Abstract
Introduction
Idiopathic pulmonary fibrosis (IPF) has high morbidity and mortality with limited treatment options. Goal-oriented management approaches, such as the “treatable traits” concept, have yet to be implemented in IPF. This study aims to identify specific treatment goals in IPF for potential interventions by analysing outcomes from two national registries.
Methods
We used data from the INSIGHTS-IPF registry, comprising 1232 IPF patients, enrolled from 2014 to 2020 as derivation cohort. Baseline and 6-month follow-up data were examined to assess clinical progression and predict 1-year mortality. Variables included forced vital capacity (FVC), diffusing capacity of the lung for carbon monoxide (DLCO), 6-min walk distance (6MWD), body mass index (BMI) and comorbidities. For validation we used data from 490 IPF patients enrolled in the Canadian Registry for Pulmonary Fibrosis (CARE-PF) and the full 2576 fibrotic interstitial lung disease (ILD) cohort of the CARE-PF registry.
Results
Multivariable analysis identified FVC, DLCO, 6MWD and BMI as independent predictors of 1-year survival. We established three risk groups based on these variables: low risk (<15% 1-year mortality), intermediate risk (15–30%) and high risk (>30%). Potential treatment goals were defined based on FVC, DLCO, 6MWD and BMI, which are readily available and may be responsive to interventions. Our risk model showed equivalent accuracy in the validation cohort, both for IPF alone and the overall population.
Conclusion
This study provides a novel risk model for IPF patients, which may also apply to a broader spectrum of fibrotic ILD. It is based on potentially actionable variables which deserve further evaluation as measurable treatment goals in interventional clinical trials.
Shareable abstract
Novel risk stratification model for fibrotic ILD based on potentially actionable variables beyond pulmonary function parameters may help maintain or achieve a low-risk status, enabling a more active treatment strategy https://bit.ly/4lWsb0b
Introduction
Idiopathic pulmonary fibrosis (IPF) is a rare, progressive and devastating lung disease characterised by irreversible scarring of lung tissue, leading to impaired lung function, reduced exercise capacity, impaired quality of life and premature death [1, 2].
While significant advancements have been made in understanding the pathogenesis and natural history of IPF, treatment options remain limited and the disease continues to carry a poor prognosis [1]. Patients with progressive pulmonary fibrosis (PPF) other than IPF are assumed to share similar pathogenetic mechanisms and do have an equally high mortality [3, 4]. Antifibrotic agents slow disease progression and potentially improve survival, but they do not provide a cure nor completely halt the fibrotic process [5]. Lung transplantation may be considered for selected IPF patients with advanced disease; however, this option is limited by donor availability and eligibility criteria [5]. Additional components of IPF management recommended by guidelines include general measures such as long-term oxygen therapy, pulmonary rehabilitation, comorbidity management and palliative care [5, 6].
Unlike management of IPF and PPF, the field of pulmonary hypertension has demonstrated the impact of risk stratification and goal-oriented therapy on patient management, leading to improved outcomes and personalised care [7–9]. This paradigm shift towards goal-oriented management has not yet been widely adopted in the realm of IPF, leaving clinicians with a substantial unmet need for a more precise and tailored approach to address the diverse and complex clinical manifestations of the disease. While risk scores for IPF abound, a goal-oriented management approach and the concept of “treatable traits” have not been established [10–19].
This study aims to define distinct risk groups based on 1-year mortality rates which could be used as a starting point for future interventional studies to establish a goal-oriented management framework across different risk profiles. The risk model is based on outcomes of two national registries (the INSIGHTS-IPF registry and the Canadian Registry for Pulmonary Fibrosis (CARE-PF)), focusing on changes occurring at 6 months that predict all-cause mortality at 1 year and that are accessible by an active, goal-oriented management strategy.
Methods
Study design and data source
The INSIGHTS-IPF registry is an investigator-initiated, observational cohort. It documented consecutive patients with newly diagnosed (incident) and known (prevalent) IPF receiving routine clinical care at 20 specialised pulmonary centres in Germany between November 2012 and June 2020. The registry's structure, methodology and regulatory aspects have been previously described, along with a comprehensive description of the baseline characteristics of the patient cohort in various publications [20–26]. Patients enrolled in CARE-PF were used for validation of key findings [27].
Ethical considerations
The INSIGHTS-IPF and the CARE-PF Registries have obtained ethical approval from the respective institutional review boards of participating centres, and all patients provided informed consent for data collection and analysis before inclusion into the registry. The present study adheres to the principles outlined in the Declaration of Helsinki and other relevant guidelines for conducting medical research involving human subjects.
Study population
Eligible participants were aged ≥18 years and had received a study site diagnosis of IPF in accordance with the 2011 American Thoracic Society/European Respiratory Society/Japanese Respiratory Society/Latin American Thoracic Association IPF guideline. There were no explicit exclusion criteria, and written informed consent was obtained from all enrolled patients. For the present post hoc analysis, patients were required to have available baseline data and at least one follow-up assessment at 6 months after enrolment or later. CARE-PF is a prospective cohort enrolling patients with fibrotic interstitial lung disease (ILD) including both IPF and non-IPF diagnoses [27, 28].
Data collection and variables
Data were prospectively collected at baseline (enrolment) and subsequently at 6- to 12-month intervals during follow-up visits. All visits between months 4 and 8 were counted as 6-month follow-up and between months 9 and 14 as 12-month follow-up. Each site documented all clinical events, including hospitalisations and acute exacerbations as determined by the treating physician, as well as any deaths occurring during the study period. Pulmonary function tests were performed at each follow-up visit as clinically indicated. These included forced vital capacity (FVC), total lung capacity (TLC, measured by body plethysmography), diffusing capacity of the lung for carbon monoxide (DLCO) and arterial oxygen tension (PaO2). The physician reported comorbid diseases (e.g. pulmonary hypertension, untreated gastrointestinal reflux), symptoms of IPF (e.g. cough, dyspnoea, anxiety, fatigue) and the 6-min walk distance (6MWD) in metres. The date of performance of pulmonary function tests and 6MWD was recorded at each visit. The study protocol included an event visit if a clinical event occurred during follow-up (e.g. death, exacerbation among others) at which the physician reported about the clinical characteristics of the patient before the event. Therefore, follow-up data were also available for patients who died within the first 6 months after enrolment. Body mass index (BMI) was calculated based on the physician-documented weight and height.
