Abstract
Aim To examine the trajectory of forced expiratory volume in 1 s (FEV1) using data from the European Cystic Fibrosis Society patient registry (ECFPR) collected from 2008 to 2016, i.e. the era before highly effective modulator therapy (HEMT). We evaluated risk factors for FEV1 decline.
Methods The study population included patients with a confirmed diagnosis of cystic fibrosis recorded in the ECFPR (2008–2016). The evolution of FEV1 % predicted (%FEV1) with age, and the yearly change in %FEV1 were evaluated. Risk factors considered were cystic fibrosis transmembrane conductance regulator (CFTR) mutation class, gender, age at diagnosis, neonatal screening, meconium ileus, sweat chloride concentration at diagnosis and country's income level.
Results We used 199 604 FEV1 recordings from 38 734 patients. The fastest decline was seen during puberty and in patients diagnosed before the age of 10 years. Males had a higher %FEV1, but a higher yearly %FEV1 loss between the ages of 15 and 25 years. We showed stabilisation and even improvement in %FEV1 over age in adults with a class III CFTR mutation, but a steady decline in patients homozygous for F508del or with both mutations of classes I/II. A faster decline in %FEV1 was found in patients from low-income countries compared to a similar %FEV1 evolution in patients from middle- and high-income countries.
Conclusions These longitudinal FEV1 data reflect the reality of cystic fibrosis across Europe in the era pre-HEMT, and can serve as baseline for comparison with the post-HEMT era. The similar evolution in middle- and high-income countries underlines opportunities for low-income countries.
Abstract
Longitudinal data in people with CF show a different FEV1 evolution in patients with class III mutation, the only group on effective modulators. Similar FEV1 evolution in middle- and high-income countries underlines opportunities for low-income countries. https://bit.ly/3YgIOIo
Introduction
Lung function decline has been consistently associated with numerous markers of disease progression among people with cystic fibrosis (CF). Low forced expiratory volume in 1 s (FEV1) sets patients at increased risk of pulmonary exacerbation, hospitalisation and chronic Pseudomonas aeruginosa infection [1]. Patient registries and large-scale epidemiological cohort studies are instrumental in understanding the disease's natural history and identifying risk factors for clinical outcomes. Much of the epidemiological evidence for FEV1-specific associations has been derived from CF patient registries, which collect clinical data on lung function and other markers of disease. The CF community has been systematically collecting longitudinal data for these purposes. The recent advances in analytical methodologies have enabled a better understanding of the mechanisms behind disease progression [1].
The European Cystic Fibrosis Society patient registry (ECFSPR) has been collecting data on people with CF annually from national registries and individual CF centres in Europe and neighbouring countries. The ECFSPR has already provided a valuable estimation of the number of people with CF in Europe in the first decade of the 21st century [2]. From 2008 to 2016, 43 786 CF patients from 26 countries reported data in the ECFSPR. This is a time period before highly effective modulator therapy (HEMT) was introduced for the majority of patients with CF. Ivacaftor was licensed by the European Medicines Agency (EMA) in 2012 for use in patients with a class III mutation, whereas the combination of ivacafor with tezacaftor and elexacaftor was only approved in 2020.
The aim of this study was to examine FEV1 progression over time among people with CF in Europe and to evaluate potential risk factors for FEV1 decline, such as cystic fibrosis transmembrane conductance regulator (CFTR) mutation class, gender, age at diagnosis, neonatal screening, meconium ileus, sweat chloride concentration at the time of diagnosis and country's income level.
Findings from the current study reflect the reality of CF across Europe, and due to their international nature, they offer a unique basis for comparison and discussion. In addition, given the huge number of data and the wide variability in lung disease and potential risk factors, they can highlight the major factors associated with FEV1. The results might help improve clinical care by indicating where there is a need to implement the recommended standards of care across European countries. Finally, these longitudinal FEV1 data reflect the reality of CF across Europe in the pre-HEMT era, and can serve as a baseline for comparison with the post-HEMT era.
