The Real-World Effectiveness of Antifungals in People with Cystic Fibrosis and Aspergillus-Positive Cultures

Abstract

Rationale

The pathogenicity of Aspergillus in the cystic fibrosis (CF) airway is debated, leading to unclear clinical benefit of antifungal therapy for Aspergillus infection.

Objective

To determine the real-world effectiveness of antifungal use in people with CF (PwCF) with Aspergillus species in the United States.

Methods

We conducted a retrospective cohort study evaluating the association of antifungal use and respiratory outcomes in PwCF and Aspergillus-positive cultures using the Cystic Fibrosis Foundation Patient Registry. Marginal structural models using inverse-probability treatment weighted estimators were used to test whether antifungal exposure was associated with forced expiratory volume in 1 second percent predicted (FEV₁pp) and pulmonary exacerbation rate while controlling for fixed and time-varying confounders. We conducted sensitivity analyses on individuals with persistent Aspergillus and without concomitant allergic bronchopulmonary aspergillosis (ABPA).

Results

A total of 14,754 individuals with Aspergillus-positive cultures between 2006 and 2019 were identified. Antifungals were prescribed to 3,575 (24.2%) unique PwCF during the study period. Antifungal use was not associated with FEV₁pp (adjusted estimate = −0.96 percentage points; 95% confidence interval [CI] = −2.21, 0.29). Antifungal use was associated with 29% increased rate of pulmonary exacerbations requiring intravenous (i.v.) antibiotics (adjusted incidence rate ratio = 1.29, 95% CI = 1.22, 1.37). In sensitivity analyses limited to individuals without ABPA, antifungals were associated with 1.88 lower FEV₁pp (95% CI = −3.35, −0.41) and an increased rate of pulmonary exacerbations (adjusted incidence rate ratio = 1.30; 95% CI = 1.21, 1.40), whereas in patients with persistent Aspergillus and persistent Aspergillus without concomitant ABPA, antifungals were not associated with FEV₁pp.

Conclusions

Antifungal therapy in PwCF and Aspergillus-positive cultures was not associated with improvements in FEV₁pp, suggesting no observed benefit. Although antifungal therapy was associated with increased risk for pulmonary exacerbations, this could reflect confounding by severity of disease. Randomized clinical trials examining the clinical efficacy of antifungals in Aspergillus infections in CF are warranted.

Aspergillus fumigatus is the most common filamentous fungi and dominant Aspergillus species recovered in the cystic fibrosis (CF) airway, whereas Aspergillus flavus, Aspergillus terreus, Aspergillus niger, Aspergillus nidulans, and cryptic species are rare and often transient (1–5). Although treatment guidelines for allergic bronchopulmonary aspergillosis (ABPA) are established in CF, the clinical indications for treatment of Aspergillus colonization and infection in CF airways without ABPA are poorly understood (6). Aspergillus fumigatus airways infection may be associated with lower lung function, increased risk for hospitalization and severe pulmonary exacerbations, and radiographic evidence of bronchiectasis and air trapping (7–11). Case series and observational studies of antifungal treatment (most commonly, itraconazole) report clinical improvements in people with CF (PwCF) and Aspergillus bronchitis in the absence of ABPA (12–14). Yet, a limited randomized clinical trial has not shown benefit of antifungal medications in chronic Aspergillus fumigatus infection in PwCF (15). The study was underpowered because of poor recruitment, and therapeutic itraconazole levels were difficult to achieve. Little is known of the clinical implications of non-fumigatus Aspergillus species in CF, but reports suggest that Aspergillus terreus may be associated with lung function decline (4). Without strong evidence to support a clinical guideline for Aspergillus infection without ABPA, clinicians’ decisions to treat are widely variable (11, 16). Triazoles are the most common antifungal class prescribed to PwCF. However, triazoles, such as itraconazole, voriconazole, and posaconazole, exhibit drug–drug interactions with CF transmembrane conductance regulator (CFTR) modulators, require therapeutic drug monitoring, and pose risk for drug toxicity. Understanding the clinical indications for antifungal use in the CF population are imperative to minimize unnecessary treatment burden, cost, and toxicity. We sought to understand the real-world comparative effectiveness of antifungal therapy in PwCF with Aspergillus-positive cultures. We hypothesized that antifungal therapy would be associated with higher forced expiratory volume in 1 second percent predicted (FEV₁pp) and reductions in pulmonary exacerbations.

