Mapping Longitudinal Lung Mycobiome Characteristics of Severe COVID-19 patients, a prospective, multicenter cohort study

Objective While many studies have confirmed a significant association between COVID-19 and invasive pulmonary aspergillosis, no study has yet characterized the longitudinal changes of the lung mycobiome in critically ill COVID-19 patients. Study design This prospective, multicenter, longitudinal cohort study included critically ill COVID-19 patients admitted to the ICU from five medical centers. We performed internal transcribed spacer (ITS) sequencing on these samples and ITS digital droplet PCR on BALF to quantify the fungal load. The study analyzed longitudinal changes of lung mycobiome in COVID-19 patients. Additionally, the characteristics of gut mycobiome have also been analyzed. Results Among the 61 patients included, 109 BALF and 72 fecal samples were collected. The absolute abundance of the mycobiome remained consistent across different hospitals, with no significant differences observed throughout the ICU stay. However, alpha diversity of the lung mycobiome increased in surviving patients, while beta diversity changes were more pronounced in deceased patients as hospitalization progressed. Although neither lung mycobiome composition nor clinical features alone could predict 28-day mortality, combining both significantly improved the prediction (AUC = 0.811). Besides, the gut mycobiome cannot predict the clinical prognosis of patients. Conclusion Our study mapped longitudinal mycobiome changes in the lower respiratory tract of critically ill patients. The mycobiome in the lungs of COVID-19 patients remains stable in the early stages after ICU admission, but significant differences emerge later. In COVID-19, the mycobiome in the lungs seems to be more predictive of a patient's clinical prognosis compared to the gut.