Statistical analysis
Patient characteristics were summarised using standard descriptive statistics: absolute and relative frequencies for categorical data, and mean±SD and median for continuous data. Changes in clinical parameters (FVC%, TLC%, DLCO%, PaO2, 6MWD, BMI) were calculated between baseline and the 6-month follow-up visit, or the clinical event visit for patients who died within 6 months after baseline.
Missing follow-up data were imputed using multiple imputation by chained equations (MICE). Predictor variables included sex, age, duration since IPF diagnosis, number of comorbidities, medical therapy, and the baseline values of BMI, FVC%, TLC%, DLCO%, PaO2 and 6MWD. The number of imputations was set to 10. Complete information (of 1232 patients) for BMI (814, 66%), FVC% (905, 73%), TLC% (905, 73%), DLCO% (881, 72%), PaO2 (843, 68%) and 6MWD (894, 73%) was available at 6-months follow-up. Cox proportional hazard models were used to estimate the association of 1-year mortality with potentially modifiable clinical parameters and their changes between baseline and the 6-month follow-up. In the first step, all modifiable clinical parameters were tested separately. For FVC%, TLC%, DLCO%, PaO2, 6MWD and BMI the baseline value and the change between baseline and 6 months were included together in the analyses. The strengths of association of the variables of interest and 1-year mortality was quantified using Harrel's C-statistic.
Model selection was performed using bootstrapping with 1000 samples drawn with replacement. A multivariable Cox proportional hazard model for 1-year mortality was derived using LASSO (least absolute shrinkage and selection operator) regression for each bootstrap sample, including all potential variables. Bootstrap inclusion frequencies were calculated for each regression coefficient included in the Cox proportional hazard models [29, 30].
All parameters with an inclusion frequency of at least 75% were included in the final multivariable Cox proportional hazard model (supplementary table S3). TLC% was not considered in this analysis because of a high multicollinearity with FVC% and DLCO%. The proportional hazard assumption for the final model was tested by Schoenfeld residuals.
Statistical analyses were performed with STATA 14 (Stata Statistical Software: Release 14; StataCorp LP, College Station, TX, USA).
Risk stratification and treatment goals
The predicted probability after fitting the final multivariable Cox proportional hazard model was estimated. The predicted probability was categorised at the 16th, 50th and 84th percentile as suggested by Royston and Altman [30]. The first and second group were combined because of the low number of events within both groups (low-risk group: ≤5% risk for mortality; medium-risk group: 5.1–16.9% risk for mortality; and high-risk group: ≥17% risk for mortality). Receiver operating characteristics curve technique was applied to estimate cut-offs for the selected variables in the final model for the three risk groups (figure 1).
FIGURE 1.
The INSIGHTS-IPF chart for 1-year mortality based on parameters that can be addressed by therapeutic interventions (based on multivariable analysis, table 3). FVC: forced vital capacity; DLCO: diffusing capacity of the lung for carbon monoxide; 6MWD: 6-min walk distance; BMI: body mass index.
In addition, the GAP point-scoring system (GAP index, a clinical prediction tool for the prognosis of patients with IPF) based on sex, age and two lung physiology variables (FVC and DLCO) was calculated for comparing the performance to predict 1-year mortality in our cohort [16]. Patients were categorised as stage I, II or III according to the GAP scoring algorithm. We also calculated the ILD-GAP for patients in the CARE-PF registry that also includes non-IPF ILD [31]. For validation, data of the open-ended cohort study CARE-PF registry were used. In that study, patients with incident and prevalent fibrotic ILD were recruited from specialised ILD clinics in Canada [27]. Recruited patients completed case report forms and clinical examinations at baseline and follow-up visits half-yearly (±3 months). The baseline physician's questionnaire included demographic information, family history of ILD, and information about ILD aetiology and treatments. The patient's questionnaire collected data about comorbidities and quality of life among others. Pulmonary function testing was performed at baseline and follow-up visits and % predicted values were calculated using consistent reference equations. Weight and height were recorded at each pulmonary function test to calculate the BMI. 6MWD was measured according to standard techniques when clinically indicated.
We validated our risk model primarily using data of the IPF population from the CARE-PF registry. In addition, we also applied our risk model to the total cohort including patients with any fibrotic ILD. The validation of our proposed model of parameters that can be addressed by therapeutic interventions in order to predict 1-year mortality was based on the publications of Royston and Altman [30] and Crowson et al. [32], as described in the supplementary material.
Results
Patient population and baseline characteristics
1232 patients of the INSIGHTS-IPF cohort comprised the derivation cohort (table 1, figure 2). The mean age of the study population was 71.7±8.8 years, with a male preponderance (81%), typical for IPF populations. The mean±sd duration of symptoms before the baseline visit was 3.4±4.0 years and time between diagnosis and study enrolment was 1.8±2.9 years. The majority of patients suffered significant functional impairment at enrolment. Detailed baseline characteristics are reported in table 1.
TABLE 1.
