Abstract
Background Progressive fibrosing interstitial lung disease (PF-ILD) is characterised by progressive physiological, symptomatic and/or radiographic worsening. The real-world prevalence and characteristics of PF-ILD remain uncertain.
Methods Patients were enrolled from the Canadian Registry for Pulmonary Fibrosis between 2015 and 2020. PF-ILD was defined as a relative forced vital capacity (FVC) decline ≥10%, death, lung transplantation or any two of: relative FVC decline ≥5% and <10%, worsening respiratory symptoms or worsening fibrosis on computed tomography of the chest, all within 24 months of diagnosis. Time-to-event analysis compared progression between key diagnostic subgroups. Characteristics associated with progression were determined by multivariable regression.
Results Of 2746 patients with fibrotic ILD (mean±sd age 65±12 years; 51% female), 1376 (50%) met PF-ILD criteria in the first 24 months of follow-up. PF-ILD occurred in 427 (59%) patients with idiopathic pulmonary fibrosis (IPF), 125 (58%) with fibrotic hypersensitivity pneumonitis (HP), 281 (51%) with unclassifiable ILD (U-ILD) and 402 (45%) with connective tissue disease-associated ILD (CTD-ILD). Compared with IPF, time to progression was similar in patients with HP (hazard ratio (HR) 0.96, 95% CI 0.79–1.17), but was delayed in patients with U-ILD (HR 0.82, 95% CI 0.71–0.96) and CTD-ILD (HR 0.65, 95% CI 0.56–0.74). Background treatment varied across diagnostic subtypes, with 66% of IPF patients receiving antifibrotic therapy, while immunomodulatory therapy was utilised in 49%, 61% and 37% of patients with CHP, CTD-ILD and U-ILD, respectively. Increasing age, male sex, gastro-oesophageal reflux disease and lower baseline pulmonary function were independently associated with progression.
Conclusions Progression is common in patients with fibrotic ILD, and is similarly prevalent in HP and IPF. Routinely collected variables help identify patients at risk for progression and may guide therapeutic strategies.
Abstract
In the setting of fibrotic ILD, disease progression was observed in 50% of prospectively evaluated patients at 24 months. Highest rates were seen in those with IPF (59%) and HP (58%), followed by U-ILD (51%) and CTD-ILD (45%). https://bit.ly/3v7T9ux
Introduction
Fibrotic interstitial lung diseases (ILDs) are a spectrum of lung disorders characterised by fibrosis of the lung parenchyma. Fibrosis represents a final common pathway for conditions that can originate through distinct pathophysiological mechanisms, including autoimmunity, granulomatous inflammation, organic and inorganic dust exposure, and other insults [1]. Such triggers precipitate the activation of fibroblasts and myofibroblasts, leading to exuberant extracellular matrix deposition and the subsequent fibrotic remodelling of the lung parenchyma. Among other risk factors, genetic predisposition and ageing-related biological mechanisms appear to affect the fibrogenic response in the lungs independent of the initial cause [2, 3].
An important subset of patients with fibrotic ILD experience progressive clinical, physiological and radiographic decline, with an associated reduction in quality of life and survival despite conventional therapies. Idiopathic pulmonary fibrosis (IPF) is often described as the prototypical fibrotic ILD; however, other ILD subtypes can have a similar poor prognosis [1]. Furthermore, the prevalence of the PF-ILD phenotype in a modern IPF cohort, managed with antifibrotic therapy, has not been robustly evaluated to date.
The INBUILD trial demonstrated the efficacy of the tyrosine kinase inhibitor nintedanib to attenuate the rate of forced vital capacity (FVC) decline in patients with non-IPF PF-ILD [4]. The rate of FVC decline measured in this trial was comparable to that observed in patients with IPF based on a comparative analysis of the placebo arms of INBUILD with INPULSIS (a randomised controlled trial studying the effect of nintedanib in IPF) [5]. Given the strength of this collective evidence, nintedanib has been approved by many regulatory bodies for patients with PF-ILD. Outside of the constraints of a clinical trial, however, robust data regarding the epidemiology and natural history of the PF-ILD phenotype are limited, and external validation in prospective cohorts is required. In a recent retrospective, single-centre analysis, Nasser et al. [6] reported a PF-ILD prevalence of 27.2% in a non-IPF ILD population. Similarly, survey data from multiple countries estimate that progressive fibrosis may occur in 14–32% of patients with non-IPF ILD [7, 8].
