Skip to main content
Download PDF
  • Research article
  • Open access
  • Published:

Azithromycin and the microbiota of cystic fibrosis sputum

BMC Microbiology volume 21, Article number: 96 (2021) Cite this article

Abstract

Background

Azithromycin is commonly prescribed drug for individuals with cystic fibrosis (CF), with demonstrated benefits in reducing lung function decline, exacerbation occurrence and improving nutrition. As azithromycin has antimicrobial activity against components of the uncultured microbiome and increasingly the CF microbiome is implicated in disease pathogenesis – we postulated azithromycin may act through its manipulation. Herein we sought to determine if the CF microbiome changed following azithromycin use and if clinical benefit observed during azithromycin use associated with baseline community structure.

Results

Drawing from a prospectively collected biobank we identified patients with sputum samples prior to, during and after initiating azithromycin and determined the composition of the CF microbial community by sequencing the V3-V4 region of the 16S rRNA gene. We categorized patients as responders if their rate of lung function decline improved after azithromycin initiation. Thirty-eight adults comprised our cohort, nine who had not utilized azithromycin in at least 3 years, and 29 who were completely naïve. We did not observe a major impact in the microbial community structure of CF sputum in the 2 years following azithromycin usage in either alpha or beta-diversity metrics. Seventeen patients (45%) were classified as Responders – demonstrating reduced lung function decline after azithromycin. Responders who were naïve to azithromycin had a modest clustering effect distinguishing them from those who were non-Responders, and had communities enriched with several organisms including Stenotrophomonas, but not Pseudomonas.

Conclusions

Azithromycin treatment did not associate with subsequent large changes in the CF microbiome structure. However, we found that baseline community structure associated with subsequent azithromycin response in CF adults.

Peer Review reports

Background

Within cystic fibrosis (CF), lung disease is responsible for the majority of morbidity and mortality associated with the genetic disease [1]. A worsened disease status has been associated with airways infection with several organisms such as Pseudomonas aeruginosa, Burkholderia cepacia complex (BBC), Stenotrophomonas maltophilia, Achromobacter spp., and Mycobacteriodes abscessus [2, 3] - cultivated through traditional techniques. However, a large body of evidence using culture-independent approaches has now established that CF airways represent polymicrobial infections, with the residing community of organisms termed as the CF microbiome. Data from our group has previously demonstrated that the CF sputum microbiome is a better predictor of disease course than traditional cultured pathogens [4], and baseline community structure associates with response to CF disease modifying therapies [5,6,7].

One of the most frequently prescribed drugs in CF is azithromycin, a 15-membered ring azalide macrolide antibiotic that inhibits bacterial protein synthesis by binding to the 50S ribosomal subunit. Placebo controlled randomized studies in CF have shown that azithromycin treatment is associated with an improvement in lung function, reduced pulmonary exacerbations (PEx), and improvement in nutritional status [8,9,10,11]. However, these benefits are more pronounced in CF patients chronically infected with P. aeruginosa– a pathogen with high-level intrinsic azithromycin resistance [12]. Accordingly, alternate explanations for this beneficial effect have focused on azithromycin’s immunomodulatory or anti-viral effects. Azithromycin has demonstrated suppression of both neutrophil oxidative metabolism and serum inflammatory markers [13, 14]. In addition, Meyer et al. [15] showed that azithromycin down-regulates inflammatory cytokine production in CF alveolar macrophages and Cigana et al. [16] found that it reduced the expression of the pro-inflammatory chemokines in CF airway epithelial cells. More recently, alternate hypotheses suggesting intrinsic anti-viral effects of azithromycin against common respiratory viruses implicated in exacerbations have been proposed [17,18,19]– rationalizing the reduced frequency of exacerbations observed in randomized placebo-controlled trials amongst individuals with CF – regardless of underlying chronic airways pathogens [8, 20].

Azithromycin’s antibacterial properties as a potential mechanism for its benefit have again come under scientific study – but rather than its effects on classical CF pathogens such as P. aeruginosa, new lines of enquiry are focusing on non-classical components within the airway’s microbiome. Indeed, recent studies have postulated that macrolides may exert at least a portion of their beneficial effects in non-CF bronchiectasis due to anti-microbiome effects [21, 22]. We postulated that a portion of the clinical benefit derived from CF individuals taking azithromycin may relate to its effects on the microbiota of CF sputum - and that microbiota composition may influence therapeutic response. In this study, we sought to determine if azithromycin associates with changes in the microbial community of sputum of a cohort of adults with CF and if there is an association between microbiome structure and subsequent patient outcomes – that might serve as novel biomarker enabling personalization of treatment for future exploration.

Results

Patients characteristics as clinical outcome

Thirty-eight patients from our single centre meeting inclusion and exclusion criteria were included in the study (14 males; 24 females), providing 58 pre-treatment azithromycin samples (median − 154 days (interquartile range (IQR) -368- 0) from start day on azithromycin) and 38 post-azithromycin samples (median + 372 days (IQR + 288 − + 420)). Twenty-nine patients with available sputum samples were naïve to azithromycin treatment and nine patients had received remote azithromycin, with a median duration of 5 years (IQR 4–6 years) off drug. In our cohort, we identified 20 patients (52.6%) with all three time points; Pre, Day 0 and Post samples, 17 patients (44.7%) with just Pre and Post sample; and one patient (2.6%) with only Day 0 and Post samples. Patients included in the study had a median age of 25.4 years (IQR 21.2–31.8) at the day of start of azithromycin treatment. Patients in the cohort were started on azithromycin because of concern for clinical deterioration (21/38; 55%), advancing to standard care (29/38; 76%), or both (10/38; 26%). Median annual rate of lung function decline in the entire cohort did not differ following azithromycin initiation: − 1.69%/year (IQR -4.12- + 1.76) pre-azithromycin and − 2.82%/year (IQR -5.5- + 2.21) post-azithromycin (p = 0.822).

Of the cohort, 45% [17] were classified as responders. Comparison of patient’s demographic and clinical characteristics at baseline (i.e. Day 0 of azithromycin treatment when available or the pre-treatment sample) between Responders vs non-Responders were similar with respect to age, gender, lung function (both percent predicted forced expiratory volume in 1 s (FEV1%) and percent predicted forced vital capacity (FVC)), CF-related co-morbidities and cultured pathogens (p > 0.05), and concurrent therapies - with the exception of the use of inhaled corticosteroids (Table 1). Categorization as Responders did not differ based on treatment cohort period from which the samples were collected (A: 54.5%; B: 50%; C: 35.2%, p = 0.267). Twenty-eight (74%) of the cohort were chronically infected with P. aeruginosa as per the Leeds definition at the time of azithromycin initiation. Rate of FEV1 decline did not differ before or after azithromycin initiation in those with chronic P. aeruginosa infection relative to those without [Before; − 2.01 (IQR: − 4.35 – 1.59) vs − 0.62 (IQR: − 4.97 – 2.6), p = 0.550 or After; − 2.82 (IQR: − 4.89 – 3.36) vs − 2.25 (IQR: − 11.6 – 1.43), p = 0.317]. Treatment rationale for azithromycin initiation did not differ in those chronically infected with P. aeruginosa versus those not [standard of care; 23/28 (82%) vs 6/10 (60%), relative risk (RR) 1.37 (95% CI 0.8–2.34), p = 0.21); concern for clinical decline; 13/28 (46%) vs 8/10 (80%), RR 0.58 (95% CI 0.35–0.96), p = 0.14].

