Ac species are gram-negative, lactose nonfermenting, catalase- and oxidase-positive bacilli that are classified as aerobic organisms, although they may also thrive in anaerobic conditions [12]. They are widely distributed in moist environments and soil and are increasingly found in hospital settings where they can be diffused through contaminated fluids. A total of 21 Ac species have been identified [12]. The most frequent isolate in CF patients worldwide is A. xylosoxidans, followed by A. ruhlandii. Other species isolates from chronic and occasional lung infection in CF patients are A. insuavis,A. deleyi, A. denitrificans, A. insolitus, A. pestifer, A. spanius and A. marplatensis. However, as the differentiation of different species requires specific molecular-based methods, the true frequency of the various species in CF patients remains poorly defined [13]. Generally, Ac species isolated with conventional methods are reported as A. xylosoxidans [14, 15].

Similar to P. aeruginosa, Ac species possess a number of intrinsic characteristics that may explain both their long-term presence in the lung microbiome and their potential ability to damage lung structure and function. Their genome is highly dynamic, and hypermutation can favour adaptation of the pathogen to the lung environment and persistent colonization/infection [16]. Ac species possess a number of protein secretion systems that allow them to deliver lethal toxins into bacterial cells [17, 18]. They are resistant to natural antimicrobial peptides contained in airway secretions [19]. Finally, they exhibit significant motility and a great ability to adhere to respiratory cells and to form biofilms, all characteristics that are important determinants of persistence, reduced sensitivity to natural defences and antibiotic activity [20, 21].

Bacteria included in the Ac genus are generally multidrug resistant pathogens [22]. This is because they possess several intrinsic and acquired resistance mechanisms that frequently simultaneously work to make antibiotics currently used against gram-negative rods completely ineffective both in vitro and in vivo. Efflux mechanisms, beta-lactamase production and mutations in target proteins are antibiotic resistance determinants in Ac species [22]. Three different efflux mechanisms have been described. AxyABM is responsible for resistance to cephalosporins (except cefepime and cefuroxime), aztreonam, nalidixic acid, fluoroquinolones, and chloramphenicol [23]. AxyXY-Oprz makes aminoglycosides, cefepime, carbapenems, some fluoroquinolones, tetracyclines, and erythromycin ineffective [24, 25]. Finally, AxyEF-OprN extrudes some fluoroquinolones (levofloxacin and ciprofloxacin), tetracyclines (doxycycline and tigecycline) and carbapenems (ertapenem and imipenem) [26]. All the efflux mechanisms are intrinsic and detectable in most of the Ac species that are commonly encountered in CF patients. Only some strains that rarely colonize these subjects do not carry the AxyXY-Oprz efflux mechanism and remain sensitive to aminoglycosides. The production of beta-lactamases may be intrinsic (as in the case of OXA114, which makes piperacillin, ticarcillin, cephalothin and benzylpenicillin totally ineffective [27]) or acquired (such as extended-spectrum beta-lactamase [ESBL] and AmpC, which inactivate all beta-lactam antibiotics except carbapenems [28, 29, 30]). Intrinsic or acquired beta-lactamases can ultimately comprise plasmidic and chromosomal carbapenemases that generally hydrolyse all beta-lactams except aztreonam [28, 29, 30]. However, strains carrying VIM-2 beta-lactamase are also resistant to aztreonam [31, 32, 33, 34].