Our study aims to evaluate the prevalence, clinical characteristics and outcomes of the PF-ILD phenotype, and its individual components, in a national, multicentre, prospective fibrotic ILD registry. We sought to identify baseline factors associated with the PF-ILD phenotype that will better inform clinical decision making for patients with fibrotic ILD.
Methods
Study population
Patients enrolled in the Canadian Registry for Pulmonary Fibrosis (CARE-PF) were studied [9]. CARE-PF is a prospective cohort of patients with fibrotic ILD of any subtype, recruited from eight specialised ILD centres, who are ≥18 years old, and able to provide consent and complete questionnaires in English or French. All patients in the registry were eligible for inclusion, starting from the date of enrolment of the first participant (November 2015) to the date of data extraction (December 2020). Ethics approval was obtained by the research ethics boards at each participating site. Informed consent was obtained from patients at the time of study enrolment.
Data collection and measurements
Baseline characteristics were collected at enrolment into CARE-PF, and included details on demographics, medical history, smoking history, medication use and family history of ILD, determined by robust clinical chart review and self-reported patient questionnaire. Lung function parameters including FVC (L), forced expiratory volume in 1 s (FEV1 (L)) and diffusing capacity of the lung for carbon monoxide (DLCO (mL·min–1·mmHg–1)) were captured serially as clinically indicated. Baseline values nearest to the date of ILD diagnosis were used to calculate the ILD-Gender–Age–Physiology (GAP) score, a validated prognostic risk score for patients with ILD [10]. 6-min walk distance (6MWD) and right ventricular systolic pressure (RVSP) on echocardiography were also collected nearest to the time of diagnosis. Immunomodulatory and antifibrotic medication use or nonuse within 24 months of diagnosis was captured. Date of ILD diagnosis was determined as the date of first evidence of fibrotic ILD on high-resolution computed tomography (HRCT) or the date of surgical lung biopsy confirming ILD diagnosis if performed.
ILD diagnoses were established by the treating ILD specialist. In the event of diagnostic uncertainty, multidisciplinary review was conducted with chest radiologists and, if applicable, lung pathologists. IPF was diagnosed according to guideline criteria available at the time of diagnosis [11, 12]. Fibrotic hypersensitivity pneumonitis (HP) was diagnosed based on clinical history, radiographic pattern and, if applicable, pathological confirmation given the absence of available clinical practice guidelines at the time of patient enrolment. Patients without a confident diagnosis (<50% confidence) were considered to have unclassifiable ILD (U-ILD) [13]. Patients meeting the proposed research criteria for interstitial pneumonia with autoimmune features (IPAF) were also considered to have U-ILD [14]. Connective tissue disease-associated ILD (CTD-ILD) required the confirmation of an underlying CTD that was thought to be associated with the fibrotic ILD. A diagnosis of idiopathic nonspecific interstitial pneumonia (NSIP) required confirmation by surgical lung biopsy [15]. Patients with fibrotic ILD secondary to other causes (e.g. sarcoidosis and asbestosis) were included in the analysis and grouped into a category labelled “Other” ILD.
Outcome assessment
The primary outcome was time to first event meeting PF-ILD criteria within the 24-month time period following ILD diagnosis [4]. PF-ILD events were defined as: a relative FVC decline ≥10%, death, lung transplantation or any two of: relative FVC decline ≥5% and <10%, worsening respiratory symptoms or worsening fibrosis on HRCT. Symptomatic progression was assessed based on the detailed review of all available clinical notes from the patient's clinical chart, and required interpretation and judgement on behalf of the site investigators. Key terms that were assessed included: breathlessness, dyspnoea, shortness of breath, respiratory symptoms, cough, functional capacity, functional ability, exercise capacity, exercise ability, increased oxygen use and increase in Medical Research Council dyspnoea scale to a higher number. A transient episode of clinical worsening <1 month in duration was not considered sufficient to meet this criterion. Patients could only meet the “radiographic progression” criteria in the event that a repeat CT within 24 months of ILD diagnosis showed worsening fibrosis (allowing observations up to 27 months to account for variable follow-up intervals). This was documented in the clinic letters/notes/referrals or in radiology reports. Direct review of the images was at the discretion of the site investigator. Key terms included: worsening fibrosis, honeycombing, interstitial changes, reticulation, architectural distortion and traction bronchiectasis.