Methods
The European Cystic Fibrosis Society patient registry
The ECFSPR collects data on people with CF annually from national registries and individual CF centres in Europe and neighbouring countries [3, 4]. The percentage of the total number of people with CF reported to the ECFSPR (self-reported coverage) varies between the participating countries and between years of follow-up; this information is presented in the ECFSPR annual reports (www.ecfs.eu/projects/ecfs-patient-registry/annual-reports). For a detailed description of this registry and its contents, refer to www.ecfs.eu/ecfspr.
Study population (dataset)
The study population included patients with a confirmed diagnosis of CF, who had at least one spirometry (FEV1) measurement in the ECFSPR from 2008 to 2016. Patients who underwent solid organ transplantation were included in the %FEV1 analysis until the year of their transplant. Patients who died during the study period are included until the year of death. Measurements before the age of 6 years and after 50 years were excluded from %FEV1 analysis, as spirometry may not be sufficiently reliable in children aged <6 years, and the number of measurements after 50 years was too low.
Definition of FEV1
The ECFSPR collects the best FEV1 value measured over the year and the corresponding height and weight of the patient [3, 4]. All FEV1 values in litres were transformed into percent predicted using the Global Lung Function Initiative reference values [5].
Age at diagnosis groups and age at follow-up
Age was calculated as the time between the date of birth and the date of the FEV1 value measured during that registry year. To define diagnosis in infants, the first decade of life, adolescents and adults, the following “age at diagnosis” groups were considered: <1, 1–10, 11–18 and >18 years.
Risk factors for the yearly change of %FEV1
These comprised type of CFTR mutation, gender, age at diagnosis, neonatal screening, meconium ileus at birth, sweat chloride concentration reported at the time of diagnosis and country's income status.
Country's income status groups
Countries were stratified into three income status groups based on tertiles of gross national income per capita, obtained from World Bank tables [6]. Bulgaria, Hungary, Lithuania, Republic of North Macedonia, Republic of Moldova, Romania, Russian Federation, Serbia, Turkey and Ukraine were assigned to the low-income group; Czech Republic, Greece, Israel, Italy, Latvia, Portugal, Slovakia, Slovenia and Spain were assigned to the middle-income group; and Austria, Belgium, Denmark, France, Germany, Ireland, Luxemburg, the Netherlands, Sweden, Switzerland and the United Kingdom were assigned to the high-income group.
Mutation groups
Patients were divided into the following mutation groups known to influence disease severity. 1) Both class I/II; 2) both F508del; 3) at least one mutation of class III; 4) at least one mutation of class IV; 5) at least one mutation of class V; 6) one of class I or II, other mutation in class unknown, or mutation unknown [7, 8, 9].
Statistical analysis
A linear mixed model was used to evaluate the evolution of %FEV1, quantify the yearly change and verify if the FEV1 progression depended on risk factors by including interaction terms in the model. In addition, nonlinear evolution of FEV1 as a function of age was allowed in the model using restricted cubic splines. The approach is similar to the model proposed by Szczesniak et al. [10], except that they did not include slope as a random effect.
Restricted cubic splines with five knots (at percentiles 5, 27.5, 50, 72.5 and 95) were used [11] to allow nonlinearity in the relation between age and %FEV1. This approach allows deviations from linearity to be captured in a flexible way. The yearly change in FEV1 (%) was calculated as the derivative of the function used to model age. The delta rule was used to calculate the 95% confidence interval for the yearly change [10].
It was evaluated whether each of the potential risk factors (gender, socioeconomic status of country, age, neonatal screening, sweat chloride concentration at time of diagnosis, meconium ileus at birth and mutation type) moderated the rate of %FEV1 change, by including interactions with age in the model. To have a uniform approach in analysis and reporting of results, the continuous moderators age at diagnosis and sweat chloride concentration have been categorised. The patients were divided into four groups according to age at diagnosis: <1, 1–10, 11–18 and >18 years, as specified earlier. The patients were divided into four groups for sweat test value based on the quantiles (<Q1, Q1–median, median–Q3 and >Q3, i.e. <80, 80–97, 98–109 and ≥110 mmol·L−1).
