Avian Flu Diary: Nature Comms: Baloxavir outperforms oseltamivir, favipiravir, and amantadine in treating lethal influenza A(H5N1) HA clade 2.3.4.4b infection in mice

CDC Influenza Antiviral Medications: Summary for Clinicians - Jan 2026

#19,062

During any novel influenza pandemic - even one as long anticipated as H5N1 - it is expected that it will take 6 months or longer before any substantial quantity of vaccine would be available to the general public (see Maggie Fox's SCI AM - A Bird Flu Vaccine Might Come Too Late to Save Us from H5N1).

During that time - as with COVID - we will have to rely heavily on NPIs (Non-Pharmaceutical Interventions) like face masks, hand-washing, and social distancing.

Unlike with COVID, there is hope that one of the currently available influenza antivirals (i.e. oseltamivir, baloxavir, favipiravir, or amantadine) might provide significant treatment benefits.

Of these 4 options, oseltamivir (Tamiflu) is the only one stockpiled in any quantity.

Amantadine is no longer used, after it lost effectiveness in 2005
Favipiravir is primarily used in Japan, and is not FDA approved (although it could be an option under an EUA).
Baloxavir is in chronic short supply, and is many times more expensive than oseltamivir.

While none of these drugs are a panacea against influenza infection, all (except amantadine) are expected to reduce the severity and duration of infection; assuming they can be delivered to the patient in the first 24-48 hrs of infection.

Logistical problems of supply and delivery aside, not all antivirals are comparable in therapeutic impact. Each is subject to different resistance mutations, and some may work better against specific flu subtypes than others.

While the U.S. government doesn't release exact numbers, oseltamivir makes up well over 90% of all (SNS) stockpiled influenza antivirals, and is the CDC's first drug of choice for treatment of influenza A.

We have seen concerns raised over the effectiveness of oseltamivir against H5N1, including last March in St. Jude Researchers: Current Antivirals Likely Less Effective Against Severe Infection Caused by Bird Flu in Cows’ Milk; excerpt below:

Our evidence suggests that it is likely going to be hard to treat people severely infected with this bovine H5N1 bird flu strain,” said corresponding author Richard Webby, PhD, St. Jude Department of Host-Microbe Interactions. “Instead, reducing infection risk by not drinking raw milk and reducing dairy farm workers’ exposures, for example, may be the most effective interventions.”

“In general, baloxavir [Xofluza] caused a greater reduction in viral levels than oseltamivir [Tamiflu], but neither was always effective,” said first author Jeremy Jones, PhD, St. Jude Department of Host-Microbe Interactions.

There is also the matter of NAI antiviral resistance - which is relatively low right now - but appears to be increasing in both seasonal and novel flu strains (see Emerg. Microbes & Inf: Oseltamivir Resistant H5N1 (Genotype D1.1) found On 8 Canadian Poultry Farms).

Since randomized antiviral trials on humans are highly problematic, nearly all evidence of efficacy against H5N1 come from either observational studies and case reports, or from animal studies.

Last month, in Nature Comms: Oseltamivir and Baloxavir Monotherapy and Combination Therapy Efficacy Against Clade 2.3.4.4b A(H5N1) Influenza Virus Infection in Ferrets, we saw a CDC study which looked at both mono and combination therapy with oseltamivir and baloxavir in ferrets infected with H5N1 (genotype D1.1).

In short, they found:

Ferrets infected with H5N1 D1.1 and treated with oseltamivir saw little or no clinical or virologic benefit compared to no treatment, with persistent high fevers, weight loss, and systemic viral replication.
Ferrets treated with Baloxavir saw significantly less fever, weight loss, and viral replication. Some ferrets, however, saw a late rise in fever (after 4 days) and viral shedding, suggesting a viral rebound.
Ferrets treated with both drugs saw similar clinical protection to baloxavir alone, but did not show signs of rebound.

Note: Ferrets are a good, but not perfect, proxy for humans in influenza research, so these results may not be 100% applicable to humans.

