Impact of Elexacaftor/Tezacaftor/Ivacaftor on Systemic Antibiotic Utilization in Pediatric Patients With Cystic Fibrosis

ABSTRACT

Introduction

Cystic fibrosis (CF) is an autosomal recessive disease resulting from mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Recurrent respiratory infections are a common problem, and patients receive frequent antibiotics. Treatment of CF involves both symptomatic and underlying disease‐targeting therapies. Elexacaftor/tezacaftor/ivacaftor (ETI), a disease‐targeting therapy, combines two correctors and a potentiator to increase functioning of the CFTR channel. With better functioning channels, symptomatic improvement is expected in the form of decreased respiratory infections.

Objective

To evaluate the change in antibiotic use following the initiation of ETI therapy in pediatric patients.

Methods

This is a single‐center, retrospective chart review conducted at Nationwide Children's Hospital (NCH) in Columbus, Ohio. Patients served as self‐controls, compared before and after initiating ETI. The primary outcome was change in antibiotic days per patient‐year. Secondary outcomes included change in use of specific antibiotics, enteral versus intravenous (IV) therapy, and different classes of antibiotics. The study was approved by the NCH institutional review board (IRB).

Results

The study included 139 patients aged 6–18 years. The median antibiotic days per patient‐year was 17 days pre‐ETI and 6 days post‐ETI (p < 0.001). There was a significant decrease in antibiotic days per patient‐year for subgroups of IV antibiotics, enteral antibiotics, IV vancomycin, linezolid and tedizolid, anti‐methicillin‐resistant Staphylococcus aureus (MRSA) tetracyclines, other anti‐MRSA medications (clindamycin and trimethoprim/sulfamethoxazole), aminoglycosides, anti‐Pseudomonas aeruginosa (Psae) beta‐lactams, and fluoroquinolones. There was no significant difference for carbapenems or non‐anti‐Psae beta‐lactams.

Conclusion

ETI therapy was associated with decreased antibiotic days for many therapeutic classes of both enteral and IV antibiotics.

1. Introduction

There are over 33,000 people in the United States living with a cystic fibrosis (CF) diagnosis, with around 1,000 new cases diagnosed every year [1]. CF is an autosomal recessive disease that occurs when there is a mutation in the cystic fibrosis transmembrane conductance regulator (CFTR) gene, which encodes a chloride channel [2]. The channel conducts chloride and bicarbonate, affecting water transport, and this functions to regulate hydration status at the epithelial membrane [3]. Main complications for people with CF (pwCF) are respiratory in nature, including bronchiectasis, pneumonia, nasal polyps, hemoptysis, pneumothorax, and respiratory failure [2]. Symptomatic treatments often involve antibiotics (inhaled or systemic), anti‐inflammatory medications, viscosity‐reducing agents, and mechanical chest physiotherapy.

Pharmacologic treatment of CF involves both symptom‐based and underlying disease‐targeting therapies. There are over 700 known, pathogenic mutations to the CFTR gene, with F508del being the most common, known mutation in pwCF in the United States [4]. Certain mutations can be targeted by modulators that correct the underlying pathology. The newest modulators on the market utilize triple therapy, the most experience being with elexacaftor combined with tezacaftor and ivacaftor, termed elexacaftor/tezacaftor/ivacaftor (ETI or Trikafta®) [3]. The combination of vanzacaftor and tezacaftor with deutivacaftor (Alyftrek®) is a once‐daily, triple modulator [5] that was not yet FDA‐approved at the time of this study but is now approved as of December, 2024 for children aged 6 years and older.

