Microbial Dynamics and Pulmonary Immune Responses in COVID-19 Secondary Bacterial Pneumonia

Microbial Dynamics and Pulmonary Immune Responses in COVID-19 Secondary Bacterial Pneumonia | 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 Microbial Dynamics and Pulmonary Immune Responses in COVID-19 Secondary Bacterial Pneumonia Charles Langelier, Natasha Spottiswoode, Alexandra Tsitsiklis, and 21 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3877429/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 29 Oct, 2024 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Secondary bacterial pneumonia (2°BP) is associated with significant morbidity following respiratory viral infection, yet mechanistically remains incompletely understood. In a prospective cohort of 112 critically ill adults intubated for COVID-19, we comparatively assessed longitudinal airway microbiome dynamics and studied the pulmonary transcriptome of patients who developed 2°BP versus controls who did not. We found that 2°BP was significantly associated with both mortality and corticosteroid treatment. The pulmonary microbiome in 2°BP was characterized by increased bacterial RNA load, dominance of culture-confirmed pathogens, and lower alpha diversity. Bacterial pathogens were detectable days prior to 2°BP clinical diagnosis, and in most cases were also present in nasal swabs. Pathogen antimicrobial resistance genes were also detectable in both the lower airway and nasal samples, and in some cases were identified prior to 2°BP clinical diagnosis. Assessment of the pulmonary transcriptome revealed suppressed TNFa signaling via NF-kB in patients who developed 2°BP, and a sub-analysis suggested that this finding was mediated by corticosteroid treatment. Within the 2°BP group, we observed a striking inverse correlation between innate and adaptive immune gene expression and bacterial RNA load. Together, our findings provide fresh insights into the microbial dynamics and host immune features of COVID-19-associated 2°BP. Health sciences/Medical research/Translational research Biological sciences/Microbiology/Microbial communities/Metagenomics Biological sciences/Microbiology/Virology/SARS-CoV-2 Health sciences/Diseases/Infectious diseases/Bacterial infection Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Secondary bacterial pneumonia (2°BP) is a morbid and often fatal complication of severe respiratory viral infections 1 – 6 . Hospital-acquired 2°BP has been especially problematic during the COVID-19 pandemic, leading to longer hospitalizations 3 , 6 , increased mortality 7 , 8 , and often involving antimicrobial-resistant (AMR) pathogens 9 , 10 . A dynamic relationship between pathogens, the lung microbiome and the host immune response underpins the pathophysiology of pneumonia 11 , 12 , yet few studies have assessed the dynamics of these biological features in patients with 2°BP, leaving gaps in our understanding of this important sequela of viral illness. The co-pathogenesis of respiratory viruses with bacterial pathogens has been recognized for decades, and best studied in the context of influenza virus 13 , 14 . In the 1918 influenza pandemic, which led to over 50 million deaths, retrospective autopsy studies revealed evidence of 2°BP in the majority of cases 14 . Influenza, COVID-19, and other viral infections lead to alterations in the upper and lower respiratory tract microbiome, which may increase susceptibility to secondary infections by creating ecological niches for pathogenic bacteria 12 , 15 . Reduced microbiome alpha diversity, for instance, is a feature of both viral and bacterial lower respiratory tract infections in critically ill patients 12 , 16 – 18 . The common practice of early empiric antimicrobial administration in critically ill patients can further disrupt the airway microbiome, and additionally select for AMR bacterial pathogens 12 , 16 . Mechanically ventilated patients in particular endure prolonged exposure to the hospital environment, which increases the risk of colonization by opportunistic pathogens in the upper airway, oropharynx, and lungs 19 , 20 . It remains unclear, however, how changes in the airway microbiome of critically ill patients with COVID-19 might precipitate 2°BP, highlighting an important knowledge gap. Severe COVID-19 is characterized by a dysregulated inflammatory response in both the airways and systemic circulation 21 , 22 , yet whether 2°BP is associated with further alterations in this pathologic immune state remains unclear. For instance, 2°BP might lead to activation of innate immune signaling pathways important for bacterial defense, which has been observed in patients with ventilator-associated pneumonia prior to the COVID-19 pandemic 23 – 25 . Alternatively, patients with 2°BP may have suppressed innate immunity, which is well described in mouse models of post-influenza 2°BP 13 , 26 , and in patients with sepsis who acquire nosocomial infections 20 , 27 . It is also possible that the host response to 2°BP may simply be overshadowed by the inflammatory state of severe SARS-CoV-2 infection. Despite their interconnected roles, few studies have assessed both lower respiratory tract microbiome dynamics and host immune responses in critically ill patients, and none for the explicit purpose of studying post-viral 2°BP. A recent elegant study showcased how lower respiratory metatranscriptomics can effectively identify connections between host and microbial factors with clinical outcomes in COVID-19 28 , however it did not focus on clinically-confirmed 2°BP. Two recent diagnostic test studies demonstrated the potential of respiratory metatranscriptomics to improve the detection of pathogens in COVID-19 patients with ventilator associated pneumonia 29 , 30 , but did not evaluate biological features of 2°BP. The burden of secondary infections in patients with COVID-19 and other viral pneumonias, as well as gaps in our mechanistic understanding of 2°BP, motivated us to carry out this study. We assessed lung microbiome dynamics and host immune responses using metatranscriptomics in a large cohort of hospitalized COVID-19 patients with rigorous 2°BP adjudication by three physicians. We observed disruption of the lung microbiome in patients with 2°BP, characterized by increased bacterial RNA load and dominance of culture-identified pathogens, as well as changes in host immune signaling involving genes important for bacterial defense. Together, our findings provide fresh insights into the biology of 2°BP, suggesting potential new therapeutic targets and approaches to 2°BP diagnosis. Results Patient Cohort and Pneumonia Adjudication We studied critically ill patients requiring mechanical ventilation for COVID-19 enrolled in the prospective observational COVID-19 Multi-Phenotyping for Effective Therapies (COMET) study between 04/2020 and 12/2021. Patients were enrolled at one tertiary care hospital and one safety net hospital in San Francisco, California under a research protocol approved by the University of California San Francisco Institutional Review Board (Methods). We collected tracheal aspirate (TA) and nasal swabs (NS) periodically following intubation, and performed metatranscriptomic sequencing (Fig. 1A). Of the 397 patients with COVID-19 enrolled in COMET, 112 had critical illness requiring invasive mechanical ventilation. Culture-confirmed secondary bacterial pneumonia (2°BP) was identified in N = 44 (39.3%) of these patients based on 3-physician adjudication using the US Centers for Disease Control and Prevention PNEU1 surveillance definition of pneumonia 31 and all available clinical data in the electronic medical record, blinded to metatranscriptomic results. Patients with no clinical evidence of bacterial pneumonia at any point during their hospitalization (N = 41, No-BP group) were also identified. 27 patients were excluded from further analysis; 22 of whom could not be confidently adjudicated into 2°BP or No-BP groups, and five with other reasons for exclusion including hospital transfer or intubation for reasons other than COVID-19. Analysis of clinical and demographic data (Table S1 A) demonstrated that hospital mortality was significantly greater in patients with 2°BP than in those without (47.7% vs 7.3%, P = 3.03e-5). Patients with 2°BP were also more likely to have received corticosteroids during their hospitalizations (97.7% vs 82.9%, P = 0.026). All patients received antibiotics during their hospitalizations, and total days of antibiotic therapy did not differ between groups. A minority of patients had received one or more SARS-CoV-2 vaccines prior to admission (9.1% vs 12.2%, P = 0.61); most patients were recruited prior to vaccine availability in early 2021. For metatranscriptomic analyses, we evaluated patients with TA samples that met baseline quality control metrics, and for 2°BP patients, those with samples available within a 5-day window of 2°BP clinical diagnosis (Fig. 1 , Methods). This left 178 TA samples from 27 2°BP and 29 No-BP patients available for metatranscriptomic analysis (Table S1 B). Within this analysis subgroup, Staphylococcus aureus was the most prevalent 2°BP pathogen identified by clinical bacterial cultures (N = 10, 37.0%) followed by Pseudomonas aeruginosa (N = 6, 22.2%) (Fig. 1 B, Table S1 B). 15 of 27 patients with 2°BP also had NS samples collected and suitable for analysis. COVID-19-associated secondary bacterial pneumonia is characterized by higher lower airway bacterial RNA load, pathogen dominance, and changes in the lung microbiome We began metatranscriptomic analyses by comparatively assessing bacterial RNA load in the lower respiratory tract microbiome of 2°BP patients versus No-BP controls. TA samples from 2°BP patients collected closest to date of clinical diagnosis had higher bacterial RNA load compared to samples from No-BP controls obtained at comparable timepoints post-intubation (P = 0.0016, Fig. 2A). We next assessed lung microbiome alpha diversity and found that while median Shannon Diversity Index (SDI) was lower in 2°BP patients compared to No-BP controls, it did not significantly differ (P = 0.10, Fig. 2B). While in most patients the 2°BP pathogen was most abundant in the TA sample collected closest to the time of clinical 2°BP diagnosis, in 10/27 (37.0%) of patients, it was top ranked by abundance in samples collected earlier or later (e.g., Patient 1427, Fig. 3 ). We thus repeated the alpha diversity analysis using the sample in which the 2°BP pathogen was highest ranked (as opposed to the sample closest to the date of 2°BP diagnosis), and found that SDI was significantly lower compared to No-BP controls (P = 0.014, Fig. 2C). Longitudinal assessment of airway microbiome dynamics demonstrated that alpha diversity decreased over time in all groups following intubation, and was consistently lower in the 2°BP group (P = 0.019, Fig. 2D). We more closely examined the relationship between the culture-confirmed 2°BP pathogen and the airway microbiome, and found that the 2°BP pathogen was detected by metatranscriptomics in all 27 2°BP cases. The 2°BP pathogen was dominant in the lower airway microbiome of at least one sample from most patients with 2°BP (Fig. 2E), with the culture-confirmed pathogen ranking in the top 3 most abundant taxa in at least one sample for 25/27 (92.6%) of cases within 7 days of clinical diagnosis (Fig. 2F). Further assessment of the most abundant taxa in lung microbiome demonstrated that Staphylococcus aureus and Pseudomonas aeruginosa were most commonly identified in 2°BP patients, while Prevotella, Mycoplasma , and Alloprevotella species were most commonly found in No-BP cases (Fig. 2G). Compositional differences in the lung microbiomes of 2°BP versus No-BP patients based on Bray Curtis dissimilarity index approached statistical significance (Fig. 2H, P = 0.07 by PERMANOVA). Lastly, we asked whether SARS-CoV-2 viral load in the lung, as measured by metatranscriptomics in reads per million (rpM), differed between patients with or without 2°BP, but found no difference (P = 0.22, Fig. S1 ). Dynamics of secondary bacterial pneumonia pathogens in the lung We next evaluated the longitudinal dynamics of the cultured-confirmed 2°BP pathogen in the lower airway by overlaying abundance rank trajectories of the cultured pathogen(s) in the airway microbiome with phenotypic susceptibility to administered antibiotics (Fig. 3 ). The culture-confirmed 2°BP pathogen could be detected in the airway prior to 2°BP diagnosis in 14/14 (100%) of the patients who had samples obtained prior to clinical diagnosis (Fig. 3 ). Assessment of longitudinal trajectories revealed cases where pathogen expansion in the lower airway prior to 2°BP diagnosis occurred in the absence of any antibiotic treatment (e.g., Patient 1254, S. aureus , Fig. 3 ), as well as following treatment with antibiotics to which the pathogen was resistant (e.g., Patient 1114, P. aeruginosa , Fig. 3 ). In patients with longitudinal sampling after clinical 2°BP diagnosis, we asked how pathogen clearance related to its susceptibility to administered antibiotics. We observed cases where initiation of antibiotics with known activity against the 2°BP pathogen resulted in a decrease in pathogen abundance, as expected (Patient 1231, E. coli , Fig. 3 ). Many pathogens, however, remained the most abundant microbe in the lower airway for days following initiation of antimicrobial therapy (e.g., Patient 1051, P. aeruginosa , Fig. 3 ). We noted that patients with P. aeruginosa infections were more likely to exhibit this impaired clearance phenotype compared to patients with other types of 2°BP (P = 0.0.017, Fig. S2). Detection of secondary bacterial pneumonia pathogens in the upper respiratory tract Among the 15 patients who also had nasal swab (NS) samples available, we did not observe differences in nasal microbiome bacterial RNA load (Fig. 4A) or alpha diversity (Fig. 4B) based on 2°BP status. We noted that in 14/15 (93.3%) of cases, the 2°BP pathogen was detected in at least one NS sample within 7 days of clinical diagnosis, and in 7/15 (46.7%) cases it ranked in the top 3 most abundant (Fig. 4C). Among the 12 patients who had NS samples collected prior to the date of clinical 2°BP diagnosis, 8/9 (88.9%) had at least one etiologic pathogen detected (Fig. 4D). The respiratory antimicrobial resistome Analysis of the lower respiratory tract resistome of 2°BP patients revealed a diversity of AMR genes representing multiple different classes, including plasmid-transmissible extended spectrum beta lactamase (ESBL) (e.g., CTX-M ) and colistin-resistance genes (e.g., MCR1 ) (Fig. 5 ). In some cases, AMR genes associated with culture-confirmed resistant pathogens were detectable before clinical diagnosis of 2°BP (e.g., CTX-M, SED-1 , patient 1196) (Fig. 5 , Fig. S3). Identification of the ampC inducible beta lactamase was noted in patient 1154, from whom K. aerogenes was identified by culture (Fig. 5 ). This pathogen continued to dominate the airway despite treatment with a beta-lactam antibiotic (piperacillin-tazobactam) to which the organism was phenotypically susceptible (Fig. 3 ). AmpC can be induced following beta lactam exposure in certain Enterobacteriaceae, resulting in reversal of phenotypic susceptibility, and in some cases clinical treatment failure 32 . Comparative assessment demonstrated that clinically relevant AMR genes detected in lower respiratory tract pathogens could also be identified in the upper airway in a subset of cases. For instance, in patient 1145 (Fig. 5 ), who had multi-drug-resistant