Background
Cystic fibrosis (CF) is the most common lethal autosomal recessive disease amongst Caucasians [
1]. Typically affected organs include the sinuses, lungs, gastrointestinal system and the male reproductive system. Pulmonary complications (both acute and chronic) are the primary cause of morbidity and mortality in the adult CF population [
2,
3]. CF lungs are classically characterized by viscous secretions as well as impaired mucociliary clearance [
4,
5]. These factors compromise airway clearance creating an optimal environment for bacterial colonization, inflammation, chronic infection, and eventually, bronchiectasis.
The cultured microbiome of the CF respiratory system is well-characterized and unique from other forms of chronic lung disease. Nevertheless, our understanding of lung microbiology in the CF population continues to evolve. Traditional perceptions focused on chronic colonization with classical CF pathogens such as
Pseudomonas aeruginosa, Staphylococcus aureus, Haemophilus influenzae, and Burkholderia cepacia complex [
6]. Other microorganisms such as
Stenotrophomonas maltophilia, Achromobacter species, and mycobacteria have been identified as emerging CF pathogens in the last two decades [
6,
7]. Whereas traditional pathogens like
P. aeruginosa and
B. cenocepacia have clearly been associated with adverse outcomes, the clinical implications are less clear with many emerging pathogens [
8‐
10]. Furthermore, it is apparent that CF airways disease may exist in both stable (classical CF pathogens) and dynamic states (non-classical pathogens whose presence in the CF airways disease is usually only temporary). How these transient infections affect CF outcomes is entirely unknown. Consequently, clinicians are posed with unique management challenges when these atypical microbes are isolated from the lower respiratory tracts of CF patients.
One such organism is Group A Streptococcus
(Streptococcus pyogenes), a common human pathogen that is rarely observed from CF lungs.
S. pyogenes are β-hemolytic Gram positive cocci and facultative anaerobes. Its ability to cause hemolysis on blood agar plates allows it to be distinguished from other more indolent streptococci [
11].
Furthermore, GAS has numerous virulence factors that enable it to evade host immune defenses, colonize epithelial surfaces, and cause infection [
12,
13].
GAS can cause a variety of different infections with a spectrum of disease severity ranging from mild to invasive and life threatening [
14]. While it most often manifests as pharyngeal, soft tissue or skin infections, it can rarely manifest as a respiratory infection. Although GAS only accounts for a small proportion of community acquired pneumonia in the general population, it tends to cause more severe and aggressive pulmonary infections and often manifest as empyema [
14]. The prevalence, natural history and clinical effect of GAS in the CF population are unknown. Herein we set out to determine the natural history of GAS and outcomes associated with GAS infection in adults with CF. Furthermore, we sought to characterize GAS to determine if genotypic or phenotypic features were associated with disease.
Discussion
Our study is the first to investigate the prevalence of GAS and its clinical effects in the adult CF population.
S. pyogenes is a common organism that can affect healthy individuals of any age [
14]. While GAS may transiently colonize the upper respiratory tract as a commensal organism (as is the case in 15 to 20 % of healthy children), it has also proven itself as a major human pathogen [
24]. It is responsible for a broad spectrum of disease including pharyngitis, scarlet fever, rheumatic fever, cellulitis, necrotizing fasciitis, toxic shock syndrome, and pneumonia [
14]. GAS infections can range from mild to life threatening; severe disease is not limited to those with chronic illness or immune compromise [
14]. Globally, GAS is responsible for a significant burden of disease; every year it accounts for 110 million skin and soft tissue infections [
11], 660 000 cases of invasive infection [
11] and over 294 000 deaths [
25].
From a pulmonary standpoint, GAS is an infrequent but important cause of pneumonia [
26]. While it only accounts for a small percentage of community acquired pneumonia (CAP) in adults,
S. pyogenes tends to cause more severe and invasive infection compared to common CAP pathogens [
14]. Furthermore, GAS has a higher likelihood of resulting in necrotizing pneumonia, progression to empyema and/or hemorrhagic pleural effusion [
13,
27]. Mortality rates for GAS pneumonia range between 20 and 38 %, which is similar to that of necrotizing fasciitis [
13,
14].
Although
S. pyogenes is a major global pathogen, little is known about its role in CF lung disease. Given the potential for GAS to cause exceptional virulence and the compromised innate immunity of CF lungs, we hypothesized that GAS may lead to adverse clinical outcomes in this population. There is a paucity of GAS epidemiologic data in the CF literature, with prior studies reporting prevalence of 0.8 % (2/258) [
28] and 0.9 % (4/495) [
29], respectively. Over 34 years, we identified that five percent of patients isolated GAS on at least one occasion. However, the presence of GAS increased the risk of PEx relative to the preceding clinic visit, particularly if present as the numerically dominant sputum pathogen. This finding may warrant treating individuals with GAS in their sputum with anti-GAS treatments in order to potentially avoid an ensuing PEx. However, other factors including exacerbation of chronically infecting pathogens, and inter-current upper respiratory viral illnesses could also have contributed (although these factors were just as likely in comparator clinical visits).
Within CF, it is clear that mere culture status may not convey the entire story. Indeed, differential pathogenic potential has been observed with the expression of a number of phenotypic traits of classical CF pathogens including
P. aeruginosa,
Bcc, and
S. aureus. For example, compared to patients with chronic methicillin-sensitive
S. aureus (MSSA) infection, those with chronic methicillin-resistant
S. aureus (MRSA) have an increased risk of death [
30]. Patients with MRSA have increased rates of lung function decline [
31] and are less likely to recover lung function following PEx [
32]. In patients with chronic
P. aeruginosa infections, its conversion to a hyper-alginate producing, mucoid phenotype is associated with progressive decline in lung function, increased risk of hospitalization and reduced survival [
8,
33‐
36]. The opposite appears true in
Bcc chronically infected patients, where mucoidy appears protective and patients with non-mucoid isolates experience an exaggerated rate of clinical decline [
37,
38]. Even the ability to persist within the CF lung seems to be influenced by specific phenotypic traits of
P. aeruginosa causing initial infections [
39]. The phenotypes that are associated with these strains may themselves not be directly involved in disproportionate lung disease, but rather they may be an indirect marker. Accordingly, we sought to characterize easily assayable and important virulence traits within infecting GAS strains to determine if these factors disproportionally modified PEx risk.
