Metagenomic features of bioburden serve as outcome indicators in combat extremity wounds

Abstract

Battlefield injury management requires specialized care, and wound infection is a frequent complication. Challenges related to characterizing relevant pathogens further complicates treatment. Applying metagenomics to wounds offers a comprehensive path toward assessing microbial genomic fingerprints and could indicate prognostic variables for future decision support tools. Wound specimens from combat-injured U.S. service members, obtained during surgical debridements before delayed wound closure, were subjected to whole metagenome analysis and targeted enrichment of antimicrobial resistance genes. Results did not indicate a singular, common microbial metagenomic profile for wound failure, instead reflecting a complex microenvironment with varying bioburden diversity across outcomes. Genus-level Pseudomonas detection was associated with wound failure at all surgeries. A logistic regression model was fit to the presence and absence of antimicrobial resistance classes to assess associations with nosocomial pathogens. A. baumannii detection was associated with detection of genomic signatures for resistance to trimethoprim, aminoglycosides, bacitracin, and polymyxin. Machine learning classifiers were applied to identify wound and microbial variables associated with outcome. Feature importance rankings averaged across models indicated the variables with the largest effects on predicting wound outcome, including an increase in P. putida sequence reads. These results describe the microbial genomic determinants in combat wound bioburden and demonstrate metagenomic investigation as a comprehensive tool for providing information toward aiding treatment of combat-related injuries.

Figure 1

Microbial profile as determined by whole metagenome sequencing and read-based taxonomic analysis. Shotgun metagenomic sequencing and taxonomic classification of sequence was performed to assess the microbial wound bioburden for (a) initial and (b) final (day of delayed wound closure) wound effluent specimens. Read abundance relative to total sequence content, including human-derived background sequence data, is binned and shown as follows: Only microbial genera for which at least one analyzed specimen demonstrated a relative abundance for that genus of > 1E−5 are shown. Abundance values were binned on a log scale. dark blue: < 1E−7, light blue: ≥ 1E−7 & < 1E−6, yellow: ≥ 1E−6 & < 1E−5, red: ≥ 1E−5. Wound outcome is annotated along the top of the heatmap, with green indicating successful healing; red indicating failed; white indicating outcome not available.

Figure 2

Alpha diversity of bioburden according to wound status. Metagenomic sequence data from wound specimens was used to calculate bioburden diversity through assessment of richness. (a) Microbial taxonomic diversity as represented by Hill’s N₂ (reciprocal Simpson index) at the genus level, per sample, in distinct samplings from wounds. Diversity trends are shown in comparison to wound outcome. (b) Effective number of genera (Hill’s N₂) shown according to increasing number of days post injury. Specimen timepoint category is indicated via color of the corresponding point. Shaded area surrounding loess trendlines indicates 95% confidence interval about the average N₂ genera.

Figure 3

Prevalence of microbial genera in samples derived from healed or failed wounds. Prevalence of detection of microbial genera at defined thresholds was calculated for wounds that either healed successfully or failed to heal. Prevalence is shown as a proportion of total healed or failed samples for each given sample category, where effluent samples are shown in panels (a-c), and tissue samples are shown in panels (d–f). Samples are shown for initial (a,d), intermediate (b,e), and final (c,f) collection timepoints. Individually significant (P < 0.05) distinctions are shown with a bolded edge. Depending on time-to-closure for each wound, a variable number of samples were available for intermediate samplings. A total of 100 effluent samples (across 46 wounds) and 94 tissue samples (across 49 wounds) were examined.

Figure 4

Association of AMR signature classes with detection of nosocomial pathogens. Logistic regression models were built to determine whether detection of defined genomic resistance signatures associates with metagenomic read abundance of specific genera corresponding to nosocomial pathogens. Resistance to drug classes are rows and nosocomial pathogens are columns. The sample collection type is included as an independent variable in each model with the type = tissue as compared to type = effluent shown as a separate column for reference. Significance (false discovery rate adjusted P-value) is indicated through color fill on a − log10 scale.

Figure 5

Association between antimicrobial treatment regimen and detection of antimicrobial resistance gene signatures in wound effluent samples. Mutual information analysis was applied to compare detection of genomic signatures for antimicrobial resistance with drug classes administered. Detection of resistance gene signatures is shown in samples obtained from both (a) healed and (b) failed wounds, to examine distinctions in associations across these patient subgroups.