E. coli, K. pneumoniae , and methicillin-resistant S. aureus grow from TA culture, MCR-1, CTX-M and mecA were detected in both NS and TA samples. As in the lower respiratory tract, we found that clinically relevant AMR genes related to the 2°BP pathogen could in some cases be detected in the nares prior to clinical recognition of bacterial pneumonia (e.g., patient 1145, Fig. S2). Host transcriptional responses in COVID-19 secondary bacterial pneumonia We next asked whether a lower respiratory transcriptional signature of 2°BP could be identified amidst the intense inflammatory state of severe COVID-19. A comparison of host gene expression between 2°BP and No-BP patients identified 226 differentially expressed genes (FDR < 0.1) (Fig. 6 A, Table S3). Gene set enrichment analysis (GSEA) revealed that 2°BP was characterized by downregulated TNFa signaling via NF-kB in (Fig. 6 B, Table S4), suggesting that a state of suppressed antibacterial defense might characterize 2°BP in COVID-19 patients. We hypothesized that corticosteroid treatment might be contributing to this, and thus performed a secondary differential gene expression analysis of only patients treated with dexamethasone and other corticosteroids. This analysis yielded 4XX differentially expressed genes (FDR < 0.1) (Table S5), with no evidence of suppressed TNFa signaling on GSEA, supporting our hypothesis. GSEA instead demonstrated non-significant upregulation of pro-inflammatory signaling pathways such as IL-6/JAK/STAT-3 in 2°BP patients (Fig. 6 C, Table S6). We further investigated connections between the airway transcriptome and microbiome in 2°BP patients by testing whether the bacterial RNA load correlated with host gene expression. Differential expression analysis identified 4784 significant genes (FDR < 0.1) (Fig. 6 D, Table S7). GSEA (Fig. 6 F, Table S8) revealed a striking inverse relationship between bacterial RNA load and immune signaling pathways important for antibacterial defense (e.g., TNFa, IL-6, IL-2), which was also evident at the individual gene level (e.g., HLA-DRB1 , C1QC ) (Fig. 6 E). Together, these results suggested that a relative state of impaired immune defense may exist in COVID-19 patients who develop 2°BP, and that this may be influenced by treatment with corticosteroids. Discussion In this prospective observational study, we assessed respiratory tract microbial dynamics and host transcriptional responses associated with 2°BP in COVID-19 patients requiring invasive mechanical ventilation. Using comparative metatranscriptomics, we found that 2°BP is characterized by disruption of the lung microbiome, dominance of pathogenic bacteria that can be co-detected in the nares, and a lower airway transcriptome globally characterized by suppressed TNFa signaling. Bacterial superinfection is a well-established contributor to influenza mortality 14 , yet in COVID-19 its role in mortality has been less clear 28 , 33 , 34 . We found that 2°BP affected 39% of mechanically ventilated patients in our multicenter cohort, and was strongly associated with mortality. Our results notably differed from two important prior studies 28 , 34 which did not find clear links between secondary bacterial infection and mortality in COVID-19 patients. This discrepancy may be explained by our use of both an established case definition 31 and microbiological criteria to define 2°BP, as opposed to simply requiring a positive bacterial respiratory culture 28 , 34 . We also found that patients who developed 2°BP were more likely to have received corticosteroids during their hospitalizations, suggesting that the therapeutic benefit of corticosteroids may come at the expense of increased 2°BP risk. Early identification and treatment of hospital-onset bacterial pneumonia can prevent adverse consequences including prolonged mechanical ventilation, inappropriate antibiotic exposure, and mortality 35 – 37 . We found that 2°BP pathogens could be detected in the lower airway up to a week before clinical recognition of infection, and were frequently amongst the most abundant taxa in the lung microbiome in the days preceding culture-based detection. In some cases, we also detected pathogen-associated AMR genes before clinical diagnosis of 2°BP. These findings suggest the potential of metatranscriptomics to enable early detection of secondary bacterial infections, adding to a growing body of literature suggesting the benefits of this technology in the intensive care unit 12 , 29 , 38 . Further work in a larger, appropriately designed cohort is needed to assess the diagnostic performance of metatranscriptomics for 2°BP diagnosis. Lower respiratory tract infections are characterized by a loss of airway microbiome alpha diversity 12 , 16 , 38 , 39 , which is also observed over time in mechanically ventilated patients 40 , 41 . We found that 2°BP further disrupts microbial community structure in patients with existing SARS-CoV-2 infection, leading to additional loss of diversity in the setting of bacterial pathogen dominance in the airway. In over a third of cases, the peak of bacterial pathogen dominance did not overlap with the date of 2°BP clinical diagnosis. This could represent a decoupling of physiologic responses and pathogen dynamics, heterogeneity in TA sampling, or reflect the challenge of identifying a new pneumonia amidst an existing severe viral lower respiratory tract infection. Our beta diversity analysis also suggested that COVID-19 patients who develop 2°BP may have differences in the composition of their airway microbiome. By characterizing relationships between 2°BP pathogens and the lung microbiome over time, we observed that in some patients, the pathogen remained dominant in the airway even after 2°BP clinical diagnosis and the initiation of appropriate antibiotics. In particular, all P. aeruginosa cases exhibited this persistence trend, which may reflect the known tendency of this pathogen to form biofilms. We found that 2°BP pathogens were not only detectable in the lower airway, but also in the upper airway in more than half of the cases, including prior to clinical pneumonia diagnosis. Most patients hospitalized for bacterial pneumonia do not require invasive mechanical ventilation, thus it is possible that minimally invasive nasal swab specimens could have diagnostic value in such cases. Because this observational study included only intubated patients, future studies with paired bronchoscopy and nasal sampling in less critically ill patients will be needed to address this question. AMR infections, in particular respiratory infections, are a major public health issue and have increased in prevalence during the COVID-19 pandemic 9 , 10 . We observed several AMR genes and classes considered potential threats to the management of bacterial infections. These include the plasmid transmissible genes MCR-1 , Qnr-S and CTX-M , which have a high potential for horizontal transfer both within patient and within hospital. We also identified several AMR genes not typically detectable by existing clinical assays. For instance, three patients from one study site were found to harbor the SED-1 class A beta lactamase (patients 1145, 1196, 1231), which was originally identified in Citrobacter sedikiae 42 , reported once in E. coli contaminating produce in China 43 , but not otherwise reported in the context of human infection. In several cases, we found that pathogen-associated AMR genes were detectable before clinical diagnosis of 2°BP, and were also found in the upper airway. While screening for mecA in the nares is routinely performed to identify patients at risk for invasive MRSA infections 44 , it has remained unclear whether this approach has value for other types of AMR infections. Our data suggest that screening for other clinically important AMR genes may prove useful for identifying patients with drug-resistant pneumonia, with implications for both antimicrobial stewardship and infection control. Severe COVID-19 is characterized by a profoundly dysregulated host response in the lower respiratory tract 21 , 45 , 46 , which complicates the detection of a transcriptional immune response to 2°BP. Nonetheless, our GSEA results suggest that impaired TNFa signaling may play a role in the pathophysiology of 2°BP. Indeed, increased rates of serious bacterial infections are well described in patients receiving anti-TNFa therapies 47 , and our findings suggest that a state of relative immunosuppression may augment susceptibility to opportunistic bacterial pathogens in COVID-19 patients. Consistent with this idea is our finding that bacterial RNA load in the lungs strongly associated with downregulation of innate and adaptative immune pathways essential for antibacterial defense. This model is also in line with murine models of post-influenza 2°BP, which demonstrate virus-induced impairment of innate immunity 13 , 26 , 48 , 49 characterized by reduced expression of TNFa, IL-6, and other cytokines. Our secondary analysis of steroid recipients did not yield this signal of downregulated TNFa signaling, suggesting that corticosteroid treatment, which is now standard of care for severe COVID-19 50 , may contribute to this immunosuppressed state. Strengths of our study include the novel use of host/microbe metatranscriptomics to study secondary bacterial infections, rigorous clinical adjudication of 2°BP, longitudinal sampling, and measurement of bacterial RNA load, a biomarker not previously evaluated in studies of pneumonia. Limitations include a relatively small sample size and incomplete longitudinal sampling for all patients. Our host transcriptomic findings require further exploration and validation in an independent cohort. While several public COVID-19 respiratory transcriptomic datasets exist, none include adjudication of 2°BP status using a rigorous and standardized definition. Taken together, our findings shed light on the microbial dynamics and host immune responses of 2°BP, a clinically important and serious complication of COVID-19 and other viral respiratory illnesses. Future studies are needed to validate these findings, clarify mechanisms in cohorts with other viral infections, and evaluate the diagnostic potential of metatranscriptomics for early detection of 2°BP. Methods Study design, enrollment, and institutional review board approval We conducted a prospective case-control study of hospitalized adults requiring mechanical ventilation for COVID-19 with or without secondary bacterial pneumonia. (Fig. 1 ). All 397 patients with clinical polymerase chain reaction (PCR)-confirmed SARS-CoV-2 infection enrolled in the UCSF COVID-19 Multiphenotyping for Effective Therapy (COMET) observational study of patients with acute respiratory illnesses were initially considered for inclusion. Seventeen patients were co-enrolled in the National Institute of Allergy and Infectious Diseases-funded Immunophenotyping Assessment in a COVID-19 Cohort (IMPACC) Network study. This study was approved by the UCSF Institutional Review Board (IRB) under protocol #20-30497. Detailed enrollment and consent protocols for both studies have been previously described 12 , 22 . Secondary bacterial pneumonia adjudication and patient inclusion Amongst the 112 critically ill COVID-19 patients requiring mechanical ventilation enrolled in COMET, N = 44 cases of culture-confirmed secondary bacterial pneumonia (2°BP) were adjudicated by 3 study team infectious disease physicians (NS, CD, CRL) based on the United States Centers for Disease Control and Prevention PNEU1 surveillance definition of pneumonia 31 and all available clinical data in the electronic medical record, blinded to metatranscriptomic results. Study team physicians also assessed the antimicrobial susceptibility patterns of cultured microbes, and identified days that patients with 2°BP had received antibiotics to which their cultured microbes were susceptible and/or resistant (Supp. Table 5). The date of 2°BP clinical diagnosis was set as the date on which the positive bacterial culture was ordered by the treating medical team. Patients with no clinical evidence of bacterial pneumonia at any point during their hospitalization (No-BP group, N = 41) were also identified. TA samples with metatranscriptomic data passing minimum quality control parameters (see below) were available for 29 2°BP patients and 29 No-BP patients. Within the 2°BP group, two additional patients were excluded from analyses as they did not have any TA samples collected within a 5-day window of 2°BP clinical diagnosis (-3 days to + 2 days). The remaining final cohort of 2°BP (N = 27) and No-BP (N = 29) patients was leveraged for all metatranscriptomic analyses, which included all samples from available within − 7 or + 7 days of clinical diagnosis. Tracheal aspirate and nasal swab sampling Following enrollment, tracheal aspirate (TA) was collected periodically following intubation without addition of saline wash, and mixed 1:1 with DNA/RNA shield in tubes containing bashing beads (Zymo Research). Nasal swabs were collected into tubes prefilled with DNA/RNA shield (Zymo Research). Samples were frozen within one hour and stored at -80C until nucleic acid extraction. RNA sequencing To evaluate host and microbial gene expression, metatranscriptomic RNA sequencing (RNA-seq) was performed on TA specimens. Following RNA extraction from 320µL of input sample (Zymo Pathogen Magbead Kit) and DNase treatment, human cytosolic and mitochondrial ribosomal RNA was depleted using FastSelect (Qiagen). To control for background contamination, we included negative controls (water and HeLa cell RNA) as well as positive controls (spike-in RNA standards from the External RNA Controls Consortium (ERCC)) 51 . RNA was then fragmented and underwent library preparation using the NEBNext Ultra II RNA-seq Kit (New England Biolabs). Libraries underwent 146 nucleotide paired-end Illumina sequencing on an Illumina Novaseq 6000. Quality control and mitigation of environmental contaminants To minimize inaccurate taxonomic assignments due to environmental and reagent derived contaminants, non-templated “water only” and HeLa cell RNA controls were processed with each group of samples that underwent nucleic acid extraction. These were included, as well as positive control clinical samples, with each sequencing run. Negative control samples enabled estimation of the number of background reads expected for each taxon. Microbes established as metagenomic contaminants were filtered out (Sphingomonas, Bradyrhizobium, Ralstonia, Delftia, Cutibacterium, Methylobacterium, Acidovorax, Chryseobacterium, Burkholderia). Samples were excluded from analysis if they had fewer than 1,000 total reads mapping to bacterial taxa or a bacterial mass of < 1 pg. A single sample per patient was utilized for analyses that involved direct comparison at a single timepoint, including bacterial RNA load, alpha diversity, beta diversity, top pathogen identification, and TA only, host gene expression analyses. The TA or NS samples collected closest to the date of clinical 2°BP diagnosis, defined as the date the positive culture was ordered by the clinical treatment team, were used in the primary timepoint analyses. To identify No-BP control samples appropriately matched by time from intubation, we selected samples collected closest to the median days post-intubation of the comparable 2°BP TA (N = 6 days) or NS (N = 5 days) samples. Because both host and microbial analyses were carried out for TA single timepoint comparison analyses, those samples were required to meet both host