**5. Discussion**

NTM are increasingly identified in respiratory specimens all over the world. In the US the estimated number of NTM pulmonary cases increased two-fold between 2010 and 2014 [8]. The growing number of NTM isolates is probably caused by the aging of populations and the increasing number of patients with chronic lung diseases predisposing to NTM colonization [9].

In CF patients, the estimated risk of NTM pulmonary infection is approximately 1000 times higher compared to the healthy population [1].

At our center, NTM was isolated in 7% of 151 hospitalized CF patients, during 10 years of follow-up. The prevalence of positive NTM isolations in other European CF centers was 4–10% [10–12]. A lower prevalence of NTM infections was noted among patients less than 16 years of age (1.3% in 2010 and 3.8% in 2015) [13]. In the US CF centers, the annual prevalence of positive NTM isolates was 2–14% and the frequency of positive isolations in CF increased from 11% per year in 2010 to 13.4% in 2014 [6,14].

The principles of NTMLD recognition have been summarized recently [15]. The diagnosis is based on clinical, radiological and microbiological data (Table 2).



Nevertheless, the problem of NTMLD diagnosis in CF is complex. Most patients present the symptoms of disease exacerbation, such as increased expectoration of purulent sputum, dyspnoea, loss of appetite and or increased body temperature. These symptoms are common for all types of infective exacerbations, irrespective of the type of cultured microorganisms.

Chest CT plays an important role in the diagnostic algorithm of both cystic fibrosis and NTMLD. In CF it is used to assess the extent of the disease, complications and response to the implemented treatment [16,17].

CF typically presents on chest CT with bilateral bronchiectasis that dominates in the upper lung lobes, bronchial wall thickening, mucus plugging, emphysema, air trapping, atelectasis, acinar nodules, thickening of interlobular and intralobular septa as well as areas of ground glass opacities [16,18]. It was recommended to perform chest CT biennially in stable CF patients [17].

Bronchiectasis is increasing in number and in size along with disease progression, due to chronic inflammation. Mucus plugging may result in partial or total obstruction

of the bronchial lumen. At the late stage of the disease, partial lung cirrhosis is described. Reactive lymphadenopathy is often observed.

NTMLD presents with two types of radiological changes: nodular-bronchiectatic and cavitary [15,19].

The bronchiectatic form was most often diagnosed in non-CF, in middle-aged females with no predisposing factors [20]. Chest CT shows bronchiectasis localized in the middle lobe and lingua, as well as small intra-lobular, poorly defined nodules, which are signs of bronchiolitis.

The cavitary form of NTMLD was mostly observed in older, white males with underlying chronic pulmonary disease [20]. Small, thin-walled cavities are usually localized in the upper lobes. On such occasions, tuberculosis has to be taken into consideration in the differential diagnosis.

The recognition of radiological features of NTMLD in CF patients is extremely difficult. The nodular-bronchiectatic form of NTMLD is practically indistinguishable from the radiologic appearance of CF. Occasionally, new nodular opacities localized in the middle lobe and lingua may be indicative of NTM infection, as was shown in two of our patients.

The cavitary form of NTMLD is easier to recognize in CF patients, as lung infiltrations with cavitation are not typically seen in the course of bacterial infections in CF. Serial analysis of chest CT in a single CF patient may reveal new nodular opacities, not resolving in the course of anti-bacterial therapy. In some of them, cavitation appears in the course of the disease. Such presentation was described in three of our patients. All of the patients with lung cavitation fulfilled the diagnostic criteria of NTMLD. Based on our experience, cavitary lesions on chest CT were the significant predictors of NTMLD not only in CF but also in COPD patients [21].

Cavitary form of NTMLD in CF patients is combined with a worsening in prognosis. Abate et al. found cavitary lesions in 30% of CF patients, this radiological presentation was combined with a three-fold increase in death risk [8].

In the majority of our CF patients with positive NTM isolates, chest CT was not indicative of NTMLD. Therefore, it is of extreme importance to perform periodic sputum cultures for mycobacteria in CF patients. Nick et al. recommended annual NTM screening in CF [6]. At our center, cultures for mycobacteria are performed in every CF patient at least once a year. Bacterial decontamination of sputum samples referred for mycobacterial diagnostics is mandatory. Serial mycobacterial cultures are the principle of screening for NTMLD in CF patients.

Respiratory samples analysis, performed at our center, revealed the presence of NTM in 7% of CF patients. *M. avium* was the dominating type, identified in 55% of patients. *M. avium*/MAC was also the most frequent isolates in the US populations of CF patients [8], whereas *M. abscessus*—in European CF patients [10,11,22].

