Azole-Resistant Aspergillus fumigatus Among Danish Cystic Fibrosis Patients: Increasing Prevalence and Dominance of TR34/L98H

Azole-resistant (azole-R) Aspergillus is an increasing challenge worldwide. Patients with cystic fibrosis (CF) are at risk of Aspergillus colonization and disease due to a favorable lung environment for microorganisms. We performed a nationwide study in 2018 of azole-non-susceptible Aspergillus in CF patients and compared with data from two prior studies. All airway samples with mold isolates from patients monitored at the two CF centers in Denmark (RH, Jan–Sept and AUH, Jan–Jun) were included. Classical species identification (morphology and thermo-tolerance) was performed and MALDI-TOF/β-tubulin sequencing was performed if needed. Susceptibility was determined using EUCAST E.Def 10.1, and E.Def 9.3.2. cyp51A sequencing and STRAf genotyping were performed for azole-non-susceptible isolates and relevant sequential isolates. In total, 340 mold isolates from 159 CF patients were obtained. The most frequent species were Aspergillus fumigatus (266/340, 78.2%) and Aspergillus terreus (26/340, 7.6%). Azole-R A. fumigatus was cultured from 7.3% (10/137) of patients, including 9.5% (9/95) of patients at RH and 2.4% at AUH (1/42), respectively. In a 10-year perspective, azole-non-susceptibility increased numerically among patients at RH (10.5% in 2018 vs 4.5% in 2007–2009). Cyp51A resistance mechanisms were found in nine azole-R A. fumigatus from eight CF patients. Five were of environmental origin (TR34/L98H), three were human medicine-driven (two M220K and one M220R), and one was novel (TR343/L98H) and found in a patient who also harbored a TR34/L98H isolate. STRAf genotyping identified 27 unique genotypes among 45 isolates and ≥2 genotypes in 8 of 12 patients. This included one patient carrying two unique TR34/L98H isolates, a rare phenomenon. Genotyping of sequential TR343/L98H and TR34/L98H isolates from the same patient showed only minor differences in 1/9 markers. Finally, azole-R A. terreus was found in three patients including two with Cyp51A alterations (M217I and G51A, respectively). Azole-R A. fumigatus is increasing among CF patients in Denmark with the environmentally associated resistance TR34/L98H mechanism being dominant. Mixed infections (wildtype/non-wildtype and several non-wildtypes) and a case of potential additional tandem repeat acquisition in vivo were found. However, similar genotypes were identified from another patient (and outside this study), potentially suggesting a predominant TR34/L98H clone in DK. These findings suggest an increasing prevalence and complexity of azole resistance in A. fumigatus.


INTRODUCTION
Azole resistance (Azole-R) in Aspergillus fumigatus is an increasing problem and complicates patient management (Lestrade et al., 2019). Azole-R A. fumigatus in patients with cystic fibrosis (CF) has been reported in several studies (Mortensen et al., 2011a;Burgel et al., 2012;Morio et al., 2012;Bader et al., 2013;Fischer et al., 2014;Stevens et al., 2016;Prigitano et al., 2017;Abdolrasouli et al., 2018;Guegan et al., 2018;Güngör et al., 2018;Seufert et al., 2018;Engel et al., 2019;Lavergne et al., 2019; Table 1). Danish CF patients are followed up monthly at the two specialized CF clinics at Copenhagen University Hospital, Rigshospitalet (RH) and at Aarhus University Hospital (AUH). From the majority of these patients, airway samples are obtained monthly. We have previously studied azole-R A. fumigatus in the Copenhagen CF cohort (Mortensen et al., 2011a,b). The first study included isolates from Jan to March (Q1) 2007 (Mortensen et al., 2011b) and the second study isolates from July to Dec (Q3-4) of 2007 and of 2009 (Mortensen et al., 2011a). These studies documented an overall azole-nonsusceptibility rate of 1.6 and of 4.5%, respectively. A lower but increasing rate from 1.8 to 3.8% of azole-R A. fumigatus among clinical samples (from CF as well as non-CF patients in Denmark) was found in a subsequent reference laboratorybased study, from the years 2010 to 2014 (Jensen et al., 2016). But the epidemiology of azole-R A. fumigatus including that specifically associated with environmental origin has not been systematically studied over a longer time in our country. Due to the increasing international and political concern related to the link between environmental azole fungicide use and azole resistance in A. fumigatus, we systematically investigated the azole resistance rate in the Danish CF population in 2018 and compared it to our previous data (Mortensen et al., 2011a).