Table 1 Patient’s demographics and clinical characteristics at baselinea as a function of azithromycin treatment response
Full size table

Overall CF microbiome

A total of 4,705,369 reads (average, 27,659.1 reads/sample; IQR, 19,278 – 72,263.7) with 344 total ASVs (Amplicon Sequence Variant) were recognized. To determine if the diversity observed in our cohort represented the overall diversity, species accumulation curve analysis was performed. We found the number of ASVs increased between 5 and 30 samples and began to plateau by the end of our sampling (See Supplementary Figure 1, Additional file 1). Of the total ASVs found, 20 ASVs (5.81%) accounted for 98.9% of the total number of the reads (Fig. 1). All these ASVs correspond to taxa commonly found in the CF airways communities: Pseudomonas (51.2%), followed by Streptococcus (12.2%), Haemophilus (10.2%), Staphylococcus (8.9%), Stenotrophomonas (3.7%), Fusobacterium (3.0%), Gemella (2.4%), Rothia (1.8%), Neisseria (1.6%), Prevotella 7 (1.4%), Parvimonas (0.5%), Veillonella (0.5%), Porphyromonas (0.3%), Prevotella 6 (0.3%), Leptotrichia (0.2%), Prevotella (0.2%), Alloprevotella (0.2%), Atopobium (0.2%), Actinobacillus (0.1%) and Eikenella (0.1%).

Fig. 1
figure 1

Taxonomic abundance comparison between responders and non-responders to the azithromycin treatment. Relative abundance at the genus level for samples collected at Pre (≤24 months pre-initiation treatment), Day 0 (start day on azithromycin) and Post (≤24 months post its initiation treatment) azithromycin initiation. The top 20 ASVs accounting for > 0.1% of total relative abundance for the whole data set are coloured and presented in the figure. R: responder and Non-responders (NR) to the azithromycin treatment, ASV: amplicon sequence variants. (*) Asterisk in the Patient ID represent patient who were not naïve to azithromycin treatment

Full size image

Baseline microbiome and primary clinical outcome

To determine possible biomarkers for response to azithromycin treatment, we compared the alpha diversity from Responders and Non-Responder samples from pre-azithromycin treatment. There were no significant differences in alpha diversity (within-sample diversity) as measured by the Shannon Diversity index (SDI) in Responders relative to Non-Responders for the entire cohort, or exclusively the naïve cohort when measured at either all pre-treatment samples or merely Day 0 samples (Fig. 2a and b). To account for the effect of having multiple samples (i.e. Pre and Day 0) per patient in the SDI between Responder and Non-Responder, we performed linear mixed effects models and found no significant differences in SDI (All cohort: p = 0.285, Naïve cohort: p = 0.165; Analysis of variance (ANOVA)). Non-metric multidimensional scaling (NMDS) plots based on Bray-Curtis dissimilarity was used to visualise potential clustering among sputum samples based on response-status to azithromycin treatment. Using our relative definition of response, sputum samples clustered by status. However, this was only statistically significant when the naïve cohort was assessed using both Pre and Day 0 samples (R2 = 7.22, p = 0.008, (Permutational multivariate analysis of variance) PERMANOVA) (Fig. 2c) or the baseline samples (R2 = 7.55, p = 0.048, PERMANOVA) (Fig. 2d). Because there were differences in the microbial community structure between Responders and Non-Responders before azithromycin treatment in the naïve cohort, we used DESeq2 to detect ASVs that show significantly differential relative abundance. When both Pre and Day 0 samples were analyzed, we found that Stenotrophomonas and Megasphaera ASVs were enriched in the responder samples (Fig. 2e) or Stenotrophomonas and Abiotrophia ASVs when only the baseline samples were analyzed (Fig. 2f) but not with Pseudomonas as was expected. As inhaled corticosteroids use differed in our cohorts – we assessed if communities clustered by use at baseline – where no difference was observed when the entire cohort (R2 = 1.72, p = 0.801, PERMANOVA) or the naïve cohort (R2 = 2.05, p = 0.858, PERMANOVA) was analyzed.

Fig. 2
figure 2

CF microbial community comparison between Responders and Non-Responder to azithromycin treatment. Alpha diversity (within patients) of Responders and Non-Responders based on Shannon diversity index (SDI) metrics for the whole cohort (n = 38 patients) and the naïve cohort (n = 29) when only the Pre (i.e. Pre and Day 0) (a) or baseline samples (b) were analyzed. Wilcoxon rank sum test was performed. NMDS plot showing beta diversity of R and Non-R based on Bray-Curtis dissimilarities when only the Pre (i.e. Pre and Day 0) (c) or baseline samples were analysed (d). ASVs that were identified by DESeq2 to be significantly different (p adjusted < 0.05) between Responders and Non-Responders, when Pre (i.e. Pre and Day 0) (e) or baseline samples were analysed (f), relative abundance is presented in the Log2 scale. Boxplots show the median with IQR and the ends of the whiskers mark the 10th and the 90th percentiles. R: responder and Non-responders (NR) to the azithromycin treatment, ASV: amplicon sequence variants

Full size image

Impact of azithromycin on the CF microbiome

We sought to understand if azithromycin induced large-scale changes in community structure by comparing communities from sputum before and after its use. Based on the Bray-Curtis dissimilarities of sputum samples collected Pre and Post-azithromycin treatment, no differences were observed with either the entire cohort (R2 = 0.98, p = 0.463, PERMANOVA) or when the naïve cohort was tested (R2 = 0.99, p = 0.667, PERMANOVA). Similar results were observed when data was stratified for patient ID (data not shown). Furthermore, no difference in SDI was noted between samples collected before and after azithromycin treatment. This was also true when either both Pre and Day 0 (1.25 vs 0.94, p = 0.264) (Fig. 3a) or only Day 0 (1.25 vs 0.94, p = 0.396) (Fig. 3b) were compared against the post samples. Similar results were observed when only patients naïve to azithromycin were analyzed (Fig. 3a and b). To account for multiple sampling for the Pre samples (i.e. Pre and Day 0) for some patients, we conducted linear mixed effects models and found similar results (p = 0.302, ANOVA).

Fig. 3
figure 3

Impact of azithromycin on the CF microbiome. a Alpha diversity (within patients) of Pre (i.e. Pre and Day 0) and Post azithromycin samples based on Shannon diversity index (SDI) metrics for the whole cohort (n = 38 patients) and the naïve cohort (n = 29). Wilcoxon rank sum test was performed. b Alpha diversity of baseline and Post azithromycin samples based on SDI metrics for the whole cohort and the naïve cohort. Wilcoxon signed-rank test was performed. Boxplots show the median with IQR and the ends of the whiskers mark the 10th and the 90th percentiles. c ASVs that were identified to be significantly different (p adjusted < 0.05) in relative abundance between Pre (i.e. Pre and Day 0) and Post groups, as detected by DESeq2

Full size image

DESeq2 was used to identify ASVs that differed after azithromycin was initiated. We found Haemophilus and Stenotrophomonas were enriched in those samples collected after azithromycin relative to before (i.e. Pre and Day 0) when the whole cohort or the naïve cohort was evaluated (p adjusted < 0.05) (Fig. 3c). Whereas clustering by azithromycin use was not observed, clustering by traditionally identified factors were. Sputum samples from the same patient were much more similar to each other than to communities of other patients (R2 = 64.9, p = 0.001, PERMANOVA) (See Supplementary Figure 2, Additional file 1). Additionally, communities clustered based on the stage of lung disease (i.e. mild (≥80%), moderate (40–80%), and advanced lung disease: (≤40%)) at the sample collection when it was analyzed the whole cohort (R2 = 3.45, p = 0.045, PERMANOVA) or only the azithromycin naïve cohort (R2 = 4.99, p = 0.03, PERMANOVA).

