Autoantibodies neutralizing type I IFNs in a fatal case of H5N1 avian influenza

Influenza A viruses (IAVs) triggered the 1918 Spanish flu (H1N1), 1957 Asian flu (H2N2), 1968 Hong Kong flu (H3N2), 1977 Russian flu (H1N1), and 2009 swine flu (H1N1) epidemics and pandemics. Zoonotic IAVs, especially H5N1 and H7N9, are among the greatest current threats to public health due to their pandemic potential (WHO, 2025; CDC, 2025). H5N1 IAV causes lethal infections in poultry, leading to important economic losses, but can also infect various mammals, including humans, causing severe disease. Since 1997, 985 cases of H5N1 infection in humans have been recorded in 25 countries by the WHO, with a fatality rate of 48% (case-fatality ratio) (WHO, 2025). From 2013 to 2019, 1,500 human cases of H7N9 infection were reported, all in China, with a fatality rate of 40% (Chen et al., 2021). Data for estimating the true infection-fatality rate (IFR) of H5N1 and H7N9 IAV infections are scarce, but this rate is probably below WHO estimates (WHO, 2025) because the rate of seroconversion in exposed individuals is only 1–2%, and even if infected, many mild cases remain undiagnosed (Wang et al., 2012). Indeed, recent studies on human H5N1 infections showed that all cases were asymptomatic or mild (Garg et al., 2025; Shimizu et al., 2016). Similar findings were reported for humans infected with H7N9 (Chen et al., 2014). So far, no human-to-human transmission of either avian IAV has been documented. However, the recent introduction of H5N1 viruses into cows resulted in several H5N1 infections in dairy workers, probably through exposure to infectious material from infected cow’s milk (Garg et al., 2025). As with other infectious agents, there is immense interindividual variability in humans exposed to H5N1 or H7N9, ranging from silent infection to lethal disease, and the IFR is not precisely known (Casanova and Abel, 2022).

Protection against avian IAVs may be mediated by cross-reactive adaptive immunity to seasonal IAVs, or by intrinsic or innate immunity. Since 2015, we and others have reported that rare monogenic inborn errors of type I IFN immunity can underlie life-threatening seasonal IAV infections in otherwise healthy individuals (Ciancanelli et al., 2015; Zhang et al., 2022a; Hernandez et al., 2018; Lim et al., 2019). Moreover, variants of MX1, an IFN-stimulated gene (ISG) that is functional in humans but not birds, were identified in Chinese individuals who developed severe H7N9 infection (Chen et al., 2021). Since 2020, we have also shown that autoantibodies neutralizing type I IFNs (AAN-I-IFNs) can underlie 5% of severe seasonal influenza pneumonia (Zhang et al., 2022b), 10% of tick-borne encephalitis (Gervais et al., 2024b), 15% of hypoxemic COVID-19 pneumonia (Bastard et al., 2020, 2021a; Stertz and Hale, 2021), 20% of severe Middle East respiratory syndrome pneumonia cases (Alotaibi et al., 2023), 30% of severe adverse reactions to live-attenuated yellow fever virus vaccine (YFV-17D) (Bastard et al., 2021b), 40% of West Nile virus encephalitis cases (Gervais et al., 2023), and most cases of the rarer Usutu or Powassan virus encephalitis (Gervais et al., 2024a). Unlike most known inborn errors of type I IFN immunity, AAN-I-IFNs are common in the general population, with a prevalence of about 0.5% under the age of 70 years and about 5% over the age of 70 years (Bastard et al., 2021a). Samples from most patients (1:10 diluted plasma or serum) neutralize 0.1–10 ng/ml IFN-α and/or -ω, resulting in a very high odds ratio (OR: 100–500) for severe disease relative to controls with mild/asymptomatic infection (Bastard et al., 2024). We thus tested the hypothesis that AAN-I-IFN can underlie severe H5N1 infection.

