Epidemiology and Azole Resistance of Clinical Isolates of Aspergillus fumigatus from a Large Tertiary Hospital in Ningxia, China

Introduction

Aspergillus fumigatus (A. fumigatus) is the most common Aspergillus species that can cause various serious diseases such as life-threatening invasive aspergillosis (IA) in immunocompromised individuals with hematopoietic stem cells and solid organ transplantation, acute leukemia, and receiving immunotherapy with/without corticosteroids.^1–3 Non-invasive infections in immunocompetent patients with underlying pulmonary conditions can lead to chronic pulmonary aspergillosis (CPA), and others can suffer from allergic bronchopulmonary aspergillosis (ABPA), fungal asthma, and Aspergillus bronchitis.^4–6 A. fumigatus has been identified as a critical priority fungal pathogen in the first-ever fungal priority pathogen list released by the World Health Organization (WHO) (https://www.who.int/publications/i/item/9789240060241). Recently, there has been a rise in reported cases of IA in patients with influenza⁷ and coronavirus disease 2019 (COVID-19).^8,9 According to existing reports, the prevalence of aspergillosis varies across different provinces in China. A study conducted by Linna Huang et al found that 29.3% of patients with viral pneumonia in Beijing from December 2022 to February 2023 also had IA.¹⁰ A retrospective cohort study revealed that between August 2016 and December 2019, a total of 617 adult viral pneumonia patients were admitted to six hospitals in China, with 14.7% patients being diagnosed with IPA.¹¹ In Nanjing, among clinically suspected patients, 26.6% of patients were ultimately diagnosed with invasive pulmonary aspergillosis (IPA),¹² which is lower compared to Shanghai, where 58.5% of patients were diagnosed with IPA.¹³ However, there is currently a lack of epidemiological statistics regarding A. fumigatus infection in Ningxia and its surrounding areas.

Treatment for A. fumigatus has heavily relied on three classes of antifungal drugs: azoles, polyenes, and echinocandins.³ However, the extensive use of azoles such as itraconazole (ITC), voriconazole (VRC), posaconazole (POS), and isavuconazole (ISV) over the past decades has led to the worldwide emergence of azole resistance in A. fumigatus. This increase in resistance, both environmentally and clinically, may be responsible for the failure of aspergillosis treatment and the consequent high mortality rates.^14,15

Azole resistance in A. fumigatus is primarily caused by alterations to the sterol biosynthesis pathway, resulting from point mutations of cyp51A and the insertion of tandem repeats (TR) in the promoter region of cyp51A (such as TR34/L98H and TR46/Y121F/T289A).^16–18 However, there are increasing reports of A. fumigatus strains that are resistant to clinical azoles despite lacking cyp51A modifications.^19,20 Extensive research has been conducted on strains with multiple mutations in hmg1 and hmg2, which encode the 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase. These mutations have provided valuable insights into the mechanism of azole drug resistance in A. fumigatus.^20–25

Several attempts have been made to correlate the mutations in hapE, which encodes one of the three subunits of the CCAAT box binding complex (CBC), with azole resistance. Interestingly, a new study investigated the combination of hapE^P88L and hmg1^F262del mutations as potential contributors to azole resistance.^26–28 In addition to these mutations, other candidate genes such as crd1B, MDR1, MDR2, and erg6 have also been extensively studied and are believed to play a role in non-cyp51A azole resistance.¹ The growing number of resistant strains and the continuous emergence of new resistance mechanisms have aroused great concern worldwide.^14,29

The epidemiology of aspergillus-disease in Ningxia has not been previously reported, according to an extensive literature search available on PubMed. A. fumigatus infection in China exhibits regional differences.³⁰ General Hospital of Ningxia Medical University, the largest tertiary hospital in Ningxia with approximately 3500 beds, serves both local patients and those from surrounding areas. We believe that our hospital may be regionally representative to some extent. Furthermore, Ningxia, located in northwest China, has a unique geographical climate and lacks research on A. fumigatus infection and drug resistance. Therefore, our study focused on assessing clinical A. fumigatus isolates collected from the General Hospital of Ningxia Medical University from November 2021 to May 2023. We aim to evaluate the susceptibility of these isolates to different antifungal agents and to investigate potential resistance mechanisms, including cyp51A mutations and expression of genes related to azole-resistance.

