The mycobiome in COPD: Descriptive and longitudinal analysis, and participation in research bronchoscopy studies

Abstract

Background: Chronic obstructive pulmonary disease (COPD) is a high-prevalent lung disease with high mortality rates. The COPD pathogenesis is only partly understood, and we do not know why some risk factor exposed subjects develop the disease. With the introduction of culture-independent sequencing techniques, it has been shown that healthy lung is not sterile, and associations between colonising bacteria, or the lung microbiota, and diseases have been proposed. Early lung microbiome studies, however, focused exclusively on the bacteria, but lungs also contain viruses, fungi, and other eukaryotes. Few studies have examined the fungal community, the mycobiome, in the lungs of COPD patients. Protected bronchoscopic sampling seems to be a well-suited method to collect samples from the COPD lung mycobiome. But the invasive nature of the bronchoscopy procedure might make the recruitment of participants challenging. To facilitate large-scale bronchoscopy studies on the lung mycobiome, we need information on participation in studies involving a bronchoscopy.

Aims: The aim of the thesis was to examine participation in research bronchoscopy studies, and to analyse the lung mycobiome of participants in the Bergen COPD Microbiome study (short name «MicroCOPD»). In paper I, we sought to look for information on motivation to participate, participation barriers, response rates, and recruitment strategies in research studies involving a bronchoscopy by reviewing the literature. In paper II, we wanted to examine participation in the MicroCOPD study by reporting response rates and participation motives and declining reasons. In paper III, our aim was to examine the lung mycobiome in subjects with or without COPD, and to look for effects on the mycobiomes from inhaled corticosteroids (ICS) use. Finally, in paper IV, we aimed to assess the stability of the lung mycobiome of the participants in the MicroCOPD study, and to determine if intercurrent antibiotic use influences the stability.

Material and methods: We performed a literature review on participation in research bronchoscopy studies using the search engines of PubMed and EMBASE. Titles and abstracts of retrieved papers were sifted according to prespecified criteria, and included papers were subject for a more in-depth review.

Individuals with and without COPD or asthma were invited to participate in the MicroCOPD study. Participants underwent at least one bronchoscopy with sample collection by bronchoalveolar lavage (BAL), and a structured interview regarding medical history, medication use, and smoking habits. Additionally, post -bronchodilator spirometry was performed. Subjects that declined participation at the screening interview were asked about their reason not to participate, while subjects accepting participation were asked about their motivation to participate just prior to the bronchoscopy using an open question. All participants also provided an oral wash (OW) sample, and we collected some of the phosphate-buffered saline used in samples to serve as a negative control sample (NCS).

Fungal DNA was extracted from the OW, BAL, and NCS samples, before sequencing of the ITS1 region on an Illumina HiSeq sequencing platform. We compared the taxonomic composition and alpha and beta diversity between participant groups and by ICS use (paper III), and between the first and the second bronchoscopy in paper IV.

Results: Only seven papers were included for in-depth reading in the literature review on research bronchoscopy studies in the literature (paper I), reflecting a paucity of information on this topic. Still, data suggested that most participated for personal benefit and altruistic reasons, while fear and inconvenience hindered participation. Response rates varied from 3 to 73%, and radio advertisement was the most effective recruitment strategy.

The response rate in the MicroCOPD study was just above 50%, and men had a significantly higher response rate than women. Procedural fear was reported as the most common non-response reason. The most frequently stated participation motive was personal health benefit, but after merging participation motives into broader categories, altruism was the most frequent main motive. Men were less likely to state altruism as their main motive.

In paper III, we found that most samples, both OW and BAL, were dominated by Candida. Malassezia and Sarocladium were also frequently found taxa in BAL samples. There was more Candida in OW samples compared to BAL samples, and analyses suggested that beta diversity differed significantly between OW and BAL samples. Neither taxonomy nor alpha or beta diversity analyses found consistent differences between participant groups. The mycobiomes did not seem to be affected by use of ICS.

Taxonomy differed more between the repeated bronchoscopies for BAL samples than OW samples, while no apparent effect was seen from participant category and intercurrent antibiotic use (paper IV). Alpha and beta diversities were consistent by time.

Conclusions: We have shown that most participants in research bronchoscopy studies participate for personal benefit or altruism, and that most decliners fear the bronchoscopy (paper II), in line with what is known from the literature (paper I). A response rate just above 50% in the MicroCOPD study means that large-scale bronchoscopy studies are feasible.

Oral and pulmonary samples differed in taxonomic composition and diversity, possibly indicating the existence of a pulmonary mycobiome (paper III). We have also shown that the lung mycobiome is less stable than the oral mycobiome, and neither COPD diagnosis, intercurrent antibiotic use, nor time between bronchoscopies seemed to influence the stability (paper IV).