Discussion
In this large retrospective cohort study, we found that individuals with CF and frequent fungal detection had higher rates of co-infection withP. aeruginosa and S. maltophilia compared to those with no or rare detection of fungi. Other clinical characteristics and outcomes did not differ between those with frequent versus rare fungi including frequency of ABPA diagnosis, nutritional status, and rate of pulmonary exacerbations. Lung function (ppFEV1) and lung function decline were also similar between groups including after controlling for chronic P. aeruginosa and ABPA.
In individuals with ABPA however, frequent fungal detection was associated with lower lung function (ppFEV1) compared to those with ABPA but rare fungal detection. Individuals with ABPA/frequent fungi had higher rates of co-infection with P. aeruginosa and S. maltophilia ; more CFRD diagnoses; and higher IgE values compared to those with ABPA/rare fungi and compared to those without ABPA regardless of fungal categorization. Individuals with ABPA/ frequent fungi also had a trend towards more rapid lung function decline during follow-up. Interestingly, we found equal proportions of ABPA among the rare and frequent fungal groups.
As expected, those with ABPA had higher absolute eosinophils and IgE values, markers of T-helper cells type 2 (Th2), typical of allergic or hypersensitivity responses. We did not detect a difference in eosinophils or IgE between the frequent and rare fungal groups. Interestingly, 60% of the rare group and 55% of the frequent group had a diagnosis of asthma over the entire study, which is higher than the general population as well as previously published literature on patients with CF.19, 20 Furthermore, in the baseline period, more individuals in the rare group had asthma, while in the follow up period, more individuals in the frequent group had asthma. . While the follow up period had more time to collect data and is likely more robust, the significance of this finding is unclear. Asthma diagnoses were documented based upon coding in our electronic medical record; thus diagnoses may reflect institutional practices, regional effects and allergens, or another unknown confounding variable.
Individuals with frequent fungal infection had higher rates of chronicP. aeruginosa co-infection and more P. aeruginosadetection overall compared to the rare fungus group. A similar pattern was seen in those with ABPA and frequent fungi, having the highest proportion of individuals with chronic P. aeruginosa infection when compared to the other three groups. This may be related to the use of aggressive antibiotic therapy in P. aeruginosa eradication protocols, allowing for shifts in biodiversity by creating a less competitive airway environment for fungal species.11, 21 Other studies and authors have suggested that biofilms created byP. aeruginosa and/or Aspergillus may be initially favorable and synergistic for fungal spores to germinate.22 However, other studies have shown thatP. aeruginosa biofilms actually inhibit Aspergillusgrowth, favoring the competition theory.23 There was no significant differences in MRSA co-infection among any of the groups, another organism possibly related to more frequent therapies and driven by microbial management.24 In the follow-up period, people in our cohort with rare fungi more often isolated P. aeruginosa prior to later fungal detection, rather than isolating fungi first, supporting the role of antimicrobial therapies and/or a synergistic environment in fungal growth, although our numbers were small.
S. maltophilia was associated with more frequent fungal infection at baseline and during follow-up. It also had a different longitudinal pattern than P. aeruginosa , showing that almost 50% isolatedS. maltophilia prior to fungus and vice versa. While one investigation in the associations with fungi and S. maltophiliahave been seen in people with CF and liver disease, the literature is still lacking in understanding of how the organisms are related.25 Our findings are both supportive of and contrary to existing data, but imply an association between the acquisition of fungus and subsequent infection with P. aeruginosaand S. maltophilia . Given this, more research is needed to identify associated risk factors as well as cause and effect relationships between these organisms.
The association between fungus and lung disease progression has been examined previously in CF. Studies have shown lower ppFEV1 in those with chronic fungal infection, yet the definition of chronic varies widely in the literature. Some investigators have adapted criteria similar to that of P. aeruginosa classifications, however there are many genera and species of fungus that make definitions difficult. Furthermore, given the need for sputum or BAL for fungal culture, detection of fungi is dependent on the person’s ability to expectorate, further complicating classification, prevalence, and disease severity. This likely explains the higher rates of bacteria and fungus isolation in our cohort when compared to registry data.26
Most studies in the literature focus on Aspergillus orAspergillus fumigatus specifically, excluding other fungal genera.5, 7, 17 We included all fungi (with the exception of Candida that is not routinely reported by our clinical laboratory) as genera such as Scedosporium may also contribute to CF lung disease.27, 28 However, over 70% of fungal positive cultures in our cohort grew Aspergillus . Our study was larger than most and included follow-up for a median of 6 years.
Our findings agree with the literature in finding higher morbidity and worse outcomes in those with ABPA. However, our study was one of the few comparing ABPA to fungal frequency both as a co-variate and as a separate group.6, 7 By doing so, we found that those with both ABPA and frequent fungus appeared to have more severe disease despite the fact that the presence of fungus in the sputum is not required for ABPA diagnosis. This dynamic emphasizes the importance of fungus on inflammation in the CF airway, yet how the pathophysiology of fungal infections needs further investigation.
This study has several limitations. Most importantly, this was a single center, retrospective study so we are unable to determine causation between fungal disease, ABPA and outcomes. Given electronic medical record limitations, we only evaluated culture results as early as 2004. It is possible that some individuals isolated fungi before this and thus meet other fungal group classification. Fungi are also difficult to isolate in culture.21, 29 Given this, there are likely individuals who have fungi that were not detected and may account for those without ABPA in the absence of positive fungal cultures. We tried to account for this by requiring at least 4 lower airway cultures for analysis. It is likely that most of the individuals in this study could spontaneously expectorate, and thus were likely more severely impacted compared to those who could not expectorate. This may explain the higher rates of fungi and bacterial pathogens given a more reliable sample than throat swabs. Furthermore, individuals in the frequent fungal group had significantly more cultures compared to those with rare fungi which may lead to surveillance bias. Prospective longitudinal studies are needed to assess the acquisition of fungus in real time and how use of antibiotics and co-infections contribute. Mechanistic studies are also needed to explore the interactions of fungi and other organisms in the CF airway as well as inflammatory host-responses, specific biomarkers of fungal disease and allergy, and clinical studies investigating characteristics regarding the development of ABPA.