CF microbiome following azithromycin initiation treatment

Review of medical records showed that 27 (71.05%) patients experienced ≥1 clinician defined PEx events in 2 years following azithromycin initiation (1 PEx; 16 (59.3%), 2 PEx; 9 (33.3%) and 3 PEx; 2 (7.4%) patients). We sought to determine if community structure following azithromycin treatment associated with PEx occurrence risk. There was no significant difference for the alpha diversity when it was analyzed in the entire cohort or the naïve subset (Fig. 4a). Similarly, we found that microbial community structure had no association with number of PEx events in the Post samples when the entire cohort (R2 = 1.60, p = 0.851, PERMANOVA) or the naïve cohort (R2 = 2.64, p = 0.637, PERMANOVA) was analyzed. However, when we determined the specific ASVs that showed significantly different relative abundance between having PEx events or not during the Post time, we observed that some taxa, normally found in CF airways communities (i.e. Haemophilus, Streptococcus, Prevotella, Gemella, Staphylococcus, Leptotrichia and Rothia), were enriched in patients who exacerbated (Fig. 4b, first panel). Paradoxically, Haemophilus was the only taxa whose relative abundance associated with patients who did not experience PEx (Fig. 4b, first panel). In addition to Fusobacterium and Parvimonas, Rothia also appeared to be enriched in the patients that experienced PEx during the post time when the naïve cohort was evaluated (Fig. 4b, second panel). Finally, we evaluated if there were differences in the microbial communities of sputum after azithromycin treatment initiation between patients that were completely naïve to the treatment and those who had not been exposed to azithromycin for three or more years. No significant difference was observed for alpha diversity (Fig. 4c) or beta diversity (R2 = 2.02, p = 0.64, PERMANOVA) in these sub-groups. However, we found that Prevotella was enriched in the subset of individuals that were naïve to azithromycin compared to those who were previously exposed three or more years prior (Fig. 4d).

Fig. 4
figure 4

CF microbiome in the Post-azithromycin sputum samples. a Alpha diversity (within patients) of samples from patients who experienced one or more than 1 PEx event after azithromycin initiation based on Shannon diversity index (SDI) metrics for the entire cohort (n = 38 patients) and those completely naïve (n = 29). b ASVs that were identified to be significantly different (p adjusted < 0.05) in relative abundance between patients with non or more than 1 PEx event during the post period, as detected by DESeq2. c Alpha diversity of samples from patients who were naïve to azithromycin or who had not been exposed to azithromycin within at least 3 years. d ASV that was identified to be significantly different (p adjusted < 0.05) in relative abundance between naive and Non-naïve patients, as detected by DESeq2. For a and c, Wilcoxon rank sum test was performed, and boxplots show the median with IQR and the ends of the whiskers mark the 10th and the 90th percentiles. ASV: amplicon sequence variants

Full size image

Discussion

Azithromycin is a critical therapy in CF and has long been included in clinical practice guidelines [23,24,25]. In those older than 6 years of age with chronic P. aeruginosa it is advocated as a recommended therapy to improve lung function and reduce PEx (strength of recommendation B; with a high certainty of net benefit - estimated as moderate) whereas those without chronic P. aeruginosa infection azithromycin is recommended to be considered to reduce PEx (strength of recommendation C; with a moderate certainty of net benefit - estimated as mild) [25]. While several investigators now suggest the benefit may be linked to its intrinsic anti-inflammatory or anti-viral properties – a wide variety of studies over the years have sought to identify anti-bacterial links [26, 27]. Indeed, some data suggests that azithromycin is protective against acquisition of new CF pathogens [28]. Owing to the recent recognition of a complex community inhabiting the airways and its association with disease outcomes – others have sought to identify a potential link with azithromycin. Recent work has established that azithromycin induces changes in the sputum microbiome in chronic lung diseases such as asthma [29] and chronic obstructive pulmonary disease (COPD) [30]. Erythromycin, another macrolide antibiotic, was found to induce changes in the microbiota composition in patients with non-CF bronchiectasis [22]. Drawing from this parallel, we hypothesized that azithromycin may induce changes within the CF microbiome - and this may explain some of its associated benefits in CF. However, as azithromycin use in CF (first reported in 1998 [31]) predates the recognition and study of the CF microbiome (first described by Rogers et al. in 2003 [32]), finding treatment naïve patients has proved difficult - limiting the ability of others to perform this important work. Owing to our unique ability to draw from a one-of-a-kind prospectively collected sputum biobank spanning more than 20 years, we were in a unique position to assess for a similar CF-associated effect of azithromycin on sputum microbial community structure of CF adults.

Our primary outcome of interest was to assess if community structure associated with a CF individual’s response to azithromycin treatment (defined by improvement in rate of lung function decline between the start day of treatment and 24 months after it). Identifying signals within the microbiome to serve as biomarkers enabling treatment personalization has been a goal of our group – and one particularly relevant for an agent such as azithromycin where different treatment responses have been observed in clinical trials based on P. aeruginosa chronic infection status. In general, we noted no significant differences between Responders and Non-Responders in terms of baseline demographics and clinical characteristics, except for the use of inhaled corticosteroids. We sought to determine whether any microbiome specific biomarkers before azithromycin treatment could be associated with subsequent response to azithromycin treatment. We did not find baseline community structure biomarker as measured by alpha-diversity correlated with azithromycin response as there were no differences in the Pre samples (i.e. Pre and Day 0) between Responders and Non-Responders. We did, however, observe a response-associated community profile based on analysis on the beta diversity using only the cohort of patients that were completely naïve to azithromycin. In particular, we found that genera such as Stenotrophomonas, Megasphaera and Abiotrophia were relatively enriched in those Responders who demonstrated improvement in rate of lung function decline after starting azithromycin. Increased risk of S. maltophilia infection have been associated in both pediatric and older CF patients with more advanced lung disease [33, 34].

Our secondary outcome of interest was to assess the frequency of PEx events that patients experience following the initiation of the azithromycin. Unlike changes in lung function, reduction of PEx events has been almost universally associated with azithromycin use – irrespective of baseline pathogens [20]. We found that neither the alpha nor beta diversity differed in the subset of patients who experienced PEx. We did observe some CF community members to be enriched in patients who had PEX after azithromycin initiation. Studies focusing on cultivated pathogens have likewise shown few microbial differences. For example, Samson et al. [35] followed a pediatric CF cohort for 12-months of azithromycin treatment and while they observed reduced PEx, they did not find changes in the prevalence of cultured CF pathogens [35].