and microbe quality control standards. Statistics Statistical significance was defined as P < 0.05 using two-tailed tests, unless stated otherwise. Categorical data were analyzed by Fisher’s exact test and nonparametric continuous variables were analyzed by Mann-Whitney. Statistical approaches used for gene expression and microbiome analyses are detailed in each respective Methods section. Lung microbiome analysis Taxonomic alignments were obtained from raw sequencing reads using the CZID pipeline 52 , 53 , which performs quality filtration and removal of human reads followed by reference-based taxonomic alignment against sequences in the National Center for Biotechnology Information (NCBI) nucleotide (NT) database, followed by assembly of reads matching each taxon detected. Taxonomic alignments underwent background correction for environmental contaminants (see below), viruses were excluded, and data was then aggregated to the genus level before calculating diversity metrics. Alpha diversity (Shannon’s Diversity Index) and beta diversity (Bray-Curtis dissimilarity) were calculated and the latter plotted using non-metric multidimensional scaling (NDMS). Bacterial RNA load/mass calculations Bacterial RNA load was calculated based on the ratio of bacterial reads in each sample to total reads aligning to the External RNA Controls Consortium (ERCC) RNA mass standards spiked into each sample 54 . The following equation was utilized for this calculation: [ERCC input mass]/[bacterial mass] = [ERCC reads]/[bacterial reads]. Host differential expression Following demultiplexing, sequencing reads were pseudo-aligned with kallisto 55 to an index consisting of all transcripts associated with human protein coding genes (ENSEMBL release 99), long non-coding RNA, cytosolic and mitochondrial ribosomal RNA sequences and the sequences of ERCC RNA standards. Gene-level counts were generated from the transcript-level abundance estimates using the R package tximport 56 , with the scaledTPM method. For quality control, we only retained samples with a total of at least 1,000,000 estimated protein-coding gene counts, and a proportion of ribosomal RNA to total RNA ≤ 50%, according to a previously described approach 38 . In addition, we only analyzed host genes with at least 10 counts in at least 20% of samples. Differential expression analysis was performed in R (v4.3.2) using the package limma-voom 57 (v3.58.1). For the comparison between 2°BP and No-BP patients, we adjusted for SARS-CoV-2 viral load rpM by adding the coefficient log10(rpM + 1) to the linear model (without adjusting for any other covariates). For the analysis of corticosteroid recipients, we restricted to patients who had received treatment with steroids at any time prior to sample collection. For the analysis of host gene expression and bacterial mass, we modelled gene expression of 2°BP patients on log10-transformed bacterial mass (without adjusting for any other covariates). Significant genes were identified using a Benjamini-Hochberg false discovery rate (FDR) < 0.1 (for 2°BP vs No-BP analysis) or < 0.05 (for the host gene vs bacterial mass analysis). Differential expression analysis results are provided in (Supp. Tables 1 and 3). Gene set enrichment analysis (GSEA) was performed using the package fgsea (v1.28.0), and the Hallmark pathways were obtained from the package msigdbr (v7.5.1). The t-statistics (obtained from limma’s differential expression analysis) were used to rank all genes, and used as input for the fgseaMultilevel function (minSize = 15, maxSize = 500). Pathways with adjusted P-value below 0.05 were considered statistically significant. Declarations Data and code availability The raw fastq files with microbial sequencing reads are available under NCBI BioProject ID: PRJNA1033689. The host gene counts are available under NCBI Gene Expression Omnibus (GEO) accession number: GSE246795. The human raw sequencing data are protected due to data privacy restrictions from the IRB protocol governing patient enrollment, which protects the release of raw genetic sequencing data from those patients enrolled under a waiver of consent. To honor this, researchers who wish to obtain raw fastq files for the purposes of independently generating gene counts can contact the corresponding author ( [email protected] ) and request to be added to the IRB protocol. All code and source data used for analyses can be found at: (https://github.com/chazlangelier/2BP). Funding NHLBI, NIAID, Chan Zuckerberg Biohub References Luyt, C.-E. et al. Ventilator-associated pneumonia in patients with SARS-CoV-2-associated acute respiratory distress syndrome requiring ECMO: a retrospective cohort study. Ann Intensive Care 10 , 158 (2020). Bardi, T. et al. Nosocomial infections associated to COVID-19 in the intensive care unit: clinical characteristics and outcome. Eur J Clin Microbiol Infect Dis (2021) doi:10.1007/s10096-020-04142-w. Maes, M. et al. Ventilator-associated pneumonia in critically ill patients with COVID-19. Crit Care 25 , 25 (2021). Søgaard, K. K. et al. Community-acquired and hospital-acquired respiratory tract infection and bloodstream infection in patients hospitalized with COVID-19 pneumonia. J Intensive Care 9 , (2021). Rouzé, A. et al. Relationship between SARS-CoV-2 infection and the incidence of ventilator-associated lower respiratory tract infections: a European multicenter cohort study. Intensive Care Med 1–11 (2021) doi:10.1007/s00134-020-06323-9. DeVoe, C. et al. Increased Rates of Secondary Bacterial Infections, Including Enterococcus Bacteremia, in Patients Hospitalized with COVID-19. Infect Control Hosp Epidemiol 1–33 (2021) doi:10.1017/ice.2021.391. Falsey, A. R. et al. Bacterial Complications of Respiratory Tract Viral Illness: A Comprehensive Evaluation. The Journal of Infectious Diseases 208 , 432–441 (2013). Reyes, L. F. et al. Risk factors for developing ventilator-associated lower respiratory tract infection in patients with severe COVID-19: a multinational, multicentre study, prospective, observational study. Sci Rep 13 , 6553 (2023). Baggs, J. et al. Antibiotic Resistant Infections among COVID-19 Inpatients in U.S. Hospitals. Clin Infect Dis 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). Kalantar, K. L. & Langelier, C. R. Host-Microbe Metagenomics: a Lens To Refocus Our Perspective on Infectious and Inflammatory Diseases. mSystems 6 , e00404-21 (2021). Langelier, C. et al. Integrating host response and unbiased microbe detection for lower respiratory tract infection diagnosis in critically ill adults. Proc Natl Acad Sci U S A 115 , E12353–E12362 (2018). McCullers, J. A. The co-pathogenesis of influenza viruses with bacteria in the lung. Nat. Rev. Microbiol. 12 , 252–262 (2014). Morens, D. M., Taubenberger, J. K. & Fauci, A. S. Predominant role of bacterial pneumonia as a cause of death in pandemic influenza: implications for pandemic influenza preparedness. J Infect Dis 198 , 962–970 (2008). Kaul, D. et al. Microbiome disturbance and resilience dynamics of the upper respiratory tract during influenza A virus infection. Nat Commun 11 , 2537 (2020). Flanagan, J. L. et al. Loss of bacterial diversity during antibiotic treatment of intubated patients colonized with Pseudomonas aeruginosa. J Clin Microbiol 45 , 1954–1962 (2007). Kalantar, K. L. et al. Metagenomic comparison of tracheal aspirate and mini-bronchial alveolar lavage for assessment of respiratory microbiota. Am. J. Physiol. Lung Cell Mol. Physiol. 316 , L578–L584 (2019). Durán-Manuel, E. M. et al. Massive sequencing of the V3-V4 hypervariable region of bronchoalveolar lavage from patients with COVID-19 and VAP reveals the collapse of the pulmonary microbiota. J Med Microbiol 71 , (2022). Ballinger, M. N. & Standiford, T. J. Innate Immune Responses in Ventilator-Associated Pneumonia. in Mucosal Immunology of Acute Bacterial Pneumonia (ed. Prince, A.) 185–212 (Springer New York, 2013). doi:10.1007/978-1-4614-5326-0_8. Wunderink, R. G. Nosocomial Pneumonia, Including Ventilator-associatedPneumonia. Proc Am Thorac Soc 2 , 440–444 (2005). Blanco-Melo, Daniel. Imbalanced host response to SARS-CoV-2 drives development of COVID-19. Cell. doi:10.1016/j.cell.2020.04.026. Sarma, A. et al. Tracheal aspirate RNA sequencing identifies distinct immunological features of COVID-19 ARDS. Nat Commun 12 , 5152 (2021). Nguyen, E. V. et al. Proteomic Profiling of Bronchoalveolar Lavage Fluid in Critically Ill Patients with Ventilator-Associated Pneumonia. PLoS ONE 8 , e58782 (2013). Conway Morris, A. et al. Diagnostic importance of pulmonary interleukin-1 and interleukin-8 in ventilator-associated pneumonia. Thorax 65 , 201–207 (2010). Pathak, K. V. et al. Molecular Profiling of Innate Immune Response Mechanisms in Ventilator-associated Pneumonia. Molecular & Cellular Proteomics 19 , 1688–1705 (2020). Robinson, K. M., Kolls, J. K. & Alcorn, J. F. The immunology of influenza virus-associated bacterial pneumonia. Current Opinion in Immunology 34 , 59–67 (2015). Conway Morris, A. et al. Cell-surface signatures of immune dysfunction risk-stratify critically ill patients: INFECT study. Intensive Care Med 44 , 627–635 (2018). Sulaiman, I. et al. Microbial signatures in the lower airways of mechanically ventilated COVID-19 patients associated with poor clinical outcome. Nat Microbiol 6 , 1245–1258 (2021). Charalampous, T. et al. Evaluating the potential for respiratory metagenomics to improve treatment of secondary infection and detection of nosocomial transmission on expanded COVID-19 intensive care units. Genome Med 13 , 182 (2021). Wu, N. et al. Rapid identification of pathogens associated with ventilator-associated pneumonia by Nanopore sequencing. Respir Res 22 , 310 (2021). United States Centers for Disease Control and Prevention. CDC/NHSN Surveillance Definitions for Specific Types of Infections. (2021). Tamma, P. D. et al. Infectious Diseases Society of America Guidance on the Treatment of AmpC β-Lactamase–Producing Enterobacterales, Carbapenem-Resistant Acinetobacter baumannii , and Stenotrophomonas maltophilia Infections. Clinical Infectious Diseases 74 , 2089–2114 (2022). Zhou, F. et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. The Lancet 395 , 1054–1062 (2020). Pickens, C. O. et al. Bacterial Superinfection Pneumonia in Patients Mechanically Ventilated for COVID-19 Pneumonia. Am J Respir Crit Care Med 204 , 921–932 (2021). Kuti, E. L., Patel, A. A. & Coleman, C. I. Impact of inappropriate antibiotic therapy on mortality in patients with ventilator-associated pneumonia and blood stream infection: A meta-analysis. Journal of Critical Care 23 , 91–100 (2008). ICU-Acquired Pneumonia Study Group & Alvarez-Lerma, F. Modification of empiric antibiotic treatment in patients with pneumonia acquired in the intensive care unit. Intensive Care Med 22 , 387–394 (1996). Celis, R. et al. Nosocomial Pneumonia. Chest 93 , 318–324 (1988). Mick, E. et al. Integrated host/microbe metagenomics enables accurate lower respiratory tract infection diagnosis in critically ill children. J Clin Invest 133 , e165904 (2023). Dickson, R. P., Erb-Downward, J. R. & Huffnagle, G. B. Towards an ecology of the lung: new conceptual models of pulmonary microbiology and pneumonia pathogenesis. Lancet Respir Med 2 , 238–246 (2014). Zakharkina, T. et al. The dynamics of the pulmonary microbiome during mechanical ventilation in the intensive care unit and the association with occurrence of pneumonia. Thorax 72 , 803–810 (2017). Pettigrew, M. M., Tanner, W. & Harris, A. D. The Lung Microbiome and Pneumonia. The Journal of Infectious Diseases 223 , S241–S245 (2021). Petrella, S., Clermont, D., Casin, I., Jarlier, V. & Sougakoff, W. Novel Class A β-Lactamase Sed-1 from Citrobacter sedlakii : Genetic Diversity of β-Lactamases within the Citrobacter Genus. Antimicrob Agents Chemother 45 , 2287–2298 (2001). Liao, N. et al. Prevalence of β-Lactam Drug-Resistance Genes in Escherichia coli Contaminating Ready-to-Eat Lettuce. Foodborne Pathogens and Disease 17 , 739–742 (2020). Parente, D. M., Cunha, C. B., Mylonakis, E. & Timbrook, T. T. The Clinical Utility of Methicillin-Resistant Staphylococcus aureus (MRSA) Nasal Screening to Rule Out MRSA Pneumonia: A Diagnostic Meta-analysis With Antimicrobial Stewardship Implications. Clinical Infectious Diseases 67 , 1–7 (2018). Sarma, A. et al. COVID-19 ARDS is characterized by a dysregulated host response that differs from cytokine storm and is modified by dexamethasone. Res Sq (2021) doi:10.21203/rs.3.rs-141578/v1. Mick, E. et al. Upper airway gene expression reveals suppressed immune responses to SARS-CoV-2 compared with other respiratory viruses. Nature Communications 11 , 5854 (2020). Hochman, D. & Wolff, B. Risk of Serious Infections and Malignancies With Anti-TNF Antibody Therapy in Rheumatoid Arthritis. JAMA 296 , 2201 (2006). Sun, K. & Metzger, D. W. Inhibition of pulmonary antibacterial defense by interferon-gamma during recovery from influenza infection. Nat Med 14 , 558–564 (2008). Goulding, J. et al. Lowering the threshold of lung innate immune cell activation alters susceptibility to secondary bacterial superinfection. J Infect Dis 204 , 1086–1094 (2011). Bhimraj, A. et al. Infectious Diseases Society of America Guidelines on the Treatment and Management of Patients With Coronavirus Disease 2019 (COVID-19). Clinical Infectious Diseases ciac724 (2022) doi:10.1093/cid/ciac724. Pine, P. S. et al. Evaluation of the External RNA Controls Consortium (ERCC) reference material using a modified Latin square design. BMC Biotechnol 16 , 54 (2016). Ramesh, A. et al. Metagenomic next-generation sequencing of samples from pediatric febrile illness in Tororo, Uganda. PLoS One 14 , e0218318 (2019). Kalantar, K. L. et al. IDseq—An open source cloud-based pipeline and analysis service for metagenomic pathogen detection and monitoring. Gigascience 9 , (2020). Pine, P. S. et al. Evaluation of the External RNA Controls Consortium (ERCC) reference material using a modified Latin square design. BMC Biotechnol 16 , 54 (2016). Bray, N. L., Pimentel, H., Melsted, P. & Pachter, L. Near-optimal probabilistic RNA-seq quantification. Nature Biotechnology 34 , 525–527 (2016). Soneson, C., Love, M. I. & Robinson, M. D. Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences. F1000Res 4 , 1521 (2015). Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Research 43 , e47–e47 (2015). Additional Declarations There is NO Competing Interest. Supplementary Files TableS2.xlsx Table S2 TableS3.csv Table S3 TableS4.csv Table S4 TableS5.csv Table S5 TableS6.csv Table S6 TableS7.csv Table S7 TableS8.csv Table S8 COVID2BPsupp011724.docx Supplementary Materials Cite Share Download PDF Status: Published Journal Publication published 29 Oct, 2024 Read the published version in Nature Communications → 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-3877429","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":269075095,"identity":"8fc42bf4-0c36-433b-8d31-2b9ebbcfbbf2","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":269075096,"identity":"92f0ee4c-0ea1-4907-9f17-be480fd80d84","order_by":1,"name":"Natasha Spottiswoode","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Natasha","middleName":"","lastName":"Spottiswoode","suffix":""},{"id":269075097,"identity":"5c4ce873-46b8-46da-9f21-917824f10f25","order_by":2,"name":"Alexandra Tsitsiklis","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Alexandra","middleName":"","lastName":"Tsitsiklis","suffix":""},{"id":269075098,"identity":"02b7ad32-c0d5-4a6d-9386-a8c6ecaae273","order_by":3,"name":"Hoang Phan","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Hoang","middleName":"","lastName":"Phan","suffix":""},{"id":269075099,"identity":"9546dd6a-e4ef-4521-b439-578d4f1fabc7","order_by":4,"name":"Victoria Chu","email":"","orcid":"https://orcid.org/0000-0002-7480-965X","institution":"UCSF","correspondingAuthor":false,"prefix":"","firstName":"Victoria","middleName":"","lastName":"Chu","suffix":""},{"id":269075100,"identity":"7a546184-9eef-48f8-a2a2-457568b1ff7c","order_by":5,"name":"Catherine Devoe","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Catherine","middleName":"","lastName":"Devoe","suffix":""},{"id":269075101,"identity":"9c7419f0-a93c-49fa-9132-868154600b6d","order_by":6,"name":"Christina Love","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Christina","middleName":"","lastName":"Love","suffix":""},{"id":269075102,"identity":"f55a4e31-1947-44cd-8812-7c64d2eb9ffd","order_by":7,"name":"Rajani Ghale","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Rajani","middleName":"","lastName":"Ghale","suffix":""},{"id":269075103,"identity":"cf570224-2b0c-43e5-bbc6-0802c2c656d8","order_by":8,"name":"Joshua Bloomstein","email":"","orcid":"","institution":"University of Washington","correspondingAuthor":false,"prefix":"","firstName":"Joshua","middleName":"","lastName":"Bloomstein","suffix":""},{"id":269075104,"identity":"8115a4c4-3267-4e6e-834b-43b540a4b1f0","order_by":9,"name":"Beth Shoshana Zha","email":"","orcid":"https://orcid.org/0000-0002-4190-2823","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Beth","middleName":"Shoshana","lastName":"Zha","suffix":""},{"id":269075105,"identity":"aa5a8348-92db-49b8-bbf6-b4b0bf08dfdd","order_by":10,"name":"Peter Mourani","email":"","orcid":"https://orcid.org/0000-0002-1829-3775","institution":"Arkansas Children's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Peter","middleName":"","lastName":"Mourani","suffix":""},{"id":269075106,"identity":"3fc0ff5c-3785-4fe4-a6c9-eb0ae414ee60","order_by":11,"name":"Katrina Kalantar","email":"","orcid":"","institution":"Chan Zuckerberg Initiative","correspondingAuthor":false,"prefix":"","firstName":"Katrina","middleName":"","lastName":"Kalantar","suffix":""},{"id":269075107,"identity":"55f17d53-92ef-472d-8606-4571db730406","order_by":12,"name":"Angela Detweiler","email":"","orcid":"https://orcid.org/0000-0001-7026-5467","institution":"Chan Zuckerberg Biohub","correspondingAuthor":false,"prefix":"","firstName":"Angela","middleName":"","lastName":"Detweiler","suffix":""},{"id":269075108,"identity":"d9e5f36e-5b0c-4b33-b0e8-248ac47921b0","order_by":13,"name":"Norma Neff","email":"","orcid":"https://orcid.org/0000-0001-7141-5420","institution":"Chan Zuckerberg Biohub","correspondingAuthor":false,"prefix":"","firstName":"Norma","middleName":"","lastName":"Neff","suffix":""},{"id":269075109,"identity":"460747d5-6c5a-41e9-9879-60e90220397a","order_by":14,"name":"Sidney Carrillo","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Sidney","middleName":"","lastName":"Carrillo","suffix":""},{"id":269075110,"identity":"f776d338-8404-4318-a507-da0c5bdc7338","order_by":15,"name":"COMET Consortium","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"COMET","middleName":"","lastName":"Consortium","suffix":""},{"id":269075111,"identity":"27eb0a1d-a360-48f2-b96a-3073ad294792","order_by":16,"name":"Joseph DeRisi","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Joseph","middleName":"","lastName":"DeRisi","suffix":""},{"id":269075112,"identity":"ebb6ef7a-4a61-46c1-9244-1c313a0d472a","order_by":17,"name":"David Erle","email":"","orcid":"https://orcid.org/0000-0002-2171-0648","institution":"UCSF","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Erle","suffix":""},{"id":269075113,"identity":"1d59ee93-fe33-4fd8-839f-df7d63058180","order_by":18,"name":"Carolyn Hendrickson","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Carolyn","middleName":"","lastName":"Hendrickson","suffix":""},{"id":269075114,"identity":"64c669b2-b7a5-4a51-ac3d-d46731b1f057","order_by":19,"name":"Kristen Kangelaris","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Kristen","middleName":"","lastName":"Kangelaris","suffix":""},{"id":269075115,"identity":"891c5887-1bc9-4b26-9299-6fc558633a93","order_by":20,"name":"Matthew Krummel","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"","lastName":"Krummel","suffix":""},{"id":269075116,"identity":"12197e80-c9cc-4665-9c2b-bc5a0d04e6b5","order_by":21,"name":"Michael Matthay","email":"","orcid":"https://orcid.org/0000-0003-3039-8155","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"","lastName":"Matthay","suffix":""},{"id":269075117,"identity":"d96209f5-12be-4b12-9213-6ba1a7bd34be","order_by":22,"name":"Prescott Woodruff","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Prescott","middleName":"","lastName":"Woodruff","suffix":""},{"id":269075118,"identity":"ef005de6-6335-453d-8b11-28913e9179d7","order_by":23,"name":"Carolyn Calfee","email":"","orcid":"","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Carolyn","middleName":"","lastName":"Calfee","suffix":""}],"badges":[],"createdAt":"2024-01-19 03:05:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3877429/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3877429/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-024-53566-x","type":"published","date":"2024-10-29T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":50751746,"identity":"8ff434e9-2853-4103-8374-447511750319","added_by":"auto","created_at":"2024-02-06 17:40:13","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":243789,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStudy overview. (A) \u003c/strong\u003ePatient flow after recruitment within 48 hours of admission.\u003cstrong\u003e \u003c/strong\u003e*Intubated for reasons other than COVID-19 or transferred to an outside hospital within 48 hours of recruitment.\u003cstrong\u003e (B)\u003c/strong\u003e Pathogens detected by bacterial culture at time of secondary bacterial pneumonia (2°BP) diagnosis. Three patients had \u0026gt;1 pathogen detected.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/d85d0ec76db5006f51adbdd5.jpg"},{"id":50752356,"identity":"731761ed-680e-4da5-a40e-100351e8be3e","added_by":"auto","created_at":"2024-02-06 17:48:13","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":507694,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLung microbiome differences in patients with or without 2°BP\u003c/strong\u003e. \u003cstrong\u003e(A) \u003c/strong\u003eBacterial RNA load and \u003cstrong\u003e(B)\u003c/strong\u003e alpha diversity measured by Shannon Diversity Index (SDI) in COVID-19 patients with 2°BP (red, N = 27) or No-BP (blue, N = 29), measured in the 2°BP TA sample closest to the date of clinical diagnosis, or No-BP patient samples obtained after a similar number of days post-intubation. \u003cstrong\u003e(C)\u003c/strong\u003e SDI measured in the 2°BP sample in which the culture-confirmed pathogen was top ranked by abundance in the lung microbiome. \u003cstrong\u003e(D) \u003c/strong\u003eLongitudinal dynamics of SDI following intubation in 2°BP and No-BP patients. \u003cstrong\u003e(E)\u003c/strong\u003e Exemplary abundance plots from two 2°BP patients demonstrating the abundance, measured in reads per million (rpM, y axis) for each of the top 15 ranked bacterial taxa in the lung microbiome (x axis). Culture-confirmed 2°BP pathogen highlighted in red. \u003cstrong\u003e(F)\u003c/strong\u003e Bar plot showing the number of 2°BP patients (y axis) in which a specific rank by rpM of the culture-confirmed pathogen (x axis) was observed within 7 days of clinical diagnosis. \u003cstrong\u003e(G) \u003c/strong\u003eHeatmap showing the number of cases in which specific bacterial genera were detected as the most abundant in the bacterial lung microbiome. Top 15 genera plotted. Staphylococcus included at species level given the intrinsic differences in potential pathogenicity between \u003cem\u003eS. aureus \u003c/em\u003eand \u003cem\u003eS. epidermidis.\u003c/em\u003e \u003cstrong\u003e(H)\u003c/strong\u003e Non-metric multidimensional scaling (NMDS) plot demonstrating compositional differences in the lung microbiome of 2°BP (red dots) versus No-BP (blue dots) patients. P values for (A-C) calculated by Wilcoxon test. P value for (D) calculated by linear mixed effects modeling treating patients as random effects. P value in (H) calculated by PERMANOVA analysis.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/d03975d1c726c382864b069f.jpg"},{"id":50751749,"identity":"ef8edc66-f952-48b5-b78a-73f0f28749e3","added_by":"auto","created_at":"2024-02-06 17:40:13","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":818836,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDynamics of 2°BP pathogen over time relative to the date of clinical diagnosis\u003c/strong\u003e. Genus level 2°BP pathogen rank based on bacterial reads per million (rpM) in the lung microbiome. Days relative to 2°BP clinical diagnosis on the X axis. Days during which patient received antibiotics to which 2°BP pathogen was phenotypically susceptible (light grey bar) or resistant (dark grey bar) shown below. Patient death during this period is shown using a vertically oriented black bar. Patient ID and pathogen genus are listed. For patients with multiple cultured pathogens (1145, 1250, and 1474), each plot represents a cultured pathogen. Open circles indicate that the pathogen was not detected in a sample.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/744ff9a5b8ddf77dce9b9162.jpg"},{"id":50752364,"identity":"096a30d0-8c49-403a-ad29-b80c83198f05","added_by":"auto","created_at":"2024-02-06 17:48:13","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":554972,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNasal microbiome differences in patients with or without 2°BP as compared to the lung microbiome. (A) \u003c/strong\u003eBacterial RNA mass or (\u003cstrong\u003eB) \u003c/strong\u003eAlpha diversity as measured by Shannon Diversity Index (SDI) in COVID-19 patients with 2°BP (purple, N = 15) or No-BP (pink, N = 20) in the 2°BP nasal swab sample closest to the date of clinical diagnosis or control samples at comparable durations on ventilator. P values are Wilcoxon tests.\u003cstrong\u003e (C) \u003c/strong\u003eRate of identification of the cultured pathogen in nasal swabs by metatranscriptomics in samples -7 to +7 days pre and post 2°BP diagnosis \u003cstrong\u003e(D)\u003c/strong\u003e Longitudinal dynamics of culture-confirmed bacterial pathogen rank in the nasal (purple) or lower respiratory (red) microbiome based on abundance (measured in rpM). Patients with multiple cultured microbes are shown as one plot per microbe. Open circles in indicate that the pathogen was not detected in the sample.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/8bd09186d4e600e683b9f739.jpg"},{"id":50751753,"identity":"90155080-c261-49a7-b054-0de9843bc84d","added_by":"auto","created_at":"2024-02-06 17:40:13","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":734481,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAntimicrobial resistance (AMR) genes in the lung and nasal microbiome of 2°BP\u003c/strong\u003e \u003cstrong\u003epatients. \u003c/strong\u003eA) AMR genes grouped and colored by class detected in tracheal aspirate (TA) +/- matched nasal swab (NS) samples. Shading corresponds to the fraction of samples collected within 7 days in which the AMR gene was detected.\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/84d7036de578fe18f2b81e8a.jpeg"},{"id":50751754,"identity":"405316c4-bb4e-490b-bcb7-7e230805b5aa","added_by":"auto","created_at":"2024-02-06 17:40:13","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":331790,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLower respiratory tract gene expression differs based on 2°BP status and correlates with bacterial mass. \u003c/strong\u003e(A) Volcano plot of differentially expressed genes between 2°BP and No-BP. (B) Bar plot of GSEA analysis showing Hallmark pathways that are downregulated in 2°BP patients. (C) Bar plot of GSEA analysis in steroid recipients showing the same Hallmark pathways as in (B). In the analyses in (A-C), we controlled for SARS-CoV-2 viral load in our differential expression analysis. (D) Volcano plot of differentially expressed genes based on bacterial mass among 2°BP patients. (E) Scatter plots showing the relationship between \u003cem\u003eHLA-DRA\u003c/em\u003e and \u003cem\u003eC1QC \u003c/em\u003egene expression and bacterial RNA mass in 2°BP patients. The black lines indicate the linear regression fit, and the ribbons indicate the 95% confidence interval of the fits. (F) Bar plot showing Hallmark pathways that are downregulated with higher bacterial mass in 2°BP patients. The P-values in (A, D, E) were calculated using linear modeling (limma package) and Benjamini-Hochberg correction. The P-values in (B, C, F) were calculated using the fgsea package and Benjamini-Hochberg correction.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/0144f948ec83a4f4607415da.png"},{"id":67829794,"identity":"c05445ff-835b-4c38-b1bb-a57db9a01697","added_by":"auto","created_at":"2024-10-30 07:06:19","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4119866,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/86e9be78-2a50-4d02-ae02-3add7d2ab724.pdf"},{"id":50751744,"identity":"0b76c604-c8e4-48ee-965e-44deb87fbdc3","added_by":"auto","created_at":"2024-02-06 17:40:12","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":28723,"visible":true,"origin":"","legend":"\u003cp\u003eTable S2\u003c/p\u003e","description":"","filename":"TableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/e950a7565afdf42456cc4d29.xlsx"},{"id":50752363,"identity":"ade3c80f-8bbc-46ab-b1c6-94455b12e035","added_by":"auto","created_at":"2024-02-06 17:48:13","extension":"csv","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":764668,"visible":true,"origin":"","legend":"\u003cp\u003eTable S3\u003c/p\u003e","description":"","filename":"TableS3.csv","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/49afc50458a2a0b7217fe41a.csv"},{"id":50752362,"identity":"8c73445f-d8ed-4476-812b-c591f9c94e01","added_by":"auto","created_at":"2024-02-06 17:48:13","extension":"csv","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":33862,"visible":true,"origin":"","legend":"\u003cp\u003eTable S4\u003c/p\u003e","description":"","filename":"TableS4.csv","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/b440f84864a9b0b54cdbdb2b.csv"},{"id":50751756,"identity":"bfc4e47b-3c42-4f9b-9f48-930d114a04a9","added_by":"auto","created_at":"2024-02-06 17:40:13","extension":"csv","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":764540,"visible":true,"origin":"","legend":"\u003cp\u003eTable S5\u003c/p\u003e","description":"","filename":"TableS5.csv","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/9c438c8b89c9ea7d1fe5cbd5.csv"},{"id":50751747,"identity":"3ffa9219-2b97-4110-9709-71c834c47823","added_by":"auto","created_at":"2024-02-06 17:40:13","extension":"csv","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":36654,"visible":true,"origin":"","legend":"\u003cp\u003eTable S6\u003c/p\u003e","description":"","filename":"TableS6.csv","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/01baa3f32327bfd7c04a7e51.csv"},{"id":50752797,"identity":"a77aee83-5693-4dc4-8e3a-97f52bc9a53b","added_by":"auto","created_at":"2024-02-06 17:56:14","extension":"csv","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":810573,"visible":true,"origin":"","legend":"\u003cp\u003eTable S7\u003c/p\u003e","description":"","filename":"TableS7.csv","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/b2f94639e6c6ccced138ca25.csv"},{"id":50751751,"identity":"b9dbb146-3054-4a0b-a8ea-957661017a2a","added_by":"auto","created_at":"2024-02-06 17:40:13","extension":"csv","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":47053,"visible":true,"origin":"","legend":"\u003cp\u003eTable S8\u003c/p\u003e","description":"","filename":"TableS8.csv","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/becedb26d7ce9e714c212de7.csv"},{"id":50751758,"identity":"097108af-93b0-485a-bcf8-4ee3002f61ff","added_by":"auto","created_at":"2024-02-06 17:40:13","extension":"docx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":5032471,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Materials\u003c/p\u003e","description":"","filename":"COVID2BPsupp011724.docx","url":"https://assets-eu.researchsquare.com/files/rs-3877429/v1/f51b1daa6fb099e1e146246a.