each subject), four consecutive BAL fluid samples were collected by the instillation of 50 ml of 0.9% saline into the right middle lobe, followed by immediate aspiration. The first aspirate (BAL1) was kept separate, and the latter three aspirates were combined (BAL2). After the recovered volumes were recorded, BAL samples were centrifuged to obtain a supernatant and a cell pellet. Due to an error in sample processing (i.e., lack of complete removal of BAL aspirate supernatant), the cell pellet samples were received diluted in various volumes of residual BAL supernatant. As a result, the planned analysis to quantitate macrophage zanamivir concentrations could not be conducted. Nasal washes were collected within 30 min prior to the collection of the BAL fluid sample. Five milliliters of 0.9% saline was instilled into each nostril, and the effluent was expelled into a sterile collection cup. The volume of nasal effluent was measured and recorded.

Assessments of safety included the collection of data on adverse events (AEs) throughout the study period. Twelve-lead electrocardiograms (ECGs) and vital signs (including heart rate and blood pressure) were evaluated before and 30 min after the administration of the second dose for intermittent i.v. and oral inhaled administrations of zanamivir or before and 8 h after the start of the continuous i.v. administration of zanamivir. Clinical laboratory evaluations were performed prior to dosing and on day 2 after the completion of dosing, including determination of urea. Urea, a small hydrophobic molecule, was to be used as an endogenous marker of ELF recovered by BAL fluid and NF, since its concentration in these matrices is assumed to be the same as that in plasma (1).

Bioanalytical assessments.

All samples (serum, BAL aspirates, and nasal effluent) were stored at −70°C until assay and analyzed by protein precipitation and liquid chromatography-tandem mass spectrometry in the positive-ion turbo ion spray mode. The calibration curve range for zanamivir in serum was 10 to 10,000 ng/ml. During method validation of the serum assay, the maximum bias observed was 9.3% and the maximum observed precision values within and between runs were 9.0% and 0.6%, respectively. The calibration curve range was 0.2 to 100 ng/ml for zanamivir in BAL fluid and nasal effluent. During method validation of the BAL fluid and nasal effluent assay, the maximum bias observed was −14.7% and the maximum within- and between-run precision values observed were 13.3% and 7.4%, respectively. Quality control (QC) samples were prepared at three different analyte concentrations and analyzed with each batch of samples against separately prepared calibration standards. For the analysis to be acceptable, no more than one-third of the QC results were to deviate from the nominal concentration by more than 15%, and at least 50% of the results from each QC concentration should be within 15% of nominal. The applicable analytical runs met all predefined run acceptance criteria.

Urea concentrations in BAL fluid and nasal wash samples were determined using a routine Bayer method (catalogue number 03040257; Bayer, Tarrytown, NY) modified by increasing the absorbance read time from 1.6 min to 20 min and by decreasing the sample-to-reagent volume ratio from 1:74 to 1:4 to give improved assay sensitivity (0.05 mg/dl), with a 10% intraday variability and 72 to 75% accuracy. QC of the assay was performed by assaying diluted control materials.

Pharmacokinetic and statistical analyses.

Serum pharmacokinetic parameters for each subject were estimated by noncompartmental analysis using WinNonlin version 5.2 (Pharsight Corporation, Cary, NC). The maximal drug concentration in serum (C_max), the time needed to reach the maximum concentration (t_max), the area under the concentration-time curve to the last quantifiable concentration (AUC_0-t), the elimination rate (λ_z), the elimination half-life (t_1/2), drug clearance (CL), and volume (V_z) were determined for all regimens as the data permitted. Computation of λ_z was based on a minimum of three time points. The AUC over a dosing interval was computed for intermittent i.v. infusion regimens, and the AUC during the infusion interval was computed for continuous-infusion regimens, both designated AUC_0-12. Trough concentrations were computed for intermittent i.v. administration, and concentrations at the end of the infusion were computed for continuous-infusion regimens, both designated C₁₂. Additionally, analyses of the AUC and C_max of zanamivir in serum were performed to provide estimates of between-subject variability. Zanamivir pharmacokinetic parameters were summarized as geometric mean values with between-subject variability estimates (percent coefficient of variation [%CV_b]).

Calculations of ELF volume and zanamivir concentrations in ELF.

ELF zanamivir concentrations were calculated based upon relative urea concentrations in BAL fluid versus serum and recovered volume of BAL fluid to derive a concentration in ELF. NF drug concentrations were similarly calculated based upon nasal effluent drug concentrations and relative urea concentrations.

