Oseltamivir and baloxavir monotherapy and combination therapy efficacy against clade 2.3.4.4b A(H5N1) influenza virus infection in ferrets

Abstract

Neuraminidase inhibitors (NAIs) and cap-dependent endonuclease inhibitors (CENIs) represent two classes of antiviral drugs recommended for early treatment of patients with seasonal influenza A virus (IAV) infections. However, only limited human data, particularly on combination antiviral treatment, are available to inform optimal dosing regimens against novel IAVs, including highly pathogenic avian influenza A(H5N1) virus, associated with severe disease. Clade 2.3.4.4b A(H5N1) viruses have caused outbreaks in avian and mammalian species worldwide, highlighting the need to assess antiviral drug efficacy against these strains. We challenged ferrets with a D1.1 genotype A(H5N1) virus and treated infected animals with the NAI oseltamivir phosphate (OST) and the CENI baloxavir acid (BXA), alone or in combination, with treatment onset commencing pre- or post-symptom onset (24- or 48-hours post-inoculation (p.i.), respectively). When administered pre- or post-illness onset, BXA, but not OST, monotherapy provided significant reduction of clinical signs and significantly decreased infectious viral levels (in both respiratory and extrapulmonary specimens) compared with mock-treated animals. Combination OST/BXA treatment, when administered pre- or post-symptom onset, resulted in significant improvements in both metrics versus OST monotherapy. These data support continued investigation of antiviral treatment modalities that include both NAI and CENI for patients with mild and severe A(H5N1) disease.

Combination treatment with two antiviral drugs (oseltamivir and baloxavir) effectively limits clade 2.3.4.4b A(H5N1) influenza virus replication and disease severity in ferrets.

Introduction

Two classes of antiviral drugs, neuraminidase inhibitors (NAIs) and cap-dependent endonuclease inhibitors (CENIs), are approved by the U.S. Food and Drug Administration (FDA) and recommended for early treatment of uncomplicated seasonal influenza¹. The NAI oseltamivir (OST) is recommended for treatment of hospitalized patients with seasonal influenza or novel influenza A virus (IAV) infections². The emergence and worldwide spread of clade 2.3.4.4b highly pathogenic avian influenza (HPAI) A(H5N1) viruses, associated with 70 confirmed human cases in the United States since 2024, has highlighted the need to assess efficacy of antivirals against these viruses³. With few exceptions, clade 2.3.4.4b A(H5N1) viruses isolated from humans in the U.S. have maintained susceptibility to all FDA-approved antiviral drugs⁴; the majority of confirmed A(H5N1) human cases received OST treatment⁵, and experienced mild to moderate illness. However, clade 2.3.4.4b viruses can cause severe and fatal illness in humans^6–9, with only limited reports regarding combination antiviral treatment in critically ill patients⁶.

Due to a lack of randomized controlled trials to assess the clinical benefit of antiviral treatment of A(H5N1) virus infection¹⁰, preclinical animal models can serve as a critical surrogate to inform antiviral efficacy against novel IAV infections. Recent studies have shown efficacy of NAI and CENI monotherapy against clade 2.3.4.4b B3.13 genotype strains using the mouse model^11–13, but no studies to date have been reported with D1.1 genotype strains, which have also been associated with human infection, including critical illness^6,14,15. D1.1 oseltamivir-resistant HPAI A(H5N1) viruses have been detected among poultry in British Columbia¹⁶, underscoring the need to include this genotype in antiviral assessments. Studies assessing the pathogenicity of novel IAV in ferrets inform pandemic risk assessment rubrics^17,18, and antiviral studies conducted in ferrets can be especially valuable due to close physiologic links and recapitulation of clinical illness of influenza between humans and ferrets¹⁹. However, while efficacy of OST has been evaluated in ferrets inoculated with early (2004–2006) human and avian isolates of A(H5N1) virus^20–22, clade 2.3.4.4b A(H5N1) viruses from the ongoing U.S. outbreak in dairy cattle and poultry⁵ have to our knowledge not been similarly evaluated in this species. Furthermore, no studies have evaluated BXA treatment following A(H5N1) virus infection in ferrets; studies assessing combination OST and BXA administration in the context of A(H5N1) virus infection have also been limited to mice²³. For these reasons, we inoculated ferrets with A/Washington/239/2024 (WA/239), a D1.1 genotype clade 2.3.4.4b A(H5N1) virus, and assessed if subsequent treatment with OST and/or BXA could ameliorate clinical signs post-inoculation and reduce viral replication in respiratory and extrapulmonary specimens during the first five days of acute infection.

