Prioritizing Therapeutic Targets for Interstitial Lung Disease: A Causal Mediation Analysis

Prioritizing Therapeutic Targets for Interstitial Lung Disease: A Causal Mediation Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Prioritizing Therapeutic Targets for Interstitial Lung Disease: A Causal Mediation Analysis Justin Oldham, Philip Molyneaux, Manoj Maddali, Chad Newton, John Kim, and 30 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8714555/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Progressive interstitial lung disease (ILD) leads to declining lung function and death. New therapies to treat ILD are urgently needed. Here we performed a secondary analysis of proteomic data from ten ILD cohorts across the United States, Canada, and United Kingdom. Causal mediation analysis was used to estimate the effect of plasma proteins previously linked to organ fibrosis in mechanistic studies (exposure) on survival (outcome) through lung function decline (mediator). Of 102 proteins tested in a discovery cohort (n = 1963), 47 were mediated by declining lung function. Of these 47 proteins, 7 showed sustained mediation in an independent validation cohort (n = 1172). Proteins with the strongest mediated effect were amphiregulin and integrin beta six. Sensitivity analysis showed that results were robust to unmeasured confounding. Here we provide epidemiological evidence implicating seven proteins as potentially causal of progressive ILD. These findings build upon mechanistic studies showing a causal link between these proteins and organ fibrosis, supporting their prioritization for therapeutic consideration. Health sciences/Biomarkers/Prognostic markers Health sciences/Medical research Interstitial Lung Disease Progressive Pulmonary Fibrosis Proteomics Biomarker Mediation Figures Figure 1 Figure 2 Figure 3 Introduction When progressive, interstitial lung disease (ILD) typically leads to declining lung function and death. 1 With a median survival of less than 5 years following diagnosis, idiopathic pulmonary fibrosis (IPF) is considered the most progressive ILD subtype. 2 However, large proportions of non-IPF ILDs also progress, including connective tissue disease-associated ILD, unclassifiable ILD, fibrotic hypersensitivity pneumonitis and idiopathic non-specific interstitial pneumonia, leading to similarly poor survival. 2 – 4 Nintedanib and, more recently, nerandomilast are approved for the treatment of progressive ILD after pivotal trials demonstrated efficacy in slowing lung function decline. 5 – 8 While these drugs, along with pirfenidone, 9 represent an important advance for the field, none stop or reverse lung fibrosis, highlighting the urgent need for new therapies. New drug development for ILD has proven challenging, as few drivers of organ fibrosis identified through mechanistic studies have translated into effective human therapies. Human-based association studies can help increase confidence in mechanistic findings, but themselves cannot establish causation. This limitation hampers translation because protein-outcome associations can arise through a disease-related causal pathway (Fig. 1 A), a disease-unrelated causal pathway (Fig. 1 B) and a non-causal pathway related to residual confounding (Fig. 1 C). Causal mediation analysis is an epidemiological method designed to overcome this uncertainty. This approach deconstructs the direct and indirect effects of an exposure on an outcome, with the indirect effects occurring through a third variable called a mediator. 10 We recently used causal mediation analysis to show that chronological age has no direct effect on ILD survival, which is instead mediated by biological age, as measured by aging biomarkers. 11 A similar approach could potentially clarify the pathways in which a circulating protein biomarker associates with ILD survival. Here we leveraged proteomic and phenotypic data from ten prospective ILD registries and cohort studies to conduct causal mediation analysis aimed at identifying circulating proteins potentially causal of ILD progression in humans, thereby prioritizing these proteins for therapeutic consideration. We hypothesized that causal mediation analysis would discriminate proteins that associate with ILD survival through declining lung function (Fig. 1 A), a cardinal feature of progressive ILD, 3,12 from those that associate with this outcome through one or more alternate pathways (Figs. 1 B-C). Proteins mediated by lung function decline across two independent, multicenter ILD cohorts and causal of organ fibrosis in laboratory models were classified as candidate therapeutic targets. Results Cohort characteristics Of 2693 and 1324 eligible individuals who underwent proteomic profiling in the discovery and validation cohorts, respectively, 1963 (73%) and 1172 (89%) were included in the analysis ( Figure E1 ). The mean age was 66 years in both cohorts, and a majority of individuals were male, white and reported a history of smoking cigarettes (Table 1 ). IPF was the predominant diagnosis in the discovery cohort, while similar proportions of IPF and CTD-ILD comprised the validation cohort. Mean percent predicted FVC and DLCO were higher and a larger proportion was treated with anti-fibrotic therapy in the discovery cohort, while a larger proportion was treated with immunosuppressant therapy in the validation cohort. Table 1 Baseline characteristics for discovery and validation cohorts Characteristic Discovery Cohort (n = 1963) Validation Cohort (n = 1172) Age, mean (± SD) 66.9 (11.3) 66.0 (11.2) Male sex, n (%) 1126 (42.6) 594 (50.7) Race, n (%) White 1659 (84.0) 990 (84.5) Black 111 (5.7) 45 (3.8) Asian 87 (4.4) 83 (7.1) Other/Unknown 106 (5.4) 54 (4.6) Diagnosis, n (%) IPF 950 (48.4) 388 (33.1) CTD-ILD 447 (22.8) 452 (38.6) fHP 275 (14.0) 111 (9.5) uILD 171 (8.7) 203 (17.3) INSIP 120 (6.1) 18 (1.5) Ever smoker, n (%) 1011 (51.5) 628 (53.6) FVC %, mean (± SD) 69.5 (18.0) 78.2 (19.2) DLCO %, mean (± SD) 46.2 (17.5) 53.9 (17.8) Anti-fibrotic exposure, n (%) At blood draw, n (%) 541 (27.6) 128 (10.9) Following blood draw, n (%) 880 (44.8) 234 (20.0) Immunosuppressant exposure, n (%) At blood draw, n (%) 374 (19.1) 391 (33.4) Following blood draw, n (%) 670 (34.1) 508 (43.3) Abbreviations: IPF = idiopathic pulmonary fibrosis; CTD = connective tissue disease; ILD = interstitial lung disease; fHp = fibrotic hypersensitivity pneumonitis; uILD = unclassifiable interstitial lung disease; INSIP = idiopathic non-specific interstitial pneumonia; FVC = forced vital capacity; DLCO = diffusion capacity of the lung for carbon monoxide Outcomes, Lung Function Decline and Composite Measure of Lung Function Decline During the 36-month observation period, 520/1963 (26.5%) and 237/1172 (20.2%) individuals died or underwent lung transplant in the discovery and validation cohorts, respectively. Mean annualized FVC decline was 8.4% (±13.8%) in the discovery cohort and 7.4% (±12.7%) in the validation cohort, while mean annualized DLCO decline was 14.9% (±20.6%) and 11.5% (±17.8%), respectively. The composite measure of lung function decline ranged from a score of 0–7, with increasing score negatively correlated with transplant-free survival and RMST (Fig. 2 ). Each unit increase in lung function decline score was associated with a 1.96 (coefficient − 1.96; 95% CI -2.14, -1.78) and 1.70 (coefficient − 1.70; 95% CI -1.93, -1.48) month decrease in RMST in the discovery and validation cohorts, respectively. Causal Mediation Analysis Of 185 proteins with previously published association with ILD survival, 102 had previously published mechanistic evidence suggesting a role in organ fibrosis ( Table E2 ). In the discovery cohort, when conducting causal mediation analysis of these 102 proteins, 67 were associated with RMST at total effect FDR p < 0.05. Declining lung function mediated the RMST association at NIE FDR p < 0.05 for 47 of these 67 proteins, with the strongest mediated effect observed for amphiregulin (AREG) and integrin beta six (ITGB6) (Table 2 , Table E3 ). When assessed in the validation cohort, 25/47 proteins showed sustained association with RMST at total effect FDR p < 0.05. Declining lung function mediated the RMST association at NIE FDR p < 0.05 for 7 of these 25 proteins, with AREG and ITGB6 again showing the strongest mediated effects (Table 2 , Table E4 ). In addition to AREG and ITGB6, granulocyte-macrophage colony-stimulating factor (CSF2), growth differentiation factor 15 (GDF15), interleukin-5 receptor subunit alpha (IL5RA), stromelysin-2 (MMP10) and group 10 secretory phospholipase A2 (PLA2G10) had RMST association that was significantly mediated by declining lung function across discovery and validation cohorts. Thus, these proteins were considered potentially causal of ILD progression and identified as high-yield therapeutic targets. Table 2 Natural indirect effects for validated proteins in discovery and validation cohorts. Symbol Discovery Cohort (n = 1963) Validation Cohort (n = 1172) Coefficient 95% CI FDR P E-value Coefficient 95% CI Low FDR P E-value AREG -2.65 -3.45, -1.85 7.02E-09 2.00 -2.24 -3.14, -1.35 2.16E-05 1.99 CSF2 -1.12 -1.73, -0.52 1.85E-03 1.51 -1.06 -1.75, -0.37 1.22E-02 1.55 GDF15 -1.59 -2.30, 0.88 1.14E-04 1.66 -1.00 -1.75, -0.25 3.14E-02 1.53 IL5RA -1.06 -1.72, -0.41 3.80E-03 1.49 -1.05 -1.69, -0.40 9.53E-03 1.54 ITGB6 -2.51 -3.33, -1.69 6.22E-08 1.95 -2.21 -3.16, -1.26 7.03E-05 1.98 MMP10 -1.25 -1.90, -0.59 1.08E-03 1.55 -1.17 -1.93, -0.40 1.16E-02 1.59 PLA2G10 -1.19 -1.87, -0.50 3.01E-03 1.53 -1.52 -2.27, -0.77 6.25E-04 1.72 Abbreviations: FDR = false discovery rate; CI = confidence interval After pooling discovery and validation cohorts, effect plots showed that mediated effects (NIE) increased with relative abundance of each validated protein (Fig. 3 ). Mediated effects predominated as AREG, ITGB6 and IL5RA relative abundance increased (Fig. 3 ). In subgroup analyses, mediated effects were similar for validated proteins when stratifying by age ( Table E6 ) and race ( Table E7 ), but higher among males ( Table E8 ) and those with lower baseline percent predicted FVC ( Table E9 ) and DLCO ( Table E10 ). Heterogeneity was also observed across diagnostic subgroups. Mediated effects were less in those CTD-ILD compared to those with IPF and other forms of fibrotic ILD ( Table E11 ). In sensitivity analyses, results were similar when using alternative quantile normalization strategies, with higher mediated effects as quantile strata increased ( Table E12 ). Results were also robust to DLCO imputation strategy, with significant mediation observed for all validated proteins after excluding those for which DLCO imputation was performed ( Table E13 ). Confounding sensitivity analysis 13 , 14 showed mediational E-values of approximately 1.5-2.0 for all validated proteins in each cohort (Table 2 ), suggesting that an unmeasured confounder of the mediator-outcome relationship with a risk ratio greater than this E-value would be required to attenuate results. Approximated risk ratios 13 , 14 for known confounders of ILD outcome risk, including age ≥65 (aRR 1.02), male sex (aRR 1.03), IPF diagnosis (aRR 1.04), CTD diagnosis (aRR 0.94), baseline FVC < 70 (aRR 1.05) and baseline DLCO < 50% (aRR 1.06) were less than E-values for all validated proteins. Discussion In this international multicohort study, we identified seven circulating plasma proteins whose associations with ILD survival are mediated through declining lung function, implicating these proteins as potential causal drivers of progressive pulmonary fibrosis. Subgroup analyses supported the biological plausibility of these findings, with stronger mediated effects observed in fibrotic-predominant ILDs (e.g., IPF and non-CTD ILDs) and among patients with more advanced disease, as measured by baseline lung function. Sensitivity analyses suggested that results were robust to quantile normalization strategy, imputation for missing DLCO and unmeasured confounding. By moving beyond traditional association analyses to interrogate causal pathways, this study provides human evidence linking these proteins to clinical outcomes through physiologic deterioration. Coupled with prior mechanistic work linking these proteins to organ fibrosis, this study supports prioritizing these proteins as promising therapeutic targets to treat progressive pulmonary fibrosis. Among validated proteins, mediation was strongest for AREG across discovery and validation cohorts, along with most key subgroups. AREG is an epidermal growth factor receptor (EGFR) ligand that can activate transforming growth factor beta 1 (TGF-β1) and lead to fibrotic remodeling through EGFR-mediated fibroblast activation. 