CF is the most common autosomal recessive disease in Caucasians (Felton and Simmonds, 2014). Mutations in the CFTR (cystic fibrosis transmembrane regulator gene) affects the chloride transportation causing dysregulated fluid transport in the epithelial cells of multiple organs (Felton and Simmonds, 2014). Clinically, CF disease is dominated by infectious pulmonary complications (Felton and Simmonds, 2014). The respiratory tract is often colonized with molds especially A. fumigatus, which is found in 16 to 56.7% of airway samples (Pihet et al., 2009). Aspergillus may cause a diversity of manifestations ranging from asymptomatic colonization to serological sensitization, allergic bronchopulmonary aspergillosis (ABPA), Aspergillus bronchitis, and aspergilloma in CF patients (Felton and Simmonds, 2014). The most common is ABPA (Pihet et al., 2009), which occurs in approximately 10% of CF patients (Burgel et al., 2016;Carsin et al., 2017) and is the cause of hypersensitivity response to Aspergillus antigens (Williams et al., 2016). Azoles are the cornerstone in the management of CF patients with Aspergillus disease requiring antifungal therapy. Itraconazole is the first choice as an antifungal drug in the treatment of ABPA to reduce the burden of A. fumigatus and minimize use of corticosteroids . Posaconazole is used as salvage therapy in ABPA or bronchitis (Skov et al., 2017;Periselneris et al., 2019), whereas voriconazole or isavuconazole are first-line agents (Maertens et al., 2016;Patterson et al., 2016) in the rare event of invasive aspergillosis (Burgel et al., 2016;Skov et al., 2017;Hamprecht et al., 2018).
Azoles target and inhibit the lanosterol 14-α-demethylase enzyme (Cyp51A) encoded by the cyp51A gene and thereby inhibit the ergosterol synthesis (Stensvold et al., 2012). Patients with recurrent or long-term need for azole therapy are at risk for azole-R Aspergillus due to selection of resistance during exposure to medical azoles . Azole-R in A. fumigatus also occurs in patients with no prior azole therapy, caused by the inhalation of resistant mutant spores from the environment presumably selected due to azole fungicide use for plant and material protection (Astvad et al., 2014;Hagiwara et al., 2016). Well-known mechanisms behind azole resistance are target gene mutations in cyp51A. Two common resistance mechanisms, TR 34 /L98H and TR 46 /Y121F/T289A, are considered to be of environmental origin (Stensvold et al., 2012). These "environmental" mechanisms have previously been found in the Danish environment (Mortensen et al., 2010;Risum et al., 2019) and in clinical samples (Mortensen et al., 2011a;Astvad et al., 2014). Furthermore, target gene upregulation, efflux, and HapE  and Hmg1 (Rybak et al., 2019) alterations have been documented as underlying mechanisms of azole resistance in selected isolates.
In this study, we investigated the azole-R rate in a 10y perspective and dissected underlying molecular resistance mechanisms and genotypes in Aspergillus in CF patients followed up at the two Danish CF centers that serve the entire country.

MATERIALS AND METHODS
The two CF centers RH and AUH follow all the Danish CF patients. A total of 522 (320 and 202 adult and children) with CF were followed up at RH and AUH, respectively, in 2018. All positive cultures with mold from airway samples from the Danish CF population were included during a 6-month (Jan-June 2018) and a 9-month (Jan-Sept 2018) period, respectively, from AUH and RH. Primary culture was performed using Sabouraud glucose agar [SSI Diagnostika, Hillerød, Denmark (RH) and bioMérieux, Marcy l'Etoile, France (AUH), respectively]. Agar plates were incubated at 35-37 • C and examined for 5 days (RH) and 2 days (AUH). Exclusion criteria were identical to our previous study (Mortensen et al., 2011a). In detail, repeat isolates from the same patient were excluded when found ≤30 days apart and confirmed as same species and with same susceptibility classification. Identification was done to the Aspergillus species complex level using classical techniques, including thermo-tolerance test for A. fumigatus specifically, followed up by MALDI-TOF applying the Mass Spectrometry Identification database (Normand et al., 2017;Imbert et al., 2019) and β-tubulin sequencing (Glass and Donaldson, 1995) when necessary.
