An alternative view of changes in lung function among people with cystic fibrosis

Cystic fibrosis (CF) airways disease is characterized by persistent inflammation and infection, with resultant loss of lung function over time (1). This process, while progressive, is non-linear and is influenced by factors such as the frequency of pulmonary exacerbations (PEx) (2), airway microbiology (3, 4) and cystic fibrosis transmembrane conductance regulator (CFTR) modulator therapy (5–7). Additionally, there is substantial variability about the rate of decline in forced expiratory volume in 1 second (FEV₁), the most commonly used physiological endpoint in CF clinical trials. Previous studies have shown that variability about a slope of decline matters (8, 9), and is a predictor of future FEV₁ decline (9). However, there is a paucity of data regarding the effect of CFTR modulators on FEV₁ variability, with the bulk of recent studies instead focusing on changes in absolute or relative FEV₁, the rate of FEV₁ decline, frequency of PEx, and health-related quality of life. Proper assessment of variability requires multiple sequential measurements over time. Since ivacaftor has now been approved for more than a decade, we have an opportunity to characterize variability in large numbers of treated individuals.

In this issue of the Journal of Cystic Fibrosis, Szczesniak and colleagues report a series of lung function variability measures in a large cohort of people with CF (PwCF) aged 6 years and older with class III CFTR mutations between 2008 and end-2020, using a combination of summary statistics and linear mixed effects (LME) models incorporating relevant covariates and fit to pediatric and adult subgroups, respectively (REF). The analyses are confined to data from the US Cystic Fibrosis Foundation Patient Registry (CFFPR) and the LME model used (as well as the composition of the population studied) is not identical to that used by others, therefore the results may not be directly comparable to prior studies examining lung function variability outcomes, but the results are interesting nonetheless. Treatment with ivacaftor was associated with reduced percent-predicted FEV₁ (ppFEV₁) variability in both pediatric and adult PwCF. The presence of Pseudomonas aeruginosa or methicillin-resistant Staphylococcus aureus was associated with greater variability in most outcomes for both children and adults, whereas lower variability was observed in those of more advanced age. Higher pre-ivacaftor levels of variability were associated with higher post-ivacaftor levels of variability, as might be expected, but were also associated with a greater reduction in variability. The greatest reductions were observed in children – this may indicate a differential effect of ivacaftor but likely reflects a combination of increased variability in the measurement itself in the very young, the fact that younger individuals typically have more lung function available to lose, and the censoring effect of death or transplantation in older PwCF. The observed differences pre- and post-ivacaftor were associated with reductions in ppFEV₁ variability of between 10.5 and 24.7% across the different measures of variability.

There are several caveats to the data. The cohort was inherently biased towards those with more severe disease, including extrapulmonary manifestations, since these individuals were more likely to undergo frequent measurement (the study required a minimum of four FEV₁ measurements in the calendar year prior to commencement of ivacaftor and a further four measurements in the year following initiation of the therapy) and to be followed up more regularly. This should be considered when interpreting the results, since those with advanced bronchiectasis are more likely to exacerbate; the deviation from a trend line that results from a PEx event is typically greater than that observed day-to-day due to the intrinsic lung function variability with a PEx. Although the authors performed sensitivity analyses that excluded measurements obtained on dates corresponding to recorded PEx, it should be noted that PEx events are imperfectly captured in the CFFPR, and even if all in-hospital measurements or measurements related to courses of intravenous antibiotic were successfully eliminated from the data set, it is possible that measurements obtained during subsequent recovery periods – as well as those obtained prior to the formal diagnosis of PEx – remained included. The same is plausible for measurements pertaining to PEx treated in the community. Therefore, although the number of individuals studied (n = 527) and the total number of FEV₁ measurements captured (n = 18,580) increase enthusiasm for the authors’ conclusions, further study of the variability in ppFEV₁ among frequent exacerbator phenotypes compared to those with infrequent exacerbations in the post-modulator era is warranted.

The decreased median number of observations post-initiation of ivacaftor aligns with its previously described clinical benefit. Interestingly, change in ppFEV₁ variability did not associate with change in rate of ppFEV₁ decline in this study, suggesting these two parameters are distinct from one another in the assessment of lung health and may serve as different endpoints for modulator effect. Future investigation of variability in larger cohorts receiving elexacaftor/tezacaftor/ivacaftor over time is merited, as is the impact of home spirometry measurements on variability (including biased readings over time and increased variance (10)). The latter is especially relevant as CF care shifts out of the hospital and clinic settings. There is also the matter of increasingly intermittent adherence to therapies given the benefits of CFTR modulators – some patients may now consider those therapies superfluous (11).

Overall, the primary results and their corresponding sensitivity analyses represent a thorough real-world investigation of lung function variability among ivacaftor-treated PwCF, while providing a different perspective on therapeutic evaluation as we seek to advance a precision medicine approach.