The amount of ELF recovered was calculated according to the urea dilution method described by Rennard et al. (15). The concentration of zanamivir in ELF was estimated from the concentration of zanamivir in BAL aspirate supernatant, the volume of BAL fluid collected, and the ratio of the urea concentration in BAL aspirate supernatant to that in plasma, as determined by previously published methods (8, 15). Thus, the concentration of zanamivir in ELF was determined as ZAN_ELF = (ZAN_BAL × V_BAL)/V_ELF, where ZAN_BAL is the concentration of zanamivir measured in the BAL aspirate supernatant, V_BAL is the volume of the BAL aspirate, and V_ELF = V_BAL × (UREA_BAL/UREA_blood), where UREA_BAL = is the concentration of urea in the BAL aspirate supernatant and UREA_blood is the concentration of urea in blood. UREA_BAL was corrected for red blood cells present in the BAL aspirate supernatant. ELF drug concentrations were calculated for BAL1 and BAL2 separately. The concentration of zanamivir in NF was estimated from the concentration in nasal effluent by using a correction for the nasal effluent urea concentration as follows: ZAN_NF = ZAN_{nasal effluent} × (UREA_blood/UREA_{nasal effluent}). The relationship between the zanamivir concentrations in ELF and serum was explored using linear regression.

RESULTS

Volunteers.

A total of 43 subjects were enrolled, and 42 subjects completed this study. One subject receiving 600 mg of zanamivir i.v. q12h was withdrawn due to an abnormal T wave on his ECG that occurred prior to the second i.v. dose of zanamivir and prior to the BAL procedure. All of the subjects in this study were male, the majority were white (86%), the mean age ± the standard deviation was 26 ± 8 (range, 18 to 47) years, and the mean body mass index was 26 ± 3 kg/m². Demographic characteristics were similar across the regimens, including the body mass index and calculated creatinine clearance.

Serum pharmacokinetics.

The t_1/2 of zanamivir was determined to be ∼2.5 h across the different i.v. regimens (Table 1). Serum zanamivir concentrations were quantifiable (i.e., >10 ng/ml) up to the last sampling time point for the 200-mg and 600-mg doses and were quantifiable up to 12 h for the 100-mg dose. Serum zanamivir concentrations were quantifiable throughout the continuous-infusion regimens. Since there were minimal apparent differences between AUC_0-t and AUC_0-12, only AUC_0-12 is reported in Table 1.

Table 1.

Serum zanamivir pharmacokinetics following i.v. or oral inhaled administration

Zanamivir regimen	n	Cmax (ng/ml)	AUC0-12 (ng · h/ml)	tmaxa (h)	C12b (ng/ml)	t1/2 (h)
Intermittent infusion
600 mg i.v. q12hc	6	39,430 (11.7)g	86,630 (11.6)	0.50 (0.50–0.50)	586 (38.2)	2.89 (3.35)
200 mg i.v. q12hc	6	13,149 (15.0)	31,671 (10.3)	0.50 (0.50–0.50)	252 (39.0)	2.68 (16.6)
100 mg i.v. q12hc	6	7,430 (12.6)	16,386 (12.6)	0.50 (0.50–0.50)	114 (28.3)	2.23 (7.66)
Continuous infusion (BAL sampling)f at:
2 h	6	574 (19.6)	5,966 (18.6)	7.00 (5.00–12.00)	581 (422–652)a	2.37 (8.26)
6 h	6	527 (15.0)	5,356 (16.1)	6.01 (5.78–8.00)	464 (378–597)a	2.13 (18.1)
12 h	6	499 (9.83)	5,136 (8.42)	7.00 (6.00–11.77)	452 (389–535)a	2.29 (9.24)
Oral inhalation, 10 mg q12hc	6	21.2 (45.3)	175 (25.8)	1.75 (0.25–4.00)	BQLd	NRe

^aPresented as median and range.

^bAt 12 h after the second dose for intermittent i.v. and oral inhaled regimens and 12 h after the start of the continuous infusion.

^cBAL sampling 12 h after the second dose.

^dBQL, below quantifiable limit of 10 ng/ml.

^eNR, not reportable due to low serum drug concentrations.

^fA 6-mg zanamivir loading dose followed by 3 mg/h.

^gEach value is the geometric mean (%CV_b) unless specified otherwise.