15 , 16 In the lungs, AREG has been implicated in macrophage-mediated tissue remodeling, with macrophages serving as a critical cellular source of AREG during tissue injury and repair. 17 , 18 AREG blockade has been shown to attenuate pulmonary fibrosis in mouse models. 15 , 19 A small interfering RNA targeting AREG is currently under development for fibrotic conditions after showing promising results in reducing kidney fibrosis, 20 with a press release suggesting an acceptable safety profile from a recent phase I trial in healthy participants (NCT05984992). A monoclonal antibody targeting AREG is also under development, with recent phase 1b results suggesting an acceptable safety profile and beneficial effect on FVC and quantitative CT fibrosis in patients with IPF. 21 Importantly, pneumonitis has not reported with AREG blockade, which remains a concern with direct EGFR inhibitors. 22 ITGB6 was the second most strongly mediated protein in our analysis. ITGB6 makes up the β6 subunit of integrin αvβ6, which has long been causally linked to fibrogenesis. αvβ6 activates latent TGF-β1, leading to fibroblast-to-myofibroblast transition and collagen deposition in the lungs and elsewhere. 23 Inhibition of αvβ6 has been shown to attenuate fibrosis in mice mouse models of fibrosis 24 and slow IPF progression in an early phase clinical trial. 25 However, recent phase II trials that targeted αvβ6 using monoclonal antibodies 26 and a small molecule inhibitor (NCT06097260) were stopped due to safety concerns, suggesting that direct αvβ6 blockade may not be possible. GDF15 is a secreted ligand of the TGF-β superfamily of proteins, which regulates energy expenditure and body weight in response to metabolic stress. 27 – 29 This protein has been shown to increase with age and has been implicated in numerous aging-relating conditions, including cardiovascular disease, diabetes and chronic lung disease, including COPD and IPF. 30 – 33 GDF15 is elevated in the lungs of patients with IPF where it likely facilitates extracellular matrix formation through direct fibroblast activation and differentiation. 34 – 36 CSF2, is a granulocyte-macrophage colony stimulating factor that plays an important role in inflammation and tissue repair. CSF2 overexpression has been shown to stimulate TGF-β1 production by alveolar macrophages, which appears to be independent of inflammation-driven changes. 37 Whether CSF2 blockade could attenuate fibrosis remains unclear however, as neutralizing anti-bodies worsened fibrosis severity in a mouse model of pulmonary fibrosis. 38 PLA2G10 belongs to the family of secretory phospholipase A2 (PLA2) enzymes, which produce free fatty acids and lysophospholipids. 39 While little is known about the role PLA2G10 may play in fibrogenesis, recent studies have shown that PLA2G10 is highly expressed in IPF lungs 40 and different types of cancer. 41 PLA2G10 upregulation also prevented T cell infiltration of cancer tissue, suggesting that PLA2G10 could represent a therapeutic target for cancer immunotherapy. 41 Lysophosphatidic acid (LPA) is a well-recognized pro-fibrotic mediator and can be produced by autotaxin and PLA2. 42 , 43 Autotaxin inhibition failed to slow IPF in a recent phase III clinical trial 44 while LPA blockade is currently being investigated in phase III clinical trials for IPF and progressive non-IPF ILD after promising phase II data. 45 MMP10 is a member of the matrix metalloproteinase family of proteins, playing a key role in cell adhesion, migration and proliferation during wound healing. 46 Lung expression of MMP10 is increased in patients with IPF and has been shown to localize to alveolar and bronchiolar epithelium, along with pulmonary macrophages. 47 While mechanistic studies establishing a causal relationship between MMP10 and pulmonary fibrosis have not been performed, a mouse model of peritoneal fibrosis suggests that MMP10 blockade may have anti-fibrotic effects 48 and a recent early phase clinical trial showed that 12-week change in circulating MMP10 after treatment with rentosertib, a small molecule TNIK inhibitor, inversely correlated with change in FVC over the same timeframe. 49 IL5RA is widely studied and well-established regulator of eosinophil activation and survival. 50 An important contributor of eosinophilic-mediated conditions such as asthma, IL5RA also appears to drive subepithelial fibrosis in this population 51 and blockade of this molecular reduces expression of several key extracellular matrix proteins, including tenascin C and procollagen III. 52 recent studies have also demonstrated IL5RA receptor expression in bronchial fibroblasts, 53 suggesting a potential role in parenchymal fibrogenesis. Single cell sequencing data support this possibility, showing upregulated IL5RA expression in pulmonary fibrosis, which promotes fibrogenesis through the Jak2/STAT3 pathway. 54 Importantly, benralizumab, an anti-IL5RA monoclonal antibody is already approved for the treatment of severe eosinophilic-mediated conditions such as asthma and eosinophilic granulomatous with polyangiitis. Our data suggest that repurposing of this safe and well tolerated drug 55 could potentially provide benefit for ILD. Our study has several limitations. First, our study design was also prone to selection bias, as only patients with serial FVC measures were included, which likely selected for individuals with less severe and progressive disease. Next, our exposure, mediator and outcome variables were each prone to measurement error. For exposure measurement error, proximity extension assays detect low abundance proteins with excellent specificity, but some degree of cross reactivity remains possible. For mediator measurement error, declining FVC and DLCO represent cardinal features of progressive ILD, 3,12 but do not fully explain this phenomenon, which can also manifest as increasing extent of fibrosis on chest imaging and worsening respiratory symptoms without lung function decline. 12 The incomplete mediation observed in this analysis underscores the difficulty of establishing an optimal measure that accurately captures a progressive phenotype. For outcome measurement error, some patients will die from a competing cause of death rather than ILD. 56 Each of these sources of measurement error likely attenuated results rather than biasing results, as none were likely differential by one another. Finally, despite a rigorous attempt to satisfy key assumptions of causal mediation analysis, residual confounding remains possible. However, our confounding sensitivity analysis suggested that unmeasured confounders with larger effect size than known confounders would be required to attenuate results. Conclusion Through causal mediation analysis, this study identified a small number of prognostic protein biomarkers that are likely to play a causal role in progressive ILD. This study provides novel insights into ILD pathobiology and helps to prioritize proteins and associated molecular pathways for therapeutic consideration. While not all candidate causal biomarkers identified here represent viable therapeutic targets, our study showcases the role causal mediation analysis can play in prioritizing molecular targets for therapeutic consideration. Methods Cohorts, Data Generation and Protein Selection Individuals with IPF, connective tissue disease-associated ILD (CTD-ILD), fibrotic hypersensitivity pneumonitis, idiopathic non-specific interstitial pneumonia and unclassifiable ILD who underwent high-throughput proteomic profiling as part of two recently published proteomic investigations 57 , 58 and a new international proteomic cohort study were eligible for inclusion ( Table E1 ). Those without baseline forced vital capacity (FVC) and diffusion capacity for carbon monoxide (DLCO) (range − 6 to + 3 months relative to blood draw), at least one FVC measure following blood draw (range 3–24 months), and complete data for covariates included in mediation modeling (see below) were excluded. Methods for proteomic data generation have been described previously. 57 , 58 Briefly, the Explore 3072 and HT arrays (Olink, Uppsala, Sweden) were used to generate proteomic data in the discovery and validation cohorts, respectively. These arrays use proximity extension assays to estimate the relative abundance of circulating plasma proteins. 59 Quantile normalization was performed to harmonize proteomic data generated across different batches, with each protein categorized according to decile of relative abundance. To increase confidence in biologically plausible results, the analysis was restricted to proteins previously linked to ILD survival in human-based studies and organ fibrosis in mechanistic studies. Causal Mediation Analysis Based on the causal framework depicted in Fig. 1 , there is no direct causal pathway from a circulating protein to death in those with ILD without an intermediate process. Instead, a protein likely influences this outcome by contributing to ILD progression (Fig. 1 A) or an unmeasured condition (Fig. 1 B). A non-causal association between protein and outcome could also exist due to unmeasured confounding (Fig. 1 C). To discriminate these pathways, causal mediation analysis was performed using the mediate package in STATA (version 18, College Station, TX). Exposure was defined as decile of relative protein abundance and modeled as a continuous variable. Mediator was defined as degree of lung function decline and modeled as a continuous variable. To capture the prognostic significance of declining FVC and DLCO, 3,12 a composite measure of annualized relative decline for both was developed ( supplementary methods ). Because missing DLCO measures can result from the inability to perform the maneuver, which has prognostic significance, 60 imputation was performed to estimate the expected rate of DLCO decline when missing for those in the discovery (6.3%; 123/1963) and validation (9.9%; 116/1172) cohorts ( supplementary methods ). Outcome was defined as three-year restricted mean transplant-free survival time (RMST), which converts time-to-event data to a continuous measure for generalized linear modeling. 61 RMST was estimated using the stpmean package in STATA, with transplant-free survival defined as the time from blood draw to death, lung transplant or censoring at 36-months or sooner if lost-to-follow-up. The mediation model framework is depicted in Fig. 1 D. To derive causal interpretations, mediation analysis assumes that there exists no confounding of the 1) exposure-outcome relationship, 2) the exposure-mediator relationship, 3) the mediator-outcome relationship, and 4) the mediator-outcome relationship caused by the exposure. 