FormalPara Key Summary Points

Why carry out this study?

Ivacaftor is a cystic fibrosis transmembrane conductance regulator (CFTR) potentiator developed to treat the underlying cause of CF. Treatment with ivacaftor improves clinical outcomes in people with CF (pwCF) with certain gating and non-gating CFTR mutations.

Since CF is a lifelong disease, it is important to understand the long-term impact of ivacaftor treatment in a real-world setting.

This observational study aimed to assess the real-world effectiveness and healthcare resource utilization after ivacaftor initiation in pwCF in France.

What was learned from the study?

At 35 CF centers in France, 129 pwCF were enrolled; 58.9% were < 18 years of age. Improvements in lung function, rate of pulmonary exacerbations, nutritional status, and growth metrics were observed with ivacaftor use and were sustained over 36 months of follow-up. Additionally, ivacaftor usage was associated with a reduction in healthcare resource utilization.

The results from this real-world study of ivacaftor use in France were consistent with prior clinical trial outcomes and confirm the clinical effectiveness of ivacaftor.

Digital Features

This article is published with digital features, including a summary slide, to facilitate understanding of the article. To view digital features for this article go to https://doi.org/10.6084/m9.figshare.14459532.

Introduction

Ivacaftor is a cystic fibrosis transmembrane conductance regulator (CFTR) modulator that potentiates the transport of chloride ions through the CFTR channel [1,2,3] and was developed to treat the underlying cause of cystic fibrosis (CF). In clinical studies, ivacaftor monotherapy has been demonstrated to improve clinical outcomes in people with CF (pwCF) with certain gating and non-gating mutations [4, 5].

The combination of multiple CFTR modulators (e.g., lumacaftor and ivacaftor, tezacaftor and ivacaftor, or elexacaftor, tezacaftor, and ivacaftor) has since broadened the population of pwCF who may benefit and has further improved clinical outcomes [6]. Based on clinical trial results [7, 8], ivacaftor was initially granted marketing authorization for the treatment of pwCF ≥ 6 years of age with a G551D-CFTR mutation by the European Medicines Agency (EMA) in 2012 [4]. An extension for other Class III non-G551D mutations (G178R, S549N, S549R, G551S, G1244E, S1251N, S1255P, or G1349D) was granted by EMA in 2014, and France followed with expanded market availability for additional mutations in 2015 [9, 10]. Indications for approved use of ivacaftor and reimbursement for ivacaftor in France have continued to expand since initial approval [11,12,13], and most recently the indication has been expanded by EMA to pwCF as young as 4 months [13]. The health authorities requested real-world data for all pwCF treated with ivacaftor in France in order to evaluate drug effectiveness and safety in pwCF treated at CF centers in France without constraints of clinical trial inclusion and exclusion criteria. Hence, BRIO (Cystic FiBRosis In Observation) was designed as an observational study to assess real-world effectiveness and healthcare resource utilization (HCRU) after initiation of ivacaftor in pwCF with G551D or non–G551D gating mutations in France.

Methods

Observational Study Design and Participants

BRIO was an observational, noninterventional, multicenter, retrospective, prospective study in France for pwCF treated with ivacaftor. The protocol for this study was developed in consultation among Vertex Pharmaceuticals, health authorities, and a scientific steering committee. PwCF were enrolled from the CF centers network (Société française de la mucoviscidose, filière muco-CFTR). Inclusion criteria included pwCF treated with ivacaftor. Two populations were recruited: pwCF already receiving ivacaftor at study inclusion, and pwCF initiating ivacaftor at study inclusion.