Materials and Methods

A. fumigatus Strains and Molecular Identification

A. fumigatus isolates were obtained from individual patients hospitalized in the General Hospital of Ningxia Medical University between November 2021 and May 2023. Af293 was kindly donated by Professor Wei Liu of Peking University First Hospital. The fungal isolates were subcultured on sabouraud dextrose agar (SDA) at 25°C for 1 week. Morphological identification was performed based on macro and microscopic features, followed by matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS, bioMerieux, France). Molecular identification was performed by amplifying the β-tubulin gene (benA) and calmodulin gene (CaM) as previously described,^31,32 and the resulting PCR products were sequenced at Tsingke Biotechnology (Beijing, China). The obtained sequences were compared to reference sequences in GenBank. This study was approved by Medical Science Research Ethics Committee IRB of the General Hospital of Ningxia Medical University (2020989 approved on 09 October 2020). Written informed consent was obtained from all study participants (consent for research).

Antifungal Susceptibility Testing

The Sensititre YeastOne YO10 (Thermo Scientific, Cleveland, OH, United States) was used for antifungal susceptibility testing (AFST) with nine antifungals: anidulafungin (AND), micafungin (MFG), caspofungin (CAS), 5-flucytosine (5-FC), posaconazole (POS), voriconazole (VRC), itraconazole (ITC), fluconazole (FLC), amphotericin B (AMB). The values of minimal inhibitory concentration (MIC) and minimum effective concentration (MEC) were determined according to the manufacturer’s instructions. Briefly, conidia were collected from the SDA plates and adjusted to a McFarland standard of 0.5. A 100 µL conidial suspension was added to RPMI-1640 medium to achieve a final working concentration of 0.4–5×10⁴ CFU/mL. 96-well plates with 100 µL working solution in each well were incubated at 35°C for 24–48 h. The minimum inhibitory concentration (MIC) values of POS, VCZ, ITC, and AMB were determined to be the lowest concentrations that produced 100% inhibition, indicated by the first blue or purple well after 48 h of incubation. The minimum effective concentration (MEC) values were determined as the lowest concentrations of AND, MFG, and CAS that produced a morphological change to small, rounded, compact hyphal forms compared to the growth control well after 24 h of incubation. The standard broth microdilution assays were performed in accordance with CLSI M38 third edition³³ to determine the isolates with MIC value(s) of VRC≥2 µg/mL or ITC>1 µg/mL base on YeastOne results. For CLSI methodology, the concentration ranges of antifungal agents were as follows: VRC, ITC, POS (0.031–16 µg/mL), CAS, and MFG (0.015–8 µg/mL). The antifungal drugs were purchased from Meilunbio, China. Candida parapsilosis ATCC 22019 and Candida krusei ATCC 6258 were used as quality controls. The epidemiological cut-offs (ECVs) of AMB (2 µg/mL), CAS (0.5 µg/mL), ITC (2 µg/mL), and the clinical breakpoint of VRC (≥2 µg/mL) referred to CLSI-M59 were used to define the isolates as wild type (WT)/Susceptible (S) or non-WT/Resistant (R).³⁴ MIC50 and MIC90 values were determined based on concentrations that inhibited 50% and 90% of the isolates, respectively. The calculation method used in this study can be referenced from the literature guide “Reference Method for Broth Dilution Antifungal Susceptibility Testing of Filamentous Fungi; Approved Standard Second Edition, M38-A2”.

Whole-Genome Resequencing for Azole-Resistant A. fumigatus

In this study, a total of 276 strains of A. fumigatus from Ningxia were labeled as NYDZY 1 to NYDZY 276. Among these strains, NYDZY 162 and NYDZY 247 were identified as being resistant to azole. For the genome resequencing of NYDZY 162 and NYDZY 247, high-purity genomic DNA was extracted using QIAamp DNA Mini Kit (Qiagen). DNA qualification, library construction, and resequencing were executed at DNBSEQ platform.³⁵ High-throughput DNA sequencing (pair-end sequencing) was performed on Pacbio platform, with a read length of 300 bp at each end and an average of 3 Gb sequencing data for each library.