Characteristics of patients at inclusion
| Characteristic | INSIGHTS-IPF patients | CARE-PF IPF patients | CARE-PF total |
|---|---|---|---|
| Patients, n | 1232 | 490 | 2576 |
| Male sex, n (%), % missing | 998 (81%), 0% | 372 (76%), 0% | 1303 (51%), 0% |
| Age, years, % missing | 71.1±8.8, 0% | 69.3±8.0, 0% | 64.4±12.3, 0% |
| BMI kg·m−2, % missing | 27.4±4.2, 0.1% | 28.4±4.9, 11% | 28.8±5.5, 9% |
| Smoking status, % missing | 4% | 2% | 2% |
| Never, n (%) | 420 (34%) | 116 (24%) | 1006 (39%) |
| Former stopped, n (%) | 783 (64%) | 360 (73%) | 1439 (56%) |
| Current, n (%) | 29 (2%) | 14 (3%) | 131 (5%) |
| Age at IPF diagnosis years, % missing | 69.4±9.5, 7% | 69.3±8.0, 0% | 62.5±12.3, 0% |
| Duration since IPF diagnosis, years, % missing | 1.8±2.9, 1% | 2.2±0.4, 0% | 2.2±0.4, 0% |
| Antifibrotics, n (%), % missing | 758 (61.5%), 0% | ||
| Current Borg dyspnoea index, % missing | 1.5±2.2, 35% | - | - |
| 6-min walk distance m, % missing | 299.7±180.2, 12% | 396.5±108.4, 72% | 419±105, 68% |
| Current NYHA class, % missing | 54% | ||
| I, n (%) | 65 (12%) | - | - |
| II, n (%) | 227 (43%) | - | - |
| III, n (%) | 216 (41%) | - | - |
| IV, n (%) | 19 (4%) | - | - |
| Lung function test, % missing | |||
| Forced vital capacity | 68.8±19.1, 4% | 77.1±18.6, 3% | 77.8±19.6, 3% |
| Total lung capacity | 70.1±15.9, 4% | 72.5±14.2, 5% | 76.5±16.2, 4% |
| Diffusion capacity for carbon monoxide | 44.9±20.9, 8% | 52.4±16.7, 12% | 59.5±19.1, 17% |
Data are reported as mean±sd or n (%). Lung function tests values are % predicted. Antifibrotics: nintedanib or pirfenidone. BMI: body mass index; IPF: idiopathic pulmonary fibrosis; NYHA: New York Heart Association.
FIGURE 2.
Study flow-chart for a) INSIGHTS-IPF and b) CARE-PF.
Unadjusted predictors of 1-year mortality
A total of 132 patients died (10.8% of 1232) within the first year after enrolment. All potential parameters included in the analysis were selected based on clinical relevance and literature. The unadjusted prediction of 1-year mortality with the baseline values and its change after 6-month follow-up from baseline was determined for each parameter. Greater 1-year mortality was associated with lower baseline values for FVC (hazard ratio (HR) 0.96, 95% CI 0.95–0.97), DLCO (HR 0.95, 95% CI 0.94–0.97), TLC (HR 0.96, 95% CI 0.95–0.97), PaO2 at rest (HR 0.96, 95% CI 0.94–0.98) and 6MWD (HR 0.94, 95% CI 0.93–0.95), but not BMI (HR 0.95, 95% CI 0.55–1.06). Highest c-statistic values were retrieved for FVC, DLCO and 6MWD. Greater 1-year mortality was also associated with 6-month change in FVC (HR 0.97, 95% CI 0.96–0.98), DLCO (HR 0.97, 95% CI 0.95–0.98), TLC (HR 0.96, 95% CI 0.94–0.97), PaO2 (HR 0.98, 95% CI 0.97–1.00) and 6MWD (HR 0.90, 95% CI 0.88–0.92, table 2). This analysis is repeated with categorised change. Change from baseline to 6-month follow-up was categorised into groups of worsening (relative decline by at least 5%), improvement (relative improvement by at least 5%) and stable (relative change between −5% and 5%). Results are reported in supplementary table S2.
TABLE 2.
Prediction of 1-year mortality by parameters that can be addressed by therapeutic interventions
| Survived | Deceased | HR# (95% CI) | p-value | |
|---|---|---|---|---|
| FVC % pred (c=0.69) | ||||
| At baseline (4% missing) | 69.9±18.7 (70.6) | 59.8±18.0 (58.0) | 0.96 (0.95–0.97) | <0.001 |
| Change within 6 months after baseline¶ (27% missing) | −0.3±10.0 (0.0) | −2.3±6.4 (−2.3) | 0.97 (0.96–0.98) | <0.001 |
| DLCO % pred (c=0.71) | ||||
| At baseline (8% missing) | 45.0±20.0 (42.9) | 34.9±18.2 (33.2) | 0.95 (0.94–0.97) | <0.001 |
| Change within 6 months after baseline¶ (28% missing) | −0.8±15.1 (0.0) | −3.6±4.7 (−3.3) | 0.97 (0.95–0.98) | <0.001 |
| TLC % pred (c=0.68) | ||||
| At baseline (4% missing) | 70.6±15.7 (70.4) | 63.7±15.1 (62.5) | 0.96 (0.95–0.97) | <0.001 |
| Change within 6 months after baseline¶ (27% missing) | 0.0±9.1 (0.0) | −3.5±5.6 (−2.7) | 0.96 (0.94–0.97) | <0.001 |
| PaO2 mmHg (c=0.64) | ||||
| At baseline (9% missing) | 69.3±12.0 (69.7) | 64.7±12.9 (63.0) | 0.96 (0.94–0.98) | <0.001 |
| Change within 6 months after baseline¶ (32% missing) | 1.3±10.6 (0.0) | 2.4±8.0 (2.6) | 0.98 (0.97–1.00) | 0.008 |
| 6MWD m (c=0.78) | ||||
| At baseline (12% missing) | 309.9±174.8 (360.0) | 253.4±151.2 (262.0) | 0.94 (0.93–0.95) | <0.001 |
| Change within 6 months after baseline¶ (27% missing) | 37.6±102.9 (7.3) | −12.0±62.5 (−5.0) | 0.90 (0.88–0.92) | <0.001 |
| BMI kg·m−2 (c=0.57) | ||||
| At baseline (0.1% missing) | 27.5±4.2 (27.1) | 26.3±4.4 (26.2) | 0.95 (0.85–1.06) | 0.336 |
| Change within 6 months after baseline¶ (34% missing) | −0.5±1.5 (0.0) | −1.3±2.2 (−0.4) | 0.85 (0.72–1.02) | 0.078 |
| Dyspnoea (c=0.55) at baseline (% missing: 0%) | 912 (82.9%) | 123 (93.2%) | 2.69 (1.36–5.30) | 0.004 |
| Cough (c=0.52) at baseline (0% missing) | 750 (68.2%) | 95 (71.9%) | 1.15 (0.79–1.69) | 0.457 |
| Anxiety (c=0.53) at baseline (0% missing) | 88 (8.0%) | 17 (12.9%) | 1.86 (1.11–3.11) | 0.019 |
| Fatigue (c=0.54) at baseline (0% missing) | 378 (34.4%) | 54 (40.9%) | 1.33 (0.94–1.89) | 0.104 |
| Number of comorbidities (c=0.58) at baseline (0% missing) | 2.3±1.6 (2.0) | 2.8±1.7 (3.0) | 1.19 (1.08–1.31) | <0.001 |
| Pulmonary hypertension (c=0.58) at baseline (0% missing) | 142 (12.9%) | 37 (28.0%) | 2.70 (1.85–3.95) | <0.001 |
| Untreated GI reflux (c=0.51) at baseline (0% missing) | 298 (27.1%) | 41 (31.1%) | 1.12 (0.77–1.62) | 0.547 |
Data are presented as mean±sd (median) or n (%), unless otherwise stated. HR: hazard ratio; c: c-statistics; p50: median; FVC: forced vital capacity; DLCO: diffusing capacity of the lung for carbon monoxide; TLC: total lung capacity; PaO2: arterial oxygen tension; 6MWD: 6-min walk distance; BMI: body mass index; GI: gastrointestinal. #: hazard ratio for one unit change except for 6MWD (per 10 m); ¶: higher values indicate improvement.