On December 9, 2024, a 71-year-old man of European descent with a 1-week history of dyspnea, fever, and confusion presented at a rural hospital in Jennings, LA. On admission, the patient had acute hypoxic respiratory failure. Chest X ray revealed multifocal pneumonia involving the right upper and lower lobes. The patient was intubated and transferred to the ICU on day 1 of hospitalization (day 1). On day 2, IAV was detected in swabs (influenza A/Louisiana/12/2024 H5N1 virus, D1.1 genotype). Oseltamivir was initiated on day 2, and baloxavir was added on day 3. The patient kept domesticated chickens and ducks, which had died about 5–7 days earlier. He was transferred to the Academic Medical Center of Ochsner-LSU Health Shreveport in Shreveport, LA. On day 3, veno-venous extracorporeal membrane oxygenation was initiated, and therapeutic plasma exchange (TPE) was performed due to concerns about thrombotic microangiopathy and viremia. Continuous renal replacement therapy (CRRT) was performed from days 4 to 13 and days 20 to 28, and an ExThera Seraph 100 Microbind Affinity Blood Filter was used (days 4 to 8). The course of the infection was complicated by atrial fibrillation with rapid ventricular response and gastrointestinal bleeding. Tracheostomy was performed on day 14. On day 25, the patient developed profound hypoxia and hypercapnia, with anuria and a return of vasoplegic shock, which became refractory to treatment. On day 29, a transition to comfort care was decided. The patient died shortly thereafter, surrounded by family. During hospitalization, the patient was tested positive for cold agglutinin, anti-Kna (Knops), and anti-JKB antibodies. He had no clinical history of severe infectious diseases or autoimmunity and no recorded influenza or COVID-19 vaccination.

We first tested the blood sample collected from the patient on day 3, before TPE/CRRT, for AAN-I-IFN neutralizing IFN-α, -β, and -ω in a luciferase reporter assay. The patient’s blood (diluted 1:20) neutralized 10 ng/ml IFN-α and 100 pg/ml IFN-ω, but not IFN-β (Fig. 1 A). It also neutralized all 12 subtypes of IFN-α tested (Fig. S1 A). ELISA and multiplex assays revealed high levels of IgG-binding IFN-α, intermediate levels of IgG binding IFN-ω, and no detectable IgG-binding IFN-β, consistent with neutralization assay results (Fig. 1, B and C). High IgG levels binding IL-17F, IL-22, possibly CXCL1, IL-5, IL-9, TSLP, and TWEAK were detected, but no IgG binding the other 36 cytokines tested (Fig. S1 B). We then tested the patient’s blood samples 18 h, 2 days, and 3 days after TPE treatment. Anti–IFN-α titer decreased but was still detectable 18 h after TPE. (Fig. 1 D). Serological tests were performed for IAVs. The patient’s blood samples were positive for A/Hong Kong/1/1968 by hemagglutination inhibition (HAI) assay, indicating exposure to seasonal H3N2 (Fig. 2 A). Finally, in the presence of blood from the patient, IFN-α failed to inhibit IAV replication in A549 cells, indicating that the AAN-I-IFN blocked the antiviral function of IFN-α (Fig. 2 B).

We report a fatal case of H5N1 pneumonia in a 71-year-old man with blood AAN-I-IFN neutralizing 10 ng/ml IFN-α and 100 pg/ml IFN-ω. In the general population, autoantibodies neutralizing these concentrations of IFN-α and IFN-ω are found in only 0.5% of individuals aged between 65 and 75 years and 1% of individuals aged >70 years (Bastard et al., 2021a). We previously showed that individuals with AAN-I-IFN have a very high risk of critical seasonal IAV pneumonia (OR: 10–100) (Zhang et al., 2022b). Causality between AAN-I-IFN and fatal H5N1 infection in this patient is therefore plausible. The patient was seropositive for seasonal H3N2, indicating that he had controlled seasonal IAV infections without vaccination. His AAN-I-IFN probably emerged when he was already seropositive for seasonal H3N2 (Fernbach et al., 2024; Bastard et al., 2023). The patient had no clinical history of severe infection or autoimmune disease, like many patients with AAN-I-IFN in this age group (Bastard et al., 2024). Similarly, like most individuals with AAN-I-IFN, he had autoantibodies neutralizing IFN-α and/or -ω, but not -β, and might therefore have benefited from IFN-β therapy, if administered early in infection and in combination with antiviral therapies.