We included all-cause mortality and lung transplantation as PF-ILD criteria to account for patients who may have had a rapid clinical deterioration that was not captured by serial physiological/clinical/radiographic assessment. The FVC measurement nearest to the ILD diagnosis date was used as the reference point for determining FVC decline. Meeting the death or transplant criterion only applied to those not previously meeting any other PF-ILD criteria. The remaining patients were classified as nonprogressors.
Statistical analysis
Descriptive analyses of patient characteristics were assessed using standard summary statistics. Differences in baseline characteristics between PF-ILD and nonprogressors were compared using the Chi-squared test for categorical variables, by the t-test for normally distributed variables and by the Mann–Whitney test for nonnormally distributed continuous variables. Time-to-event models, to determine time to progression from diagnosis, were constructed using Cox proportional hazards models. Exploratory analyses were conducted to identify factors associated with PF-ILD. Unadjusted analyses followed by multivariable analysis were performed including age, sex, ethnicity, smoking history, family history, comorbidities, history of surgical lung biopsy and baseline pulmonary function testing as covariates. Thresholds used to categorise physiological variables were based on guideline recommendations and key values derived from the existing fibrotic ILD literature. The relationship between ILD diagnosis and time to PF-ILD event was evaluated by Kaplan–Meier time-to-event curves. The relative contribution of each component of the PF-ILD definition was also assessed. A sensitivity analysis excluded mortality and lung transplantation in the criteria for PF-ILD. The proportion of patients excluded from the analysis due to missing data was compared across ILD diagnoses to determine if missing data were balanced across these subgroups. Subgroup analyses were performed to identify variables associated with time to progression for individual ILD subtypes. Statistical analyses were performed using Stata version 15 (StataCorp, College Station, TX, USA).
Results
Baseline characteristics and incidence of PF-ILD
In total, 2746 patients (mean±sd age 65±12 years; 51% female) had fibrotic ILD with data available for assessment of PF-ILD as defined earlier. Criteria for PF-ILD were met in 1376 (50%) within 24 months of diagnosis, including 59% of all patients with IPF, 58% with fibrotic HP, 51% with U-ILD and 45% with CTD-ILD. Patients with diagnoses other than these major categories were least likely to show progression (39%). Table 1 displays and compares the baseline characteristics of PF-ILD and nonprogressors.
Baseline characteristics
Contribution of individual components of the PF-ILD definition
The PF-ILD phenotype was most commonly established based on the presence of an FVC decline ≥10% over 24 months (675 PF-ILD patients (49%)). Death occurred in 61 patients who did not meet any other PF-ILD criteria prior to their death, accounting for 4% of PF-ILD cases. Contributions of the other criteria are outlined in table 2. There were 85 patients classified as PF-ILD using symptom and radiographic progression criteria who were missing serial FVC data.
Individual progressive fibrosing interstitial lung disease (PF-ILD) criteria met within 24 months of ILD diagnosis
Clinical characteristics of PF-ILD
Compared with nonprogressors, patients with PF-ILD were slightly older, more often male, had a higher cumulative pack-year smoking history in ever-smokers, were more likely to have a history of coronary artery disease or gastro-oesophageal reflux disease (GORD), and had lower baseline FVC % pred, DLCO % pred and 6MWD. In the 1140 patients with echocardiographic data, patients with PF-ILD had higher median RVSP. Baseline ILD-GAP scores were higher in patients with PF-ILD (table 1).
