Influenza A virus (IAV) infection predisposes to subsequent bacterial superinfection in the lungs that promotes a greater pneumonia severity. Among host innate immune players, alveolar macrophages (AM) are essential for lung protection against bacteria. From mice studies, the prominent paradigm suggests that influenza infection induces almost total mouse AM (mAM) depletion, resulting in an increased vulnerability to secondary bacterial infection. Nevertheless, the potential IAV-induced AM depletion as never been study in human. The aim of the study is to characterize the AM number in newly influenza-infected human and to compare the ability of IAV to infect AM from mouse and human origin.

Because the observation of AM depletion in mice was described with non-translational influenza inoculum (too high), we infected c57BL/6J mice with a mild inoculum of influenza A virus (H3N2, 10 PFU) by intranasal route and measured the AM count. Next, we described a case report of a hospital-acquired influenza infection in a patient with pre and post-infection bronchoalveolar lavage (BAL) procedures. Then, we extracted data from 2 published clinical investigations assessing BAL in critical care unit and we performed an ancillary study included patients with a PCR-confirmed influenza infection and a BAL performed within the four firth days of ICU admission (AM expressed in mean±SEM). Finally, using ex vivo AMs models, we compared the ability of IAV to infect mouse and human AMs and the subsequence consequences regarding inflammatory mediator production.

Using mice model of mild IAV infection, we demonstrated that despite the low IAV inoculum the AM count dropped drastically in the BAL fluid from the fourth day (reduction of 64% from baseline) and was maximal at the seventh day post-infection (reduction of 90%). At the contrary, a patient newly infected by IAV (4days after the onset of the symptoms) did not present significant modification of the AM count (>200,000 cells/mL before and after the IAV infection). Next, we assessed the cellularity of BAL of influenza-infected critically ill patients (n=15) and controls (n=7), thanks to the ancillary study described. The absolute AM count was respectively 3.8×10⁵±1.5×10⁵ cells/mL in influenza-infected and 2.8×10⁵±0.6×10⁵ cells/mL in controls. In ex vivo AM models, we demonstrated that IAV infects, replicates, and induces inflammatory response similarly in AM of mouse and human origins.

We demonstrated that IAV-induced AM depletion in mouse but not in human. We did not observe difference in vitro regarding the viral replication and subsequent inflammatory response in mAM and mAM. It is unlikely that the secondary bacterial infections are explained by AM depletion in influenza-infected human.