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Microbial Dynamics and Pulmonary Immune Responses in COVID-19 Secondary Bacterial Pneumonia","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSecondary bacterial pneumonia (2\u0026deg;BP) is a morbid and often fatal complication of severe respiratory viral infections\u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3 CR4 CR5\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Hospital-acquired 2\u0026deg;BP has been especially problematic during the COVID-19 pandemic, leading to longer hospitalizations\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, increased mortality\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, and often involving antimicrobial-resistant (AMR) pathogens\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. A dynamic relationship between pathogens, the lung microbiome and the host immune response underpins the pathophysiology of pneumonia\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, yet few studies have assessed the dynamics of these biological features in patients with 2\u0026deg;BP, leaving gaps in our understanding of this important sequela of viral illness.\u003c/p\u003e \u003cp\u003eThe co-pathogenesis of respiratory viruses with bacterial pathogens has been recognized for decades, and best studied in the context of influenza virus\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. In the 1918 influenza pandemic, which led to over 50\u0026nbsp;million deaths, retrospective autopsy studies revealed evidence of 2\u0026deg;BP in the majority of cases\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Influenza, COVID-19, and other viral infections lead to alterations in the upper and lower respiratory tract microbiome, which may increase susceptibility to secondary infections by creating ecological niches for pathogenic bacteria\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Reduced microbiome alpha diversity, for instance, is a feature of both viral and bacterial lower respiratory tract infections in critically ill patients\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe common practice of early empiric antimicrobial administration in critically ill patients can further disrupt the airway microbiome, and additionally select for AMR bacterial pathogens\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Mechanically ventilated patients in particular endure prolonged exposure to the hospital environment, which increases the risk of colonization by opportunistic pathogens in the upper airway, oropharynx, and lungs\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. It remains unclear, however, how changes in the airway microbiome of critically ill patients with COVID-19 might precipitate 2\u0026deg;BP, highlighting an important knowledge gap.\u003c/p\u003e \u003cp\u003eSevere COVID-19 is characterized by a dysregulated inflammatory response in both the airways and systemic circulation\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, yet whether 2\u0026deg;BP is associated with further alterations in this pathologic immune state remains unclear. For instance, 2\u0026deg;BP might lead to activation of innate immune signaling pathways important for bacterial defense, which has been observed in patients with ventilator-associated pneumonia prior to the COVID-19 pandemic\u003csup\u003e\u003cspan additionalcitationids=\"CR24\" citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Alternatively, patients with 2\u0026deg;BP may have suppressed innate immunity, which is well described in mouse models of post-influenza 2\u0026deg;BP\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, and in patients with sepsis who acquire nosocomial infections\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. It is also possible that the host response to 2\u0026deg;BP may simply be overshadowed by the inflammatory state of severe SARS-CoV-2 infection.\u003c/p\u003e \u003cp\u003eDespite their interconnected roles, few studies have assessed both lower respiratory tract microbiome dynamics and host immune responses in critically ill patients, and none for the explicit purpose of studying post-viral 2\u0026deg;BP. A recent elegant study showcased how lower respiratory metatranscriptomics can effectively identify connections between host and microbial factors with clinical outcomes in COVID-19\u003csup\u003e28\u003c/sup\u003e, however it did not focus on clinically-confirmed 2\u0026deg;BP. Two recent diagnostic test studies demonstrated the potential of respiratory metatranscriptomics to improve the detection of pathogens in COVID-19 patients with ventilator associated pneumonia\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, but did not evaluate biological features of 2\u0026deg;BP.\u003c/p\u003e \u003cp\u003eThe burden of secondary infections in patients with COVID-19 and other viral pneumonias, as well as gaps in our mechanistic understanding of 2\u0026deg;BP, motivated us to carry out this study. We assessed lung microbiome dynamics and host immune responses using metatranscriptomics in a large cohort of hospitalized COVID-19 patients with rigorous 2\u0026deg;BP adjudication by three physicians. We observed disruption of the lung microbiome in patients with 2\u0026deg;BP, characterized by increased bacterial RNA load and dominance of culture-identified pathogens, as well as changes in host immune signaling involving genes important for bacterial defense. Together, our findings provide fresh insights into the biology of 2\u0026deg;BP, suggesting potential new therapeutic targets and approaches to 2\u0026deg;BP diagnosis.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n\u003ch2\u003ePatient Cohort and Pneumonia Adjudication\u003c/h2\u003e\n\u003cp\u003eWe studied critically ill patients requiring mechanical ventilation for COVID-19 enrolled in the prospective observational COVID-19 Multi-Phenotyping for Effective Therapies (COMET) study between 04/2020 and 12/2021. Patients were enrolled at one tertiary care hospital and one safety net hospital in San Francisco, California under a research protocol approved by the University of California San Francisco Institutional Review Board (Methods). We collected tracheal aspirate (TA) and nasal swabs (NS) periodically following intubation, and performed metatranscriptomic sequencing (Fig. 1A).\u003c/p\u003e\n\u003cp\u003eOf the 397 patients with COVID-19 enrolled in COMET, 112 had critical illness requiring invasive mechanical ventilation. Culture-confirmed secondary bacterial pneumonia (2\u0026deg;BP) was identified in N\u0026thinsp;=\u0026thinsp;44 (39.3%) of these patients based on 3-physician adjudication using the US Centers for Disease Control and Prevention PNEU1 surveillance definition of pneumonia\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e and all available clinical data in the electronic medical record, blinded to metatranscriptomic results. Patients with no clinical evidence of bacterial pneumonia at any point during their hospitalization (N\u0026thinsp;=\u0026thinsp;41, No-BP group) were also identified. 27 patients were excluded from further analysis; 22 of whom could not be confidently adjudicated into 2\u0026deg;BP or No-BP groups, and five with other reasons for exclusion including hospital transfer or intubation for reasons other than COVID-19.\u003c/p\u003e\n\u003cp\u003eAnalysis of clinical and demographic data (Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003eA) demonstrated that hospital mortality was significantly greater in patients with 2\u0026deg;BP than in those without (47.7% vs 7.3%, P\u0026thinsp;=\u0026thinsp;3.03e-5). Patients with 2\u0026deg;BP were also more likely to have received corticosteroids during their hospitalizations (97.7% vs 82.9%, P\u0026thinsp;=\u0026thinsp;0.026). All patients received antibiotics during their hospitalizations, and total days of antibiotic therapy did not differ between groups. A minority of patients had received one or more SARS-CoV-2 vaccines prior to admission (9.1% vs 12.2%, P\u0026thinsp;=\u0026thinsp;0.61); most patients were recruited prior to vaccine availability in early 2021.\u003c/p\u003e\n\u003cp\u003eFor metatranscriptomic analyses, we evaluated patients with TA samples that met baseline quality control metrics, and for 2\u0026deg;BP patients, those with samples available within a 5-day window of 2\u0026deg;BP clinical diagnosis (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e, Methods). This left 178 TA samples from 27 2\u0026deg;BP and 29 No-BP patients available for metatranscriptomic analysis (Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003eB). Within this analysis subgroup, \u003cem\u003eStaphylococcus aureus\u003c/em\u003e was the most prevalent 2\u0026deg;BP pathogen identified by clinical bacterial cultures (N\u0026thinsp;=\u0026thinsp;10, 37.0%) followed by \u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e (N\u0026thinsp;=\u0026thinsp;6, 22.2%) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eB, Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003eB). 15 of 27 patients with 2\u0026deg;BP also had NS samples collected and suitable for analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCOVID-19-associated secondary bacterial pneumonia is characterized by higher lower airway bacterial RNA load, pathogen dominance, and changes in the lung microbiome\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe began metatranscriptomic analyses by comparatively assessing bacterial RNA load in the lower respiratory tract microbiome of 2\u0026deg;BP patients versus No-BP controls. TA samples from 2\u0026deg;BP patients collected closest to date of clinical diagnosis had higher bacterial RNA load compared to samples from No-BP controls obtained at comparable timepoints post-intubation (P\u0026thinsp;=\u0026thinsp;0.0016, Fig.\u0026nbsp;2A). We next assessed lung microbiome alpha diversity and found that while median Shannon Diversity Index (SDI) was lower in 2\u0026deg;BP patients compared to No-BP controls, it did not significantly differ (P\u0026thinsp;=\u0026thinsp;0.10, Fig.\u0026nbsp;2B).\u003c/p\u003e\n\u003cp\u003eWhile in most patients the 2\u0026deg;BP pathogen was most abundant in the TA sample collected closest to the time of clinical 2\u0026deg;BP diagnosis, in 10/27 (37.0%) of patients, it was top ranked by abundance in samples collected earlier or later (e.g., Patient 1427, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). We thus repeated the alpha diversity analysis using the sample in which the 2\u0026deg;BP pathogen was highest ranked (as opposed to the sample closest to the date of 2\u0026deg;BP diagnosis), and found that SDI was significantly lower compared to No-BP controls (P\u0026thinsp;=\u0026thinsp;0.014, Fig.\u0026nbsp;2C). Longitudinal assessment of airway microbiome dynamics demonstrated that alpha diversity decreased over time in all groups following intubation, and was consistently lower in the 2\u0026deg;BP group (P\u0026thinsp;=\u0026thinsp;0.019, Fig.\u0026nbsp;2D).\u003c/p\u003e\n\u003cp\u003eWe more closely examined the relationship between the culture-confirmed 2\u0026deg;BP pathogen and the airway microbiome, and found that the 2\u0026deg;BP pathogen was detected by metatranscriptomics in all 27 2\u0026deg;BP cases. The 2\u0026deg;BP pathogen was dominant in the lower airway microbiome of at least one sample from most patients with 2\u0026deg;BP (Fig.\u0026nbsp;2E), with the culture-confirmed pathogen ranking in the top 3 most abundant taxa in at least one sample for 25/27 (92.6%) of cases within 7 days of clinical diagnosis (Fig.\u0026nbsp;2F).\u003c/p\u003e\n\u003cp\u003eFurther assessment of the most abundant taxa in lung microbiome demonstrated that \u003cem\u003eStaphylococcus aureus\u003c/em\u003e and \u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e were most commonly identified in 2\u0026deg;BP patients, while \u003cem\u003ePrevotella, Mycoplasma\u003c/em\u003e, and \u003cem\u003eAlloprevotella\u003c/em\u003e species were most commonly found in No-BP cases (Fig.\u0026nbsp;2G). Compositional differences in the lung microbiomes of 2\u0026deg;BP versus No-BP patients based on Bray Curtis dissimilarity index approached statistical significance (Fig.\u0026nbsp;2H, P\u0026thinsp;=\u0026thinsp;0.07 by PERMANOVA). Lastly, we asked whether SARS-CoV-2 viral load in the lung, as measured by metatranscriptomics in reads per million (rpM), differed between patients with or without 2\u0026deg;BP, but found no difference (P\u0026thinsp;=\u0026thinsp;0.22, Fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n\u003ch2\u003eDynamics of secondary bacterial pneumonia pathogens in the lung\u003c/h2\u003e\n\u003cp\u003eWe next evaluated the longitudinal dynamics of the cultured-confirmed 2\u0026deg;BP pathogen in the lower airway by overlaying abundance rank trajectories of the cultured pathogen(s) in the airway microbiome with phenotypic susceptibility to administered antibiotics (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). The culture-confirmed 2\u0026deg;BP pathogen could be detected in the airway prior to 2\u0026deg;BP diagnosis in 14/14 (100%) of the patients who had samples obtained prior to clinical diagnosis (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). Assessment of longitudinal trajectories revealed cases where pathogen expansion in the lower airway prior to 2\u0026deg;BP diagnosis occurred in the absence of any antibiotic treatment (e.g., Patient 1254, \u003cem\u003eS. aureus\u003c/em\u003e, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e), as well as following treatment with antibiotics to which the pathogen was resistant (e.g., Patient 1114, \u003cem\u003eP. aeruginosa\u003c/em\u003e, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). In patients with longitudinal sampling after clinical 2\u0026deg;BP diagnosis, we asked how pathogen clearance related to its susceptibility to administered antibiotics. We observed cases where initiation of antibiotics with known activity against the 2\u0026deg;BP pathogen resulted in a decrease in pathogen abundance, as expected (Patient 1231, \u003cem\u003eE. coli\u003c/em\u003e, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). Many pathogens, however, remained the most abundant microbe in the lower airway for days following initiation of antimicrobial therapy (e.g., Patient 1051, \u003cem\u003eP. aeruginosa\u003c/em\u003e, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). We noted that patients with \u003cem\u003eP. aeruginosa\u003c/em\u003e infections were more likely to exhibit this impaired clearance phenotype compared to patients with other types of 2\u0026deg;BP (P\u0026thinsp;=\u0026thinsp;0.0.017, Fig. S2).