10 To satisfy assumption two, the mediator model was adjusted for center, proteomic batch, age, sex, race, ILD diagnosis, smoking history, baseline percent predicted FVC and DLCO, pulmonary hypertension risk and exposure to anti-fibrotic (nintedanib or pirfenidone) and immunosuppressant (mycophenolate mofetil, azathioprine, rituximab or cyclophosphamide) therapy at the time of blood draw. To satisfy assumptions one and three, the outcome model was adjusted for these covariates plus new anti-fibrotic and immunosuppressant exposure following blood draw. To address assumption four, we utilized relatively short windows between exposure, mediator and outcome, 10 which reduced the likelihood of death due to a competing condition. 56 When reporting results, the total effect represents the RMST difference in months between groups in the first and tenth deciles of protein relative abundance. The natural indirect effect (NIE), also referred to as the mediated effect, represents the difference in RMST between these groups due to declining lung function (Fig. 1 A). The natural direct effect (NDE) represents the difference in RMST between these groups due to an unmeasured pathway (Figs. 1 B-C). Exposure-mediator interaction was allowed in all analyses, and robust standard errors were used when estimating effect estimates. Because mediation analysis requires an exposure-outcome association and a plausible biological relationship between exposure, mediator and outcome, only proteins with total effect p < 0.05 after false discovery rate (FDR) adjustment using the Benjamini Hochberg procedure 62 and previously linked to organ fibrosis in mechanistic studies were considered. Proteins associated with RMST through the lung function decline pathway (NIE FDR p < 0.05) in the discovery cohort were advanced for validation cohort testing. Those with sustained mediation by declining lung function in the validation cohort at NIE FDR p < 0.05 were considered potentially causal of progressive ILD and classified as candidate therapeutic targets. Discovery and validation cohorts were then pooled and effect plots generated to visualize mediated effects over the full range of protein values. Subgroup analyses were performed after stratification by key demographic, physiological, and diagnostic subgroups. Sensitivity analyses were performed to evaluate the effect of different quantile normalization strategies and exclusion of those with imputed DLCO decline values. Confounding sensitivity analysis was performed to estimate the mediational E-value for each protein, which estimates amount of residual confounding that would be required to attenuate results. 13 , 14 Declarations Abstract word count 158 Manuscript word count 3085 Data Sharing Statement Individual level data and summary statistics for this study will be made available within 6 months of publication through BioLINCC ( https://biolincc.nhlbi.nih.gov/home/ ). Investigators interested in accessing individual-level data prior to BioLINCC release should contact Dr. Justin Oldham ( [email protected] ). Declaration of Interest JMO reports grant funding from the national heart, lung and blood institute related to the submitted work, unrelated grant funding from Boehringer Ingelheim and personal fees from Boehringer Ingelheim, Genentech, Roche, Mediar Therapeutics, Oorja Bio, BMS, Insmed and GSK. PLM reports grant funding from AstraZeneca, GSK, Asthma and Lung UK and Action for Pulmonary Fibrosis and personal fees from Roche, Boehringer Ingelheim, AstraZeneca, Trevi, Qureight, Endeavor BioMedicines, United Therapeutics and Redx. CAN reports personal fees from Boehringer Ingelheim, BMS and Medpace, Inc. AA reports consulting fees from Boehringer Ingelheim, Genentech, Roche, Ingen, Medscape, Abbvie and Patient Mpower. LVW holds a GSK/British Lung Foundation Chair in Respiratory research. RGJ are funded by a UK National Institute for Health and Care Research (NIHR) Research Professorship. MES reports personal fees from Boehringer Ingelheim, BMS, Fibrogen and the Pulmonary Fibrosis Foundation. WAF is an employee of Glaxo Smith Kline. DK reports grant funding from the National Institutes of Health, Department of Defense, Boehringer Ingelheim and Merck and personal fees from Abbvie, Amgen, Argenx, Boehringer Ingelheim, BMS, Cabeletta, Merck, NKarta, Novartis and Zura Bio. CJR reports grant funding from Boehringer Ingelheim and personal fees from Boehringer Ingelheim, Pliant, AstraZeneca, Trevi, Avalyn, Abbvie, Veracyte, Merck, BMS and JAMP. KRF reports grant funding from Boehringer Ingelheim and personal fees from Fibrogen, Pliant, United Therapeutics, PureTech, CSL Behring, Dawoong, Insilico, Vicore, GSK, Avalyn, Boehringer Ingelheim, Trevi and Tempus. AMHV reports personal fees from Abbvie, Avalyn, Boehringer Ingelheim, Calluna, Roche, Genentech, Janssen Biotech, Medscape, Merck, MSD, Novartis, Pliant and Werfen. TMM reports personal fees from Amgen, AstraZeneca, Boehringer Ingelheim, BMS, Celgene, FibroGen, Genentech, GSK, Merck, PureTech, Sanofi, Trevi, United Therapeutics. CKG reports an unrelated grant from Boehringer Ingelheim and prior advisory board service for Pliant Therapeutics, unrelated to the submitted work. PJW reports grants from Sanofi, grants and personal fees from Boehringer Ingelheim and Roche/Genentech, personal fees from Gossamer Bio, Blade Therapeutics, and Pliant, unrelated to the submitted work. FJM reports personal fees from Altos Labs, AstraZeneca, Biogen, Boehringer Ingelheim, BMS, Chiesi, DevPro, Endeavor, Excalibur, GSK, Lung Therapeutics, Medtronic, Nitto, Novartis, Regeneron, Respivant, Roche, RS BioTherapeutics, Sanofi, Teva, Tvardi, Vicore and Zambon. IN reports personal fees from Veractye, Boehringer Ingelheim and Sanofi. All remaining authors report no disclosures. The PROFILE study was funded by the Medical Research Council (grant number G0901226), the NIHR (grant numberRP-2017-08-ST2-014), and GSK R&D (grant number CRT114316), and was sponsored by the University of Nottingham and the Royal Brompton and Harefield NHS Foundation Trust. Funding NHLBI – R01HL169166 (JMO), R01HL166290 (JMO), T15LM007033 (MVM), K23HL171871 (JVP), K23HL148498 (CAN), K23HL150301 (JSK), K23HL146942 (AA), R35HL176572 (BBM), R01HL093096 (CKG), R01HL139897 (PJW), UG3HL145266 (FJM, IN) Author Contributions JMO and JAS designed and supervised the study. JMO, PLM, CAN, JSK, JVP, GCG, AA, SRJ, RGJ, MES, DNO, DK, CJR, KRF, AMHV, TMM, CKG, PJW, FJM and IN recruited patients for the study. JMO, PLM, CAN, SK, JSK, JVP, GYL, GCG, AA, DB, RBH, RGJ, DK, CJR, KRF, AMHV, TMM, CKG and PJW collected clinical data. JMO, ALL, CHC, WAF, MVS and SFM processed plasma samples and generated proteomic data. JMO, MVM, SM, and JAS performed all analyses with scientific input from PLM, LVW, RGJ, GS, IS, RLZ, BBM, TMM and PJW. JMO and JAS wrote the manuscript with scientific input from PLM, RLZ, BBM, TMM and PJW. All authors interpreted results, reviewed the manuscript, and approved the final version of this manuscript. Acknowledgements We thank all the patients who contributed blood samples for this work, along with site investigators and research staff who helped recruit for these registries and cohort studies. References Wijsenbeek M, Cottin V (2020) Spectrum of Fibrotic Lung Diseases. N Engl J Med 383:958–968. https://doi.org:10.1056/NEJMra2005230 Cottin V et al (2018) Presentation, diagnosis and clinical course of the spectrum of progressive-fibrosing interstitial lung diseases. Eur Respir Rev 27. https://doi.org:10.1183/16000617.0076-2018 Pugashetti JV et al (2023) Validation of Proposed Criteria for Progressive Pulmonary Fibrosis. Am J Respir Crit Care Med 207:69–76. https://doi.org:10.1164/rccm.202201-0124OC Oldham JM et al (2022) Lung function trajectory in progressive fibrosing interstitial lung disease. Eur Respir J 59. https://doi.org:10.1183/13993003.01396-2021 Maher TM et al (2025) Nerandomilast in Patients with Progressive Pulmonary Fibrosis. N Engl J Med 392:2203–2214. https://doi.org:10.1056/NEJMoa2503643 Flaherty KR et al (2019) Nintedanib in Progressive Fibrosing Interstitial Lung Diseases. N Engl J Med. https://doi.org:10.1056/NEJMoa1908681 Richeldi L et al (2025) Nerandomilast in Patients with Idiopathic Pulmonary Fibrosis. N Engl J Med 392:2193–2202. https://doi.org:10.1056/NEJMoa2414108 Richeldi L et al (2014) Efficacy and safety of nintedanib in idiopathic pulmonary fibrosis. N Engl J Med 370:2071–2082. https://doi.org:10.1056/NEJMoa1402584 King TE Jr. et al (2014) A phase 3 trial of pirfenidone in patients with idiopathic pulmonary fibrosis. N Engl J Med 370:2083–2092. https://doi.org:10.1056/NEJMoa1402582 VanderWeele TJ, Mediation Analysis (2016) A Practitioner's Guide. Annu Rev Public Health 37:17–32. https://doi.org:10.1146/annurev-publhealth-032315-021402 Pugashetti JV et al (2024) Biological Age, Chronological Age and Survival in Pulmonary Fibrosis: A Causal Mediation Analysis. Am J Respir Crit Care Med. https://doi.org:10.1164/rccm.202310-1887OC Raghu G et al (2022) 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 205:e18–e47. https://doi.org:10.1164/rccm.202202-0399ST VanderWeele TJ, Ding P (2017) Sensitivity Analysis in Observational Research: Introducing the E-Value. Ann Intern Med 167:268–274. https://doi.org:10.7326/m16-2607 Smith LH, VanderWeele TJ (2019) Mediational E-values: Approximate Sensitivity Analysis for Unmeasured Mediator-Outcome Confounding. Epidemiology 30:835–837. https://doi.org:10.1097/ede.0000000000001064 Zhao R et al (2024) Sustained amphiregulin expression in intermediate alveolar stem cells drives progressive fibrosis. Cell Stem Cell 31:1344–1358e1346. https://doi.org:10.1016/j.stem.2024.07.004 Liu T, De Los Santos FG, Ding L, Wu Z, Phan SH (2016) Amphiregulin Promotes Fibroblast Activation in Pulmonary Fibrosis. FASEB J 30 50.56–50.56. https://doi.org:https://doi.org/10.1096/fasebj.30.1_supplement.50.6 Ding L et al (2016) Bone Marrow CD11c+ Cell-Derived Amphiregulin Promotes Pulmonary Fibrosis. J Immunol 197:303–312. https://doi.org:10.4049/jimmunol.1502479 Minutti CM et al (2019) A Macrophage-Pericyte Axis Directs Tissue Restoration via Amphiregulin-Induced Transforming Growth Factor Beta Activation. Immunity 50, 645–654.e646 https://doi.org:10.1016/j.immuni.2019.01.008 Lindzen M et al (2021) Targeting autocrine amphiregulin robustly and reproducibly inhibits ovarian cancer in a syngeneic model: roles for wildtype p53. Oncogene 40:3665–3679. https://doi.org:10.1038/s41388-021-01784-8 Son B et al (2022) SAMiRNA Targeting Amphiregulin Alleviate Total-Body-Irradiation-Induced Renal Fibrosis. Radiat Res 197:471–479. https://doi.org:10.1667/rade-21-00220.1 Maher TM et al Late Breaking Abstract - PMG1015 demonstrates well-tolerated safety and favorable FVC changes after 12 weeks of treatment in IPF Patients. Eur Respir J 66, PA6092 https://doi.org:10.1183/13993003.congress-2025.PA6092 Ohmori T et al (2021) Molecular and Clinical Features of EGFR-TKI-Associated Lung Injury. Int J Mol Sci 22. https://doi.org:10.3390/ijms22020792 Saini G et al (2015) αvβ6 integrin may be a potential prognostic biomarker in interstitial lung disease. Eur Respir J 46:486–494. https://doi.org:10.1183/09031936.00210414 Horan GS et al (2008) Partial inhibition of integrin alpha(v)beta6 prevents pulmonary fibrosis without exacerbating inflammation. Am J Respir Crit Care Med 177:56–65. https://doi.org:10.1164/rccm.200706-805OC Lancaster L et al (2024) Bexotegrast in Patients with Idiopathic Pulmonary Fibrosis: The INTEGRIS-IPF Clinical Trial. Am J Respir Crit Care Med 210:424–434. https://doi.org:10.1164/rccm.202403-0636OC Raghu G et al (2022) A Phase IIb Randomized Clinical Study of an Anti-α(v)β(6) Monoclonal Antibody in Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med 206:1128–1139. https://doi.org:10.1164/rccm.202112-2824OC Hsu JY et al (2017) Non-homeostatic body weight regulation through a brainstem-restricted receptor for GDF15. Nature 550:255–259. https://doi.org:10.1038/nature24042 Tsai VW et al (2013) TGF-b superfamily cytokine MIC-1/GDF15 is a physiological appetite and body weight regulator. PLoS ONE 8:e55174. https://doi.org:10.1371/journal.pone.0055174 Breit SN, Brown DA, Tsai VW (2021) The GDF15-GFRAL Pathway in Health and Metabolic Disease: Friend or Foe? Annu Rev Physiol 83:127–151. https://doi.org:10.1146/annurev-physiol-022020-045449 Welsh P et al (2022) Reference ranges for GDF-15, and risk factors associated with GDF-15, in a large general population cohort. Clin Chem Lab Med 60:1820–1829. https://doi.org:10.1515/cclm-2022-0135 Verhamme FM, Freeman CM, Brusselle GG, Bracke KR, Curtis JL (2019) GDF-15 in Pulmonary and Critical Care Medicine. Am J Respir Cell Mol Biol 60:621–628. https://doi.org:10.1165/rcmb.2018-0379TR Zhang Y et al (2019) GDF15 is an epithelial-derived biomarker of idiopathic pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 317:L510–l521. https://doi.org:10.1152/ajplung.00062.2019 Lehallier B et al (2019) Undulating changes in human plasma proteome profiles across the lifespan. Nat Med 25:1843–1850. https://doi.org:10.1038/s41591-019-0673-2 Radwanska A et al (2022) Increased expression and accumulation of GDF15 in IPF extracellular matrix contribute to fibrosis. JCI Insight 7. https://doi.org:10.1172/jci.insight.153058 Zhang Y et al (2019) GDF15 is an epithelial-derived biomarker of idiopathic pulmonary fibrosis. Am J Physiology-Lung Cell Mol Physiol 317:L510–L521. https://doi.org:10.1152/ajplung.00062.2019 Aversa Z et al (2023) Biomarkers of cellular senescence in idiopathic pulmonary fibrosis. Respir Res 24. https://doi.org:10.1186/s12931-023-02403-8 Xing Z, Tremblay GM, Sime PJ, Gauldie J (1997) Overexpression of granulocyte-macrophage colony-stimulating factor induces pulmonary granulation tissue formation and fibrosis by induction of transforming growth factor-beta 1 and myofibroblast accumulation. Am J Pathol 150:59–66 Christensen PJ et al (2000) Role of diminished epithelial GM-CSF in the pathogenesis of bleomycin-induced pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 279:L487–495. https://doi.org:10.1152/ajplung.2000.279.3.L487 Yamamoto K, Hakoi H, Nomura S, Murakami M (2023) The Roles of sPLA(2)s in Skin Homeostasis and Disease. Biomolecules 13. https://doi.org:10.3390/biom13040668 Jaiswal A et al (2023) Cellular Distribution of Secreted Phospholipase A2 in Lungs of IPF Patients and Its Inhibition in Bleomycin-Induced Pulmonary Fibrosis in Mice. Cells 12 https://doi.org:10.3390/cells12071044 Zhang T et al (2024) Up-regulated PLA2G10 in cancer impairs T cell infiltration to dampen immunity. Sci Immunol 9:eadh2334. https://doi.org:10.1126/sciimmunol.adh2334 Murakami M et al (2016) The Roles of the Secreted Phospholipase A(2) Gene Family in Immunology. Adv Immunol 132:91–134. https://doi.org:10.1016/bs.ai.2016.05.001 Aoki J, Inoue A, Okudaira S (2008) Two pathways for lysophosphatidic acid production. Biochim Biophys Acta 1781:513–518. https://doi.org:10.1016/j.bbalip.2008.06.005 Richeldi L et al (2024) Zinpentraxin Alfa for Idiopathic Pulmonary Fibrosis: The Randomized Phase III STARSCAPE Trial. Am J Respir Crit Care Med 209:1132–1140. https://doi.org:10.1164/rccm.202401-0116OC Corte TJ et al (2025) Efficacy and Safety of Admilparant, an LPA(1) Antagonist, in Pulmonary Fibrosis: A Phase 2 Randomized Clinical Trial. Am J Respir Crit Care Med 211:230–238. https://doi.org:10.1164/rccm.202405-0977OC Schlage P, Kockmann T, Sabino F, Kizhakkedathu JN (2015) Auf dem Keller, U. Matrix Metalloproteinase 10 Degradomics in Keratinocytes and Epidermal Tissue Identifies Bioactive Substrates With Pleiotropic Functions. Mol Cell Proteom 14:3234–3246. https://doi.org:10.1074/mcp.M115.053520 Sokai A et al (2015) Matrix metalloproteinase-10: a novel biomarker for idiopathic pulmonary fibrosis. Respir Res 16:120. https://doi.org:10.1186/s12931-015-0280-9 Ishimura T et al (2023) Matrix metalloproteinase-10 deficiency has protective effects against peritoneal inflammation and fibrosis via transcription factor NFκΒ pathway inhibition. Kidney Int 104:929–942. https://doi.org:10.1016/j.kint.2023.08.010 Xu Z et al (2025) A generative AI-discovered TNIK inhibitor for idiopathic pulmonary fibrosis: a randomized phase 2a trial. Nat Med 31:2602–2610. https://doi.org:10.1038/s41591-025-03743-2 Matucci A, Maggi E, Vultaggio A, Eosinophils (2019) the IL-5/IL-5Rα axis, and the biologic effects of benralizumab in severe asthma. Respir Med 160:105819. https://doi.org:10.1016/j.rmed.2019.105819 Blyth DI, Wharton TF, Pedrick MS, Savage TJ, Sanjar S (2000) Airway subepithelial fibrosis in a murine model of atopic asthma: suppression by dexamethasone or anti-interleukin-5 antibody. Am J Respir Cell Mol Biol 23:241–246. https://doi.org:10.1165/ajrcmb.23.2.3999 Flood-Page P et al (2003) Anti-IL-5 treatment reduces deposition of ECM proteins in the bronchial subepithelial basement membrane of mild atopic asthmatics. J Clin Invest 112:1029–1036. https://doi.org:10.1172/jci17974 Bajbouj K et al (2023) IL-5 receptor expression in lung fibroblasts: Potential role in airway remodeling in asthma. Allergy 78:882–885. https://doi.org:10.1111/all.15627 Chen S, Zhao T, Xie S, Wan X (2024) Epithelial IL5RA promotes epithelial-mesenchymal transition in pulmonary fibrosis via Jak2/STAT3 cascade. Pulm Pharmacol Ther 84:102286. https://doi.org:10.1016/j.pupt.2024.102286 Nair P et al (2017) Oral Glucocorticoid–Sparing Effect of Benralizumab in Severe Asthma. N Engl J Med 376:2448–2458. https://doi.org:doi:10.1056/NEJMoa1703501 Thenappan A et al (2024) Competing Causes of Death in Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med 210:938–940. https://doi.org:10.1164/rccm.202403-0595RL Oldham JM et al (2023) Proteomic Biomarkers of Survival in Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med. https://doi.org:10.1164/rccm.202301-0117OC Alqalyoobi S et al (2025) Proteomic Biomarkers of Survival in Non-IPF Interstitial Lung Disease. Am J Respir Crit Care Med. https://doi.org:10.1164/rccm.202407-1506OC Lundberg M, Eriksson A, Tran B, Assarsson E, Fredriksson S (2011) Homogeneous antibody-based proximity extension assays provide sensitive and specific detection of low-abundant proteins in human blood. Nucleic Acids Res 39:e102. https://doi.org:10.1093/nar/gkr424 Ley B et al (2012) A multidimensional index and staging system for idiopathic pulmonary fibrosis. Ann Intern Med 156:684–691. https://doi.org:10.7326/0003-4819-156-10-201205150-00004 Royston P, Parmar MK (2013) B. Restricted mean survival time: an alternative to the hazard ratio for the design and analysis of randomized trials with a time-to-event outcome. BMC Med Res Methodol 13:152. https://doi.org:10.1186/1471-2288-13-152 Benjamini Y, Hochberg Y (1995) Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J Roy Stat Soc: Ser B (Methodol) 57:289–300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x . https://doi.org: Additional Declarations Yes there is potential Competing Interest. JMO reports grant funding from the national heart, lung and blood institute related to the submitted work, unrelated grant funding from Boehringer Ingelheim and personal fees from Boehringer Ingelheim, Genentech, Roche, Mediar Therapeutics, Oorja Bio, BMS, Insmed and GSK. PLM reports grant funding from AstraZeneca, GSK, Asthma and Lung UK and Action for Pulmonary Fibrosis and personal fees from Roche, Boehringer Ingelheim, AstraZeneca, Trevi, Qureight, Endeavor BioMedicines, United Therapeutics and Redx. CAN reports personal fees from Boehringer Ingelheim, BMS and Medpace, Inc. AA reports consulting fees from Boehringer Ingelheim, Genentech, Roche, Ingen, Medscape, Abbvie and Patient Mpower. LVW holds a GSK/British Lung Foundation Chair in Respiratory research. RGJ are funded by a UK National Institute for Health and Care Research (NIHR) Research Professorship. MES reports personal fees from Boehringer Ingelheim, BMS, Fibrogen and the Pulmonary Fibrosis Foundation. WAF is an employee of Glaxo Smith Kline. DK reports grant funding from the National Institutes of Health, Department of Defense, Boehringer Ingelheim and Merck and personal fees from Abbvie, Amgen, Argenx, Boehringer Ingelheim, BMS, Cabeletta, Merck, NKarta, Novartis and Zura Bio. CJR reports grant funding from Boehringer Ingelheim and personal fees from Boehringer Ingelheim, Pliant, AstraZeneca, Trevi, Avalyn, Abbvie, Veracyte, Merck, BMS and JAMP. KRF reports grant funding from Boehringer Ingelheim and personal fees from Fibrogen, Pliant, United Therapeutics, PureTech, CSL Behring, Dawoong, Insilico, Vicore, GSK, Avalyn, Boehringer Ingelheim, Trevi and Tempus. AMHV reports personal fees from Abbvie, Avalyn, Boehringer Ingelheim, Calluna, Roche, Genentech, Janssen Biotech, Medscape, Merck, MSD, Novartis, Pliant and Werfen. TMM reports personal fees from Amgen, AstraZeneca, Boehringer Ingelheim, BMS, Celgene, FibroGen, Genentech, GSK, Merck, PureTech, Sanofi, Trevi, United Therapeutics. CKG reports an unrelated grant from Boehringer Ingelheim and prior advisory board service for Pliant Therapeutics, unrelated to the submitted work. PJW reports grants from Sanofi, grants and personal fees from Boehringer Ingelheim and Roche/Genentech, personal fees from Gossamer Bio, Blade Therapeutics, and Pliant, unrelated to the submitted work. FJM reports personal fees from Altos Labs, AstraZeneca, Biogen, Boehringer Ingelheim, BMS, Chiesi, DevPro, Endeavor, Excalibur, GSK, Lung Therapeutics, Medtronic, Nitto, Novartis, Regeneron, Respivant, Roche, RS BioTherapeutics, Sanofi, Teva, Tvardi, Vicore and Zambon. IN reports personal fees from Veractye, Boehringer Ingelheim and Sanofi. All remaining authors report no disclosures. The PROFILE study was funded by the Medical Research Council (grant number G0901226), the NIHR (grant numberRP-2017-08-ST2-014), and GSK R&D (grant number CRT114316), and was sponsored by the University of Nottingham and the Royal Brompton and Harefield NHS Foundation Trust. Supplementary Files SupplementCMAPFinal.docx Online supplement Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8714555","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":585828940,"identity":"169c3d2a-8b73-41b4-b657-7be3cf39a7fb","order_by":0,"name":"Justin Oldham","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4UlEQVRIiWNgGAWjYDACdgYzIMkmByQMgFgCSuIDzBAtxiRrYUhsgGhhIKyFn5l524OfOXzpa9ubNzDz5ljImzMwH7zNg0eLZDNbuWHvNrbcbWeOFTDzbpMw3NnAlmyNT4vBYR4zCV6Qlhs5BiAtCQYHeMykCWmR/LuNLd3s/huYFv5vBLVIA21JMLvBA7eFDa8WoF/KpGW3sRluO5NWcHAu0C8bDrMZW87Bo4WfvXmb5Nttx+TNjh/e+ODttjp5g+PND2+8waMFCo6ByQNgkpmwchCoIU7ZKBgFo2AUjEwAACe9RFY3EfYJAAAAAElFTkSuQmCC","orcid":"","institution":"University of Michigan","correspondingAuthor":true,"prefix":"","firstName":"Justin","middleName":"","lastName":"Oldham","suffix":""},{"id":585828941,"identity":"a036d99c-84a8-4744-bb16-3d8477701408","order_by":1,"name":"Philip Molyneaux","email":"","orcid":"https://orcid.org/0000-0003-1301-8800","institution":"National Heart and Lung Institute, Imperial College London","correspondingAuthor":false,"prefix":"","firstName":"Philip","middleName":"","lastName":"Molyneaux","suffix":""},{"id":585828942,"identity":"15aafaeb-e437-4f65-a3f8-80ce83792959","order_by":2,"name":"Manoj Maddali","email":"","orcid":"","institution":"Piedmont