Determination of Azole-Resistance Related Genes Expression Levels by RT-qPCR

The expression levels of cyp51A, cyp51B, cdr1b, atrR, atrF, srbA, erg4A/4B/24A/25A, and MDR1/2/3/4 were determined by real-time qPCR. Briefly, conidia (1.0×10⁶ cells/mL) were collected from SDA plates and incubated at 35°C and 200 rpm for 48 h in 5 mL Sabouraud liquid medium. After centrifugation, the culture supernatant was discarded, mycelia were harvested, and they were washed twice with sterile distilled water. RNA extraction was performed using the RNAsimple Total RNA Kit (TIANGEN, Biotech, China) following the manufacturer’s instructions. The concentration of RNA was determined using NanoDrop spectrophotometer (ThermoFisher), and cDNA was synthesized using a PrimeScript RT Master Mix kit (TaKaRa Biotechnology, China) according to the manufacturer’s instructions. qPCR was conducted with specific primers (Table 1) using LightCycler 480 SYBR green I master mix (Roche). Af293 was used as susceptible control. The difference in gene expression was calculated using the 2-ΔΔCT method. Each sample was performed in triplicate.

Table 1 Sequence Information of the qPCR Primers Used in This Study

Statistical Analysis

Statistical differences between multiple groups were performed by a one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparisons tests, using GraphPad Prism 8.0 software. Data were presented as medians or means ± standard deviation. P < 0.05 was considered statistically significant.

Results

Fungal Isolates

A total of 276 strains identified as A. fumigatus using morphological and molecular methods and MALDI-TOF MS were collected in this study. The number of A. fumigatus isolates during the first 5 months of 2023 was 192, which is 3.4 times higher than the number during the same period of 2022. Notably, there was a significant increase in January and March of 2023 (Figure 1A). The A. fumigatus were collected from patients mainly distributed in department of respiratory (n=119, 43.12%), emergency department (n=49, 17.75%), intensive care medicine (ICU) (n=22, 7.97%), cardiology (n=12, 4.35%), geriatric (n=9, 3.26%), nephrology (n=7, 2.54%), hematology (n=7, 2.54%), and infection (n=6, 2.17%) (Figure 1B). In terms of specimen types, over 50% of the A. fumigatus isolates (n=246, 89.13%) were recovered from sputum, 9.78% (n=27) were from bronchoalveolar lavage fluid (BALF), 0.72% (n = 2) were from wounds, 0.37% (n = 1) were from puncture fluid (Figure 1C).

Figure 1 Distribution of the 276 A. fumigatus isolates from November 2021-May 2023: (A) Number of A. fumigatus isolates by month of detection; (B) Distribution by department type; (C) Distribution by specimen type.

Antifungal Susceptibility Testing and Detection of cyp51A

The MIC50, MIC90, range, ECVs, and categories are summarized in Table 2. Among the A. fumigatus isolates, 49.28% (n=136) exhibited high MIC values of 5-FC (>64 µg/mL), while more than 50% of isolates (n=232, 84.06%) demonstrated higher MIC values for FLC (≥256 µg/mL). These findings suggest the inherent resistance of A. fumigatus to these antifungal agents. All strains were wild type (WT) to AMB, with MIC values ranging from 0.25 to 2 µg/mL, and both MIC50 and MIC90 = 1 µg/mL. The MIC values of CAS for three strains (NYDZY 39, NYDZY 65, NYDZY 200) were 8 μg/mL, 8 μg/mL, 4 μg/mL, respectively, which are categorized as Non-WT. However, all three strains showed low MIC values for VRC, ITC, and POS. Two isolates of A. fumigatus (NYDZY 162, NYDZY 247) exhibited higher MIC values of >16 μg/mL for ITC and 2 μg/mL for POS. In the case of VRC, NYDZY 162 exhibited high MIC values of 8 μg/mL, while NYDZY 247 showed values of 4 μg/mL, indicating azole resistance. The distribution of VRC, ITZ, and POS MICs for clinical A. fumigatus isolates is shown in Figure 2. The MIC/MEC values of VRC, ITC, POS, CAS, MFG, and AMB against two azole-resistant isolates, determined by YeastOne YO10 and CLSI methods, are summarized in Table 3. Both methods showed high categorical agreement. Compared to the CLSI method, NYDZY 162 exhibited one-dilution higher MIC values of VRC and one-dilution lower MIC values of AMB. Sequencing results revealed no mutations in the cyp51A coding and promoter region in NYDZY 162, but amid acid substitution was observed in NYDZY 247.