Table 3 describes the distribution of PF-ILD by underlying diagnosis. In the CTD-ILD group, criteria for PF-ILD were met in a similar percentage of patients with systemic sclerosis, rheumatoid arthritis, myositis, undifferentiated CTD and mixed CTD (42–49% of patients progressed). Progression was less common in patients with Sjögren syndrome and systemic lupus erythematosus (25–37% of patients progressed). Among patients with other types of fibrosing ILD, those with idiopathic NSIP, occupational ILD and smoking-related ILD had higher rates of progression (41–56%) compared with those with sarcoidosis and drug-induced ILD (31–32%). Supplementary table S1 details the distribution of immunosuppressive and antifibrotic use among diagnostic subgroups. As expected in a real-world population, treatment varied across diagnostic subtypes. Antifibrotic therapy was only utilised in the setting of IPF where 66% of patients received therapy with either nintedanib or pirfenidone. Immunomodulatory therapy was utilised in 49%, 61% and 37% of patients with CHP, CTD-ILD and U-ILD, respectively. Statistical analyses further exploring these findings were not performed given the presence of significant confounding by indication.
Progressive fibrosing interstitial lung disease (PF-ILD) by diagnosis
Factors associated with progression in PF-ILD
Compared with patients with IPF, time to progression was similar in HP (hazard ratio (HR) 0.96, 95% CI 0.79–1.17), but was delayed in CTD-ILD (HR 0.65, 95% CI 0.56–0.74) and U-ILD (HR 0.82, 95% CI 0.71–0.96) (table 4). Kaplan–Meier curves for risk of progression are shown in figure 1. Progression rates were similar for all ILD subtypes in a sensitivity analysis that excluded death within 24 months as a PF-ILD event (supplementary table S2). There were 219 patients excluded from the analysis due to missing data (supplementary figure S1), with missingness balanced across ILD subtypes (supplementary table S3).
Hazard ratios (HRs) for progression to progressive fibrosing interstitial lung disease (PF-ILD) by diagnosis
Kaplan–Meier curves for risk of progression. IPF: idiopathic pulmonary fibrosis; CTD: connective tissue disease; ILD: interstitial lung disease.
Variables associated with progression on unadjusted analysis included increasing age, male sex, higher pack-year smoking history, history of GORD, and reduced baseline FVC and DLCO (table 5). The median time from initial lung function measurement to baseline time-point was 19 days. In a multivariable model, increasing age, male sex, history of GORD, and reduced baseline FVC <70% predicted and DLCO <75% predicted remained associated with progression. When assessing factors associated with progression, there was no detectable difference in the rate of progression comparing patients with HP who had or did not have an identifiable exposure. Similar results were observed across all relevant diagnostic subgroups.
Unadjusted and multivariable analyses evaluating risk factors for progressive fibrosing interstitial lung disease
Discussion
This study represents the largest analysis evaluating ILD progression, and the PF-ILD phenotype, across the spectrum of all fibrotic ILDs. Our results show that progression of fibrotic ILD, as defined by clinical, radiographic and physiological criteria, occurs in ∼50% patients at 24 months, with the highest rates in those with IPF and HP, followed by U-ILD and CTD-ILD. Variables associated with progression include increasing age, male sex, a history of GORD, baseline FVC <70% predicted and baseline DLCO <75% predicted.