Some of these data have been previously reported in abstract form (17).

Methods

Study Design and Population

This was a retrospective cohort study using the Cystic Fibrosis Foundation Patient Registry (CFFPR), an encounter-based database of approximately 38,000 PwCF followed in CFF-accredited centers in the United States from January 1, 2006, to December 31, 2019 (18, 19). The CFFPR comprises demographic data, anthropometric measures, medications, pulmonary function tests, CF-related complications (including ABPA), detailed information of intravenous (i.v.) antibiotic use for pulmonary exacerbations, and respiratory microbiology, including Aspergillus species. Real-world clinical data are collected with a high level of accuracy and reliability (18). We aimed to study PwCF and Aspergillus infection. Thus, we included participants ages 6 years and older who had positive Aspergillus culture(s) without a history of solid organ transplantation. We excluded children who were younger than 6 years old because of less reliable spirometry measurements and transplant recipients because of immunosuppressed health status. We included subjects from the index encounter date of first Aspergillus-positive culture until transplantation, death, and/or last time observed.

Study Variables

The Aspergillus species variable in CFFPR does not distinguish between Aspergillus fumigatus or non-fumigatus Aspergillus species, but CF epidemiology has showed that Aspergillus fumigatus accounts for over 80% of Aspergillus species isolated in PwCF (2, 3, 20). The exposure of interest was receipt of antifungal therapy (yes/no), which was measured at each patient encounter and modeled as a time-varying exposure. The primary outcome was FEV₁pp as calculated by the registry using the Global Lung Initiative reference equations and was available at each patient encounter (18, 21). The secondary outcome was the rate of pulmonary exacerbation events requiring intravenous antibiotic therapy, captured as a care episode in the CFFPR. The start and end dates for each care episodes are available for each patient, enabling us to link the care episode data to the encounter data to examine the number of care episodes between patient encounters. Age at index Aspergillus-positive culture, sex assigned at birth (male or female), race (non-Hispanic White, non-Hispanic Black, or non-Hispanic other), Hispanic ethnicity, and CFTR mutation class (Class I–III or Class IV–VI) were time-invariant variables in our analysis. Time-varying variables included body mass index category (underweight, normal weight, overweight, obese), pancreatic insufficiency (defined as pancreatic enzyme usage), ABPA, CF-related diabetes, CFTR modulator use (defined as ivacaftor or lumacaftor/ivacaftor or tezacaftor/ivacaftor), azithromycin, oral corticosteroid use, number of care episodes in the preceding year, Pseudomonas aeruginosa coinfection, methicillin-resistant Staphylococcus aureus coinfection, Burkholderia cepacia complex coinfection, Scedosporium coinfection, CF care center (to account for center-level practices), and insurance status. We determined persistent Aspergillus status as at least two Aspergillus-positive cultures 30 days apart in the index year, on the basis of previously described chronic Aspergillus definitions (7, 15, 22).

Statistical Analysis

Cohort characteristics at the index Aspergillus encounter were described using frequency counts (and percentages of the total cohort or exposure group) or means (and standard deviations) for continuous variables. To highlight potential differences between the populations who receive antifungal treatment, we compared baseline characteristics in subjects who ever received antifungals (defined as antifungal therapy at any encounter) and those who never received antifungals during the study period using t tests and chi-square tests.

For our primary analysis, we used the inverse probability of treatment weights (IPTWs) to control for confounding and marginal structural models to model the relationship between antifungal therapy and outcomes while accounting for time-varying variables. First, using all encounters from the index date of Aspergillus diagnosis to end of follow-up, we calculated the IPTW for antifungal exposure at each visit. We used the time-invariant and time-varying covariates measured from the prior visit to generate the IPTWs. Second, these weights were applied to a generalized estimating equations (GEE) linear regression model for FEV₁pp using time-invariant covariates as fixed effects and an independence working correlation structure for repeated measures. We report the effect estimate of antifungal medication and 95% confidence intervals (CIs). For the secondary outcome, we fit a GEE Poisson model to the weighted sample and used an offset term to account for the time observed so that we were modeling the number of care episodes in the time observed between encounters. We report the adjusted incidence rate ratios (IRRs) for the association between antifungal treatment compared with no antifungal treatment on the rate of pulmonary exacerbations. (For additional details regarding the methods, see the data supplement.)