The EUCAST E.Def 10.1 method (Arendrup et al., 2017) was used for azole-R screening for A. fumigatus, and EUCAST E.Def 9.3.1 susceptibility testing  was performed for amphotericin B for the majority of the isolates and for itraconazole, posaconazole, and voriconazole for screening-positive A. fumigatus isolates and Aspergillus species other than A. fumigatus. Isolates with azole MIC(s) above the ECOFF(s) underwent cyp51A sequencing as previously described (Mortensen et al., 2011a). EUCAST clinical breakpoints v 9.0 were adopted for susceptibility classification into susceptible, non-susceptible (intermediate and resistant), and azole-R (Arendrup et al., 2013). For species and agents without breakpoints, EUCAST ECOFFs were adopted and non-wildtype isolates were regarded resistant. Sequential isolates from all patients harboring resistant A. fumigatus underwent STRAf genotyping (De Valk et al., 2005).
Results were compared to our previous Danish studies on azole-R in CF patients followed up at RH allowing a 10-year perspective (Mortensen et al., 2011a). Comparison of groups was performed with a contingency chi-square test using GraphPad Prism version 8.0.2.
(Preliminary results from RH have been presented at the European Congress on Clinical Microbiology and Infectious Diseases in 2019).

RESULTS
In total, 340 unique mold isolates from 159 CF patients were obtained, of which 240 isolates were derived from 110 CF patients at RH (2.2 isolates per patient) and 100 isolates from 49 CF patients at AUH (2.0 isolates per patient). The median age was 30 years (6-68 years) at RH and 22 years (6-50 years) at AUH among patients with a mold isolate.
STRAf genotyping identified 27 unique STRAf genotypes among the 45 A. fumigatus isolates from the 12 patients with azole non-susceptible A. fumigatus (Table 3). Eight patients harbored isolates with more than one genotype. Three of these patients, carried isolates that differed only in a single marker (RH-5, RH-7, and RH-8), and six patients carried isolates that were clearly unrelated (including two patients with both related and unrelated genotypes RH-7). Thus, patient RH-5 had six isolates with 8/9 identical STRAf markers, while marker 3A ranged from 95 repeats (TR 34 /L98H) to 96-101 repeats (five TR 34 3 /L98H) ( Table 3). In contrast, Patient AUH-2 harbored two TR 34 /L98H isolates but with clearly different STRAf profiles. Patient RH-2 harbored 10 A. fumigatus with three different genotypes, while patient RH-8 had four isolates with 8/9 identical markers and a fifth isolate with 7/9 identical markers. Isolates that shared 8-9 markers were also found across patients. Thus, patient RH-7 had two TR 34 /L98H isolates with identical STRAf profiles as two TR 34 3 /L98H isolates from RH-5. Azole resistance was also detected in other Aspergillus species. At AUH, such isolates were found in 3/49 (6.1%) patients including two patients with A. thermomutatus, and one patient with a voriconazole-resistant A. terreus isolate with a wild-type cyp51A ( Table 2). At RH, two out of eight CF patients with A. terreus (1.8% of CF patients at RH) had nonsusceptible isolates. One patient (RH-11) had both resistant and intermediate isolates recovered, which had Cyp51A amino acid substitutions M217I and Y491H, respectively. The other patient had a resistant A. terreus with a G51A mutation (RH-12, Table 2). MIC distributions for all Aspergillus isolates with reduced azole susceptibility at the two centers are shown in Supplementary Table 1.

DISCUSSION
We report detailed and nationwide data on azole nonsusceptibility and mold species distribution in respiratory isolates from Danish CF patients. Azole-R A. fumigatus with environmental origin was dominating and found at both centers suggesting a wide geographic distribution of TR 34 /L98H in Denmark. However, although the proportions of patients with A. fumigatus (85.7 and 86.3%, respectively) were similar at the two centers, an almost four-fold higher rate of azoleresistant A. fumigatus was observed at RH compared to AUH. Moreover, the resistance pattern was more diverse at RH and included both the resistance deriving from the environment and resistance mutations associated with azole treatment selection. Unfortunately, data on azole use in these patients could not be retrieved. However, it is most likely that the differences in azole-R between the two centers may be related to different prescription practices with a more extensive and longer duration of azole treatment at RH, in part due to an overall higher age and number of patients with chronic aspergillus bronchitis at RH. This is supported by the observation that human-driven target gene mutations were more common in patients at RH than at AUH.