We did not find an obvious effect of azithromycin on the CF microbiome in either the naïve or the whole cohort of CF adults. This was somewhat of a surprise as previous studies have shown that azithromycin altered the airway microbiota in other chronic lung diseases such as asthma, COPD and non-CF Bronchiectasis (nCFB) [22, 29, 30]. One possible explanation could be related to the differences of our cohort including, baseline demographics and clinical characteristics, such as age and lung disease stage, both of which are known to have an effect on airways microbiota [36]. In particular, our cohort of CF adults had more advanced lung disease and a microbial community which was much less diverse (median SDI 1.21 (IQR 0.5–1.8) – and thus less likely to experience significant perturbations [37, 38]. It is possible a younger cohort of individuals with CF with more diverse microbial communities – more in keeping with those of asthma, COPD, and nCFB [39, 40] – may be more likely to demonstrate a change. However, other studies focusing on classical CF methods have also found that azithromycin treatment had no major impact on microbiological outcomes in a CF pediatric cohort with early P. aeruginosa infection [11] or in sputum bioburden [9]. Although we found that the overall microbial community structure had no association with azithromycin treatment, we found that some specific taxa such as Stenotrophomonas and Haemophilus were enriched in the Post samples when both the whole and naïve cohort were evaluated. Other work, has previously shown that H. influenzae and S. aureus were the only two cultured microorganisms that showed an increase in macrolides resistance in a controlled trial of azithromycin in 6–18 years of age CF patients who were uninfected with P. aeruginosa [14]. It is important to note that those results were based on qualitative data (i.e. airway culture), since most of the patients included were unable to produce sputum. Lastly, it has been previously shown that CF isolates of S. maltophilia are resistant to macrolides – and their recovery following its initiation is thus not unexpected [41, 42].

Several additional limitations of our study warrant consideration. Studies evaluating how CF therapies affect the microbiome can best be done in patients who are naïve to the treatment – a difficult prospect – when studying existing treatments in a population that is already appropriately managed with standard of care treatments like azithromycin. A significant limitation of our study was the modest number of patients in our biobank that met our inclusion and exclusion criteria. This limitation is true to most microbiome studies – where our cohort size is typical. While 76% of our 38 individuals were completely naïve to azithromycin, 9 had prior exposure - a median 5 years earlier. Given the very transient nature of antibacterial effects on the CF microbiome [37, 38, 43, 44] we were comfortable to include those in our study – although we evaluated them both together and separately to account for this. Additionally, although our study had a modest number of patients, when we performed a species accumulation curves analysis to analyze the adequacy of sample size in relation with the bacterial diversity observed, it was found that sequencing depth was appropriate. We suggest that for all future investigational agents that might be perceived to have an effect on the CF microbiome – studies should prospectively collect and store samples during patient follow up enabling the potential for subsequent analysis in a controlled fashion [45]. For example, Segal and collaborators [30] upon exploring azithromycin in COPD were able to assess both the impact on the lung microbiome and metabolome. In order to have a sufficient sample size to analyze amongst a single centre population of adults with CF we utilized samples from three different cohorts of time – introducing cohort heterogeneity given the changes that have been observed over time in CF. We attempted to minimize this by analyzing each cohort together and separately. Furthermore, prior work from our group specifically explored how microbial communities changed over time in different birth cohorts with CF – and observed that the only significant changes in community structure over time related to improving respiratory lung function in successive cohorts [46]. Another limitation of our study is related with the usage of sputum samples as an indicator of the lower airway microbiome. Although studies that use sputum samples have identified the general representation of predominant taxa in the lung microbiota [47], it is important to highlight that some taxa identified in the sputum microbiome could be a result of contamination of oral cavity microbiota.

While the most significant determinant of community clustering in the analysis of serial sputum samples from a cohort is the specific-individual from which a sample is derived [48], increasing data suggests that modest day-to-day changes do exist within an individuals’ expectorated sputum [48, 49]. Accordingly, future efforts to understand the association between specific therapeutics and community structure would benefit from serial sampling – something that was not possible in the retrospective analysis of a clinically collected biobank such as ours despite our efforts to assess pre samples and Day 0 samples separately and together. While we attempted to control for confounders by excluding the most egregious – i.e. parenteral antibiotics [43, 50]– we were not able to control for changes in therapy intrinsic to CF - such as cycled inhaled antibiotics - that can have modest impacts on the CF microbiome in this opportunistically collected biobank of clinical samples [5,6,7, 51]. Furthermore, increasingly robust data suggests that some of these – notably inhaled tobramycin – may be negatively impacted through azithromycin use (via upregulation of the mexXY-efflux pump manifesting in abrogation of tobramycin clinical benefit) – further confounding a treatment-response association that could not be adjusted for in our small study [52, 53]. Inhaled corticosteroid use was found to be more common in our responder population – something that does warrant further exploration given its potential role in respiratory immunomodulation. Despite common usage, the role of inhaled corticosteroids in CF remains uncertain given that there is no clear evidence they reduce inflammation in people with CF or provide clinical benefit [54,55,56]. Notably, communities did not differ based on their use at baseline. Further studies are required to determine the role of inhaled corticosteroids in individuals with CF and how it may impact response to azithromycin. Indeed, animal models of asthma suggest changes in community structure have been associated with their use [57], whereas no gross changes were observed in humans between high and low dose inhaled corticosteroids [58]. Finally, future studies should not only be focused on how chronic therapies influences the microbiome but also how it influences induction of antimicrobial resistance within different components of the community [59].

Conclusions

Herein, we found that azithromycin does not appear to induce significant changes in the sputum microbiome of CF adults that were naïve or were not exposed to azithromycin within at least a 3-year period. While samples clustered by patient response status to azithromycin, this effect was small. Our data suggests that the modification of microbial community structure is not a significant mechanism by which azithromycin exerts its clinical benefits in adults with CF.

Methods

Patients and sample collection

Patients and samples were retrospectively drawn from a single CF center (the Calgary Adult CF Clinic and its associated Biobank). Patients have prospectively contributed samples to this regional ethics board approved collection from 1998-present (REB15–0854) and provide consent for ongoing research purposes (REB15–2744). This longitudinal collection contains more than 18,000 serial CF sputum samples and has been used for many projects exploring the relationship between the CF microbiome and patient outcomes. Patients were included if they were completely naïve/ or had been off azithromycin for at least 3 years and had at least one sputum before collected and after the initiation of azithromycin. Samples were classified as: before-azithromycin (Pre) (from – 24 months to the day of its initiation), and post-azithromycin (Post) (up to 24 months following initiation treatment). Samples collected at Day 0 (azithromycin start day) were included when available. Sputum samples were excluded from the study if they were collected within 14 days of a PEx and/or the receipt of new systemic antimicrobial agents owing to the potential for confounding effects on the microbiome [43, 50]. Because the study included samples spanning over 15 years – during which many changes in CF care have occurred, samples were categorized in three cohorts based on the year of sample collection: cohort A (2000–2005), cohort B (2006–2010) and cohort C (2011–2015) – and separately analyzed as well. Patient demographics (i.e. sex, age, genotype and nutritional status as measured by body mass index), dynamic variables of disease (i.e. percent predicted forced expiratory volume in 1 s (FEV1), percent predicted forced vital capacity (FVC) and CF related diseases), medical therapies and cultured pathogens – at the time of azithromycin initiation - were recorded. FEV1 values were normalized using the National Health and Nutrition Examination Survey III references and the FEV1% predicted at sample collection was used to categorize patients in the following stages of lung disease: mild (≥80%), moderate (40–80%), and advanced (≤40%).