Results were reported from models for each moderator separately (univariable model) and a multivariable model containing all specified potential moderators. Models were based on all available cases. Differences between subjects with and without complete moderator information were evaluated. Given the large number of missing values for sweat chloride concentration, a multivariable model without sweat test value as moderator was considered additionally. Subgroup analyses (univariable models) were also performed in each mutation group separately. A detailed description of the methodology can be found in the supplementary material.
Software
All analyses were performed using SAS software (version 9.4 of the SAS System for Windows; SAS Institute, Cary, NC, USA).
Results
Study population
During the study period 2008–2016, 43 786 patients aged ≥6 years had data reported in the ECFSPR; 5052 subjects were excluded from the analysis (supplementary figure S1). 15 582 and 29 254 patients were included in the multivariable analysis with and without the parameter sweat chloride concentration at diagnosis, respectively. Clinical and demographic characteristics of the study population (univariable and multivariable model) are shown in table 1.
Clinical and demographic characteristics of the study population (study population and population included in the multivariable model)
Rate of FEV1 decline
199 604 FEV1 recordings from 38 734 patients were analysed to describe the evolution of FEV1 % predicted over age (%FEV1) and the yearly change of %FEV1 (figure 1, table 2). At every age there was a decline in mean FEV1 % predicted, with the highest rate of decline found during puberty (figure 1).
a) Overall evolution of forced expiratory volume in 1 s (FEV1); b) yearly changes in FEV1. Data are presented as mean (95% CI). Note that the yearly rate of change is constant before 7.40 and after 41.88 years (dashed vertical lines), since by using restricted cubic splines, the relationship is assumed to be linear in the outer intervals (defined by quantiles 0.05 and 0.95, respectively).
Forced expiratory volume in 1 s (FEV1) % predicted and yearly rate of FEV1 % pred change at specific ages
Effect of moderators on FEV1 evolution
Mutation groups
The most substantial and quite comparable FEV1 decline was found in the mutation groups “both class I/II” and “both F508del” (figure 2, supplementary table S1). The multivariable model confirmed the results from the univariable analysis (table 3). A different mean course was seen in people with CF with a class III mutation, with %FEV1 improving from age 20 years to 25 years onwards.
Evolution of forced expiratory volume in 1 s (FEV1) as a function of mutation group (univariable model, only considering effect of mutation).
Results of the multivariable model analysis
Gender
A significant, but slight, difference in the evolution of FEV1 between males and females was found. Males showed, on average, a higher %FEV1 (figure 3), especially when pre-pubertal and in later adulthood, but they had a higher rate of decline during puberty and even more so during young adulthood. A comparable pattern was observed in analyses performed separately in each mutation group (data not shown). Moreover, a similar difference in evolution between males and females was seen in the multivariable model (table 3).
a) Evolution and b) yearly rate of change of forced expiratory volume in 1 s (FEV1) as a function of gender (univariable model, only considering effect of gender). Note that a test for an interaction equals a test for a difference in change. Vertical dashed lines represent 5th and 95th percentiles.
Country's income status
A much worse FEV1 decline was found in the group of countries with low income compared to the middle- and high-income groups who follow a similar evolution. In the univariable analysis, the middle-income group has slightly higher %FEV1 values (figure 4a), but this difference disappears in the multivariable model (figure 4b, table 3). The same separation of low versus middle- and high-income countries is seen in the subgroups of both mutations class I/II (supplementary figure S2a), and in patients with both mutations F508del (supplementary figure S2b).
Evolution of forced expiratory volume in 1 s (FEV1) as a function of the country's socioeconomic status: a) univariable and b) multivariable model.