We conducted several additional analyses to evaluate the robustness of our findings. Because antifungals may be used as adjunctive medications for ABPA treatment, we performed analyses excluding encounter visits with ABPA. Transient Aspergillus may not be representative of infection, so we analyzed only those subjects with persistent Aspergillus. We also conducted our analyses while omitting people receiving antifungal treatment on or before index Aspergillus encounter. We examined the effect modification of age at Aspergillus onset, biological sex, Pseudomonas aeruginosa coinfection, and pancreatic insufficiency on antifungal and FEV₁pp by testing the significance of interaction terms, which were defined as the product of antifungal treatment and the covariate of interest. We defined “child” as those who less than 18 years old and “adult” as those who were ages 18 years and older at first Aspergillus-positive culture.

For the individuals who received an antifungal treatment at some point in the follow-up period, we compared the rate of pulmonary exacerbations before and after the first antifungal treatment, using a multivariable negative binomial model with GEE to account for repeated encounters and an offset for the period of time observed. This model was adjusted for the same covariates as those used to create the IPTWs in the primary analyses. Adjusted IRRs and 95% confidence intervals are presented for the period after (vs. before) the first antifungal treatment and for the rate of pulmonary exacerbations.

The University of Pennsylvania Institutional Review Board deemed this study as research that did not involve human subjects. The analysis was conducted using both SAS software, Version 9.4 (SAS Institute), and R statistical software, Version 4.2.1 (packages include ipw, survey, and ggplot2; R Project for Statistical Computing).

Results

Study Cohort

Figure 1 depicts the flow diagram for our study population. The cohort comprised 14,754 PwCF and at least one positive Aspergillus culture. Antifungals were prescribed to 3,575 (24.2%) unique subjects during the study period. Table 1 summarizes the demographic and clinical characteristics of the study sample at the index Aspergillus-positive encounter and stratified by antifungal use at any point. The mean age of index Aspergillus was 21.5 ± 12.7 years, and 51% of the subjects were male. The mean FEV₁pp was 71.9 ± 24.9%. Persistent Aspergillus in the index year was observed in 5,384 (36.5%) individuals. Pseudomonas aeruginosa colonization was observed in 6,298 (42.7) individuals. Subjects who received antifungals were younger, female, and had pancreatic insufficiency. The antifungal group had a lower mean FEV₁pp (66.2 vs. 73.6%) and a higher mean number of pulmonary exacerbations in the preceding year of index Aspergillus (1.47 vs. 1.17) compared with the group who never received antifungals. ABPA, oral corticosteroid use, and persistent Aspergillus were more common in subjects who received antifungal therapy, whereas baseline CFTR modulator use was less common among those who received antifungal therapy during the study period.

Figure 1.

Study flow diagram for patient population. CFFPR = Cystic Fibrosis Foundation Patient Registry.

Table 1.

Patient characteristics at index Aspergillus-positive encounter, stratified by antifungal use at any point during the study period

Characteristic	Full Cohort (N = 14,754)	Patients Who Received No Antifungals (n = 11,179)	Patients Who Received Antifungals (n = 3,575)
Age (yr) at index Aspergillus-positive culture, mean (SD)	21.5 (12.5)	21.8 (12.7)	20.8 (11.8)
Age group at index Aspergillus, n (%)
Children, ages 6–17 yr	7,219 (48.9)	5,397 (48.3)	1,822 (51.0)
Adults, ages 18 yr and older	7,535 (51.1)	5,782 (51.7)	1,753 (49.0)
Male sex, n (%)	7,521 (51.0)	5,921 (53.0)	1,600 (45.8)
Race-ethnicity, n (%)
Hispanic	1,057 (7.2)	765 (6.8)	292 (8.2)
Non-Hispanic White	13,049 (88.4)	9,905 (88.6)	3,144 (87.9)
Non-Hispanic Black	446 (3.0)	354 (3.1)	92 (2.6)
Non-Hispanic other	202 (1.4)	155 (1.4)	47 (1.3)
Two CFTR variants in Mutation Class I, II, or III, n (%)	13,430 (91.0)	10,091 (90.7)	3,339 (93.4)
BMI category, n (%)
Underweight	7,043 (47.7)	5,394 (48.3)	1,649 (46.1)
Normal weight	6,257 (42.4)	4,721 (42.2)	1,536 (43.0)
Overweight	445 (3.0)	336 (3.0)	109 (3.1)
Obese	149 (1.0)	131 (1.2)	18 (0.5)
FEV1pp, mean (SD)	71.9 (24.9)	73.6 (24.8)	66.2 (24.5)
Pancreatic insufficiency, n (%)	13,238 (89.7)	9,961 (89.1)	3,277 (91.7)
CF-related diabetes, n (%)	2,459 (16.7)	1,769 (15.8)	690 (19.3)
Oral corticosteroid use, n (%)	762 (5.2)	404 (3.6)	358 (10.0)
CFTR modulator use, n (%)	890 (6.0)	777 (7.0)	113 (3.2)
Pseudomonas aeruginosa, n (%)	6,298 (42.7)	4,772 (42.7)	1,526 (42.7)
Scedosporium, n (%)	118 (0.8)	74 (0.7)	44 (1.2)
MRSA, n (%)	2,746 (18.6)	2,097 (18.8)	649 (18.2)
ABPA, n (%)	753 (5.1)	250 (2.2)	503 (14.1)
Persistent Aspergillus, n (%)*	5,384 (36.5)	3,728 (33.4)	1,656 (46.3)
Number of intravenous antibiotic courses in the preceding year (care episodes), mean (SD)	1.24 (1.44)	1.17 (1.41)	1.47 (1.50)