Multivariable predictors of 1-year mortality
The identification of a multivariable model was based on variables reported in table 2 and on a bootstrap approach. FVC, DLCO, 6MWD and BMI at baseline and change in DLCO and 6MWD between baseline and 6-month follow-up predicted 1-year mortality (table 3, figure 1). This identified multivariable model showed a good discrimination performance for 1-year mortality (c-statistic 0.83). In addition, the proportional hazard assumption was not violated (p=0.342). The calibration of the score seems to be good (calibration slope of 0.013, p=0.411). We performed a sensitivity analysis in patients with all available information, i.e. a complete case analysis rather than in imputed data. The results were comparable to the main analysis (supplementary table S4). We also performed a sensitivity analysis by starting the survival time at the second assessment of pulmonary function test and 6MWD in order to address a potential immortal time bias. The resulting effect estimates (supplementary table S5) are comparable to the results reported in table 3, except for the non-significant HR for FVC.
TABLE 3.
Multivariable prediction of 1-year mortality by parameters that can be addressed by therapeutic interventions with 1-year mortality (c0.83)
| HR+ (95% CI) | p-value | |
|---|---|---|
| FVC % pred# at baseline | 0.83 (0.73–0.94) | 0.018 |
| DLCO % pred# | ||
| At baseline | 0.79 (0.68–0.92) | 0.017 |
| Change within 6 months after baseline¶ | 0.80 (0.65–0.97) | 0.002 |
| 6MWD | ||
| At baseline | 0.95 (0.96–0.99) | <0.001 |
| Change within 6 months after baseline+ | 0.92 (0.94–0.96) | <0.001 |
| BMI kg·m−2 at baseline¶ | 0.92 (0.96–0.99) | 0.003 |
HR: hazard ratio; FVC: forced vital capacity; DLCO: diffusing capacity of the lung for carbon monoxide; 6MWD: 6-min walk distance; BMI: body mass index. #: hazard ratio for change per 10%; ¶: higher values indicate improvement; +: hazard ratio for one unit change except for 6MWD (per 10 m).
External validation in the CARE-PF cohort
Our multivariable Cox-proportional hazard model was externally validated in the CARE-PF cohort as described in detail in the supplementary material. A total of 490 IPF patients were included in the validation cohort. Baseline characteristics of the validation cohort (CARE-PF, IPF patients and total cohort) are shown in table 1. Firstly, the distribution of the prognostic score, the weighted sum of selected variables in the risk model, weighted by the regression coefficients (logarithm of hazard ratios), were compared between the derivation and validation cohorts. The prognostic scores showed no outliers and a comparable distribution between both cohorts. Secondly, the prognostic score was categorised at the 16th, 50th and 84th centiles. The percentage of patients in the determined risk groups are shown in supplementary table S2 and figure 3. It shows that the prognostic score had a poorer prognosis in CARE-PF than in INSIGHTS-IPF with higher numbers of patients in the medium- and high-risk groups. Thirdly, the discrimination performance by calculating Harrell's c-index was tested and rated as good (c=0.80). Lastly, calibration performance was tested in the validation cohort, the ability of the prognostic score to accurately predict the absolute risk level. The so-called calibration slope was −0.001 (p=0.995), which indicates that our proposed risk score may have a similar effect in IPF patients of CARE-PF. The proportional hazard assumption was also not violated in IPF patients of CARE-PF (p=0.499).
FIGURE 3.
Kaplan–Meier curves for 1-year mortality for each risk group estimated by the 50th and 84th centiles of the Prognostic Index in the derivation cohorts.
Comparison to GAP score
Higher GAP scores were positively associated with 1-year mortality in the INSIGHTS-IPF cohort (HR 1.96, 95% CI 1.68–2.30, p<0.001) and in the CARE-PF IPF cohort (HR 2.17, 95% CI 1.64–2.86, p<0.001). However, the discrimination performance of GAP (c=0.74) was lower than for our model, which showed good discrimination performance in INSIGHTS-IPF (c=0.83), IPF patients in CARE-PF (c=0.80) and the CARE-PF total cohort (c=0.84). In addition, the ILD-GAP was positively associated with 1-year mortality (HR 1.71, 95% CI 1.54–1.91, p<0.001, c=0.85) in the CARE-PF total cohort including non-IPF ILD patients.
Discussion
A more individualised approach should be taken to the management of fibrotic lLD that considers the heterogeneity of the disease and its multifaceted impact on patients’ lives. In this study, we developed a risk stratification model based on the entire data set of 1232 patients from the INSIGHTS-IPF Registry as a derivation cohort. The derived model enables us to differentiate patients into low-, intermediate- and high-risk groups based on specific parameters, including FVC, DLCO, 6MWD and BMI. Moreover, this model was successfully reproduced and validated in the total cohort and in the IPF subgroup and the full cohort of the Canadian CARE-PF registry. Consequently, treatment goals beyond traditional pulmonary function test parameters, including measures to stabilise body weight and enhance exercise tolerance, may be considered actionable and measurable targets for therapeutic interventions in future clinical trials.
Treatment effects of currently approved anti-fibrotic drugs on FVC and DLCO are limited, but stabilisation in a majority and slight improvements in a minority of patients have been documented [33–37].
The general treatment strategy for antifibrotics is to perpetuate therapy as long as tolerability of the drug allows and more or less irrespective of FVC decline during therapy [5]. Recent data, however, suggest that a switch of antifibrotics in case of diseases progression may be more beneficial as previously thought [38–40].