Results and discussion

With few exceptions, ferrets inoculated with clade 2.3.4.4b A(H5N1) viruses isolated from 2024 have manifested with fever, weight loss, and lethargy (decreased activity), often reaching humane endpoints during the first week post-inoculation^13,24–26. Similarly, mock-treated ferrets inoculated with WA/239 virus developed elevated and sustained fevers, measurable weight loss, and pronounced lethargy; 75% of animals met humane endpoints for euthanasia on day 5 p.i. (Table 1, Fig. 1A and B). To examine the capacity of antiviral drugs to mitigate these clinical signs, OST and BXA (administered as monotherapy or in combination) were administered to ferrets commencing 24 h (prior to onset of clinical signs in mock-treated animals) or at 48 h p.i. (after clinical signs were detected).

Table 1.

Clinical signs and frequency of viral detection in respiratory and gastrointestinal tract specimens among ferrets following antiviral treatment

Treatmenta	Onseta	Weight loss (%)b	Temp rise (°C)c	RIId	Nas Dise	AR shedding (log10 RNA copy #)f	Diare	RS (log10 PFU/ml)g	Int titer (log10 PFU/g)g	Lethality day 5 p.i.h
Mock	NA	7.2	2.2	1.7	2/4	5/6 (1.9 ± 0.5)	1/4	7/15 (2.9 ± 0.7)	4/4 (4.9 ± 0.6)	3/4*
OST	24	9.3	1.7	1.9	3/4	4/6 (1.7 ± 0.6)	2/4	10/15 (2.2 ± 0.9)	4/4 (3.2 ± 1.3)	4/4*
BXA	24	0	0.6	1.2	0/4	2/6 (3.3 ± 0.3)	0/4	0/16	0/4	0/4
OST/BXA	24	0	0.5	1.0	0/4	5/6 (2.4 ± 0.4)	0/4	1/16 (2)	0/4	0/4
OST	48	11.6	1.8	2.0	4/4	1/4 (1.5)	1/4	2/15 (2.9)	2/4 (3.3 ± 0.8)	4/4*
BXA	48	2.2	1.8	1.1	0/4	2/4 (2.8 ± 0.4)	0/4	3/16 (4.2)	1/4 (3.6)	0/4
OST/BXA	48	2.3	1.3	1.1	1/8	4/8 (2.5 ± 0.4)	0/8	1/31 (2.9)	0/8	1/8*

^aFerrets (n = 4 for all groups except 48 h OST/BXA which is n = 8) were administered OST monotherapy (twice-daily 5 mg/kg per os (by mouth)), BXA monotherapy (one 4 mg/kg subcutaneous injection), or combination OST and BXA treatment, commencing 24 h or 48 h p.i. with WA/239 A(H5N1) virus (intranasal inoculation with 100 PFU/1 ml). NA, mock-treated ferrets received per os and subcutaneous vehicle only on the same schedule as 24 h OST/BXA group.

^bPercentage mean maximum weight loss from baseline (934–1480 g) between days 1–5 p.i.

^cMean maximum temperature rise (in degrees °C) between days 1–5 p.i. over baseline (38.1–39.2 °C) between days 1–5 p.i.

^dRelative Inactivity Index (RII) calculated on daily lethargy scoring.

^eNumber of ferrets with detected nasal discharge or diarrhea between days 1–5 p.i./total number of ferrets.