Table 2 Susceptibility Profiles of 276 A. Fumigatus Isolates

Table 3 The MIC/MEC Values (µg/mL) of Different Antifungals Against Two A. Fumigatus Isolates as Determined by the YeastOne YO10 and CLSI Methods

Figure 2 MICs’ distributions of antifungals tested for 276 clinical A. fumigatus isolates with YeastOne YO10. (A) voriconazole; (B) itraconazole; (C) posaconazole. Abbreviations: S, susceptible; I, intermediate; R, resistant; ECV, epidemiological cut-offs.

The Expression Levels of Azole-Resistance Related Genes

RT-PCR results revealed that A. fumigatus clinical isolates (NYDZY 162) resistant to triazoles showed increased expression of cyp51A (P < 0.05), cdr1B (P < 0.0001), MDR1 (P < 0.0001), MDR2 (P < 0.01), artR (P < 0.01), srbA (P < 0.001), erg24A (P < 0.01), and erg4B (P < 0.01) compared to the wild-type Af293. However, there was no significant difference in the expression of these genes in NYDZY 247 (P > 0.05). On the contrary, NYDZY 247 exhibited significantly lower levels of erg24A (P < 0.001) and erg4B (P < 0.05) expression compared to both Af293 and NYDZY 162. Additionally, the expression levels of genes MDR3 (P < 0.0001, P < 0.0001), erg25A (P < 0.01, P < 0.001), and erg4A (P < 0.0001, P < 0.0001) significantly decreased in both NYDZY 162 and NYDZY 247 (Figure 3).

Figure 3 RT-qPCR of azole-resistance related genes’ expression (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 in one-way ANOVA with Tukey’s multiple comparison). Abbreviation: ns, non-significant.

The Clinical Profiles of Patients with Azole-Resistant A. fumigatum

NYDZY 162 was isolated from the sputum of a severe COVID-19 patient (Case 1) who also had non-Hodgkin’s lymphoma, diabetes, hypertension, arrhythmia, and other conditions. Chest CT scans revealed images of lungs with abnormal infiltrates, reticular and linear opacities, pleural thickening, and the accumulation of lung lesions on both sides. After receiving treatment with voriconazole and ceftazidime, the patient’s condition improved significantly, leading to his discharge. NYDZY 247 was isolated from the sputum of a patient (Case 2) with novel coronavirus pneumonia who also had co-infection with bacteria and fungi during the acute exacerbation stage of chronic obstructive pulmonary disease. This patient had a history of old pulmonary tuberculosis, old bronchial tuberculosis, pulmonary hypertension, and other diseases. The chest CT scan results showed the accumulation of lung lesions on both sides, and features like reticular and linear opacities, nodule, pleural thickening, and pleural effusion. The condition of this patient improved after treatment with voriconazole, carbofenozin, imipenem/cilastatin, cefoperazone/sulbactam. The clinical profiles of patients with azole-resistant A. fumigatus was shown in Table 4. It is worth noting that the serum galactomannan (GM) tests were negative in both patients infected with azole-resistant A. fumigatus strains.

Table 4 The Clinical Profile and Examinations of Patients Infected by Resistant A. Fumigatus

Discussion

A. fumigatus is a globally prevalent environmental mold that can cause a variety of human diseases, leading to notable mortality. Around 3,000,000 cases of chronic pulmonary aspergillosis (CPA) and invasive aspergillosis (IA) are reported worldwide annually.^39,40 Moreover, the extensive use of triazole antifungals has resulted in the emergence of azole-resistant A. fumigatus, which poses significant clinical and economic challenges.^41–44 Our study conducted in Ningxia province, China, observed an increase in the number of identified A. fumigatus strains from 2021 to 2023. Additionally, we discovered two azole-resistant strains during our research. These findings highlight a concerning trend of A. fumigatus infection in the area.

