Lung Microbiota Is Related to Smoking Status and to Development of Acute Respiratory Distress Syndrome in Critically Ill Trauma Patients

Abstract

Rationale: Cigarette smoking is associated with increased risk of acute respiratory distress syndrome (ARDS) in patients after severe trauma; however, the mechanisms underlying this association are unknown.

Objectives: To determine whether cigarette smoking contributes to ARDS development after trauma by altering community composition of the lung microbiota.

Methods: We studied the lung microbiota of mechanically ventilated patients admitted to the ICU after severe blunt trauma. To do so, we used 16S ribosomal RNA gene amplicon sequencing of endotracheal aspirate samples obtained on ICU admission (n = 74) and at 48 hours after admission (n = 30). Cigarette smoke exposure (quantified using plasma cotinine), ARDS development, and other clinical parameters were correlated with lung microbiota composition.

Measurements and main results: Smoking status was significantly associated with lung bacterial community composition at ICU admission (P = 0.007 by permutational multivariate ANOVA [PERMANOVA]) and at 48 hours (P = 0.03 by PERMANOVA), as well as with significant enrichment of potential pathogens, including Streptococcus, Fusobacterium, Prevotella, Haemophilus, and Treponema. ARDS development was associated with lung community composition at 48 hours (P = 0.04 by PERMANOVA) and was characterized by relative enrichment of Enterobacteriaceae and of specific taxa enriched at baseline in smokers, including Prevotella and Fusobacterium.

Conclusions: After severe blunt trauma, a history of smoking is related to lung microbiota composition, both at the time of ICU admission and at 48 hours. ARDS development is also correlated with respiratory microbial community structure at 48 hours and with taxa that are relatively enriched in smokers at ICU admission. The data derived from this pilot study suggest that smoking-related changes in the lung microbiota could be related to ARDS development after severe trauma.

Keywords: acute lung injury; acute respiratory distress syndrome; cigarette smoking; microbiota; trauma.

Figure 1.

(A) Principal coordinates analysis plot (Canberra distance matrix) of 0-hour samples colored by smoking status on the basis of cotinine level. The r² value represents the percent variance explained, and the P value denotes the significance of this percentage. (B) Discriminatory taxa (0-h three-model analysis comparing smokers with nonsmokers) labeled by genus. *Fusobacterium OTU4319899; **Prevotella OTU4447248. OTU = operational taxonomic unit; PC = principal component.

Figure 2.

(A) Procrustes analysis comparing 0-hour and 48-hour sample weighted UniFrac coordinate matrices (n = 25). (B) A comparison of mean community distance between all 0-hour samples on the left and between all 48-hour samples on the right, indicating that 48-hour microbiota exhibit significantly greater mean distance between samples (i.e., greater microbiota compositional variability at this time point). The box-and-whisker plot represents median, minimum, and maximum values. The dashed lines represent mean and SD. ****P < 0.0001. PC = principal component.

Figure 3.

(A) Principal coordinates analysis plot (weighted UniFrac distance) of 48-hour samples colored by acute respiratory distress syndrome (ARDS) status. The r² value represents the percent variance explained, and the P value denotes the significance of this percentage. (B) Discriminatory taxa (48-h three-model analysis comparing ARDS-positive [ARDS+] with ARDS-negative [ARDS−] patients) labeled by genus. OTU = operational taxonomic unit; PC = principal component.