{"paper_id":"4e1bb476-a1bd-4e7a-8978-d1c32c13c07c","body_text":"Empiric Azithromycin in COVID-19 Impacts the Respiratory Microbiome and Antimicrobial Resistome without Anti-inflammatory Benefit | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Empiric Azithromycin in COVID-19 Impacts the Respiratory Microbiome and Antimicrobial Resistome without Anti-inflammatory Benefit Charles Langelier, Abigail Glascock, Cole Maguire, Hoang Van Phan, and 50 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6875205/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 16 Mar, 2026 Read the published version in Nature Microbiology → Version 1 posted You are reading this latest preprint version Abstract Azithromycin is often prescribed unnecessarily for respiratory infections, many of which are viral. During the COVID-19 pandemic, its use was widespread, in part due to alleged therapeutic benefits, which have since been disproven. Here, we sought to understand the impact of azithromycin exposure on the respiratory microbiome, antimicrobial resistome, and host immune response in a prospective multicenter cohort of 1164 patients hospitalized for SARS-CoV-2 infection. Using longitudinal nasal metatranscriptomics, we compared patients treated with azithromycin (n=366, 31.4%) to those who received no antibiotics (n=474, 40.7%) or antibiotics other than azithromycin (n=324, 27.8%). We found that azithromycin treatment altered the community composition of the nasal microbiome, reducing bacterial relative abundance, increasing fungal relative abundance, and increasing potentially pathogenic taxa such as Klebsiella and Staphylococcus. Azithromycin treatment was most notably associated with increases in the number of detectably expressed macrolide/lincosamide/streptogramin (MLS) antimicrobial resistance genes, as well as their relative proportion in the resistome, with changes observable after one day of exposure. Of the MLS resistance genes, the expression of ermC , msrA and ermX increased the most in patients receiving azithromycin. Correlation analyses demonstrated that MLS resistance gene expression was significantly associated with the abundance of several taxa, including both commensal (e.g., Dolosigranulum, Corynebacterium ) and potentially pathogenic genera (e.g., Streptococcus, Staphylococcus). Assessment of the peripheral blood and upper airway host transcriptome demonstrated no differences in the expression of inflammatory genes. Taken together, our findings demonstrate that azithromycin treatment in COVID-19 leads to dysbiosis of the upper respiratory microbiome and changes in the expression of MLS resistance genes, without apparent anti-inflammatory benefit. Biological sciences/Microbiology/Microbial communities/Metagenomics Health sciences/Diseases/Infectious diseases/Viral infection Biological sciences/Microbiology/Antimicrobials/Antimicrobial resistance Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Antimicrobial resistance (AMR) is one of the top global health threats facing humanity 1 and is increasingly hindering the effective treatment of respiratory infections 2 . Rates of hospital-onset AMR infections dramatically increased during the SARS-CoV-2 pandemic 3 , complicating the treatment of COVID-19 and reversing the prior downward trend in deaths from drug resistant pathogens 4 . While the underlying reasons for this were multifactorial, the overuse of broad spectrum antibiotics in COVID-19 patients was a notable contributor 3,5–9 . Azithromycin, a World Health Organization essential medicine 10 , is one of the most widely used antibiotics in human healthcare with > 40 million prescriptions annually in the United States (U.S.) alone 11 . Azithromycin overuse has been well documented in the outpatient setting 12 , where an estimated 30% of antibiotic prescriptions are inappropriate 13 . During the first year of the COVID-19 pandemic, azithromycin became one of the most commonly used antibiotics in hospitalized patients as well 9,14 . This was driven in part by early studies suggesting possible antiviral activity 15,16 , and prior work demonstrating anti-inflammatory properties of macrolide antibiotics 17,18 . Randomized clinical trials, however, subsequently demonstrated that azithromycin conferred no clinical benefit in the treatment of COVID-19 19,20 . Nonetheless, many medical centers initially incorporated azithromycin into their COVID-19 treatment guidelines, and public misinformation continues to drive overprescription of the drug 21 . Recent work has found that azithromycin exposure can alter the human microbiome and its reservoir of antimicrobial resistance genes, termed the resistome 22–24 . For instance, secondary analyses of the MORDOR (Macrolide Oraux pour Réduire les Décès avec un Oeil sur la Résistance) clinical trials found that biannual mass azithromycin distribution to African children led to an increase in the abundance of both macrolide and other AMR gene (ARGs) classes in the gut microbiome 22 . In addition, adults with asthma randomized to thrice weekly azithromycin over 12 months had an increase in PCR copy number of macrolide resistance genes in sputum samples compared to controls 25 . Despite being the most common scenario for its use, no studies have yet assessed the impact of azithromycin on the respiratory microbiome in the context of empiric prescription for acute respiratory infection. Furthermore, no studies of azithromycin exposure have yet incorporated metatranscriptomics, which can assess both bacterial 16S rRNA abundance and ARG expression, providing a functional profile of the actively expressed resistome 26,27 . To address these gaps, we carry out respiratory metatranscriptomics in a prospective cohort of 1164 adults hospitalized for COVID-19, and study the impacts of azithromycin exposure. We find marked changes in the respiratory microbiome, including increases in detectably expressed macrolide resistance genes and their proportional representation in the airway resistome, without evidence of antiviral or immune modulating benefit. Taken together, our findings offer new insights into the adverse effects and biological consequences of empiric azithromycin exposure during viral infection. Results Cohort We carried out a prospective observational study of 1164 adults hospitalized for COVID-19 enrolled in the multicenter IMmuno Phenotyping Assessment in a COVID-19 Cohort (IMPACC) 28–30 between May 2020 and March 2021 ( Fig. 1a ). Previously established COVID-19 outcome trajectory groups (TGs) 29 were utilized to group patients based on disease severity. TGs ranged from 1 (lowest severity) to 5 (death within 28 days) 29 . Administration of azithromycin and other antibiotics was tracked following admission and throughout hospitalization. Of 1164 COVID-19 patients studied, 366 (31.4%) were treated empirically with azithromycin ± other antibiotics (Azithro group), 474 (40.7%) received no antibiotics (No-Abx group), and 324, (27.8%) received antibiotics other than azithromycin (Other-Abx group) ( Supp. Table 1 and 2 ). Empiric azithromycin administration was greatest among patients with the highest COVID-19 severity ( Fig. 1b ), although compared to those in the Other-Abx group, azithromycin-treated patients had less severe disease ( Supp. Table 1 ). The median number of azithromycin treatment days was 2 (IQR 1-4 days, range 1-35 days), which significantly differed across TGs (p = 0.03, Supp. Fig. 1 ). Azithromycin was administered in most (98.2%) patients within 1 week of hospital admission ( Fig. 1c ). Patients treated with azithromycin were most likely to have been co-administered ceftriaxone (77.2%) or vancomycin (19.4%) ( Fig. 1d ). Sex or race did not differ based on azithromycin usage ( Supp. Table 1 ). Impact of azithromycin exposure on the respiratory microbiome We first examined the impact of azithromycin on the upper respiratory tract microbiome using metatranscriptomic RNA sequencing (RNA-seq) of nasal swab (NS) samples collected at six timepoints over 28 days following hospital admission. These analyses were adjusted for age quintile, sex, severity TG, days of hospitalization, patient, receipt of corticosteroids, and receipt of the six most common antibiotics aside from azithromycin. We found that azithromycin treatment for 5 ±1 days, a common duration of prescription 31 , was associated with a significant decrease in bacterial abundance in the airway (adjusted p value (p adj ) = 0.026), with an effect observable within 1 ±1 day (p adj = 0.0019, Fig. 2a ). Other antibiotics also led to a decrease in upper respiratory bacterial abundance after 1 ±1 days (p adj = 0.036) but not at the later timepoint ( Fig. 2a ). Assessment of the mycobiome demonstrated that receipt of azithromycin was associated with an increase in fungal relative abundance in the upper airway after 1 ±1 days (padj = 0.038, Fig. 2b ), with a time-dependent increasing trend observed over 5 days of azithromycin administration (Supp. Fig. 2b ). No differences in upper airway microbiome alpha diversity were observed based on azithromycin treatment status after 5 ±1 days, although a significant increase was seen early on (p adj = 0.029, Fig. 2c ). Significant differences were found in the microbiome community composition based on azithromycin expsoure, measured by Bray-Curtis dissimilarity index (PERMANOVA p = 0.001, Fig. 2d ). A comparison of the trajectories for Bray-Curtis distances versus the earliest timepoint for each patient demonstrated marked shifts in community composition over time, independent of antibiotic exposure ( Fig. 2e ). Differential taxonomic abundance analysis demonstrated that azithromycin exposure was associated with enrichment of potentially pathogenic taxa in the upper airway including Staphylococcus and Klebsiella species, and depletion of several typically commensal taxa such as Neisseria and Fusobacterium ( Fig. 2f ). We also tested whether azithromycin treatment associated with any changes in SARS-CoV-2 relative abundance in the upper airway. We observed no differences with respect to either the Other Abx or No Abx groups after 5 ±1 days of treatment ( Supp. Fig. 3 ). Impact of azithromycin exposure on the respiratory antimicrobial resistome To understand the impact of azithromycin exposure on the respiratory antimicrobial resistome, we first compared Shannon Diversity Index across groups. Azithromycin exposure was associated with an increase in the alpha diversity of the upper airway resistome at the early (1 ±1 day) treatment timepoint compared with the Other-Abx group (p adj = 5.8e-3) but not with the No Abx controls ( Fig. 3a ). A shift in the composition of the antimicrobial resistome upon azithromycin exposure was also evident, with significant differences in Bray Curtis dissimilarity indices observed at 1 ±1 day and 5 ±1 days of azithromycin treatment versus controls (PERMANOVA p = 0.001, Fig. 3b ). Since azithromycin is a macrolide class antibiotic, we next focused on ARGs conferring resistance to the macrolide, lincosamide and streptogramin (MLS) class of antibiotics. We found that azithromycin exposure was associated with a significantly greater number of detectably expressed MLS genes compared to the No-Abx or Other-Abx groups at both the early (p adj = 3.5e-4 and 6.2e-7, respectively) and late (p adj = 1.3e-3 and 4.3 e-6, respectively) timepoints ( Fig. 3c ). Longitudinal modeling subsequently demonstrated that days of azithromycin exposure resulted in a significant increase in the number of detectably expressed MLS ARGs in the upper respiratory microbiome (p adj = 7.6e-7, Fig. 3d ). To comparatively evaluate the impact of azithromycin exposure across different ARG classes, we assessed longitudinal changes in the fraction of the resistome represented by each class. We found that after 5 ±1 days of azithromycin exposure, MLS ARGs increased from 24.5% to 42.9% of the resistome (p adj = 1.7e-4, Fig. 3e ). Notably, the enrichment in MLS ARGs, both in terms of ARG richness and as a percent of the resistome, persisted even 7-10 days after cessation of azithromycin ( Fig. 3f and Supp. Fig. 4 ). Correlations within the resistome and microbiome To assess relationships between macrolide resistance genes and bacterial taxa within the upper airway microbiome, we performed multi-dimensional correlation analyses ( Fig. 4b ). Significant positive correlations were found between several MLS genes and both potentially pathogenic (e.g., Staphylococcus, Streptococcus) as well as common commensal (e.g., Corynebacterium 32 , Dolosigranulum 33 ) genera ( Figs. 4b, 4c ). Impact of azithromycin on host inflammatory responses Prior studies have found that azithromycin can confer anti-inflammatory properties, leading to off-label use of the antibiotic as an immune modulatory agent 17,18 . We thus sought to understand whether azithromycin exposure in the setting of acute COVID-19 was associated with changes in host inflammatory gene expression in the airway or blood by carrying out differential gene expression analyses. No differentially expressed genes were identified (false discovery rate < 0.05) in either anatomical compartment after 5 ±1 days of treatment ( Supp. Tables 3, 4 ), suggesting that azithromycin does not meaningfully attenuate pathologic inflammatory responses in hospitalized COVID-19 patients. Discussion In a large multicenter cohort of hospitalized COVID-19 patients, empiric azithromycin treatment was associated with changes in the upper respiratory tract microbiome, mycobiome and antimicrobial resistome. We observed a significant expansion of detectably expressed macrolide resistance genes after 5 ±1 days of azithromycin treatment, with effects in some cases observed within a few days. In addition, we found that azithromycin treatment was associated with changes in the composition of the upper airway microbiota including enrichment of Klebsiella and Staphylococcus species, and an increase in the burden of fungal taxa in the upper respiratory tract. Together, our findings demonstrate that inappropriate azithromycin use in patients with viral respiratory infections can drive expansion of macrolide resistance determinants and disrupt the composition of the airway microbiome. Prior work examining mass azithromycin treatment in African children 22,24 found a concerning relationship between exposure to this drug and an increase in macrolide resistance genes in the gut microbiome after four years. We build on these important findings by demonstrating effects in the respiratory tract and at the transcriptional level, detectable within a few days of antibiotic treatment. Importantly, we find that azithromycin leads to not only an increase in the potential for resistance within the microbiome, but a functional impact on the expression of MLS resistance genes. Macrolide-resistant Streptococcus pnuemoniae and Streptococcus pyogenes are considered urgent threats by the U.S. Centers for Disease Control and Prevention 34 . Perhaps it is not surprising that macrolide resistance in these species has increased over the past decade given our results, and considering that 30% of antibiotics prescribed in outpatient settings have been deemed inappropriate or unnecessary 13 . Given that a large fraction of azithromycin prescriptions are written for children 12 , this is particularly concerning, as they may become colonized with macrolide-resistance bacteria at an early age due to unnecessary exposure to this drug. Azithromycin is used prophylactically for chronic obstructive pulmonary disease 35 , cystic fibrosis/bronchiectasis 36 , lung transplantation 37 , HIV/AIDS 38 and other conditions. While few studies have examined the impact of azithromycin prophylaxis on the respiratory resistome, a recent study of asthma patients found an increase in macrolide resistance genes using multiplex PCR in the sputum microbiome after 12 months 25 . Further work is needed to understand the impact of azithromycin prophylaxis on the upper and lower respiratory microbiome and resistome in these patient populations. A sub-analysis of the MORDOR trial found that four years of biannual azithromycin treatment in African children reduced mortality but led to an increase in macrolide resistant Streptococcus pneumoniae cultured from the nasopharynx 24 . Consistent with these prior microbiological observations, our correlation analysis suggested that both Streptococcus and Staphylococcus species, encompassing some of the most important bacterial pneumonia pathogens, may harbor these resistance determinants. In addition, we found relationships between MLS resistance gene expression and the abundance of commensal and contextually pathogenic taxa, such as Corynebacterium species. In the MORDOR trial, mass azithromycin treatment was also found to cause an increase in the burden of non-macrolide resistance genes in the gut microbiome 22 . In contrast, we found relatively few off-target effects on other classes of ARGs in the upper airway. One possible explanation may lie in the age and demographic differences of the studied populations. For instance, the burden of ARGs in both the gut 39,40 and respiratory microbiome increases with age, potentially due to lifetime antibiotic exposures 26 . It is possible that our cohort of hospitalized adults in the U.S. may have had a higher baseline burden of ARGs compared to African children living in rural settings. Alternatively, differences may be attributable to sampling of the respiratory versus gut microbiome, or the use of metatranscriptomics versus metagenomic DNA sequencing. Prior work has demonstrated that azithromycin has immune-modulating potential 17,18 , findings that have encouraged its use in patients with cystic fibrosis, chronic obstructive pulmonary disease 41 , and other inflammatory diseases. Azithromycin use early during the COVID-19 pandemic was in part driven by the idea that it might attenuate harmful inflammatory responses, as well as a now-retracted publication purporting clinical benefit when combined with hydroxychloroquine 42 . Randomized clinical trials, however, subsequently found no therapeutic benefit of azithromycin for COVID-19 19,20 . Consistent with this, we observed no significant associations between azithromycin treatment and inflammatory gene expression, or viral load, in either the airway or blood of hospitalized COVID-19 patients. Strengths of our study include a large multicenter cohort, detailed clinical phenotyping, use of respiratory metatranscriptomics, and