In our study, we collected a total of 276 strains of A. fumigatus. Among these, 192 strains were collected in 2023, which represents a more than 2.5 times increase compared to the previous year. This significant rise in isolates was strongly correlated with the surge in the number of COVID-19 infected patients following the lifting of epidemic control restrictions. This is similar to studies conducted in other provinces in China.⁴⁵ It has been reported worldwide that there is a high incidence of COVID-19-Associated Pulmonary Aspergillosis (CAPA) among critically ill COVID-19 patients,^46–49 which is believed to be caused by severe immunomodulation and lymphocyte depletion resulting from the virus and the subsequent use of immune-suppressive drugs.⁵⁰ The most common methods used for diagnosing CAPA include recovering Aspergillus spp. in culture media of bronchoalveolar fluid (BALF) and tracheal aspirate, as well as detecting conventional GM from BALF, tracheal aspirate, and serum specimens.⁵¹ However, in our study, diagnosing and confirming CAPA in two patients with isolated azole-resistant A. fumigatus remains challenging due to the patients’ serious underlying conditions, the overlap of clinical manifestations and imaging findings between COVID-19 and CAPA, the limited use of bronchoscopies to avoid the risk of virus transmission, and the imperfect sensitivity of serum GM.^1,52,53 Fortunately, both patients recovered and were discharged after receiving treatment with voriconazole and antibacterials, highlighting the importance of combination therapy and appropriate dosage for severe infections. But other researchers have indicated that patients infected with azole-resistant strains have a higher mortality rate compared to those infected with sensitive strains.^1,54 A 5-year retrospective cohort study showed a 21% higher day-42 mortality in azole-resistant invasive aspergillosis compared with wild-type infections cases.⁵⁵ In another multicentre retrospective study, it was also observed that the mortality rate at 6 and 12 weeks was higher in haematology patients with voriconazole-resistant cases compared to those with susceptible cases.⁵⁶ Therefore, it is an urgent issue to monitor drug resistance in clinical isolates and unravel the molecular mechanisms underlying azole resistance.

In addition to patients with associated diseases, strains with azole-resistance have also been identified in patients with no previous history of antifungal treatment.^57,58 Furthermore, environmental isolates have been found to develop azole-resistance due to the use of fungicides. Molecular epidemiological studies have demonstrated that people can acquire infections from their immediate surroundings.⁵⁹ The emergence of these resistant A. fumigatus in both natural and clinical settings has led to an expansion of the population of patients at risk, which has generated well-founded concerns in the face of limited treatment options.

The prevalence of azole-resistant Aspergillus species varies geographically. Our study found that 2 out of 276 samples (0.72%) were resistant to VRC and non-wild type for ITC, which is lower compared to Beijing (4.3%),⁶⁰ Shanghai (3.57–7.02%),^61,62 and Anhui (5.79%)⁶³ in China. In a separate study conducted in China, A. fumigatus isolates were collected from various regions, revealing a 4.4% frequency of azole resistance.³⁰ However, when compared to several other countries, the rates of azole-non-wild type A. fumigatus were found to be higher in Europe (9.5%) and North America (9.1%) than in Latin America and the Asia-Pacific region (5.3%) over a 5-year surveillance period.⁴²