At RH, cyp51A mutations of environmental origin accounted for half of the detected resistance, and the TR 34 /L98H rate has doubled over the past decade since the first detection in Q3-4 of 2007-9 ( Mortensen et al., 2011a). Additionally, a subsequent laboratory study of A. fumigatus isolates received at the national reference center reported an increase during 2010-2014 (Jensen et al., 2016). Taken together, these studies suggest that TR 34 /L98H has gradually become more prevalent in Denmark since 2007 despite the fact that two of three environmental sampling studies in Denmark failed to detect TR 34 /L98H and TR 46 /Y121F/T289A in soil and air samples (Astvad et al., 2014;Jensen et al., 2016). This suggests either significant fluctuations in the number of resistant spores in the environment or that even low levels of resistant A. fumigatus can contribute to resistant infections in a predisposed lung environment. Our observations of alternating or mixed resistant and susceptible isolates recovered from the same patient is a wellknown phenomenon and highlight that a single sample may not be a representation for the entire lung flora (van Leer-Buter et al., 2007;Astvad et al., 2014). Not only may different phenotypes dominate in different lung sections but mixed A. fumigatus strains are also very challenging to identify and separate unless molecular analyses are performed. Of note, two TR 34 /L98H isolates with different and unique STRAf genotypes among our collection were recovered from patient AUH-2, a case which we have not previously seen in DK.
Of particular interest, patient RH-5 harbored five pan-azoleresistant isolates with a novel TR 34 3 /L98H resistance mechanism, which to our knowledge has not previously been found in clinical specimens. However, exposure in vitro of A. fumigatus conidia already containing a 34-bp insertion in the cyp51A-gene promoter to 8 mg/L of tebuconazole resulted in one clone with a 34-bp triplicate repeat . In addition, the TR 46 /Y121F/T289A has also been found with additional 46-bp repeats in the promoter region in compost as well during sexual mating in in vitro studies (Zhang et al., 2017). The question remains whether this TR 34 3 /L98H resistance variant is novel in the environment and thus acquired de novo, as suggested by being isolated first, or whether the TR 34 /L98H was in fact first (but undiscovered in the first three specimens) and the additional TR 34 repeat acquired in vivo. The STRAf profiles suggest that the five TR 34 3 /L98H isolates are isogenic with a classical example of microevolution. It is noteworthy that the TR 34 /L98H isolate from this patient shared 8/9 markers and had 95 repeats at marker 3A while an increasing number of repeats (96-101) were seen in the TR 34 3 /L98H isolates over time. Increasing repeat numbers have previously been found over time in vitro and in vivo (Mortensen et al., 2011a;De Groot and Meis, 2019). Furthermore, the TR 34 /L98H strain was discovered in a sample mixed with the TR 34 3 /L98H strain and had a white and slow-growing phenotype, which could help explain why it could potentially have been overlooked in earlier samples. Indeed, the in vivo acquisition of a tandem repeat in the promotor region has been reported from our group, where a 120-basepair tandem repeat evolved in a patient during azole therapy, supported by whole-genome sequencing (WGS) . It has also been suggested that the TR 34 helps compensate for loss of fitness associated with the L98H change (Verweij et al., 2016) and thus the additional TR 34 could potentially further improve fitness and outgrow the TR 34 /L98H, which in this patient appeared with a weaker phenotype (Verweij et al., 2016). A third hypothesis could be that this is a random coincidence of similar STRAf genotypes. Indeed, the finding of two TR 34 /L98H isolates from another patient (RH-7) displaying identical STRAf profiles as two TR 34 3 /L98H isolates was surprising and further complicates the interpretation. One concern would be lab contamination, but since the two RH-7 isolates were received months apart and with different cyp51A profiles, this seems unlikely. A final, and worrying, theory is that we may have encountered a dominating TR 34 /L98H clone in DK similar to the study from India (Chowdhary et al., 2012). Indeed, outside the study period we have encountered a total of 20 isolates from 12 different patients from all around DK and also in two air samples sharing the same 8/9 STRAf markers, exclusively differing in marker 3A, ranging from 35 to more than 130 repeats. The high variation in 3A (in our two patients) indicates a highly mutagenic strain type, which may help explain the rare development of the TR 34 3 /L98H variant. Further studies including WGS are desirable to further explore the origin of this novel resistance mechanism as well as the potentially novel dominating genotype.