Healthcare","correspondingAuthor":false,"prefix":"","firstName":"Manoj","middleName":"","lastName":"Maddali","suffix":""},{"id":585828943,"identity":"f8d4c0c9-3153-4c24-b09f-281de1ffc08a","order_by":3,"name":"Chad Newton","email":"","orcid":"https://orcid.org/0000-0001-5256-9029","institution":"University of Texas Southwestern","correspondingAuthor":false,"prefix":"","firstName":"Chad","middleName":"","lastName":"Newton","suffix":""},{"id":585828944,"identity":"1a05dcfa-8c3e-4032-85aa-5ee325dda7fa","order_by":4,"name":"John Kim","email":"","orcid":"https://orcid.org/0000-0002-8887-150X","institution":"University of Virginia","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"","lastName":"Kim","suffix":""},{"id":585828945,"identity":"c35b6a25-3c20-4cb5-9296-e8fe35c696ee","order_by":5,"name":"Sam Konkol","email":"","orcid":"","institution":"University of Virginia","correspondingAuthor":false,"prefix":"","firstName":"Sam","middleName":"","lastName":"Konkol","suffix":""},{"id":585828946,"identity":"e7d0f73f-c4ef-4623-9db1-1af05202ef33","order_by":6,"name":"Janelle Pugashetti","email":"","orcid":"","institution":"University of Michigan","correspondingAuthor":false,"prefix":"","firstName":"Janelle","middleName":"","lastName":"Pugashetti","suffix":""},{"id":585828947,"identity":"a1ec025f-4a48-41bc-b18f-ea011e06084f","order_by":7,"name":"Gabrielle Liu","email":"","orcid":"","institution":"University of California at Davis","correspondingAuthor":false,"prefix":"","firstName":"Gabrielle","middleName":"","lastName":"Liu","suffix":""},{"id":585828948,"identity":"6609e8b5-e2a8-446d-999f-78d9e3c2ce4c","order_by":8,"name":"Gillian Goobie","email":"","orcid":"","institution":"University of British Columbia","correspondingAuthor":false,"prefix":"","firstName":"Gillian","middleName":"","lastName":"Goobie","suffix":""},{"id":585828949,"identity":"42645176-ad27-467c-94d1-2cbce7441497","order_by":9,"name":"Ayodeji Adegunsoye","email":"","orcid":"","institution":"University of Chicago","correspondingAuthor":false,"prefix":"","firstName":"Ayodeji","middleName":"","lastName":"Adegunsoye","suffix":""},{"id":585828950,"identity":"792c047b-cd9d-478c-b6f3-3d4ac8629f2e","order_by":10,"name":"Shwu-Fan Ma","email":"","orcid":"https://orcid.org/0000-0002-3058-9480","institution":"University of Virginia","correspondingAuthor":false,"prefix":"","firstName":"Shwu-Fan","middleName":"","lastName":"Ma","suffix":""},{"id":585828951,"identity":"1d41176f-3c0c-47b4-9974-bb7f292e2ca1","order_by":11,"name":"Drew Bornstein","email":"","orcid":"","institution":"University of Michigan","correspondingAuthor":false,"prefix":"","firstName":"Drew","middleName":"","lastName":"Bornstein","suffix":""},{"id":585828952,"identity":"669e545c-008c-4fdb-8264-e7af083e17a7","order_by":12,"name":"Susan Murray","email":"","orcid":"","institution":"University of Michigan","correspondingAuthor":false,"prefix":"","firstName":"Susan","middleName":"","lastName":"Murray","suffix":""},{"id":585828953,"identity":"11524b88-fb93-4e35-895f-5c5f9aaf1458","order_by":13,"name":"Louise Wain","email":"","orcid":"https://orcid.org/0000-0003-4951-1867","institution":"University of Leicester","correspondingAuthor":false,"prefix":"","firstName":"Louise","middleName":"","lastName":"Wain","suffix":""},{"id":585828954,"identity":"6350102c-4066-4dce-b186-963cc83afb80","order_by":14,"name":"Gauri Saini","email":"","orcid":"","institution":"University of Nottingham","correspondingAuthor":false,"prefix":"","firstName":"Gauri","middleName":"","lastName":"Saini","suffix":""},{"id":585828955,"identity":"dd7d340f-1def-4f66-abae-f5766e70f8bd","order_by":15,"name":"Iain Stewart","email":"","orcid":"","institution":"Imperial College London","correspondingAuthor":false,"prefix":"","firstName":"Iain","middleName":"","lastName":"Stewart","suffix":""},{"id":585828956,"identity":"31c3524e-e253-406e-8f63-dad6ddd54185","order_by":16,"name":"Simon Johnson","email":"","orcid":"https://orcid.org/0000-0002-9837-2763","institution":"University of Nottingham","correspondingAuthor":false,"prefix":"","firstName":"Simon","middleName":"","lastName":"Johnson","suffix":""},{"id":585828957,"identity":"4cd7ace1-e048-4a92-90c2-d0d74ca1ff78","order_by":17,"name":"Gisli Jenkins","email":"","orcid":"","institution":"National Heart and Lung Institute, Imperial College","correspondingAuthor":false,"prefix":"","firstName":"Gisli","middleName":"","lastName":"Jenkins","suffix":""},{"id":585828958,"identity":"f93fd56a-9724-4469-a77d-3a2d8aa9c6a9","order_by":18,"name":"Mary Strek","email":"","orcid":"","institution":"University of Chicago","correspondingAuthor":false,"prefix":"","firstName":"Mary","middleName":"","lastName":"Strek","suffix":""},{"id":585828959,"identity":"608ab5cd-d9d6-4bb5-9ef1-47b6449388ad","order_by":19,"name":"Angela Linderholm","email":"","orcid":"","institution":"University of California at Davis","correspondingAuthor":false,"prefix":"","firstName":"Angela","middleName":"","lastName":"Linderholm","suffix":""},{"id":585828960,"identity":"134d487c-8ea1-4ef3-b9d1-0540d6132915","order_by":20,"name":"Ching-Hsien Chen","email":"","orcid":"https://orcid.org/0000-0002-4211-9988","institution":"University of California Davis","correspondingAuthor":false,"prefix":"","firstName":"Ching-Hsien","middleName":"","lastName":"Chen","suffix":""},{"id":585828961,"identity":"68427296-d799-48f3-a7e8-d03e30e87d22","order_by":21,"name":"William Fahy","email":"","orcid":"","institution":"GlaxoSmithKline (United Kingdom)","correspondingAuthor":false,"prefix":"","firstName":"William","middleName":"","lastName":"Fahy","suffix":""},{"id":585828962,"identity":"a9e1e5b7-9b09-45cf-ba26-22e9e0d9181f","order_by":22,"name":"Rachel Zemans","email":"","orcid":"https://orcid.org/0000-0003-3537-9869","institution":"University of Michigan–Ann Arbor","correspondingAuthor":false,"prefix":"","firstName":"Rachel","middleName":"","lastName":"Zemans","suffix":""},{"id":585828963,"identity":"2a099455-3460-4212-bbc7-98a87fed68c8","order_by":23,"name":"Bethany Moore","email":"","orcid":"https://orcid.org/0000-0003-3051-745X","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Bethany","middleName":"","lastName":"Moore","suffix":""},{"id":585828964,"identity":"5285385f-b392-47b4-8f54-9babc5dcb13c","order_by":24,"name":"Dinesh Khanna","email":"","orcid":"https://orcid.org/0000-0003-1412-4453","institution":"University of Michigan","correspondingAuthor":false,"prefix":"","firstName":"Dinesh","middleName":"","lastName":"Khanna","suffix":""},{"id":585828965,"identity":"3638ea6c-3c56-4e59-98fd-a3827a271457","order_by":25,"name":"Christopher Ryerson","email":"","orcid":"","institution":"University of British Columbia","correspondingAuthor":false,"prefix":"","firstName":"Christopher","middleName":"","lastName":"Ryerson","suffix":""},{"id":585828966,"identity":"8653dba4-baf3-44a9-a956-ea3224a4eb7e","order_by":26,"name":"Kevin Flaherty","email":"","orcid":"","institution":"University of Michigan","correspondingAuthor":false,"prefix":"","firstName":"Kevin","middleName":"","lastName":"Flaherty","suffix":""},{"id":585828967,"identity":"9c2625fd-3631-42ba-87e0-e4f6ace645b9","order_by":27,"name":"Madathilparambil Suresh","email":"","orcid":"","institution":"University of Michigan–Ann Arbor","correspondingAuthor":false,"prefix":"","firstName":"Madathilparambil","middleName":"","lastName":"Suresh","suffix":""},{"id":585828968,"identity":"3fcde470-dd52-472b-a245-0660fad012c7","order_by":28,"name":"Anna-Maria Hoffmann-Vold","email":"","orcid":"","institution":"Oslo University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Anna-Maria","middleName":"","lastName":"Hoffmann-Vold","suffix":""},{"id":585828969,"identity":"87f04012-97c4-405a-bdaa-22d7389e86c3","order_by":29,"name":"Toby Maher","email":"","orcid":"https://orcid.org/0000-0001-7192-9149","institution":"Royal Brompton Hospital","correspondingAuthor":false,"prefix":"","firstName":"Toby","middleName":"","lastName":"Maher","suffix":""},{"id":585828970,"identity":"33464e55-63b5-4c87-b1cb-68f2dfb75326","order_by":30,"name":"Christine Garcia","email":"","orcid":"https://orcid.org/0000-0002-0771-1249","institution":"Columbia University Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Christine","middleName":"","lastName":"Garcia","suffix":""},{"id":585828971,"identity":"d9b36723-45ef-4deb-9468-1c9fde91f10b","order_by":31,"name":"Paul Wolters","email":"","orcid":"https://orcid.org/0000-0002-3108-6164","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Paul","middleName":"","lastName":"Wolters","suffix":""},{"id":585828972,"identity":"ee99447c-2552-4e35-9b70-c7fffd993dd3","order_by":32,"name":"Fernando Martinez","email":"","orcid":"https://orcid.org/0000-0002-2412-3182","institution":"University of Massachusetts Chan Medical School","correspondingAuthor":false,"prefix":"","firstName":"Fernando","middleName":"","lastName":"Martinez","suffix":""},{"id":585828973,"identity":"b52dfc3d-7915-4c09-864d-f5cc7ef9f38a","order_by":33,"name":"Imre Noth","email":"","orcid":"","institution":"University of Virginia School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Imre","middleName":"","lastName":"Noth","suffix":""},{"id":585828974,"identity":"7ed6e03d-f687-44e0-a621-dce581de55b2","order_by":34,"name":"Jennifer A. Smith","email":"","orcid":"https://orcid.org/0000-0002-3575-5468","institution":"Department of Epidemiology, School of Public Health, University of Michigan, Ann Arbor, MI, USA","correspondingAuthor":false,"prefix":"","firstName":"Jennifer","middleName":"A.","lastName":"Smith","suffix":""}],"badges":[],"createdAt":"2026-01-27 22:10:46","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8714555/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8714555/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101925771,"identity":"bab28d79-c0ae-44d0-bd23-cd9c5aaecc31","added_by":"auto","created_at":"2026-02-05 06:11:48","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":517810,"visible":true,"origin":"","legend":"\u003cp\u003eCausal framework depicting an association between a protein and ILD survival through an ILD-related causal pathway (a), ILD-unrelated causal pathway (b), and non-causal pathway due to residual confounding (c), along with causal diagram depicting modeling approach to satisfy key assumptions of mediation analysis. Abbreviation: ILD = interstitial lung disease; L\u003csub\u003e1\u003c/sub\u003e = measured confounders of exposure-outcome relationship, L\u003csub\u003e2\u003c/sub\u003e = measured confounders of exposure-mediator relationship, L\u003csub\u003e3\u003c/sub\u003e = measured confounders of mediator-outcome relationship, U\u003csub\u003eM\u003c/sub\u003e = unmeasured mediator; U\u003csub\u003eL1-3\u003c/sub\u003e = unmeasured confounders of exposure-outcome, exposure-mediator and mediator-outcome relationship. Mediator model covariates selected were age, sex, race, smoking history, body mass index, ILD diagnosis, baseline FVC, baseline DLCO and anti-fibrotic and immunosuppressant exposure at the time of blood draw. Outcome model covariates selected were mediator model covariates plus anti-fibrotic and immunosuppressant exposure following blood draw.\u003c/p\u003e","description":"","filename":"fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8714555/v1/503185dc0ebc135b21435b99.jpg"},{"id":101925786,"identity":"c7688b66-0ae9-476c-8f6b-f8c807103e3d","added_by":"auto","created_at":"2026-02-05 06:11:55","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":609581,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between lung function decline score and three-year transplant-free survival and restricted mean survival time in the discovery (A-B), validation (C-D) and pooled (E-F) cohorts.