^fNumber of samples with >1 viral RNA copies detected in the >4 µm fraction of a 2-stage NIOSH bioaerosol sampler days 2–4 p.i. (mock and 24-onset groups) or days 3–4 p.i. (48-onset groups)/total samples collected. Parentheses report the mean ± standard deviation of log₁₀ RNA copies per hour among all samples with positive detection. Samples were collected daily for one hour from cages holding two unanesthetized ferrets from the same group.

^gNumber of ferrets with infectious virus detected in rectal swabs (RS) collected daily days 2–5 p.i., or intestinal tissue (pooled duodenum, jejunum-ileum, and descending colon) collected the day of euthanasia (day 5 p.i. unless otherwise specified), at or above the limit of detection (10 PFU)/total samples collected. Parentheses report mean ± standard deviation of log₁₀ PFU/ml or g among all samples with positive detection.

^hNumber of ferrets out of total per group that reached humane endpoint criteria by day 5 p.i. *, one ferret in each group reached humane endpoints day 3 p.i. (48-OST/BXA) or day 4 p.i. (mock, 24-OST, 48-OST).

Fig. 1. Clinical signs and viral levels in A(H5N1) virus-infected ferrets following antiviral treatment.

Ferrets (n = 4 for all groups except 48 h OST/BXA which is n = 8) were inoculated with WA/239 virus, and antiviral treatment commenced 24 h (left panels) or 48 h (right panels) post-inoculation with the drug(s) specified. A, C Temperature change post-inoculation is shown as mean rise in °C over pre-inoculation temperature. Line color indicates group (maroon = mock, brown = OST, dark teal = BXA, light teal = OST/BXA). The same mock group was used to assess antiviral efficacy in both study arms. The red dashed vertical line indicates the day of first antiviral treatment. B, D Weight change post-inoculation is shown as percentage mean weight loss from pre-inoculation body weight. E, F Nasal washes (NW) were collected on each day indicated and are reported as log₁₀ PFU/ml (day 1 p.i. NW were only sporadically detected at titers <2 log₁₀ PFU/ml and are not shown). Dots represent individual ferrets sampled in each group. Boxes depict median with first and third quartile; whiskers extend from each hinge to the largest/smallest value no more than 1.5 of the interquartile range. G–J Tissues were collected day 5 p.i. from all ferrets unless otherwise indicated and are shown as circles; ferrets that met humane endpoints and were euthanized before day 5 p.i. are shown with a square (day 3 p.i., n = 1 ferret) or triangles (day 4 p.i., n = 3 ferrets); these 4 animals were excluded from the statistical analyses shown in (G–J). Viral titers are reported as log₁₀ PFU/ml (nasal turbinate, blood) or g (all others). The limit of detection (LOD) was 10 PFU; ferrets with no viral titer were assigned the LOD. *p < 0.05; **p < 0.01; ***p < 0.001; ****, p < 0.00001 by one-way ANOVA with Tukey post-hoc test (E, F) or by Kruskal–Wallis with Benjamini–Hochberg post-hoc test (G–J). Comparisons p > 0.05 are not shown. See Supplemental Table 1-3 for all supporting statistical evaluations.

OST monotherapy did not modulate fever onset (time of first rise over baseline), magnitude (maximum elevated temperature), or duration (calculated via area under the curve, AUC) relative to mock-treated animals, even when administered prior to first temperature elevation (Table 1, Fig. 1A). In contrast, BXA monotherapy significantly reduced fever detection relative to mock-treated ferrets when administered at 24 h p.i., supported by reduced mean maximum temperature (p < 0.0001, see Supplementary Table 1–3 for all statistical comparisons in this study, and Supplementary Tables 4–7 for supporting source data) and lower overall temperature elevation compared to mock-treated ferrets (p < 0.02). Combination OST/BXA treatment initiated at 24 h p.i. yielded comparable statistically significant differences as BXA monotherapy relative to mock. However, delaying administration of BXA monotherapy or OST/BXA combination treatment to 48 h p.i. (at which point fever was present in all animals) did not yield significant differences between groups using either metric (Fig. 1C).

