Stenotrophomonas maltophilia in people with Cystic Fibrosis: a systematic review of prevalence, risk factors and management

A systematic review of the literature was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guideline recommendations [11]. A search of the literature in medical databases, including MEDLINE by PubMed, Cochrane Library and EMBASE, for articles published in English from 2000 to 2022, was performed. Used keywords, limited to Title, were as follows: “(Stenotrophomonas maltophilia) AND (Cystic Fibrosis)”. Duplicates were removed, references of selected articles were included if pertinent and if they fulfilled the inclusion criteria. National and International CF Registries were searched on Google Scholar database.

For completeness, we subsequently expanded the search strings by adding the following keywords: limited to title: "(Stenotrophomonas maltophilia) AND (Cystic Fibrosis) AND (lung function)"; "(Stenotrophomonas maltophilia) AND (Cystic Fibrosis) AND (FEV1)"; "(Stenotrophomonas maltophilia) AND (Cystic Fibrosis) AND (treatment)"; "(Stenotrophomonas maltophilia) AND (Cystic Fibrosis) AND (antibiotic)"; "(Stenotrophomonas maltophilia) AND (Cystic Fibrosis) AND (genotype)"; "(Stenotrophomonas maltophilia) AND (Cystic Fibrosis) AND (phenotype)"; "(Stenotrophomonas maltophilia) AND (Cystic Fibrosis) AND (heterogeneity)". We also extended the search to include: "(Stenotrophomonas maltophilia) AND (lung function)"; "(Stenotrophomonas maltophilia) AND (FEV1)"; "(Stenotrophomonas maltophilia) AND (treatment)"; "(Stenotrophomonas maltophilia) AND (antibiotic)"; "(Stenotrophomonas maltophilia) AND (genotype)"; "(Stenotrophomonas maltophilia) AND (phenotype)"; "(Stenotrophomonas maltophilia) AND (heterogeneity)". After this further search and after removing duplicates, no additional information was found beyond what had already been selected in the previous articles.

The search was restricted to the English language. Articles reporting SM prevalence in patients with CF, risk factors for SM infection, effect of SM on lung function and therapeutic strategies available for SM eradication in acute and chronic settings were initially included. Review articles, commentaries, editorials, and letters to the author with no original data were excluded.

Duplicate publications were removed, then three authors separately (MT, GG and ID) checked the titles and abstracts and removed irrelevant studies according to the inclusion and exclusion criteria. For each article with original data: author, country, year of publication, type of study, type and number of included participants were analysed and summarized in Table 1.

For observational studies, adherence to Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations [36] was assessed and reported in Table 2. The results of the assessment suggest that the overall quality of studies included in the systematic review was medium. Particularly, most of them did not evaluate bias of the study and did not give an adequate characterization of the quantitative variables and of the data sources. However, almost all studies clearly defined study design, setting, outcome data and adequately discuss the key results.

Overall, 184 records were initially identified and duplicates were removed. One hundred and five articles were screened, of which 70 were excluded by title and abstract because they didn’t match the inclusion criteria. The remaining 35 records were analysed in full text and we selected 15 articles with original data focused on clinically relevant aspects such as SM prevalence in patients with CF, risk factors for SM infection, effect of SM on lung function and therapeutic strategies.

Out of the resulted articles, the 2021 Cystic Fibrosis Foundation Patient Registry (CFFPR) [37], the 2021 European Cystic Fibrosis Society Patient Registry (ECFSR)[38], the 2020 Italian Cystic Fibrosis Registry (ICFR) [39], the 2021 French Cystic Fibrosis Registry (FCFR) [40], the 2021 Australian Cystic Fibrosis Data Registry (ACFDR)[41], the 2021 Canadian Cystic Fibrosis Registry (CCFR) [42], were included in the Review. Nine pertinent articles from the references of the studies selected were also considered. Figure 1 shows the flow diagram of literature search and data extraction.

Characteristics of the original articles included in the review are summarized in Table 1.

The original articles and the registries included in the systematic review were related to 4 mail topics: (1) prevalence and risk factors for SM infection and colonization, (2) impact of SM on lung function in patients with SM, (3) genotype and phenotype heterogeneity of SM, (4) antimicrobial therapy available against SM infection.

Main outcomes of each original article included in the review are summarized in Table 3 and discussed more in detail in the dedicated paragraphs in the Discussion section.