Zanamivir was readily absorbed following oral inhaled administration, with the maximum concentration achieved between 0.25 and 4 h postdose (Table 1); however, serum drug concentrations were quantifiable only up to 4 to 8 h postdose and were variable in the terminal phase, resulting in poor estimation of λ_z and t_1/2. Therefore, C₁₂, λ_z, and t_1/2 were not reported for oral inhalation.

Estimates of between-subject variability for selected zanamivir pharmacokinetic parameters (AUC_0-12, AUC_0-τ, and C_max) following i.v. administration were low, with %CV_b values ranging from 10% to 20%.

Zanamivir concentrations in ELF and NF.

Following i.v. administration, calculated median zanamivir concentrations in ELF were generally higher when using BAL1 than when using BAL2 (data not shown). Given the localized distribution of an initial BAL and the potential for contamination during insertion of the bronchoscope, zanamivir concentrations in ELF were calculated and reported using BAL2 for all of the i.v. regimens as planned. This is consistent with other studies evaluating ELF drug concentrations of other systemic anti-infectives (8, 15). Since the oral inhaled zanamivir regimen represents localized delivery, ELF zanamivir concentrations calculated from BAL1 were anticipated to be higher than ELF zanamivir concentrations calculated from BAL2. The median ELF zanamivir concentration calculated using BAL1 was indeed approximately 3-fold (range, 1.3- to 6.6-fold) higher than that calculated using BAL2. As a result, zanamivir concentrations in ELF were calculated using both BAL1 and BAL2 for the oral inhaled regimen as a conservative measure.

Descriptive statistics of zanamivir concentrations in serum, ELF, and NF, along with individual ELF/serum and NF/serum zanamivir concentration ratios, are presented in Table 2. For 100 mg, 200 mg, and 600 mg zanamivir, the median 12-h postdose ELF zanamivir concentrations (calculated from BAL2) were 74, 146, and 419 ng/ml, respectively, suggesting dose proportionality. For continuous infusions, the median ELF zanamivir concentrations were 141, 209, and 197 ng/ml when BAL sampling was conducted at 2, 6, and 12 h after the start of the infusion, respectively. Individual zanamivir concentrations in ELF for continuous infusions are displayed in Fig. 1. From linear regression analysis, zanamivir concentrations in ELF were well correlated with serum zanamivir concentrations following intermittent administration (slope = 1.01, intercept = 74.8 ng/ml, r² = 0.75; data not shown). Following the intermittent i.v. administration of zanamivir, the median ELF/serum zanamivir concentration ratios ranged from 0.55 to 0.79. For continuous-infusion regimens, ELF/serum zanamivir concentration ratios were 0.24, 0.43, and 0.45 for samples taken at 2, 6, and 12 h after the start of the infusion. For oral inhaled administration, the median 12-h postdose ELF zanamivir concentration was 891 ng/ml when calculated from BAL1 and 326 ng/ml when calculated from BAL2. Serum zanamivir concentrations were below the level of quantitation (i.e., <0.2 ng/ml) at the time of BAL fluid sampling for oral inhaled zanamivir, prohibiting computation of the ELF/serum zanamivir concentration ratio. The absolute zanamivir concentrations in ELF in relation to the IC₅₀s for neuraminidases of various subtypes of influenza virus (12) are shown in Fig. 2.

Table 2.

Zanamivir concentrations in serum, ELF, and NF following i.v. or inhaled zanamivir administration

Zanamivir regimen [BAL sampling time (h)]	Zanamivir concn (ng/ml)			Individual-subject ratio
	Seruma	ELFb	NF	ELF/serum	NF/serum
600 mg i.v. q12h (12)	642 (290–825)f	419 (216–1,163)	498 (212–1,547)	0.73 (0.6–1.4)	0.84 (0.3–2.3)
200 mg i.v. q12h (12)	247 (143–438)	146 (70.9–299)	101 (58.1–296)	0.55 (0.3–1.2)	0.47 (0.2–112)
100 mg i.v. q12h (12)	117 (75–169)	74.0 (59.2–114)	46 (20.7–98.9)c	0.79 (0.4–0.9)	0.60 (0.1–0.7)
Continuous i.v. infusion (2)	536 (394–644)	141 (66.0–190)	BQLd	0.24 (0.1–0.3)	NCe
Continuous i.v. infusion (6)	536 (423–630)	209 (129–259)	BQL (BQL–13.3)d	0.43 (0.3–0.5)	NCe
Continuous i.v. infusion (12)	452 (389–535)	197 (187–282)	BQL (BQL–87.5)d	0.45 (0.4–0.6)	NCe
Oral inhalation, 10 mg (12)	BQLd	BAL1, 891 (116–3,189); BAL2, 326 (90–760)c	613 (42.0–1,911)c	NCe	NCe

^aC₁₂ at 12 h after the second dose for intermittent i.v. and oral inhaled and 12 h after the start of the continuous infusion.