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n\u003ch2\u003eDetection of secondary bacterial pneumonia pathogens in the upper respiratory tract\u003c/h2\u003e\n\u003cp\u003eAmong the 15 patients who also had nasal swab (NS) samples available, we did not observe differences in nasal microbiome bacterial RNA load (Fig.\u0026nbsp;4A) or alpha diversity (Fig.\u0026nbsp;4B) based on 2\u0026deg;BP status. We noted that in 14/15 (93.3%) of cases, the 2\u0026deg;BP pathogen was detected in at least one NS sample within 7 days of clinical diagnosis, and in 7/15 (46.7%) cases it ranked in the top 3 most abundant (Fig.\u0026nbsp;4C). Among the 12 patients who had NS samples collected prior to the date of clinical 2\u0026deg;BP diagnosis, 8/9 (88.9%) had at least one etiologic pathogen detected (Fig.\u0026nbsp;4D).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n\u003ch2\u003eThe respiratory antimicrobial resistome\u003c/h2\u003e\n\u003cp\u003eAnalysis of the lower respiratory tract resistome of 2\u0026deg;BP patients revealed a diversity of AMR genes representing multiple different classes, including plasmid-transmissible extended spectrum beta lactamase (ESBL) (e.g., \u003cem\u003eCTX-M\u003c/em\u003e) and colistin-resistance genes (e.g., \u003cem\u003eMCR1\u003c/em\u003e) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e). In some cases, AMR genes associated with culture-confirmed resistant pathogens were detectable before clinical diagnosis of 2\u0026deg;BP (e.g., \u003cem\u003eCTX-M, SED-1\u003c/em\u003e, patient 1196) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e, Fig. S3). Identification of the \u003cem\u003eampC\u003c/em\u003e inducible beta lactamase was noted in patient 1154, from whom \u003cem\u003eK. aerogenes\u003c/em\u003e was identified by culture (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e). This pathogen continued to dominate the airway despite treatment with a beta-lactam antibiotic (piperacillin-tazobactam) to which the organism was phenotypically susceptible (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). \u003cem\u003eAmpC\u003c/em\u003e can be induced following beta lactam exposure in certain Enterobacteriaceae, resulting in reversal of phenotypic susceptibility, and in some cases clinical treatment failure\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eComparative assessment demonstrated that clinically relevant AMR genes detected in lower respiratory tract pathogens could also be identified in the upper airway in a subset of cases. For instance, in patient 1145 (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e), who had multi-drug-resistant \u003cem\u003eE. coli, K. pneumoniae\u003c/em\u003e, and methicillin-resistant \u003cem\u003eS. aureus\u003c/em\u003e grow from TA culture, \u003cem\u003eMCR-1, CTX-M\u003c/em\u003e and \u003cem\u003emecA\u003c/em\u003e were detected in both NS and TA samples. As in the lower respiratory tract, we found that clinically relevant AMR genes related to the 2\u0026deg;BP pathogen could in some cases be detected in the nares prior to clinical recognition of bacterial pneumonia (e.g., patient 1145, Fig. S2).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n\u003ch2\u003eHost transcriptional responses in COVID-19 secondary bacterial pneumonia\u003c/h2\u003e\n\u003cp\u003eWe next asked whether a lower respiratory transcriptional signature of 2\u0026deg;BP could be identified amidst the intense inflammatory state of severe COVID-19. A comparison of host gene expression between 2\u0026deg;BP and No-BP patients identified 226 differentially expressed genes (FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.1) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eA, Table S3). Gene set enrichment analysis (GSEA) revealed that 2\u0026deg;BP was characterized by downregulated TNFa signaling via NF-kB in (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eB, Table S4), suggesting that a state of suppressed antibacterial defense might characterize 2\u0026deg;BP in COVID-19 patients.\u003c/p\u003e\n\u003cp\u003eWe hypothesized that corticosteroid treatment might be contributing to this, and thus performed a secondary differential gene expression analysis of only patients treated with dexamethasone and other corticosteroids. This analysis yielded 4XX differentially expressed genes (FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.1) (Table S5), with no evidence of suppressed TNFa signaling on GSEA, supporting our hypothesis. GSEA instead demonstrated non-significant upregulation of pro-inflammatory signaling pathways such as IL-6/JAK/STAT-3 in 2\u0026deg;BP patients (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eC, Table S6).\u003c/p\u003e\n\u003cp\u003eWe further investigated connections between the airway transcriptome and microbiome in 2\u0026deg;BP patients by testing whether the bacterial RNA load correlated with host gene expression. Differential expression analysis identified 4784 significant genes (FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.1) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eD, Table S7). GSEA (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eF, Table S8) revealed a striking inverse relationship between bacterial RNA load and immune signaling pathways important for antibacterial defense (e.g., TNFa, IL-6, IL-2), which was also evident at the individual gene level (e.g., \u003cem\u003eHLA-DRB1\u003c/em\u003e, \u003cem\u003eC1QC\u003c/em\u003e) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eE). Together, these results suggested that a relative state of impaired immune defense may exist in COVID-19 patients who develop 2\u0026deg;BP, and that this may be influenced by treatment with corticosteroids.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this prospective observational study, we assessed respiratory tract microbial dynamics and host transcriptional responses associated with 2\u0026deg;BP in COVID-19 patients requiring invasive mechanical ventilation. Using comparative metatranscriptomics, we found that 2\u0026deg;BP is characterized by disruption of the lung microbiome, dominance of pathogenic bacteria that can be co-detected in the nares, and a lower airway transcriptome globally characterized by suppressed TNFa signaling.\u003c/p\u003e \u003cp\u003eBacterial superinfection is a well-established contributor to influenza mortality\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, yet in COVID-19 its role in mortality has been less clear\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. We found that 2\u0026deg;BP affected 39% of mechanically ventilated patients in our multicenter cohort, and was strongly associated with mortality. Our results notably differed from two important prior studies\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e which did not find clear links between secondary bacterial infection and mortality in COVID-19 patients. This discrepancy may be explained by our use of both an established case definition\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e and microbiological criteria to define 2\u0026deg;BP, as opposed to simply requiring a positive bacterial respiratory culture\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. We also found that patients who developed 2\u0026deg;BP were more likely to have received corticosteroids during their hospitalizations, suggesting that the therapeutic benefit of corticosteroids may come at the expense of increased 2\u0026deg;BP risk.\u003c/p\u003e \u003cp\u003eEarly identification and treatment of hospital-onset bacterial pneumonia can prevent adverse consequences including prolonged mechanical ventilation, inappropriate antibiotic exposure, and mortality\u003csup\u003e\u003cspan additionalcitationids=\"CR36\" citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. We found that 2\u0026deg;BP pathogens could be detected in the lower airway up to a week before clinical recognition of infection, and were frequently amongst the most abundant taxa in the lung microbiome in the days preceding culture-based detection. In some cases, we also detected pathogen-associated AMR genes before clinical diagnosis of 2\u0026deg;BP. These findings suggest the potential of metatranscriptomics to enable early detection of secondary bacterial infections, adding to a growing body of literature suggesting the benefits of this technology in the intensive care unit\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Further work in a larger, appropriately designed cohort is needed to assess the diagnostic performance of metatranscriptomics for 2\u0026deg;BP diagnosis.\u003c/p\u003e \u003cp\u003eLower respiratory tract infections are characterized by a loss of airway microbiome alpha diversity\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e, which is also observed over time in mechanically ventilated patients\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. We found that 2\u0026deg;BP further disrupts microbial community structure in patients with existing SARS-CoV-2 infection, leading to additional loss of diversity in the setting of bacterial pathogen dominance in the airway. In over a third of cases, the peak of bacterial pathogen dominance did not overlap with the date of 2\u0026deg;BP clinical diagnosis. This could represent a decoupling of physiologic responses and pathogen dynamics, heterogeneity in TA sampling, or reflect the challenge of identifying a new pneumonia amidst an existing severe viral lower respiratory tract infection. Our beta diversity analysis also suggested that COVID-19 patients who develop 2\u0026deg;BP may have differences in the composition of their airway microbiome.\u003c/p\u003e \u003cp\u003eBy characterizing relationships between 2\u0026deg;BP pathogens and the lung microbiome over time, we observed that in some patients, the pathogen remained dominant in the airway even after 2\u0026deg;BP clinical diagnosis and the initiation of appropriate antibiotics. In particular, all \u003cem\u003eP. aeruginosa\u003c/em\u003e cases exhibited this persistence trend, which may reflect the known tendency of this pathogen to form biofilms.\u003c/p\u003e \u003cp\u003eWe found that 2\u0026deg;BP pathogens were not only detectable in the lower airway, but also in the upper airway in more than half of the cases, including prior to clinical pneumonia diagnosis. Most patients hospitalized for bacterial pneumonia do not require invasive mechanical ventilation, thus it is possible that minimally invasive nasal swab specimens could have diagnostic value in such cases. Because this observational study included only intubated patients, future studies with paired bronchoscopy and nasal sampling in less critically ill patients will be needed to address this question.\u003c/p\u003e \u003cp\u003eAMR infections, in particular respiratory infections, are a major public health issue and have increased in prevalence during the COVID-19 pandemic\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. We observed several AMR genes and classes considered potential threats to the management of bacterial infections. These include the plasmid transmissible genes \u003cem\u003eMCR-1\u003c/em\u003e, \u003cem\u003eQnr-S\u003c/em\u003e and \u003cem\u003eCTX-M\u003c/em\u003e, which have a high potential for horizontal transfer both within patient and within hospital. We also identified several AMR genes not typically detectable by existing clinical assays. For instance, three patients from one study site were found to harbor the \u003cem\u003eSED-1\u003c/em\u003e class A beta lactamase (patients 1145, 1196, 1231), which was originally identified in \u003cem\u003eCitrobacter sedikiae\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, reported once in \u003cem\u003eE. coli\u003c/em\u003e contaminating produce in China\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, but not otherwise reported in the context of human infection.\u003c/p\u003e \u003cp\u003eIn several cases, we found that pathogen-associated AMR genes were detectable before clinical diagnosis of 2\u0026deg;BP, and were also found in the upper airway. While screening for \u003cem\u003emecA\u003c/em\u003e in the nares is routinely performed to identify patients at risk for invasive MRSA infections\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e, it has remained unclear whether this approach has value for other types of AMR infections. Our data suggest that screening for other clinically important AMR genes may prove useful for identifying patients with drug-resistant pneumonia, with implications for both antimicrobial stewardship and infection control.\u003c/p\u003e \u003cp\u003eSevere COVID-19 is characterized by a profoundly dysregulated host response in the lower respiratory tract\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, which complicates the detection of a transcriptional immune response to 2\u0026deg;BP. Nonetheless, our GSEA results suggest that impaired TNFa signaling may play a role in the pathophysiology of 2\u0026deg;BP. Indeed, increased rates of serious bacterial infections are well described in patients receiving anti-TNFa therapies\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e, and our findings suggest that a state of relative immunosuppression may augment susceptibility to opportunistic bacterial pathogens in COVID-19 patients.\u003c/p\u003e \u003cp\u003eConsistent with this idea is our finding that bacterial RNA load in the lungs strongly associated with downregulation of innate and adaptative immune pathways essential for antibacterial defense. This model is also in line with murine models of post-influenza 2\u0026deg;BP, which demonstrate virus-induced impairment of innate immunity\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e,\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e characterized by reduced expression of TNFa, IL-6, and other cytokines. Our secondary analysis of steroid recipients did not yield this signal of downregulated TNFa signaling, suggesting that corticosteroid treatment, which is now standard of care for severe COVID-19\u003csup\u003e50\u003c/sup\u003e, may contribute to this immunosuppressed state.\u003c/p\u003e \u003cp\u003eStrengths of our study include the novel use of host/microbe metatranscriptomics to study secondary bacterial infections, rigorous clinical adjudication of 2\u0026deg;BP, longitudinal sampling, and measurement of bacterial RNA load, a biomarker not previously evaluated in studies of pneumonia. Limitations include a relatively small sample size and incomplete longitudinal sampling for all patients. Our host transcriptomic findings require further exploration and validation in an independent cohort. While several public COVID-19 respiratory transcriptomic datasets exist, none include adjudication of 2\u0026deg;BP status using a rigorous and standardized definition.\u003c/p\u003e \u003cp\u003eTaken together, our findings shed light on the microbial dynamics and host immune responses of 2\u0026deg;BP, a clinically important and serious complication of COVID-19 and other viral respiratory illnesses. Future studies are needed to validate these findings, clarify mechanisms in cohorts with other viral infections, and evaluate the diagnostic potential of metatranscriptomics for early detection of 2\u0026deg;BP.