Clinical outcome definitions

Treatment initiation rationale was reviewed and rationale for initiation classified as; addition of therapy consistent with standard care, acceleration of treatment owing to clinical deterioration or both. Clinical outcomes in the 2 years before and after azithromycin initiation were collected from a detailed review of each clinical encounter in every patient’s medical record. Dynamic changes in FEV1 and PEx occurrence were specifically extracted for each time period. The rate of lung function decline in each period was determined through the construction of subject-specific linear regressions of FEV1values collected in those 2 years of study. We included a median of 9 FEV1 measurements (IQR 7.2–10.7) per patient, for the 24 months before, and a median of 9 FEV1 measurements (IQR 6–10) per patient for the 24 months after in each construct. The primary outcome sought was to differentiate patients based on azithromycin-associated improvements in lung function. Patients were classified as Responders if their net rate of FEV1 decline improved (difference between 24 months after treatment and 24 months before treatment is greater than 0) and classified as Non-Responders if no improvement was observed (less than 0). For all the analysis performed in this study, patients were examined in aggregate and a separate sub-group analysis was performed on those patients who were completely naïve to azithromycin. The secondary outcome of interest was the occurrence of PEx. A PEx event was defined clinically as periods of acute worsening with the resultant administration of systemic antimicrobial agents [60]. Clinical charts were retrospectively reviewed, and we verified PEx frequency in the 24 months before and after azithromycin initiation. Patients with chronic P. aeruginosa were defined as per the modified Leed’s criteria [61].

DNA isolation, sequencing and sequence analysis

Total genomic DNA was extracted as previously described [46]. Reagent controls were sequenced and analyzed as quality controls. Barcoded primers were used for paired-end sequencing of the 16S rRNA variable 3 and 4 region using the Illumina MiSeq platform (Illumina, Inc., San Diego, CA, USA) as previously described [4]. After sequencing, adapter and barcode sequences were trimmed using Cutadapt v1.2.1 [62]. Sequences were processed following the Divisive amplicon denoising algorithm (DADA) 2 pipeline (open R package), which sample inference was based on error models constructed for the amplicon dataset [63]. Taxonomic assignments for the non-chimeric amplicon sequence variants (ASVs) table were generated using the Silva SSU r132 reference database [64] and performing the IDTAXA algorithm [65].

Microbiota analysis

Species accumulation curves (SAC) were determined using the specaccum function of the vegan package in R [66]. Alpha diversity was calculated using the Shannon Diversity index (SDI). Beta diversity was calculated after proportionally normalizing all samples. Beta diversity clustering plots (i.e. non-metric multidimensional scaling (NMDS)) were generated using Bray-Curtis (BC) dissimilarities metrics to visualise potential clustering patterns among samples. Permutational multivariate analysis of variance (PERMANOVA) statistical analysis [67] were performed as previously described [4] to determine factors that may shape the dynamics of the CF microbiome in patients treated with azithromycin. Permutations were stratified to patient ID to account for the random effects of having multiple measures per patient in the pre-treatment category (i.e. Pre and day 0). DESeq2 [68] was used to identify differences in taxa relative abundance among the primary and secondary outcomes and for the impact of azithromycin in the CF microbiome. Taxa with overall relative abundance < 1% were excluded from DESeq2 analysis.

Statistics

Non-parametric Wilcoxon rank-sum (Mann-Whitney) tests and Fisher exact probability tests were performed for comparisons of patients’ demographics, clinical characteristics, alpha diversity and relative abundance, between Responders and Non-Responders to azithromycin at baseline. Paired statistics (i.e. Wilcoxon signed-rank test) were used for pre-treatment and post-treatment comparisons of changes in alpha diversity and clinical comparisons performed using a two-sided Fisher’s exact test using STATA 16.1 (Stata Corp, Texas, USA). To further examine the effect of multiple samples before treatment (i.e. Pre and day 0) on alpha diversity, linear mixed effects models were generated using the lme4 package in R [69]. We controlled the repeated sampling of the patients by including the patient ID as a random intercept term. False discovery rate was used to control for multiple testing by using Benjamini–Hochberg P-value correction in R 3.6.1 (R Core Team, 2019).

Availability of data and materials

The dataset supporting the conclusions of this article are available in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) repository, BioProject ID PRJNA666737 (http://www.ncbi.nlm.nih.gov/bioproject/666737). The data will be released to the public when the manuscript is formally accepted for publication. Also, the datasets supporting the conclusions of this article is included within the article and its additional file.

Abbreviations

ANOVA:

Analysis of variance

ASV:

Amplicon Sequence Variant

CF:

Cystic fibrosis

CFTR:

Cystic fibrosis transmembrane conductance regulator

CI:

Confidence intervals

COPD:

Chronic obstructive pulmonary disease

FEV1 :

Forced expiratory volume in 1 s

FVC:

Forced vital capacity

IQR:

Inter quartile range

nCFB:

Non-CF bronchiectasis

NMDS:

Non-metric multi-dimensional scaling

PERMANOVA:

Permutational multivariate analysis of variance

PEx:

Pulmonary exacerbation

SDI:

Shannon Diversity Index

RR:

Relative risk

References

  1. Lyczak JB, Cannon CL, Pier GB. Lung infections associated with cystic fibrosis. Clin Microbiol Rev. 2002;15(2):194–222. https://doi.org/10.1128/CMR.15.2.194-222.2002.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. LiPuma JJ. The changing microbial epidemiology in cystic fibrosis. Clin Microbiol Rev. 2010;23(2):299–323. https://doi.org/10.1128/CMR.00068-09.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Surette MG. The cystic fibrosis lung microbiome. Ann Am Thorac Soc. 2014;11(Supplement 1):S61–5.

    Article  PubMed  Google Scholar 

  4. Acosta N, Heirali A, Somayaji R, Surette MG, Workentine ML, Sibley CD, et al. Sputum microbiota is predictive of long-term clinical outcomes in young adults with cystic fibrosis. Thorax. 2018;73(11):1016–25. https://doi.org/10.1136/thoraxjnl-2018-211510.

    Article  PubMed  Google Scholar 

  5. Heirali AA, Workentine ML, Acosta N, Poonja A, Storey DG, Somayaji R, et al. The effects of inhaled aztreonam on the cystic fibrosis lung microbiome. Microbiome. 2017;5(1):51. https://doi.org/10.1186/s40168-017-0265-7.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Heirali A, Thornton C, Acosta N, Somayaji R, Laforest Lapointe I, Storey D, et al. Sputum microbiota in adults with CF associates with response to inhaled tobramycin. Thorax. 2020;75(12):1058–64. https://doi.org/10.1136/thoraxjnl-2019-214191.

    Article  PubMed  Google Scholar 

  7. Heirali AA, Acosta N, Storey DG, Workentine ML, Somayaji R, Laforest-Lapointe I, et al. The effects of cycled inhaled aztreonam on the cystic fibrosis (CF) lung microbiome. J Cyst Fibros. 2019;18(6):829–37. https://doi.org/10.1016/j.jcf.2019.02.010.