Like fever, OST monotherapy did not modulate weight loss onset, magnitude, or duration relative to mock-treated animals when first administered 24 h p.i. (Table 1, Fig. 1B). In comparison to OST monotherapy, ferrets receiving BXA monotherapy or OST/BXA combination treatment initiated at 24 h p.i. were protected from weight loss (measured by maximum weight loss detected p < 0.02, or overall AUC of weight loss p < 0.05 for both treatments). When treatment initiation was delayed to 48 h p.i., mean maximum weight loss among ferrets receiving BXA monotherapy or OST/BXA combination treatment was significantly reduced compared to ferrets receiving OST monotherapy only (2.2% or 2.3%, respectively, vs. 11.6%; p < 0.02), but differences in overall AUC were not observed (Fig. 1D).

Ferrets that exhibit reduced activity levels will have a higher Relative Inactivity Index (RII) compared to ferrets that remain alert and playful throughout the observation period. The highest RII was recorded for mock-treated and OST monotherapy-treated ferrets, with no significant differences in mean values between these groups (Table 1). In contrast, BXA (either monotherapy or in combination with OST, initiated either 24 or 48 h p.i.) significantly reduced the overall mean RII versus OST-monotherapy treated ferrets (p < 0.05) (Table 1). Lethargy scores are a critical component in humane endpoint criteria; in agreement, by day 5 p.i., 100% of ferrets receiving OST monotherapy first initiated at either 24 or 48 h p.i. had met endpoint criteria, in contrast to none of the ferrets receiving BXA monotherapy (Table 1). Collectively, these data support that BXA monotherapy given prior to or after illness onset protected ferrets from severe disease compared with OST monotherapy; the addition of BXA to OST treatment when delayed until after onset of clinical signs still offered protection relative to OST monotherapy, though reduced compared to combination treatment initiated prior to onset of clinical signs.

We next examined modulation of viral titers detected in nasal wash (NW) specimens collected daily p.i. Despite a low initial challenge dose, WA/239 virus was detected at high (>10⁴ PFU/ml) mean titers in ferret NW specimens on days 2–5 p.i. from mock-treated animals (Fig. 1E), with a peak mean NW titer of 4.5 ± 0.6 log₁₀ PFU/ml. NW viral titers from ferrets receiving OST monotherapy initiated at 24 h p.i. had comparable peak NW titers (4.5 ± 1.2 log₁₀ PFU/ml) and overall similar viral shedding as measured by AUC as mock-treated ferrets (p > 0.05), whereas BXA monotherapy initiated at 24 h p.i. significantly reduced viral titers in NW compared with mock-treated animals (either peak titer (2.6 ± 0.8 log₁₀ PFU/ml) or overall shedding measured by AUC; p < 0.05). Unlike clinical signs of infection, where combination OST/BXA treatment yielded comparable outcomes to BXA monotherapy, combination OST/BXA treatment initiated 24 h p.i. resulted in reduced NW viral shedding compared with either OST or BXA monotherapy as measured by peak titer (1.4 ± 0.5 log₁₀ PFU/ml). When treatment was initiated at 48 h p.i., (1–3 h prior to the day 2 p.i. NW collection, such that antiviral efficacy is most rigorously assessed days 3–5 p.i. in this specimen), peak NW titers were reduced 10-fold in ferrets receiving OST monotherapy (3.6 ± 1.2 log₁₀ PFU/ml) or >100-fold in ferrets receiving BXA monotherapy (1.8 ± 1.1 log₁₀ PFU/ml) or OST/BXA combination therapy (1.8 ± 0.9 log₁₀ PFU/ml) relative to mock (Fig. 1F). BXA monotherapy resulted in significantly reduced peak NW titers compared to OST monotherapy (p < 0.05) when treatment was initiated either 24 or 48 h p.i.