Age at diagnosis
A small, but significantly lower, mean %FEV1 and a higher rate of decline was found in groups with age ≤10 years at diagnosis (<1 years and 1–10 years) compared to those with age 11–18 years at diagnosis in the univariable model (p<0.001) (table 3). The group with diagnosis at age >18 years had higher %FEV1 values and showed less decline than those with earlier diagnoses (figure 5a, table 3).
Evolution of forced expiratory volume in 1 s (FEV1) as a function of age at diagnosis: result from a) univariable and b) multivariable model.
However, this pattern was less clear when evaluating the effect of age at diagnosis in each mutation group separately. Within the groups of patients diagnosed at <1 years and between 1 and 10 years (approximately), 73% and 60% had both class I/II or both F508del mutation, respectively. For the groups diagnosed between 11 and 18 years and >18 years, these percentages were 25% and 9%, respectively (supplementary table S2). A subanalysis of the group of patients diagnosed at <1 year and between 1 and 10 years indicated that the difference between the mutation groups was still present (supplementary figure S3).
In the multivariable model, the differences in mean FEV1 were attenuated compared to the univariable result (after correction for the other variables, the %FEV1 was lower in the diagnosis at age >18 years group compared to the univariable results). However, the differences in decline still had the same order of magnitude as in the univariable analysis (figure 5b, table 3).
Neonatal screening
Up to the age of 30 years, there was a small, but significantly more negative yearly change of %FEV1 when neonatal screening was not performed versus “result positive” (table 3, supplementary figure S4).
Meconium ileus
In the univariable analysis, patients with meconium ileus had lower %FEV1 values and showed a stronger decline (supplementary figure S5a). No difference was found in the yearly change within the mutation groups “both class I/II” and “both F508del” (supplementary figure S6a and b). In the multivariable model, there was neither evidence for a difference in decline as a function of meconium ileus nor a difference in mean %FEV1 value (supplementary figure S5b).
Sweat chloride concentration
In the univariable analysis, higher %FEV1 values and a lower rate of decline were found in the lowest quartile group for sweat chloride concentration (<Q1) (supplementary figure S7a). However, this difference disappeared in most subanalyses according to mutation groups. For example, there was no/hardly any difference in FEV1 according to sweat chloride concentration in mutation groups “both class I/II” and “both F508del” groups, probably also due to the low number of subjects with sweat chloride values <Q1 (supplementary figure S8a and b). In the multivariable model, there remained a significant difference in the mean %FEV1 value between the four groups, i.e. the lower the sweat test value, the higher the mean %FEV1. Only the comparison of Q4 versus Q3 was not significant (supplementary figure S7b, table 3). However, there was no evidence for a difference in decline in the multivariable model as a function of sweat chloride concentration.
Discussion
Longitudinal analysis of FEV1 change over age among CF patients in Europe over 9 years from 2008 to 2016 was performed, that is, the era before HEMT for the majority of patients with CF. Our study showed a vastly different FEV1 evolution in patients with a class III mutation, the only class with effective modulator treatment during the data collection. Ivacaftor was licensed by the EMA in 2012 for use in patients with a class III mutation, whereas the combination of ivacafor with tezacaftor and elexacaftor was only approved in 2020. A similar evolution in FEV1 in middle- and high-income countries and a much worse course of FEV1 with age in countries with low socioeconomic status was found. Finally, we confirmed several known risk factors for worse %FEV1 outcome such as female sex, diagnosis prior to the age of 10 years, and both mutations class I/II or F508del.
This study has great strengths. First, we performed a longitudinal analysis of FEV1 change over age among people with CF in Europe (ECFSPR) over 9 years, in the era before HEMT. The ECFSPR is the world's largest CF database, with information on >52 000 subjects from 40 countries in Europe [2, 12]. This is the most extensive dataset (according to our knowledge) analysing longitudinal FEV1 data among patients with CF.
