Tweet
Thuy-Lan Lite, M.D., Ph.D.
Stephanie Goya, Ph.D.
Margaret L. Davis, M.D., M.P.H.
Amy E. Morris, M.D.
Chloe Bryson-Cahn, M.D.
Alexander Vengerovsky, M.D.
Carlos G. Corpuz, M.D.
B. Ethan Nunley, B.S.
Janine R. Maenza, M.D.
Thomas R. Hawn, M.D., Ph.D.
Andrew M. Luks, M.D.
John B. Lynch, M.D., M.P.H.
University of Washington, Seattle
Timothy M. Uyeki, M.D., M.P.H.
Centers for Disease Control and Prevention, Atlanta
Helen Y. Chu, M.D., M.P.H.
Alexander L. Greninger, M.D., Ph.D.
Clifford C. Sung, M.D.
University of Washington, Seattle
Sporadic human infections with avian influenza A viruses have caused a wide spectrum of illness. As these novel influenza A viruses pose pandemic potential, timely detection and characterization of such human infections are important for global public health.
In October 2025, a 75-year-old woman with Waldenström macroglobulinemia presented to a local hospital with fever and diarrhea (illness day 1). She had had lymphopenia for several years and had received treatment with rituximab and bendamustine 7 months before presentation (see the Supplementary Appendix, available with the full text of this letter at NEJM.org). She had not received the 2025–2026 influenza vaccine. Real-time reverse-transcriptase–polymerase-chain-reaction (RT-PCR) assay of a nasal swab was negative for influenza A virus, and her chest radiograph was unremarkable. She was discharged home with instructions for supportive care, with no scheduled follow-up.
Cough, pharyngitis, and progressive dyspnea subsequently developed, and the patient returned to the hospital 9 days after she had been discharged; she was admitted for acute hypoxemic respiratory failure. Computed tomography of the chest revealed bilateral ground-glass and consolidative opacities in her lungs (Fig. S1 in the Supplementary Appendix), and her trachea was intubated on illness day 11. Repeat nasal swabs were again negative for influenza A virus on days 10 and 13. Despite treatment with broad-spectrum antimicrobials and glucocorticoids, her hypoxemia worsened, and severe acute respiratory distress syndrome developed. She was transferred twice for escalating levels of care, and she ultimately arrived at a tertiary center in Seattle, Washington, on day 15 (Table S1).
In Seattle, influenza A virus was detected by RT-PCR in both nasal-swab and bronchoalveolar-lavage specimens (cycle-threshold values, 34.6 and 23.8, respectively) (Figure 1A and Table S1). Both specimens subsequently tested positive for influenza A(H5).1 An extended respiratory viral PCR panel was otherwise negative. Oseltamivir therapy was started on illness day 16, along with baloxavir and amantadine on day 19 (Table S2) and tocilizumab on day 22 (Fig. S2). Pneumocystis jirovecii DNA was also detected by PCR in a bronchoalveolar-lavage specimen, although a direct fluorescent antibody test was negative. This finding was thought to represent colonization or a false positive result; however, the patient received trimethoprim–sulfamethoxazole owing to the severity of her respiratory failure. Despite antimicrobial therapy, lung-protective ventilation, prone positioning, neuromuscular blockade, and inhaled epoprostenol, she remained critically ill. Her hypoxemia worsened, and she died on illness day 28, after she had been transitioned to comfort-focused care.
Influenza A virus genome sequencing of the bronchoalveolar-lavage specimen, which was completed on illness day 21, confirmed highly pathogenic avian influenza (HPAI) A(H5N5) virus, Eurasian lineage clade 6 (EA6). Viral segment sequences had approximately 99% identity to contemporary HPAI A(H5N5) viruses without evidence of reassortment (Figure 1B and Fig. S3). The A(H5) hemagglutinin was consistent with clade 2.3.4.4b, the dominant hemagglutinin clade associated with currently spreading A(H5N1) viruses. No antiviral resistance mutations were detected. Intrahost viral dynamics from serial tracheal aspirates over 9 days revealed limited adaptation to human hosts. The allele frequency of mammalian-adaptation marker polymerase basic protein 2 (PB2) E627K decreased from 42% to 9% over the same period, a finding that is in contrast to A(H5N1) spillover events, in which PB2 and hemagglutinin mutations associated with enhanced binding to human receptors have emerged (Figure 1C and Fig. S4).2,3
Clinical history revealed that the patient lived on a rural property, where she had fed and handled ducks and their eggs daily, without gloves or masks. Two ducks had become ill and exhibited abnormal neurologic behaviors shortly before symptoms had developed in the patient (see video). Epidemiologic investigations are ongoing. No other cases have been identified among the patient’s close contacts or health care personnel who cared for her.
This case highlights several diagnostic and clinical challenges. First, multiple upper-respiratory-tract specimens were negative for influenza A virus despite high-burden lower respiratory tract infection, a finding that emphasizes the importance of obtaining bronchoalveolar-lavage or tracheal specimens in patients with severe pneumonia and relevant zoonotic exposures. Because of the delayed diagnosis of HPAI A(H5N5) virus infection, antiviral treatments were not initiated until late in the clinical course.4 Second, the identification of an HPAI A(H5N5) virus with a transient, low-frequency mammalian-associated change underscores the fact that fully avian influenza A viruses remain capable of zoonotic spillover.5 Finally, severe acute respiratory distress syndrome in the context of underlying hematologic cancer arouses concern that such patients may be particularly vulnerable to adverse outcomes from novel influenza A virus infections.
This report shows the importance of considering avian influenza A virus infection in patients with pneumonia and direct bird exposure, even when testing for influenza virus in upper-respiratory-tract specimens is negative.