A randomized, placebo‐controlled, phase 3 study from 2019 showed ETI improved percent predicted forced expiratory volume in 1 s (ppFEV₁), decreased pulmonary exacerbations, improved CF‐related quality of life, increased body mass index, and decreased sweat chloride concentration in patients 12 years and older with one copy of F508del [6]. Another phase 3 study found improved ppFEV₁, sweat chloride, and respiratory domain of the Cystic Fibrosis Questionnaire–Revised (CFQ‐R RD) score compared with tezacaftor/ivacaftor in patients 12 years and older with 2 copies of F508del [7]. Efficacy and safety results were gathered for patients aged 6–11 with at least one F508del mutation in a 2021 phase 3 study, which led to approval in this age group [8]. Safety and efficacy were similar to what was observed in patients 12 years and older. Approval in ages 2–5 followed in 2023 after a phase 3 trial evaluating pharmacokinetics, safety, and efficacy [9]. ETI is currently FDA‐approved for patients aged 2 years and older with at least one copy of the F508del mutation or another mutation that is susceptible based on in vitro data or results in production of CFTR protein [10].

With approval for increasingly younger ages, more data are available to analyze the effects on these younger populations. Landmark phase 3 clinical trials reported primary outcomes through just 24 weeks of therapy [6, 8, 9]. At the time of our study, over 1 year of data post‐initiation of ETI were available for pwCF aged 6 and older. While signs such as lung function and long‐term outcomes such as mortality rates have been shown to improve in patients using ETI, more outcomes data are needed to evaluate other clinical improvements in functional status that affect patients' everyday lives as well as overall CF care. One of these measures is pulmonary infections, for which systemic antibiotic use can serve as a surrogate marker. Other studies have identified decreased antibiotic use following ETI initiation in patients 12 years and older [11, 12, 13], and just recently, one study was published showing similar results in those 6 years and older [14]. Previous research at our institution demonstrated decreased intravenous (IV), oral, and inhaled antibiotic use, decreased rates of bacterial isolation from respiratory cultures, and the ability to de‐escalate use of chronic therapies [15, 16, 17]. There is a need for further analysis of antibiotic use in a younger CF population and in the context of antibiotic classes, which can inform antimicrobial stewardship programs. This study aimed to evaluate the change from baseline in systemic antibiotic use at 1 year post‐initiation of ETI in pediatric patients who initiated ETI at age 6 years or older.

2. Methods

2.1. Study Design and Participants

This is a single‐center, retrospective chart review conducted at Nationwide Children's Hospital (NCH) in Columbus, Ohio. The pediatric and adult CF centers at the hospital are accredited by the Cystic Fibrosis Foundation. Patients served as self‐controls, compared before and after initiation of ETI. Patients were identified from an internal NCH database as maintained in the Triple‐Combo Database study and the Cystic Fibrosis Longitudinal Study. All patients who started therapy at our institution are included in this database. Patients were included if they had a CF diagnosis, initiated ETI at NCH between the ages of 6 and 18 years, and had at least 1 year of antibiotic data in the NCH system both prior to and after initiation of ETI. Patients were excluded if they had a lung transplant, were a participant in an ETI clinical trial, or had a known interruption in ETI therapy of 14 or more continuous days during the study period. The study was approved by the NCH institutional review board (IRB) as STUDY00003821, and informed consent was waived due to determination of no greater than minimal risk.

Patient demographic and baseline data included age and weight at initiation, sex assigned at birth, race, ethnicity, gene profile for CFTR, baseline ppFEV₁, and modulator use prior to initiation. The decision was made to include patients previously on ivacaftor, an already highly‐effective modulator therapy, due to previous data that ETI can still improve outcomes over ivacaftor alone [18]. Antibiotic data were collected 1 year prior to initiation of ETI and for as long after as data were available up until the IRB submission date of November 10th, 2023. To account for the potentially longer data collection period after initiation, antibiotic use was scaled by observed person‐time. Antibiotic data collected included name of the antibiotic, if it was enteral or IV, and length of therapy. Antibiotics were only included if they were systemic and being used for a respiratory infection. Antibiotics were not included if they were aerosolized or used for other infections, anti‐inflammatory purposes, or motility indications.