Several studies have analysed longitudinal data from CF registries, but they are not as extensive as ours [13–17]. In addition, the heterogeneous European CF population showed a wide variability for some baseline factors potentially associated with FEV1, and hence is ideally suited for evaluation.
Furthermore, we used the appropriate methodology to evaluate CF lung function decline. Decline in FEV1 is the strongest predictor of mortality in people with CF [14, 15], but recent epidemiological studies point out that FEV1 progression is not constant (linear) over age [1]. Szczesniak et al. [15] substantiated the growing body of evidence that CF lung function decline is nonlinear. Our study underpins this and used appropriate methodology to evaluate this outcome. Using splines enabled us to obtain a smooth longitudinal evolution of FEV1 without imposing a simplified relationship (such as linear or quadratic), either for the mean evolution nor for the rates of change. Furthermore, given the importance of appropriate modelling of the variance and the correlation of the FEV1 values, serial correlation and correlated random intercepts and slopes were included in the statistical model. Finally, using multivariable analysis enabled us to point out the major fixed determinants of FEV1 decline among people with CF.
Many risk factors influence the progression of CF lung disease and early mortality [16, 17], but their relative impact over time has been uncertain. Previous studies have typically focused on a limited number of intrinsic (e.g. meconium ileus [18], genotype [19, 20]) or extrinsic (e.g. pulmonary exacerbations) [21] risk factors. Our study evaluated a battery of fixed risk factors, i.e. gender, age at diagnosis, neonatal screening, sweat test value, meconium ileus, mutation type and economic status of the country in the era before HEMT.
FEV1 progression with age
Our study showed a rapid lung function decline during adolescence and early adulthood, which is a frequent observation in the clinical course of CF, mainly attributed to low medication adherence during puberty [15]. In a retrospective study using the United States Cystic Fibrosis Foundation Patient Registry data, Szczesniak et al. [10] found that individuals experienced the most rapid decline at a median of 16.3 years, and the average maximal FEV1 loss was 2.0% pred per year. Higher risk of decreases in lung function during adolescence was also reported by Liou et al. [22] and by Welsh et al. [23].
In our study, as in other studies [24], FEV1 seems to be better maintained in the advanced age groups (figure 1). This is potentially a survivor effect: patients with the most severe lung disease are not included in cross-sectional studies because they died or underwent a lung transplant, leaving an increased proportion of “less severely ill” patients in the eldest age groups.
Mutation groups
Our study showed the most substantial decline in FEV1 in mutation groups “both class I/II” and “both F508del” (both groups being comparable). A better outcome in people with CF and at least one class IV or V mutation has been reported [9, 25]. However, we describe a different evolution in people with CF with a class III mutation: from adulthood onwards, mean lung function increased. This distinct FEV1 course in this subgroup reflects the benefit from treatment with the CFTR modulator ivacaftor licensed by the EMA since July 2012. De Boeck and Zolin [7], using ECFSPR data from the years 2008–2010, did not find differences in year-to-year change in FEV1 between patients with a stop-codon mutation, homozygous for F508del or at least one class III mutation. However, they confirmed the less severe disease in patients with class IV and V mutations, reflected by a higher median age and a lower proportion of subjects with severe FEV1 impairment [7]. The substantial benefit from the CFTR modulator ivacaftor [26–28] was also reported by Kawala et al. [29], who analysed retrospective data from the Canadian CF Registry and showed that FEV1 increased significantly by 5.7% pred (p<0.001) after initiation of ivacaftor. Moreover, Volkova et al. [30], using data from national United States and United Kingdom registries, showed that relative to comparators, ivacaftor-treated patients had better-preserved lung function. However, we are the first to demonstrate reversal of the downward slope of FEV1 over age in this longitudinal analysis of patients with a class III mutation.
Gender
Several studies have identified a “gender gap”, with less morbidity and mortality in males [31]. Our study showed a significant difference in evolution between males and females, and males having a higher %FEV1 on average. This same difference in evolution between males and females remained in the multivariable model. We also showed a more rapid decline of FEV1 in males during puberty and even more during young adulthood than females. This could be attributed to reduced treatment adherence among males over puberty.