rigorous quality control of clinical and biological data. As with any research, our study also has limitations. These include the observational study design and the use of short read sequencing, which precluded definitively linking macrolide resistance genes to specific taxa. In addition, our analyses were limited to the upper airway and thus may not reflect microbial changes occurring in the lungs. Antibiotic administration data were extracted manually by clinical research coordinators at each study site, an approach susceptible to human error. Azithromycin treatment, however, was confirmed by an independent adjudicator for every patient. Future randomized clinical trials - ideally that include airway and gut microbiome sampling as well as bacterial culture - are needed to more fully characterize the impact of exposure to azithromycin and other antibiotics on the human microbiome and resistome. In sum, we find that azithromycin exposure in hospitalized COVID-19 patients is associated with compositional changes in the airway microbiome and expansion of macrolide resistance genes. Taken together our findings suggest that empiric use of azithromycin in patients with viral respiratory infections may lead to adverse impacts and contribute to antimicrobial resistance. Methods Study Design, Clinical Cohort, Inclusion and Ethics IMPACC is a prospective longitudinal study that enrolled 1164 hospitalized COVID-19 patients, as previously described in detail 28–30,43,44 . Participants 18 years and older were recruited from 20 hospitals across 15 academic biomedical centers within the United States and confirmed to be SARS-CoV-2 positive by reverse transcription PCR (RT-PCR) testing. No participants were vaccinated for SARS-CoV-2 at time of enrollment nor during their hospitalization. To better categorize patients into different COVID-19 severity groups, they were classified into one of five trajectory groups using latent class mixed modeling of the degree of respiratory illness and external oxygen administration 29 . The Department of Health and Human Services Office for Human Research Protections (OHRP) and NIAID concurred that the IMPACC study qualified for public health surveillance exemption. The study protocol was reviewed by each site’s institutional review board (IRB), with twelve sites conducting as a public health surveillance study, and three sites integrating the IMPACC study into IRB-approved protocols (The University of Texas at Austin, IRB 2020-04-0117; University of California San Francisco, IRB 20-30497; Case Western Reserve University, IRB STUDY20200573) with participants providing informed consent. Participants enrolled at sites operating as a public health surveillance study were provided information sheets describing the study including the samples to be collected and plans for analysis and data de-identification. Participants who requested not to participate after review of the study plan and information were not enrolled. Participants were not compensated while hospitalized but were subsequently compensated for outpatient visits and surveys. This study was registered at clinicaltrials.gov (NCT0438777) and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. Metatranscriptomic Sequencing Mid-turbinate nasal swabs were collected withing 72 hours of hospital admission and at subsequent visits with target dates of 4, 7, 14, 21, and 28 days post hospital admission. The nasal swabs were stored in 1 mL of Zymo-DNA/RNA shield reagent (Zymo Research), before RNA was extracted twice in parallel from 250 uL of sample. The RNA was then purified with the KingFisher Flex sample purification system (ThermoFisher) and the quick DNA-RNA MagBead kit (Zymo Research). The duplicated RNA was pooled and aliquoted at 20 μL for the downstream RNA-sequencing. The isolated RNA was subsequently sequenced on a NovaSeq 6000 (Illumina) at 100 bp paired-end read length. The data was aligned using STAR (v2.4.3) against the GRCh38 reference genome and host gene counts were quantified using HTSeq-count (v0.4.1). Microbiome and Resistome Profiling Metatranscriptomic data was processed using the open-source CZ ID pipeline (https://czid.org/) 45 . Microbiome profiling was performed within CZ ID using the Illumina mNGS pipeline (v7.1). CZ ID first removes host reads by aligning to the GRCh38 human reference genome using STAR 46 . Adapters are removed using Trimmomatic 47 and reads are filtered for low quality using PriceSeq 48 . Duplicates are identified using czid-dedup. Low complexity reads are filtered out using the Lempel-Ziv-Welch algorithm and any remaining host reads are removed by alignment to GRCh38 using Bowtie2 49 . After performing the filtering steps, de-duplicated reads are subsampled to 2 million total reads. Taxonomic classification of remaining reads is performed on reads and assembled contigs (SPAdes) 50 by aligning to the NCBI nucleotide (NT) and protein (NR) databases. Resistome profiles were generated using the CZ ID AMR pipeline (v0.2), which leverages the Resistance Gene Identifier (RGI) tool and the Comprehensive Antibiotic Resistance Database (CARD) 51 . Following microbiome and resistome profiling with CZ ID, background and batch correction were performed to remove contaminants and adjust for batch effects (see below). To limit the contribution of spurious hits to downstream analyses, additional filtering of microbial taxa and ARGs was performed on a per sample basis. Microbial taxa were excluded if they did not meet all of the following criteria: (1) ten hits to the NT database, (2) one hit to the NR database, (3) a minimum alignment length of 50 bases. ARGs were excluded if they did not have ≥5% read coverage breadth and meet one of the following two criteria: (1) present in ≥5% of all samples with detectable ARGs (2) depth per million (DPM) ≥1 and ≥10 hits. Background & Batch Correction Negative controls (double distilled water) were processed and sequenced alongside samples in the IMPACC cohort to enable the characterization and subtraction of background contamination. The sequencing data generated from these samples was analyzed using the CZ ID metagenomic and AMR pipelines as described above. Background and batch correction was performed on the microbiome and resistome datasets separately. A negative binomial model was used to model the distribution of reads of microbial taxa/ARGs in the negative controls. Mean and dispersion parameters were then fitted to these data. Mean estimates were generated for each batch:taxon or batch:ARG pair in the negative controls, where batch corresponds to the phase of the IMPACC study (1, 2, 3A or 3B). The MASS package (v7.3.58.1) in R was used to generate a single dispersion parameter across all taxa/ARGs. P-values were adjusted for multiple testing using the Benjamini-Hochberg False Discovery Rate algorithm. Microbial taxa/ARGs that were present at a significantly higher abundance in participant samples than in negative controls (FDR <0.1) were retained for downstream analyses. Single Time Point Analyses Participants were assigned to one of three groups: those who received Azithromycin (Azithro) +/- other antibiotics, those who took only non-Azithromycin antibiotics (Other-Abx), or participants who did not receive antibiotics (No-Abx). Participants with only partially captured antibiotic start and stop dates were excluded from these analyses. The 2 time points studied for these groups were samples collected with 1 ±1 day of exposure and 5 ±1 days of exposure, applicable to the Azithromycin and Other-Abx groups. The second time point (5 ±1 days) was chosen based on the average course of azithromycin, approximately 5 days. In an attempt to match the No-Abx samples to the Azithromycin and Other-Abx samples at each time point, the distribution of days of hospitalization across those groups was examined. Based on the findings, we selected samples for the No-Abx group as follows: In the 1 ±1 day group, the samples had no antibiotic exposure and up to 40 days of hospitalization. In the 5 ±1 days group, the samples had no antibiotic exposure, and between 3-40 days of hospitalization. Samples that qualified for both time points were split evenly between time points. The number of \"No-Abx\" samples was also rarified (50%) to make the numbers in each group more comparable. In addition to this sample matching process for the No-Abx group, days from hospital admission was included as a covariate in statistical models to further control for slight differences. A linear mixed-effects model (using the lme4 package) was used to calculate differences between the groups at both timepoints while using age quintile, sex, severity TG, days from hospitalization, and receipt of steroids as fixed effects in the model, in combination with the participant’s enrollment site included as a random effect. For only the Azithromycin group, the number of days of the six most common co-administered antibiotics (vancomycin, ceftriaxone, cefepime, piperacillin-tazobactam, doxycycline and meropenem) were also included as fixed effects. Resulting p values were adjusted using the Benjamini-Hochberg False Discovery Rate algorithm via the p_adjust() function of the stats (v4.2.3) package. Generalized Additive Mixed Model (GAMM) Analyses Various metrics were modeled over days of azithromycin exposure using the nlme (v3.1-162), lme4 (v1.1-35.5), lmerTest (v3.1-3) and ggeffects (v1.7.2) packages in R. Samples were limited to the first 10 days of hospitalization and the first 5 days of antibiotic exposure due to few patients with longer antibiotic courses. The GAMM models included fixed effects of severity TG, sex, age quintile of the participant, and whether they ever received steroids, in addition to a smoothed term for days of azithromycin usage and days from hospitalization, and participant as a mixed effect. Additionally, days of administration for the six most prevalent antibiotics in the cohort aside from azithromycin (vancomycin, ceftriaxone, cefepime, piperacillin-tazobactam, doxycycline, and meropenem) were also added to the models as smooth terms. R model formula: ~ s(Azithromycin, bs = 'cr', k=4)+s(Ceftriaxone, bs = 'cr', k=4)+s(Vancomycin, bs = 'cr', k=4)+s(Cefepime, bs = 'cr', k=4)+s(Meropenem, bs = 'cr', k=4)+s(Piperacillin_Tazobactam, bs = 'cr', k=4)+s(Doxycycline, bs = 'cr', k=4)+s(event_date, bs = 'cr')+ trajectory_group + sex + discretized_admit_age_quantile + ever_steroids Microbiome Diversity Metrics Alpha diversity (Shannon Diversity Index) was calculated using the diversity() function in the R package vegan(2.6-6.1). Beta diversity (Bray-Curtis dissimilarity) analysis was performed using the vegan functions vegdist(), betadisper(), permutest() and adonis2(). The beta diversity analysis was adjusted for age quintile, sex, days since hospitalization, severity TG, patient, and receipt of corticosteroids in the adonis2() function. For the resistome analysis, receipt of the six most common antibiotics aside from azithromycin (vancomycin, ceftriaxone, cefepime, piperacillin-tazobactam, doxycycline and meropenem) was also included in the model. Principal Component Analysis (PCoA) of the resistome and bacterial microbiome was also performed using cmdscale() function of the stats(v4.2.3) package. Differential Abundance Analysis Bacterial microbiome profile data was converted into a phyloseq object using R packages ape (v5.8.1), ade4 (v1.7.22), and phyloseq (v1.34.0). Differential abundance analysis was performed using the R package ANCOMBC (v1.0.5) with an alpha level of 0.05 and prevalence filter (zero_cut argument) of 0.97. The analysis was adjusted for the following covariates: age quintile, sex, days from hospitalization, severity TG, patient, receipt of corticosteroids and receipt of the six most prevalent antibioticsaside from azithromycin (vancomycin, ceftriaxone, cefepime, piperacillin-tazobactam, doxycycline, and meropenem). P-values were adjusted for multiple testing using the Benjamini-Hochberg correction. Correlation Analyses The 50 most abundant bacterial taxa by reads per million (RPM) were included in the correlation analysis. RPM was used as an abundance metric in lieu of raw reads to adjust for variations in sequencing depth between samples. The abundance metric used for ARGs was depth per million (DPM), which adjusts for both variations in sequencing depth between samples and variations in gene length between ARGs. Spearman correlation was performed between bacterial taxa RPM and ARG DPM using the cor() function in the stats (v4.2.3) package and the cor_pmat() function in the rstatix (v0.7.2) package in R. Results were visualized using the corrplot (v0.95) package. PBMC and Nasal Swab Host Transcriptional Profiling RNA-seq and alignment against the host transcriptome was performed as previously described, 52,53 and the deidentified, quality-controlled raw gene count files and metadata were obtained from the IMPACC study. We filtered for host protein-coding genes that had at least 10 counts in at least 20% of the samples. To identify genes that were associated with azithromycin exposure compared to No-Abx patients, we utilized the limma (v3.58.1) package in R. We modeled gene expression as a function of azithromycin exposure, corrected for age, sex, trajectory group, days since hospitalization, and exposure to corticosteroids. Then, we applied voom and duplicateCorrelation for two iterations to calculate the correlation between multiple samples from the same patients. Next, we applied the lmFit and eBayes functions to estimate the effects of Azithromycin exposure on gene expression and their P-values. Finally, the P-values were adjusted with Benjamini-Hochberg correction. Statistics All code was written in R v4.2.3 or R 4.0.3. Data processing was performed using R packages dplyr (v1.1.4) and tidyverse (v2.0.0). Plots were generated using R packages ggplot2 (v3.5.1), ggpubr (v0.6.0), scales (1.3.0) and ggbeeswarm (v0.7.2). Package versions for each analysis are reported in the code repository. Declarations Data and code availability Data used in this study is available at ImmPort Shared Data under the accession number SDY1760 and in the NLM’s Database of Genotypes and Phenotypes (dbGaP) under the accession number phs002686.v2.p2. Source data for each figure are provided with this manuscript in the Source Data file. All code is deposited in the following Bitbucket repository: https://bitbucket.org/kleinstein/impacc-public-code/src/master/azithromycin_manuscript/ . Funding NIH (5R01AI135803-03, 5U19AI118608-04, 5U19AI128910-04, 4U19AI090023-11, 4U19AI118610-06, R01AI145835-01A1S1, 5U19AI062629-17, 5U19AI057229-17, 5U19AI125357-05, 5U19AI128913-03, 3U19AI077439-13, 5U54AI142766-03, 5R01AI104870-07, 3U19AI089992-09, 3U19AI128913-03, 5T32DA018926-18, and K0826161611); NIAID, NIH (3U19AI1289130, U19AI128913-04S1, R01AI122220, and 1K23AI185326-01); NCATS (UM1TR004528), and National Science Foundation (DMS2310836). Funding sources did not have a direct role in the design, analysis, or approval of this manuscript. IMPACC Network Competing Interests The Icahn School of Medicine at Mount Sinai has filed patent applications relating to SARS-CoV-2 serological assays and NDV-based SARS-CoV-2 vaccines which list Florian Krammer as co-inventor. Mount Sinai has spun out a company, Kantaro, to market serological tests for SARS-CoV-2. Florian Krammer has consulted for Merck and Pfizer (before 2020), and is currently consulting for Pfizer, Seqirus, 3rd Rock Ventures, Merck and Avimex. The Krammer laboratory is also collaborating with Pfizer on animal models of SARS-CoV-2. Viviana Simon is a co-inventor on a patent filed relating to SARS-CoV-2 serological assays (the \"Serology Assays\"). Ofer Levy is a named inventor on patents held by Boston Children’s Hospital relating to vaccine adjuvants and human in vitro platforms that model vaccine action. His laboratory has received research support from GlaxoSmithKline (GSK) and is a co-founder of and advisor to ARMR Sciences (formerly Ovax, Inc) that develops technologies to detect illicit substances and prevent overdose. He is also an advisor to GSK and Sanofi. Charles Cairns serves as a consultant to bioMerieux and is funded for a grant from Bill & Melinda Gates Foundation. James A Overton is a consultant at Knocean Inc. Jessica Lasky-Su serves as a scientific advisor of Precion Inc. Scott R. Hutton, Greg Michelloti and Kari Wong are employees of Metabolon Inc. Vicki Seyfer- Margolis is a current employee of MyOwnMed. Nadine Rouphael reports grants or contracts with Merck, Sanofi, Pfizer, Vaccine Company, Quidel, Lilly and Immorna, and has participated on data safety monitoring boards for Moderna, Sanofi, Seqirus, Pfizer, EMMES, ICON, BARDA, Imunon, CyanVac and Micron. Nadine Rouphael has also received support for meetings/travel from Sanofi and Moderna and honoraria from Virology Education. Adeeb Rahman is a current employee of Immunai Inc. Steven Kleinstein is a consultant related to ImmPort data repository for Peraton. Nathan Grabaugh is a consultant for Tempus Labs and the National Basketball Association. Akiko Iwasaki is a consultant for 4BIO, Blue Willow Biologics, Revelar Biotherapeutics, RIGImmune, Xanadu Bio, Paratus Sciences. Monika Kraft receives research funds paid to her institution from NIH, ALA; Sanofi, Astra-Zeneca for work in asthma, serves as a consultant for Astra-Zeneca, Sanofi, Chiesi, GSK for severe asthma; is a co-founder and CMO for RaeSedo, Inc, a company created to develop peptidomimetics for treatment of inflammatory lung disease. Esther Melamed received research funding from Babson Diagnostics and honorarium from Multiple Sclerosis Association of America and has served on the advisory boards of Genentech, Horizon, Teva, and Viela Bio. Carolyn Calfee receives research funding from NIH, FDA, DOD, Roche-Genentech and Quantum Leap Healthcare Collaborative as well as consulting services for Janssen, Vasomune, Gen1e Life Sciences, NGMBio, and Cellenkos. Wade Schulz was an investigator for a research agreement, through Yale University, from the Shenzhen Center for Health Information for work to advance intelligent disease prevention and health promotion; collaborates with the National Center for Cardiovascular Diseases in Beijing; is a technical consultant to Hugo Health, a personal health information platform; cofounder of Refactor Health, an AI-augmented data management platform for health care; and has received grants from Merck and Regeneron Pharmaceutical for research related to COVID-19. Grace A McComsey received research grants from Redhill, Cognivue, Pfizer, and Genentech, and served as a research consultant for Gilead, Merck, Viiv/GSK, and Jenssen. Linda N. Geng received research funding paid to her institution from Pfizer, Inc. Acknowledgments We thank the participants of the study for their voluntary enrollment and contribution of samples for this work. See the supplement for details on the IMPACC Network. We acknowledge the assistance of the following individuals: Sanya Thomas, Mitchell Cooney, Shun Rao, Sofia Vignolo, and Elena Morrocchi (all from the CDCC); Arash Naeim, Marianne Bernardo, Sarahmay Sanchez, Shannon Intluxay, Clara Magyar, Jenny Brook, Estefania Ramires-Sanchez, Megan Llamas, Claudia Perdomo, Clara E. Magyar, and Jennifer A. Fulcher (all from the David Geffen School of Medicine at UCLA); members of the UCLA Center for Pathology Research Services and the Pathology Research Portal; M. Catherine Muenker, Dimitri Duvilaire, Maxine Kuang, William Ruff, Khadir Raddassi, Denise Shepherd, Haowei Wang, Omkar Chaudhary, Syim Salahuddin, John Fournier, Michael Rainone, and Maxine Kuang (all from the Yale School of Medicine). We thank the leadership of Boston Children’s Hospital including Drs. Wendy Chung, Gary Fleisher, Nancy Andrews, and Kevin Churchwell for their support for the Precision Vaccines Program. Dr. Augustine’s and Becker’s co-authorship of this report does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases, the National Institutes of Health or any other agency of the United States Government. References EClinicalMedicine. Antimicrobial resistance: a top ten global public health threat. EClinicalMedicine 41 , 101221 (2021). Murray, C. J. et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. The Lancet 399 , 629–655 (2022). Baggs, J. et al. Antibiotic Resistant Infections among COVID-19 Inpatients in U.S. Hospitals. Clin. Infect. Dis. Off. Publ. Infect. Dis. Soc. Am. ciac517 (2022) doi:10.1093/cid/ciac517. Centers for Disease Control and Prevention. U.S. Impact on Antimicrobial Resistance, Special Report 2022. 