2.2. Outcome Measures

The primary outcome was change in antibiotic days per patient‐year after initiation of ETI. The secondary outcomes included change in antibiotic days per patient‐year for enteral antibiotics, IV antibiotics, and various classes of antibiotics. Secondary outcomes were analyzed as each individual antibiotic contributing to the day count. For example, if a patient took two antibiotics for the same 10‐day period, antibiotic days were denoted as 10 for the primary outcome but 20 for the secondary outcomes. In addition to antibiotic days per patient‐year, the number of patients in each time frame with at least one antibiotic day is presented to demonstrate overall change in antibiotic usage for rarely‐prescribed antibiotics when both the median and interquartile range (IQR) may be zero. These numbers are not scaled to person‐year. Subgroup analyses included evaluation of the primary outcome by prior modulator use, age at initiation, CFTR gene profile, and baseline ppFEV₁. Evaluated antibiotics are listed in Appendix A.

2.3. Statistical Analysis

Variables were summarized using count (percentage) for categorical variables and median (IQR) for continuous skewed variables. All variables were visualized with bar or violin plots to check data quality and evaluate distributional assumptions. Antibiotic usage was quantified in days per patient‐year to account for different lengths of observation time before and after initiating ETI; data for the post‐ETI period exceeded 1 year if the patient had longer follow‐up, while 1 year of pre‐ETI antibiotic data were extracted for all patients. Wilcoxon signed‐rank tests were used to analyze the primary endpoint based on the skewed distribution of paired differences. Similar analyses were conducted for individual antibiotics and groups of antibiotics. Subgroup analyses were conducted by baseline ppFEV1 functional categories (< 50%, 50%–69%, ≥ 70%), use of prior CFTR modulators (prior modulator or naïve and if prior, which modulator), CFTR genotype (F508del homozygous, F508del heterozygous, or other), and age group (6–11 or 12–18 years old at ETI initiation). Two‐sided p‐values < 0.05 were considered statistically significant. All statistical analyses were performed in R software version 4.0 (R Core Team, Vienna, Austria) with reproducible programming in R Markdown.

3. Results

3.1. Demographics

One hundred thirty‐nine patients were included in the final analysis. Reasons for exclusion are outlined in Figure 1. Patient demographics and baseline characteristics are presented in Table 1. The majority of patients were female (57%) and identified as white (99%). The racial distribution is consistent with the typical distribution of patients known to have CF in the United States [19]. Fifty‐three percent of patients used a different modulator prior to initiation of ETI. One patient in the study had a nontuberculous mycobacteria infection.

Figure 1.

Patient selection. CF = cystic fibrosis, ETI = elexacaftor/tezacaftor/ivacaftor. [Color figure can be viewed at wileyonlinelibrary.com]

Table 1.

Demographics and baseline characteristics.

Characteristic	N	N = 139
Age at initiation (yr), median (IQR)	139	11 (9, 15)
Sex, female	139	79 (57%)
Race	139
White		138 (99%)
Bi‐racial		1 (0.7%)
Ethnicity	139
Hispanic/Latino		2 (1.4%)
Not Hispanic/Latino		137 (99%)
Weight at initiation (kg), median (IQR)	138	38 (27, 52)
CFTR gene profile	139
F508del homozygous		83 (60%)
F508del heterozygous		49 (35%)
Other		7 (5.0%)
Other CFTR gene profile	7
E193X, N1303K		1 (14%)
I517del, S945L		1 (14%)
P705L, M470V		1 (14%)
R1066C, R560T		1 (14%)
R117H (5 T/7 T), W1282X		1 (14%)
R74Q, R297Q		2 (29%)
ppFEV1 at initiation, median (IQR)	128	90 (82, 99)
Modulator used prior to initiation	139	74 (53%)
Prior modulatora
Tezacaftor/ivacaftor	74	20 (27%)
Lumacaftor/ivacaftor	74	54 (73%)
Ivacaftor	74	13 (18%)

Abbreviations: CFTR, cystic fibrosis transmembrane conductance regulator; IQR, interquartile range; ppFEV₁, percent predicted forced expiratory volume in 1 s.

^a Multiple responses possible.

3.2. Primary Outcome

The primary outcome is depicted in Figure 2. All 139 patients were included in the primary analysis. There was a decrease in median total antibiotic days per patient‐year after ETI initiation from 17 days (IQR 2, 46) to 6 days (IQR 1, 13), which was statistically significant (p < 0.001). This encompassed all antibiotic days for respiratory infections.