Prevalence of SM in patients with CF was analysed based on the national and international registries from countries with high prevalence of CF. Some registries, such as the French and the North America’s ones [37, 40], showed a declining trend in prevalence of SM, on contrast the ICFR reported an increasing in SM prevalence in patients with CF [39]. Differences in prevalence of SM infection by country are represented in Fig. 2. Which factors could play a role as risk factor for SM infection is unclear, although some studies suggested: younger age, lung function decline, use of antimicrobial drugs, exposure do oral steroids and Aspergillus fumigatus co-infection [12‐17].

The impact of SM infection on lung function in patients with CF is still under debate. Most studies included in the systematic review related to this topic (seven out of nine articles) showed a worst pulmonary outcome in CF patients with SM infection, particularly with a higher decrease in Forced Expiratory Volume in the 1st second (FEV1), increase in hospitalization, in mortality and an in need for lung transplantation [18‐24]. A higher proportion of fungal co-infections, mainly by Aspergillus fumigatus, was also reported in CF patients with SM infection [23].

All the articles about genotype and phenotype heterogeneity of SM reported a high heterogeneity, although the role of this heterogeneity in the pulmonary outcome has not been investigated [25‐28].

No antimicrobial drug nor antimicrobial regimen’s duration for treatment of SM infections, both acute and chronic, has been established. Related to the antimicrobial drugs, minocycline, doxycycline, trimethoprim-sulfamethoxazole, ticarcillin-clavulanate and aerosolized levofloxacin were suggested by the original articles included in the systematic review [27, 29, 30]. The antibiotic strategies available against SM infection in CF are summarized in Table 4. Very little is available about the antimicrobial regimen’s duration, although no differences from a 10-day regimen to a 21-day regimen was reported [31].

The reported prevalence of SM among patients with CF is various, with many differences mainly depending on the country. For the 2021 CFFPR Annual Data Report, including CF patients from North America, the prevalence of the bacteria was 5.6%, declining from the reported prevalence of 12.7% and 13.1% respectively in 2006 and 2016 [37]. Also the 2021 CCFR reported a decrease in SM prevalence in the last few years, from 14.2% in 2017 to 13.7% in 2021 [42, 43]. Interestingly, in contrast with these results, in France an increasing trend in the prevalence of colonization by SM in CF patients has been reported, from 4.7% in 1999 to 10.5% in 2016 [32]. The proportion of colonization by SM in CF patients in France is still decreasing, as the 2021 FCFR reported 9.3% of patients to be colonized [40]. According to the 2021 ECFSR data, the prevalence of the SM infections was 6.6% in children and 7.7% in adults, considering both chronic and not chronic or intermittent infections. The highest proportions were observed in Northern Europe, reaching 25.0% in the paediatric population in Iceland [38]. Differently from the proportion reported by the 2021 ECFPR, the 2021 ACFDR reported a higher prevalence of SM in children than in adults affected by CF, respectively 7.6% and 6.5% [41]. In Italy in 2020 the prevalence of SM reported by the ICFR was 7.6% in the adult population and 8.6% in children. The data reported by the ICFR, in contrast with that of North America, showed an increasing prevalence of SM, with and increasing from 2.9% in 2018 to 7.6% in 2020 in adults and from 2.6% in 2018 to 8.6% in 2020 in paediatrics [39].

Figure 2 summarizes the average SM prevalence in adults and children affected by CF, considering the proportions reported by the Registries included in the review.

While a seasonal variation for some common bacterial pathogens in CF patients has been described, as for PA, methicillin-susceptible Staphylococcus aureus, Achromobacter xylosoxidans and Haemophilus influenzae, such variation was excluded for SM [33].

The differences in prevalence are likely due to variations in local ecology [32], even if some other factors influencing the acquisition of the infection might play a role. Little is known about factors which could impact on the colonization of SM, and consequently may influence the prevalence in patients with CF. Firstly, some studies described an association between the severity of lung disease and the acquisition of SM infection, with a higher risk in patients with a faster decline in FEV1 [12]. It is debated whether the use of antibiotics could represent a risk factor for the infection. Some studies reported a higher risk of SM isolation in case of antibiotic therapy [13, 14], nevertheless, other authors described exposure to antibiotic courses as a protective factor through the preservation of lung function [12]. Particularly, Denton et al. described an increase in SM isolation in patients who received anti-PA antibiotic courses, suggesting that the treatment of common infections in CF patients could raise the risk of colonization by SM [15]. A study by Marchac et al. described an association between the isolation of Aspergillus fumigatus and subsequent SM infection, although this finding has not been well supported [16]. In agreement with Marchac’s study, Paugam et al. observed a higher proportion of SM colonization in CF patients with Aspergillus fumigatus [17].