Viral loads in NW specimens have been linked with the magnitude of influenza A viral RNA in expelled air of inoculated ferrets²⁷, but quantification of expelled air from antiviral-treated animals to our knowledge has not been previously performed. We used a size-fractionating bioaerosol sampler to collect airborne particles from pair-housed ferrets, with daily collections starting from the day of first antiviral treatment (day 1 or 2 p.i. for 24 h or 48 h treatment groups, respectively) through day 4 p.i. On day 1 p.i., immediately prior to first antiviral treatment in the 24 h treatment groups, viral RNA was detected from all cages (3.0 ± 1.2 log₁₀ viral RNA copies/h in particles >4 µm). Despite significant reductions in infectious virus detected in NW specimens among ferrets treated with BXA monotherapy or in combination with OST, viral RNA was still detected in cages holding animals from all treatment groups, with frequency and magnitude of detection independent of specific antiviral treatment administered or timing of treatment onset (Table 1). However, mean viral RNA levels in all groups were generally low (2.2 ± 0.6 log₁₀ viral RNA copies/h among all samples collected post-antiviral dosing onset), likely reflective of the low initial viral challenge dose, low infectious viral loads in NW specimens present among many of the sampled ferrets, and other considerations.

To assess viral loads in discrete sites throughout the ferret respiratory tract, all ferrets were humanely euthanized by day 5 p.i. for collection of systemic tissues, with few exceptions when ferrets met endpoint criteria days 3 or 4 p.i. (see Table 1). High viral titers (>5 log₁₀ PFU/ml or g) were detected in nasal turbinate (NT), trachea (Tr), and lung (Lg) specimens in mock-treated animals (Fig. 1G and H). Among tissue specimens collected day 5 p.i., ferrets receiving OST monotherapy (initiated at 24 or 48 h p.i.) had no significant difference in viral titers in either NT, Tr, or Lg specimens compared to mock. In contrast, BXA monotherapy initiated at 24 or 48 h p.i. (with or without combination OST treatment) resulted in low to undetectable viral loads in all respiratory tract tissues examined (Fig. 1G and H).

Due to reports of gastrointestinal (GI) symptoms in some confirmed human A(H5N1) cases^5,28 and frequent detection of diarrhea among A(H5N1) virus-infected ferrets²⁴, we next examined the capacity of antiviral treatments to mitigate viral replication in the GI tract. WA/239 virus was detected in both rectal swabs and intestinal tissue among all mock-treated ferrets, with comparable frequency and magnitude among ferrets receiving OST monotherapy initiated 24 h p.i. (Table 1). BXA monotherapy administered at either 24 or 48 h p.i. (with or without concurrent OST administration) significantly reduced viral load in intestinal tissue versus mock-treated ferrets (p < 0.03); diarrhea was not observed in ferrets receiving BXA treatment (alone or in combination with OST) but was detected in 25–50% of ferrets receiving mock- or OST-treatment (Table 1).

Clade 2.3.4.4b A(H5N1) viruses are capable of viremia and systemic spread in ferrets²⁴. By day 5 p.i., 4/4 mock-treated ferrets had infectious virus recovered from blood, reaching mean titers of 2.9 ± 1.7 log₁₀ PFU/ml. OST monotherapy reduced the frequency of ferrets with detectable virus detected in blood (2/4 ferrets following either 24 or 48 h p.i. treatment onset); BXA monotherapy or OST/BXA combination therapy abrogated viremia detection in all ferrets tested at day 5 p.i. (Fig. 1I-J). While mean titers in olfactory bulb and brain tissues were generally reduced in ferrets receiving BXA or combination OST/BXA-treated ferrets versus mock-treated ferrets with few exceptions, these differences did not reach statistical significance (Fig. 1I and J).

Systemic spread to other tissues, such as the liver, spleen, and kidney, have important consequences for overall mammalian health during HPAI A(H5N1) virus infection, but these tissues are rarely examined during antiviral studies in vivo. Mock-treated ferrets had a high frequency of viral detection in all three tissues, with mean titers >6.5 (liver) and >3.5 (spleen, kidney) log₁₀ PFU/g. OST monotherapy resulted in generally comparable mean titers as mock-treated ferrets (Fig. 1I and J). However, BXA monotherapy initiated 24 or 48 h p.i., (with or without concurrent OST treatment), resulted in an absence of infectious virus detection in all three tissues among ferrets tested day 5 p.i. (Supplementary Table 7).These data support that systemic spread of A(H5N1) virus during infection (through day 5 p.i.) is significantly reduced in ferrets receiving BXA (with or without concurrent OST treatment) as compared to mock-treated or OST monotherapy-treated ferrets, even when treatment administration is delayed until after emergence of clinical signs.