Increasing the isolation rate of *M. abscessus* in CF centers may indicate in-hospital transmission between the patients or environmental source of acquisition [4,5]. Recent molecular epidemiologic investigations (whole genome sequencing g-WGS) indicate that health-care related transmission of *M. abscessus* is rare [22,23]. Nevertheless, Hassan et al. found that 15% of clustering of MAC isolates concerned patients sharing CF centers, thus indicating a common source of infection [24].

In 2 of our patients (18%), *M. chimaera* was identified. The clinical course of the disease was fulminant in one patient. High virulence of *M. chimaera* was also confirmed by Cohen-Bacrie et al. in an 11-year-old boy with CF [25]. Larcher et al. reported four cases of *M. chimaera* NTMLD in CF, clinical improvement and stabilization of spirometry values were observed in three of them, in the course of treatment, and marked worsening of spirometry in one, that was not treated [12].

In all of our study groups, respiratory specimens were also positive for other bacteria, cultured at the time of NTM isolation. The bacterial cultures of sputum revealed *P. aeruginosa* in 63% of patients and *S. aureus*, in 52%. Such bacterial species are typical for the adult population of CF patients, and they are cultured with increasing frequency at the time of

advanced lung disease. In the previous publication from our CF center, concerning 89 CF patients observed between 2008 and 2011, *P. aeruginosa* was cultured in 55.6% of specimens and *S. aureus*—in 37.8% [26].

Fungal infections are diagnosed frequently in CF patients, due to the presence of thick mucus and impaired mechanism of bronchial clearance, the mechanisms that prolong exposition to inhaled spores. The common use of broad-spectrum antibiotics is regarded as a risk factor for yeas<sup>t</sup> infection [27]. A previous study conducted in our center, which concerned 217 respiratory specimens obtained from 42 CF patients, revealed 205 (68%) strains of yeas<sup>t</sup> (mainly *C. albicans*) and 96 (32%) strains of filamentous fungi (mainly *A. fumigatus*) [28].

The most interesting microbiological phenomenon in the patients with CF and in non-CF bronchiectasis is the frequent co-habitation of NTM and fungi. The results of a recently published meta-analysis indicated that the risk of NTM isolation in adult CF patients was increased by 2.75 times in those with *A. fumigatus* colonization, compared to the noncolonized group [29]. It is not clear whether the increased risk of NTM isolation in patients colonized with *A. fumigatus,* depends on direct interactions between both microorganisms. *A. fumigatus* colonization favors T-helper 2 CD4+ T cell response, downregulating the cytokines responsible for NTM eradication [29]. It is also possible that the cohabitation of both pathogens concerns patients with profound lung structural disturbances [10,29]. Additionally, some CF patients colonized with *A. fumigatus*, are recognized with ABPA and treated with oral steroids. ABPA was listed as another possible risk factor for NTM isolation [10,13,30]. In the present study group, 46% of the patients were diagnosed with ABPA, prior to NTM isolation.

The other fungal pathogens frequently isolated in CF patients are members of *Candida* spp. Cuthbertson et al. reported that, based on PCR analysis of the fungal microbiome, *Candida* spp. was found more frequently in CF patients compared to *A. fumigatus*, and that it was one of the risk factors of NTM infection [31]. In the present study, *Candida* spp. was cultured simultaneously with NTM in 82% of patients.

In summary, the diagnosis of NTMLD in CF is very difficult. Most patients present with new clinical signs of infection, but they improve with standard antibacterial therapy. Hemoptysis, which may indicate NTMLD, is frequently observed in CF patients without NTMLD. Weight loss listed as one of the signs of NTM infection can be present in CF due to PEX or malabsorption syndrome. Thus, there are no specific complaints that could indicate NTM infection in CF patients. The early radiologic appearance may be non-specific, as discussed already. Therefore, the role of frequent sputum cultures for NTM and comparison of previous chest CT examinations are of extreme importance.

The treatment of NTMLD in CF patients is complicated, due to many drug interactions and the length of therapy. According to recent publications, CF treatment with CFTR modifiers reduces the frequency of PEX [32]. Preliminary studies indicate also the possibility of a lower risk of NTM infections in patients receiving CFTR modifiers [33].

**Author Contributions:** D.W. writing–original draft, L.O. radiological scans preparation, and description; D.F. data collection E.J., data collection, patient's medical care W.S., data collection, patient's medical care E.A.-K. supervision and M.S. writing—review and editing, supervision All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The project was accepted by the Institutional Ethical Committee (No. 9/2015).

**Informed Consent Statement:** Patient consent was waived due to retrospective character of published data.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.