The effect of chronic colonization of SM on lung function and clinical status in CF patients is still unclear. Goss et al. analysed data from the CF Foundation National Patient Registry (CFNPR) registers from 1994 to 1999. In this extensive cohort study on 2739 CF patients, there was no association between SM and a decrease of lung function after controlling for confounders (age, sex, weight, height, pancreatic insufficiency, PA and Burkolderia cepacia colonization, use of intravenous antibiotics) [34]. A subsequent cohort study from 2008 to 2009 compared 82 CF patients with at least one positive culture of SM to a CF control group with no chronic gram-negative infections. In this study, patients with SM positive cultures every month for 6 consecutive months or, less often, when combined with an increase in number of specific, precipitating antibodies were defined as chronically infected. They found that patients who had been chronically infected with SM for at least 2 years, had a significantly larger decline in lung function, demonstrated as change in FEV1% of predicted value per year. However, no change was detected in the rate of FEV1 decline when those patients were compared to themselves in the previous 3 years before they became chronically infected [18]. A similar retrospective cohort study showed that chronic SM status (defined as 2 or more positive sputum or bronchoalveolar cultures in the previous 12 months) does not affect FEV1 recovery and SM antibiotic treatment does not influence the recovery or the gain in FEV1 after a pulmonary exacerbation [35]. In contrast, the same group found increased rates of mortality and lung transplantation among patients with SM chronic infection, although this effect was no longer significant in a time-varying model that includes lung function [19].

Recently, some observational studies suggested that SM chronic infection may be associated with worse respiratory outcomes and accelerated lung function decline. In a retrospective review of medical records with CF in the USA, Com et al. compared children with low and high initial FEV1, in order to analyse their baseline characteristics. The authors described a significant correlation between low initial FEV1 measurements and positive respiratory culture for SM (p<0.05) [20]. In 2015, Cogen et al. in a multicenter longitudinal observational study, in order to identify a high-risk group in PA–negative and ≤12 years of age children with CF, described SM as a risk factor for FEV1 decline [21]. A subsequent longitudinal retrospective study of 88 patients demonstrated that the acquisition of SM is associated with an acceleration in lung function decline. More interestingly, the effect persisted after controlling for confounders. In this study, chronic infection was defined as two or more positive cultures within a 12-month time period following acquisition, otherwise infection was classified as intermittent. Interestingly, both the intermittent and chronic subgroups were associated with lung function decline, and the change in rate of decline did not significantly differ between them. Chronic SM infection was also associated with an almost twofold increase in mean annual hospitalizations (p=0.007) [22].

In a recent retrospective study Poore et al. noticed an association between SM colonization and frequent fungal infection, especially Aspergillus (70% of fungal positive cultures in this cohort). Furthermore, they found that patients with SM and frequent fungal isolation had lower average lung function by almost 10% compared to controls [23].

We found several limitations in these studies, such as small cohorts of patients, type of study design (lack of prospective studies) and different clinical characteristics among patients included. Moreover, the definition of chronic colonization is based on different criteria among the studies, which makes them barely comparable. However, the most recent studies suggest a more active role of SM in influencing the progression of lung disease rather than simply being an indicator of disease severity. This can probably be explained by the fact that SM was considered a classical but infrequent bacterium in CF patients until the 2000's, but its incidence appears to be increasing in recent decades [24].

While the genetic adaptations and resulting phenotypic variations in PA and Staphylococcus aureus colonization of CF lungs are well-documented, the specific adaptive characteristics of SM that contribute to its persistence in CF patient only recently gained interest among many authors[25‐28]. Genetic studies have revealed significant genotypic diversity within SM chronically infected CF patients. Multiple strains of SM can coexist within an individual patient, suggesting ongoing acquisition and colonization events [26, 28]. Genotyping techniques, such as pulsed-field gel electrophoresis and multilocus sequence typing have provided insights into the clonal relatedness and genetic variation among different isolates. The phenotypic variability is observed in various aspects, including antibiotic resistance patterns, biofilm formation, and virulence factors.