2022. https://www.cdc.gov/drugresistance/pdf/covid19-impact-report-508.pdf (2022). Rose, A. N. et al. Trends in Antibiotic Use in United States Hospitals During the Coronavirus Disease 2019 Pandemic. Open Forum Infect. Dis. 8 , ofab236 (2021). Romaszko-Wojtowicz, A., Tokarczyk-Malesa, K., Doboszyńska, A. & Glińska-Lewczuk, K. Impact of COVID-19 on antibiotic usage in primary care: a retrospective analysis. Sci. Rep. 14 , 4798 (2024). Nandi, A., Pecetta, S. & Bloom, D. E. Global antibiotic use during the COVID-19 pandemic: analysis of pharmaceutical sales data from 71 countries, 2020–2022. eClinicalMedicine 57 , (2023). Sulis, G., Batomen, B., Kotwani, A., Pai, M. & Gandra, S. Sales of antibiotics and hydroxychloroquine in India during the COVID-19 epidemic: An interrupted time series analysis. PLoS Med. 18 , e1003682 (2021). O’Leary, E. N. et al. Impact of the COVID-19 Pandemic on Inpatient Antibiotic Use in the United States, January 2019 Through July 2022. Clin. Infect. Dis. 78 , 24–26 (2024). World Health Organization. WHO Model List of Essential Medicines. https://www.who.int/publications/i/item/WHO-MHP-HPS-EML-2023.02 (2023). U.S. Food and Drug Administration. FDA Drug Safety Communication: Azithromycin (Zithromax or Zmax) and the risk of potentially fatal heart rhythms. https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-azithromycin-zithromax-or-zmax-and-risk-potentially-fatal-heart (2018). Hersh, A. L. et al. Frequency of First-line Antibiotic Selection Among US Ambulatory Care Visits for Otitis Media, Sinusitis, and Pharyngitis. JAMA Intern. Med. 176 , 1870 (2016). Fleming-Dutra, K. E. et al. Prevalence of Inappropriate Antibiotic Prescriptions Among US Ambulatory Care Visits, 2010-2011. JAMA 315 , 1864–1873 (2016). Bogdanić, N., Močibob, L., Vidović, T., Soldo, A. & Begovać, J. Azithromycin consumption during the COVID-19 pandemic in Croatia, 2020. PLOS ONE 17 , e0263437 (2022). Touret, F. et al. In vitro screening of a FDA approved chemical library reveals potential inhibitors of SARS-CoV-2 replication. Sci. Rep. 10 , 13093 (2020). Gautret, P. et al. Hydroxychloroquine and azithromycin as a treatment of COVID-19: results of an open-label non-randomized clinical trial. Int. J. Antimicrob. Agents 56 , 105949 (2020). Kanoh, S. & Rubin, B. K. Mechanisms of action and clinical application of macrolides as immunomodulatory medications. Clin. Microbiol. Rev. 23 , 590–615 (2010). Speer, E. M. et al. Pentoxifylline, dexamethasone and azithromycin demonstrate distinct age-dependent and synergistic inhibition of TLR- and inflammasome-mediated cytokine production in human newborn and adult blood in vitro. PloS One 13 , e0196352 (2018). Oldenburg, C. E. et al. Effect of Oral Azithromycin vs Placebo on COVID-19 Symptoms in Outpatients With SARS-CoV-2 Infection: A Randomized Clinical Trial. JAMA 326 , 490 (2021). Hinks, T. S. C. et al. Azithromycin versus standard care in patients with mild-to-moderate COVID-19 (ATOMIC2): an open-label, randomised trial. Lancet Respir. Med. 9 , 1130–1140 (2021). Gagliotti, C. et al. Use of Azithromycin Attributable to Acute SARS ‐ CoV ‐2 Infection. Pharmacoepidemiol. Drug Saf. 33 , e5857 (2024). Doan, T. et al. Macrolide and Nonmacrolide Resistance with Mass Azithromycin Distribution. N. Engl. J. Med. 383 , 1941–1950 (2020). Arzika, A. M. et al. Prolonged mass azithromycin distributions and macrolide resistance determinants among preschool children in Niger: A sub-study of a cluster-randomized trial (MORDOR). PLoS Med. 21 , e1004386 (2024). Doan, T. et al. Macrolide Resistance in MORDOR I — A Cluster-Randomized Trial in Niger. N. Engl. J. Med. 380 , 2271–2273 (2019). Taylor, S. L. et al. Long-Term Azithromycin Reduces Haemophilus influenzae and Increases Antibiotic Resistance in Severe Asthma. Am. J. Respir. Crit. Care Med. 200 , 309–317 (2019). Chu, V. T. et al. The antibiotic resistance reservoir of the lung microbiome expands with age in a population of critically ill patients. Nat. Commun. 15 , 92 (2024). Chu, V. T. et al. Impact of doxycycline post-exposure prophylaxis for sexually transmitted infections on the gut microbiome and antimicrobial resistome. Nat. Med. (2024) doi:10.1038/s41591-024-03274-2. Arce, Joann et al. Multi-omic longitudinal study reveals immune correlates of clinical course among hospitalized COVID-19 patients. Cell Rep. Med. (2023). Ozonoff, A. et al. Phenotypes of disease severity in a cohort of hospitalized COVID-19 patients: Results from the IMPACC study. EBioMedicine 83 , 104208 (2022). IMPACC Manuscript Writing Team & IMPACC Network Steering Committee. Immunophenotyping assessment in a COVID-19 cohort (IMPACC): A prospective longitudinal study. Sci. Immunol. 6 , eabf3733 (2021). Hopkins, S. & Williams, D. Five-day azithromycin in the treatment of patients with community-acquired pneumonia. Curr. Ther. Res. 56 , 915–925 (1995). Ortiz-Pérez, A. et al. High frequency of macrolide resistance mechanisms in clinical isolates of Corynebacterium species. Microb. Drug Resist. Larchmt. N 16 , 273–277 (2010). Laclaire, L. & Facklam, R. Antimicrobial susceptibility and clinical sources of Dolosigranulum pigrum cultures. Antimicrob. Agents Chemother. 44 , 2001–2003 (2000). Centers for Disease Control and Prevention (U.S.). Antibiotic Resistance Threats in the United States, 2019 . https://stacks.cdc.gov/view/cdc/82532 (2019) doi:10.15620/cdc:82532. Venkatesan, P. GOLD COPD report: 2024 update. Lancet Respir. Med. 12 , 15–16 (2024). Mogayzel, P. J. et al. Cystic fibrosis pulmonary guidelines. Chronic medications for maintenance of lung health. Am. J. Respir. Crit. Care Med. 187 , 680–689 (2013). Vos, R. et al. A randomised controlled trial of azithromycin to prevent chronic rejection after lung transplantation. Eur. Respir. J. 37 , 164–172 (2011). National Institutes of Health, Centers for Disease Control and Prevention, HIV Medicine Association, and Infectious Diseases Society of America. Panel on Guidelines for the Prevention and Treatment of Opportunistic Infections in Adults and Adolescents with HIV. Guidelines for the Prevention and Treatment of Opportunistic Infections in Adults and Adolescents with HIV. https://clinicalinfo.hiv.gov/en/guidelines/adult-and-adolescent-opportunistic-infection (2024). Lu, N. et al. DNA microarray analysis reveals that antibiotic resistance-gene diversity in human gut microbiota is age related. Sci. Rep. 4 , 4302 (2014). Wu, L. et al. Metagenomics-Based Analysis of the Age-Related Cumulative Effect of Antibiotic Resistance Genes in Gut Microbiota. Antibiotics 10 , 1006 (2021). Albert, R. K. et al. Azithromycin for Prevention of Exacerbations of COPD. N. Engl. J. Med. 365 , 689–698 (2011). Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43 , e47–e47 (2015). Maguire, C. et al. Chronic Viral Reactivation and Associated Host Immune Response and Clinical Outcomes in Acute COVID-19 and Post-Acute Sequelae of COVID-19. BioRxiv Prepr. Serv. Biol. 2024.11.14.622799 (2024) doi:10.1101/2024.11.14.622799. Phan, H. V. et al. Host-microbe multiomic profiling reveals age-dependent immune dysregulation associated with COVID-19 immunopathology. Sci. Transl. Med. 16 , eadj5154 (2024). Kalantar, K. L. et al. IDseq-An open source cloud-based pipeline and analysis service for metagenomic pathogen detection and monitoring. GigaScience 9 , giaa111 (2020). Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinforma. Oxf. Engl. 29 , 15–21 (2013). Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinforma. Oxf. Engl. 30 , 2114–2120 (2014). Ruby, J. G., Bellare, P. & Derisi, J. L. PRICE: software for the targeted assembly of components of (Meta) genomic sequence data. G3 Bethesda Md 3 , 865–880 (2013). Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat Methods 9 , (2012). Bankevich, A. et al. SPAdes: A New Genome Assembly Algorithm and Its Applications to Single-Cell Sequencing. J. Comput. Biol. 19 , 455–477 (2012). Alcock, B. P. et al. CARD 2023: expanded curation, support for machine learning, and resistome prediction at the Comprehensive Antibiotic Resistance Database. Nucleic Acids Res. 51 , D690–D699 (2023). Ozonoff, A. et al. Phenotypes of disease severity in a cohort of hospitalized COVID-19 patients: Results from the IMPACC study. EBioMedicine 83 , 104208 (2022). IMPACC Manuscript Writing Team & IMPACC Network Steering Committee. Immunophenotyping assessment in a COVID-19 cohort (IMPACC): A prospective longitudinal study. Sci. Immunol. 6 , eabf3733 (2021). Additional Declarations Yes there is potential Competing Interest. The Icahn School of Medicine at Mount Sinai has filed patent applications relating to SARS-CoV-2 serological assays and NDV-based SARS-CoV-2 vaccines which list Florian Krammer as co-inventor. Mount Sinai has spun out a company, Kantaro, to market serological tests for SARS-CoV-2. Florian Krammer has consulted for Merck and Pfizer (before 2020), and is currently consulting for Pfizer, Seqirus, 3rd Rock Ventures, Merck and Avimex. The Krammer laboratory is also collaborating with Pfizer on animal models of SARS-CoV-2. Viviana Simon is a co-inventor on a patent filed relating to SARS-CoV-2 serological assays (the \"Serology Assays\"). Ofer Levy is a named inventor on patents held by Boston Children’s Hospital relating to vaccine adjuvants and human in vitro platforms that model vaccine action. His laboratory has received research support from GlaxoSmithKline (GSK) and is a co-founder of and advisor to ARMR Sciences (formerly Ovax, Inc) that develops technologies to detect illicit substances and prevent overdose. He is a consultant to GSK and Sanofi. Charles Cairns serves as a consultant to bioMerieux and is funded for a grant from Bill & Melinda Gates Foundation. James A Overton is a consultant at Knocean Inc. Jessica Lasky-Su serves as a scientific advisor of Precion Inc. Scott R. Hutton, Greg Michelloti and Kari Wong are employees of Metabolon Inc. Vicki Seyfer- Margolis is a current employee of MyOwnMed. Nadine Rouphael reports grants or contracts with Merck, Sanofi, Pfizer, Vaccine Company, Quidel, Lilly and Immorna, and has participated on data safety monitoring boards for Moderna, Sanofi, Seqirus, Pfizer, EMMES, ICON, BARDA, Imunon, CyanVac and Micron. Nadine Rouphael has also received support for meetings/travel from Sanofi and Moderna and honoraria from Virology Education. Adeeb Rahman is a current employee of Immunai Inc. Steven Kleinstein is a consultant related to ImmPort data repository for Peraton. Nathan Grabaugh is a consultant for Tempus Labs and the National Basketball Association. Akiko Iwasaki is a consultant for 4BIO, Blue Willow Biologics, Revelar Biotherapeutics, RIGImmune, Xanadu Bio, Paratus Sciences. Monika Kraft receives research funds paid to her institution from NIH, ALA; Sanofi, Astra-Zeneca for work in asthma, serves as a consultant for Astra-Zeneca, Sanofi, Chiesi, GSK for severe asthma; is a co-founder and CMO for RaeSedo, Inc, a company created to develop peptidomimetics for treatment of inflammatory lung disease. Esther Melamed received research funding from Babson Diagnostics and honorarium from Multiple Sclerosis Association of America and has served on the advisory boards of Genentech, Horizon, Teva, and Viela Bio. Carolyn Calfee receives research funding from NIH, FDA, DOD, Roche-Genentech and Quantum Leap Healthcare Collaborative as well as consulting services for Janssen, Vasomune, Gen1e Life Sciences, NGMBio, and Cellenkos. Wade Schulz was an investigator for a research agreement, through Yale University, from the Shenzhen Center for Health Information for work to advance intelligent disease prevention and health promotion; collaborates with the National Center for Cardiovascular Diseases in Beijing; is a technical consultant to Hugo Health, a personal health information platform; cofounder of Refactor Health, an AI-augmented data management platform for health care; and has received grants from Merck and Regeneron Pharmaceutical for research related to COVID-19. Grace A McComsey received research grants from Redhill, Cognivue, Pfizer, and Genentech, and served as a research consultant for Gilead, Merck, Viiv/GSK, and Jenssen. Linda N. Geng received research funding paid to her institution from Pfizer, Inc. Supplementary Files SupplementalTable3NS.csv Supplemental Table 3 SupplementalTable4PBMC.csv Supplemental Table 4 nreditorialpolicychecklistNMEDA142753.pdf Editorial Policy Checklist SupplementaryMaterials.docx Cite Share Download PDF Status: Published Journal Publication published 16 Mar, 2026 Read the published version in Nature Microbiology → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-6875205\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":false,\"archivedVersions\":[],\"articleType\":\"Article\",\"associatedPublications\":[],\"authors\":[{\"id\":472169759,\"identity\":\"53ee8288-2bce-4975-9e1a-75fabf32f9df\",\"order_by\":0,\"name\":\"Charles Langelier\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDUlEQVRIie3QMUvDQBTA8fcIOFWyHqj0K6Q4Kl39Ch3zCJiluBS6OPiKoEuh6235CimuDi8c6CJxDXSpCM7p1qFCcw5SxBwdBe8/PJLwfnAXAJ/vDxbqZiDbx4AF4Ny+LgEO2omqvglacmlt5CTRYIc0w+xBjm7fa+T+VfdiMpHV4yvNZg2rx6adHD9FEXIy6r0UXOiPBekqAdSlg6g4ipEDmmti05EF5SqU4PDORdJakG9onr2x2UhJeWgg+HSRwbDHyIYy1UwQoRwSCNBBVDUcAfEz5R3iYirJqb1LMS3TVhLq9KFeba4puzemXkv/pPljuFyPz1rJV7E9oex8kF/XftblvdZ8Pp/vP7YF8Nxi5JeT2+IAAAAASUVORK5CYII=\",\"orcid\":\"https://orcid.org/0000-0002-6708-4646\",\"institution\":\"University of California, San Francisco\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Charles\",\"middleName\":\"\",\"lastName\":\"Langelier\",\"suffix\":\"\"},{\"id\":472169760,\"identity\":\"e8b40215-11bf-4953-aa6c-7538cd55a22b\",\"order_by\":1,\"name\":\"Abigail Glascock\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Chan Zuckerberg Biohub\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Abigail\",\"middleName\":\"\",\"lastName\":\"Glascock\",\"suffix\":\"\"},{\"id\":472169761,\"identity\":\"0e82e5d6-495e-43e9-8b23-0df7e3d7a732\",\"order_by\":2,\"name\":\"Cole Maguire\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-8986-9762\",\"institution\":\"The University of Texas at Austin, Austin, TX 78712, USA\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Cole\",\"middleName\":\"\",\"lastName\":\"Maguire\",\"suffix\":\"\"},{\"id\":472169762,\"identity\":\"34b5a8b3-34b1-491b-97a1-36ce1e6838ea\",\"order_by\":3,\"name\":\"Hoang Van Phan\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"UCSF\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Hoang\",\"middleName\":\"Van\",\"lastName\":\"Phan\",\"suffix\":\"\"},{\"id\":472169763,\"identity\":\"db091f8f-46ef-4129-b79b-ebb187375d3f\",\"order_by\":4,\"name\":\"Emily Lydon\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of California San Francisco\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Emily\",\"middleName\":\"\",\"lastName\":\"Lydon\",\"suffix\":\"\"},{\"id\":472169764,\"identity\":\"5aff5f85-58f6-4990-8177-a9e01ca917cb\",\"order_by\":5,\"name\":\"Carolyn Calfee\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of California San Francisco\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Carolyn\",\"middleName\":\"\",\"lastName\":\"Calfee\",\"suffix\":\"\"},{\"id\":472169765,\"identity\":\"63e63881-0a89-4a9e-b68e-402dc6e38dd4\",\"order_by\":6,\"name\":\"IMPACC Network\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"NA\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"IMPACC\",\"middleName\":\"\",\"lastName\":\"Network\",\"suffix\":\"\"},{\"id\":472169766,\"identity\":\"3a300381-543d-4763-b5b6-0424ee6fdf8d\",\"order_by\":7,\"name\":\"David Corry\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-9729-1016\",\"institution\":\"Baylor College of Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"David\",\"middleName\":\"\",\"lastName\":\"Corry\",\"suffix\":\"\"},{\"id\":472169767,\"identity\":\"18b21856-f409-4faa-b18a-d19186160c6f\",\"order_by\":8,\"name\":\"Farrah Kheradmand\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-5343-103X\",\"institution\":\"Baylor College of Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Farrah\",\"middleName\":\"\",\"lastName\":\"Kheradmand\",\"suffix\":\"\"},{\"id\":472169768,\"identity\":\"d7ad46f3-ef8b-4598-a5c9-b877000c227e\",\"order_by\":9,\"name\":\"Lindsey Baden\",\"email\":\"\",\"orcid\":\"https://orcid.org/0009-0004-1752-1926\",\"institution\":\"Brigham and Women’s