We and others have shown that AAN-I-IFN underlie a growing range of severe infections of emerging and circulating viruses (Bastard et al., 2024; Hale, 2023; Busnadiego et al., 2022). This fatal case of H5N1 infection has broad clinical and biological implications. It suggests that type I IFNs may contribute to innate and intrinsic immunity to emerging viruses, including zoonotic viruses, such as avian IAV. AAN-I-IFNs are common in the general population and therefore constitute a substantial threat to public health (Bastard et al., 2021a). We also recently identified a dominant-negative mutation in IFNAR1 (p.Pro335del), encoding the type I IFN receptor, which is remarkably common in South China (0.6–2%). This mutation impairs the response to IFN-α and -ω, but not -β, as in patients with AAN-I-IFN. Heterozygous carriers of this mutation are vulnerable to infections with various viruses, including SARS-CoV-2 (Al Qureshah et al., 2025). They may also be vulnerable to H5N1, H7N9, or other avian IAVs. New cases of H5N1 infections are increasing in the recent years (Siegers et al., 2025). Since October 2024, four other hospitalized cases of H5N1 infection have been documented in North America (Jassem et al., 2025; CDC, 2025), some of which might perhaps have been caused by inborn errors of type I IFN immunity or AAN-I-IFN (Langlois and Casanova, 2025).

The initial replication of avian IAV in humans is suboptimal due to the lack of mammal-adaptive mutations of the viral genes. However, crippled type I IFN immunity, due to inborn errors or AAN-I-IFN, may boost replication sufficiently for an avian IAV to generate mutations facilitating adaptation to mammals (Langlois and Casanova, 2025; Spieler et al., 2020). Sequencing of the influenza A/Louisiana/12/2024 H5N1 virus responsible for this case revealed low-frequency mutations in the HA segment (mixed amino acid populations at positions A134V, N182K, and E186D) associated with an increase in binding to human-type cell receptors (CDC, 2024). Previous reports of MX1 variants in patients with severe H7N9 infections suggest that variants of a single ISG may facilitate cross-species transmission (Chen et al., 2021; Hale et al., 2010; Casanova and Abel, 2024). However, the very rare human MX1 variants are unlikely to facilitate human-to-human transmission, as the virus remains MX1 sensitive and cannot infect the predominantly MX1 wildtype population. By contrast, the common presence of AAN-I-IFN and IFNAR1 p.Pro335del may enable the virus to undergo selection for more efficient human-to-human transmission if propagated in a cluster of affected individuals, paving the way for zoonotic virus pandemics.

Written informed consent was obtained in LSU Health Shreveport in accordance with the approval of the institute review board (protocol number 2899). AAN-I-IFN neutralization (Groen et al., 2024, 2025), ELISA (Gervais et al., 2024a), multiplex assays (Bastard et al., 2020), IAV infection assay (Zhang et al., 2022b), and HAI assays (Aydillo et al., 2021), were performed as previously described.

Fig. S1 provides more information on the autoantibodies neutralizing different type I IFN subtypes (A) and cytokines (B).