The primary objective was to describe the characteristics of pwCF treated with ivacaftor in a real-world setting and to evaluate the impact of ivacaftor on percent predicted forced expiratory volume in 1 s (ppFEV1). Secondary objectives were to evaluate the impact of ivacaftor on the number of pulmonary exacerbations (PEx); HCRU in terms of hospitalization, intravenous (IV) antibiotics courses, and concomitant symptomatic treatments; nutritional status (body mass index [BMI]) and growth metrics (BMI-for-age z score); and pulmonary microbiology. BMI-for-age z score was calculated for those aged < 20 years (based on the Centers for Disease Control growth chart from 2000 [14]) because it allows for standardization of BMI to normal populations of children of different ages (≤ 20 years) and both sexes. The occurrence of adverse events (AEs) and other safety results were collected. All AEs, serious and nonserious, observed in pwCF exposed to ivacaftor during this study were collected. Pulmonary exacerbations were collected as an effectiveness outcome; in this study, PEx was defined as worsening of the clinical state requiring acute antibiotics treatment and defined by the clinician as an exacerbation. PEx could also be spontaneously reported as an AE. The proportions of pwCF treated with ivacaftor presenting with abnormally high levels of serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), γ-glutamyl transferase, alkaline phosphatase, and total bilirubin were assessed. Values for ppFEV1 were based on the Global Lung Function Initiative equations for White individuals [15].

All participant data were collected from paper or electronic medical records. Data were collected and transferred by local site staff onto a secure web-based electronic case report form. Data were collected retrospectively and/or prospectively for up to 12 months prior to ivacaftor initiation through up to 24 months after enrollment (Figure S1). Retrospective data were collected at study entry. Post-enrollment study data were collected approximately every 6 months for 2 years of participation. For pwCF already being treated with ivacaftor at study inclusion, the collection of on-treatment data was both retrospective and prospective. For those initiating ivacaftor at study inclusion, on-treatment data were only prospective.

Statistical Analysis

Baseline value was defined as the last available value at or before the date of ivacaftor initiation. Analyses of lung function and nutritional status summarized the average changes from baseline within each 6-month interval after ivacaftor initiation. Analyses of clinical events evaluated rates in the 12 months prior to ivacaftor (pre-ivacaftor) compared with the 12 months after ivacaftor initiation (post-ivacaftor). Clinical events in the overall post-ivacaftor initiation period were summarized to provide an annualized rate.

A mixed model for repeated measures (MMRM) for continuous outcomes (e.g., ppFEV1) and a negative binomial model for PEx were used for analysis. Least squares (LS) means with 95% CIs were obtained from an MMRM model, with the average change from baseline in a time interval as the dependent variable, categorical time interval as a fixed effect, continuous baseline value as a covariate, the pwCF as a random effect, and an unstructured covariance matrix for repeated measures. The estimated event rates and estimated number of days per person-year with 95% CIs for the 12 months pre-ivacaftor and first 12 months post-ivacaftor, and the corresponding ratio (with 95% CI) of 12 months post-ivacaftor relative to 12 months pre-ivacaftor, were obtained from a log-negative binomial regression model. The model specified the number of events in each time interval as the dependent variable, categorical time interval as a fixed effect, and continuous ppFEV1 at baseline as a covariate with an unstructured covariance matrix for repeated measures. The estimated annualized event rate with 95% CI for the overall post-ivacaftor period was obtained from a similar negative binomial regression model for the number of events in the 12 months pre-ivacaftor period and the overall post-ivacaftor period. The CIs were nominal and did not control for multiplicity. All analyses were performed using SAS® version 9.4 (SAS Institute, Cary, NC, USA).

Ethics

This study was performed in accordance with the Helsinki Declaration of 1964, and its later amendments. Each investigator used International Society for Pharmacoepidemiology guidelines for good pharmacoepidemiology practice (2007) and good pharmacovigilance practices. In accordance with French regulatory requirements for observational studies, written informed consent was obtained from each pwCF or their parent or legal guardian, as appropriate, and assent was obtained from each child (if applicable).

Results

Population

A total of 129 pwCF were enrolled at 35 tertiary CF centers in France from March 16, 2016, to December 2, 2019 (Table 1). Of the 57 CF centers in France contacted regarding participation in the study, most centers (35) participated; those that declined did so because they did not have any pwCF who met inclusion criteria. The mean age at ivacaftor initiation was 19.1 years (range, 2–64 years), and 53 pwCF (41.1%) were < 12 years of age (Table 2). Most pwCF (64.3% [n = 83]) had a G551D allele, and G551D/F508del was the most common CFTR genotype (39.5% [n = 51]). The baseline severity of CF disease covered a broad range, with ppFEV1 ranging from 17.7 percentage points (pp) to 125.8 pp (mean [SD], 75.22 [24.86] pp). The range of ivacaftor exposure was 20–92 months, with a mean (SD) of 52.7 (20.8) months.