The CF gender gap in mortality/survival has been described before, and the majority of data indicate that it still exists in the modern era of aggressive treatment of CF lung disease [32]. Sweezey [32] emphasised the increasing evidence suggesting a role for the effects of gender, particularly the female sex hormone oestrogen, on infection, inflammation and transepithelial ion transport, all significant determinants of CF lung disease.
Country's income status
Socioeconomic status is another perspective worth considering affecting the progression of lung disease, with several reports indicating that a low socioeconomic status is associated with significantly poorer outcomes in CF [2], and is shown to affect survival among people with CF [13, 33, 34, 35].
Our study showed the most substantial decline in FEV1 with age in countries in the low-income group. Middle- and high-income groups followed a similar evolution, with the middle-income group having slightly higher %FEV1 values. The “unexpected” higher values in the middle-income group were due to the higher prevalence of “both class I/II” and “both F508del” in the higher income group, and were therefore cancelled out in the multivariable model.
Mehta et al. [2] showed that there were fewer children and young adults with CF than expected in non-European Union countries. This finding is reinforced by the increased chance of patients surviving to the age of 40 years in European Union countries, even if they have the severe Phe508del mutation [2]. In contrast, Schechter et al. [34] showed that children with CF in the United States who could not afford health insurance were reported to have a more significant risk of death, poorer pulmonary function, poorer nutrition and were more likely to require hospitalisation for pulmonary exacerbation when compared to CF patients with health insurance. In another study, Schechter et al. [35] showed that CF health outcomes are correlated with the socioeconomic status spectrum.
Age at diagnosis, newborn screening, meconium ileus and sweat chloride concentration
Our study showed a small but significant difference in mean %FEV1 and rate of decline between groups with age at diagnosis at age ≤10 years. The group with a diagnosis aged >18 years had higher %FEV1 values and showed less decline than groups with earlier diagnoses. However, this pattern was attenuated when evaluating the effect of age at diagnosis in each mutation group separately (supplementary table S1). Adult diagnosis of CF is typically considered a milder form of the disease [36]. Keating et al. [36] showed that lung function declines more slowly in adults diagnosed with CF. Nick et al. [31] showed that people with CF diagnosed in childhood and who survive to the age of 40 years have more severe CFTR genotypes and phenotypes than adult-diagnosed patients.
Moreover, our study showed a significant difference in the yearly change between the two groups “newborn screening not done” versus “result positive”. There is clear evidence supporting newborn screening for CF. Early diagnosis allows the early management to improve clinical outcomes [37, 38].
Patients with meconium ileus had lower %FEV1 values and showed a more robust decline, but there was neither evidence for a difference in decline as a function of meconium ileus nor for a difference in mean %FEV1 value in the multivariable model. The latter coincides with the absence of a difference in the yearly change within the mutation groups “both class I/II” and “both F508del”. Johnson et al. [39] also showed that babies presenting with meconium ileus had no worse long-term outcomes than those presenting symptomatically later in infancy.
Our study showed higher %FEV1 values and a lower rate of decline in the group with the lowest values in the sweat test (supplementary figure S6a). However, this difference disappeared in most analyses performed separately in the mutation groups. In addition, Davis et al. [40] showed that sweat chloride concentration did not correlate with the severity index, either in the population as a whole or in the population of patients with alleles associated with pancreatic sufficiency, thought to have some residual CFTR function. These data suggested that sweat chloride concentration per se does not necessarily predict a milder pulmonary course in people with CF.
Limitations
Our study has some limitations that should be considered, including potential survivor bias. This is a common problem in datasets of this type [41]. In addition, disease registries may have problems with adherence to the definitions proposed, data quality process, missing data (such as many missing values for sweat chloride concentration in the ECFSPR), data entry errors and differences between countries with varied socioeconomic status [42]. In recent years, the ECFSPR introduced a data quality control project to check and overcome these limitations. Moreover, in 2018, the EMA qualified the ECFSPR as a resource for collecting CF-specific data for pharmacoepidemiology studies [43]. However, given longitudinal observations over a mean of 9 years and analysing such a big dataset, it is unlikely that any other possible Type II errors occurred for clinically significant determinants of lung disease.