In the light of the data presented here this may provide a basis for future interventional studies at various levels including training measures and pulmonary rehabilitation, nutrition counselling and food supplementation as well as switching of anti-fibrotic drug in patients whose FVC has not stabilised at 6 months. While the data on DLCO changes are limited, it could be used here as a confirmatory parameter for FVC changes.
The 6-min walk test has been used as an outcome parameter in numerous controlled clinical trials in pulmonary fibrosis and has been associated with disease progression and mortality. Data from pulmonary rehabilitation [41–43] clearly demonstrate that it is possible to stabilise and even improve the 6MWD in IPF populations, even with advanced disease [44–47]. Using pulmonary rehab or supervised training programmes to maintain or improve 6MWD may therefore be implemented in the treatment of IPF patients.
Recent data also suggest that IPF patients are at risk to lose weight and develop sarcopenia even on short course [48–50]. Interestingly, in our model a lower BMI at baseline and decline of BMI during the first 6 months (univariate analysis) may indicate a higher risk of deterioration and death. Therefore, our data may suggest that maintaining body composition and weight applying targeted nutritional interventions should be studied in more detail in prospective clinical trials.
It is generally acknowledged that IPF and PPF have many clinical and pathophysiological commonalities. Interestingly, the risk model originally developed in IPF seems to also apply to a large mixed group of prevailing PPF and IPF patients from the Canadian CARE-PF registry. This observation may be expected but needs further exploration in future studies.
Our study has significant limitations as it is retrospective in nature, and despite applying a validation cohort it up to now lacks independent prospective confirmation. It is in the nature of clinical registries that there are missing values and imputation methods unavoidably reduce the robustness of the statistical analysis. Nonetheless, the INSIGHTS-IPF registry and CARE-PF provide a comprehensive and robust database comprising large cohorts of IPF and pulmonary fibrosis patients. By utilising the wealth of data within these registries as discovery and validation cohorts, we sought to investigate the changes that occur at 6 months after baseline and their predictive value for the subsequent course of the disease, focusing variables that are potentially accessible for interventions. Acknowledging the importance of thorough model checking, several validation steps were taken to ensure the robustness of the prediction model. As described in the methods, we assessed linearity, threshold effects and multicollinearity, employing LASSO regression and bootstrapping to select and validate variables. In addition, the external validation of our model in the CARE-PF cohort already demonstrates good performance, suggesting its generalisability across datasets.
While other models, such as the GAP score, focus on risk estimation, our study extends the utility of mortality prediction tools by not only stratifying patients based on risk but also explicitly linking these predictions to potentially actionable and measurable treatment goals, such as stabilising body weight and improving exercise capacity. We demonstrated that changes in 6MWD and BMI at 6 months can influence 1-year mortality. This integration of risk prediction with specific management strategies represents a novel application in IPF care. By providing a structured framework that clinicians can use to guide therapy—whether through rehabilitation, nutritional interventions, or adjustments or switches of antifibrotic treatments—this model may be used as a starting point to shift from static prognostication to dynamic, goal-oriented management. Importantly, we focused on those parameters which showed statistical significance in multivariate testing. However, some of the parameters which were predictors of 1-year mortality in univariate testing might also be treatable traits, e.g. dyspnoea or anxiety, and should be included in future studies.
Obviously, the parameters and potential treatment goals discussed here are not novel and have been studied in isolation before. But looking at these parameters in combination may change the spirit of how physicians judge whether a treatment is appropriate and successful or not. While stabilisation of FVC and DLCO has been the prioritised treatment goal in the past, our data suggest that exercise tolerance and BMI might be equally relevant, warranting confirmation in well-designed prospective interventional trials to confirm the validity of this new, more comprehensive management approach.
The INSIGHTS-IPF registry is a cohort study with potential sources of bias. The risk period for death is defined from enrolment to 12 months follow-up. Patients who died within the first 6 months after inclusion have no measure of change in pulmonary function test and 6MWD by definition. However, follow-up assessments of these parameters existed for 85% of patients who died, but the time between enrolment and follow-up is <6 months for patients who died within the first 6 months after enrolment. Multiple imputation with chained equations was performed to estimate missing values in follow-up. Preliminary analyses showed that the missing data occurred at random and depended on disease characteristics that were included in the imputation model. However, we cannot fully exclude the possibility that missing data depended on an unmeasured variable in our cohort. Multiple imputation of follow-up data may inflate the discriminant performance of the prediction model. We re-analysed our model by categorising the change from baseline to 6 months follow-up at 5% thresholds (relative decline by at least 5%, relative improvement by at least 5% and relative change between −5% and 5%). These cut-offs were selected based on the distribution of the change scores and do not present minimal clinical important differences.
In conclusion, our study provides a novel risk stratification model which offers a roadmap to a first goal-oriented treatment approach in IPF, which may also apply to PPF. Besides physiological parameters (i.e. FVC, DLCO), exercise tolerance and nutritional status contribute to the overall mortality risk. Consequently, besides anti-fibrotic drugs, physical and nutritional interventions should be included in future clinical trials aiming to achieve and maintain a low-risk status for fibrotic ILD patients including IPF and PPF and establishing a more active treatment strategy.
Acknowledgements
The authors thank the patients for their participation in the registry.
Footnotes
Provenance: Submitted article, peer reviewed.
This study is registered at www.clinicaltrials.gov with identifier number NCT01695408.
Ethics statement: The study materials were approved by the Ethics Committee of the Medical Faculty, Technical University of Dresden, on 15 September 2012, and by further local ethic committees as per local requirements. All patients provided informed consent before they were included in the study.