\u003c/p\u003e","description":"","filename":"fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8714555/v1/d908d366c2b5e42c674ce076.jpg"},{"id":101925781,"identity":"2e9d3bb6-67f0-4949-b92b-cec38af93634","added_by":"auto","created_at":"2026-02-05 06:11:50","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":400354,"visible":true,"origin":"","legend":"\u003cp\u003eEffect plots depicting the total effect (TE), natural indirect effect (NIE) and natural direct effect (NDE) on restricted mean survival time (RMST).\u003c/p\u003e","description":"","filename":"fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8714555/v1/1f9319d41f172f2f38f4ae5b.jpg"},{"id":105565294,"identity":"54f98951-6fd1-430c-a513-bdaa2c550966","added_by":"auto","created_at":"2026-03-27 12:52:48","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2554071,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8714555/v1/000d13bf-c2d5-43a7-ad67-f431875a99fb.pdf"},{"id":101925783,"identity":"9a43d3c1-c249-4c67-9efd-f2e86b204f49","added_by":"auto","created_at":"2026-02-05 06:11:53","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":620504,"visible":true,"origin":"","legend":"Online supplement","description":"","filename":"SupplementCMAPFinal.docx","url":"https://assets-eu.researchsquare.com/files/rs-8714555/v1/416c91d742468d81d996351b.docx"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nJMO reports grant funding from the national heart, lung and blood institute related to the submitted work, unrelated grant funding from Boehringer Ingelheim and personal fees from Boehringer Ingelheim, Genentech, Roche, Mediar Therapeutics, Oorja Bio, BMS, Insmed and GSK. PLM reports grant funding from AstraZeneca, GSK, Asthma and Lung UK and Action for Pulmonary Fibrosis and personal fees from Roche, Boehringer Ingelheim, AstraZeneca, Trevi, Qureight, Endeavor BioMedicines, United Therapeutics and Redx. CAN reports personal fees from Boehringer Ingelheim, BMS and Medpace, Inc. AA reports consulting fees from Boehringer Ingelheim, Genentech, Roche, Ingen, Medscape, Abbvie and Patient Mpower. LVW holds a GSK/British Lung Foundation Chair in Respiratory research. RGJ are funded by a UK National Institute for Health and Care Research (NIHR) Research Professorship. MES reports personal fees from Boehringer Ingelheim, BMS, Fibrogen and the Pulmonary Fibrosis Foundation. WAF is an employee of Glaxo Smith Kline. DK reports grant funding from the National Institutes of Health, Department of Defense, Boehringer Ingelheim and Merck and personal fees from Abbvie, Amgen, Argenx, Boehringer Ingelheim, BMS, Cabeletta, Merck, NKarta, Novartis and Zura Bio. CJR reports grant funding from Boehringer Ingelheim and personal fees from Boehringer Ingelheim, Pliant, AstraZeneca, Trevi, Avalyn, Abbvie, Veracyte, Merck, BMS and JAMP. KRF reports grant funding from Boehringer Ingelheim and personal fees from Fibrogen, Pliant, United Therapeutics, PureTech, CSL Behring, Dawoong, Insilico, Vicore, GSK, Avalyn, Boehringer Ingelheim, Trevi and Tempus. AMHV reports personal fees from Abbvie, Avalyn, Boehringer Ingelheim, Calluna, Roche, Genentech, Janssen Biotech, Medscape, Merck, MSD, Novartis, Pliant and Werfen. TMM reports personal fees from Amgen, AstraZeneca, Boehringer Ingelheim, BMS, Celgene, FibroGen, Genentech, GSK, Merck, PureTech, Sanofi, Trevi, United Therapeutics. CKG reports an unrelated grant from Boehringer Ingelheim and prior advisory board service for Pliant Therapeutics, unrelated to the submitted work. PJW reports grants from Sanofi, grants and personal fees from Boehringer Ingelheim and Roche/Genentech, personal fees from Gossamer Bio, Blade Therapeutics, and Pliant, unrelated to the submitted work. FJM reports personal fees from Altos Labs, AstraZeneca, Biogen, Boehringer Ingelheim, BMS, Chiesi, DevPro, Endeavor, Excalibur, GSK, Lung Therapeutics, Medtronic, Nitto, Novartis, Regeneron, Respivant, Roche, RS BioTherapeutics, Sanofi, Teva, Tvardi, Vicore and Zambon. IN reports personal fees from Veractye, Boehringer Ingelheim and Sanofi. All remaining authors report no disclosures. The PROFILE study was funded by the Medical Research Council (grant number G0901226), the NIHR (grant numberRP-2017-08-ST2-014), and GSK R\u0026D (grant number CRT114316), and was sponsored by the University of Nottingham and the Royal Brompton and Harefield NHS Foundation Trust.","formattedTitle":"Prioritizing Therapeutic Targets for Interstitial Lung Disease: A Causal Mediation Analysis","fulltext":[{"header":"Introduction","content":"\u003cp\u003eWhen progressive, interstitial lung disease (ILD) typically leads to declining lung function and death.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e With a median survival of less than 5 years following diagnosis, idiopathic pulmonary fibrosis (IPF) is considered the most progressive ILD subtype.\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e However, large proportions of non-IPF ILDs also progress, including connective tissue disease-associated ILD, unclassifiable ILD, fibrotic hypersensitivity pneumonitis and idiopathic non-specific interstitial pneumonia, leading to similarly poor survival.\u003csup\u003e\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e Nintedanib and, more recently, nerandomilast are approved for the treatment of progressive ILD after pivotal trials demonstrated efficacy in slowing lung function decline.\u003csup\u003e\u003cspan additionalcitationids=\"CR6 CR7\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e While these drugs, along with pirfenidone,\u003csup\u003e9\u003c/sup\u003e represent an important advance for the field, none stop or reverse lung fibrosis, highlighting the urgent need for new therapies.\u003c/p\u003e \u003cp\u003eNew drug development for ILD has proven challenging, as few drivers of organ fibrosis identified through mechanistic studies have translated into effective human therapies. Human-based association studies can help increase confidence in mechanistic findings, but themselves cannot establish causation. This limitation hampers translation because protein-outcome associations can arise through a disease-related causal pathway (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA), a disease-unrelated causal pathway (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB) and a non-causal pathway related to residual confounding (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). Causal mediation analysis is an epidemiological method designed to overcome this uncertainty. This approach deconstructs the direct and indirect effects of an exposure on an outcome, with the indirect effects occurring through a third variable called a mediator.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe recently used causal mediation analysis to show that chronological age has no direct effect on ILD survival, which is instead mediated by biological age, as measured by aging biomarkers.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e A similar approach could potentially clarify the pathways in which a circulating protein biomarker associates with ILD survival. Here we leveraged proteomic and phenotypic data from ten prospective ILD registries and cohort studies to conduct causal mediation analysis aimed at identifying circulating proteins potentially causal of ILD progression in humans, thereby prioritizing these proteins for therapeutic consideration. We hypothesized that causal mediation analysis would discriminate proteins that associate with ILD survival through declining lung function (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA), a cardinal feature of progressive ILD,\u003csup\u003e3,12\u003c/sup\u003e from those that associate with this outcome through one or more alternate pathways (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB-C). Proteins mediated by lung function decline across two independent, multicenter ILD cohorts and causal of organ fibrosis in laboratory models were classified as candidate therapeutic targets.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eCohort characteristics\u003c/h2\u003e \u003cp\u003eOf 2693 and 1324 eligible individuals who underwent proteomic profiling in the discovery and validation cohorts, respectively, 1963 (73%) and 1172 (89%) were included in the analysis (\u003cb\u003eFigure E1\u003c/b\u003e). The mean age was 66 years in both cohorts, and a majority of individuals were male, white and reported a history of smoking cigarettes (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). IPF was the predominant diagnosis in the discovery cohort, while similar proportions of IPF and CTD-ILD comprised the validation cohort. Mean percent predicted FVC and DLCO were higher and a larger proportion was treated with anti-fibrotic therapy in the discovery cohort, while a larger proportion was treated with immunosuppressant therapy in the validation cohort.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline characteristics for discovery and validation cohorts\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDiscovery Cohort (n\u0026thinsp;=\u0026thinsp;1963)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eValidation Cohort (n\u0026thinsp;=\u0026thinsp;1172)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, mean (\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e66.9 (11.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66.0 (11.2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale sex, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1126 (42.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e594 (50.7)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eRace, n (%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWhite\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1659 (84.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e990 (84.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBlack\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e111 (5.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e45 (3.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAsian\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e87 (4.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e83 (7.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther/Unknown\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e106 (5.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e54 (4.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eDiagnosis, n (%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIPF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e950 (48.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e388 (33.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCTD-ILD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e447 (22.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e452 (38.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003efHP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e275 (14.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e111 (9.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003euILD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e171 (8.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e203 (17.3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eINSIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e120 (6.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e18 (1.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEver smoker, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1011 (51.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e628 (53.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFVC %, mean (\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e69.5 (18.