Abstract Rationale Approximately 50-70% of patients will receive antibiotics during their ICU stay. The use of antimicrobials is not without consequences. Beyond concerns regarding antibiotic resistance, recent technological advances in the field of metagenomic next generation sequencing (mNGS) have highlighted a pernicious effect of antibiotic use: disruption of the lung microbiome, or dysbiosis. The lung microbiome refers to the resident community of microbes that exist in healthy lungs. Dysbiosis is associated with a variety of pulmonary diseases and an altered host immune response. The causal relationship between dysbiosis, host immune dysregulation, and various pulmonary diseases is currently undefined; however, given the widespread, empiric use of antibiotics and their inherent ability to disrupt the lung microbiome, we hypothesize that antibiotic exposure will adversely affect clinical outcomes. Methods We recruited a cohort of 245 patients requiring invasive mechanical ventilation secondary to COVID-19. Each patient underwent lower airway sampling with collection of bronchoalveolar lavage fluid (BAL) or tracheal aspirates (TA) throughout their hospitalization. The exposure to antibiotics for a period of seven days leading up to intubation was quantified using the Antibiotic Exposure Score (AES). We performed metatranscriptomic and metagenomic analysis by RNA sequencing to characterize the lung microbiome, antimicrobial resistance genes (resistome), and host immunity. We used Mann-Whitney comparisons and logistical regression controlling for covariance to determine if antibiotic exposure was independently associated with clinical outcomes. Results We compared patients who survived to hospital discharge with those who did not and analyzed the AES of the first lower airway sample after intubation from survivors (Figure 1A survivors median AES 92.125 vs. 142.5 for non-survivors, p = 0.003). Greater exposure to antibiotics prior to intubation was associated with increased mortality. The difference between survival could not be explained by disease severity according to Acute Physiology and Chronic Health Evaluation (APACHE) score at time of intubation nor did it correlate with hospital length of stay (LOS), ventilator days, or ICU days (Figure 1B). Conclusions We demonstrated that early antibiotic exposure in a cohort of critically ill patients requiring invasive mechanical ventilation is associated with increased overall mortality. To elucidate other contributors or confounders, we are analyzing and comparing the lung microbiome, lung resistome, and host immune response between these groups. Further analysis will highlight the prognostic, diagnostic, and therapeutic role of the lung microbiome and response to broad-spectrum antibiotics in this cohort. This abstract is funded by: Veterans Affairs IK2BX005309-01A2 (BGW); National Center for Advancing Translational Sciences (NCATS), National Institutes of Health, through Grant Award Number KL2TR001446 (BGW); NYU CTSA grant (TL1 TR001447) from the National Center for Advancing Translational Sciences, National Institutes of Health (CJC);CHEST Foundation Research Grant in Chronic Obstructive Pulmonary Disease (BGW); R37 CA244775 (LNS, NCI/NIH); R33 GM147800 (LNS, NIGMS/NIH); U01 AG088351 (LNS, NIA/NIH)
Gordon et al. (Fri,) studied this question.