Table 1 Participant disposition
Full size table
Table 2 Demographics and clinical characteristics at baseline
Full size table

Data for ≥ 24 months of treatment with ivacaftor was available for 123 pwCF, and 120 participants had initiated ivacaftor prior to enrollment (Figure S2). Seven pwCF (5.4%) discontinued ivacaftor after enrollment; five pwCF (3.9%) withdrew after discontinuation of ivacaftor (participation in another interventional clinical study [n = 2], lung transplantation [n = 1], moving to another city [n = 1], and missed visit [n = 1]); two pwCF discontinued ivacaftor but remained in the study (artificial insemination [n = 1] and excessive weight gain [n = 1]). Twenty-two pwCF (17.1%) interrupted ivacaftor, with AE being the most common reason (n = 4 [3.1%]). Other reasons included the individual deciding not to take or forgetting to take ivacaftor, difficulties obtaining drug supply, and travel. Interruptions ranged from 1–177 days.

Clinical Outcomes

The LS mean (SE) ppFEV1 increased by 8.49 (1.08) pp through the first 6 months of ivacaftor from a baseline mean (SD) of 75.22 (24.86) pp; this improvement was sustained through up to 36 months of ivacaftor (Fig. 1a).

Fig. 1
figure 1

Absolute change from baseline in a ppFEV1, b BMI-for-age z score, and c BMI. BMI body mass index, LS least squares, ppFEV1 percent predicted forced expiratory volume in 1 s

Full size image

The estimated event rates for all PEx were 0.86 events per person-year in the 12 months of the pre-ivacaftor period and 0.49 events per person-year in the first 12 months of the post-ivacaftor period; the associated estimated rate ratio was 0.57 (95% CI 0.43–0.75). For the overall post-ivacaftor period (i.e., the rates that occurred during overall exposure to ivacaftor), the PEx event rate was 0.57 per person-year. In the 12 months of the pre-ivacaftor period, the number of days of PEx per person-year was 19.22 days and 10.92 days for the first 12 months on ivacaftor; the estimated rate ratio was 0.57 (95% CI 0.43–0.75). The number of days of PEx was 12.46 days per person-year for the overall post-ivacaftor period (Table S1).

For PEx requiring hospitalization, the estimated event rate decreased from 0.11 events per person-year in the 12 months of the pre-ivacaftor period to 0.03 events per person-year during the first 12 months of the post-ivacaftor period; the associated estimated rate ratio was 0.25 (95% CI 0.13–0.48). The event rate of PEx requiring hospitalization was 0.06 events per person-year during the overall post-ivacaftor period (Table 3).

Table 3 Number of exacerbations, hospitalizations, and courses of medication before and post ivacaftor treatment
Full size table

For pwCF aged < 20 years at ivacaftor initiation (n = 79), the mean (SD) BMI-for-age z score was −0.48 (1.04) at baseline; the LS mean (SE) z score increased by 0.31 (0.05) during the first 12 months following ivacaftor initiation and then remained relatively stable (Fig. 1b). In pwCF aged ≥ 20 years at ivacaftor initiation, the mean (SD) BMI at baseline was 22.04 (2.61) kg/m2; the LS mean (SE) of BMI increased by 1.01 (0.18) kg/m2 during the first 12 months following ivacaftor initiation and was sustained through 36 months of ivacaftor (Fig. 1c).

Healthcare Resource Utilization

The estimated event rate for all hospitalizations (regardless of cause) was 0.47 hospitalizations per person-year in the 12 months of the pre-ivacaftor and 0.22 hospitalizations per person-year in the first 12 months of the post-ivacaftor period; the associated estimated rate ratio was 0.48 (95% CI 0.32–0.71). For the overall post-ivacaftor period, the event rate of hospitalizations was 0.22 per person-year (Table 3). The estimated number of days of hospitalization per person-year was 3.65 days for the 12-months pre-ivacaftor period and 1.43 days for the first 12 months on ivacaftor. The estimated rate ratio was 0.39 (95% CI 0.24–0.63). The number of days of hospitalization was 1.58 days per person-year for the overall post-ivacaftor period (Table S1).