Figure 2.

Change in antibiotic days per person‐year by patient.

3.3. Secondary Outcomes

Secondary outcomes are depicted in Table 2. There was a decrease in both IV and enteral antibiotic use after ETI initiation (p < 0.001 for both). Individual antibiotics and classes of antibiotics were also evaluated.

Table 2.

Secondary outcomes.

	Patients with ≥ 1 antibiotic day, n (%)		Antibiotic days per person‐year, median (IQR); range
Characteristic	Pre‐ETI N = 139	Post‐ETI N = 139	Pre‐ETI N = 139	Post‐ETI N = 139	p a
IV antibiotics	35 (25%)	26 (19%)	0 (0, 3); 0–515	0 (0, 0); 0–40	< 0.001
Enteral antibiotics	104 (75%)	105 (76%)	21 (0, 52); 0–601	7 (2, 15); 0–330	< 0.001
IV vancomycin	16 (12%)	13 (9%)	0 (0, 0); 0–45	0 (0, 0); 0–12	0.003
Linezolid and tedizolid	13 (9%)	5 (4%)	0 (0, 0); 0–89	0 (0, 0); 0–102	0.012
Anti‐MRSA tetracyclines	43 (31%)	31 (22%)	0 (0, 13); 0–324	0 (0, 0); 0–113	< 0.001
Other anti‐MRSA medicationsb	43 (31%)	49 (35%)	0 (0, 10); 0–90	0 (0, 4); 0–30	< 0.001
Anti‐Psae beta‐lactams	34 (24%)	25 (18%)	0 (0, 0); 0–70	0 (0, 0); 0–16	< 0.001
Non anti‐Psae beta‐lactams	36 (26%)	59 (42%)	0 (0, 3); 0–47	0 (0, 5); 0–32	0.3
Carbapenems	5 (4%)	0 (0%)	0 (0, 0); 0–159	0 (0, 0); 0–0	0.059
Aminoglycosides	32 (23%)	22 (16%)	0 (0, 0); 0–80	0 (0, 0); 0–13	< 0.001
Fluoroquinolones	59 (42%)	44 (32%)	0 (0, 14); 0–75	0 (0, 4); 0–30	< 0.001

Note: Bolded p‐values indicate statistical significance.

Abbreviations: ETI, elexacaftor/tezacaftor/ivacaftor; IQR, interquartile range; IV, intravenous; MRSA, methicillin‐resistant Staphylococcus aureus; Psae, Pseudomonas aeruginosa.

^a Wilcoxon signed‐rank test.

^b Clindamycin, trimethoprim/sulfamethoxazole.

Anti‐methicillin‐resistant Staphylococcus aureus (MRSA) medications were divided into four different categories: IV vancomycin, linezolid and tedizolid, tetracyclines, and other (ceftaroline, clindamycin, trimethoprim/sulfamethoxazole). There was a decrease in IV vancomycin (p = 0.003), linezolid and tedizolid (p = 0.012), tetracyclines (p < 0.001), and other anti‐MRSA medication use (p < 0.001). No patients received ceftaroline during the study period.

Beta‐lactams were evaluated as whether or not they were effective against Pseudomonas aeruginosa (Psae). There was a decrease in anti‐Psae beta‐lactam use (p < 0.001), but a non‐significant increase in non‐anti‐Psae beta‐lactam use (p = 0.3).

There was a decrease in carbapenem use, but this did not reach statistical significance (0.059), likely due to overall low use. In fact, no patients received carbapenems post‐ETI. There was a significant decrease in fluoroquinolone and IV aminoglycoside use (p < 0.001 for both).