Azole resistance in A. fumigatus is primarily linked to mutations in cyp51A and overexpression of cyp51A, as well as overexpression of efflux pumps. Point mutations in the coding gene in cyp51A combined with tandem repeats in the promoter region have been frequently observed, such as TR34/L98H, which means that the sequence of tandem repeats 34 times, followed by the presence of the amino acid substitution from leucine (L) to histidine (H) at position 98. Mechanically, tandem repeats within the promoter region of cyp51A is associated with overexpression of cyp51A and the L98H substitution decreases hydrogen bonding between residue 98 and the polar side chains of neighboring residues, thereby preventing the triazole from fitting into the binding pocket.^64,65 Additionally, strains carrying a TR46 allele in conjunction with Y121F and T289A exchanges were also currently the most prevalent isolates worldwide.¹ The sequencing analysis results in our study revealed that NYDZY 247 carried L98H/S297T mutations in the cyp51A gene without the presence of TR in the promoter region (TR34 and TR46), which was quite different from other previous studies where the occurrence of such amino acid substitution was always accompanied by TR variations.^66–69 The mRNA level of cyp51A did not show any increase in NYDZY 247, which could be explained by the presence of non-TR mutations in cyp51A. Other single nucleotide polymorphisms (SNPs), such as G54, M220, and G448, in the cyp51A gene have been linked to drug resistance.²⁹ These polymorphisms are frequently observed in patients with chronic pulmonary aspergillosis, invasive aspergillosis bronchitis, aspergilloma, and chronic cavitary pulmonary aspergillosis (CCPA) who have undergone long-term treatment with azole antifungals and have been associated with several clinical treatment failures.^37,70–72 The specific amino acid substitution in the cyp51A sequence without tandem repeats determines the alteration in the interaction between 14α-demethylase and the azole drugs, leading to different patterns of azole cross-resistance.⁷³ Therefore, Non-TR mutation of cyp51A in A. fumigatus can still lead to drug resistance and have significant clinical consequences, which highlights the importance of closely monitoring such mutations.

Findings from several groups have shown that the transcription factors AtrR and SrbA play a significant role in azole resistance in A. fumigatus. These factors co-regulate the expressions of cyp51A and efflux transporter cdr1B,^36,74,75 which aligns with our main finding that NYDZY 247 exhibited co-upregulation of these genes. Moreover, several clinical azole-resistant A. fumigatus isolates have shown high levels of cyp51B expression, even in the absence of cyp51A mutations,^76,77 similar to what has been observed in NYDZY 162. Recent evidence suggests that cyp51B may act as a backup system when cyp51A expression is compromised.⁷⁸ Consistent with the results of other studies, our findings also showed higher levels of cyp51A and cdr1B expression in the azole-resistant strain NYDZY 162. However, there were some observed differences: NYDZY 162 exhibited a greater increase in cdr1B expression compared to cyp51A, whereas other studies reported a >500-fold induction of cyp51A and only a modest change in cdr1B expression (>5-fold).⁷⁹ These results imply that the increased expression of the efflux transporter gene in NYDZY 162 may be the primary cause of drug resistance.

Although antifungal susceptibility testing of A. fumigatus is not routinely performed in most clinical laboratories, regular assessment of it can be helpful in guiding clinical management and understanding changes in drug resistance. Prior to our study, no reports had been made regarding azole-resistant A. fumigatus in Ningxia. Although we only identified two azole-resistant strains, this finding represents the first occurrence in the entire northwest region of China. Given the increasing trend in the number of A. fumigatus isolates from 2022 to 2023, it is likely that more azole-resistant strains will be found in the future. Moreover, we report a novel resistance mechanism in one of the azole-resistant strains, which involves the non-TR L98H/S297T mutations of the cyp51A gene. However, it should be noted that our investigation is not exhaustive, and further research will be conducted to delve into the underlying mechanisms in our future work.

Conclusions

This study, for the first time, revealed a prevalence of azole resistance in A. fumigatus at 0.72% in Ningxia province of China. Among the isolates, NYDZY 162 and NYDZY 247 exhibited the same azole resistance phenotype with different resistance mechanisms. NYDZY 162 exhibited overexpression of cyp51A without any mutations, while NYDZY 247 showed non-TR point mutations (L98H/S297T) in cyp51A that have not been previously reported. Our study emphasized the severity of A. fumigus spread in the region and the emergence of azole resistance. Therefore, it is crucial to consistently enhance the surveillance of antifungal susceptibility in clinical strains and implement effective measures to eliminate transmission and nosocomial infections, which may be helpful to prevent further exacerbation of clinical azole resistance.