Emergence of oseltamivir-resistance during or after OST treatment has been reported in A(H5N1) patients with fatal outcomes²⁹. Emergence of viruses with amino acid substitutions associated with antiviral resistance is an infrequent but documented event during A(H5N1) in vivo efficacy studies^21,30. The WA/239 virus used in this study displayed similar sensitivity to OST and BXA as other viruses from the 2.3.4.4b A(H5N1) outbreak in the United States. This was shown by in vitro phenotypic testing^4,31, and by the lack of substitutions in the NA (e.g., E119D, H275Y)^32,33 and PA cap-dependent endonuclease (CEN) domain (e.g., E23K, I38T/M/F/N/S/L…)^34,35, which are known or suspected to reduce susceptibility to either antiviral. Notably, NW, NT, and Lg specimens collected from ferrets days 3–5 p.i. with sufficiently high viral titers were sequenced, and did not contain any substitutions in the vicinity of the NA active site or the PA CEN domain, supporting that antiviral treatment under the conditions of this study was not associated with emergence of viruses with reduced antiviral susceptibility, though the low viral titers recovered from some specimens precluded robust sequence reads from all ferrets. Overall, sequence analyses of contemporary clade 2.3.4.4b A(H5N1) viruses (including both B3.13 and D1.1 genotypes) support a low percentage of virus isolates with amino acid mutations associated with reduced NAI and CENI susceptibility^3,16, highlighting the need for continued surveillance and evaluation of antiviral drug susceptibility of A(H5N1) viruses.

This study assessed antiviral treatment initiated 24 or 48 h p.i., using drug concentrations and administration routes that represent human standard-of-care equivalents for seasonal IAV infection and comparable pharmacokinetics between ferrets and humans at administered doses^36,37. Collectively, studies conducted in vivo support that OST and BXA efficacy is both time- and dose-dependent, with earlier administration and higher concentrations of drugs associated with reduced detection of clinical signs p.i. and decreased viral replication in A(H5N1) virus infected animals¹⁰. While BXA monotherapy was associated with significantly improved health outcomes in A(H5N1) virus-infected ferrets compared with OST monotherapy, by day 5 p.i., some ferrets receiving BXA monotherapy (with treatment onset at either 24 or 48 h p.i.) did display elevated temperatures and/or viral levels in NW specimens relative to day 4 p.i. levels (Fig. 1A, C, E and F), suggesting a potential viral rebound in these animals. Animals receiving combination treatment with both OST and BXA (with treatment onset at either 24 or 48 h p.i.) did not present with increased temperature or viral levels in NW specimens by day 5 p.i., supporting increased efficacy of combination treatment in controlling the infection at this time relative to BXA monotherapy, possibly attributable to repeated OST dosing facilitating increased control of low viral levels in the upper respiratory tract. Future studies assessing higher doses of OST, and repeated administration of BXA, would provide greater insight into potential dosing of antiviral treatment for A(H5N1) patients in clinical settings.

This study had several limitations. Only one representative antiviral drug for each class of NAI and CENI was evaluated. All animals were humanely euthanized at day 5 p.i. to assess systemic spread of virus; while 15/16 mock- or OST monotherapy-treated ferrets had reached humane endpoints by this time, we did not include additional ferrets to assess survival among ferrets receiving BXA or combination treatment. While systemic tissues were collected for quantification of infectious virus, histopathology was not performed; delayed and reduced pathological changes in respiratory tract tissues has been shown previously in ferrets inoculated with a low dose of IAV receiving antiviral treatment³⁸, warranting further histopathological investigations in A(H5N1) virus-infected ferrets receiving combination treatment with OST/BXA. BXA monotherapy has been shown to reduce airborne virus transmission in ferrets following seasonal IAV infection³⁶; while selected clade 2.3.4.4b viruses have shown a capacity to transmit via multiple modes in ferrets^13,24,39, transmission assessments were not performed. However, detection and quantification of viral RNA in airborne particles shed from inoculated, drug-treated ferrets (Table 1) supports the need for future investigations with both seasonal and novel IAV strains. To align with recommendations for post-exposure prophylaxis for persons at increased risk for exposures to A(H5)-virus infected animals⁴⁰, prophylactic (pre-IAV challenge) antiviral administration was not performed.

