The Upper-Airway Microbiome as a Biomarker of Asthma Exacerbations despite Inhaled Corticosteroid Treatment
-
Perez-Garcia, Javier
1
-
González-Carracedo, Mario
2
-
Espuela-Ortiz, Antonio
1
-
Hernández-Pérez, José M.
3
-
González-Pérez, Ruperto
4
-
Sardón-Prado, Olaia
5
-
Martin-Gonzalez, Elena
1
-
Mederos-Luis, Elena
6
-
Poza-Guedes, Paloma
4
-
Corcuera-Elosegui, Paula
5
-
Callero, Ariel
7
-
Sánchez-Machín, Inmaculada
6
-
Korta-Murua, Javier
5
-
Pérez-Pérez, José A.
2
-
Villar, Jesús
8
-
Pino-Yanes, Maria
1
-
Lorenzo-Diaz, Fabian
1
- 1 Genomics and Health Group, Department of Biochemistry, Microbiology, Cell Biology and Genetics, Universidad de La Laguna (ULL), La Laguna, Tenerife, Spain
- 2 Department of Biochemistry, Microbiology, Cell Biology and Genetics, Universidad de La Laguna (ULL), La Laguna, Tenerife, Spain
- 3 Pulmonary Medicine Service, Hospital Universitario de N.S de Candelaria, La Laguna, Tenerife, Spain
- 4 Severe Asthma Unit, Allergy Department, Hospital Universitario de Canarias, La Laguna, Tenerife, Spain
- 5 Division of Pediatric Respiratory Medicine, Hospital Universitario Donostia, San Sebastián, Spain
- 6 Allergy Department, Hospital Universitario de Canarias, La Laguna, Tenerife, Spain
- 7 Allergy Service, Hospital Universitario N.S. de Candelaria, La Laguna, Tenerife, Spain
- 8 CIBER de Enfermedades Respiratorias, Instituto de Salud Carlos III, Madrid, Spain
Editor: Zenodo
Año de publicación: 2023
Tipo: Dataset
Resumen
The response to inhaled corticosteroids (ICS) in asthma is affected by the interplay of several factors. Among these, the role of the upper-airway microbiome has been scarcely investigated. We aimed to evaluate the association between the salivary, pharyngeal, and nasal microbiome with asthma exacerbations despite ICS use. Samples from 250 asthma patients from the Genomics and Metagenomics of Asthma Severity (GEMAS) study treated with ICS were analyzed. Controls/cases were defined by the absence/presence of asthma exacerbations in the past six months despite being treated with ICS. The bacterial microbiota was profiled by sequencing the V3-V4 region of the 16S rRNA gene. Differences between groups were assessed by PERMANOVA and regression models adjusted for potential confounders. A false discovery rate (FDR) of 5% was used to correct for multiple comparisons. Classification models of asthma exacerbations despite ICS treatment were built with machine learning approaches based on clinical, genetic, and microbiome data. In nasal and saliva samples, cases had lower bacterial diversity (Richness, Shannon, and Faith indexes) than controls (0.007≤p≤0.037). Asthma exacerbations accounted for 8-9% of the interindividual variation of the salivary and nasal microbiomes (0.003≤p≤0.046). Three, four, and eleven bacterial genera from the salivary, pharyngeal, and nasal microbiomes were differentially abundant between groups (4.09x10<sup>-12</sup>≤FDR≤0.047). Integrating clinical, genetic, and microbiome data showed good discrimination for the development of asthma exacerbations despite ICS use (AUC<sub>training</sub>:0.82 and AUC<sub>validation</sub>:0.77). The diversity and composition of the upper-airway microbiome are associated with asthma exacerbations despite ICS treatment. The salivary microbiome has a potential application as a biomarker of asthma exacerbations despite ICS use. The summary results of the association between bacterial genera and asthma exacerbations despite ICS use are published in this repository. Analyses were conducted using the DESeq2 package adjusting for potential confounders (see more details in the manuscript). The submitted summary tables contains the following columns: Kingdom: Taxonomic classification at Kingdom level Phylum: Taxonomic classification at Phylum level Class: Taxonomic classification at Class level Order: Taxonomic classification at Order level Family: Taxonomic classification at Family level Genus: Taxonomic classification at Genus level log2FC: log<sub>2</sub>(fold-change). Positive values indicate higher abundance in cases and negative values in controls SE: standard error of log<sub>2</sub>(fold-change) P.Value: p-value of the regression model FDR: adjusted p-value by false discovery rate (Benjamin-Hochberg method)