Due to the different distributions of resistance mechanisms among Ac species and within the same species, minimum inhibitory concentration (MIC) breakpoints for these pathogens have not been definitively established. Only recently have clinical minimal inhibitory concentration breakpoints for several antibiotics been proposed by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) [35]. Moreover, sensitivity to antibiotics can significantly vary from strain to strain and seems to gradually decrease, probably because of within-host bacterial genome evolution during chronic colonization [36]. This can explain why studies carried out in different countries and in different periods of time have reported different results. The global evaluation of nine case series published from 2003 to 2014 showed that the most effective in vitro antibiotics were ticarcillin (99.5%), cefoperazone-sulbactam (98.7%), piperacillin-tazobactam (97.2%), and imipenem (86.4%), whereas aztreonam and tetracyclines displayed poor sensitivity (1.5% and 17.5%, respectively) [37]. In a study carried out in Argentina in which 59 strains of Ac species were collected from clinical specimens of patients, similar to what had been previously reported by other authors [38, 39], the most active antibiotics were piperacillin alone or with tazobactam and carbapenems, with meropenem significantly more effective than imipenem and ertapenem [40]. Co-trimoxazole, minocycline, and colistin were at least partially active. Among cephalosporins, ceftazidime was more effective than cefepime [40]. In UK, in a sample of 112 Ac spp. from CF patients, piperacillin-tazobactam (70.2%) and cotrimoxazole (69.7%) were the most active antibiotics [41]. Finally, in France, colistin was found to be effective in 19 out of 22 Ac strains collected from CF patients [42].

Combinations of antibiotics were found to exert a synergistic interaction that in some cases significantly increased the bactericidal activity of the single agent, providing evidence for potentially effective in vivo therapies. A very good example in this regard has been reported by Daman-Çelik et al. [43]. These authors tested meropenem alone or in combination with colistin, levofloxacin, and chloramphenicol and found that when meropenem was combined with colistin, the combination was effective against bacteria susceptible to meropenem and colistin but also against colistin resistance. In contrast, the meropenem-levofloxacin combination had a synergistic but not bactericidal effect, whereas the meropenem-chloramphenicol combination was neither synergistic nor bactericidal [43].

In the last 20 years, Ac species have been detected with increased frequency in the respiratory secretions of CF patients [44]. Several factors could explain the emergence of these pathogens [45, 45]. The use of aggressive antibiotic therapy in an attempt to eliminate typical CF pathogens, such as S. aureus and P. aeruginosa, may have strongly induced bacterial selection. Evidence of Ac species may also have been favoured by the introduction in clinical practice of advanced bacterial detection methods, more careful CF patient follow-up, and prolongation of the average lifespan. Finally, it cannot be excluded that the increased detection may be ascribed to some Ac species characteristics, mainly constitutive resistance to many antibiotics and the ability to adapt to surrounding pressures by means of within-host genome evolution, all factors that may favour persistence in respiratory secretions [44, 45, 46].

Despite detection in almost all studies, the incidence of Ac species identification in CF patients varies significantly from study to study, ranging from 3% to 30% [47, 48, 49]. The criteria used to define colonization (sporadic, intermittent, or chronic) and the time of evaluation can explain the differences. The lowest values were associated with chronic colonization and with collection of respiratory samples in the first years of this century [50]. A study carried out in France clearly showed that the detection of Ac species in respiratory secretions of CF patients increased from 1999 to 2018 from 3.1% to 6.7% [51]. Similar data have been collected in the USA [52].

The clinical relevance of Ac species in CF patients remains debated. Several retrospective studies have shown that CF patients with severe disease are more frequently infected by A. xylosoxidans than are patients with less severe signs and symptoms of lung involvement [53, 54, 55]. Worse lung function, more frequent pulmonary exacerbations and the need for hospitalization and antibiotic treatment were more common among A. xylosoxidans-infected patients than among noninfected controls [53, 54, 55]. This explains why the detection of this pathogen in a CF patient is considered a marker of severe CF. However, as few studies for species other than A. xylosoxidans exist, the greater clinical relevance of this pathogen is not definitively established. On the other hand, the role of Ac species infection as a cause of primary deterioration of lung function is far from definitively ascertained. Support for this hypothesis is given by the evidence that Ac species can cause a significant inflammatory status similar to that caused by P. aeruginosa and are potentially able to cause progressive lung damage per se. Hansen et al. measured cytokine levels in the serum and sputum of 11 healthy controls and 60 CF patients, 11 with A. xylosoxidans, 11 with the B. cepacia complex, 21 with P. aeruginosa and 17 without infection with these pathogens [56]. They found that all of the chronically infected patients had significantly higher serum levels of interferon (IFN)-$\gamma$ and interleukin (IL)-6 than noninfected CF patients. However, only A. xylosoxidans and P. aeruginosa patients had significantly higher sputum tumour necrosis factor (TNF)-$\alpha$ and serum granulocyte colony stimulating factor (G-CSF) levels, suggesting that these bacteria could cause greater lung damage.