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n\u003ch2\u003eStudy design, enrollment, and institutional review board approval\u003c/h2\u003e\n\u003cp\u003eWe conducted a prospective case-control study of hospitalized adults requiring mechanical ventilation for COVID-19 with or without secondary bacterial pneumonia. (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). All 397 patients with clinical polymerase chain reaction (PCR)-confirmed SARS-CoV-2 infection enrolled in the UCSF COVID-19 Multiphenotyping for Effective Therapy (COMET) observational study of patients with acute respiratory illnesses were initially considered for inclusion. Seventeen patients were co-enrolled in the National Institute of Allergy and Infectious Diseases-funded Immunophenotyping Assessment in a COVID-19 Cohort (IMPACC) Network study. This study was approved by the UCSF Institutional Review Board (IRB) under protocol #20-30497. Detailed enrollment and consent protocols for both studies have been previously described\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n\u003ch2\u003eSecondary bacterial pneumonia adjudication and patient inclusion\u003c/h2\u003e\n\u003cp\u003eAmongst the 112 critically ill COVID-19 patients requiring mechanical ventilation enrolled in COMET, N\u0026thinsp;=\u0026thinsp;44 cases of culture-confirmed secondary bacterial pneumonia (2\u0026deg;BP) were adjudicated by 3 study team infectious disease physicians (NS, CD, CRL) based on the United States Centers for Disease Control and Prevention PNEU1 surveillance definition of pneumonia\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e and all available clinical data in the electronic medical record, blinded to metatranscriptomic results. Study team physicians also assessed the antimicrobial susceptibility patterns of cultured microbes, and identified days that patients with 2\u0026deg;BP had received antibiotics to which their cultured microbes were susceptible and/or resistant (Supp. Table\u0026nbsp;5). The date of 2\u0026deg;BP clinical diagnosis was set as the date on which the positive bacterial culture was ordered by the treating medical team. Patients with no clinical evidence of bacterial pneumonia at any point during their hospitalization (No-BP group, N\u0026thinsp;=\u0026thinsp;41) were also identified. TA samples with metatranscriptomic data passing minimum quality control parameters (see below) were available for 29 2\u0026deg;BP patients and 29 No-BP patients. Within the 2\u0026deg;BP group, two additional patients were excluded from analyses as they did not have any TA samples collected within a 5-day window of 2\u0026deg;BP clinical diagnosis (-3 days to +\u0026thinsp;2 days). The remaining final cohort of 2\u0026deg;BP (N\u0026thinsp;=\u0026thinsp;27) and No-BP (N\u0026thinsp;=\u0026thinsp;29) patients was leveraged for all metatranscriptomic analyses, which included all samples from available within \u0026minus;\u0026thinsp;7 or +\u0026thinsp;7 days of clinical diagnosis.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n\u003ch2\u003eTracheal aspirate and nasal swab sampling\u003c/h2\u003e\n\u003cp\u003eFollowing enrollment, tracheal aspirate (TA) was collected periodically following intubation without addition of saline wash, and mixed 1:1 with DNA/RNA shield in tubes containing bashing beads (Zymo Research). Nasal swabs were collected into tubes prefilled with DNA/RNA shield (Zymo Research). Samples were frozen within one hour and stored at -80C until nucleic acid extraction.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n\u003ch2\u003eRNA sequencing\u003c/h2\u003e\n\u003cp\u003eTo evaluate host and microbial gene expression, metatranscriptomic RNA sequencing (RNA-seq) was performed on TA specimens. Following RNA extraction from 320\u0026micro;L of input sample (Zymo Pathogen Magbead Kit) and DNase treatment, human cytosolic and mitochondrial ribosomal RNA was depleted using FastSelect (Qiagen). To control for background contamination, we included negative controls (water and HeLa cell RNA) as well as positive controls (spike-in RNA standards from the External RNA Controls Consortium (ERCC))\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. RNA was then fragmented and underwent library preparation using the NEBNext Ultra II RNA-seq Kit (New England Biolabs). Libraries underwent 146 nucleotide paired-end Illumina sequencing on an Illumina Novaseq 6000.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n\u003ch2\u003eQuality control and mitigation of environmental contaminants\u003c/h2\u003e\n\u003cp\u003eTo minimize inaccurate taxonomic assignments due to environmental and reagent derived contaminants, non-templated \u0026ldquo;water only\u0026rdquo; and HeLa cell RNA controls were processed with each group of samples that underwent nucleic acid extraction. These were included, as well as positive control clinical samples, with each sequencing run. Negative control samples enabled estimation of the number of background reads expected for each taxon. Microbes established as metagenomic contaminants were filtered out (Sphingomonas, Bradyrhizobium, Ralstonia, Delftia, Cutibacterium, Methylobacterium, Acidovorax, Chryseobacterium, Burkholderia). Samples were excluded from analysis if they had fewer than 1,000 total reads mapping to bacterial taxa or a bacterial mass of \u0026lt;\u0026thinsp;1 pg.\u003c/p\u003e\n\u003cp\u003eA single sample per patient was utilized for analyses that involved direct comparison at a single timepoint, including bacterial RNA load, alpha diversity, beta diversity, top pathogen identification, and TA only, host gene expression analyses. The TA or NS samples collected closest to the date of clinical 2\u0026deg;BP diagnosis, defined as the date the positive culture was ordered by the clinical treatment team, were used in the primary timepoint analyses. To identify No-BP control samples appropriately matched by time from intubation, we selected samples collected closest to the median days post-intubation of the comparable 2\u0026deg;BP TA (N\u0026thinsp;=\u0026thinsp;6 days) or NS (N\u0026thinsp;=\u0026thinsp;5 days) samples. Because both host and microbial analyses were carried out for TA single timepoint comparison analyses, those samples were required to meet both host and microbe quality control standards.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n\u003ch2\u003eStatistics\u003c/h2\u003e\n\u003cp\u003eStatistical significance was defined as P\u0026thinsp;\u0026lt;\u0026thinsp;0.05 using two-tailed tests, unless stated otherwise. Categorical data were analyzed by Fisher\u0026rsquo;s exact test and nonparametric continuous variables were analyzed by Mann-Whitney. Statistical approaches used for gene expression and microbiome analyses are detailed in each respective \u003cspan class=\"InternalRef\"\u003eMethods\u003c/span\u003e section.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n\u003ch2\u003eLung microbiome analysis\u003c/h2\u003e\n\u003cp\u003eTaxonomic alignments were obtained from raw sequencing reads using the CZID pipeline\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e52\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e, which performs quality filtration and removal of human reads followed by reference-based taxonomic alignment against sequences in the National Center for Biotechnology Information (NCBI) nucleotide (NT) database, followed by assembly of reads matching each taxon detected. Taxonomic alignments underwent background correction for environmental contaminants (see below), viruses were excluded, and data was then aggregated to the genus level before calculating diversity metrics. Alpha diversity (Shannon\u0026rsquo;s Diversity Index) and beta diversity (Bray-Curtis dissimilarity) were calculated and the latter plotted using non-metric multidimensional scaling (NDMS).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n\u003ch2\u003eBacterial RNA load/mass calculations\u003c/h2\u003e\n\u003cp\u003eBacterial RNA load was calculated based on the ratio of bacterial reads in each sample to total reads aligning to the External RNA Controls Consortium (ERCC) RNA mass standards spiked into each sample\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. The following equation was utilized for this calculation: [ERCC input mass]/[bacterial mass] = [ERCC reads]/[bacterial reads].\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n\u003ch2\u003eHost differential expression\u003c/h2\u003e\n\u003cp\u003eFollowing demultiplexing, sequencing reads were pseudo-aligned with kallisto\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e to an index consisting of all transcripts associated with human protein coding genes (ENSEMBL release 99), long non-coding RNA, cytosolic and mitochondrial ribosomal RNA sequences and the sequences of ERCC RNA standards. Gene-level counts were generated from the transcript-level abundance estimates using the R package tximport\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e, with the scaledTPM method. For quality control, we only retained samples with a total of at least 1,000,000 estimated protein-coding gene counts, and a proportion of ribosomal RNA to total RNA \u0026le; 50%, according to a previously described approach\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. In addition, we only analyzed host genes with at least 10 counts in at least 20% of samples.\u003c/p\u003e\n\u003cp\u003eDifferential expression analysis was performed in R (v4.3.2) using the package limma-voom\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e (v3.58.1). For the comparison between 2\u0026deg;BP and No-BP patients, we adjusted for SARS-CoV-2 viral load rpM by adding the coefficient log10(rpM\u0026thinsp;+\u0026thinsp;1) to the linear model (without adjusting for any other covariates). For the analysis of corticosteroid recipients, we restricted to patients who had received treatment with steroids at any time prior to sample collection. For the analysis of host gene expression and bacterial mass, we modelled gene expression of 2\u0026deg;BP patients on log10-transformed bacterial mass (without adjusting for any other covariates). Significant genes were identified using a Benjamini-Hochberg false discovery rate (FDR)\u0026thinsp;\u0026lt;\u0026thinsp;0.1 (for 2\u0026deg;BP vs No-BP analysis) or \u0026lt;\u0026thinsp;0.05 (for the host gene vs bacterial mass analysis). Differential expression analysis results are provided in (Supp. Tables\u0026nbsp;1 and 3).\u003c/p\u003e\n\u003cp\u003eGene set enrichment analysis (GSEA) was performed using the package fgsea (v1.28.0), and the Hallmark pathways were obtained from the package msigdbr (v7.5.1). The t-statistics (obtained from limma\u0026rsquo;s differential expression analysis) were used to rank all genes, and used as input for the fgseaMultilevel function (minSize\u0026thinsp;=\u0026thinsp;15, maxSize\u0026thinsp;=\u0026thinsp;500). Pathways with adjusted P-value below 0.05 were considered statistically significant.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData and code availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe raw fastq files with microbial sequencing reads are available under NCBI BioProject ID: PRJNA1033689. The host gene counts are available under NCBI Gene Expression Omnibus (GEO) accession number: GSE246795. The human raw sequencing data are protected due to data privacy restrictions from the IRB protocol governing patient enrollment, which protects the release of raw genetic sequencing data from those patients enrolled under a waiver of consent. To honor this, researchers who wish to obtain raw fastq files for the purposes of independently generating gene counts can contact the corresponding author ([email protected]) and request to be added to the IRB protocol. All code and source data used for analyses can be found at: (https://github.com/chazlangelier/2BP).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNHLBI, NIAID, Chan Zuckerberg Biohub\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eLuyt, C.-E. \u003cem\u003eet al.\u003c/em\u003e Ventilator-associated pneumonia in patients with SARS-CoV-2-associated acute respiratory distress syndrome requiring ECMO: a retrospective cohort study. \u003cem\u003eAnn Intensive Care\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 158 (2020).\u003c/li\u003e\n\u003cli\u003eBardi, T. \u003cem\u003eet al.\u003c/em\u003e Nosocomial infections associated to COVID-19 in the intensive care unit: clinical characteristics and outcome. \u003cem\u003eEur J Clin Microbiol Infect Dis\u003c/em\u003e (2021) doi:10.1007/s10096-020-04142-w.\u003c/li\u003e\n\u003cli\u003eMaes, M. \u003cem\u003eet al.\u003c/em\u003e Ventilator-associated pneumonia in critically ill patients with COVID-19. \u003cem\u003eCrit Care\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 25 (2021).\u003c/li\u003e\n\u003cli\u003eS\u0026oslash;gaard, K. K. \u003cem\u003eet al.\u003c/em\u003e Community-acquired and hospital-acquired respiratory tract infection and bloodstream infection in patients hospitalized with COVID-19 pneumonia. \u003cem\u003eJ Intensive Care\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, (2021).\u003c/li\u003e\n\u003cli\u003eRouz\u0026eacute;, A. \u003cem\u003eet al.\u003c/em\u003e Relationship between SARS-CoV-2 infection and the incidence of ventilator-associated lower respiratory tract infections: a European multicenter cohort study. \u003cem\u003eIntensive Care Med\u003c/em\u003e 1\u0026ndash;11 (2021) doi:10.1007/s00134-020-06323-9.\u003c/li\u003e\n\u003cli\u003eDeVoe, C. \u003cem\u003eet al.\u003c/em\u003e Increased Rates of Secondary Bacterial Infections, Including Enterococcus Bacteremia, in Patients Hospitalized with COVID-19. \u003cem\u003eInfect Control Hosp Epidemiol\u003c/em\u003e 1\u0026ndash;33 (2021) doi:10.1017/ice.2021.391.\u003c/li\u003e\n\u003cli\u003eFalsey, A. R. \u003cem\u003eet al.\u003c/em\u003e Bacterial Complications of Respiratory Tract Viral Illness: A Comprehensive Evaluation. \u003cem\u003eThe Journal of Infectious Diseases\u003c/em\u003e \u003cstrong\u003e208\u003c/strong\u003e, 432\u0026ndash;441 (2013).\u003c/li\u003e\n\u003cli\u003eReyes, L. F. \u003cem\u003eet al.\u003c/em\u003e Risk factors for developing ventilator-associated lower respiratory tract infection in patients with severe COVID-19: a multinational, multicentre study, prospective, observational study. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 6553 (2023).\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\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\u003eKalantar, K. L. \u0026amp; Langelier, C. R. Host-Microbe Metagenomics: a Lens To Refocus Our Perspective on Infectious and Inflammatory Diseases. \u003cem\u003emSystems\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, e00404-21 (2021).\u003c/li\u003e\n\u003cli\u003eLangelier, C. \u003cem\u003eet al.\u003c/em\u003e Integrating host response and unbiased microbe detection for lower respiratory tract infection diagnosis in critically ill adults. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e \u003cstrong\u003e115\u003c/strong\u003e, E12353\u0026ndash;E12362 (2018).\u003c/li\u003e\n\u003cli\u003eMcCullers, J. A. The co-pathogenesis of influenza viruses with bacteria in the lung. \u003cem\u003eNat. Rev. Microbiol.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 252\u0026ndash;262 (2014).\u003c/li\u003e\n\u003cli\u003eMorens, D. M., Taubenberger, J. K. \u0026amp; Fauci, A. S. Predominant role of bacterial pneumonia as a cause of death in pandemic influenza: implications for pandemic influenza preparedness. \u003cem\u003eJ Infect Dis\u003c/em\u003e \u003cstrong\u003e198\u003c/strong\u003e, 962\u0026ndash;970 (2008).