3.4. Exploratory Analyses

Various subgroup analyses of the primary outcome were evaluated and are depicted in Table 3. The significant decrease in antibiotic days per patient‐year was consistent for patients whether or not they used a different modulator before initiating ETI (p < 0.001), though the decrease was more robust in those naïve to modulator therapy (75% vs. 50%). There was a significant decrease for patients using tezacaftor/ivacaftor and lumacaftor/ivacaftor (p < 0.001 for both), but not for ivacaftor (p = 0.4). A significant decrease was observed for patients that were homozygous or heterozygous for F508del (p < 0.001 for both), but not for those with other mutations (p = 0.078). A significant decrease in antibiotic days was not dependent on age at ETI initiation (p < 0.001 for both 6–11 years and 12–18 years). For baseline ppFEV₁ as a marker of lung dysfunction, this was divided into mild (≥ 70%), moderate (50%–69%), and severe (< 50%). The total number of patients was less than 139 due to missing baseline data, potentially due to telehealth appointments during more restrictive periods of the COVID‐19 pandemic. Significant decreases were observed in the mild and moderate group (p < 0.001 and p = 0.031, respectively), but not in the severe group (p = 0.3).

Table 3.

Antibiotic days per person‐year: subgroup analyses.

Subgroup	Pre‐ETIa	Post‐ETIa	p b
Any modulator pre‐ETI
Yes (n = 74)	14 (0, 45)	7 (3, 13)	< 0.001
No (n = 65)	24 (10, 48)	6 (0, 12)	< 0.001
Modulator used
Tezacaftor/ivacaftor (n = 20)	20 (14, 59)	6 (0, 12)	< 0.001
Lumacaftor/ivacaftor (n = 54)	16 (3, 50)	7 (3, 13)	< 0.001
Ivacaftor (n = 13)	4 (0, 14)	4 (0, 7)	0.4
CFTR gene profile
F508del homozygous (n = 83)	15 (5, 48)	7 (3, 13)	< 0.001
F508del heterozygous (n = 49)	20 (0, 49)	5 (0, 11)	< 0.001
Other (n = 7)	14 (12, 22)	12 (4, 15)	0.078
Age at initiation (years)
6–11 (n = 74)	14 (0, 33)	8 (4, 16)	< 0.001
12–18 (n = 65)	26 (14, 58)	4 (1, 11)	< 0.001
Baseline ppFEV 1
≥ 70% (n = 119)	14 (0, 42)	6 (0, 11)	< 0.001
50–69% (n = 6)	61 (58, 68)	8 (5, 13)	0.031
< 50% (n = 3)	130 (112, 134)	38 (25, 40)	0.3

Note: Bolded p‐values indicate statistical significance.

Abbreviations: CFTR, cystic fibrosis transmembrane conductance regulator; ETI, elexacaftor/tezacaftor/ivacaftor; IQR, interquartile range; ppFEV₁, percent predicted forced expiratory volume in 1 s.

^a Median (IQR).

^b Wilcoxon signed‐rank test.

4. Discussion

While there are many manifestations of CF, patients are frequently impacted by respiratory infections and the need for antibiotics. PwCF utilize various symptom‐based therapies including, but not limited to, antibiotics. We hypothesized that one of the newest underlying disease‐targeting therapies could reduce the need for one of these symptom‐targeting therapies.

Our institution recently published benefits of ETI therapy in increasing ppFEV₁, reducing MRSA and Psae culture positivity, decreasing antibiotic use, and decreasing markers of lung dysfunction [15]. The present study more than doubles the number of patients and now includes predominately pediatric patients. This study also breaks down patients into subgroups and antibiotics into various categories, which is informative to antibiotic stewardship programs. To our knowledge, this is only the second study to analyze change in antibiotic use post‐ETI in pediatric patients as young as 6 years of age, the first being the recently‐published article by Williams et al. [14].

Overall, this study observed a decrease in use of IV, enteral, and all antibiotics combined. This study also showed a decrease in IV vancomycin, linezolid and tedizolid, anti‐MRSA tetracyclines and other anti‐MRSA medications, IV aminoglycosides, anti‐Psae beta‐lactams, and fluoroquinolones. The decrease in use of antibiotics targeting MRSA and Psae is particularly noteworthy as these are some of the most common and most severe infections that pwCF can present with [20]. Use of non‐anti‐Psae beta‐lactams and carbapenems did not show a significant decrease, for carbapenems likely due in part to the small sample sizes of patients using this class of medication. The lack of decrease in anti‐Psae beta‐lactams could be due to a shift in using more of these agents as compared to the broader anti‐Psae agents, which did show a decrease. Overall, this favors a trend towards utilization of more narrow‐spectrum agents.