The estimated number of antibiotics courses per person-year was 1.43 courses during the pre-ivacaftor period and 0.79 courses during the first 12 months of the post-ivacaftor period; the associated estimated rate ratio was 0.55 (95% CI 0.40–0.75). For the overall post-ivacaftor period, the rate was 0.88 courses per person-year (Table 3). For IV antibiotics, the estimated rate ratio for the first 12 months after ivacaftor initiation vs the 12 months pre-ivacaftor was 0.39 (95% CI 0.23–0.66). Similar trends were seen for numbers of days of antibiotics use for PEx (Table S1).

Pulmonary Microbiology

The percentage of pwCF with a positive culture for ≥ 1 of the assessed respiratory pathogens was higher during the 12 months of the pre-ivacaftor period than during the first 48 months of the overall post-ivacaftor period (90.9% of pwCF pre-ivacaftor and a range of 75.8% to 83.3% annually of pwCF post-ivacaftor; Figure S3). Methicillin-sensitive Staphylococcus aureus, Pseudomonas aeruginosa, Aspergillus fumigatus, and Haemophilus influenzae were the most frequently identified pathogens in positive cultures. Stenotrophomonas maltophilia and methicillin-resistant S. aureus were less common, while Achromobacter xylosoxidans, nontuberculous mycobacteria, and Burkholderia cepacia complex were positive in < 4% of pwCF. The prevalence of a positive culture of methicillin-sensitive S. aureus and P. aeruginosa in pwCF was higher during the 12 months of the pre-ivacaftor period than during the first 12 months of the overall post-ivacaftor period (58.2% vs. 49.6% and 48.2% vs. 33.9%, respectively); the percentage of pwCF with a positive culture for these two organisms generally remained lower throughout the first 48 months of the overall post-ivacaftor period. For the other pathogens cultured, prevalence was variable or too low to draw conclusions about changes over time (Figure S3).

Safety

Overall, 252 AEs were reported in 75 pwCF (58.1%) (Table S2). Six pwCF (4.7%) had AEs (serious or nonserious) considered by the investigator to be related to ivacaftor, including weight increase (n = 4), hemorrhage (cheek bleeding; n = 1), and rash (n = 1). Of the 252 AEs, 47 serious AEs were reported in 25 pwCF (19.4%). The most common serious AEs were PEx (n = 3 [2.3%]; three events), hemoptysis (n = 2 [1.6%]; five events), and influenza (n = 2 [1.6%]; two events). Most serious AEs were individual cases reported in only 1 pwCF. One serious AE was reported to be related to ivacaftor; the case of hemorrhage was further described as “cheek bleeding” by the investigator, and the event was reported as resolved 1 day after onset. No deaths were reported.

Hepatic Enzymes

One pwCF had hepatitis with highly elevated ALT (> 8 × upper limit of normal [ULN]) and AST (> 8 × ULN) while taking ivacaftor (Table S3). This individual recovered within 2 weeks and continued ivacaftor for > 4 years with no recurrence of signs of hepatitis. No other pwCF had ALT or AST > 8 × ULN.

Other Endpoints

The most commonly reported comorbidities were exocrine pancreatic insufficiency, gastroesophageal reflux disease, nasal polyps, and CF-related diabetes. Little difference was observed in prevalence of these comorbidities between the pre- and post-ivacaftor periods. One pwCF underwent pulmonary transplantation > 5 years after ivacaftor initiation. The individual discontinued ivacaftor and withdrew from the study. No notable changes in chronic medication use or pancreatic enzyme use occurred with ivacaftor (data not shown).

