Blood cultures and blood microbiota analysis as surrogates for bronchoalveolar lavage fluid analysis in dogs with bacterial pneumonia

Abstract

Background: Diagnosis of canine bacterial pneumonia relies on airway lavage to confirm septic, suppurative inflammation, and a positive bacterial culture. Considering risks of bronchoalveolar lavage fluid (BALF) collection, minimally invasive methods like culture or next generation sequencing of blood would be appealing. In dogs with bacterial pneumonia, our study aims included (1): determining proportion of agreement between cultivable bacteria in BALF and blood (2); characterizing BALF, blood, and oropharyngeal (OP) microbiota and determining if bacteria cultured from BALF were present in these communities; and (3) comparing relatedness of microbial community composition at all three sites. Bacterial cultures were performed on BALF and blood. After DNA extraction of BALF, blood and OP, 16S rRNA amplicon libraries were generated, sequenced, and compared to a bacterial gene sequence database.

Results: Disregarding one false positive, blood cultures were positive in 2/9 dogs (5 total isolates), all 5 isolates were present in BALF cultures (16 total isolates). Based on sequencing data, all sites had rich and diverse microbial communities. Comparing cultured BALF bacterial genera with sequenced taxa, all dogs had ≥1 cultured isolate present in their microbiota: cultured BALF isolates were found in microbiota of BALF (12/16), blood (7/16), and OP (6/11; only 7 dogs had OP swabs). Of 394 distinct taxa detected in BALF, these were present in 75% OP and 45% blood samples. BALF community composition was significantly different than OP (p = 0.0059) and blood (p = 0.0009).

Conclusions: Blood cultures are insensitive but specific for cultured BALF bacteria in canine bacterial pneumonia. Cultivable BALF bacteria were present in BALF, blood and OP microbiota to differing degrees.

Fig. 1

Box and whisker plot displaying richness (total number of unique taxonomic units in a sample) observed at each site. The black bars in the box structures represent the median value, the bottom line of the box to the black line represents the first quartile and the black line to the outer box line represents the third quartile. The line extended vertically from the box (both on the bottom and the top) to the solid horizontal line, represents the extent of the area where outliers are found. Richness was significantly higher in oropharyngeal swabs (OP) compared to bronchoalveolar lavage fluid (BALF) (p = 0.018) and blood (p < 0.001)

Fig. 2

Bar graph displaying the percentage relative abundance of different bacterial taxa at the level of the operational taxonomic unit (OTU) in bronchoalveolar lavage fluid, oropharyngeal swabs and blood. Each color represents a different OTU and with all OTUs from a single sample (BALF or blood) adding up to 100%. Numbers at the bottom of the graph represent a different dog

Fig. 3

Venn diagram showing the distribution of operational taxonomic units (OTUs) detected in at least one sample of bronchoalveolar lavage fluid (BALF), oropharyngeal swab (OP) or blood of dogs with bacterial pneumonia

Fig. 4

β-diversity as shown via principal component analysis. Unweighted principal component analysis of samples from all sample sites (BALF, OP and blood) PC1 versus PC2; legends at right. The ellipses represent 95% intervals