Hospital\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Lindsey\",\"middleName\":\"\",\"lastName\":\"Baden\",\"suffix\":\"\"},{\"id\":472169769,\"identity\":\"bf57437b-42cf-4656-8a3b-6347973871f4\",\"order_by\":10,\"name\":\"Rafick-Pierre Sekaly\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Case Western Reserve University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Rafick-Pierre\",\"middleName\":\"\",\"lastName\":\"Sekaly\",\"suffix\":\"\"},{\"id\":472169770,\"identity\":\"a4f9addb-ead3-4bdb-978c-2f83c7509670\",\"order_by\":11,\"name\":\"Grace McComsey\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-2690-8888\",\"institution\":\"Case Western Reserve University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Grace\",\"middleName\":\"\",\"lastName\":\"McComsey\",\"suffix\":\"\"},{\"id\":472169771,\"identity\":\"107b6046-909a-4a1b-aba0-e49baf721976\",\"order_by\":12,\"name\":\"Elias Haddad\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-6924-4300\",\"institution\":\"Drexel University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Elias\",\"middleName\":\"\",\"lastName\":\"Haddad\",\"suffix\":\"\"},{\"id\":472169772,\"identity\":\"5f6a672e-3b9f-4c9f-a0fa-dd003b9877f7\",\"order_by\":13,\"name\":\"Charles Cairns\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-2291-5415\",\"institution\":\"Drexel University/Tower Health Hospital\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Charles\",\"middleName\":\"\",\"lastName\":\"Cairns\",\"suffix\":\"\"},{\"id\":472169773,\"identity\":\"440c122e-9eef-4a32-9000-3a23cc7f9474\",\"order_by\":14,\"name\":\"Bali Pulendran\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-6517-4333\",\"institution\":\"Stanford University School of Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Bali\",\"middleName\":\"\",\"lastName\":\"Pulendran\",\"suffix\":\"\"},{\"id\":472169774,\"identity\":\"cd34cbbe-59ed-4313-b00c-14515b596043\",\"order_by\":15,\"name\":\"Ana Fernandez-Sesma\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-3405-6702\",\"institution\":\"Icahn School of Medicine at Mount Sinai\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Ana\",\"middleName\":\"\",\"lastName\":\"Fernandez-Sesma\",\"suffix\":\"\"},{\"id\":472169775,\"identity\":\"e14ebb36-cf05-4ac2-8b38-b2b04d4e4c53\",\"order_by\":16,\"name\":\"Viviana Simon\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-6416-5096\",\"institution\":\"Icahn School of Medicine at Mount Sinai\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Viviana\",\"middleName\":\"\",\"lastName\":\"Simon\",\"suffix\":\"\"},{\"id\":472169776,\"identity\":\"84b7408e-9800-40bd-a893-01ff3839e241\",\"order_by\":17,\"name\":\"Jordan Metcalf\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-4359-5270\",\"institution\":\"University of Oklahoma Health Sciences Center\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Jordan\",\"middleName\":\"\",\"lastName\":\"Metcalf\",\"suffix\":\"\"},{\"id\":472169777,\"identity\":\"2c5acb1a-0b6a-40ff-bda6-e23a69d85d28\",\"order_by\":18,\"name\":\"Nelson Higuita\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"U Oklahoma\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Nelson\",\"middleName\":\"\",\"lastName\":\"Higuita\",\"suffix\":\"\"},{\"id\":472169778,\"identity\":\"65e86054-4f14-4a5c-9f99-48d8ed92054f\",\"order_by\":19,\"name\":\"William Messer\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-4381-6347\",\"institution\":\"OHSU\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"William\",\"middleName\":\"\",\"lastName\":\"Messer\",\"suffix\":\"\"},{\"id\":472169779,\"identity\":\"29208dad-f0c3-441c-a417-795f10074e26\",\"order_by\":20,\"name\":\"Mark Davis\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-6868-657X\",\"institution\":\"Stanford University School of Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Mark\",\"middleName\":\"\",\"lastName\":\"Davis\",\"suffix\":\"\"},{\"id\":472169780,\"identity\":\"a3a38c60-9f4d-4fc3-9b53-8bd75974fd8e\",\"order_by\":21,\"name\":\"Kari C. Nadeau\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-2146-2955\",\"institution\":\"Stanford University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Kari\",\"middleName\":\"C.\",\"lastName\":\"Nadeau\",\"suffix\":\"\"},{\"id\":472169781,\"identity\":\"4ed69eb5-3bac-487f-866a-8f84268317a1\",\"order_by\":22,\"name\":\"Monica Kraft\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of Arizona\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Monica\",\"middleName\":\"\",\"lastName\":\"Kraft\",\"suffix\":\"\"},{\"id\":472169782,\"identity\":\"e964432c-5f32-47a0-b1d7-395e5f92cea4\",\"order_by\":23,\"name\":\"Chris Bime\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of Arizona\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Chris\",\"middleName\":\"\",\"lastName\":\"Bime\",\"suffix\":\"\"},{\"id\":472169783,\"identity\":\"0c405b34-868f-4fb7-a667-26062de637e3\",\"order_by\":24,\"name\":\"Joanna Schaenman\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"UCLA\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Joanna\",\"middleName\":\"\",\"lastName\":\"Schaenman\",\"suffix\":\"\"},{\"id\":472169784,\"identity\":\"eabccce8-5ae4-490a-be5c-4810860659b2\",\"order_by\":25,\"name\":\"David Erle\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-2171-0648\",\"institution\":\"UCSF\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"David\",\"middleName\":\"\",\"lastName\":\"Erle\",\"suffix\":\"\"},{\"id\":472169785,\"identity\":\"75576852-6ee5-4499-b5c3-bcfb5e64cf47\",\"order_by\":26,\"name\":\"Mark Atkinson\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-8489-4782\",\"institution\":\"University of Florida\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Mark\",\"middleName\":\"\",\"lastName\":\"Atkinson\",\"suffix\":\"\"},{\"id\":472169786,\"identity\":\"77dbca32-431a-4467-bb71-c62bc74a8f79\",\"order_by\":27,\"name\":\"Lauren I R Ehrlich\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-1697-1755\",\"institution\":\"University of Texas at Austin\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Lauren\",\"middleName\":\"I R\",\"lastName\":\"Ehrlich\",\"suffix\":\"\"},{\"id\":472169787,\"identity\":\"90bd9faf-6fa3-4d1d-9758-04c755b7b76c\",\"order_by\":28,\"name\":\"Esther Melamed\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"The University of Texas at Austin\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Esther\",\"middleName\":\"\",\"lastName\":\"Melamed\",\"suffix\":\"\"},{\"id\":472169788,\"identity\":\"5a672355-e112-4ded-b9f5-b0c883e8d04e\",\"order_by\":29,\"name\":\"Ruth Montgomery\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-8661-4454\",\"institution\":\"Yale School of Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Ruth\",\"middleName\":\"\",\"lastName\":\"Montgomery\",\"suffix\":\"\"},{\"id\":472169789,\"identity\":\"985c5713-94db-4faa-9bad-9e236b58ed88\",\"order_by\":30,\"name\":\"Albert Shaw\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Yale University School of Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Albert\",\"middleName\":\"\",\"lastName\":\"Shaw\",\"suffix\":\"\"},{\"id\":472169790,\"identity\":\"6644a53a-dc63-4e31-bd19-633a118d89cc\",\"order_by\":31,\"name\":\"Catherine Hough\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Oregon Health Sciences University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Catherine\",\"middleName\":\"\",\"lastName\":\"Hough\",\"suffix\":\"\"},{\"id\":472169791,\"identity\":\"3300b445-994d-44f4-b907-ff855926bc2a\",\"order_by\":32,\"name\":\"Linda Geng\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-9891-5705\",\"institution\":\"Stanford University School of Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Linda\",\"middleName\":\"\",\"lastName\":\"Geng\",\"suffix\":\"\"},{\"id\":472169792,\"identity\":\"a46ac35a-3219-4147-9da1-b029b82f8b74\",\"order_by\":33,\"name\":\"Annmarie Hoch\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-0836-2164\",\"institution\":\"Harvard\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Annmarie\",\"middleName\":\"\",\"lastName\":\"Hoch\",\"suffix\":\"\"},{\"id\":472169793,\"identity\":\"d07b2ad4-5ae0-4963-93c2-f77733e42fc5\",\"order_by\":34,\"name\":\"David Hafler\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-4664-535X\",\"institution\":\"Yale University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"David\",\"middleName\":\"\",\"lastName\":\"Hafler\",\"suffix\":\"\"},{\"id\":472169794,\"identity\":\"143ad859-29f2-47be-a940-2430f4f3d80a\",\"order_by\":35,\"name\":\"Alison Augustine\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"National Institute of Allergy and Infectious Diseases/National Institutes of Health\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Alison\",\"middleName\":\"\",\"lastName\":\"Augustine\",\"suffix\":\"\"},{\"id\":472169795,\"identity\":\"d4bf773e-1a45-44e3-8830-eb14fd9ff6c0\",\"order_by\":36,\"name\":\"Patrice Becker\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-1170-8587\",\"institution\":\"National Institute of Allergy and Infectious Diseases/National Institutes of Health\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Patrice\",\"middleName\":\"\",\"lastName\":\"Becker\",\"suffix\":\"\"},{\"id\":472169796,\"identity\":\"7c745985-551b-4558-a31b-fc595df8fde2\",\"order_by\":37,\"name\":\"Bjoern Peters\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-8457-6693\",\"institution\":\"La Jolla Institute for Allergy and Immunology\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Bjoern\",\"middleName\":\"\",\"lastName\":\"Peters\",\"suffix\":\"\"},{\"id\":472169797,\"identity\":\"4b2533c6-2daf-4c22-b988-617eb006af98\",\"order_by\":38,\"name\":\"Al Ozonoff\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-4233-5899\",\"institution\":\"Boston Children's Hospital\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Al\",\"middleName\":\"\",\"lastName\":\"Ozonoff\",\"suffix\":\"\"},{\"id\":472169798,\"identity\":\"fe8252a8-872a-478d-9153-4c594385d9f6\",\"order_by\":39,\"name\":\"Seunghee Kim-Schulze\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-0192-4400\",\"institution\":\"Icahn School of Medicine at Mount Sinai\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Seunghee\",\"middleName\":\"\",\"lastName\":\"Kim-Schulze\",\"suffix\":\"\"},{\"id\":472169799,\"identity\":\"f90504e8-cdc7-4a22-b01c-490e1a873779\",\"order_by\":40,\"name\":\"Florian Krammer\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-4121-776X\",\"institution\":\"Icahn School of Medicine at Mount Sinai\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Florian\",\"middleName\":\"\",\"lastName\":\"Krammer\",\"suffix\":\"\"},{\"id\":472169800,\"identity\":\"ad742c2a-3ae1-4bc1-95dc-b2e1a39ef595\",\"order_by\":41,\"name\":\"Steven Bosinger\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-2116-5061\",\"institution\":\"Emory University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Steven\",\"middleName\":\"\",\"lastName\":\"Bosinger\",\"suffix\":\"\"},{\"id\":472169801,\"identity\":\"0fea0878-9211-47ff-a980-c011e83f7b8e\",\"order_by\":42,\"name\":\"Walter Eckalbar\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of California, San Francisco\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Walter\",\"middleName\":\"\",\"lastName\":\"Eckalbar\",\"suffix\":\"\"},{\"id\":472169802,\"identity\":\"cc70f790-3fe4-4bf2-a19f-e488ad569090\",\"order_by\":43,\"name\":\"Matthew Altman\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-1784-8505\",\"institution\":\"Benaroya Research Institute\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Matthew\",\"middleName\":\"\",\"lastName\":\"Altman\",\"suffix\":\"\"},{\"id\":472169803,\"identity\":\"6888e557-e7e3-4f04-ba48-fc6c628ab2bf\",\"order_by\":44,\"name\":\"Michael Wilson\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-8705-5084\",\"institution\":\"University of California, San Francisco\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Michael\",\"middleName\":\"\",\"lastName\":\"Wilson\",\"suffix\":\"\"},{\"id\":472169804,\"identity\":\"10245694-3bac-41a3-a2e6-6a789c82df0c\",\"order_by\":45,\"name\":\"Leying Guan\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-0609-1073\",\"institution\":\"Yale University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Leying\",\"middleName\":\"\",\"lastName\":\"Guan\",\"suffix\":\"\"},{\"id\":472169805,\"identity\":\"2fbd2551-1373-48d4-9906-434d8aee28e6\",\"order_by\":46,\"name\":\"Holden Maecker\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-0795-9946\",\"institution\":\"Stanford University School of Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Holden\",\"middleName\":\"\",\"lastName\":\"Maecker\",\"suffix\":\"\"},{\"id\":472169806,\"identity\":\"33122db2-4c69-484a-8a7f-79a7d60f6d40\",\"order_by\":47,\"name\":\"Hanno Steen\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-0179-6648\",\"institution\":\"Boston Children's Hospital\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Hanno\",\"middleName\":\"\",\"lastName\":\"Steen\",\"suffix\":\"\"},{\"id\":472169807,\"identity\":\"eda363da-6ace-475e-8995-34a68ef573ba\",\"order_by\":48,\"name\":\"Joann Diray-Arce\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-2183-4269\",\"institution\":\"Harvard\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Joann\",\"middleName\":\"\",\"lastName\":\"Diray-Arce\",\"suffix\":\"\"},{\"id\":472169808,\"identity\":\"d2dfcb06-536f-4708-a7ce-ab385aa737e4\",\"order_by\":49,\"name\":\"Nadine Rouphael\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Emory University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Nadine\",\"middleName\":\"\",\"lastName\":\"Rouphael\",\"suffix\":\"\"},{\"id\":472169809,\"identity\":\"1e4852b8-dfde-41e2-a544-6db997dd4289\",\"order_by\":50,\"name\":\"Steven Kleinstein\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0003-4957-1544\",\"institution\":\"Yale University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Steven\",\"middleName\":\"\",\"lastName\":\"Kleinstein\",\"suffix\":\"\"},{\"id\":472169810,\"identity\":\"3e0fae8b-268c-4f1b-ab8f-cdff3fe6ec9f\",\"order_by\":51,\"name\":\"Elaine Reed\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of California, Los Angeles\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Elaine\",\"middleName\":\"\",\"lastName\":\"Reed\",\"suffix\":\"\"},{\"id\":472169811,\"identity\":\"8dc66a51-a08c-4391-8c24-ebea2be8297f\",\"order_by\":52,\"name\":\"Ofer Levy\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-5859-1945\",\"institution\":\"Boston Children's Hospital\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Ofer\",\"middleName\":\"\",\"lastName\":\"Levy\",\"suffix\":\"\"},{\"id\":472169812,\"identity\":\"8ac90096-f40c-49b9-bd87-ba1c038dda90\",\"order_by\":53,\"name\":\"Victoria Chu\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-7480-965X\",\"institution\":\"UCSF\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Victoria\",\"middleName\":\"\",\"lastName\":\"Chu\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-06-11 23:40:11\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-6875205/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-6875205/v1\",\"draftVersion\":[],\"editorialEvents\":[{\"content\":\"https://doi.org/10.1038/s41564-026-02285-8\",\"type\":\"published\",\"date\":\"2026-03-16T04:00:00+00:00\"}],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":85071514,\"identity\":\"4eaaaed7-d55c-4a53-b1a9-62384aed8823\",\"added_by\":\"auto\",\"created_at\":\"2025-06-20 15:40:36\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":150212,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eStudy overview and cohort demographics. a \\u003c/strong\\u003eGraphical overview demonstrating geographic location of IMPACC cohort study sites, sampling approach, and antimicrobial exposure groups studied including (n=366, 31.4%) treated empirically with azithromycin +/- other antibiotics, (n=474, 40.7%) who received no antibiotics, and (n=324, 27.8%) received antibiotics other than azithromycin. \\u003cstrong\\u003eb \\u003c/strong\\u003eBar plot demonstrating percent of patients within each COVID-19 trajectory group (colors) exposed to azithromycin. Grey reference bar indicates percent of patients within the cohort treated with azithromycin. \\u003cstrong\\u003ec\\u003c/strong\\u003eDensity plot highlighting distribution of azithromycin treatment with respect to days of hospitalization. \\u003cstrong\\u003ed\\u003c/strong\\u003e Bar plot depicting antimicrobials most frequently co-prescribed with azithromycin.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/d70efda5d4fe1c1f20f9ca17.png\"},{\"id\":85069833,\"identity\":\"0fc05839-f383-4d96-b077-ad03f205827c\",\"added_by\":\"auto\",\"created_at\":\"2025-06-20 15:32:36\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":317051,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eAzithromycin treatment disrupts the respiratory microbiome and mycobiome. \\u003c/strong\\u003e\\u0026nbsp;\\u003cstrong\\u003ea \\u003c/strong\\u003eTotal bacterial abundance in the nasal microbiome, measured by reads per million (RPM), comparing patients treated with azithromycin (orange), other antibiotics (green) or no antibiotics (blue) after 1 ±1 day (left) or 5 ±1 days (right) of antimicrobial treatment or hospitalization (controls). \\u003cstrong\\u003eb \\u003c/strong\\u003eTotal fungal abundance in the nasal microbiome, measured by reads per million (RPM), highlighting differences between antimicrobial treatment groups. \\u003cstrong\\u003ec \\u003c/strong\\u003eAlpha diversity of the nasal microbiome highlighting differences between antimicrobial treatment groups. \\u003cstrong\\u003ed\\u003c/strong\\u003e Principal coordinate analysis of Bray-Curtis dissimilarity reveals compositional differences of the nasal microbiome based on azithromycin treatment for 1 ±1 day (orange) or 5 ±1 days (red) in comparison to no antibiotic exposure (blue). Significance calculated with PERMANOVA. \\u003cstrong\\u003ee\\u003c/strong\\u003e Bray-Curtis dissimilarity distances over time within the nasal microbiome compared to the earliest date of sampling following hospital admission. \\u003cstrong\\u003ef \\u003c/strong\\u003eHeatmap highlighting differentially abundant genera in patients treated with azithromycin for 5 ±1 days compared to those who received no antibiotics or other antibiotics. Taxa enriched with azithromycin treatment are in red, those depleted are in blue. FC = fold change. Stars correspond to statistical significance (*** = p \\u0026lt; 0.001). For boxplots, the box limits correspond to the interquartile range (IQR) and the center line the median. The lower whisker extends to the smallest value within 1.5 * IQR below Q1, and the upper whisker extends to the largest value within 1.5 * IQR above Q3.