Definition of abbreviations: ABPA = allergic bronchopulmonary aspergillosis; BMI = body mass index; CFTR = cystic fibrosis transmembrane conductance regulator gene; FEV₁pp = forced expiratory volume in 1 second percent predicted; MRSA = methicillin-resistant Staphylococcus aureus; SD = standard deviation.

^*Persistent Aspergillus is defined as two or more Aspergillus-positive cultures (at least 30 d apart) in the index year.

Effect of antifungal treatment on FEV₁pp

After creating the time-varying IPTWs, we plotted the distribution of weights at each visit number to ensure that there were no extreme weights and that the mean of the weights was approximately one at each visit (see Figure E1 in the data supplement).

The results from our primary analysis and subgroup analyses for FEV₁pp are presented in Figure 2. Antifungal treatment was not associated with FEV₁pp (−0.96 percentage points; 95% CI = −2.21, 0.29; P = 0.13; Figure 2). We did not observe a relationship in the persistent Aspergillus subcohort (−0.33 percentage points; 95% CI = −2.08, 1.42; P = 0.72). When limiting the study population to those with Aspergillus without ABPA (by excluding encounter visits with recorded ABPA), we observed that PwCF on antifungals had 1.88 lower FEV₁pp compared with those not on antifungals (−1.88 percentage points; 95% CI = −3.35, −0.41; P = 0.012; Figure 2). However, antifungal treatment was not associated with FEV₁pp in subjects with persistent Aspergillus cultures while excluding ABPA encounters (−1.11 percentage points; 95% CI = −3.12, 0.89; P = 0.28; Figure 2). We also found that antifungals were associated with FEV₁pp when restricting to individuals naive to antifungal therapy before the index Aspergillus culture (see data supplement for further discussion). We did not detect effect modification by age at index Aspergillus culture, sex, Pseudomonas aeruginosa, or pancreatic insufficiency (all P values for interaction terms were >0.05).

Figure 2.

The association of antifungal treatment on FEV₁pp, primary and subgroup analyses while excluding ABPA encounters. ABPA = allergic bronchopulmonary aspergillosis; CI = confidence interval; FEV₁pp = forced expiratory volume in 1 second percent predicted.

Effect of antifungal treatment on pulmonary exacerbations

The results from the generalized linear marginal structural models for the primary analysis and additional analyses for the rate of pulmonary exacerbations are presented in Figure 3. Antifungal treatment was associated with a 29% increased rate of pulmonary exacerbations before the next clinical encounter (adjusted IRR = 1.29; 95% CI = 1.22, 1.37; P < 0.001; Figure 3). We obtained similar conclusions from the model restricted to Aspergillus-positive participants, omitting ABPA encounters (IRR = 1.30; 95% CI = 1.21, 1.40; P < 0.001; Figure 3). There was no effect modification by age, sex, pancreatic insufficiency, or Pseudomonas aeruginosa coinfection on the relationship between antifungal and pulmonary exacerbation rate (all P values for interaction terms were >0.05). Antifungal use was also associated with an increased rate of pulmonary exacerbations in patients with persistent Aspergillus in the index year (IRR = 1.23; 95% CI = 1.13, 1.33; P < 0.001) and when restricted to patients with persistent Aspergillus without ABPA (IRR = 1.26; 95% CI = 1.15, 1.38; P < 0.001; Figure 3).