The primary outcome of antibiotic days per patient‐year was further analyzed by subgroups. Decrease in antibiotic use was observed independent of age at initiation of ETI. Benefit was evident only for those who previously received a combination modulator: not surprisingly, this was not realized in patients who were previously on a highly‐effective modulator therapy with ivacaftor monotherapy. Patients that were homozygous or heterozygous for F508del exhibited significant decreases in antibiotic use, but those with a different mutation (5% of the study population) did not. This study mirrors results from a phase 2 study from 2018 in which patients that had one copy of F508del (with a minimal function second mutation) and patients homozygous for F508del had a significant increase in ppFEV₁ with ETI therapy. In that study, heterozygous patients were compared to placebo, and homozygous patients were compared to tezacaftor/ivacaftor therapy [21]. Another retrospective, single‐center study demonstrated improvements in lung clearance index and ppFEV₁ with ETI, a change that was not seen in the patients using lumacaftor/ivacaftor [22]. In 2025, another retrospective study found decreased antibiotic courses per year, decreased antibiotics per course, and decreased use of IV antibiotics with ETI [14]. The present study adds to past literature by correlating real‐world, clinical efficacy outcome improvement with ETI therapy compared to other combination modulators and includes the youngest patients eligible for ETI with at least 1 year of data at the time of the study.

In the present study, F508del heterozygous patients experienced a larger decrease in antibiotic usage than F508del homozygous patients (75% vs. 53%), likely due to fewer of the heterozygous patients being on prior modulator therapy due to fewer approved modulator options for patients with only one copy of F508del. Similarly, a significant decrease in antibiotic use was seen for patients with mild and moderate lung dysfunction at baseline, but not in those with severe dysfunction (2% of patients analyzed). In most of the aforementioned scenarios where significance was not reached, it would be reasonable that small sample size contributed to this lack of significance.

This study is limited in its design as a retrospective chart review. Certain data were not available for all patients, including baseline lung function and, potentially, outside antibiotic usage not captured within the NCH patient record. The study was also unable to ensure compliance with ETI therapy, and it is possible that interruptions in ETI therapy were unaccounted for if not documented in the patient chart. Finally, the study spanned the period of the COVID‐19 pandemic. This could have impacted infection rates due to social distancing and influenced differences between age groups, as older patients were more likely to have more data available during more stringent isolation periods. This limitation will affect many ETI studies as the initial approval time periods overlap with the pandemic.

Overall, use of ETI was associated with decreased antibiotic use for many different classes of both enteral and IV antibiotics. Decrease in use was not dependent on age at initiation or whether a patient had previously used a different modulator. Decrease did appear dependent on which modulator was used previously, CFTR genotype, and baseline lung function. However, this may be due to smaller sample size in certain groups and whether the previous modulator was a highly‐effective modulator for the patient's gene mutations. Evaluating a larger sample size with additional patients meeting these criteria may lead to different findings. A future direction of this study will include evaluating patients who initiated ETI between ages 2 and 5 years old, currently the youngest group for which this medication is approved.

Author Contributions

Julianne Fava: writing – original draft, investigation, methodology, writing – review and editing; project administration, visualization, conceptualization, data curation. Emily Stephan: Writing – review and editing, methodology, conceptualization. Jessica Tansmore: conceptualization, methodology, writing – review and editing. Shahid Sheikh: conceptualization, writing – review and editing, methodology. Mariah Eisner: writing – review and editing, Formal analysis. Kimberly J. Novak: conceptualization, writing – review and editing, methodology, supervision, project administration.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1

Fava J., Stephan E., Tansmore J., Sheikh S., Eisner M., and Novak K. J., “Impact of Elexacaftor/Tezacaftor/Ivacaftor on Systemic Antibiotic Utilization in Pediatric Patients With Cystic Fibrosis,” Pediatric Pulmonology 61 (2026): e71749. 10.1002/ppul.71749.

Data Availability Statement

The authors have nothing to report.