The Adenoids but Not the Palatine Tonsils Serve as a Reservoir for Bacteria Associated with Secretory Otitis Media in Small Children

Abstract

Acute otitis media (AOM), secretory otitis media (SOM), and acute pharyngotonsillitis are the most frequent reasons for visits to general practitioners, pediatricians, and otolaryngologists. Microbial colonization of the epithelial lining of Waldeyer's lymphatic tissues, consisting of the palatine tonsils, lingual tonsils, adenoids, and Eustachian tube tonsil, is a well-known clinical challenge during infancy due to frequent episodes of upper respiratory tract infections. However, no previous studies have investigated the combined role of the palatine tonsils and the adenoids as a reservoir for pathogens associated with SOM in small children. We analyzed the combined crypt microbiome of the palatine tonsils and adenoids from 14 small children with hyperplasia of the tonsils or adenoids and 14 small children with SOM using 16S rRNA gene pyrosequencing. Our study demonstrated a significant difference between the microbiome of the adenoids and that of the palatine tonsils in the two groups but not between the two anatomical locations within the two groups. In particular, the potential pathogens Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis were almost exclusively found in the adenoids of both patient groups, indicating that the adenoids and not the palatine tonsils are the main reservoir for potential pathogens leading to AOM and SOM. IMPORTANCE Our findings that the microbiome differs between crypts of the adenoids and crypts of the palatine tonsils, including the relative abundances of potential pathogens such as Haemophilus influenzae, Streptococcus pneumoniae, and Moraxella catarrhalis, may be the stepping stone for further investigation of individual microbiomes in a longitudinal design that includes recording of the fluctuating health status of the child. Such studies may have the potential to lead to new preventive measurements such as implantation of protective nonpathogens at the nasopharynx as an alternative to adenoidectomy.

Keywords: 16S rRNA genes; adenoids; adenotonsillectomy; microbiome; otitis media with effusion; tonsils.

FIG 1

Schematic presentation of the sample collection.

FIG 2

Venn diagram showing overlap of observed OTUs for the four groups clustered at 98.5% similarity. A total of 187 OTUs were found in at least one sample from each of the four groups. A total of 1,036 OTUs were found across all 56 samples.

FIG 3

Unweighted pair group method using average linkages (UPGMA) tree showing the phylogenetic relationship of the 200 most abundant OTUs across all samples. The tree file was generated using representative sequences from all 200 OTUs in MEGA 7 and visualized using iToL software. Each phylum is color highlighted, and the average percent abundancies of each OTU for the four groups are shown.

FIG 4

Box plot of the 30 most abundant OTUs. Whiskers show the range of the relative abundancies of each OTU in the four groups. Median and average values are shown as vertical lines and a cross, respectively.

FIG 5

Bar plot (A) and box plot (B) of the relative abundances of the phyla found and the 20 most abundant genera in each of the four groups. Median and average values are shown as vertical lines and crosses, respectively, in panel B.

FIG 6

PCoA plots comparing the bacterial communities from the four groups using weighted UniFrac distances (A) and the OTU structure (thetaYC calculator) using 98.5% sequence similarity for clustering (B). The percentage of variation explained by each principal coordinate (PC) is indicated on the axes. Each point represents a microbial community. Weighted UniFrac distances from the phylogenetic tree generated from the Clearcut program implemented in the Mothur software. Confidence ellipses (95%) are shown for each group.

FIG 7

Differentially abundant bacterial taxa (phyla, families, genera, and OTUs) identified by linear discriminant analysis (LDA) coupled with effect size measurements (LEfSe) in comparisons of the HP group and the SOM group from the tonsils (A) and the adenoids (B) and from the two body sites (C). Only taxa that met the significant linear discriminant analysis threshold value of 3.5 are shown.

FIG 8

Relative abundances of the three species M. catarrhalis, H. influenzae, and S. pneumoniae in each of the 56 samples grouped according to the group definition.