We applied pragmatic criteria to define progression, similar to what was previously used in the INBUILD clinical trial, which demonstrated the efficacy of nintedanib in attenuating the rate of FVC decline in the PF-ILD population [4]. Mortality and lung transplantation were selected as PF-ILD criteria in order to account for patients who may have had a rapid clinical deterioration that was not captured by serial physiological/clinical/radiographic assessment, in order to clearly capture our primary intent of describing disease behaviour in the setting of fibrotic ILD. The prevalence of PF-ILD in our cohort was 50% at 2 years, greater than that reported by Nasser et al. [6] in a recent publication from a large European centre that applied comparable criteria to define PF-ILD. In their analysis, Nasser et al. [6] reported that 168 out of 617 patients (27%), assessed over a 7-year period, met PF-ILD criteria. Key differences that distinguish our CARE-PF cohort from this previous analysis include CARE-PF's design as a prospective multicentre study, inclusion of patients with IPF in the analysed cohort, inclusion of death and lung transplantation within 24 months as a PF-ILD event, and inclusion of patients managed with off-label antifibrotic therapy. Another retrospective study, conducted across nine specialist centres in the UK by Simpson et al. [8], applied the INBUILD PF-ILD definition to all new incident cases of non-IPF fibrotic ILD assessed over a 2-year period starting in 2017. The authors identified 1749 patients with non-IPF fibrotic ILD, of whom 14.5% met INBUILD PF-ILD criteria. They similarly found progression to be most common in HP, followed by U-ILD and then CTD-ILD [8]. Other reports assessing PF-ILD have used varying definitions and follow-up periods, and have often studied specific diseases rather than the spectrum of all fibrotic ILDs, limiting comparisons across ILD subtypes [16]. International surveys have estimated the real-world prevalence of non-IPF PF-ILD to be in the range of 18–32% [7].
It is widely accepted that IPF is the prototypical PF-ILD. Rates of progression have been estimated to be as high as 95%, although such estimates use varying criteria and timelines to define progression [1]. Our prospective longitudinal data demonstrate the prevalence of PF-ILD in our IPF population is much lower at only 59% within 24 months of the time of diagnosis. Although somewhat surprising, these data speak to the clinical heterogeneity of real-world populations, most notably our as-treated IPF population, the majority of whom had received antifibrotic therapy at some point in their disease course. These data provide novel insights of the natural history of a contemporary IPF cohort, the relevance of which is heightened as we move past the era of placebo-controlled trials in fibrotic lung disease. Even after excluding patients with IPF, however, we found that progression occurred in 46% of non-IPF patients managed with conventional therapies, as outlined in supplementary table S1. The rate of PF-ILD was greatest in the patient population with fibrotic HP (58%), followed by U-ILD (51%), CTD-ILD (45%) and other ILDs (31–56%). Within the CTD-ILD group, patients with systemic sclerosis demonstrated the highest rate of progression (49%), similar to previous estimates [17, 18]. The comparable nature of these prevalence data to the IPF population emphasises the critical importance of identifying the PF-ILD phenotype across the spectrum of fibrotic lung disease.
Independent risk factors for progression included increasing age, male sex, history of GORD, baseline FVC <70% predicted and baseline DLCO <75% predicted. The highest risk was observed in those patients with the most compromised lung function. One notable difference between the prognostic risk factors assessed in the ILD-GAP index and the risk factors identified in our study is that HP had similar risk of progression compared with IPF and U-ILD. Prospective validation is required to further delineate the relevance of this finding. Although we have identified clinical factors associated with increased progression, there are likely additional factors that further contribute to this risk. Other factors, including genetic predisposition, molecular signatures and undocumented environmental exposures, are likely of importance, and represent an area of evolving research and understanding. This is particularly relevant as it relates to the development of reliable biomarkers that predict the PF-ILD phenotype [19, 20].
Several criteria have been used to define PF-ILD [21]. Our study incorporated physiological, symptomatic and radiographic worsening, comparable to the definition used in the INBUILD trial [4]. Other trials have used different criteria to define PF-ILD. Two recent studies have assessed the role of pirfenidone in reducing disease progression in fibrosing ILD and defined PF-ILD by an absolute FVC decline of ≥5% on at least three measurements over 6–24 months [22], or defined PF-ILD in patients with U-ILD as an absolute FVC decline of ≥5% or symptomatic worsening within a 6-month period [23]. Strong trends towards reducing FVC progression with pirfenidone were observed in both studies. Such encouraging results, together with the INBUILD study, emphasise the critical importance of identifying the PF-ILD phenotype and the associated therapeutic implications. For the purpose of our study, we used a definition of PF-ILD similar to that described in the INBUILD trial, providing an external and real-world application of this definition. A relative decline in FVC ≥10% over 2 years was the primary factor defining progression (49%) in our population, similar to the percentage that was reported in the INBUILD study [4]. Consensus regarding the optimal criteria for PF-ILD remains to be determined.