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/248898629e54e1b463e747de.png\"},{\"id\":85071515,\"identity\":\"58248197-83e7-49a6-a3b1-47fa194e5799\",\"added_by\":\"auto\",\"created_at\":\"2025-06-20 15:40:36\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":332648,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eAzithromycin exposure disrupts the respiratory resistome.\\u003c/strong\\u003e \\u003cstrong\\u003ea \\u003c/strong\\u003eAlpha diversity of the nasal antimicrobial resistome highlighting differences between patients treated with azithromycin (orange), other antibiotics (green) or no antibiotics (blue) at 1 ±1 day (left) or 5 ±1 days (right) of antimicrobial treatment or hospitalization (controls). \\u003cstrong\\u003eb\\u003c/strong\\u003e Compositional differences of the nasal resistome based on azithromycin treatment for 1 ±1 day (orange), azithromycin treatment for 5 ±1 days (dark orange), or no antibiotic treatment (blue). \\u003cstrong\\u003ec\\u003c/strong\\u003e Number of detectably expressed MLS resistance genes (ARGs) based on antibiotic treatment groups. \\u003cstrong\\u003ed\\u003c/strong\\u003e Generalized additive mixed model (GAMM) demonstrating changes in the number of detectably expressed MLS resistance genes over time. \\u003cstrong\\u003ee \\u003c/strong\\u003eGAMM model demonstrating longitudinal changes in the proportional representation of MLS resistance genes (orange) in the nasal resistome over time. \\u003cstrong\\u003ef \\u003c/strong\\u003eNumber of detectably expressed MLS resistance genes after 5 ±1 days of exposure compared to 7 to 10 days after azithromycin cessation. For boxplots, the box limits correspond to the interquartile range (IQR) and the center line the median. The lower whisker extends to the smallest value within 1.5 * IQR below Q1, and the upper whisker extends to the largest value within 1.5 * IQR above Q3. For violin plots, the shape of the violin represents the kernel density estimate of the data, with tails trimmed to the upper and lower ranges of the data. For GAMM plots, a smoothed curve representing the estimated non-linear relationship between variables is plotted. The center line represents the predicted value, and the shaded area denotes the 95% confidence interval.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/facdd6947060fb6133e93059.png\"},{\"id\":85071516,\"identity\":\"7eca2676-71f2-4231-b37e-fd85e8e54067\",\"added_by\":\"auto\",\"created_at\":\"2025-06-20 15:40:36\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":369493,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eCorrelations within the respiratory resistome and microbiome. a \\u003c/strong\\u003ePercent of the MLS resistome across the full cohort comprised by individual MLS ARGs. \\u003cstrong\\u003eb \\u003c/strong\\u003eCorrelations between relative abundance of MLS resistance genes (DPM) and bacterial genera (RPM). Color bar reflects Spearman correlation coefficient. ***p\\u003csub\\u003eadj\\u003c/sub\\u003e \\u0026lt; 0.001; **p\\u003csub\\u003eadj \\u003c/sub\\u003e\\u0026lt; 0.01; *p\\u003csub\\u003eadj \\u003c/sub\\u003e\\u0026lt; 0.05. \\u003cstrong\\u003ec\\u003c/strong\\u003e Network plot demonstrating significant correlations between bacterial taxa and MLS resistance genes. Taxa depicted by blue ovals. Resistance genes depicted by orange hexagons. Weight of edges and color represent spearman correlation coefficient. Solid lines show relationships supported by literature. Dashed lines show relationships where no association has previously been described in the literature.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/5968135fa5c210c2331f2f9b.png\"},{\"id\":104777036,\"identity\":\"5f447448-faf1-4b1e-84ac-c3ec16ad0bec\",\"added_by\":\"auto\",\"created_at\":\"2026-03-17 07:07:02\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":2735157,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/bb21389c-f115-4d5e-ba41-fddb61e039b6.pdf\"},{\"id\":85069837,\"identity\":\"abca62db-687c-4d6e-b0d6-aec4bb5c744c\",\"added_by\":\"auto\",\"created_at\":\"2025-06-20 15:32:36\",\"extension\":\"csv\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1059230,\"visible\":true,\"origin\":\"\",\"legend\":\"Supplemental Table 3\",\"description\":\"\",\"filename\":\"SupplementalTable3NS.csv\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/0ad1d967952eb6ac5559d876.csv\"},{\"id\":85069835,\"identity\":\"b2001e37-7990-461e-9090-444a881998b9\",\"added_by\":\"auto\",\"created_at\":\"2025-06-20 15:32:36\",\"extension\":\"csv\",\"order_by\":2,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":989128,\"visible\":true,\"origin\":\"\",\"legend\":\"Supplemental Table 4\",\"description\":\"\",\"filename\":\"SupplementalTable4PBMC.csv\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/44f177874ec130eed86784db.csv\"},{\"id\":85069836,\"identity\":\"e325ab87-a135-4f3c-bdd4-0f1b504c5f10\",\"added_by\":\"auto\",\"created_at\":\"2025-06-20 15:32:36\",\"extension\":\"pdf\",\"order_by\":3,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1668557,\"visible\":true,\"origin\":\"\",\"legend\":\"Editorial Policy Checklist\",\"description\":\"\",\"filename\":\"nreditorialpolicychecklistNMEDA142753.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/0c380724377c681e2ff64e1e.pdf\"},{\"id\":85069839,\"identity\":\"cbc2ae88-3db7-4704-9cb0-b0c328c25610\",\"added_by\":\"auto\",\"created_at\":\"2025-06-20 15:32:36\",\"extension\":\"docx\",\"order_by\":4,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":353604,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"SupplementaryMaterials.docx\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6875205/v1/4f2d9292f4fff16bccfe1510.docx\"}],\"financialInterests\":\"\\u003cb\\u003eYes\\u003c/b\\u003e there is potential Competing Interest.\\nThe Icahn School of Medicine at Mount Sinai has filed patent applications relating to SARS-CoV-2 serological assays and NDV-based SARS-CoV-2 vaccines which list Florian Krammer as co-inventor. Mount Sinai has spun out a company, Kantaro, to market serological tests for SARS-CoV-2. Florian Krammer has consulted for Merck and Pfizer (before 2020), and is currently consulting for Pfizer, Seqirus, 3rd Rock Ventures, Merck and Avimex. The Krammer laboratory is also collaborating with Pfizer on animal models of SARS-CoV-2. Viviana Simon is a co-inventor on a patent filed relating to SARS-CoV-2 serological assays (the \\\"Serology Assays\\\"). Ofer Levy is a named inventor on patents held by Boston Children’s Hospital relating to vaccine adjuvants and human in vitro platforms that model vaccine action. His laboratory has received research support from GlaxoSmithKline (GSK) and is a co-founder of and advisor to ARMR Sciences (formerly Ovax, Inc) that develops technologies to detect illicit substances and prevent overdose. He is a consultant to GSK and Sanofi. Charles Cairns serves as a consultant to bioMerieux and is funded for a grant from Bill \\u0026 Melinda Gates Foundation. James A Overton is a consultant at Knocean Inc. Jessica Lasky-Su serves as a scientific advisor of Precion Inc. Scott R. Hutton, Greg Michelloti and Kari Wong are employees of Metabolon Inc. Vicki Seyfer- Margolis is a current employee of MyOwnMed. Nadine Rouphael reports grants or contracts with Merck, Sanofi, Pfizer, Vaccine Company, Quidel, Lilly and Immorna, and has participated on data safety monitoring boards for Moderna, Sanofi, Seqirus, Pfizer, EMMES, ICON, BARDA, Imunon, CyanVac and Micron. Nadine Rouphael has also received support for meetings/travel from Sanofi and Moderna and honoraria from Virology Education. Adeeb Rahman is a current employee of Immunai Inc. Steven Kleinstein is a consultant related to ImmPort data repository for Peraton. Nathan Grabaugh is a consultant for Tempus Labs and the National Basketball Association. Akiko Iwasaki is a consultant for 4BIO, Blue Willow Biologics, Revelar Biotherapeutics, RIGImmune, Xanadu Bio, Paratus Sciences. Monika Kraft receives research funds paid to her institution from NIH, ALA; Sanofi, Astra-Zeneca for work in asthma, serves as a consultant for Astra-Zeneca, Sanofi, Chiesi, GSK for severe asthma; is a co-founder and CMO for RaeSedo, Inc, a company created to develop peptidomimetics for treatment of inflammatory lung disease. Esther Melamed received research funding from Babson Diagnostics and honorarium from Multiple Sclerosis Association of America and has served on the advisory boards of Genentech, Horizon, Teva, and Viela Bio. Carolyn Calfee receives research funding from NIH, FDA, DOD, Roche-Genentech and Quantum Leap Healthcare Collaborative as well as consulting services for Janssen, Vasomune, Gen1e Life Sciences, NGMBio, and Cellenkos. Wade Schulz was an investigator for a research agreement, through Yale University, from the Shenzhen Center for Health Information for work to advance intelligent disease prevention and health promotion; collaborates with the National Center for Cardiovascular Diseases in Beijing; is a technical consultant to Hugo Health, a personal health information platform; cofounder of Refactor Health, an AI-augmented data management platform for health care; and has received grants from Merck and Regeneron Pharmaceutical for research related to COVID-19. Grace A McComsey received research grants from Redhill, Cognivue, Pfizer, and Genentech, and served as a research consultant for Gilead, Merck, Viiv/GSK, and Jenssen. Linda N. Geng received research funding paid to her institution from Pfizer, Inc.\",\"formattedTitle\":\"Empiric Azithromycin in COVID-19 Impacts the Respiratory Microbiome and Antimicrobial Resistome without Anti-inflammatory Benefit\",\"fulltext\":[{\"header\":\"Introduction\",\"content\":\"\\u003cp\\u003eAntimicrobial resistance (AMR) is one of the top global health threats facing humanity\\u003csup\\u003e1\\u003c/sup\\u003e and is increasingly hindering the effective treatment of respiratory infections\\u003csup\\u003e2\\u003c/sup\\u003e. Rates of hospital-onset AMR infections dramatically increased during the SARS-CoV-2 pandemic\\u003csup\\u003e3\\u003c/sup\\u003e, complicating the treatment of COVID-19 and reversing the prior downward trend in deaths from drug resistant pathogens\\u003csup\\u003e4\\u003c/sup\\u003e. While the underlying reasons for this were multifactorial, the overuse of broad spectrum antibiotics in COVID-19 patients was a notable contributor\\u003csup\\u003e3,5–9\\u003c/sup\\u003e.\\u003c/p\\u003e\\n\\u003cp\\u003eAzithromycin, a World Health Organization essential medicine\\u003csup\\u003e10\\u003c/sup\\u003e, is one of the most widely used antibiotics in human healthcare with \\u0026gt; 40 million prescriptions annually in the United States (U.S.) alone\\u003csup\\u003e11\\u003c/sup\\u003e. Azithromycin overuse has been well documented in the outpatient setting\\u003csup\\u003e12\\u003c/sup\\u003e, where an estimated 30% of antibiotic prescriptions are inappropriate\\u003csup\\u003e13\\u003c/sup\\u003e. During the first year of the COVID-19 pandemic, azithromycin became one of the most commonly used antibiotics in hospitalized patients as well\\u003csup\\u003e9,14\\u003c/sup\\u003e. This was driven in part by early studies suggesting possible antiviral activity\\u003csup\\u003e15,16\\u003c/sup\\u003e, and prior work demonstrating anti-inflammatory properties of macrolide antibiotics\\u003csup\\u003e17,18\\u003c/sup\\u003e. Randomized clinical trials, however, subsequently demonstrated that azithromycin conferred no clinical benefit in the treatment of COVID-19\\u003csup\\u003e19,20\\u003c/sup\\u003e. Nonetheless, many medical centers initially incorporated azithromycin into their COVID-19 treatment guidelines, and public misinformation continues to drive overprescription of the drug\\u003csup\\u003e21\\u003c/sup\\u003e.\\u003c/p\\u003e\\n\\u003cp\\u003eRecent work has found that azithromycin exposure can alter the human microbiome and its reservoir of antimicrobial resistance genes, termed the resistome\\u003csup\\u003e22–24\\u003c/sup\\u003e. For instance, secondary analyses of the MORDOR (Macrolide Oraux pour Réduire les Décès avec un Oeil sur la Résistance) clinical trials found that biannual mass azithromycin distribution to African children led to an increase in the abundance of both macrolide and other AMR gene (ARGs) classes in the gut microbiome\\u003csup\\u003e22\\u003c/sup\\u003e. In addition, adults with asthma randomized to thrice weekly azithromycin over 12 months had an increase in PCR copy number of macrolide resistance genes in sputum samples compared to controls\\u003csup\\u003e25\\u003c/sup\\u003e.\\u003c/p\\u003e\\n\\u003cp\\u003eDespite being the most common scenario for its use, no studies have yet assessed the impact of azithromycin on the respiratory microbiome in the context of empiric prescription for acute respiratory infection. Furthermore, no studies of azithromycin exposure have yet incorporated metatranscriptomics, which can assess both bacterial 16S rRNA abundance and ARG expression, providing a functional profile of the actively expressed resistome\\u003csup\\u003e26,27\\u003c/sup\\u003e.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eTo address these gaps, we carry out respiratory metatranscriptomics in a prospective cohort of 1164 adults hospitalized for COVID-19, and study the impacts of azithromycin exposure. We find marked changes in the respiratory microbiome, including increases in detectably expressed macrolide resistance genes and their proportional representation in the airway resistome, without evidence of antiviral or immune modulating benefit. Taken together, our findings offer new insights into the adverse effects and biological consequences of empiric azithromycin exposure during viral infection.\\u003c/p\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eCohort\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe carried out a prospective observational study of 1164 adults hospitalized for COVID-19 enrolled in the multicenter IMmuno Phenotyping Assessment in a COVID-19 Cohort (IMPACC)\\u003csup\\u003e28\\u0026ndash;30\\u003c/sup\\u003e between May 2020 and March 2021 (\\u003cstrong\\u003eFig. 1a\\u003c/strong\\u003e). Previously established COVID-19 outcome trajectory groups (TGs)\\u003csup\\u003e29\\u003c/sup\\u003e were utilized to group patients based on disease severity. TGs ranged from 1 (lowest severity) to 5 (death within 28 days)\\u003csup\\u003e29\\u003c/sup\\u003e. Administration of azithromycin and other antibiotics was tracked following admission and throughout hospitalization. Of 1164 COVID-19 patients studied, 366 (31.4%) were treated empirically with azithromycin \\u0026plusmn; other antibiotics (Azithro group), 474 (40.7%) received no antibiotics (No-Abx group), and 324, (27.8%) received antibiotics other than azithromycin (Other-Abx group) (\\u003cstrong\\u003eSupp. Table 1\\u003c/strong\\u003e \\u003cstrong\\u003eand 2\\u003c/strong\\u003e).\\u003c/p\\u003e\\n\\u003cp\\u003eEmpiric azithromycin administration was greatest among patients with the highest COVID-19 severity (\\u003cstrong\\u003eFig. 1b\\u003c/strong\\u003e), although compared to those in the Other-Abx group, azithromycin-treated patients had less severe disease (\\u003cstrong\\u003eSupp. Table 1\\u003c/strong\\u003e). The median number of azithromycin treatment days was 2 (IQR 1-4 days, range 1-35 days), which significantly differed across TGs (p = 0.03, \\u003cstrong\\u003eSupp. Fig. 1\\u003c/strong\\u003e). Azithromycin was administered in most (98.2%) patients within 1 week of hospital admission (\\u003cstrong\\u003eFig. 1c\\u003c/strong\\u003e). Patients treated with azithromycin were most likely to have been co-administered ceftriaxone (77.2%) or vancomycin (19.4%) (\\u003cstrong\\u003eFig. 1d\\u003c/strong\\u003e). Sex or race did not differ based on azithromycin usage (\\u003cstrong\\u003eSupp. Table 1\\u003c/strong\\u003e).\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eImpact of azithromycin exposure on the respiratory microbiome\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe first examined the impact of azithromycin on the upper respiratory tract microbiome using metatranscriptomic RNA sequencing (RNA-seq) of nasal swab (NS) samples collected at six timepoints over 28 days following hospital admission. These analyses were adjusted for age\\u0026nbsp;quintile, sex, severity TG, days of hospitalization, patient, receipt of corticosteroids, and receipt of the six most common antibiotics aside from azithromycin. We found that azithromycin treatment for 5 \\u0026plusmn;1 days, a common duration of prescription\\u003csup\\u003e31\\u003c/sup\\u003e, was associated with a significant decrease in bacterial abundance in the airway (adjusted p value (p\\u003csub\\u003eadj\\u003c/sub\\u003e) = 0.026), with an effect observable within 1 \\u0026plusmn;1 day (p\\u003csub\\u003eadj\\u003c/sub\\u003e = 0.0019, \\u003cstrong\\u003eFig. 2a\\u003c/strong\\u003e). Other antibiotics also led to a decrease in upper respiratory bacterial abundance after 1 \\u0026plusmn;1 days (p\\u003csub\\u003eadj\\u003c/sub\\u003e = 0.036) but not at the later timepoint (\\u003cstrong\\u003eFig. 2a\\u003c/strong\\u003e).\\u003c/p\\u003e\\n\\u003cp\\u003eAssessment of the mycobiome demonstrated that receipt of azithromycin was associated with an increase in fungal relative abundance in the upper airway after 1 \\u0026plusmn;1 days (padj = 0.038, \\u003cstrong\\u003eFig. 2b\\u003c/strong\\u003e), with a time-dependent increasing trend observed over 5 days of azithromycin administration (Supp. \\u003cstrong\\u003eFig. 2b\\u003c/strong\\u003e). No differences in upper airway microbiome alpha diversity were observed based on azithromycin treatment status after 5 \\u0026plusmn;1 days, although a significant increase was seen early on (p\\u003csub\\u003eadj\\u003c/sub\\u003e = 0.029, \\u003cstrong\\u003eFig. 2c\\u003c/strong\\u003e). Significant differences were found in the microbiome community composition based on azithromycin expsoure, measured by Bray-Curtis dissimilarity index (PERMANOVA p = 0.001, \\u003cstrong\\u003eFig. 2d\\u003c/strong\\u003e). A comparison of the trajectories for Bray-Curtis distances versus the earliest timepoint for each patient demonstrated marked shifts in community composition over time, independent of antibiotic exposure (\\u003cstrong\\u003eFig. 2e\\u003c/strong\\u003e).