    Article  CAS  PubMed  Google Scholar 

  8. Saiman L, Marshall BC, Mayer-Hamblett N, Burns JL, Quittner AL, Cibene DA, et al. Azithromycin in patients with cystic fibrosis chronically infected with Pseudomonas aeruginosa a randomized controlled trial. JAMA. 2003;290(13):1749–56. https://doi.org/10.1001/jama.290.13.1749.

    Article  CAS  PubMed  Google Scholar 

  9. Equi A, Balfour-Lynn IM, Bush A, Rosenthal M. Long term azithromycin in children with cystic fibrosis: a randomised, placebo-controlled crossover trial. Lancet. 2002;360(9338):978–84. https://doi.org/10.1016/S0140-6736(02)11081-6.

    Article  CAS  PubMed  Google Scholar 

  10. Clement A, Tamalet A, Leroux E, Ravilly S, Fauroux B, Jais J-P. Long term effects of azithromycin in patients with cystic fibrosis: a double blind, placebo controlled trial. Thorax. 2006;61(10):895–902. https://doi.org/10.1136/thx.2005.057950.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Mayer-Hamblett N, Retsch-Bogart G, Kloster M, Accurso F, Rosenfeld M, Albers G, et al. Azithromycin for early Pseudomonas infection in cystic fibrosis. The OPTIMIZE randomized trial. Am J Respir Crit Care Med. 2018;198(9):1177–87. https://doi.org/10.1164/rccm.201802-0215OC.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Chalmers JD. Macrolide resistance in Pseudomonas aeruginosa: implications for practice. Eur Respir J. 2017;49(5):1700689. https://doi.org/10.1183/13993003.00689-2017.

    Article  PubMed  Google Scholar 

  13. Levert H, Gressier B, Moutard I, Brunet C, Dine T, Luyckx M, et al. Azithromycin impact on neutrophil oxidative metabolism depends on exposure time. Inflammation. 1998;22(2):191–201. https://doi.org/10.1023/A:1022340107017.

    Article  CAS  PubMed  Google Scholar 

  14. Ratjen F, Saiman L, Mayer-Hamblett N, Lands LC, Kloster M, Thompson V, et al. Effect of azithromycin on systemic markers of inflammation in patients with cystic fibrosis uninfected with Pseudomonas aeruginosa. Chest. 2012;142(5):1259–66. https://doi.org/10.1378/chest.12-0628.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Meyer M, Huaux F, Gavilanes X, Van Den Brûle S, Lebecque P, Re SL, et al. Azithromycin reduces exaggerated cytokine production by M1 alveolar macrophages in cystic fibrosis. Am J Respir Cell Mol Biol. 2009;41(5):590–602. https://doi.org/10.1165/rcmb.2008-0155OC.

    Article  CAS  PubMed  Google Scholar 

  16. Cigana C, Nicolis E, Pasetto M, Assael BM, Melotti P. Anti-inflammatory effects of azithromycin in cystic fibrosis airway epithelial cells. Biochem Biophys Res Commun. 2006;350(4):977–82. https://doi.org/10.1016/j.bbrc.2006.09.132.

    Article  CAS  PubMed  Google Scholar 

  17. Gielen V, Johnston SL, Edwards MR. Azithromycin induces anti-viral responses in bronchial epithelial cells. Eur Respir J. 2010;36(3):646–54. https://doi.org/10.1183/09031936.00095809.

    Article  CAS  PubMed  Google Scholar 

  18. Schögler A, Kopf BS, Edwards MR, Johnston SL, Casaulta C, Kieninger E, et al. Novel antiviral properties of azithromycin in cystic fibrosis airway epithelial cells. Eur Respir J. 2015;45(2):428–39. https://doi.org/10.1183/09031936.00102014.

    Article  CAS  PubMed  Google Scholar 

  19. Kiedrowski MR, Bomberger JM. Viral-bacterial co-infections in the cystic fibrosis respiratory tract. Front Immunol. 2018;9:3067.

    Article  CAS  PubMed Central  PubMed  Google Scholar 

  20. Saiman L, Anstead M, Mayer-Hamblett N, Lands LC, Kloster M, Hocevar-Trnka J, et al. Effect of azithromycin on pulmonary function in patients with cystic fibrosis uninfected with Pseudomonas aeruginosa: a randomized controlled trial. JAMA. 2010;303(17):1707–15. https://doi.org/10.1001/jama.2010.563.

    Article  CAS  PubMed  Google Scholar 

  21. Sibley CD, Parkins MD, Rabin HR, Surette MG. The relevance of the polymicrobial nature of airway infection in the acute and chronic management of patients with cystic fibrosis. Curr Opin Investig Drugs. 2009;10(8):787–94.

    CAS  PubMed  Google Scholar 

  22. Rogers GB, Bruce KD, Martin ML, Burr LD, Serisier DJ. The effect of long-term macrolide treatment on respiratory microbiota composition in non-cystic fibrosis bronchiectasis: an analysis from the randomised, double-blind, placebo-controlled BLESS trial. Lancet Respir Med. 2014;2(12):988–96. https://doi.org/10.1016/S2213-2600(14)70213-9.

    Article  CAS  PubMed  Google Scholar 

  23. Castellani C, Duff AJA, Bell SC, Heijerman HGM, Munck A, Ratjen F, et al. ECFS best practice guidelines: the 2018 revision. J Cyst Fibros. 2018;17(2):153–78. https://doi.org/10.1016/j.jcf.2018.02.006.

    Article  PubMed  Google Scholar 

  24. Kanoh S, Rubin BK. Mechanisms of action and clinical application of macrolides as immunomodulatory medications. Clin Microbiol Rev. 2010;23(3):590–615. https://doi.org/10.1128/CMR.00078-09.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Peter J, Mogayzel J, Naureckas ET, Robinson KA, Mueller G, Hadjiliadis D, et al. Cystic fibrosis pulmonary guidelines. Am J Respir Crit Care Med. 2013;187(7):680–9.

    Article  Google Scholar 

  26. Imperi F, Leoni L, Visca P. Antivirulence activity of azithromycin in Pseudomonas aeruginosa. Front Microbiol. 2014;5:178.

    Article  PubMed Central  PubMed  Google Scholar 

  27. Mulet X, Maciá MD, Mena A, Juan C, Pérez JL, Oliver A. Azithromycin in Pseudomonas aeruginosa biofilms: bactericidal activity and selection of nfxB mutants. Antimicrob Agents Chemother. 2009;53(4):1552–60. https://doi.org/10.1128/AAC.01264-08.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Cogen JD, Onchiri F, Emerson J, Gibson RL, Hoffman LR, Nichols DP, et al. Chronic azithromycin use in cystic fibrosis and risk of treatment-emergent respiratory pathogens. Ann Am Thorac Soc. 2018;15(6):702–9. https://doi.org/10.1513/AnnalsATS.201801-012OC.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Slater M, Rivett DW, Williams L, Martin M, Harrison T, Sayers I, et al. The impact of azithromycin therapy on the airway microbiota in asthma. Thorax. 2014;69(7):673–4. https://doi.org/10.1136/thoraxjnl-2013-204517.

    Article  PubMed  Google Scholar 

  30. Segal LN, Clemente JC, Wu BG, Wikoff WR, Gao Z, Li Y, et al. Randomised, double-blind, placebo-controlled trial with azithromycin selects for anti-inflammatory microbial metabolites in the emphysematous lung. Thorax. 2017;72(1):13–22. https://doi.org/10.1136/thoraxjnl-2016-208599.