\u003c/li\u003e\n\u003cli\u003eKaul, D. \u003cem\u003eet al.\u003c/em\u003e Microbiome disturbance and resilience dynamics of the upper respiratory tract during influenza A virus infection. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 2537 (2020).\u003c/li\u003e\n\u003cli\u003eFlanagan, J. L. \u003cem\u003eet al.\u003c/em\u003e Loss of bacterial diversity during antibiotic treatment of intubated patients colonized with Pseudomonas aeruginosa. \u003cem\u003eJ Clin Microbiol\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, 1954\u0026ndash;1962 (2007).\u003c/li\u003e\n\u003cli\u003eKalantar, K. L. \u003cem\u003eet al.\u003c/em\u003e Metagenomic comparison of tracheal aspirate and mini-bronchial alveolar lavage for assessment of respiratory microbiota. \u003cem\u003eAm. J. Physiol. Lung Cell Mol. Physiol.\u003c/em\u003e \u003cstrong\u003e316\u003c/strong\u003e, L578\u0026ndash;L584 (2019).\u003c/li\u003e\n\u003cli\u003eDur\u0026aacute;n-Manuel, E. M. \u003cem\u003eet al.\u003c/em\u003e Massive sequencing of the V3-V4 hypervariable region of bronchoalveolar lavage from patients with COVID-19 and VAP reveals the collapse of the pulmonary microbiota. \u003cem\u003eJ Med Microbiol\u003c/em\u003e \u003cstrong\u003e71\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eBallinger, M. N. \u0026amp; Standiford, T. J. Innate Immune Responses in Ventilator-Associated Pneumonia. in \u003cem\u003eMucosal Immunology of Acute Bacterial Pneumonia\u003c/em\u003e (ed. Prince, A.) 185\u0026ndash;212 (Springer New York, 2013). doi:10.1007/978-1-4614-5326-0_8.\u003c/li\u003e\n\u003cli\u003eWunderink, R. G. Nosocomial Pneumonia, Including Ventilator-associatedPneumonia. \u003cem\u003eProc Am Thorac Soc\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 440\u0026ndash;444 (2005).\u003c/li\u003e\n\u003cli\u003eBlanco-Melo, Daniel. Imbalanced host response to SARS-CoV-2 drives development of COVID-19. \u003cem\u003eCell.\u003c/em\u003e doi:10.1016/j.cell.2020.04.026.\u003c/li\u003e\n\u003cli\u003eSarma, A. \u003cem\u003eet al.\u003c/em\u003e Tracheal aspirate RNA sequencing identifies distinct immunological features of COVID-19 ARDS. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 5152 (2021).\u003c/li\u003e\n\u003cli\u003eNguyen, E. V. \u003cem\u003eet al.\u003c/em\u003e Proteomic Profiling of Bronchoalveolar Lavage Fluid in Critically Ill Patients with Ventilator-Associated Pneumonia. \u003cem\u003ePLoS ONE\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, e58782 (2013).\u003c/li\u003e\n\u003cli\u003eConway Morris, A. \u003cem\u003eet al.\u003c/em\u003e Diagnostic importance of pulmonary interleukin-1 and interleukin-8 in ventilator-associated pneumonia. \u003cem\u003eThorax\u003c/em\u003e \u003cstrong\u003e65\u003c/strong\u003e, 201\u0026ndash;207 (2010).\u003c/li\u003e\n\u003cli\u003ePathak, K. V. \u003cem\u003eet al.\u003c/em\u003e Molecular Profiling of Innate Immune Response Mechanisms in Ventilator-associated Pneumonia. \u003cem\u003eMolecular \u0026amp; Cellular Proteomics\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 1688\u0026ndash;1705 (2020).\u003c/li\u003e\n\u003cli\u003eRobinson, K. M., Kolls, J. K. \u0026amp; Alcorn, J. F. The immunology of influenza virus-associated bacterial pneumonia. \u003cem\u003eCurrent Opinion in Immunology\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 59\u0026ndash;67 (2015).\u003c/li\u003e\n\u003cli\u003eConway Morris, A. \u003cem\u003eet al.\u003c/em\u003e Cell-surface signatures of immune dysfunction risk-stratify critically ill patients: INFECT study. \u003cem\u003eIntensive Care Med\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, 627\u0026ndash;635 (2018).\u003c/li\u003e\n\u003cli\u003eSulaiman, I. \u003cem\u003eet al.\u003c/em\u003e Microbial signatures in the lower airways of mechanically ventilated COVID-19 patients associated with poor clinical outcome. \u003cem\u003eNat Microbiol\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 1245\u0026ndash;1258 (2021).\u003c/li\u003e\n\u003cli\u003eCharalampous, T. \u003cem\u003eet al.\u003c/em\u003e Evaluating the potential for respiratory metagenomics to improve treatment of secondary infection and detection of nosocomial transmission on expanded COVID-19 intensive care units. \u003cem\u003eGenome Med\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 182 (2021).\u003c/li\u003e\n\u003cli\u003eWu, N. \u003cem\u003eet al.\u003c/em\u003e Rapid identification of pathogens associated with ventilator-associated pneumonia by Nanopore sequencing. \u003cem\u003eRespir Res\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 310 (2021).\u003c/li\u003e\n\u003cli\u003eUnited States Centers for Disease Control and Prevention. CDC/NHSN Surveillance Definitions for Specific Types of Infections. (2021).\u003c/li\u003e\n\u003cli\u003eTamma, P. D. \u003cem\u003eet al.\u003c/em\u003e Infectious Diseases Society of America Guidance on the Treatment of AmpC \u0026beta;-Lactamase\u0026ndash;Producing Enterobacterales, Carbapenem-Resistant \u003cem\u003eAcinetobacter baumannii\u003c/em\u003e , and \u003cem\u003eStenotrophomonas maltophilia\u003c/em\u003e Infections. \u003cem\u003eClinical Infectious Diseases\u003c/em\u003e \u003cstrong\u003e74\u003c/strong\u003e, 2089\u0026ndash;2114 (2022).\u003c/li\u003e\n\u003cli\u003eZhou, F. \u003cem\u003eet al.\u003c/em\u003e Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. \u003cem\u003eThe Lancet\u003c/em\u003e \u003cstrong\u003e395\u003c/strong\u003e, 1054\u0026ndash;1062 (2020).\u003c/li\u003e\n\u003cli\u003ePickens, C. O. \u003cem\u003eet al.\u003c/em\u003e Bacterial Superinfection Pneumonia in Patients Mechanically Ventilated for COVID-19 Pneumonia. \u003cem\u003eAm J Respir Crit Care Med\u003c/em\u003e \u003cstrong\u003e204\u003c/strong\u003e, 921\u0026ndash;932 (2021).\u003c/li\u003e\n\u003cli\u003eKuti, E. L., Patel, A. A. \u0026amp; Coleman, C. I. Impact of inappropriate antibiotic therapy on mortality in patients with ventilator-associated pneumonia and blood stream infection: A meta-analysis. \u003cem\u003eJournal of Critical Care\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 91\u0026ndash;100 (2008).\u003c/li\u003e\n\u003cli\u003eICU-Acquired Pneumonia Study Group \u0026amp; Alvarez-Lerma, F. Modification of empiric antibiotic treatment in patients with pneumonia acquired in the intensive care unit. \u003cem\u003eIntensive Care Med\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 387\u0026ndash;394 (1996).\u003c/li\u003e\n\u003cli\u003eCelis, R. \u003cem\u003eet al.\u003c/em\u003e Nosocomial Pneumonia. \u003cem\u003eChest\u003c/em\u003e \u003cstrong\u003e93\u003c/strong\u003e, 318\u0026ndash;324 (1988).\u003c/li\u003e\n\u003cli\u003eMick, E. \u003cem\u003eet al.\u003c/em\u003e Integrated host/microbe metagenomics enables accurate lower respiratory tract infection diagnosis in critically ill children. \u003cem\u003eJ Clin Invest\u003c/em\u003e \u003cstrong\u003e133\u003c/strong\u003e, e165904 (2023).\u003c/li\u003e\n\u003cli\u003eDickson, R. P., Erb-Downward, J. R. \u0026amp; Huffnagle, G. B. Towards an ecology of the lung: new conceptual models of pulmonary microbiology and pneumonia pathogenesis. \u003cem\u003eLancet Respir Med\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 238\u0026ndash;246 (2014).\u003c/li\u003e\n\u003cli\u003eZakharkina, T. \u003cem\u003eet al.\u003c/em\u003e The dynamics of the pulmonary microbiome during mechanical ventilation in the intensive care unit and the association with occurrence of pneumonia. \u003cem\u003eThorax\u003c/em\u003e \u003cstrong\u003e72\u003c/strong\u003e, 803\u0026ndash;810 (2017).\u003c/li\u003e\n\u003cli\u003ePettigrew, M. M., Tanner, W. \u0026amp; Harris, A. D. The Lung Microbiome and Pneumonia. \u003cem\u003eThe Journal of Infectious Diseases\u003c/em\u003e \u003cstrong\u003e223\u003c/strong\u003e, S241\u0026ndash;S245 (2021).\u003c/li\u003e\n\u003cli\u003ePetrella, S., Clermont, D., Casin, I., Jarlier, V. \u0026amp; Sougakoff, W. Novel Class A \u0026beta;-Lactamase Sed-1 from \u003cem\u003eCitrobacter sedlakii\u003c/em\u003e : Genetic Diversity of \u0026beta;-Lactamases within the \u003cem\u003eCitrobacter\u003c/em\u003e Genus. \u003cem\u003eAntimicrob Agents Chemother\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, 2287\u0026ndash;2298 (2001).\u003c/li\u003e\n\u003cli\u003eLiao, N. \u003cem\u003eet al.\u003c/em\u003e Prevalence of \u0026beta;-Lactam Drug-Resistance Genes in \u003cem\u003eEscherichia coli\u003c/em\u003e Contaminating Ready-to-Eat Lettuce. \u003cem\u003eFoodborne Pathogens and Disease\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 739\u0026ndash;742 (2020).\u003c/li\u003e\n\u003cli\u003eParente, D. M., Cunha, C. B., Mylonakis, E. \u0026amp; Timbrook, T. T. The Clinical Utility of Methicillin-Resistant Staphylococcus aureus (MRSA) Nasal Screening to Rule Out MRSA Pneumonia: A Diagnostic Meta-analysis With Antimicrobial Stewardship Implications. \u003cem\u003eClinical Infectious Diseases\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 1\u0026ndash;7 (2018).\u003c/li\u003e\n\u003cli\u003eSarma, A. \u003cem\u003eet al.\u003c/em\u003e COVID-19 ARDS is characterized by a dysregulated host response that differs from cytokine storm and is modified by dexamethasone. \u003cem\u003eRes Sq\u003c/em\u003e (2021) doi:10.21203/rs.3.rs-141578/v1.\u003c/li\u003e\n\u003cli\u003eMick, E. \u003cem\u003eet al.\u003c/em\u003e Upper airway gene expression reveals suppressed immune responses to SARS-CoV-2 compared with other respiratory viruses. \u003cem\u003eNature Communications\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 5854 (2020).\u003c/li\u003e\n\u003cli\u003eHochman, D. \u0026amp; Wolff, B. Risk of Serious Infections and Malignancies With Anti-TNF Antibody Therapy in Rheumatoid Arthritis. \u003cem\u003eJAMA\u003c/em\u003e \u003cstrong\u003e296\u003c/strong\u003e, 2201 (2006).\u003c/li\u003e\n\u003cli\u003eSun, K. \u0026amp; Metzger, D. W. Inhibition of pulmonary antibacterial defense by interferon-gamma during recovery from influenza infection. \u003cem\u003eNat Med\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 558\u0026ndash;564 (2008).\u003c/li\u003e\n\u003cli\u003eGoulding, J. \u003cem\u003eet al.\u003c/em\u003e Lowering the threshold of lung innate immune cell activation alters susceptibility to secondary bacterial superinfection. \u003cem\u003eJ Infect Dis\u003c/em\u003e \u003cstrong\u003e204\u003c/strong\u003e, 1086\u0026ndash;1094 (2011).\u003c/li\u003e\n\u003cli\u003eBhimraj, A. \u003cem\u003eet al.\u003c/em\u003e Infectious Diseases Society of America Guidelines on the Treatment and Management of Patients With Coronavirus Disease 2019 (COVID-19). \u003cem\u003eClinical Infectious Diseases\u003c/em\u003e ciac724 (2022) doi:10.1093/cid/ciac724.\u003c/li\u003e\n\u003cli\u003ePine, P. S. \u003cem\u003eet al.\u003c/em\u003e Evaluation of the External RNA Controls Consortium (ERCC) reference material using a modified Latin square design. \u003cem\u003eBMC Biotechnol\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 54 (2016).\u003c/li\u003e\n\u003cli\u003eRamesh, A. \u003cem\u003eet al.\u003c/em\u003e Metagenomic next-generation sequencing of samples from pediatric febrile illness in Tororo, Uganda. \u003cem\u003ePLoS One\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, e0218318 (2019).\u003c/li\u003e\n\u003cli\u003eKalantar, K. L. \u003cem\u003eet al.\u003c/em\u003e IDseq\u0026mdash;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, (2020).\u003c/li\u003e\n\u003cli\u003ePine, P. S. \u003cem\u003eet al.\u003c/em\u003e Evaluation of the External RNA Controls Consortium (ERCC) reference material using a modified Latin square design. \u003cem\u003eBMC Biotechnol\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 54 (2016).\u003c/li\u003e\n\u003cli\u003eBray, N. L., Pimentel, H., Melsted, P. \u0026amp; Pachter, L. Near-optimal probabilistic RNA-seq quantification. \u003cem\u003eNature Biotechnology\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 525\u0026ndash;527 (2016).\u003c/li\u003e\n\u003cli\u003eSoneson, C., Love, M. I. \u0026amp; Robinson, M. D. Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences. \u003cem\u003eF1000Res\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 1521 (2015).\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 Research\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, e47\u0026ndash;e47 (2015).\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":"[email protected]","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-3877429/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3877429/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSecondary bacterial pneumonia (2\u0026deg;BP) is associated with significant morbidity following respiratory viral infection, yet mechanistically remains incompletely understood. In a prospective cohort of 112 critically ill adults intubated for COVID-19, we comparatively assessed longitudinal airway microbiome dynamics and studied the pulmonary transcriptome of patients who developed 2\u0026deg;BP versus controls who did not. We found that 2\u0026deg;BP was significantly associated with both mortality and corticosteroid treatment. The pulmonary microbiome in 2\u0026deg;BP was characterized by increased bacterial RNA load, dominance of culture-confirmed pathogens, and lower alpha diversity. Bacterial pathogens were detectable days prior to 2\u0026deg;BP clinical diagnosis, and in most cases were also present in nasal swabs. Pathogen antimicrobial resistance genes were also detectable in both the lower airway and nasal samples, and in some cases were identified prior to 2\u0026deg;BP clinical diagnosis. Assessment of the pulmonary transcriptome revealed suppressed TNFa signaling via NF-kB in patients who developed 2\u0026deg;BP, and a sub-analysis suggested that this finding was mediated by corticosteroid treatment. Within the 2\u0026deg;BP group, we observed a striking inverse correlation between innate and adaptive immune gene expression and bacterial RNA load. Together, our findings provide fresh insights into the microbial dynamics and host immune features of COVID-19-associated 2\u0026deg;BP.\u003c/p\u003e","manuscriptTitle":"Microbial Dynamics and Pulmonary Immune Responses in COVID-19 Secondary Bacterial Pneumonia","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-02-06 17:40:08","doi":"10.21203/rs.3.rs-3877429/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"9a384e93-a3aa-480f-8459-d225ebb01c3b","owner":[],"postedDate":"February 6th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":28352035,"name":"Health sciences/Medical research/Translational research"},{"id":28352036,"name":"Biological sciences/Microbiology/Microbial communities/Metagenomics"},{"id":28352037,"name":"Biological sciences/Microbiology/Virology/SARS-CoV-2"},{"id":28352038,"name":"Health sciences/Diseases/Infectious diseases/Bacterial infection"}],"tags":[],"updatedAt":"2024-10-30T07:06:13+00:00","versionOfRecord":{"articleIdentity":"rs-3877429","link":"https://doi.org/10.1038/s41467-024-53566-x","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2024-10-29 04:00:00","publishedOnDateReadable":"October 29th, 2024"},"versionCreatedAt":"2024-02-06 17:40:08","video":"","vorDoi":"10.1038/s41467-024-53566-x","vorDoiUrl":"https://doi.org/10.1038/s41467-024-53566-x","workflowStages":[]},"version":"v1","identity":"rs-3877429","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3877429","identity":"rs-3877429","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}