^bOnly BAL2 reported for i.v. regimens.

^cn = 5.

^dBQL, below quantifiable limit of 10 ng/ml for serum or 0.2 ng/ml for NF.

^eNC, not calculated due to low concentrations.

^fResults are medians (ranges).

Fig. 1.

Scatterplot of ELF zanamivir concentrations (derived from BAL2) for continuous-infusion regimens according to sampling time.

Fig. 2.

Scatterplot of ELF zanamivir concentrations (on a log scale) according to zanamivir regimen in relation to the IC₅₀s for subtype A/H1N1 (N1), A/H3N2 (N2), and B neuraminidases reported by the Neuraminidase Inhibitor Network (12). ELF drug concentrations were derived from BAL2 for i.v. regimens and from BAL1 and BAL2 for oral inhaled regimens. BID, twice daily.

Safety.

All of the zanamivir regimens were well tolerated during the study. The overall safety profile of the various zanamivir regimens in this study was similar to the previously reported safety profile of i.v. zanamivir (2). Zanamivir administered as a continuous infusion had a safety profile similar to that observed with intermittent infusion. No serious AEs or deaths were reported.

The nature and frequency of AEs appeared similar across the various regimens. The AEs most commonly reported by the 43 subjects who received zanamivir were leukocytosis (6 subjects), neutrophilia (5 subjects), postprocedural complications (5 subjects), pharyngolaryngeal pain (4 subjects), dizziness (4 subjects), and cough (3 subjects). Only three subjects reported four AEs (all mild in severity) that were considered by the investigator to be drug related: an abnormal T wave in the ECG, back pain, pain in an extremity, and headache.

One subject, a 21-year-old healthy Caucasian male, received 600 mg of zanamivir i.v. q12h and was withdrawn due to the AE of an ECG abnormality. After the first i.v. dose of zanamivir and just prior to the second dose, occurred an abnormal T-wave inversion in the predose ECGAE that was confirmed by repeat ECG assessment. This was considered by the investigator to be clinically significant and potentially related to the study drug, and 325 mg of aspirin was administered. A screening ECG 6 days prior to dosing had been normal, but no ECG assessments were performed prior to or immediately following the administration of the first dose and therefore the precise onset of the T-wave abnormality remained unknown. The subject had no associated symptoms and had stable vital signs throughout. Creatinine kinase-MB and troponin tests were normal. Further medical evaluation did not reveal any underlying cardiac illness or etiology. This was the only clinically significant ECG value reported during this study.

Three subjects who received the continuous zanamivir infusion had clinically significant abnormal laboratory values approximately 20 h after the start of the infusions. One subject had increased total bilirubin (1.8 times the upper limit of normal [1.8 × ULN]), one subject had increased white blood cell counts (1.4 × ULN), and one subject had both increased total bilirubin (2.2 × ULN) and increased white blood cell counts (1.8 × ULN). All of these values had resolved at follow-up.

DISCUSSION

This study was designed to evaluate zanamivir penetration into the pulmonary compartment as measured by ELF concentrations following i.v. administration of zanamivir. I.v. administration of 600 mg of zanamivir q12h has been demonstrated to have a significant prophylactic effect in an experimental challenge with influenza A virus (2), and i.v. zanamivir showed evidence of efficacy against H5N1 in a macaque model (16). However, efficacy in clinical influenza virus infection has not been established. Recently, several case reports regarding the successful use of i.v. zanamivir to treat pandemic H1N1 2009 pneumonitis unresponsive to oseltamivir have been published (5, 6, 9, 11).