Clinical findings do not definitively solve the problem of the causative role of Ac species. Indeed, data collected in several studies have not answered this question. In one investigation, respiratory function and exacerbation frequency were compared in CF patients with at least one sputum culture positive for A. xylosoxidans and in control uninfected patients [57]. Data collected between 1 year prior to and 3 years after A. xylosoxidans isolation were considered. Compared to negative patients, positive subjects showed a greater decline in forced expiratory volume in 1 second (FEV1) in the first year (–153.6 $\pm$ 16.1 mL$\cdot$year${}^{-1}$ vs. –63.8 $\pm$ 18.5 mL$\cdot$year${}^{-1}$; p = 0.0003), together with more exacerbations in the first 3 years after pathogen detection (9 vs. 7; p = 0.03). However, these findings do not definitively demonstrate a causal relationship between the presence of the pathogen and the worsening of pulmonary disease because the subjects with A. xylosoxidans had, at the beginning of the study, a worse clinical situation and greater FEV1 decline, and more exacerbations were mainly evidenced in A. xylosoxidans cases co-colonized with P aeruginosa (75%) [56]. Similar inconclusive findings were reported by Edwards et al. [58], Firmida et al. [59], Somayaji et al. [55], and Marsac et al. [60]. The first authors found that if CF patients were more likely to experience pulmonary exacerbation with incident Ac species-positive cultures (42% vs. 21%; odds ratio [OR], 2.7; 95% confidence interval [CI], 1.1–6.7; p = 0.03), persistent infection was associated with neither annual lung function decline (–1.08% [95% CI, –2.73 to 0.57%] vs. –2.74% [95% CI, –4.02 to 1.46%]; p = 0.12) nor the risk of pulmonary exacerbations (OR, 1.21 [95% CI, 0.45 to 3.28]; p = 0.70) [55]. In a case-control retrospective study, the clinical course of a group of chronically or intermittently A. xylosoxidans-colonized/infected CF patients was compared with that of never colonized/infected subjects during two periods that were 2 years apart [61]. No differences in lung function among groups over 2 years were evidenced, although in the chronically colonized/infected subjects, a trend towards a greater decrease in lung function was observed (51.7% in the chronic colonization/infection group vs. 82.7% in the intermittent colonization/infection group vs. 76% in the never colonized/infected group) [55]. Somayaji et al. [55] studied 88 patients who had one or more cultures positive for Ac species during the course of 18 years. They found that pulmonary exacerbations and risk of death or transplantation were more common in these subjects than in uninfected CF patients. However, further evaluations did not show any independent association between chronic Ac species infection and worsening of the risk of pulmonary exacerbation, lung function deterioration or the time lag for death or lung transplantation. Finally, a recent case control study by Marsac et al. carried out in two French paediatric centres compared 45 patients infected by A. xylosoxidans with the same number of never infected controls matched for age, sex, pancreatic status and genotype [60]. Clinical data collected in the two years immediately preceding and following the first identification of the pathogen were evaluated. CF severity was significantly greater in A. xylosoxidans patients than in controls both before and after pathogen detection. Pulmonary exacerbations, hospitalization, and the need for antibiotic courses were significantly more frequent in positive than in negative patients. Moreover, lung function decline tended to be faster in cases (–5.5% vs. –0.5% per year). However, even in this study, the greater colonization with Pseudomonas aeruginosa in A. xylosoxidans-positive patients (p = 0.0002) makes these findings difficult to interpret [60].