Discussion

Here we report on ivacaftor use in a real-world setting in France. The main results were substantial improvements in ppFEV1, growth metrics, and nutritional status after initiating ivacaftor. These were observed early within the first 6 months following ivacaftor initiation and sustained through up to 36 months of treatment. The changes observed were consistent with the previously reported improvements of 5.0–13.5 pp in ppFEV1, 0.30–1.5 kg/m2 in BMI, and 0.4–0.45 in BMI-for-age z score in clinical studies of ivacaftor and in real-world observational and registry studies [7, 8, 16,17,18,19]. The sustained ppFEV1 improvement over 24 months is highly relevant since progressive lung damage is normally observed in CF [20]. The improvement in growth metrics and nutritional status observed in both age groups (aged < 20 years and ≥ 20 years) is important contrasted with the poor growth trajectory commonly observed in pwCF [21, 22], and optimized growth metrics and nutritional status are well-known factors for improving survival in pwCF.

Treatment with ivacaftor was associated with a reduction of 43% for both annualized incidence and number of days of PEx, and it was associated with a 75% reduction of annualized incidence of PEx requiring hospitalization in accordance with previous reports [18, 19]. Interestingly, we also observed a decrease in the annual percentage of pwCF who had a positive respiratory culture after initiating ivacaftor. We observed a lower annual prevalence of infections of two main pathogens (methicillin-sensitive S. aureus and P. aeruginosa) after initiation of ivacaftor. These microbiology results supported previous observations [16, 17, 23]. S. aureus and P. aeruginosa infections are associated with poor outcomes and increase treatment burden, often requiring repeated antibiotics courses [24]. Our findings are congruent with observations made in a recently published 5-year follow-up of a US cohort of pwCF treated with ivacaftor [25]; the current study enrolled a larger population, including children as young as 2 years of age, but had a shorter follow-up duration. Taken together, these results are highly clinically relevant and further support the benefits of ivacaftor and the potential to alter the clinical course of CF disease.

Monitoring of hepatic safety in this population did not reveal any untoward findings in a real-world setting. Only one person reported highly elevated hepatic enzymes while taking ivacaftor, albeit transitorily. The individual recovered and continued ivacaftor for > 4 years with no recurrence of liver disease. Overall, adverse event data were generally consistent with underlying CF disease and complemented the known safety profile of ivacaftor [7, 8, 18, 19, 26,27,28], and no deaths occurred in this study.

This study comprised the largest population of pwCF in France receiving ivacaftor to date; it included pwCF with genotypes beyond G551D and followed them for the longest duration of real-world ivacaftor treatment. This is the first study of ivacaftor with real-life data in French children < 12 years of age (~ 40% of participants). While this study was designed to include 24 months of prospective data for each pwCF, 76.7% of pwCF were exposed to > 36 months of ivacaftor. This apparent discrepancy can be explained by the inclusion of retrospective data from those already receiving ivacaftor.

In this observational study, the participating population was more heterogeneous than a trial population, and the baseline severity of CF disease was broad. As such, these results provide important data from outside of clinical trials and are reflective of the standard of care in France.

Limitations

A limitation of this study was the enrollment of pwCF who were initiating ivacaftor and pwCF already receiving ivacaftor at enrollment; thus, data were analyzed relative to date of ivacaftor initiation rather than study entry date. Consequently, results beyond 24 months should be interpreted with caution. Additionally, incompleteness of pre-enrollment data is an inherent limitation of observational studies that collect retrospective data from medical files, which however were mainly electronic. Also, lack of a comparator cohort did not allow controlling for changes over time due to natural disease progression. In this observational study, a PEx was defined as a worsening of the clinical state requiring acute antibiotics treatment and defined by the clinician as an exacerbation. In the ivacaftor clinical trials, a PEx was defined as a requirement for new or changed antibiotic therapy (IV, inhaled, or oral) for ≥ 4 signs and symptoms [8, 19, 29]; due to these differences in collection of PEx outcomes, direct comparisons should not be made between this observational study and other clinical studies. The pwCF enrolled in this real-world observational study were from a single country; thus caution must be exercised in extrapolating findings to other parts of the world due to differences in standards of care and health systems.

Conclusions

This observational study demonstrated the clinical effectiveness and safety of ivacaftor in a real-world setting in France. In addition, ivacaftor was associated with a reduction in HCRU. These data suggest that treatment with CFTR modulators can modify the outcomes of CF disease in pwCF with indicated mutations.