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e78.2 (19.2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDLCO %, mean (\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e46.2 (17.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e53.9 (17.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAnti-fibrotic exposure, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAt blood draw, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e541 (27.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e128 (10.9)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFollowing blood draw, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e880 (44.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e234 (20.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eImmunosuppressant exposure, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAt blood draw, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e374 (19.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e391 (33.4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFollowing blood draw, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e670 (34.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e508 (43.3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"3\"\u003eAbbreviations: IPF\u0026thinsp;=\u0026thinsp;idiopathic pulmonary fibrosis; CTD\u0026thinsp;=\u0026thinsp;connective tissue disease; ILD\u0026thinsp;=\u0026thinsp;interstitial lung disease; fHp\u0026thinsp;=\u0026thinsp;fibrotic hypersensitivity pneumonitis; uILD\u0026thinsp;=\u0026thinsp;unclassifiable interstitial lung disease; INSIP\u0026thinsp;=\u0026thinsp;idiopathic non-specific interstitial pneumonia; FVC\u0026thinsp;=\u0026thinsp;forced vital capacity; DLCO\u0026thinsp;=\u0026thinsp;diffusion capacity of the lung for carbon monoxide\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eOutcomes, Lung Function Decline and Composite Measure of Lung Function Decline\u003c/h3\u003e\n\u003cp\u003eDuring the 36-month observation period, 520/1963 (26.5%) and 237/1172 (20.2%) individuals died or underwent lung transplant in the discovery and validation cohorts, respectively. Mean annualized FVC decline was 8.4% (\u0026plusmn;13.8%) in the discovery cohort and 7.4% (\u0026plusmn;12.7%) in the validation cohort, while mean annualized DLCO decline was 14.9% (\u0026plusmn;20.6%) and 11.5% (\u0026plusmn;17.8%), respectively. The composite measure of lung function decline ranged from a score of 0\u0026ndash;7, with increasing score negatively correlated with transplant-free survival and RMST (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Each unit increase in lung function decline score was associated with a 1.96 (coefficient\u0026thinsp;\u0026minus;\u0026thinsp;1.96; 95% CI -2.14, -1.78) and 1.70 (coefficient\u0026thinsp;\u0026minus;\u0026thinsp;1.70; 95% CI -1.93, -1.48) month decrease in RMST in the discovery and validation cohorts, respectively.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eCausal Mediation Analysis\u003c/h3\u003e\n\u003cp\u003eOf 185 proteins with previously published association with ILD survival, 102 had previously published mechanistic evidence suggesting a role in organ fibrosis (\u003cb\u003eTable E2\u003c/b\u003e). In the discovery cohort, when conducting causal mediation analysis of these 102 proteins, 67 were associated with RMST at total effect FDR p\u0026thinsp;\u0026lt;\u0026thinsp;0.05. Declining lung function mediated the RMST association at NIE FDR p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 for 47 of these 67 proteins, with the strongest mediated effect observed for amphiregulin (AREG) and integrin beta six (ITGB6) (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, \u003cb\u003eTable E3\u003c/b\u003e). When assessed in the validation cohort, 25/47 proteins showed sustained association with RMST at total effect FDR p\u0026thinsp;\u0026lt;\u0026thinsp;0.05. Declining lung function mediated the RMST association at NIE FDR p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 for 7 of these 25 proteins, with AREG and ITGB6 again showing the strongest mediated effects (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, \u003cb\u003eTable E4\u003c/b\u003e). In addition to AREG and ITGB6, granulocyte-macrophage colony-stimulating factor (CSF2), growth differentiation factor 15 (GDF15), interleukin-5 receptor subunit alpha (IL5RA), stromelysin-2 (MMP10) and group 10 secretory phospholipase A2 (PLA2G10) had RMST association that was significantly mediated by declining lung function across discovery and validation cohorts. Thus, these proteins were considered potentially causal of ILD progression and identified as high-yield therapeutic targets.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eNatural indirect effects for validated proteins in discovery and validation cohorts.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026minus;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026minus;\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSymbol\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e \u003cp\u003eDiscovery Cohort (n\u0026thinsp;=\u0026thinsp;1963)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c9\" namest=\"c6\"\u003e \u003cp\u003eValidation Cohort (n\u0026thinsp;=\u0026thinsp;1172)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFDR P\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eE-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e95% CI Low\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eFDR P\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eE-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAREG\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-2.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c3\"\u003e \u003cp\u003e-3.45, -1.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7.02E-09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-2.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c7\"\u003e \u003cp\u003e-3.14, -1.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2.16E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCSF2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-1.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c3\"\u003e \u003cp\u003e-1.73, -0.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.85E-03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-1.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c7\"\u003e \u003cp\u003e-1.75, -0.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1.22E-02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.55\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGDF15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-1.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-2.30, 0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.14E-04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c7\"\u003e \u003cp\u003e-1.75, -0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e3.14E-02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.53\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIL5RA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-1.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c3\"\u003e \u003cp\u003e-1.72, -0.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.80E-03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-1.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c7\"\u003e \u003cp\u003e-1.69, -0.40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e9.53E-03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.54\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eITGB6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-2.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c3\"\u003e \u003cp\u003e-3.33, -1.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6.22E-08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-2.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c7\"\u003e \u003cp\u003e-3.16, -1.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e7.03E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.98\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMMP10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-1.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c3\"\u003e \u003cp\u003e-1.90, -0.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.08E-03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-1.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c7\"\u003e \u003cp\u003e-1.93, -0.40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1.16E-02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.59\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePLA2G10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-1.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c3\"\u003e \u003cp\u003e-1.87, -0.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.01E-03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-1.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026minus;\" colname=\"c7\"\u003e \u003cp\u003e-2.27, -0.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e6.25E-04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.72\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"9\"\u003eAbbreviations: FDR\u0026thinsp;=\u0026thinsp;false discovery rate; CI\u0026thinsp;=\u0026thinsp;confidence interval\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eAfter pooling discovery and validation cohorts, effect plots showed that mediated effects (NIE) increased with relative abundance of each validated protein (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Mediated effects predominated as AREG, ITGB6 and IL5RA relative abundance increased (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In subgroup analyses, mediated effects were similar for validated proteins when stratifying by age (\u003cb\u003eTable E6\u003c/b\u003e) and race (\u003cb\u003eTable E7\u003c/b\u003e), but higher among males (\u003cb\u003eTable E8\u003c/b\u003e) and those with lower baseline percent predicted FVC (\u003cb\u003eTable E9\u003c/b\u003e) and DLCO (\u003cb\u003eTable E10\u003c/b\u003e). Heterogeneity was also observed across diagnostic subgroups. Mediated effects were less in those CTD-ILD compared to those with IPF and other forms of fibrotic ILD (\u003cb\u003eTable E11\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn sensitivity analyses, results were similar when using alternative quantile normalization strategies, with higher mediated effects as quantile strata increased (\u003cb\u003eTable E12\u003c/b\u003e). Results were also robust to DLCO imputation strategy, with significant mediation observed for all validated proteins after excluding those for which DLCO imputation was performed (\u003cb\u003eTable E13\u003c/b\u003e). Confounding sensitivity analysis\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e showed mediational E-values of approximately 1.5-2.0 for all validated proteins in each cohort (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), suggesting that an unmeasured confounder of the mediator-outcome relationship with a risk ratio greater than this E-value would be required to attenuate results. Approximated risk ratios\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e for known confounders of ILD outcome risk, including age \u0026ge;65 (aRR 1.02), male sex (aRR 1.03), IPF diagnosis (aRR 1.04), CTD diagnosis (aRR 0.94), baseline FVC\u0026thinsp;\u0026lt;\u0026thinsp;70 (aRR 1.05) and baseline DLCO\u0026thinsp;\u0026lt;\u0026thinsp;50% (aRR 1.06) were less than E-values for all validated proteins.