There is a paucity of studies evaluating BXA in combination with NAIs to mitigate clinical signs and viral levels following IAV inoculation compared with monotherapy in animal models, with even fewer studies using novel IAVs. Our study found when administered at 24 or 48 h p.i., combination OST/BXA treatment offered generally minimal improvements in reduction of clinical signs and viral levels relative to BXA monotherapy, but substantial improvements in both metrics relative to OST monotherapy. Current recommendations specify prompt oseltamivir treatment of patients with A(H5N1)², and consideration of combination antiviral therapy for immunocompromised outpatients and hospitalized patients. Our data support continued investigation of BXA monotherapy and combination OST/BXA therapy in patients with A(H5N1). Further investigations of antiviral treatment modalities are needed, including combination NAI and CENI, in vivo and in patients with mild and severe illness due to A(H5N1) virus infection, due to clade 2.3.4.4b and other virus clades.

Methods

Virus

A/Washington/239/2024 (WA/239) virus was propagated in MDCK cells as previously described⁴¹. All experiments were performed at BSL3 containment, including enhancements, as required by the U.S. Department of Agriculture and the Federal Select Agent Program⁴².

Ferret inoculation and specimen collection

All animal experiments and procedures were approved by the Institutional Care and Use Committee (IACUC) of the Centers for Disease Control Prevention in an AAALAC International-accredited animal facility; we have complied with all relevant ethical regulations for animal use. 7 to 12-month-old male Fitch ferrets (Mustela putorius furo, Triple F Farms), were serologically negative to currently circulating influenza A and B viruses (as confirmed by standard hemagglutination inhibition assay) at the time of use. All animals were housed in Duo-Flo Bioclean mobile units (Lab Products Incorporated) during experimentation. Prior to inoculation, a subcutaneous temperature transponder (IPTT-300, BMDS) was inserted into the dorsal space between the scapulae of each animal. 32 ferrets were inoculated with 100 PFU of WA/239 virus by the intranasal route (1 ml total volume in PBS). Ferrets were anesthetized for inoculation and all subsequent samplings with a ketamine:xylazine cocktail given intramuscularly.

All ferrets were observed daily or twice-daily post-inoculation for clinical signs of infection (including weight loss, temperature, lethargy assessments). Lethargy was scored as previously described to calculate a Relative Inactivity Index^43,44. Animals that exhibited sustained diarrhea, lethargy, and/or labored breathing in addition to other signs of severe disease were humanely euthanized. Nasal washes and rectal swabs were collected from all inoculated animals daily as previously described⁴⁵. Specimens were frozen at −80 °C until titration by standard plaque assay⁴¹. All animals were humanely euthanized on day 5 p.i. (earlier if humane endpoint criteria were met) by intracardiac administration of 1 mg/kg, 390 mg pentobarbital sodium and 50 mg phenytoin sodium per 100 ml (Euthanasia Solution, Med-Pharmex) for postmortem necropsy, where blood, kidney, spleen, liver, intestine (pooled from duodenum, jejuno-ileum, and descending colon), olfactory bulb, brain, lung (pooled from all lobes proximal to bronchi), trachea, and nasal turbinates were collected and titrated for infectious virus. Nasal wash, nasal turbinate, and lung specimens >2 log₁₀ PFU/ml or g were inactivated in AVL buffer (Qiagen), and frozen at −80 °C before RNA extraction and quantification.