Conflict of interest: J. Behr reports grants from Atemweg-Foundation, BMBF, DFG and Boehringer Ingelheim; and consultancy fees from AstraZeneca, Biogen, BMS, Boehringer Ingelheim, Ferrer, Galapagos, Gossamer Bio, Johnson&Johnson, Novartis, Pliant, Pulmovant, Roche and Sanofi-Genzyme. A. Prasse reports grants from Roche/InterMune and Boehringer Ingelheim. H. Wirtz reports consultancy fees from Boehringer Ingelheim, AstraZeneca, Insmed and Novocure. D. Koschel payment or honoraria for lectures, presentations, manuscript writing or educational events from AstraZeneca, Boehringer Ingelheim, Pharma GmbH, Chiesi, GSK, Novartis, Roche Pharma and Sanofi Aventis; and participation on a data safety monitoring board or advisory board with Boehringer Ingelheim. D. Pittrow reports consultancy fees from Actelion, Bayer, Boehringer Ingelheim, Sanofi, Biogen, Shield and MSD. S. Andreas reports support for the present study from Boehringer Ingelheim, and consultancy fees from Boehringer Ingelheim and Roche; and is an associate editor of this journal. C. Grohé reports payment or honoraria for lectures, presentations, manuscript writing or educational events from AstraZeneca, Boehringer Ingelheim, Chiesi, GSK, MSD, Janssen and Sanofi. H. Wilkens reports consultancy fees from AOP, Bayer HealthCare, Biotest, Boehringer Ingelheim, GSK, Janssen and MSD. L. Hagmeyer reports consultancy fees from Boehringer Ingelheim, Roche, BMS and Pfizer. D. Skowasch reports payment or honoraria for lectures, presentations, manuscript writing or educational events from AstraZeneca, Boehringer Ingelheim, Chiesi, GSK, MSD, Janssen and Sanofi. J.F. Meyer reports payment or honoraria for lectures, presentations, manuscript writing or educational events from Actelion, Janssen-Cliag and Pfizer. S. Gläser reports payment or honoraria for lectures, presentations, manuscript writing or educational events from Boehringer Ingelheim, AstraZeneca and Berlin Chemie. C. Neurohr reports payment or honoraria for lectures, presentations, manuscript writing or educational events from Boehringer Ingelheim and Roche Pharma; and participation on advisory boards with Boehringer Ingelheim and Roche Pharma. M. Schwaiblmair reports payment or honoraria for lectures, presentations, manuscript writing or educational events from AstraZeneca, Bayer, Berlin-Chemie, Boehringer Ingelheim, Chiesi, CSL Behring, GSK, Grifols, Janssen-Cilag, Novartis, Pfizer and Sanofi. K. Buschulte reports grants from Sarkoidose-Netzwerk e.V; and payment or honoraria for lectures, presentations, manuscript writing or educational events from Boehringer Ingelheim, Astra Zeneca and Chiesi. T. Veit reports consultancy fees from Boehringer-Ingelheim, Roche and MSD. C.J. Ryerson reports consultancy fees from AstraZeneca, Avalyn, Boehringer Ingelheim, Cipla Ltd, Pliant Therapeutics, Trevi Therapeutics and Veracyte. K.A. Johannson reports consultancy fees from Abbvie, Boehringer Ingelheim, Pliant Therapeutics and Hoffman La Roche. V. Marcoux reports grants from AstraZeneca, Boehringer Ingelheim and the University of Saskatchewan Respiratory Research Center; and consultancy fees from Boehringer Ingelheim and Roche. J.H. Fisher reports consultancy fees from AstraZeneca and Boehringer Ingelheim; and participation on the Medical Advisory Board of the Canadian Pulmonary Fibrosis Foundation. D. Assayag reports grants from the Canadian Institute for Health Research and Boehringer Ingelheim; and consultancy fees from Avalyn and Boehringer Ingelheim. H. Manganas reports consultancy fees from Boehringer Ingelheim; and payment or honoraria for lectures, presentations, manuscript writing or educational events from Boehringer Ingelheim. M. Kolb reports grants from Canadian Institute for Health Research, Boehringer Ingelheim, United Therapeutics and Structure Therapeutics; and consultancy fees from Boehringer Ingelheim, Roche, European Respiratory Journal, Fortrea, United Therapeutics, Abbvie, Avalyn, DevPro Biopharma, Horizon/Amgen, CSL Behring, AZ, GSK, Sanofi, Structure Therapeutics and Glenmark. M. Kreuter reports grants from Roche/InterMune, Boehringer Ingelheim, AstraZeneca, Chiesi, GSK and Galapagos. M. Held, J. Klotsche, M. Claussen, J. Kirschner, M. Frankenberger and N. Khalil report no conflicts of interest to disclose.
Support statement: The INSIGHTS-ILD registry was funded by Boehringer Ingelheim. The company had no influence on the conduct of the study or interpretation of data, or reporting. Costs associated with the development and publication of this manuscript were borne by the authors’ institutions. Funding information for this article has been deposited with the Open Funder Registry.