In a study by Vidigal et al. genotypic diversity, mutation frequency, and antibiotic resistance were examined in 90 SM isolates from 19 CF patients with chronic colonization [25]. The findings revealed that SM undergoes significant genetic diversity during chronic CF lung infection, although a decreased mutation rate was observed in the later isolates. In a more focused investigation, Pompilio et al. 2016 evaluated 13 SM strains isolated from a single CF patient with chronic infection over a 10-year period [26]. They examined various traits including growth rate, biofilm formation, motility, mutation frequencies, antibiotic resistance, and pathogenicity. The results demonstrated that SM adaptation led to increased antibiotic resistance but decreased in vivo pathogenicity and biofilm formation. However, it is important to note this study's limitation of only considering one chronically infected patient. Interestingly, according to Esposito et al. and Alcaraz et al. the wide range of phenotypes exhibited by SM strains, only marginally correlates with the distribution of mutations across their genomes [27, 28].

These studies collectively emphasize the remarkable adaptability of SM during chronic infection in CF patients. This heterogeneity likely arises from the microorganism's need to adapt to a highly challenging CF lung environment, while facing diverse selection pressures based on the host's unique conditions. The mechanisms that drive the development of high genomic heterogeneity, resulting in a wide range of phenotypes, is still unclear and further studies are needed in order to better understand it [27]. Although the discussion of this topic is beyond the scope of our review, which is focused on clinical aspects of SM in CF, it will be important to clarify the mechanisms of development of genotypic and phenotypic heterogeneity, giving the possible impact on diagnosis, treatment, and infection control strategies.

A Cochrane Intervention Review by Amin et al. was conducted to assess the effectiveness of antibiotic treatment in people with CF, primarily in the setting of acute pulmonary exacerbations and then in chronic colonization of SM. However, there was no evidence since no randomized control trial met the inclusion criteria for the review [10].

The objective of administering antibiotics during a CF pulmonary exacerbation is twofold: to decrease the bacterial presence in the airways, potentially eliminating the bacteria altogether, and to reduce inflammation, consequently enhancing lung function and extending the period before another exacerbation occurs [44].

A retrospective cohort study showed that antibiotic therapy targeting SM during pulmonary exacerbations in patients with chronic SM infection did not affect the degree of FEV1 recovery or the time to subsequent exacerbation [35]. It is worth noting, however, that the majority of patients received treatment with a single antimicrobial drug targeting SM, resulting in successful elimination of SM from the airways in only a quarter of chronic SM pulmonary exacerbations. It is widely known that SM exhibits intrinsic resistance to a wide range of antimicrobial agents, and it is often recommended to employ a combination of antibiotics to effectively treat SM infections. Even the authors suggested that the antimicrobial monotherapy may not be sufficient.

SM is a multidrug-resistant opportunistic bacteria and can rapidly develop antimicrobial resistance mutations [45]. Trimethoprim-sulfamethoxazole has been historically considered the first line of treatment for SM infections due to high susceptibility rates and large clinical experience [45‐47]. According to Esposito et al. the most effective antibiotics against SM were minocycline, doxycycline and trimethoprim-sulfamethoxazole, showing comparable susceptibility rates [27]. Whereas San Gabriel et al. demonstrated that in vitro, SM appears to be most susceptible to trimethoprim-sulfamethoxazole, ticarcillin-clavulanate and doxycycline [29].

The Infectious Diseases Society of America (IDSA) provided guidance for SM infections management in CF and non-CF patients, consisting of "suggested approaches" based on clinical experience, expert opinion, and a review of the available literature [48]. In case of moderate to severe infections, and also considering the multiple mechanism of antibiotic resistance, they recommended a combination therapy. They suggested 3 approaches: (1) the use of combination therapy, with trimethoprim-sulfamethoxazole and minocycline as the preferred combination; (2) the initiation of trimethoprim-sulfamethoxazole monotherapy with the addition of a second agent (minocycline [preferred], tigecycline, levofloxacin, or cefiderocol) if there is a delay in clinical improvement with trimethoprim-sulfamethoxazole alone; (3) the combination of ceftazidime avibactam and aztreonam, when intolerance or inactivity of other agents are anticipated. For mild infections and polymicrobial infections where the role of SM is unclear, they suggested monotherapy with trimethoprim-sulfamethoxazole, levofloxacin, minocycline, tigecycline or cefiderocol. Table 4 reports the suggested dosages for antimicrobial therapy.

The multicenter randomized controlled clinical trial STOP2 evaluated the antimicrobial therapy duration during acute exacerbations in CF adults, regardless of the bacterial species involved. The outcome was similar for the 10 day, the 14 day and the 21 day regimens [31].

No suppressive therapy for CF patient with SM chronic infection is available, despite aerosolized levofloxacin as a potential future strategy [30].