Antiviral administration

Oseltamivir phosphate (OST, MedChemExpress) was dissolved in sterile sugar syrup (15% fructose in water) to a concentration of 5 mg/kg and administered by syringe per os (by mouth) to unsedated ferrets (1 ml total volume). Ferrets received OST twice-daily starting 24 or 48 h p.i. for the duration of the experiment (10–12 h apart). Mock-treated ferrets received sterile sugar syrup only on the same schedule. Baloxavir acid (BXA, MedChemExpress) was dissolved in 0.5% (wt/vol) methylcellulose to a concentration of 2 mg/ml in sterile water. Anesthetized ferrets received four subcutaneous injections (1 mg/kg per site) of BXA on the dorsal region for a total dose of 4 mg/kg per ferret as previously described³⁶, administered either 24 or 48 h p.i. Mock-treated ferrets received subcutaneous injections with 0.5% methylcellulose at 24 h p.i. only.

Aerosol collection

Aerosol samples were collected daily from unsedated ferrets starting the day of first antiviral administration (24 or 48 h p.i.) through day 4 p.i., using a NIOSH BC 251 two-stage cyclone aerosol sampler as previously described²⁷. Briefly, air was sampled from the cages of pair-housed ferrets for 1 h at 3.5 L/min, with collected aerosols size-fractionated into >4 µm, 1–4 µm, and <1 µm stages. Viral material from all stages was immediately collected in PBS, inactivated in AVL buffer (Qiagen), and frozen at −80 °C before RNA extraction and quantification. Aerosol samplers were decontaminated with 70% ethanol after each use, flushed with water, and allowed to air dry.

RNA quantification and sequencing

RNA from all inactivated samples was extracted using the QIAamp 96 viral RNA mini-extraction kit with a 100 µl elution volume. All aerosol samples were tested in duplicate by real-time RT-PCR as previously described²⁴, using universal primers described previously⁴⁶. Mean viral RNA copy numbers were normalized and expressed as RNA copy number per hour; specimens with gene copy numbers of <1 were declared negative. All ferret nasal wash and tissue specimens were subjected to next generation sequencing as previously described⁴⁷ on an Illumina iSeq100 (Illumina) using universal primers. Variants were called at 5% frequency using a minimum coverage of 100x and excluding paired reads mapped more than 30% from the expected distance of 500 bp.

Statistics, data analysis, and reproducibility

All analyses were conducted in R v.4.4.0 using the package tidyverse v2.0.0⁴⁸. Figures were generated using ggplot2 v3.5.2⁴⁹, wesanderson v0.3.7⁵⁰, ggpubr v0.6.0⁵¹, and patchwork v1.3.0⁵². All analyses were conducted with the log₁₀ of the measured viral titer. Area under the curve (AUC) was calculated as previously described⁵³ starting at the time of antiviral treatment onset. Multivariate comparisons (either Welch’s ANOVA with Games-Howell post-test, one-way ANOVA with Tukey’s HSD post-hoc test, or Kruskal-Wallis with Benjamini-Hochberg post-hoc test, as specified in Supplementary Table 1) were performed using base R stats or the packages PMCMRplus⁵⁴ or FSA⁵⁵. Sample size was n = 4 ferrets for all groups except 48 h OST/BXA which is n = 8.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

J.A.B. designed the experiment (with input from T.M.U., L.V.G., and T.R.M.), performed ferret experiments, processed samples, analyzed data (using code written by T.J.K.), and wrote the first draft of the manuscript. J.A.B., N.B., X.S., J.P.P., C.P., and T.R.M. performed ferret experiments and processed ferret samples. H.Z. assisted with plaque assays. T.J.K performed sequence analysis. T.R.M supervised the work. All authors contributed to, reviewed, and approved the final manuscript.

Peer review

Peer review information

Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editor: Johannes Stortz.

Data availability

The manuscript includes no previously published data, and all data is reported in the manuscript or supplemental files; clinical and virological data analyzed for the study are provided in the Supplemental Tables 4-7. Sequence data have been deposited to the NCBI Sequence Read Archive (BioProject identification number PRJNA1339024).

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-026-09607-w.