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
00570-2025.SUPPLEMENT
Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
References
- 1.Podolanczuk AJ, Thomson CC, Remy-Jardin M, et al. Idiopathic pulmonary fibrosis: state of the art for 2023. Eur Respir J 2023; 61: 2200957. doi: 10.1183/13993003.00957-2022. [DOI] [PubMed] [Google Scholar]
- 2.Maher TM, Bendstrup E, Dron L, et al. Global incidence and prevalence of idiopathic pulmonary fibrosis. Respir Res 2021; 22: 197. doi: 10.1186/s12931-021-01791-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wijsenbeek M, Kreuter M, Olson A, et al. Progressive fibrosing interstitial lung diseases: current practice in diagnosis and management. Curr Med Res Opin 2019; 35: 2015–2024. doi: 10.1080/03007995.2019.1647040. [DOI] [PubMed] [Google Scholar]
- 4.Hambly N, Farooqi MM, Dvorkin-Gheva A, et al. Prevalence and characteristics of progressive fibrosing interstitial lung disease in a prospective registry. Eur Respir J 2022; 60: 2102571. doi: 10.1183/13993003.02571-2021. [DOI] [PubMed] [Google Scholar]
- 5.Raghu G, Remy-Jardin M, Richeldi L, et al. Idiopathic pulmonary fibrosis (an Update) and progressive pulmonary fibrosis in adults: an official ATS/ERS/JRS/ALAT clinical practice guideline. Am J Respir Crit Care Med 2022; 205: e18–e47. doi: 10.1164/rccm.202202-0399ST. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Raghu G, Collard HR, Egan JJ, et al. An official ATS/ERS/JRS/ALAT statement: idiopathic pulmonary fibrosis: evidence-based guidelines for diagnosis and management. Am J Respir Crit Care Med 2011; 183: 788–824. doi: 10.1164/rccm.2009-040GL. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hoeper MM, Markevych I, Spiekerkoetter E, et al. Goal-oriented treatment and combination therapy for pulmonary arterial hypertension. Eur Respir J 2005; 26: 858–863. doi: 10.1183/09031936.05.00075305. [DOI] [PubMed] [Google Scholar]
- 8.McLaughlin VV, Gaine SP, Howard LS, et al. Treatment goals of pulmonary hypertension. J Am Coll Cardiol 2013; 62: D73–D81. doi: 10.1016/j.jacc.2013.10.034. [DOI] [PubMed] [Google Scholar]
- 9.Sitbon O, Galiè N. Treat-to-target strategies in pulmonary arterial hypertension: the importance of using multiple goals. Eur Respir Rev Off J Eur Respir Soc 2010; 19: 272–278. doi: 10.1183/09059180.00008210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Watters LC, King TE, Schwarz MI, et al. A clinical, radiographic, and physiologic scoring system for the longitudinal assessment of patients with idiopathic pulmonary fibrosis. Am Rev Respir Dis 1986; 133: 97–103. doi: 10.1164/arrd.1986.133.1.97. [DOI] [PubMed] [Google Scholar]
- 11.King TE, Tooze JA, Schwarz MI, et al. Predicting survival in idiopathic pulmonary fibrosis: scoring system and survival model. Am J Respir Crit Care Med 2001; 164: 1171–1181. doi: 10.1164/ajrccm.164.7.2003140. [DOI] [PubMed] [Google Scholar]
- 12.Wells AU, Desai SR, Rubens MB, et al. Idiopathic pulmonary fibrosis: a composite physiologic index derived from disease extent observed by computed tomography. Am J Respir Crit Care Med 2003; 167: 962–969. doi: 10.1164/rccm.2111053. [DOI] [PubMed] [Google Scholar]
- 13.Best AC, Meng J, Lynch AM, et al. Idiopathic pulmonary fibrosis: physiologic tests, quantitative CT indexes, and CT visual scores as predictors of mortality. Radiology 2008; 246: 935–940. doi: 10.1148/radiol.2463062200. [DOI] [PubMed] [Google Scholar]
- 14.du Bois RM, Weycker D, Albera C, et al. Ascertainment of individual risk of mortality for patients with idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 2011; 184: 459–466. doi: 10.1164/rccm.201011-1790OC. [DOI] [PubMed] [Google Scholar]
- 15.Lee SH, Shim HS, Cho SH, et al. Prognostic factors for idiopathic pulmonary fibrosis: clinical, physiologic, pathologic, and molecular aspects. Sarcoidosis Vasc Diffuse Lung Dis Off J WASOG 2011; 28: 102–112. [PubMed] [Google Scholar]
- 16.Ley B, Ryerson CJ, Vittinghoff E, et al. A multidimensional index and staging system for idiopathic pulmonary fibrosis. Ann Intern Med 2012; 156: 684–691. doi: 10.7326/0003-4819-156-10-201205150-00004. [DOI] [PubMed] [Google Scholar]
- 17.Rozanski C, Mura M. Multi-dimensional indices to stage idiopathic pulmonary fibrosis: a systematic review. Sarcoidosis Vasc Diffuse Lung Dis 2014; 31: 8–18. [PubMed] [Google Scholar]
- 18.Ley B, Elicker BM, Hartman TE, et al. Idiopathic pulmonary fibrosis: CT and risk of death. Radiology 2014; 273: 570–579. doi: 10.1148/radiol.14130216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Khor YH, Cottin V, Holland AE, et al. Treatable traits: a comprehensive precision medicine approach in interstitial lung disease. Eur Respir J 2023; 62: 2300404. doi: 10.1183/13993003.00404-2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Behr J, Hoeper MM, Kreuter M, et al. Charakteristika und Versorgung der idiopathischen Lungenfibrose. “[Characteristics and management of idiopathic pulmonary fibrosis: INSIGHTS-IPF registry.]” Dtsch Med Wochenschr 2012; 137: 2586–2588. doi: 10.1055/s-0032-1327244. [DOI] [PubMed] [Google Scholar]
- 21.Behr J, Hoeper MM, Kreuter M, et al. Investigating significant health trends in idiopathic pulmonary fibrosis (INSIGHTS-IPF): rationale, aims and design of a nationwide prospective registry. BMJ Open Respir Res 2014; 1: e000010. doi: 10.1136/bmjresp-2013-000010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Behr J, Kreuter M, Hoeper MM, et al. Management of patients with idiopathic pulmonary fibrosis in clinical practice: the INSIGHTS-IPF registry. Eur Respir J 2015; 46: 186–196. doi: 10.1183/09031936.00217614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kreuter M, Swigris J, Pittrow D, et al. Health related quality of life in patients with idiopathic pulmonary fibrosis in clinical practice: insights-IPF registry. Respir Res 2017; 18: 139. doi: 10.1186/s12931-017-0621-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kreuter M, Swigris J, Pittrow D, et al. The clinical course of idiopathic pulmonary fibrosis and its association to quality of life over time: longitudinal data from the INSIGHTS-IPF registry. Respir Res 2019; 20: 59. doi: 10.1186/s12931-019-1020-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Behr J, Prasse A, Wirtz H, et al. Survival and course of lung function in the presence or absence of antifibrotic treatment in patients with idiopathic pulmonary fibrosis: long-term results of the INSIGHTS-IPF registry. Eur Respir J 2020; 56: 1902279. doi: 10.1183/13993003.02279-2019. [DOI] [PubMed] [Google Scholar]