The results of our study are limited by factors mostly relating to the use of registry data. First, there were 219 patients excluded from our study due to the unavailability of progression data within 2 years of ILD diagnosis. These missing data were balanced across diagnostic subgroups and thus less likely to bias comparisons of risk of progression of any particular ILD diagnosis. As standard practice in Canada involves the routine assessment of patients with fibrotic ILD at 3–6 months intervals, we do not feel that patients with a milder phenotype of disease were preferentially excluded from the analysis [24]. Second, relevant criterion such as acute exacerbation of ILD and respiratory death were not included in the PF-ILD definition. Given Canada's large geographic area, patients travel large distances to access specialty care. As such, data relating to cause of death, hospitalisation and acute exacerbation, from sites remote to the study centre, are extremely difficulty to capture reliably and accurately. Third, evidence of progression was only assessed up to 24 months following diagnosis and prolongation of follow-up would lead to increased prevalence of meeting PF-ILD criteria over time. The frequency of this long-term progression is worthy of further evaluation in longer-term cohorts. Fourth, although we collected information on the use of immunosuppression and antifibrotic therapy, we did not pursue cause–effect analyses due to the certainty of confounding by indication and challenges in analysing such data in a retrospective cohort. Our results should therefore be considered applicable to similar “as-treated” real-world populations. As patients in our registry were recruited from tertiary care academic referral centres, it is possible that referral bias may have led to an overestimation of the prevalence of PF-ILD. Such bias is commonly encountered in ILD cohorts, given the subspecialty nature of disease management, and has influenced our traditional understanding of the natural history of IPF. The relatively low rates of progression observed in our IPF population, however, suggest that the influence of this inherent bias was minimised.
Conclusions
Progression is common in fibrotic ILD, regardless of the underlying mechanism and trigger for lung injury, although with a lower frequency of progression in a real-world as-treated population of patients with IPF compared with conventional wisdom. Our results provide real-world context to the previously described pragmatic criteria for assessing progression that are based on serial assessment of FVC decline, worsening symptoms and radiographic progression; variables that are routinely collected in clinical practice. Future studies identifying additional risk factors for progression such as genetic and molecular profiles are required to better characterise risk in individual patients and further inform management decisions.
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 material ERJ-02571-2021.SUPPLEMENT
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-02571-2021.Shareable
Acknowledgements
We would like to acknowledge the commitment and enormous contributions of our collaborative research team, especially the research coordinators who work tirelessly to ensure the success of our project. We would also like to thank our patients and their families; their incredible commitment and devotion is inspiring, without which this research would not be possible.
Footnotes
This article has an editorial commentary: https://doi.org/10.1183/13993003.01449-2022
Author contributions: N. Hambly, M.M. Farooqi, A. Dvorkin-Gheva, C.J. Ryerson and M. Kolb contributed to the conception and design, acquisition of data, analyses and interpretation of the data, drafted the article, revised it critically for important intellectual content, and gave final approval of the version to be published. K. Donohoe, K. Garlick, C. Scallan, S.G. Chong, S. MacIsaac, D. Assayag, K.A. Johannson, C.D. Fell, V. Marcoux, H. Manganas, J. Morisset, A. Comes, J.H. Fisher, S. Shapera, A.S. Gershon, T. To, A.W. Wong, M. Sadatsafavi, P.G. Wilcox, A.J. Halayko, N. Khalil, G. Cox and L. Richeldi contributed to the conception and design, acquisition of local data, revised the drafted manuscript critically for important intellectual content, and gave final approval of the version to be published.
Conflict of interest: M.M. Farooqi reports no conflicts of interest relevant to this manuscript.
Conflict of interest: A. Dvorkin-Gheva reports no conflicts of interest relevant to this manuscript.
Conflict of interest: K. Donohoe reports no conflicts of interest relevant to this manuscript.
Conflict of interest: K. Garlick was an employee for the medical department of Boehringer Ingelheim at the time of the preparation of this manuscript.
Conflict of interest: C. Scallan reports personal fees from Boehringer Ingelheim.