Figure 3.

The IRR of antifungal treatment on rate of pulmonary exacerbations (defined as care episodes), primary analyses, and subgroup analyses. ABPA = allergic bronchopulmonary aspergillosis; CI = confidence interval; IRR = incidence risk ratio.

We then examined the rate of pulmonary exacerbations before and after antifungal therapy in the subjects who received antifungal treatment anytime during the study period, using a multivariable negative binomial GEE model. In 3,096 subjects, we found an increased risk for pulmonary exacerbations after antifungal treatment in the unadjusted and adjusted models, respectively (unadjusted IRR = 1.20; 95% CI = 1.15, 1.25; P < 0.001; and adjusted IRR = 1.18; 95% CI = 1.14, 1.23; P < 0.001).

Discussion

In this comparative effectiveness study of over 14,000 children and adults with CF and Aspergillus-positive cultures, we found that antifungal treatment was not associated with improvements in pulmonary outcomes. In fact, we observed that antifungal use was potentially linked with an increased rate of severe pulmonary exacerbations requiring intravenous antibiotic therapy after accounting for confounders. Individuals who were initiated on antifungal therapy during the study period exhibited greater parameters of disease at Aspergillus onset, including lower FEV₁pp and higher number of pulmonary exacerbations in the previous year. However, we controlled for these factors, and our weighting mitigated the differences in these factors between groups, providing valid estimates for the relationship of antifungal treatment and pulmonary outcomes.

Antifungal treatment was not associated with FEV₁pp in a CF population with positive Aspergillus cultures. These data correspond with the unclear causal relationship between Aspergillus fumigatus and lung function in PwCF (10, 23, 24). However, we found that antifungal use was associated with a 1.88 percentage point lower FEV₁pp in those with Aspergillus-positive cultures, but without ABPA, most likely representing possible treatment for Aspergillus bronchitis. We also observed that antifungal therapy was associated with higher rates of intravenous antibiotic therapy without clear reductions in pulmonary exacerbation events. Single-center studies have suggested that Aspergillus fumigatus infection is associated with greater frequency of severe pulmonary exacerbations requiring intravenous antibiotic therapy and hospitalization in PwCF (7, 10). The proposed definition of Aspergillus bronchitis or infection suggests consideration of antifungal treatment when respiratory symptoms and worsening pulmonary function persist despite antibiotic therapy (1, 12). However, diagnostic criteria or biomarkers to differentiate Aspergillus bronchitis and colonization have yet to be characterized. Antifungal therapy targeted at filamentous fungi may reduce Aspergillus burden in the airways and improve respiratory symptoms (12–14). However, the level of evidence for antifungal treatment in Aspergillus infection in PwCF is weak. Antifungal therapy is often considered for PwCF with worsening lung function and pulmonary symptoms, after systemic and inhaled antibacterials have been attempted. It is possible that our observations suggest that antifungal treatment may be representative of unmeasured (or inadequately measured) disease factors that may prompt a clinician to treat. Although our IPTW modeling approach mitigated confounding by severity of disease, residual confounding may still be possible. Antifungal therapy could negatively impact health status by bacterial selection, drug–drug interactions, or idiosyncratic side effects. Another potential explanation is that the antifungal therapy was insufficiently treating the Aspergillus contributing to respiratory symptom burden, because of poor drug bioavailability, drug toxicity, antifungal resistance patterns, or the need for prolonged duration of therapy.

Among the 3,575 individuals who were given antifungals during the study period, 51% were children at onset of the Aspergillus infection, suggesting a substantial proportion of children being treated for Aspergillus-related lung disease. Aspergillus infection is classically associated with older age among PwCF; yet sampling bias (e.g., oropharyngeal swab sampling in children) may confound this relationship. Although age did not modify the relationship between antifungal use and FEV₁pp, we must consider that children tend to take liquid formulations of medications and drug metabolism may be different in children, and that antifungal pharmacokinetics and pharmacodynamics may affect achievement of therapeutic drug level and, thereby, efficacy (25).