\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eDifferential taxonomic abundance analysis demonstrated that azithromycin exposure was associated with enrichment of potentially pathogenic taxa in the upper airway including \\u003cem\\u003eStaphylococcus\\u0026nbsp;\\u003c/em\\u003eand \\u003cem\\u003eKlebsiella\\u003c/em\\u003e species, and depletion of several typically commensal taxa such as \\u003cem\\u003eNeisseria\\u003c/em\\u003e and \\u003cem\\u003eFusobacterium\\u003c/em\\u003e (\\u003cstrong\\u003eFig. 2f\\u003c/strong\\u003e). We also tested whether azithromycin treatment associated with any changes in SARS-CoV-2 relative abundance in the upper airway. We observed no differences with respect to either the Other Abx or No Abx groups after 5 \\u0026plusmn;1 days of treatment (\\u003cstrong\\u003eSupp. Fig. 3\\u003c/strong\\u003e).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eImpact of azithromycin exposure on the respiratory antimicrobial resistome\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eTo understand the impact of azithromycin exposure on the respiratory antimicrobial resistome, we first compared Shannon Diversity Index across groups. Azithromycin exposure was associated with an increase in the alpha diversity of the upper airway resistome at the early (1 \\u0026plusmn;1 day) treatment timepoint compared with the Other-Abx group (p\\u003csub\\u003eadj\\u0026nbsp;\\u003c/sub\\u003e= 5.8e-3) but not with the No Abx controls (\\u003cstrong\\u003eFig. 3a\\u003c/strong\\u003e). A shift in the composition of the antimicrobial resistome upon azithromycin exposure was also evident, with significant differences in Bray Curtis dissimilarity indices observed at 1 \\u0026plusmn;1 day and 5 \\u0026plusmn;1 days of azithromycin treatment versus controls (PERMANOVA p = 0.001, \\u003cstrong\\u003eFig. 3b\\u003c/strong\\u003e).\\u003c/p\\u003e\\n\\u003cp\\u003eSince azithromycin is a macrolide class antibiotic, we next focused on ARGs conferring resistance to the macrolide, lincosamide and streptogramin (MLS) class of antibiotics. We found that azithromycin exposure was associated with a significantly greater number of detectably expressed MLS genes compared to the No-Abx or Other-Abx groups at both the early (p\\u003csub\\u003eadj\\u003c/sub\\u003e = 3.5e-4 and 6.2e-7, respectively) and late (p\\u003csub\\u003eadj\\u003c/sub\\u003e = 1.3e-3 and 4.3 e-6, respectively) timepoints (\\u003cstrong\\u003eFig. 3c\\u003c/strong\\u003e). Longitudinal modeling subsequently demonstrated that days of azithromycin exposure resulted in a significant increase in the number of detectably expressed MLS ARGs in the upper respiratory microbiome (p\\u003csub\\u003eadj\\u003c/sub\\u003e = 7.6e-7, \\u003cstrong\\u003eFig. 3d\\u003c/strong\\u003e).\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eTo comparatively evaluate the impact of azithromycin exposure across different ARG classes, we assessed longitudinal changes in the fraction of the resistome represented by each class. We found that after 5 \\u0026plusmn;1 days of azithromycin exposure, MLS ARGs increased from 24.5% to 42.9% of the resistome (p\\u003csub\\u003eadj\\u003c/sub\\u003e = 1.7e-4, \\u003cstrong\\u003eFig. 3e\\u003c/strong\\u003e). Notably, the enrichment in MLS ARGs, both in terms of ARG richness and as a percent of the resistome, persisted even 7-10 days after cessation of azithromycin (\\u003cstrong\\u003eFig. 3f and Supp. Fig. 4\\u003c/strong\\u003e). \\u003cstrong\\u003e\\u0026nbsp;\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCorrelations within the resistome and microbiome\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eTo assess relationships between macrolide resistance genes and bacterial taxa within the upper airway microbiome, we performed multi-dimensional correlation analyses (\\u003cstrong\\u003eFig. 4b\\u003c/strong\\u003e). Significant positive correlations were found between several MLS genes and both potentially pathogenic (e.g., \\u003cem\\u003eStaphylococcus, Streptococcus)\\u0026nbsp;\\u003c/em\\u003eas well as common commensal (e.g., \\u003cem\\u003eCorynebacterium\\u003c/em\\u003e\\u003csup\\u003e32\\u003c/sup\\u003e\\u003cem\\u003e, Dolosigranulum\\u003c/em\\u003e\\u003csup\\u003e33\\u003c/sup\\u003e\\u003cem\\u003e)\\u0026nbsp;\\u003c/em\\u003egenera (\\u003cstrong\\u003eFigs. 4b, 4c\\u003c/strong\\u003e).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eImpact of azithromycin on host inflammatory responses\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003ePrior studies have found that azithromycin can confer anti-inflammatory properties, leading to off-label use of the antibiotic as an immune modulatory agent\\u003csup\\u003e17,18\\u003c/sup\\u003e. We thus sought to understand whether azithromycin exposure in the setting of acute COVID-19 was associated with changes in host inflammatory gene expression in the airway or blood by carrying out differential gene expression analyses. No differentially expressed genes were identified (false discovery rate \\u0026lt; 0.05) in either anatomical compartment after 5 \\u0026plusmn;1 days of treatment (\\u003cstrong\\u003eSupp. Tables 3, 4\\u003c/strong\\u003e), suggesting that azithromycin does not meaningfully attenuate pathologic inflammatory responses in hospitalized COVID-19 patients.\\u003c/p\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cp\\u003eIn a large multicenter cohort of hospitalized COVID-19 patients, empiric azithromycin treatment was associated with changes in the upper respiratory tract microbiome, mycobiome and antimicrobial resistome. We observed a significant expansion of detectably expressed macrolide resistance genes after 5 ±1 days of azithromycin treatment, with effects in some cases observed within a few days. In addition, we found that azithromycin treatment was associated with changes in the composition of the upper airway microbiota including enrichment of \\u003cem\\u003eKlebsiella\\u0026nbsp;\\u003c/em\\u003eand \\u003cem\\u003eStaphylococcus\\u0026nbsp;\\u003c/em\\u003especies, and an increase in the burden of fungal taxa in the upper respiratory tract. Together, our findings demonstrate that inappropriate azithromycin use in patients with viral respiratory infections can drive expansion of macrolide resistance determinants and disrupt the composition of the airway microbiome.\\u003c/p\\u003e\\n\\u003cp\\u003ePrior work examining mass azithromycin treatment in African children\\u003csup\\u003e22,24\\u003c/sup\\u003e found a concerning relationship between exposure to this drug and an increase in macrolide resistance genes in the gut microbiome after four years. We build on these important findings by demonstrating effects in the respiratory tract and at the transcriptional level, detectable within a few days of antibiotic treatment. Importantly, we find that azithromycin leads to not only an increase in the potential for resistance within the microbiome, but a functional impact on the expression of MLS resistance genes.\\u003c/p\\u003e\\n\\u003cp\\u003eMacrolide-resistant \\u003cem\\u003eStreptococcus pnuemoniae\\u003c/em\\u003e and \\u003cem\\u003eStreptococcus pyogenes\\u003c/em\\u003e are considered urgent threats by the U.S. Centers for Disease Control and Prevention\\u003csup\\u003e34\\u003c/sup\\u003e. Perhaps it is not surprising that macrolide resistance in these species has increased over the past decade given our results, and considering that 30% of antibiotics prescribed in outpatient settings have been deemed inappropriate or unnecessary\\u003csup\\u003e13\\u003c/sup\\u003e. Given that a large fraction of azithromycin prescriptions are written for children\\u003csup\\u003e12\\u003c/sup\\u003e, this is particularly concerning, as they may become colonized with macrolide-resistance bacteria at an early age due to unnecessary exposure to this drug.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eAzithromycin is used prophylactically for chronic obstructive pulmonary disease\\u003csup\\u003e35\\u003c/sup\\u003e, cystic fibrosis/bronchiectasis\\u003csup\\u003e36\\u003c/sup\\u003e, lung transplantation\\u003csup\\u003e37\\u003c/sup\\u003e, HIV/AIDS\\u003csup\\u003e38\\u003c/sup\\u003e and other conditions. While few studies have examined the impact of azithromycin prophylaxis on the respiratory resistome, a recent study of asthma patients found an increase in macrolide resistance genes using multiplex PCR in the sputum microbiome after 12 months\\u003csup\\u003e25\\u003c/sup\\u003e. Further work is needed to understand the impact of azithromycin prophylaxis on the upper and lower respiratory microbiome and resistome in these patient populations.\\u003c/p\\u003e\\n\\u003cp\\u003eA sub-analysis of the MORDOR trial found that four years of biannual azithromycin treatment in African children reduced mortality but led to an increase in macrolide resistant \\u003cem\\u003eStreptococcus pneumoniae\\u003c/em\\u003e cultured from the nasopharynx\\u003csup\\u003e24\\u003c/sup\\u003e. Consistent with these prior microbiological observations, our correlation analysis suggested that both \\u003cem\\u003eStreptococcus\\u0026nbsp;\\u003c/em\\u003eand \\u003cem\\u003eStaphylococcus\\u0026nbsp;\\u003c/em\\u003especies, encompassing some of the most important bacterial pneumonia pathogens, may harbor these resistance determinants. In addition, we found relationships between MLS resistance gene expression and the abundance of commensal and contextually pathogenic taxa, such as \\u003cem\\u003eCorynebacterium\\u003c/em\\u003e species.\\u003c/p\\u003e\\n\\u003cp\\u003eIn the MORDOR trial, mass azithromycin treatment was also found to cause an increase in the burden of non-macrolide resistance genes in the gut microbiome\\u003csup\\u003e22\\u003c/sup\\u003e. In contrast, we found relatively few off-target effects on other classes of ARGs in the upper airway. One possible explanation may lie in the age and demographic differences of the studied populations. For instance, the burden of ARGs in both the gut\\u003csup\\u003e39,40\\u003c/sup\\u003e and respiratory microbiome increases with age, potentially due to lifetime antibiotic exposures\\u003csup\\u003e26\\u003c/sup\\u003e. It is possible that our cohort of hospitalized adults in the U.S. may have had a higher baseline burden of ARGs compared to African children living in rural settings. Alternatively, differences may be attributable to sampling of the respiratory versus gut microbiome, or the use of metatranscriptomics versus metagenomic DNA sequencing.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003ePrior work has demonstrated that azithromycin has immune-modulating potential\\u003csup\\u003e17,18\\u003c/sup\\u003e, findings that have encouraged its use in patients with cystic fibrosis, chronic obstructive pulmonary disease\\u003csup\\u003e41\\u003c/sup\\u003e, and other inflammatory diseases. Azithromycin use early during the COVID-19 pandemic was in part driven by the idea that it might attenuate harmful inflammatory responses, as well as a now-retracted publication purporting clinical benefit when combined with hydroxychloroquine\\u003csup\\u003e42\\u003c/sup\\u003e. Randomized clinical trials, however, subsequently found no therapeutic benefit of azithromycin for COVID-19\\u003csup\\u003e19,20\\u003c/sup\\u003e. Consistent with this, we observed no significant associations between azithromycin treatment and inflammatory gene expression, or viral load, in either the airway or blood of hospitalized COVID-19 patients.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eStrengths of our study include a large multicenter cohort, detailed clinical phenotyping, use of respiratory metatranscriptomics, and rigorous quality control of clinical and biological data. As with any research, our study also has limitations. These include the observational study design and the use of short read sequencing, which precluded definitively linking macrolide resistance genes to specific taxa. In addition, our analyses were limited to the upper airway and thus may not reflect microbial changes occurring in the lungs.\\u003c/p\\u003e\\n\\u003cp\\u003eAntibiotic administration data were extracted manually by clinical research coordinators at each study site, an approach susceptible to human error. Azithromycin treatment, however, was confirmed by an independent adjudicator for every patient. \\u0026nbsp; Future randomized clinical trials - \\u0026nbsp;ideally that include airway and gut microbiome sampling as well as bacterial culture - are needed to more fully characterize the impact of exposure to azithromycin and other antibiotics on the human microbiome and resistome.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eIn sum, we find that azithromycin exposure in hospitalized COVID-19 patients is associated with compositional changes in the airway microbiome and expansion of macrolide resistance genes. Taken together our findings suggest that empiric use of azithromycin in patients with viral respiratory infections may lead to adverse impacts and contribute to antimicrobial resistance.\\u003cstrong\\u003e\\u003cbr\\u003e\\u0026nbsp;\\u003c/strong\\u003e\\u003c/p\\u003e\"},{\"header\":\"Methods\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eStudy Design, Clinical Cohort, Inclusion and Ethics\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eIMPACC is a prospective longitudinal study that enrolled 1164 hospitalized COVID-19 patients, as previously described in detail\\u003csup\\u003e28–30,43,44\\u003c/sup\\u003e. Participants 18 years and older were recruited from 20 hospitals across 15 academic biomedical centers within the United States and confirmed to be SARS-CoV-2 positive by reverse transcription PCR (RT-PCR) testing. No participants were vaccinated for SARS-CoV-2 at time of enrollment nor during their hospitalization. To better categorize patients into different COVID-19 severity groups, they were classified into one of five trajectory groups using latent class mixed modeling of the degree of respiratory illness and external oxygen administration\\u003csup\\u003e29\\u003c/sup\\u003e.\\u003c/p\\u003e\\n\\u003cp\\u003eThe Department of Health and Human Services Office for Human Research Protections (OHRP) and NIAID concurred that the IMPACC study qualified for public health surveillance exemption. The study protocol was reviewed by each site’s institutional review board (IRB), with twelve sites conducting as a public health surveillance study, and three sites integrating the IMPACC study into IRB-approved protocols (The University of Texas at Austin, IRB 2020-04-0117; University of California San Francisco, IRB 20-30497; Case Western Reserve University, IRB STUDY20200573) with participants providing informed consent. Participants enrolled at sites operating as a public health surveillance study were provided information sheets describing the study including the samples to be collected and plans for analysis and data de-identification. Participants who requested not to participate after review of the study plan and information were not enrolled. Participants were not compensated while hospitalized but were subsequently compensated for outpatient visits and surveys. This study was registered at clinicaltrials.gov (NCT0438777) and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eMetatranscriptomic Sequencing\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eMid-turbinate nasal swabs were collected withing 72 hours of hospital admission and at subsequent visits with target dates of 4, 7, 14, 21, and 28 days post hospital admission. The nasal swabs were stored in 1 mL of Zymo-DNA/RNA shield reagent (Zymo Research), before RNA was extracted twice in parallel from 250 uL of sample. The RNA was then purified with the KingFisher Flex sample purification system (ThermoFisher) and the quick DNA-RNA MagBead kit (Zymo Research). The duplicated RNA was pooled and aliquoted at 20 μL for the downstream RNA-sequencing. The isolated RNA was subsequently sequenced on a NovaSeq 6000 (Illumina) at 100 bp paired-end read length. The data was aligned using STAR (v2.4.3) against the GRCh38 reference genome and host gene counts were quantified using HTSeq-count (v0.4.1). \\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eMicrobiome and Resistome Profiling\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eMetatranscriptomic data was processed using the open-source CZ ID pipeline (https://czid.org/)\\u003csup\\u003e45\\u003c/sup\\u003e. Microbiome profiling was performed within CZ ID using the Illumina mNGS pipeline (v7.1). CZ ID first removes host reads by aligning to the GRCh38 human reference genome using STAR\\u003csup\\u003e46\\u003c/sup\\u003e. Adapters are removed using Trimmomatic\\u003csup\\u003e47\\u003c/sup\\u003e and reads are filtered for low quality using PriceSeq\\u003csup\\u003e48\\u003c/sup\\u003e. Duplicates are identified using czid-dedup. Low complexity reads are filtered out using the Lempel-Ziv-Welch algorithm and any remaining host reads are removed by alignment to GRCh38 using Bowtie2\\u003csup\\u003e49\\u003c/sup\\u003e. After performing the filtering steps, de-duplicated reads are subsampled to 2 million total reads. Taxonomic classification of remaining reads is performed on reads and assembled contigs (SPAdes)\\u003csup\\u003e50\\u003c/sup\\u003e by aligning to the NCBI nucleotide (NT) and protein (NR) databases. Resistome profiles were generated using the CZ ID AMR pipeline (v0.2), which leverages the Resistance Gene Identifier (RGI) tool and the Comprehensive Antibiotic Resistance Database (CARD)\\u003csup\\u003e51\\u003c/sup\\u003e. Following microbiome and resistome profiling with CZ ID, background and batch correction were performed to remove contaminants and adjust for batch effects (see below). To limit the contribution of spurious hits to downstream analyses, additional filtering of microbial taxa and ARGs was performed on a per sample basis. Microbial taxa were excluded if they did not meet all of the following criteria: (1) ten hits to the NT database, (2) one hit to the NR database, (3) a minimum alignment length of 50 bases. ARGs were excluded if they did not have ≥5% read coverage breadth and meet one of the following two