The GAS isolated from CF airways were typical of GAS reported in other diseases [
40]. We identified variable expression of virulence factors, which has previously been reported [
41]. Of the limited virulence factors assessed, expression did not increase risk for occurrence of PEx at the time of isolation. Almost all GAS isolates produced AI-2, a diffusible cell-cell signaling molecule enabling inter-species bacterial communication on cell density [
42,
43]. Either directly (GAS mediated primary effects) or indirectly (through induction of quorum sensing in patients chronically colonized with pathogens such as
P. aeruginosa), GAS may be able to trigger a PEx [
44].
The antibiotic sensitivity profile of our strains was similar to that of other
S. pyogenes epidemiologic studies [
40,
45], with the exception of a complete absence of clindamycin resistance in our few isolates. Interestingly, our study along with several small studies [
40,
45], have not found high rates of macrolide or fluoroquinolone resistance, as reported in larger studies (despite their frequent use in CF). Indeed, this may suggest that those GAS in CF are not unrecognized chronic endogenous lower respiratory tract flora in these individuals but rather newly acquired transient organisms not previously exposed to antibiotics. Importantly, our GAS strains were sensitive to antibiotics commonly used in the empiric treatment of CF PEx, although ceftazidime, an antibiotic commonly used in the empiric PEx management, should not be used where GAS is involved. Nearly all strains were sensitive to azithromycin; this raises the possibility that chronic azithromycin may suppress GAS and prevent initial colonization. Indeed, registry data suggests that 60–70 % of CF patients with chronic
P. aeruginosa infection and 22 % of CF patients without chronic
P. aeruginosa receive chronic macrolide therapy and this may account for its low observed incidence in our cohort [
46,
47].
PFGE has been shown to be similarly effective at differentiating commonly infecting clones of GAS as other established typing modalities including
emm gene typing [
48‐
50]. Using PFGE we demonstrated strain persistence in those patients with repeated positive cultures, rather than repeated new infections with different strains. We did identify two patients with the same GAS isolate by PFGE, but propose this was unlikely to be patient-patient spread. Whereas typical CF pathogens are rare and opportunistic of the general population, GAS commonly colonizes the upper respiratory tract in the general population and common strains persist in locals for extended period of times [
51,
52]. Furthermore, GAS from these patients were identified > 2.5 years apart with multiple negative cultures in the ensuing time period making CF patient-patient transmission biologically implausible.
Members of the genus
Streptococcus have not traditionally been considered CF respiratory pathogens. However, using a combination of semi-selective agars and high density sampling, high rates of Viridans Group Streptococci (VGS) have universally been identified [
53‐
55]. The
Streptococcus anginosus group, in particular, is increasingly thought to have a role in CF and its emergence as numerically dominant organism has been observed in a large subset of PEx [
54]. Traditional clinical microbiology protocols have been developed to overlook VGS. However, GAS is easily identifiable as beta-hemolysis is a defining feature (present in 99 % of isolates) [
56], and as such is more likely to be distinguished from oropharyngeal streptococci using traditional culture techniques.
The role of bacteria in CF airways disease has been viewed through the lens of contributing to chronic progressive lung disease [
57,
58]. Indeed, when assessing traditional CF pathogens such as
P. aeruginosa and
B. cepacia complex, as well as emerging organisms such as
S. maltophilia, this model holds true. However, increasingly organisms not typically associated with CF airways disease including the Enterobactericeae, Pneumococcus, and GAS are seen to transiently colonize the airways [
8,
59,
60]. The impact of these organisms on short-term and long-term outcomes are for the most part unknown. Given that transient colonization/infection with respiratory viruses has been shown to produce short term deleterious effects either through direct pathogenesis or indirectly through resident microflora, so too might transient bacterial pathogens [
61]. Indeed, while an acute impact on patient well-being was observed with incident GAS infection, long term effects were not noted, nor should they be expected. This highlights the concept that emerging CF pathogens may not necessarily manifest as chronic infections, as is often seen with
P. aeruginosa, S. aureus, and
Bcc.
The main limitations of the study are the retrospective design and wide confidence intervals relating to small sample size and few events. Despite trying to account for factors that could influence the risk of PEx, we were unable to control for differences in chest physiotherapy quality and frequency as well as adherence to medications. We also lacked documentation regarding potential confounding viral and/or environmental triggers for PEx. Future studies assessing potential short-term impact of transient airway colonizers are warranted based on the data herein.
Competing interests
MDP and MGS are supported through research grants from Cystic Fibrosis Canada and Gilead. MDP, HRR have performed advisory board duties for Gilead, Novartis, Roche and Vertex. None of these relate to the work contained herein and the authors declare that they have no competing interests.
Authors’ contributions
KS was responsible for collection and analysis of all patient related data, and drafting the manuscript. AN, BW, CST performed GAS genotyping and phenotyping and assisted in the revision of the manuscript. RS assisted in the statistical analysis, data collection and revision of the manuscript. MGS, HRR and MDP were responsible for the conception of the project, supervision of the collection and analysis of patient and microbiologic data, and assisted with revising the manuscript. All authors read and approved the final manuscript.