In this study, while the degree of air trapping on CT imaging improved over the two years, bronchiectasis progressed and baseline abnormalities in the ratio of airway lumen artery worsened over the course of the study. We did not detect a significant change in LCI, likely due to the small number of measurements collected on participants. As expected, our study demonstrated improvements in nutritional and growth parameters in children aged 6–11 years with CF treated with LUM/IVA. Similar to the clinical trial in children of this age [20], we noted no significant change in FEV1.
Improvements in end organ function with LUM/IVA are modest compared to the improvements noted with previous (Ivacaftor in gating mutations) [26] and subsequent (Elexacaftor/Tezacaftor/Ivacaftor [ETI] for Phe508del) therapies, sometime referred to as ‘highly effective modulators’ in people with CF [27]. In many jurisdictions, ETI has replaced LUM/IVA as the modulator of choice in people with the Phe508del mutation; however, LUM/IVA is still in use in younger children and may remain in use in jurisdictions where approval or funding is not in place for other modulators. The data reported here remains relevant therefore and establishes some important information about the real-world use of modulators in children. While we demonstrated improvements in nutritional parameters, our study did not detect improvement in lung function and, despite the improvement in air trapping; bronchiectasis progressed in children treated with this modulator combination.
In relation to the impact of LUM/IVA on structural lung disease as measured by CT scores, there are no similar studies to date in children with CF aged 6–11 years. Real-world studies have demonstrated improvements in CT scores in association with Ivacaftor treatment in children aged six years or older with CF and the G511D mutation [34–36]. More recently, studies showed that CT scores improved following the introduction LUM/IVA in adolescents and adults [37–39]. Campredon et al reported a significant decrease in mucus plugging and peribronchial thickening but not bronchiectasis using Bhalla CT scores in 283 adolescents and adults treated with LUM/IVA for one year [38]. A retrospective study of 34 adolescents and adults with CF described improvement in mucus plugging, but not any other outcome measures, using the Brody scoring system [39]. These studies did not employ spirometry control for image capture or standardisation of scanners to homogenise image quality. We used a validated spirometry controlled CT protocol [22] and the well-standardised PRAGMA CF scoring system [23] ensuring that our methods were sensitive for quantifying and monitoring structural lung changes [13]. The mean baseline % disease in our cohort differed when compared with other studies examining children with CF in this age group (Table E3 and Figure E1). The differences seen between the different cohorts is most likely related to the different image acquisition methods used as well as varying clinical practices at different centres.
We did observe improvements in trapped air scores over two years, but despite this saw a progression of bronchiectasis. This may suggest that LUM/IVA has some effect in relieving small airway obstruction caused by inflammation and infection but is unable to prevent the development or worsening of bronchiectasis caused by established or persistent infection and inflammation. In order to corroborate our findings, we examined AA dimensions throughout the segmental generations and found widening of the airways of our participants. This method has been shown to correlate well with bronchiectasis detected by PRAGMA CF scoring [17]. The increases in AA dimensions in respect to AlumenA-ratio alone over two years, indicates a subtle cleaning out of airways as described in a recent study [40]. The AA dimension findings correlate with the PRAGMA CT scores demonstrating that LUM/IVA had some effect on relieving small airway obstruction but the progression of structural lung damage continued in the smaller airways of our study participants.
While clinical trials have reported improvements in pulmonary function with LUM/IVA in different age groups [7, 20]; there have been mixed findings in real-world settings. The PROSPECT study examined the impact of LUM/IVA in children aged six and above and adults [41]. In contrast to the clinical trials, they did not report improvement in ppFEV1, but did show an improvement of 0.55 units in LCI values at twelve months among 49 participants over six years of age [42]. Other real-world studies have reported improvements in ppFEV1 in those with impaired lung function at baseline but not in those with preserved lung function [29] [43]. In the French real-world study of outcomes with LUM/IVA, adolescents with impaired pulmonary function and raised LCI values (mean 12.3) did not demonstrate improvements in LCI or ppFEV1 over one year [44]. Similar to the clinical trials and real-world studies that included children aged 6–11 years; our study did not identify any improvement in ppFEV1. As this is a cohort of children with well-preserved lung function, our ability to detect significant changes in ppFEV1 was likely limited.
MBW measurements in our study did not detect any differences in LCI2.5 in children aged 6–11 years with CF over two years. The low numbers, natural variability in disease severity and relatively small effect size suggest that any effect of LUM/IVA has on ventilation inhomogeneity may be insufficient to be detected in a clinical population this size. Most participants only had MBW measurements taken at one or two time points at varying intervals after commencement of LUM/IVA rather than the prespecified four time points over two years as originally planned (Supplementary table S1). We conclude that the small sample size and low number of children with follow up measurements significantly hampered our ability to draw any assumptions on the lack of significant changes in LCI2.5 seen in our study.
Improvements in weight, height and BMI identified in this study are similar to those reported in clinical trials [20, 28] and real world studies examining the impact of LUM/IVA in the same age group [29]. Similarly, clinical trials demonstrated a decrease in exacerbation rates requiring IV antibiotics in adolescents and adults on LUM/IVA [7, 30, 31]. While this finding was replicated in one recent real-world study (8), several others failed to demonstrate a reduction in exacerbations requiring IV antibiotics in adolescents and adults [29, 32, 33]. The low number of exacerbations in our study, the relatively mild or early nature of lung disease of study participants and the relatively small number of participants may explain why we found no difference.
This study has a number of limitations. The COVID-19 pandemic, staff shortages and sub-optimal testing environment affected the collection of our MBW measurements in clinical settings. As this was an observational follow up study, we did not have a control group that would have strengthened our findings. In particular the absence of a CT control group means that we cannot determine the precise effect of LUM/IVA on structural lung disease, other than to say that while it may improve trapped air it does not appear to prevent the progression of bronchiectasis in this age group over two years. Only a small number of children were hospitalised for pulmonary exacerbations prior to or during the study period reflecting the clinical stability in most young children with CF and making detection of any treatment effect challenging. In addition, interpretation of FEV1 data would have been more robust if larger numbers were included. National registries would be better suited to analysis of data in relation to hospitalisations, FEV1 and other clinical data collected as part of routine care.
Notwithstanding its limitations, this is an important study. The demonstration of improvements in air trapping but worsening of bronchiectasis scores on LUM/IVA are original and underline the importance of collecting real world imaging data and the ongoing development and testing of more effective CFTR modulators. The negative findings in relation to lung function measures are perhaps not surprising in the context of the previous literature, the small number of participants in the study who had MBW measurements and the fact that the study was carried out on an unselected group of children in a real-world setting. The positive findings in relation to nutritional parameters corroborate other trial and real-world data and underline the clinical benefit of introduction of LUM/IVA in this age cohort. Ongoing work by our group will examine the subsequent impact of ETI on LCI, spirometry-controlled CT scores and other outcomes in this group of children and others as part of the RECOVER trial (NCT04602468).