Two CF patients had resistant A. fumigatus with wild-type cyp51A, as reported in other CF studies (Mortensen et al., 2011a;Burgel et al., 2012;Guegan et al., 2018;Seufert et al., 2018;Lavergne et al., 2019) at similar rates (Mortensen et al., 2011a;Burgel et al., 2012;Lavergne et al., 2019) as well as in patients with chronic pulmonary aspergillosis (Howard et al., 2013). Phenotypic susceptibility testing therefore remains crucial because molecular detection of resistance mechanism enables the detection of resistance, but not susceptibility. Moreover, alternating findings of susceptible and non-susceptible isolates in the same patient demonstrate the need of repeated sampling and susceptibility testing of several colonies when present in patients requiring azole therapy, as recommended in current guidelines (Ullmann et al., 2018;Guinea et al., 2019).
Azole-R A. terreus constitutes a significant challenge since A. terreus has intrinsic reduced susceptibility to amphotericin B, rendering it multidrug-resistant Zoran et al., 2018;Rivero-Menendez et al., 2019). The present finding of A. terreus of 6.3% (10/159) nationally at patient level is quite high compared to previous studies (Mortensen et al., 2011a;Fischer et al., 2014;Engel et al., 2019), reporting 1.9% (Mortensen et al., 2011a) and 2.4% A. terreus at the isolate level (Fischer et al., 2014) and 3.9% of all Aspergillus spp. (Engel et al., 2019). Whereas M217I has been reported previously (Rivero-Menendez et al., 2019), G51A is, to our knowledge, novel. We also detected A. thermomutatus, another inherently voriconazoleresistant Aspergillus spp., at AUH. This species has been detected in one CF patient at RH previously (Mortensen et al., 2011a) illustrating that resistant Aspergillus infection is not limited to A. fumigatus in this setting.
When we compare the present study's result to others, the current overall azole-R rate of 7.3% in the total CF population corresponded well with CF studies from other European countries [France, Germany, the Netherlands (Bader et al., 2013;Seufert et al., 2018;Engel et al., 2019;Lavergne et al., 2019), and the United States (Table 1)] (Stevens et al., 2016). Internationally, published azole resistance rates have varied greatly. No azole resistance was reported from a Portuguese center (Amorim et al., 2010), in one of two Italian centers (Prigitano et al., 2017), and in five out of 12 German centers (Seufert et al., 2018; Table 3). In contrast, Abdolrasouli et al. found a concerning high resistance rate of 16.2% in CF patients specifically, which could be reflected upon the patient group at a cardiothoracic center in United Kingdom following up CF patients (Abdolrasouli et al., 2018). Guegan et al. also found a high azole-R rate of 15.2%, but in a limited CF population of 33 patients (Guegan et al., 2018; Table 1).
The major strength of the present study is the fact that it allowed a 10-year perspective on azole-resistant Aspergillus and a nationwide surveillance perspective of the current mold epidemiology in CF patients in Denmark. Since we also included Aspergillus species other than A. fumigatus, we also reported mutations in A. terreus and furthermore detailed information on STRAf genotyping in one patient with A. fumigatus. A limitation, however, is that we do not have information regarding the clinical relevance of the retrieved A. fumigatus, nor do we have any information regarding preceding antifungal use.
In conclusion, azole-R Aspergillus is increasing in proportion and complexity among Danish CF patients. The larger and increasing proportion involved resistant A. fumigatus of environmental origin, and novel genotypes in both A. fumigatus and A. terreus were found. Although the isolation of Aspergillus may reflect contamination or transient colonization and thus include patients in whom antifungal therapy is not indicated, the continuously emerging reports of azole-resistant Aspergillus is worrisome, and resistance remains a significant challenge. This is of concern as effective alternative treatments to azoles are lacking and as it suggests that azole-resistant A. fumigatus may also be an increasing challenge in other patient populations at risk for aspergillus disease.

DATA AVAILABILITY STATEMENT
The raw data can be provided from the corresponding author according to the Danish law.

ETHICS STATEMENT
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent from the patients was not required to participate in this study in accordance with the national legislation and the institutional requirements.

AUTHOR CONTRIBUTIONS
MR, MA, and HJ designed the study. HJ, JG, and LK were responsible for primary cultures and isolation. MA, JG, and LK were responsible for the susceptibility testing. RH and NA-C performed the molecular analysis. MR performed the data management. MR and MA wrote and revised the manuscript after review from all co-authors. All authors contributed to the article and approved the submitted version.