    Article  PubMed  Google Scholar 

  31. Jaffé A, Francis J, Rosenthal M, Bush A. Long-term azithromycin may improve lung function in children with cystic fibrosis. Lancet. 1998;351(9100):420. https://doi.org/10.1016/S0140-6736(05)78360-4.

    Article  PubMed  Google Scholar 

  32. Rogers GB, Hart CA, Mason JR, Hughes M, Walshaw MJ, Bruce KD. Bacterial diversity in cases of lung infection in cystic fibrosis patients: 16S ribosomal DNA (rDNA) length heterogeneity PCR and 16S rDNA terminal restriction fragment length polymorphism profiling. J Clin Microbiol. 2003;41(8):3548–58. https://doi.org/10.1128/JCM.41.8.3548-3558.2003.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Stanojevic S, Ratjen F, Stephens D, Lu A, Yau Y, Tullis E, et al. Factors influencing the acquisition of Stenotrophomonas maltophilia infection in cystic fibrosis patients. J Cyst Fibros. 2013;12(6):575–83. https://doi.org/10.1016/j.jcf.2013.05.009.

    Article  PubMed  Google Scholar 

  34. Goss CH, Otto K, Aitken ML, Rubenfeld GD. Detecting Stenotrophomonas maltophilia does not reduce survival of patients with cystic fibrosis. Am J Respir Crit Care Med. 2002;166(3):356–61. https://doi.org/10.1164/rccm.2109078.

    Article  PubMed  Google Scholar 

  35. Samson C, Tamalet A, Thien HV, Taytard J, Perisson C, Nathan N, et al. Long-term effects of azithromycin in patients with cystic fibrosis. Respir Med. 2016;117:1–6. https://doi.org/10.1016/j.rmed.2016.05.025.

    Article  PubMed  Google Scholar 

  36. Bevivino A, Bacci G, Drevinek P, Nelson MT, Hoffman L, Mengoni A. Deciphering the ecology of cystic fibrosis bacterial communities: towards systems-level integration. Trends Mol Med. 2019;25(12):1110–22. https://doi.org/10.1016/j.molmed.2019.07.008.

    Article  PubMed  Google Scholar 

  37. Stressmann FA, Rogers GB, van der Gast CJ, Marsh P, Vermeer LS, Carroll MP, et al. Long-term cultivation-independent microbial diversity analysis demonstrates that bacterial communities infecting the adult cystic fibrosis lung show stability and resilience. Thorax. 2012;67(10):867–73. https://doi.org/10.1136/thoraxjnl-2011-200932.

    Article  PubMed  Google Scholar 

  38. Daniels TWV, Rogers GB, Stressmann FA, van der Gast CJ, Bruce KD, Jones GR, et al. Impact of antibiotic treatment for pulmonary exacerbations on bacterial diversity in cystic fibrosis. J Cyst Fibros. 2013;12(1):22–8. https://doi.org/10.1016/j.jcf.2012.05.008.

    Article  CAS  PubMed  Google Scholar 

  39. Boutin S, Depner M, Stahl M, Graeber SY, Dittrich SA, Legatzki A, et al. Comparison of Oropharyngeal microbiota from children with asthma and cystic fibrosis. Mediat Inflamm. 2017;2017:5047403.

    Article  Google Scholar 

  40. Tiew PY, Jaggi TK, Chan LLY, Chotirmall SH. The airway microbiome in COPD, bronchiectasis and bronchiectasis-COPD overlap. Clin Respir J. 2020;15(2):123–33.

  41. Steinkamp G, Wiedemann B, Rietschel E, Krahl A, Gielen J, Bärmeier H, et al. Prospective evaluation of emerging bacteria in cystic fibrosis. J Cyst Fibros. 2005;4(1):41–8. https://doi.org/10.1016/j.jcf.2004.10.002.

    Article  CAS  PubMed  Google Scholar 

  42. Saiman L, Chen Y, Gabriel PS, Knirsch C. Synergistic activities of macrolide antibiotics against Pseudomonas aeruginosa, Burkholderia cepacia, Stenotrophomonas maltophilia, and Alcaligenes xylosoxidans isolated from patients with cystic fibrosis. Antimicrob Agents Chemother. 2002;46(4):1105–7. https://doi.org/10.1128/AAC.46.4.1105-1107.2002.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Carmody LA, Zhao J, Kalikin LM. The daily dynamics of cystic fibrosis airway microbiota during clinical stability and at exacerbation. Microbiome. 2015;3(1):12. https://doi.org/10.1186/s40168-015-0074-9.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Tunney MM, Klem ER, Fodor AA, Gilpin DF, Moriarty TF, McGrath SJ, et al. Use of culture and molecular analysis to determine the effect of antibiotic treatment on microbial community diversity and abundance during exacerbation in patients with cystic fibrosis. Thorax. 2011;66(7):579–84. https://doi.org/10.1136/thx.2010.137281.

    Article  CAS  PubMed  Google Scholar 

  45. Dickson RP, Morris A. Macrolides, inflammation and the lung microbiome: untangling the web of causality. Thorax. 2017;72(1):10–2. https://doi.org/10.1136/thoraxjnl-2016-209180.

    Article  PubMed  Google Scholar 

  46. Acosta N, Whelan FJ, Somayaji R, Poonja A, Surette MG, Rabin HR, et al. The evolving cystic fibrosis microbiome: a comparative cohort study spanning 16 years. Ann Am Thorac Soc. 2017;14(8):1288–97. https://doi.org/10.1513/AnnalsATS.201609-668OC.

    Article  PubMed  Google Scholar 

  47. Hogan DA, Willger SD, Dolben EL, Hampton TH, Stanton BA, Morrison HG, et al. Analysis of lung microbiota in Bronchoalveolar lavage, protected brush and sputum samples from subjects with mild-to-moderate cystic fibrosis lung disease. PLoS One. 2016;11(3):e0149998. https://doi.org/10.1371/journal.pone.0149998.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Whelan FJ, Heirali AA, Rossi L, Rabin HR, Parkins MD, Surette MG. Longitudinal sampling of the lung microbiota in individuals with cystic fibrosis. PLoS One. 2017;12(3):e0172811. https://doi.org/10.1371/journal.pone.0172811.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Caverly LJ, Lu J, Carmody LA, Kalikin LM, Shedden K, Opron K, et al. Measures of cystic fibrosis airway microbiota during periods of clinical stability. Ann Am Thorac Soc. 2019;16(12):1534–42. https://doi.org/10.1513/AnnalsATS.201903-270OC.

    Article  PubMed  PubMed Central  Google Scholar 

  50. Carmody LA, Zhao J, Schloss PD, Petrosino JF, Murray S, Young VB, et al. Changes in cystic fibrosis airway microbiota at pulmonary exacerbation. Ann Am Thorac Soc. 2013;10(3):179–87. https://doi.org/10.1513/AnnalsATS.201211-107OC.

    Article  PubMed  PubMed Central  Google Scholar 

  51. Nelson MT, Wolter DJ, Eng A, Weiss EJ, Vo AT, Brittnacher MJ, et al. Maintenance tobramycin primarily affects untargeted bacteria in the CF sputum microbiome. Thorax. 2020;75(9):780–90. https://doi.org/10.1136/thoraxjnl-2019-214187.