Finally, this was an observational study based on registry data, so confounding factors that were not measured in the registry are possible. Moreover, patient variables that change over time, such as allergic bronchopulmonary aspergillosis, CF-related diabetes, and liver disease were not analysed in the study. Therefore, our ability to assess the impact of extrinsic factors known to affect disease progression is limited.
Conclusions and clinical implications of the study
In this retrospective study using longitudinal data from a large dataset from the heterogeneous European CF population in the ECFSPR, we found that individuals with CF experienced the most rapid decline during puberty, with males showing a more potent drop during puberty and even more so during young adulthood. Furthermore, we showed that key predictors of the early FEV1 decline included both mutation classes I/II and F508del. However, we described a different evolution in people with CF with a class III mutation, as from adulthood onwards, mean lung function increased. This was attributed to HEMT. Finally, our study showed a similar evolution in middle- and high-income countries, underlining opportunities for low-income countries.
Our study is the first to objectively analyse lung disease trajectory in an extensive European CF population dataset. Finding predictors and/or treatments associated with a slower rate of FEV1 decline remains a significant clinical objective within epidemiologic and comparative effectiveness, especially in the era of new therapies in CF, with the establishment of CFTR modulators. This study has the potential to be the last in line of publications defining CF disease prior to CFTR modulator therapies, except for class III mutation group, as results already have shown. This is of great importance for objective validation of CFTR modulator effects on a population scale, or even evaluating the risk factors for low FEV1 in people with CF on CFTR modulator therapy.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary methods 00449-2022.SUPPLEMENT
Supplementary tables 00449-2022.SUPPLEMENT2
Figure S1 00449-2022.SUPPLEMENT
Figure S2 00449-2022.SUPPLEMENT2
Figure S3 00449-2022.SUPPLEMENT3
Figure S4 00449-2022.SUPPLEMENT4
Figure S5 00449-2022.SUPPLEMENT5
Figure S6 00449-2022.SUPPLEMENT6
Figure S7 00449-2022.SUPPLEMENT7
Figure S8 00449-2022.SUPPLEMENT8
Acknowledgement
We thank the people with cystic fibrosis (CF) and their families for consenting to their data being included in the European CF Society Patient Registry (ECFSPR). We thank the centres and individual country representatives for allowing the use of the data, and the ECFSPR for providing access to anonymised patient data.
Footnotes
Provenance: Submitted article, peer reviewed.
ECFSPR Collaborative Group: We would like to thank the following people representing the countries that provided the data and the members of the ECFSPR Scientific Committee. Andreas Pfleger, Department of Paediatrics and Adolescent Medicine, Division of Paediatric Pulmonology and Allergology, Medical University of Graz, Graz, Austria; Géraldine Daneau, Sciensano, Epidemiology and Public Health, Health Services Research, Brussels, Belgium; Elise Lammertijn (ECFSPR Scientific Committee), Cystic Fibrosis Europe, and Association Muco A.S.B.L. – Mucovereniging V.Z.W., Brussels, Belgium; Guergana Petrova, Alexandrovska University Hospital, Paediatric Clinic, Sofia, Bulgaria; Pavel Drevinek, Department of Medical Microbiology, Second Faculty of Medicine, Charles University and Motol University Hospital, Prague, Czech Republic; Milan Macek Jr (ECFSPR Scientific Committee), Department of Biology