Given the cellular surface location of the sialidase enzyme, ELF is likely to be the most relevant matrix to consider for zanamivir from a pharmacodynamic standpoint, whereas alveolar macrophage drug concentrations are important for intracellular infection (1). Characterization of alveolar macrophage zanamivir concentrations was included in the protocol to gauge the potential for these cells to serve as a source of zanamivir in ELF via intracellular-extracellular efflux. However, an error in sample processing (i.e., lack of complete removal of BAL fluid supernatant) prevented our quantitation of zanamivir concentrations within the cell pellets from the BAL fluid. Consistent with other studies measuring ELF drug concentrations, rapid processing and short-term storage of BAL samples on ice were employed to minimize the potential for efflux from macrophages into the BAL supernatant. However, without being able to quantitate intracellular zanamivir concentrations, the potential contribution of intracellular efflux to the measured ELF drug concentrations cannot be estimated. Theoretically, efflux from macrophages and other cells during processing may have contributed to the measured ELF drug concentrations, potentially resulting in a slight overestimation of ELF zanamivir concentrations (the total volume of macrophages accounted for less than 1% of the recovered BAL fluid).

In this study, two BAL samples were collected from each subject and analyzed; BAL1 was the first aliquot, and BAL2 consisted of the second, third, and fourth aliquots combined. A previous study evaluated the anatomic distribution of BAL following sequential aliquots of saline and found that the first aliquot was aspirated almost exclusively from an anatomic region close to the bronchoscope, whereas fluid movement occurred throughout the whole lung segment for the second and third aliquots (10). As a result, the initial BAL wash has been discarded in most studies that have quantified antimicrobial concentrations in the pulmonary compartment following systemic administration (1). However, it was uncertain whether or not it would be appropriate to discard the initial BAL wash for oral inhaled zanamivir due to the localized drug delivery via the airways. As a result, both BAL1 and BAL2 were analyzed in this study. Generally, zanamivir concentrations in ELF were higher when calculated using BAL1 than when calculated using BAL2, especially for oral inhaled administration (median value, approximately 3-fold higher for BAL1 than for BAL2). Considering the localized administration, both BAL1 and BAL2 were reported for oral inhaled zanamivir, whereas only BAL2 was reported for i.v. administration, consistent with previously published methods. Hence, the ELF zanamivir concentration calculated from BAL1 for oral inhaled zanamivir may reflect a nonuniform distribution in the lung, such that higher concentrations are evident in a region of the lung that is more proximal to the position of the bronchoscope.

Following the administration of 600 mg of zanamivir i.v. twice daily, the median concentration of zanamivir in ELF evaluated 12 h after dosing was 419 ng/ml (73% of the serum drug concentration at the same time point), which is 552 to 1,653 times the in vitro IC₅₀s for influenza virus neuraminidases reported by the Neuraminidase Inhibitor Network [0.76, 1.82, and 2.28 nM for the subtype A/H1N1 (N1), A/H3N2 (N2), and B neuraminidases, respectively] (12). The median zanamivir concentration in NF was similar to that in ELF (498 versus 419 ng/ml) but higher than historic NF zanamivir concentrations reported for this regimen (3), where the median NF zanamivir concentration was 184 ng/ml following twice daily i.v. administration of 600 mg of zanamivir for 5 days (12 h after dosing).

Following the administration of 100 mg and 200 mg of zanamivir twice daily, the median concentrations of zanamivir in ELF at 12 h were 74.0 ng/ml (79% of the concentration in serum) and 146 ng/ml (55% of the concentration in serum), respectively, which are 96 to 291 and 192 to 576 times the in vitro IC₅₀s for neuraminidases of various influenza virus subtypes reported by the Neuraminidase Inhibitor Network (12). The median zanamivir concentrations in NF were 46.0 and 101 ng/ml, respectively, following the administration of 100 mg and 200 mg of zanamivir twice daily, with similar ranges of NF and ELF zanamivir concentrations at the same time points.

Following the administration of 10 mg of zanamivir by oral inhalation, zanamivir was readily absorbed, with a median serum C_max of 21.2 ng/ml and a median t_max value of 1.75 h postdose. Serum drug concentrations quickly declined to below the limit of quantification at 4 to 8 h postdose. Although serum zanamivir concentrations at 12 h were too low to be measured (i.e., <10 ng/ml), the median zanamivir concentration in ELF calculated from BAL1 at 12 h was 891 ng/ml, which represents 1,172 to 3,517 times the IC₅₀ for influenza virus neuraminidases of various influenza virus subtypes reported by the Neuraminidase Inhibitor Network (12). The corresponding median ELF zanamivir concentration calculated using BAL2 was 326 ng/ml, which was 429 to 1,287 times the in vitro IC₅₀. Zanamivir concentrations in NF were comparable to those in ELF for oral inhalation. The zanamivir concentration in NF in this study (613 ng/ml) appeared to be approximately 5-fold higher than that in a previous study of single 10-mg doses of oral inhaled zanamivir using the same Diskhaler (14), where the median zanamivir concentration in NF was 122 (range, <0.5 to 212) ng/ml at 12 h. The reason for this discrepancy is uncertain, as the nasal sampling procedures used in these studies were similar and the same urea corrections were applied. The discrepancy may be due in part to the fact that two doses were administered in this study, in contrast to the previous single-dose study, or may reflect the variable nature of these measurements (14).