This study was supported in part by the St. Giles Foundation; the Rockefeller University; Howard Hughes Medical Institute; Institut National de la Santé et de la Recherche Médicale (INSERM); the Imagine Institute; Paris Cité University; the National Center for Research Resources; the National Center for Advancing Sciences of the National Institutes of Health (NIH) (UL1TR001866); NIH (R01AI163029); American Lung Association (COVID-1026207); the Stavros Niarchos Foundation (SNF) as part of its grant to the SNF Institute for Global Infectious Disease Research at The Rockefeller University; the Square Foundation; Grandir - Fonds de solidarité pour l’enfance; the Fondation du Souffle; the SCOR Corporate Foundation for Science; Battersea and Bowery Advisory Group; the French National Research Agency (ANR) under the France 2030 program (ANR-10-IAHU-01); the Integrative Biology of Emerging Infectious Diseases Laboratory of Excellence (ANR-10-LABX-62-IBEID); ANR GENVIR (ANR-20-CE93-003); ANR AAILC (ANR-21-LIBA-0002); ANR AI2D (ANR-22-CE15-0046); ANR GENFLU (ANR-22-CE92-0004); the ANR-RHU COVIFERON program (ANR-21-RHUS-08); the HORIZON-HLTH-2021-DISEASE-04 program under grant agreement 101057100 (UNDINE); the European Union’s Horizon 2020 research and innovation program under grant agreement No. 824110 (EASI-genomics); W. E. Ford, General Atlantic’s Chairman and Chief Executive Officer, G. Caillaux, General Atlantic’s Co-President, Managing Director and Head of business in EMEA, and the General Atlantic Foundation; the French Ministry of Higher Education, Research, and Innovation (MESRI-COVID-19); REACTing-INSERM. Jean-Laurent Casanova is an inventor on patent application PCT/US2021/042741, filed July 22, 2021, submitted by The Rockefeller University and covering the diagnosis of susceptibility to, and the treatment of, viral disease, and viral vaccines, including COVID-19 and vaccine-associated diseases. Paul Bastard was supported by the French Foundation for Medical Research (EA20170638020), the MD-PhD program of the Imagine Institute (with the support of the Fondation Bettencourt-Schueller), and a “Poste CCA-INSERM-Bettencourt” (with the support of the Fondation Bettencourt-Schueller). This study was also partly funded by Center for Research on Influenza Pathogenesis and Transmission, a NIAID funded Center of Excellence for Influenza Research and Response (contract # 75N93021C00014) to Adolfo García-Sastre. The Adolfo García-Sastre laboratory has received research support from Avimex, Dynavax, Pharmamar, 7Hills Pharma, ImmunityBio, and Accurius, outside of the reported work. Adolfo García-Sastre has consulting agreements for the following companies involving cash and/or stock: CastleVax, Amovir, Vivaldi Biosciences, ContraFect, 7Hills Pharma, Avimex, Pagoda, Accurius, Esperovax, Applied Biological Laboratories, Pharmamar, CureLab Oncology, CureLab Veterinary, Synairgen, Paratus, Pfizer, Virofend, and Prosetta, outside of the reported work. Adolfo García-Sastre has been an invited speaker in meeting events organized by Seqirus, Janssen, Abbott, AstraZeneca, and Novavax. Adolfo García-Sastre is inventor on patents and patent applications on the use of antivirals and vaccines for the treatment and prevention of virus infections and cancer, owned by the Icahn School of Medicine at Mount Sinai, New York, outside of the reported work.

Author contributions: Qian Zhang: conceptualization, data curation, funding acquisition, investigation, methodology, project administration, resources, supervision, validation, visualization, and writing—original draft, review, and editing. Taylor S. Conrad: investigation, resources, and writing—review and editing. Marcela Moncada-Velez: formal analysis, investigation, methodology, validation, and writing—review and editing. Kaijun Jiang: investigation and writing—review and editing. Anastasija Cupic: formal analysis and investigation. Jonathan Eaton: conceptualization and investigation. Kimberley Hutchinson: investigation. Adrian Gervais: conceptualization, investigation, validation, visualization, and writing—original draft, review, and editing. Ruyue Chen: investigation. Anne Puel: funding acquisition, resources, and writing—review and editing. Paul Bastard: investigation, resources, supervision, and writing—review and editing. Aurelie Cobat: resources. Theresa Sokol: resources. Ryan A. Langlois: conceptualization and writing—review and editing. Lisa Miorin: methodology, project administration, resources, supervision, writing—review and editing. Adolfo García-Sastre: conceptualization, funding acquisition, investigation, methodology, resources, supervision, and writing—review and editing. John A. Vanchiere: conceptualization, data curation, investigation, project administration, resources, supervision, and writing—review and editing. Jean-Laurent Casanova: conceptualization, funding acquisition, project administration, resources, supervision, validation, and writing—review and editing.