criteria: (1) present in ≥5% of all samples with detectable ARGs (2) depth per million (DPM) ≥1 and ≥10 hits. \\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eBackground \\u0026amp; Batch Correction\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eNegative controls (double distilled water) were processed and sequenced alongside samples in the IMPACC cohort to enable the characterization and subtraction of background contamination. The sequencing data generated from these samples was analyzed using the CZ ID metagenomic and AMR pipelines as described above. Background and batch correction was performed on the microbiome and resistome datasets separately. A negative binomial model was used to model the distribution of reads of microbial taxa/ARGs in the negative controls. Mean and dispersion parameters were then fitted to these data. Mean estimates were generated for each batch:taxon or batch:ARG pair in the negative controls, where batch corresponds to the phase of the IMPACC study (1, 2, 3A or 3B). The MASS package (v7.3.58.1) in R was used to generate a single dispersion parameter across all taxa/ARGs. P-values were adjusted for multiple testing using the Benjamini-Hochberg False Discovery Rate algorithm. \\u003c/p\\u003e\\n\\u003cp\\u003eMicrobial taxa/ARGs that were present at a significantly higher abundance in participant samples than in negative controls (FDR \\u0026lt;0.1) were retained for downstream analyses. \\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cbr\\u003e \\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eSingle Time Point Analyses\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eParticipants were assigned to one of three groups: those who received Azithromycin (Azithro) +/- other antibiotics, those who took only non-Azithromycin antibiotics (Other-Abx), or participants who did not receive antibiotics (No-Abx). Participants with only partially captured antibiotic start and stop dates were excluded from these analyses. The 2 time points studied for these groups were samples collected with 1 ±1 day of exposure and 5 ±1 days of exposure, applicable to the Azithromycin and Other-Abx groups. The second time point (5 ±1 days) was chosen based on the average course of azithromycin, approximately 5 days. In an attempt to match the No-Abx samples to the Azithromycin and Other-Abx samples at each time point, the distribution of days of hospitalization across those groups was examined. Based on the findings, we selected samples for the No-Abx group as follows: In the 1 ±1 day group, the samples had no antibiotic exposure and up to 40 days of hospitalization. In the 5 ±1 days group, the samples had no antibiotic exposure, and between 3-40 days of hospitalization. Samples that qualified for both time points were split evenly between time points. The number of \\\"No-Abx\\\" samples was also rarified (50%) to make the numbers in each group more comparable. In addition to this sample matching process for the No-Abx group, days from hospital admission was included as a covariate in statistical models to further control for slight differences. A linear mixed-effects model (using the lme4 package) was used to calculate differences between the groups at both timepoints while using age quintile, sex, severity TG, days from hospitalization, and receipt of steroids as fixed effects in the model, in combination with the participant’s enrollment site included as a random effect. For only the Azithromycin group, the number of days of the six most common co-administered antibiotics (vancomycin, ceftriaxone, cefepime, piperacillin-tazobactam, doxycycline and meropenem) were also included as fixed effects. Resulting p values were adjusted using the Benjamini-Hochberg False Discovery Rate algorithm via the p_adjust() function of the stats (v4.2.3) package.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cbr\\u003e \\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eGeneralized Additive Mixed Model (GAMM) Analyses\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003ch4\\u003eVarious metrics were modeled over days of azithromycin exposure using the nlme (v3.1-162), lme4 (v1.1-35.5), lmerTest (v3.1-3) and ggeffects (v1.7.2) packages in R. Samples were limited to the first 10 days of hospitalization and the first 5 days of antibiotic exposure due to few patients with longer antibiotic courses. The GAMM models included fixed effects of severity TG, sex, age quintile of the participant, and whether they ever received steroids, in addition to a smoothed term for days of azithromycin usage and days from hospitalization, and participant as a mixed effect. Additionally, days of administration for the six most prevalent antibiotics in the cohort aside from azithromycin (vancomycin, ceftriaxone, cefepime, piperacillin-tazobactam, doxycycline, and meropenem) were also added to the models as smooth terms. \\u003c/h4\\u003e\\n\\u003cp\\u003eR model formula:\\u003c/p\\u003e\\n\\u003cp\\u003e~ s(Azithromycin, bs = 'cr', k=4)+s(Ceftriaxone, bs = 'cr', k=4)+s(Vancomycin, bs = 'cr', k=4)+s(Cefepime, bs = 'cr', k=4)+s(Meropenem, bs = 'cr', k=4)+s(Piperacillin_Tazobactam, bs = 'cr', k=4)+s(Doxycycline, bs = 'cr', k=4)+s(event_date, bs = 'cr')+ trajectory_group + sex + discretized_admit_age_quantile + ever_steroids\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eMicrobiome Diversity Metrics\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAlpha diversity (Shannon Diversity Index) was calculated using the diversity() function in the R package vegan(2.6-6.1). Beta diversity (Bray-Curtis dissimilarity) analysis was performed using the vegan functions vegdist(), betadisper(), permutest() and adonis2(). The beta diversity analysis was adjusted for age quintile, sex, days since hospitalization, severity TG, patient, and receipt of corticosteroids in the adonis2() function. For the resistome analysis, receipt of the six most common antibiotics aside from azithromycin (vancomycin, ceftriaxone, cefepime, piperacillin-tazobactam, doxycycline and meropenem) was also included in the model. Principal Component Analysis (PCoA) of the resistome and bacterial microbiome was also performed using cmdscale() function of the stats(v4.2.3) package. \\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eDifferential Abundance Analysis\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eBacterial microbiome profile data was converted into a phyloseq object using R packages ape (v5.8.1), ade4 (v1.7.22), and phyloseq (v1.34.0). Differential abundance analysis was performed using the R package ANCOMBC (v1.0.5) with an alpha level of 0.05 and prevalence filter (zero_cut argument) of 0.97. The analysis was adjusted for the following covariates: age quintile, sex, days from hospitalization, severity TG, patient, receipt of corticosteroids and receipt of the six most prevalent antibioticsaside from azithromycin (vancomycin, ceftriaxone, cefepime, piperacillin-tazobactam, doxycycline, and meropenem). P-values were adjusted for multiple testing using the Benjamini-Hochberg correction. \\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCorrelation Analyses\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe 50 most abundant bacterial taxa by reads per million (RPM) were included in the correlation analysis. RPM was used as an abundance metric in lieu of raw reads to adjust for variations in sequencing depth between samples. The abundance metric used for ARGs was depth per million (DPM), which adjusts for both variations in sequencing depth between samples and variations in gene length between ARGs. Spearman correlation was performed between bacterial taxa RPM and ARG DPM using the cor() function in the stats (v4.2.3) package and the cor_pmat() function in the rstatix (v0.7.2) package in R. Results were visualized using the corrplot (v0.95) package.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003ePBMC and Nasal Swab Host Transcriptional Profiling\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eRNA-seq and alignment against the host transcriptome was performed as previously described,\\u003csup\\u003e52,53\\u003c/sup\\u003e and the deidentified, quality-controlled raw gene count files and metadata were obtained from the IMPACC study. We filtered for host protein-coding genes that had at least 10 counts in at least 20% of the samples. To identify genes that were associated with azithromycin exposure compared to No-Abx patients, we utilized the limma (v3.58.1) package in R. We modeled gene expression as a function of azithromycin exposure, corrected for age, sex, trajectory group, days since hospitalization, and exposure to corticosteroids. Then, we applied voom and duplicateCorrelation for two iterations to calculate the correlation between multiple samples from the same patients. Next, we applied the lmFit and eBayes functions to estimate the effects of Azithromycin exposure on gene expression and their P-values. Finally, the P-values were adjusted with Benjamini-Hochberg correction.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eStatistics\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAll code was written in R v4.2.3 or R 4.0.3. Data processing was performed using R packages dplyr (v1.1.4) and tidyverse (v2.0.0). Plots were generated using R packages ggplot2 (v3.5.1), ggpubr (v0.6.0), scales (1.3.0) and ggbeeswarm (v0.7.2). Package versions for each analysis are reported in the code repository.\\u003c/p\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eData and code availability\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eData used in this study is available at ImmPort Shared Data under the accession number SDY1760 and in the NLM\\u0026rsquo;s Database of Genotypes and Phenotypes (dbGaP) under the accession number phs002686.v2.p2. Source data for each figure are provided with this manuscript in the Source Data file. All code is deposited in the following Bitbucket repository: https://bitbucket.org/kleinstein/impacc-public-code/src/master/azithromycin_manuscript/ .\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eFunding\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eNIH (5R01AI135803-03, 5U19AI118608-04, 5U19AI128910-04, 4U19AI090023-11, 4U19AI118610-06, R01AI145835-01A1S1, 5U19AI062629-17, 5U19AI057229-17, 5U19AI125357-05, 5U19AI128913-03, 3U19AI077439-13, 5U54AI142766-03, 5R01AI104870-07, 3U19AI089992-09, 3U19AI128913-03, 5T32DA018926-18, and K0826161611); NIAID, NIH (3U19AI1289130, U19AI128913-04S1, R01AI122220, and 1K23AI185326-01); NCATS (UM1TR004528), and National Science Foundation (DMS2310836). Funding sources did not have a direct role in the design, analysis, or approval of this manuscript.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eIMPACC Network Competing Interests\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe Icahn School of Medicine at Mount Sinai has filed patent applications relating to SARS-CoV-2 serological assays and NDV-based SARS-CoV-2 vaccines which list Florian Krammer as co-inventor. Mount Sinai has spun out a company, Kantaro, to market serological tests for SARS-CoV-2. Florian Krammer has consulted for Merck and Pfizer (before 2020), and is currently consulting for Pfizer, Seqirus, 3rd Rock Ventures, Merck and Avimex. The Krammer laboratory is also collaborating with Pfizer on animal models of SARS-CoV-2. Viviana Simon is a co-inventor on a patent filed relating to SARS-CoV-2 serological assays (the \\u0026quot;Serology Assays\\u0026quot;). Ofer Levy is a named inventor on patents held by Boston Children\\u0026rsquo;s Hospital relating to vaccine adjuvants and human in vitro platforms that model vaccine action. His laboratory has received research support from GlaxoSmithKline (GSK) and is a co-founder of and advisor to \\u003cem\\u003eARMR Sciences\\u003c/em\\u003e (formerly Ovax, Inc) that develops technologies to detect illicit substances and prevent overdose. He is also an advisor to GSK and Sanofi. Charles Cairns serves as a consultant to bioMerieux and is funded for a grant from Bill \\u0026amp; Melinda Gates Foundation. James A Overton is a consultant at Knocean Inc. Jessica Lasky-Su serves as a scientific advisor of Precion Inc. Scott R. Hutton, Greg Michelloti and Kari Wong are employees of Metabolon Inc. Vicki Seyfer- Margolis is a current employee of MyOwnMed. Nadine Rouphael reports grants or contracts with Merck, Sanofi, Pfizer, Vaccine Company, Quidel, Lilly and Immorna, and has participated on data safety monitoring boards for Moderna, Sanofi, Seqirus, Pfizer, EMMES, ICON, BARDA, Imunon, CyanVac and Micron. Nadine Rouphael has also received support for meetings/travel from Sanofi and Moderna and honoraria from Virology Education. Adeeb Rahman is a current employee of Immunai Inc. Steven Kleinstein is a consultant related to ImmPort data repository for Peraton. Nathan Grabaugh is a consultant for Tempus Labs and the National Basketball Association. Akiko Iwasaki is a consultant for 4BIO, Blue Willow Biologics, Revelar Biotherapeutics, RIGImmune, Xanadu Bio, Paratus Sciences. Monika Kraft receives research funds paid to her institution from NIH, ALA; Sanofi, Astra-Zeneca for work in asthma, serves as a consultant for Astra-Zeneca, Sanofi, Chiesi, GSK for severe asthma; is a co-founder and CMO for RaeSedo, Inc, a company created to develop peptidomimetics for treatment of inflammatory lung disease. Esther Melamed received research funding from Babson Diagnostics and honorarium from Multiple Sclerosis Association of America and has served on the advisory boards of Genentech, Horizon, Teva, and Viela Bio. Carolyn Calfee receives research funding from NIH, FDA, DOD, Roche-Genentech and Quantum Leap Healthcare Collaborative as well as consulting services for Janssen, Vasomune, Gen1e Life Sciences, NGMBio, and Cellenkos. Wade Schulz was an investigator for a research agreement, through Yale University, from the Shenzhen Center for Health Information for work to advance intelligent disease prevention and health promotion; collaborates with the National Center for Cardiovascular Diseases in Beijing; is a technical consultant to Hugo Health, a personal health information platform; cofounder of Refactor Health, an AI-augmented data management platform for health care; and has received grants from Merck and Regeneron Pharmaceutical for research related to COVID-19. Grace A McComsey received research grants from Redhill, Cognivue, Pfizer, and Genentech, and served as a research consultant for Gilead, Merck, Viiv/GSK, and Jenssen. Linda N. Geng received research funding paid to her institution from Pfizer, Inc.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eAcknowledgments\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe thank the participants of the study for their voluntary enrollment and contribution of samples for this work. See the supplement for details on the IMPACC Network. We acknowledge the assistance of the following individuals: Sanya Thomas, Mitchell Cooney, Shun Rao, Sofia Vignolo, and Elena Morrocchi (all from the CDCC); Arash Naeim, Marianne Bernardo, Sarahmay Sanchez, Shannon Intluxay, Clara Magyar, Jenny Brook, Estefania Ramires-Sanchez, Megan Llamas, Claudia Perdomo, Clara E. Magyar, and Jennifer A. Fulcher (all from the David Geffen School of Medicine at UCLA); members of the UCLA Center for Pathology Research Services and the Pathology Research Portal; M. Catherine Muenker, Dimitri Duvilaire, Maxine Kuang, William Ruff, Khadir Raddassi, Denise Shepherd, Haowei Wang, Omkar Chaudhary, Syim Salahuddin, John Fournier, Michael Rainone, and Maxine Kuang (all from the Yale School of Medicine). We thank the leadership of Boston Children\\u0026rsquo;s Hospital including Drs. Wendy Chung, Gary Fleisher, Nancy Andrews, and Kevin Churchwell for their support for the Precision Vaccines Program. Dr. Augustine\\u0026rsquo;s and Becker\\u0026rsquo;s co-authorship of this report does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases, the National Institutes of Health or any other agency of the United States Government.\\u003cstrong\\u003e\\u003cbr\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n \\u003cli\\u003eEClinicalMedicine. Antimicrobial resistance: a top ten global public health threat. \\u003cem\\u003eEClinicalMedicine\\u003c/em\\u003e \\u003cstrong\\u003e41\\u003c/strong\\u003e, 101221 (2021).\\u003c/li\\u003e\\n \\u003cli\\u003eMurray, C. J. \\u003cem\\u003eet al.\\u003c/em\\u003e Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. \\u003cem\\u003eThe Lancet\\u003c/em\\u003e \\u003cstrong\\u003e399\\u003c/strong\\u003e, 629\\u0026ndash;655 (2022).\\u003c/li\\u003e\\n \\u003cli\\u003eBaggs, J. \\u003cem\\u003eet al.\\u003c/em\\u003e Antibiotic Resistant Infections among COVID-19 Inpatients in U.S. Hospitals. \\u003cem\\u003eClin. Infect. Dis. Off. Publ. Infect. Dis. Soc. Am.\\u003c/em\\u003e ciac517 (2022) doi:10.1093/cid/ciac517.\\u003c/li\\u003e\\n \\u003cli\\u003eCenters for Disease Control and Prevention. U.S. Impact on Antimicrobial Resistance, Special Report 2022. 2022. https://www.cdc.gov/drugresistance/pdf/covid19-impact-report-508.pdf (2022).\\u003c/li\\u003e\\n \\u003cli\\u003eRose, A. N. \\u003cem\\u003eet al.\\u003c/em\\u003e Trends in Antibiotic Use in United States Hospitals During the Coronavirus Disease 2019 Pandemic. \\u003cem\\u003eOpen Forum Infect. Dis.\\u003c/em\\u003e \\u003cstrong\\u003e8\\u003c/strong\\u003e, ofab236 (2021).\\u003c/li\\u003e\\n \\u003cli\\u003eRomaszko-Wojtowicz, A., Tokarczyk-Malesa, K., Doboszyńska, A. \\u0026amp; Glińska-Lewczuk, K. Impact of COVID-19 on antibiotic usage in primary care: a retrospective analysis. \\u003cem\\u003eSci. Rep.\\u003c/em\\u003e \\u003cstrong\\u003e14\\u003c/strong\\u003e, 4798 (2024).\\u003c/li\\u003e\\n \\u003cli\\u003eNandi, A., Pecetta, S. \\u0026amp; Bloom, D. E. Global antibiotic use during the COVID-19 pandemic: analysis of pharmaceutical sales data from 71 countries, 2020\\u0026ndash;2022. \\u003cem\\u003eeClinicalMedicine\\u003c/em\\u003e \\u003cstrong\\u003e57\\u003c/strong\\u003e, (2023).\\u003c/li\\u003e\\n \\u003cli\\u003eSulis, G., Batomen, B., Kotwani, A., Pai, M. \\u0026amp; Gandra, S. Sales of antibiotics and hydroxychloroquine in India during the COVID-19 epidemic: An interrupted time series analysis. \\u003cem\\u003ePLoS Med.\\u003c/em\\u003e \\u003cstrong\\u003e18\\u003c/strong\\u003e, e1003682 (2021).\\u003c/li\\u003e\\n \\u003cli\\u003eO\\u0026rsquo;Leary, E. N. \\u003cem\\u003eet al.