- 26.Leuschner G, Klotsche J, Kreuter M, et al. Idiopathic pulmonary fibrosis in elderly patients: analysis of the INSIGHTS-IPF observational study. Front Med 2020; 7: 601279. doi: 10.3389/fmed.2020.601279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ryerson CJ, Tan B, Fell CD, et al. The Canadian Registry for Pulmonary Fibrosis: design and rationale of a national pulmonary fibrosis registry. Can Respir J 2016; 2016: 3562923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Fisher JH, Kolb M, Algamdi M, et al. Baseline characteristics and comorbidities in the CAnadian REgistry for Pulmonary Fibrosis. BMC Pulm Med 2019; 19: 223. doi: 10.1186/s12890-019-0986-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sauerbrei W, Schumacher M. A bootstrap resampling procedure for model building: application to the Cox regression model. Stat Med 1992; 11: 2093–2109. doi: 10.1002/sim.4780111607. [DOI] [PubMed] [Google Scholar]
- 30.Royston P, Altman DG. External validation of a Cox prognostic model: principles and methods. BMC Med Res Methodol 2013; 13: 33. doi: 10.1186/1471-2288-13-33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ryerson CJ, Vittinghoff E, Ley B, et al. Predicting survival across chronic interstitial lung disease: the ILD-GAP model. Chest 2014; 145: 723–728. doi: 10.1378/chest.13-1474. [DOI] [PubMed] [Google Scholar]
- 32.Crowson CS, Atkinson EJ, Therneau TM. Assessing calibration of prognostic risk scores. Stat Methods Med Res 2016; 25: 1692–1706. doi: 10.1177/0962280213497434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Brown KK, Flaherty KR, Cottin V, et al. Lung function outcomes in the INPULSIS® trials of nintedanib in idiopathic pulmonary fibrosis. Respir Med 2019; 146: 42–48. doi: 10.1016/j.rmed.2018.11.012. [DOI] [PubMed] [Google Scholar]
- 34.Nakano A, Ohkubo H, Fukumitsu K, et al. Remarkable improvement in a patient with idiopathic pulmonary fibrosis after treatment with nintedanib. Intern Med Tokyo Jpn 2019; 58: 1141–1144. doi: 10.2169/internalmedicine.1890-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Flaherty KR, Kolb M, Vancheri C, et al. Stability or improvement in forced vital capacity with nintedanib in patients with idiopathic pulmonary fibrosis. Eur Respir J 2018; 52: 1702593. doi: 10.1183/13993003.02593-2017. [DOI] [PubMed] [Google Scholar]
- 36.Avdeev S, Ilkovich M, Terpigorev S, et al. Effects of pirfenidone on idiopathic pulmonary fibrosis progression and safety: results of multicenter prospective observational study. Life Basel Switz 2023; 13: 483. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.King TE, Bradford WZ, Castro-Bernardini S, et al. A phase 3 trial of pirfenidone in patients with idiopathic pulmonary fibrosis. N Engl J Med 2014; 370: 2083–2092. doi: 10.1056/NEJMoa1402582. [DOI] [PubMed] [Google Scholar]
- 38.Cilli A, Uzer F, Sevinç C, et al. Tolerability and efficacy of second-line antifibrotics in patients with idiopathic pulmonary fibrosis. Pulm Pharmacol Ther 2021; 71: 102099. doi: 10.1016/j.pupt.2021.102099. [DOI] [PubMed] [Google Scholar]
- 39.Suzuki Y, Mori K, Aono Y, et al. Switching antifibrotics in patients with idiopathic pulmonary fibrosis: a multi-center retrospective cohort study. BMC Pulm Med 2021; 21: 221. doi: 10.1186/s12890-021-01587-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Aono Y, Nakamura Y, Kono M, et al. Prognostic significance of forced vital capacity decline prior to and following antifibrotic therapy in idiopathic pulmonary fibrosis. Ther Adv Respir Dis 2020; 14: 1753466620953783. doi: 10.1177/1753466620953783. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Harari S, Wells AU, Wuyts WA, et al. The 6-min walk test as a primary end-point in interstitial lung disease. Eur Respir Rev 2022; 31: 220087. doi: 10.1183/16000617.0087-2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.du Bois RM, Weycker D, Albera C, et al. Six-minute-walk test in idiopathic pulmonary fibrosis: test validation and minimal clinically important difference. Am J Respir Crit Care Med 2011; 183: 1231–1237. doi: 10.1164/rccm.201007-1179OC. [DOI] [PubMed] [Google Scholar]
- 43.du Bois RM, Albera C, Bradford WZ, et al. 6-Minute walk distance is an independent predictor of mortality in patients with idiopathic pulmonary fibrosis. Eur Respir J 2014; 43: 1421–1429. doi: 10.1183/09031936.00131813. [DOI] [PubMed] [Google Scholar]
- 44.Raghu G, Ghazipura M, Fleming TR, et al. Meaningful endpoints for idiopathic pulmonary fibrosis (IPF) clinical trials: emphasis on “feels, functions, survives”. Report of a collaborative discussion in a symposium with direct engagement from representatives of patients, investigators, the National Institutes of Health, a patient advocacy organization, and a regulatory agency. Am J Respir Crit Care Med 2024; 209: 647–669. doi: 10.1164/rccm.202312-2213SO. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Brunetti G, Malovini A, Maniscalco M, et al. Pulmonary rehabilitation in patients with interstitial lung diseases: correlates of success. Respir Med 2021; 185: 106473. doi: 10.1016/j.rmed.2021.106473. [DOI] [PubMed] [Google Scholar]
- 46.Jackson RM, Gómez-Marín OW, Ramos CF, et al. Exercise limitation in IPF patients: a randomized trial of pulmonary rehabilitation. Lung 2014; 192: 367–376. doi: 10.1007/s00408-014-9566-9. [DOI] [PubMed] [Google Scholar]
- 47.Jarosch I, Schneeberger T, Gloeckl R, et al. Short-term effects of comprehensive pulmonary rehabilitation and its maintenance in patients with idiopathic pulmonary fibrosis: a randomized controlled trial. J Clin Med 2020; 9: 1567. doi: 10.3390/jcm9051567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Kenn K, Gloeckl R, Behr J. Pulmonary rehabilitation in patients with idiopathic pulmonary fibrosis: a review. Respir Int Rev Thorac Dis 2013; 86: 89–99. [DOI] [PubMed] [Google Scholar]
- 49.Li J, Lu Y, Deng M, et al. Prevalence and risk factors of sarcopenia in idiopathic pulmonary fibrosis: a systematic review and meta-analysis. Front Med 2023; 10: 1187760. doi: 10.3389/fmed.2023.1187760. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Faverio P, Fumagalli A, Conti S, et al. Short-term evolution of nutritional status in patients with idiopathic pulmonary fibrosis. Ann Am Thorac Soc 2023; 20: 1066–1070. doi: 10.1513/AnnalsATS.202211-935RL. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material
00570-2025.SUPPLEMENT
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.