    Article  PubMed  Google Scholar 

  52. Nick JA, Moskowitz SM, Chmiel JF, Forssén AV, Kim SH, Saavedra MT, et al. Azithromycin may antagonize inhaled tobramycin when targeting Pseudomonas aeruginosa in cystic fibrosis. Ann Am Thorac Soc. 2014;11(3):342–50. https://doi.org/10.1513/AnnalsATS.201310-352OC.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Nichols DP, Happoldt CL, Bratcher PE, Caceres SM, Chmiel JF, Malcolm KC, et al. Impact of azithromycin on the clinical and antimicrobial effectiveness of tobramycin in the treatment of cystic fibrosis. J Cyst Fibros. 2017;16(3):358–66. https://doi.org/10.1016/j.jcf.2016.12.003.

    Article  CAS  PubMed  Google Scholar 

  54. Ross KR, Chmiel JF, Konstan MW. The role of inhaled corticosteroids in the management of cystic fibrosis. Paediatr Drugs. 2009;11(2):101–13. https://doi.org/10.2165/00148581-200911020-00002.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Balfour-Lynn IM, Welch K, Smith S. Inhaled corticosteroids for cystic fibrosis. Cochrane Database Syst Rev. 2019;(7).

  56. Balfour-Lynn IM, Lees B, Hall P, Phillips G, Khan M, Flather M, et al. Multicenter randomized controlled trial of withdrawal of inhaled corticosteroids in cystic fibrosis. Am J Respir Crit Care Med. 2006;173(12):1356–62. https://doi.org/10.1164/rccm.200511-1808OC.

    Article  CAS  PubMed  Google Scholar 

  57. Chen Y-F, Hou H-H, Cheng S-L, Wang H-C, Yu C-J. Different inhaled corticosteroids altered lung microbiome and regulating the expression of PAF receptor as well as the phagocytosis function of macrophage. Eur Respir J. 2019;54(suppl 63):PA2380.

    Google Scholar 

  58. Martin MJ, Zain NMM, Hearson G, Rivett DW, Koller G, Wooldridge DJ, et al. The airways microbiome of individuals with asthma treated with high and low doses of inhaled corticosteroids. PLoS One. 2021;15(12):e0244681.

    Article  Google Scholar 

  59. Waters V. Chronic antibiotic use in cystic fibrosis: a fine balance. Ann Am Thorac Soc. 2018;15(6):667–8. https://doi.org/10.1513/AnnalsATS.201803-172ED.

    Article  PubMed  Google Scholar 

  60. Lam JC, Somayaji R, Surette MG, Rabin HR, Parkins MD. Reduction in Pseudomonas Aeruginosa sputum density during a cystic fibrosis pulmonary exacerbation does not predict clinical response. BMC Infect Dis. 2015;15(1):145. https://doi.org/10.1186/s12879-015-0856-5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  61. Proesmans M, Balinska-Miskiewicz W, Dupont L, Bossuyt X, Verhaegen J, Høiby N, et al. Evaluating the “Leeds criteria” for Pseudomonas aeruginosa infection in a cystic fibrosis centre. Eur Respir J. 2006;27(5):937–43.

  62. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011;17(1):10–2. https://doi.org/10.14806/ej.17.1.200.

    Article  Google Scholar 

  63. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13(7):581–3. https://doi.org/10.1038/nmeth.3869.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  64. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013;41(Database issue):D590–D6. https://doi.org/10.1093/nar/gks1219.

    Article  CAS  PubMed  Google Scholar 

  65. Murali A, Bhargava A, Wright ES. IDTAXA: a novel approach for accurate taxonomic classification of microbiome sequences. Microbiome. 2018;6(1):140. https://doi.org/10.1186/s40168-018-0521-5.

    Article  PubMed  PubMed Central  Google Scholar 

  66. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. Vegan: community ecology package. R package version 24–5. 2017. https://CRAN.R-project.org/package=vegan. Accessed 1 Feb 2021.

    Google Scholar 

  67. Anderson MJ. A new method for non-parametric multivariate analysis of variance. Austral Ecol. 2001;26(1):32–46.

    Google Scholar 

  68. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):1–21.

    Article  Google Scholar 

  69. Bates D, Mächler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. arXiv preprint arXiv:14065823; 2014.

    Google Scholar 

Download references

Acknowledgements

The authors gratefully acknowledge the staff of the Calgary Adult CF Clinic and Calgary Laboratory Services for their continued efforts to stock and maintain the CAC FC Biobank. The authors thank Matthew L. Workentine for providing support with bioinformatics analysis.

Funding

This work was supported by a grant from CIHR to MDP.

Author information

Authors and Affiliations

  1. Department of Microbiology, Immunology and Infectious Diseases, University of Calgary, 3330 Hospital Drive, NW, Calgary, Alberta, Canada

    Nicole Acosta, Christina S. Thornton, Michael G. Surette, Ranjani Somayaji, Harvey R. Rabin & Michael D. Parkins

  2. Department of Biochemistry and Biomedical Sciences, McMaster University, Hamilton, Ontario, Canada

    Michael G. Surette & Laura Rossi

  3. Department of Medicine, University of Calgary, 3330 Hospital Drive, NW, Calgary, Alberta, Canada

    Ranjani Somayaji, Harvey R. Rabin & Michael D. Parkins

Authors
  1. Nicole Acosta
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Christina S. Thornton
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Michael G. Surette
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Ranjani Somayaji
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Laura Rossi
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Harvey R. Rabin
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Michael D. Parkins
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

NA and CST were responsible for sample collection. NA was primarily responsible for DNA extraction and microbiome analysis. Sample analysis and statistical analyses were performed by NA, RS, LR, HRR and MGS. MDP, RS and NA extracted and analyzed clinical data. NA wrote the initial draft of the manuscript, and all authors contributed to its revision. MDP is the guarantor of this work. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Michael D. Parkins.

Ethics declarations

Ethics approval and consent to participate

This study was performed in accordance with the Declaration of Helsinki and was approved by the University of Calgary Conjoint Regional Health Ethics Board with the following reference numbers: REB15–0854 and REB15–2744. Patients were recruited from the Calgary Adult CF Clinic and all patients provide written informed consent for the collection and storage of specimens and subsequent analysis. All samples were de-identified. Patients have contributed samples to this regional ethics board (REB) collection from 1998-present (REB15–0854) and provide consent for ongoing research purposes (REB15–2744).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Supplementary Figure 1.

Species accumulation curves (SAC) determined using the specaccum function from vegan package in R using method = “random” and permutations = 500. SAC plot shows the increase in ASVs detected with the addition of each patient sample. Supplementary Figure 2. NMDS plot showing beta diversity of CF patient treated with azithromycin based on Bray-Curtis dissimilarities. Each patient is color coded and was sampled up to three time points: Pre (≤24 months pre-initiation treatment), day 0 (start day on azithromycin) and Post (≤24 months post its initiation treatment). Arrows indicate the timewise sequence of samples (i.e. Pre to day 0 to Post).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Acosta, N., Thornton, C.S., Surette, M.G. et al. Azithromycin and the microbiota of cystic fibrosis sputum. BMC Microbiol 21, 96 (2021). https://doi.org/10.1186/s12866-021-02159-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12866-021-02159-5

Keywords

BMC Microbiology

ISSN: 1471-2180

Contact us