\\u003c/em\\u003e Impact of the COVID-19 Pandemic on Inpatient Antibiotic Use in the United States, January 2019 Through July 2022. \\u003cem\\u003eClin. Infect. Dis.\\u003c/em\\u003e \\u003cstrong\\u003e78\\u003c/strong\\u003e, 24\\u0026ndash;26 (2024).\\u003c/li\\u003e\\n \\u003cli\\u003eWorld Health Organization. WHO Model List of Essential Medicines. https://www.who.int/publications/i/item/WHO-MHP-HPS-EML-2023.02 (2023).\\u003c/li\\u003e\\n \\u003cli\\u003eU.S. Food and Drug Administration. FDA Drug Safety Communication: Azithromycin (Zithromax or Zmax) and the risk of potentially fatal heart rhythms. https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-azithromycin-zithromax-or-zmax-and-risk-potentially-fatal-heart (2018).\\u003c/li\\u003e\\n \\u003cli\\u003eHersh, A. L. \\u003cem\\u003eet al.\\u003c/em\\u003e Frequency of First-line Antibiotic Selection Among US Ambulatory Care Visits for Otitis Media, Sinusitis, and Pharyngitis. \\u003cem\\u003eJAMA Intern. Med.\\u003c/em\\u003e \\u003cstrong\\u003e176\\u003c/strong\\u003e, 1870 (2016).\\u003c/li\\u003e\\n \\u003cli\\u003eFleming-Dutra, K. E. \\u003cem\\u003eet al.\\u003c/em\\u003e Prevalence of Inappropriate Antibiotic Prescriptions Among US Ambulatory Care Visits, 2010-2011. \\u003cem\\u003eJAMA\\u003c/em\\u003e \\u003cstrong\\u003e315\\u003c/strong\\u003e, 1864\\u0026ndash;1873 (2016).\\u003c/li\\u003e\\n \\u003cli\\u003eBogdanić, N., Močibob, L., Vidović, T., Soldo, A. \\u0026amp; Begovać, J. Azithromycin consumption during the COVID-19 pandemic in Croatia, 2020. \\u003cem\\u003ePLOS ONE\\u003c/em\\u003e \\u003cstrong\\u003e17\\u003c/strong\\u003e, e0263437 (2022).\\u003c/li\\u003e\\n \\u003cli\\u003eTouret, F. \\u003cem\\u003eet al.\\u003c/em\\u003e In vitro screening of a FDA approved chemical library reveals potential inhibitors of SARS-CoV-2 replication. \\u003cem\\u003eSci. Rep.\\u003c/em\\u003e \\u003cstrong\\u003e10\\u003c/strong\\u003e, 13093 (2020).\\u003c/li\\u003e\\n \\u003cli\\u003eGautret, P. \\u003cem\\u003eet al.\\u003c/em\\u003e Hydroxychloroquine and azithromycin as a treatment of COVID-19: results of an open-label non-randomized clinical trial. \\u003cem\\u003eInt. J. Antimicrob. Agents\\u003c/em\\u003e \\u003cstrong\\u003e56\\u003c/strong\\u003e, 105949 (2020).\\u003c/li\\u003e\\n \\u003cli\\u003eKanoh, S. \\u0026amp; Rubin, B. K. Mechanisms of action and clinical application of macrolides as immunomodulatory medications. \\u003cem\\u003eClin. Microbiol. Rev.\\u003c/em\\u003e \\u003cstrong\\u003e23\\u003c/strong\\u003e, 590\\u0026ndash;615 (2010).\\u003c/li\\u003e\\n \\u003cli\\u003eSpeer, E. M. \\u003cem\\u003eet al.\\u003c/em\\u003e Pentoxifylline, dexamethasone and azithromycin demonstrate distinct age-dependent and synergistic inhibition of TLR- and inflammasome-mediated cytokine production in human newborn and adult blood in vitro. \\u003cem\\u003ePloS One\\u003c/em\\u003e \\u003cstrong\\u003e13\\u003c/strong\\u003e, e0196352 (2018).\\u003c/li\\u003e\\n \\u003cli\\u003eOldenburg, C. E. \\u003cem\\u003eet al.\\u003c/em\\u003e Effect of Oral Azithromycin vs Placebo on COVID-19 Symptoms in Outpatients With SARS-CoV-2 Infection: A Randomized Clinical Trial. \\u003cem\\u003eJAMA\\u003c/em\\u003e \\u003cstrong\\u003e326\\u003c/strong\\u003e, 490 (2021).\\u003c/li\\u003e\\n \\u003cli\\u003eHinks, T. S. C. \\u003cem\\u003eet al.\\u003c/em\\u003e Azithromycin versus standard care in patients with mild-to-moderate COVID-19 (ATOMIC2): an open-label, randomised trial. \\u003cem\\u003eLancet Respir. Med.\\u003c/em\\u003e \\u003cstrong\\u003e9\\u003c/strong\\u003e, 1130\\u0026ndash;1140 (2021).\\u003c/li\\u003e\\n \\u003cli\\u003eGagliotti, C. \\u003cem\\u003eet al.\\u003c/em\\u003e Use of Azithromycin Attributable to Acute SARS ‐ CoV ‐2 Infection. \\u003cem\\u003ePharmacoepidemiol. Drug Saf.\\u003c/em\\u003e \\u003cstrong\\u003e33\\u003c/strong\\u003e, e5857 (2024).\\u003c/li\\u003e\\n \\u003cli\\u003eDoan, T. \\u003cem\\u003eet al.\\u003c/em\\u003e Macrolide and Nonmacrolide Resistance with Mass Azithromycin Distribution. \\u003cem\\u003eN. Engl. J. Med.\\u003c/em\\u003e \\u003cstrong\\u003e383\\u003c/strong\\u003e, 1941\\u0026ndash;1950 (2020).\\u003c/li\\u003e\\n \\u003cli\\u003eArzika, A. M. \\u003cem\\u003eet al.\\u003c/em\\u003e Prolonged mass azithromycin distributions and macrolide resistance determinants among preschool children in Niger: A sub-study of a cluster-randomized trial (MORDOR). \\u003cem\\u003ePLoS Med.\\u003c/em\\u003e \\u003cstrong\\u003e21\\u003c/strong\\u003e, e1004386 (2024).\\u003c/li\\u003e\\n \\u003cli\\u003eDoan, T. \\u003cem\\u003eet al.\\u003c/em\\u003e Macrolide Resistance in MORDOR I \\u0026mdash; A Cluster-Randomized Trial in Niger. \\u003cem\\u003eN. Engl. J. Med.\\u003c/em\\u003e \\u003cstrong\\u003e380\\u003c/strong\\u003e, 2271\\u0026ndash;2273 (2019).\\u003c/li\\u003e\\n \\u003cli\\u003eTaylor, S. L. \\u003cem\\u003eet al.\\u003c/em\\u003e Long-Term Azithromycin Reduces Haemophilus influenzae and Increases Antibiotic Resistance in Severe Asthma. \\u003cem\\u003eAm. J. Respir. Crit. Care Med.\\u003c/em\\u003e \\u003cstrong\\u003e200\\u003c/strong\\u003e, 309\\u0026ndash;317 (2019).\\u003c/li\\u003e\\n \\u003cli\\u003eChu, V. T. \\u003cem\\u003eet al.\\u003c/em\\u003e The antibiotic resistance reservoir of the lung microbiome expands with age in a population of critically ill patients. \\u003cem\\u003eNat. Commun.\\u003c/em\\u003e \\u003cstrong\\u003e15\\u003c/strong\\u003e, 92 (2024).\\u003c/li\\u003e\\n \\u003cli\\u003eChu, V. T. \\u003cem\\u003eet al.\\u003c/em\\u003e Impact of doxycycline post-exposure prophylaxis for sexually transmitted infections on the gut microbiome and antimicrobial resistome. \\u003cem\\u003eNat. Med.\\u003c/em\\u003e (2024) doi:10.1038/s41591-024-03274-2.\\u003c/li\\u003e\\n \\u003cli\\u003eArce, Joann et al. Multi-omic longitudinal study reveals immune correlates of clinical course among hospitalized COVID-19 patients. \\u003cem\\u003eCell Rep. Med.\\u003c/em\\u003e (2023).\\u003c/li\\u003e\\n \\u003cli\\u003eOzonoff, A. \\u003cem\\u003eet al.\\u003c/em\\u003e Phenotypes of disease severity in a cohort of hospitalized COVID-19 patients: Results from the IMPACC study. \\u003cem\\u003eEBioMedicine\\u003c/em\\u003e \\u003cstrong\\u003e83\\u003c/strong\\u003e, 104208 (2022).\\u003c/li\\u003e\\n \\u003cli\\u003eIMPACC Manuscript Writing Team \\u0026amp; IMPACC Network Steering Committee. Immunophenotyping assessment in a COVID-19 cohort (IMPACC): A prospective longitudinal study. \\u003cem\\u003eSci. Immunol.\\u003c/em\\u003e \\u003cstrong\\u003e6\\u003c/strong\\u003e, eabf3733 (2021).\\u003c/li\\u003e\\n \\u003cli\\u003eHopkins, S. \\u0026amp; Williams, D. Five-day azithromycin in the treatment of patients with community-acquired pneumonia. \\u003cem\\u003eCurr. Ther. Res.\\u003c/em\\u003e \\u003cstrong\\u003e56\\u003c/strong\\u003e, 915\\u0026ndash;925 (1995).\\u003c/li\\u003e\\n \\u003cli\\u003eOrtiz-P\\u0026eacute;rez, A. \\u003cem\\u003eet al.\\u003c/em\\u003e High frequency of macrolide resistance mechanisms in clinical isolates of Corynebacterium species. \\u003cem\\u003eMicrob. Drug Resist. Larchmt. N\\u003c/em\\u003e \\u003cstrong\\u003e16\\u003c/strong\\u003e, 273\\u0026ndash;277 (2010).\\u003c/li\\u003e\\n \\u003cli\\u003eLaclaire, L. \\u0026amp; Facklam, R. Antimicrobial susceptibility and clinical sources of Dolosigranulum pigrum cultures. \\u003cem\\u003eAntimicrob. Agents Chemother.\\u003c/em\\u003e \\u003cstrong\\u003e44\\u003c/strong\\u003e, 2001\\u0026ndash;2003 (2000).\\u003c/li\\u003e\\n \\u003cli\\u003eCenters for Disease Control and Prevention (U.S.). \\u003cem\\u003eAntibiotic Resistance Threats in the United States, 2019\\u003c/em\\u003e. https://stacks.cdc.gov/view/cdc/82532 (2019) doi:10.15620/cdc:82532.\\u003c/li\\u003e\\n \\u003cli\\u003eVenkatesan, P. GOLD COPD report: 2024 update. \\u003cem\\u003eLancet Respir. Med.\\u003c/em\\u003e \\u003cstrong\\u003e12\\u003c/strong\\u003e, 15\\u0026ndash;16 (2024).\\u003c/li\\u003e\\n \\u003cli\\u003eMogayzel, P. J. \\u003cem\\u003eet al.\\u003c/em\\u003e Cystic fibrosis pulmonary guidelines. Chronic medications for maintenance of lung health. \\u003cem\\u003eAm. J. Respir. Crit. Care Med.\\u003c/em\\u003e \\u003cstrong\\u003e187\\u003c/strong\\u003e, 680\\u0026ndash;689 (2013).\\u003c/li\\u003e\\n \\u003cli\\u003eVos, R. \\u003cem\\u003eet al.\\u003c/em\\u003e A randomised controlled trial of azithromycin to prevent chronic rejection after lung transplantation. \\u003cem\\u003eEur. Respir. J.\\u003c/em\\u003e \\u003cstrong\\u003e37\\u003c/strong\\u003e, 164\\u0026ndash;172 (2011).\\u003c/li\\u003e\\n \\u003cli\\u003eNational Institutes of Health, Centers for Disease Control and Prevention, HIV Medicine Association, and Infectious Diseases Society of America. Panel on Guidelines for the Prevention and Treatment of Opportunistic Infections in Adults and Adolescents with HIV. Guidelines for the Prevention and Treatment of Opportunistic Infections in Adults and Adolescents with HIV. https://clinicalinfo.hiv.gov/en/guidelines/adult-and-adolescent-opportunistic-infection (2024).\\u003c/li\\u003e\\n \\u003cli\\u003eLu, N. \\u003cem\\u003eet al.\\u003c/em\\u003e DNA microarray analysis reveals that antibiotic resistance-gene diversity in human gut microbiota is age related. \\u003cem\\u003eSci. Rep.\\u003c/em\\u003e \\u003cstrong\\u003e4\\u003c/strong\\u003e, 4302 (2014).\\u003c/li\\u003e\\n \\u003cli\\u003eWu, L. \\u003cem\\u003eet al.\\u003c/em\\u003e Metagenomics-Based Analysis of the Age-Related Cumulative Effect of Antibiotic Resistance Genes in Gut Microbiota. \\u003cem\\u003eAntibiotics\\u003c/em\\u003e \\u003cstrong\\u003e10\\u003c/strong\\u003e, 1006 (2021).\\u003c/li\\u003e\\n \\u003cli\\u003eAlbert, R. K. \\u003cem\\u003eet al.\\u003c/em\\u003e Azithromycin for Prevention of Exacerbations of COPD. \\u003cem\\u003eN. Engl. J. Med.\\u003c/em\\u003e \\u003cstrong\\u003e365\\u003c/strong\\u003e, 689\\u0026ndash;698 (2011).\\u003c/li\\u003e\\n \\u003cli\\u003eRitchie, M. E. \\u003cem\\u003eet al.\\u003c/em\\u003e limma powers differential expression analyses for RNA-sequencing and microarray studies. \\u003cem\\u003eNucleic Acids Res.\\u003c/em\\u003e \\u003cstrong\\u003e43\\u003c/strong\\u003e, e47\\u0026ndash;e47 (2015).\\u003c/li\\u003e\\n \\u003cli\\u003eMaguire, C. \\u003cem\\u003eet al.\\u003c/em\\u003e Chronic Viral Reactivation and Associated Host Immune Response and Clinical Outcomes in Acute COVID-19 and Post-Acute Sequelae of COVID-19. \\u003cem\\u003eBioRxiv Prepr. Serv. Biol.\\u003c/em\\u003e 2024.11.14.622799 (2024) doi:10.1101/2024.11.14.622799.\\u003c/li\\u003e\\n \\u003cli\\u003ePhan, H. V. \\u003cem\\u003eet al.\\u003c/em\\u003e Host-microbe multiomic profiling reveals age-dependent immune dysregulation associated with COVID-19 immunopathology. \\u003cem\\u003eSci. Transl. Med.\\u003c/em\\u003e \\u003cstrong\\u003e16\\u003c/strong\\u003e, eadj5154 (2024).\\u003c/li\\u003e\\n \\u003cli\\u003eKalantar, K. L. \\u003cem\\u003eet al.\\u003c/em\\u003e IDseq-An open source cloud-based pipeline and analysis service for metagenomic pathogen detection and monitoring. \\u003cem\\u003eGigaScience\\u003c/em\\u003e \\u003cstrong\\u003e9\\u003c/strong\\u003e, giaa111 (2020).\\u003c/li\\u003e\\n \\u003cli\\u003eDobin, A. \\u003cem\\u003eet al.\\u003c/em\\u003e STAR: ultrafast universal RNA-seq aligner. \\u003cem\\u003eBioinforma. Oxf. Engl.\\u003c/em\\u003e \\u003cstrong\\u003e29\\u003c/strong\\u003e, 15\\u0026ndash;21 (2013).\\u003c/li\\u003e\\n \\u003cli\\u003eBolger, A. M., Lohse, M. \\u0026amp; Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. \\u003cem\\u003eBioinforma. Oxf. Engl.\\u003c/em\\u003e \\u003cstrong\\u003e30\\u003c/strong\\u003e, 2114\\u0026ndash;2120 (2014).\\u003c/li\\u003e\\n \\u003cli\\u003eRuby, J. G., Bellare, P. \\u0026amp; Derisi, J. L. PRICE: software for the targeted assembly of components of (Meta) genomic sequence data. \\u003cem\\u003eG3 Bethesda Md\\u003c/em\\u003e \\u003cstrong\\u003e3\\u003c/strong\\u003e, 865\\u0026ndash;880 (2013).\\u003c/li\\u003e\\n \\u003cli\\u003eLangmead, B. \\u0026amp; Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. \\u003cem\\u003eNat Methods\\u003c/em\\u003e \\u003cstrong\\u003e9\\u003c/strong\\u003e, (2012).\\u003c/li\\u003e\\n \\u003cli\\u003eBankevich, A. \\u003cem\\u003eet al.\\u003c/em\\u003e SPAdes: A New Genome Assembly Algorithm and Its Applications to Single-Cell Sequencing. \\u003cem\\u003eJ. Comput. Biol.\\u003c/em\\u003e \\u003cstrong\\u003e19\\u003c/strong\\u003e, 455\\u0026ndash;477 (2012).\\u003c/li\\u003e\\n \\u003cli\\u003eAlcock, B. P. \\u003cem\\u003eet al.\\u003c/em\\u003e CARD 2023: expanded curation, support for machine learning, and resistome prediction at the Comprehensive Antibiotic Resistance Database. \\u003cem\\u003eNucleic Acids Res.\\u003c/em\\u003e \\u003cstrong\\u003e51\\u003c/strong\\u003e, D690\\u0026ndash;D699 (2023).\\u003c/li\\u003e\\n \\u003cli\\u003eOzonoff, A. \\u003cem\\u003eet al.\\u003c/em\\u003e Phenotypes of disease severity in a cohort of hospitalized COVID-19 patients: Results from the IMPACC study. \\u003cem\\u003eEBioMedicine\\u003c/em\\u003e \\u003cstrong\\u003e83\\u003c/strong\\u003e, 104208 (2022).\\u003c/li\\u003e\\n \\u003cli\\u003eIMPACC Manuscript Writing Team \\u0026amp; IMPACC Network Steering Committee. Immunophenotyping assessment in a COVID-19 cohort (IMPACC): A prospective longitudinal study. \\u003cem\\u003eSci. Immunol.\\u003c/em\\u003e \\u003cstrong\\u003e6\\u003c/strong\\u003e, eabf3733 (2021).\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":true,\"hideJournal\":false,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":true,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"nature-portfolio\",\"isNatureJournal\":true,\"hasQc\":false,\"allowDirectSubmit\":false,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"\",\"title\":\"Nature Portfolio\",\"twitterHandle\":\"\",\"acdcEnabled\":false,\"dfaEnabled\":false,\"editorialSystem\":\"ejp\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false},\"keywords\":\"\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-6875205/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-6875205/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eAzithromycin is often prescribed unnecessarily for respiratory infections, many of which are viral. During the COVID-19 pandemic, its use was widespread, in part due to alleged therapeutic benefits, which have since been disproven. Here, we sought to understand the impact of azithromycin exposure on the respiratory microbiome, antimicrobial resistome, and host immune response in a prospective multicenter cohort of 1164 patients hospitalized for SARS-CoV-2 infection. Using longitudinal nasal metatranscriptomics, we compared patients treated with azithromycin (n=366, 31.4%) to those who received no antibiotics (n=474, 40.7%) or antibiotics other than azithromycin (n=324, 27.8%). We found that azithromycin treatment altered the community composition of the nasal microbiome, reducing bacterial relative abundance, increasing fungal relative abundance, and increasing potentially pathogenic taxa such as \\u003cem\\u003eKlebsiella\\u003c/em\\u003eand \\u003cem\\u003eStaphylococcus. \\u003c/em\\u003eAzithromycin treatment was most notably associated with increases in the number of detectably expressed macrolide/lincosamide/streptogramin (MLS) antimicrobial resistance genes, as well as their relative proportion in the resistome, with changes observable after one day of exposure. Of the MLS resistance genes, the expression of \\u003cem\\u003eermC\\u003c/em\\u003e, \\u003cem\\u003emsrA\\u003c/em\\u003e and \\u003cem\\u003eermX \\u003c/em\\u003eincreased the most in patients receiving azithromycin. Correlation analyses demonstrated that MLS resistance gene expression was significantly associated with the abundance of several taxa, including both commensal (e.g., \\u003cem\\u003eDolosigranulum, Corynebacterium\\u003c/em\\u003e) and potentially pathogenic genera (e.g., \\u003cem\\u003eStreptococcus, Staphylococcus).\\u003c/em\\u003eAssessment of the peripheral blood and upper airway host transcriptome demonstrated no differences in the expression of inflammatory genes. Taken together, our findings demonstrate that azithromycin treatment in COVID-19 leads to dysbiosis of the upper respiratory microbiome and changes in the expression of MLS resistance genes, without apparent anti-inflammatory benefit.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Empiric Azithromycin in COVID-19 Impacts the Respiratory Microbiome and Antimicrobial Resistome without Anti-inflammatory Benefit\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-06-20 15:32:31\",\"doi\":\"10.21203/rs.3.rs-6875205/v1\",\"editorialEvents\":[],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"nature-microbiology\",\"isNatureJournal\":true,\"hasQc\":false,\"allowDirectSubmit\":false,\"externalIdentity\":\"nmicrobiol\",\"sideBox\":\"Learn more about [Nature Microbiology](http://www.nature.com/nmicrobiol/)\",\"snPcode\":\"\",\"submissionUrl\":\"\",\"title\":\"Nature Microbiology\",\"twitterHandle\":\"\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"ejp\",\"reportingPortfolio\":\"Nature Research\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false}}],\"origin\":\"\",\"ownerIdentity\":\"9a384e93-a3aa-480f-8459-d225ebb01c3b\",\"owner\":[],\"postedDate\":\"June 20th, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"published-in-journal\",\"subjectAreas\":[{\"id\":50143722,\"name\":\"Biological sciences/Microbiology/Microbial communities/Metagenomics\"},{\"id\":50143723,\"name\":\"Health sciences/Diseases/Infectious diseases/Viral infection\"},{\"id\":50143724,\"name\":\"Biological sciences/Microbiology/Antimicrobials/Antimicrobial resistance\"}],\"tags\":[],\"updatedAt\":\"2026-03-17T07:06:54+00:00\",\"versionOfRecord\":{\"articleIdentity\":\"rs-6875205\",\"link\":\"https://doi.org/10.1038/s41564-026-02285-8\",\"journal\":{\"identity\":\"nature-microbiology\",\"isVorOnly\":false,\"title\":\"Nature Microbiology\"},\"publishedOn\":\"2026-03-16 04:00:00\",\"publishedOnDateReadable\":\"March 16th, 2026\"},\"versionCreatedAt\":\"2025-06-20 15:32:31\",\"video\":\"\",\"vorDoi\":\"10.1038/s41564-026-02285-8\",\"vorDoiUrl\":\"https://doi.org/10.1038/s41564-026-02285-8\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-6875205\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-6875205\",\"identity\":\"rs-6875205\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}