and Medical Genetics, Second Faculty of Medicine, Charles University and Motol University Hospital, Prague, Czech Republic; Hanne Vebert Olesen, Department of Pediatrics and Adolescent Medicine, Cystic Fibrosis Center, Aarhus University Hospital, Aarhus, Denmark; Pierre-Régis Burgel (ECFSPR Scientific Committee), Respiratory Medicine and National Cystic Fibrosis Reference Center, Cochin Hospital, Assistance Publique-Hôpitaux de Paris, Université de Paris, Institut Cochin, INSERM U1016, Paris, France; Lydie Lemonnier-Videau, Vaincre la Mucoviscidose, Paris, France; Andrea Párniczky, Heim Pál National Pediatric Institute, Budapest, Hungary; Godfrey Fletcher, The Cystic Fibrosis Registry of Ireland, Dublin, Ireland; Rita Padoan, Cystic Fibrosis Regional Support Centre, Department of Pediatrics, University of Brescia, Brescia, and Scientific Board of Italian CF Registry, Rome, Italy; Anna Zolin, Department of Clinical Sciences and Community Health, Laboratory of Medical Statistics, Epidemiology and Biometry G.A. Maccacaro, University of Milan, Milan, Italy; Elina Aleksejeva, Department of Pneumology, Children's Clinical University Hospital, Rīga Stradinš University, Riga, Latvia; Kestutis Malakauskas, Adult Cystic Fibrosis Center, Department of Pulmonology, Lithuanian University of Health Sciences, Kaunas, Lithuania; Anna-Maria Charatsi, Paediatric Department, Centre Hospitalier Luxembourg, Luxembourg; Vincent Gulmans, Dutch Cystic Fibrosis Foundation (NCFS), Baarn, the Netherlands; Stojka Fustik, Centre for Cystic Fibrosis, University Children's Hospital, Skopje, North Macedonia; Ivana Arnaudova Danevskai, Centre for Cystic Fibrosis, Children and Adults, Institute for Respiratory Diseases in Children, Kozle, North Macedonia; Oxana Turcu, Ambulatory Cystic Fibrosis and Other Rare Diseases Center, Institute for Maternal and Child Healthcare, State University of Medicine and Pharmacy “Nicolae Testemitanu”, Department of Pediatrics, Chisinau, Republic of Moldova; Luísa Pereira, Centre for Cystic Fibrosis, Hospital de Santa Maria, Lisbon, Portugal; Liviu Pop, Victor Babes University of Medicine and Pharmacy Timisoara, National Cystic Fibrosis Centre Timisoara, Romania; Elena Kondratyeva, Research Centre for Medical Genetics, Moscow, Russian Federation; Milan Rodić, National Centre for Cystic Fibrosis, Mother and Child Health Institute of Serbia “Dr Vukan Čupić”, Belgrade, Serbia; Hana Kayserová, Cystic Fibrosis Centre, University Hospital of Bratislava, Bratislava, Slovakia; M. Dolores Pastor-Vivero, Paediatric Pneumology and Cystic Fibrosis Unit, Osakidetza, Hospital Universitario Cruces, Bizkaia, Spain; Isabelle de Monestrol, Stockholm CF centre, Karolinska University Hospital, Karolinska Institutet, Stockholm, Sweden; Anders Lindblad (ECFSPR Scientific Committee), Gothenburg Cystic Fibrosis Centre, Queen Silvia Children's Hospital, The Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden; Deniz Dogru, Cystic Fibrosis Registry of Turkey, Ankara, Turkey; Halyna Makukh, Institute of Hereditary Pathology Ukrainian National Academy of Medical Sciences, Lviv, Ukraine; Siobhán B. Carr (ECFSPR Scientific Committee), Department of Respiratory Paediatrics, Royal Brompton Hospital, and NHLI, Imperial College, London, United Kingdom; and Rebecca Cosgriff, Cystic Fibrosis Trust, London, United Kingdom.
Conflict of interest: The authors declare no conflict of interest.
- Received January 10, 2023.
- Accepted February 2, 2023.
- Copyright ©The authors 2023
This version is distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0. For commercial reproduction rights and permissions contact permissions{at}ersnet.org
References