Today we've a new study- again from Webby & Jones et al. - this time using highly susceptible female BALB/c mice; which suggests that baloxavir outperforms oseltamivir, favipiravir, and amantadine in treating clade 2.3.4.4b (circa 2022) H5N1 avian viruses (in mice).

I've only reproduced the abstract and conclusion (reformatted for readability), so those wanting more specifics will want to download and read the PDF file.

I'll have a bit more after the break.

Article Open access
Published: 19 February 2026
Baloxavir outperforms oseltamivir, favipiravir, and amantadine in treating lethal influenza A(H5N1) HA clade 2.3.4.4b infection in mice
Konstantin Andreev, Jeremy C. Jones, Ahmed Kandeil, Peter Vogel, Richard J. Webby & Elena A. Govorkova
Nature Communications , Article number: (2026) Cite this article

Abstract

Intercontinental spread of highly pathogenic avian influenza A(H5N1) viruses poses significant pandemic risks and necessitates strong protective countermeasures.

We evaluated the therapeutic efficacy of the neuraminidase inhibitor oseltamivir, the polymerase inhibitors baloxavir and favipiravir, and an ion-channel blocker amantadine, against severe influenza A(H5N1) virus infection in female BALB/c mice.
Baloxavir (≥10 mg/kg, 1 dose) fully protected mice from death, significantly reduced virus respiratory replication, and prevented neuroinvasion.
Oseltamivir (≥100 mg/kg/day for 5 days) provided limited survival benefits, reduced lung titers but failed to prevent viral neuroinvasion.
Favipiravir (≥100 mg/kg/day for 5 days) provided partial protection, although did not reduce viral titers in lungs and brain.
Amantadine provided no benefits.
Although all drugs inhibited A(H5N1) viruses in vitro, in vivo correlations did not extend beyond baloxavir. Our results indicate that baloxavir is the most reliable treatment to address both respiratory replication and systemic spread of contemporary A(H5N1) viruses in mice and should be considered in pandemic planning.

(SNIP)

In summary, our study demonstrates that among four classes of influenza antivirals with wide or limited approval, only the polymerase inhibitor baloxavir provides consistent and robust protection and reduction in viral titers after lethal challenge with A(H5N1) clade 2.3.4.4b in mice.
The polymerase inhibitor favipiravir provided partial protection against both disease criteria. In contrast, oseltamivir afforded limited survival benefits and did not control viral titers. These results may in part be due to targeting the early steps of the viral replication cycle and may be of significance for A(H5N1) viruses that have increased polymerase activity. For amantadine and oseltamivir, we found little to no correlation between their antiviral efficacy in mice and susceptibility in vitro, confirming previously published data44.
At present, we encourage prioritization of baloxavir or, potentially, baloxavir drug combinations as a potential frontline therapeutic approach for A(H5N1) clade 2.3.4.4b infections. We also suggest that more prognostic approaches are warranted to evaluate antivirals against emerging influenza viruses.

(Continue. . . )

Of note, this study used 4-year-old strains instead of more contemporary `Bovine' B3.13 or avian D1.1 genotypes, which together have caused dozens of human infections in the United States over the past 2 years.

Both of these 2022 H5N1 isolates, however, are "well-characterized" for producing 100% lethality and neurotropism in BALB/C mice; making differences in outcomes easier to identify.

Hopefully, B3.13, D1.1 - and any future genotypes that emerge - will undergo similar testing as time permits, as the results could differ.

Over the past year we've seen increased calls for a shift in pandemic antiviral strategy towards Baloxavir (or combination therapy), but most countries have invested heavily in oseltamivir, making any change likely slow in coming.

Oseltamivir is now a generic drug, manufactured by at least a dozen companies, while Baloxavir is made only by Roche/Shionogi plants. The global supply of oseltamivir is probably 50 times greater than baloxavir, while the cost is roughly 1/10th.

The reality is, even oseltamivir will be hard to get - at least during the first critical 24-48 hrs of infection - during a severe global influenza pandemic.

While I would be grateful for any antiviral (except, maybe Amantadine) in a flu pandemic, we will once again have to rely heavily on preventing infection; wearing face masks, hand washing, improved indoor ventilation, staying home while sick, and avoiding crowds.