Conflict of interest: S.G. Chong reports no conflicts of interest relevant to this manuscript.
Conflict of interest: S. MacIsaac reports no conflicts of interest relevant to this manuscript.
Conflict of interest: D. Assayag reports personal fees from Roche and Boehringer Ingelheim.
Conflict of interest: K.A. Johannson reports grants from Boehringer Ingelheim, Pulmonary Fibrosis Society of Calgary and University of Calgary School of Medicine; personal fees from Boehringer Ingelheim, Roche, Three Lakes Foundation, Pliant Therapeutics, Theravance and Blade Therapeutics.
Conflict of interest: C.D. Fell reports educational grants and research grants from Boehringer Ingelheim, Roche and the Canadian Pulmonary Fibrosis Foundation; personal fees from Roche and Boehringer Ingelheim; Chair of the Boards for the Canadian Pulmonary Fibrosis Foundation.
Conflict of interest: V. Marcoux reports grants from Boehringer Ingelheim, AstraZeneca and Roche; personal fees from Boehringer Ingelheim and Roche.
Conflict of interest: H. Manganas reports grants from Boehringer Ingelheim and Gilead.
Conflict of interest: J. Morisset reports personal fees from Boehringer Ingelheim and Roche.
Conflict of interest: A. Comes reports no conflicts of interest relevant to this manuscript.
Conflict of interest: J.H. Fisher reports no conflicts of interest relevant to this manuscript.
Conflict of interest: S. Shapera reports educational grants and research grants from Boehringer Ingelheim, Roche and the Canadian Pulmonary Fibrosis Foundation; personal fees from AstraZeneca, Boehringer Ingelheim Canada and Hoffman La-Roche Canada; served on the medical advisory board for the Canadian Pulmonary Fibrosis Foundation.
Conflict of interest: A.S. Gershon reports a research grant from the Canadian Pulmonary Fibrosis Foundation.
Conflict of interest: T. To reports no conflicts of interest relevant to this manuscript.
Conflict of interest: A.W. Wong reports personal fees from AstraZeneca and Boehringer Ingelheim.
Conflict of interest: M. Sadatsafavi reports grants from Boehringer Ingelheim; personal fees from Boehringer Ingelheim.
Conflict of interest: P.G. Wilcox reports personal fees from Boehringer Ingelheim.
Conflict of interest: A.J. Halayko reports no conflicts of interest relevant to this manuscript.
Conflict of interest: N. Khalil reports no conflicts of interest relevant to this manuscript.
Conflict of interest: G. Cox reports personal fees from Boehringer Ingelheim and Roche.
Conflict of interest: L. Richeldi reports grants and personal fees from Boehringer Ingelheim; personal fees from Biogen, Sanofi-Aventis, Celgene, RespiVant, CSL Behring, Nitto, Pliant Therapeutics, Cipla, Zambon, Promedior, Boehringer Ingelheim, Roche and Fibrogen.
Conflict of interest: C.J. Ryerson reports grants from Boehringer Ingelheim; personal fees from Boehringer Ingelheim, Roche, Pliant Therapeutics, Cipla and Veracyte.
Conflict of interest: M. Kolb reports grants from the Canadian Institute for Health Research, Roche, Boehringer Ingelheim, Pieris and Prometic; personal fees from Boehringer Ingelheim, Roche, European Respiratory Journal, Belerophon, United Therapeutics, Nitto Denko, MitoImmune, Pieris, AbbVie, DevPro Biopharma, Horizon, Algernon and CSL Behring.
Conflict of interest: N. Hambly reports grants from Roche and Boehringer Ingelheim; personal fees from Roche, Boehringer Ingelheim and Janssen.
Support statement: CARE-PF is funded by Boehringer Ingelheim. The study sponsor did not have any input on the study design, access to data or interpretation of the results for this substudy. Funding information for this article has been deposited with the Crossref Funder Registry.
- Received September 27, 2021.
- Accepted February 17, 2022.
- Copyright ©The authors 2022. For reproduction rights and permissions contact permissions{at}ersnet.org
References