Following the continuous infusion of zanamivir at a rate of 3 mg/h (6-mg loading dose), the serum pharmacokinetics of zanamivir showed a slight biexponential disposition postinfusion and cross-cohort serum zanamivir concentrations at 2, 6, and 12 h were similar to the C₁₂ of 600 mg q12 i.v. administration, as expected. Despite similar serum zanamivir concentrations, ELF zanamivir concentrations were lowest when BAL fluid sampling was conducted at 2 h postdose, whereas those obtained with BAL sampling at 6 and 12 h postdose were higher, suggesting that distribution into the pulmonary compartment was incomplete 2 h after continuous infusion. Zanamivir concentrations in NF were generally similar to both ELF and serum zanamivir concentrations for intermittent i.v. administration and oral inhaled administration, but NF concentrations were below the limit of quantification for most samples following continuous infusion of zanamivir.

The pulmonary and nasal zanamivir concentrations across the regimens in this study suggest differences in pulmonary penetration, according to the i.v. administration schedule used. Pulmonary drug concentrations increased in proportion to the dose for intermittent i.v. administration of zanamivir, with similar ELF/serum concentration ratios across a 6-fold range of doses (55 to 79%). Excluding the BAL samples collected at 2 h, the median ELF/serum zanamivir concentration ratios (43 to 45%) for continuous infusion appeared lower than those obtained with intermittent administration, suggesting less pulmonary penetration during continuous infusion. However, given the longer duration of dosing prior to BAL sampling for the intermittent regimens (24 h versus 12 h), the pulmonary penetration following continuous infusion may have been higher with continued dosing for an additional 12 h.

Following the i.v. administration of zanamivir across the various regimens in this study, the median ELF zanamivir concentration was highest for the 600-mg regimen (419 ng/ml). The median ELF zanamivir concentration following 600 mg i.v. was lower than that obtained by oral inhaled administration when calculated using BAL1 (891 ng/ml) but not when the ELF zanamivir concentration was calculated using BAL2 (326 ng/ml). However, the high variability of calculated ELF zanamivir concentrations precluded definitive comparisons between regimens. Given the magnitude of the ELF concentration above the IC₅₀ for the intermittent i.v. infusion and oral inhaled regimens, the differences in absolute ELF concentrations may not be clinically significant for susceptible strains (Fig. 2). However, the clinical activity of i.v. zanamivir doses lower than 600 mg has not been published to date, although a 300-mg regimen is being evaluated in an ongoing phase III study (ClinicalTrials.gov identifier NCT01231620).

The overall safety profile of the i.v. zanamivir regimens used in this study was similar to the previously reported safety profile of zanamivir (2). Zanamivir administered as a continuous infusion had a safety profile similar to that observed with intermittent infusion. An ongoing phase II study is investigating the safety and tolerability of 600 mg administered twice daily by infusion over 30 min in hospitalized patients with influenza, including patients requiring intensive care and mechanical ventilation (ClinicalTrials.gov identifier NCT01014988). In addition, an ongoing phase III study will evaluate the efficacy, antiviral activity, and safety of i.v. zanamivir at 300 mg or 600 mg twice daily compared to those of oral oseltamivir at 75 mg twice daily in hospitalized subjects with laboratory-confirmed or suspected influenza virus infections (ClinicalTrials.gov identifier NCT01231620).

In conclusion, the data from this study demonstrate that zanamivir administered by intermittent i.v. infusion results in ELF zanamivir concentrations that are many-fold higher than the IC₅₀ for a range of influenza A and B viruses (12), including H5N1 (4, 16), pandemic H1N1 2009 (6), and currently circulating seasonal influenza virus strains. The demonstrated pulmonary penetration of zanamivir following i.v. administration in this study supports the zanamivir dosing under investigation for the treatment of hospitalized patients with influenza.