We conducted subgroup analyses focused on PwCF with persistent Aspergillus to examine a potential benefit of antifungal treatment on FEV₁pp and pulmonary exacerbation rates. Although persistent Aspergillus cultures cannot distinguish colonization or infection, repeated fungal isolation is more suggestive of chronic Aspergillus fumigatus that may exhibit “pathoadaptive” mechanisms contributing to disease (26). Transient Aspergillus cultures are less likely to represent a true infection that may benefit from antifungals. However, antifungals were associated with increased pulmonary exacerbation risk and a trend in lower FEV₁pp in those with persistent Aspergillus (Figures 2 and 3). These findings suggest that repeated positive Aspergillus cultures alone may not indicate Aspergillus infection warranting treatment. Novel biomarkers for the diagnosis of Aspergillus infection are needed.

Overall, our data do not support real-world clinical effectiveness of antifungals for Aspergillus-positive cultures in PwCF. The only randomized controlled trial of itraconazole for the treatment of chronic Aspergillus fumigatus infection in PwCF showed no effect on pulmonary exacerbations, FEV₁pp, or patient-reported outcomes. This trial was very small, however (n = 35), and subtherapeutic itraconazole levels were found in 43% of participants in the treatment arm, limiting its interpretation (15). Taken together, these data highlight the need for more robust data to inform clinicians on the role of antifungals for the management of Aspergillus-related lung disease in CF. Existing data support antifungal use in ABPA in patients without CF, whereas the benefit for antifungal therapy in ABPA in PwCF has been less clear (6, 27–29). Data are lacking for antifungal use in Aspergillus fumigatus or other fungal infections in the absence of ABPA, except in case series. However, over 3,500 individuals were prescribed antifungals in our real-world cohort in the United States, showing the clinical importance of studying treatment for Aspergillus infection. Survey data of CF clinicians’ approach to Aspergillus-related lung disease suggest that other regions of the world may consider antifungal therapy for PwCF to a greater extent than American clinicians do (11).

We acknowledge several limitations of our study. Most important, we recognize that our observational study design limits our ability to infer causality. However, observational studies using IPTW methodology are the standard approach for comparative effectiveness analyses using “real-world” data. As a benefit of our approach, our IPTW models included time-lagged variables (see the data supplement) to best model the receipt of antifungal treatment at each visit. We attempted to mitigate confounding by severity of illness, although unmeasured or residual confounding is still possible. As fungal surveillance of CF respiratory samples for Aspergillus detection is not universal standard of care, our Aspergillus-positive study population determined by the CF registry is subject to detection bias. Furthermore, respiratory sample type may influence Aspergillus detection. However, we were unable to distinguish cultures results from different specimen types in our analysis. We recognize Aspergillus-positive culture as a surrogate for Aspergillus infection is limited, yet, diagnostic biomarkers to distinguish Aspergillus colonization and infection are not established, leaving cultures and clinical features to characterize Aspergillus bronchitis. Risk for misclassification of persistent Aspergillus is possible, because our definition requires more than one encounter with culture data in the index Aspergillus year. However, individuals with only one encounter during the index Aspergillus year were rare. Because of the limitations of the CFFPR, the specific antifungal medication name, dosing, or clinical rationale for the antifungal cannot be determined. We rely on the assumption that the antifungal treatment is for, and has antifungal activity against, Aspergillus. However, we also know that poor drug bioavailability of antifungals can lead to variable serum drug levels, affecting drug efficacy (25). We recognize that our left-truncated data because of the age restriction of ages 6 years and older may introduce bias. However, Aspergillus-related lung disease and antifungal use are rare in young children who are younger than 6 years of age; therefore, it is unlikely that the conclusions are affected. These data are based on CF centers in the United States, limiting its generalizability to other regions. We attempted to account for CF center-level practice patterns by incorporating CF center to generate our treatment weights, but we recognize that the treatment practices by center may influence the respiratory outcomes observed. Finally, the study period of this analysis precedes the introduction of elexacaftor/tezacaftor/ivacaftor, which has dramatically improved respiratory outcomes in a significant proportion of the CF population and is associated with reductions in Aspergillus fumigatus density in PwCF (30, 31). The assessment of antifungal usage and effectiveness in the current CF population remains warranted.

In conclusion, antifungals were not associated with clinical respiratory benefit in a real-life CF population in the United States. Despite our IPTW approach to account for confounders, we observed that antifungals were associated with increased rate of pulmonary exacerbations, concerning for residual indication bias. Randomized controlled trials for antifungal medications for Aspergillus infection in PwCF are necessary to guide clinicians with regard to who may benefit from treatment.