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this international multicohort study, we identified seven circulating plasma proteins whose associations with ILD survival are mediated through declining lung function, implicating these proteins as potential causal drivers of progressive pulmonary fibrosis. Subgroup analyses supported the biological plausibility of these findings, with stronger mediated effects observed in fibrotic-predominant ILDs (e.g., IPF and non-CTD ILDs) and among patients with more advanced disease, as measured by baseline lung function. Sensitivity analyses suggested that results were robust to quantile normalization strategy, imputation for missing DLCO and unmeasured confounding. By moving beyond traditional association analyses to interrogate causal pathways, this study provides human evidence linking these proteins to clinical outcomes through physiologic deterioration. Coupled with prior mechanistic work linking these proteins to organ fibrosis, this study supports prioritizing these proteins as promising therapeutic targets to treat progressive pulmonary fibrosis.\u003c/p\u003e \u003cp\u003eAmong validated proteins, mediation was strongest for AREG across discovery and validation cohorts, along with most key subgroups. AREG is an epidermal growth factor receptor (EGFR) ligand that can activate transforming growth factor beta 1 (TGF-β1) and lead to fibrotic remodeling through EGFR-mediated fibroblast activation.\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e In the lungs, AREG has been implicated in macrophage-mediated tissue remodeling, with macrophages serving as a critical cellular source of AREG during tissue injury and repair.\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e AREG blockade has been shown to attenuate pulmonary fibrosis in mouse models.\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e A small interfering RNA targeting AREG is currently under development for fibrotic conditions after showing promising results in reducing kidney fibrosis,\u003csup\u003e20\u003c/sup\u003e with a press release suggesting an acceptable safety profile from a recent phase I trial in healthy participants (NCT05984992). A monoclonal antibody targeting AREG is also under development, with recent phase 1b results suggesting an acceptable safety profile and beneficial effect on FVC and quantitative CT fibrosis in patients with IPF.\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e Importantly, pneumonitis has not reported with AREG blockade, which remains a concern with direct EGFR inhibitors.\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eITGB6 was the second most strongly mediated protein in our analysis. ITGB6 makes up the β6 subunit of integrin αvβ6, which has long been causally linked to fibrogenesis. αvβ6 activates latent TGF-β1, leading to fibroblast-to-myofibroblast transition and collagen deposition in the lungs and elsewhere.\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e Inhibition of αvβ6 has been shown to attenuate fibrosis in mice mouse models of fibrosis\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e and slow IPF progression in an early phase clinical trial.\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e However, recent phase II trials that targeted αvβ6 using monoclonal antibodies\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e and a small molecule inhibitor (NCT06097260) were stopped due to safety concerns, suggesting that direct αvβ6 blockade may not be possible.\u003c/p\u003e \u003cp\u003eGDF15 is a secreted ligand of the TGF-β superfamily of proteins, which regulates energy expenditure and body weight in response to metabolic stress.\u003csup\u003e\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e This protein has been shown to increase with age and has been implicated in numerous aging-relating conditions, including cardiovascular disease, diabetes and chronic lung disease, including COPD and IPF.\u003csup\u003e\u003cspan additionalcitationids=\"CR31 CR32\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e GDF15 is elevated in the lungs of patients with IPF where it likely facilitates extracellular matrix formation through direct fibroblast activation and differentiation.\u003csup\u003e\u003cspan additionalcitationids=\"CR35\" citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eCSF2, is a granulocyte-macrophage colony stimulating factor that plays an important role in inflammation and tissue repair. CSF2 overexpression has been shown to stimulate TGF-β1 production by alveolar macrophages, which appears to be independent of inflammation-driven changes.\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e Whether CSF2 blockade could attenuate fibrosis remains unclear however, as neutralizing anti-bodies worsened fibrosis severity in a mouse model of pulmonary fibrosis.\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003ePLA2G10 belongs to the family of secretory phospholipase A2 (PLA2) enzymes, which produce free fatty acids and lysophospholipids.\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e While little is known about the role PLA2G10 may play in fibrogenesis, recent studies have shown that PLA2G10 is highly expressed in IPF lungs\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e and different types of cancer.\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e PLA2G10 upregulation also prevented T cell infiltration of cancer tissue, suggesting that PLA2G10 could represent a therapeutic target for cancer immunotherapy.\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e Lysophosphatidic acid (LPA) is a well-recognized pro-fibrotic mediator and can be produced by autotaxin and PLA2.\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e Autotaxin inhibition failed to slow IPF in a recent phase III clinical trial\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e while LPA blockade is currently being investigated in phase III clinical trials for IPF and progressive non-IPF ILD after promising phase II data.\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eMMP10 is a member of the matrix metalloproteinase family of proteins, playing a key role in cell adhesion, migration and proliferation during wound healing.\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e Lung expression of MMP10 is increased in patients with IPF and has been shown to localize to alveolar and bronchiolar epithelium, along with pulmonary macrophages.\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e While mechanistic studies establishing a causal relationship between MMP10 and pulmonary fibrosis have not been performed, a mouse model of peritoneal fibrosis suggests that MMP10 blockade may have anti-fibrotic effects\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e and a recent early phase clinical trial showed that 12-week change in circulating MMP10 after treatment with rentosertib, a small molecule TNIK inhibitor, inversely correlated with change in FVC over the same timeframe. \u003csup\u003e49\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eIL5RA is widely studied and well-established regulator of eosinophil activation and survival.\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e An important contributor of eosinophilic-mediated conditions such as asthma, IL5RA also appears to drive subepithelial fibrosis in this population\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e and blockade of this molecular reduces expression of several key extracellular matrix proteins, including tenascin C and procollagen III.\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e recent studies have also demonstrated IL5RA receptor expression in bronchial fibroblasts,\u003csup\u003e53\u003c/sup\u003e suggesting a potential role in parenchymal fibrogenesis. Single cell sequencing data support this possibility, showing upregulated IL5RA expression in pulmonary fibrosis, which promotes fibrogenesis through the Jak2/STAT3 pathway.\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e Importantly, benralizumab, an anti-IL5RA monoclonal antibody is already approved for the treatment of severe eosinophilic-mediated conditions such as asthma and eosinophilic granulomatous with polyangiitis. Our data suggest that repurposing of this safe and well tolerated drug\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e could potentially provide benefit for ILD.\u003c/p\u003e \u003cp\u003eOur study has several limitations. First, our study design was also prone to selection bias, as only patients with serial FVC measures were included, which likely selected for individuals with less severe and progressive disease. Next, our exposure, mediator and outcome variables were each prone to measurement error. For exposure measurement error, proximity extension assays detect low abundance proteins with excellent specificity, but some degree of cross reactivity remains possible. For mediator measurement error, declining FVC and DLCO represent cardinal features of progressive ILD,\u003csup\u003e3,12\u003c/sup\u003e but do not fully explain this phenomenon, which can also manifest as increasing extent of fibrosis on chest imaging and worsening respiratory symptoms without lung function decline.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e The incomplete mediation observed in this analysis underscores the difficulty of establishing an optimal measure that accurately captures a progressive phenotype. For outcome measurement error, some patients will die from a competing cause of death rather than ILD.\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e Each of these sources of measurement error likely attenuated results rather than biasing results, as none were likely differential by one another. Finally, despite a rigorous attempt to satisfy key assumptions of causal mediation analysis, residual confounding remains possible. However, our confounding sensitivity analysis suggested that unmeasured confounders with larger effect size than known confounders would be required to attenuate results.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThrough causal mediation analysis, this study identified a small number of prognostic protein biomarkers that are likely to play a causal role in progressive ILD. This study provides novel insights into ILD pathobiology and helps to prioritize proteins and associated molecular pathways for therapeutic consideration. While not all candidate causal biomarkers identified here represent viable therapeutic targets, our study showcases the role causal mediation analysis can play in prioritizing molecular targets for therapeutic consideration.\u003c/p\u003e\n\n "},{"header":"Methods","content":"\u003ch2\u003eCohorts, Data Generation and Protein Selection\u003c/h2\u003e\u003cp\u003eIndividuals with IPF, connective tissue disease-associated ILD (CTD-ILD), fibrotic hypersensitivity pneumonitis, idiopathic non-specific interstitial pneumonia and unclassifiable ILD who underwent high-throughput proteomic profiling as part of two recently published proteomic investigations\u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e,\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e and a new international proteomic cohort study were eligible for inclusion (\u003cb\u003eTable E1\u003c/b\u003e). Those without baseline forced vital capacity (FVC) and diffusion capacity for carbon monoxide (DLCO) (range − 6 to + 3 months relative to blood draw), at least one FVC measure following blood draw (range 3–24 months), and complete data for covariates included in mediation modeling (see below) were excluded.\u003c/p\u003e\u003cp\u003eMethods for proteomic data generation have been described previously.\u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e,\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e Briefly, the Explore 3072 and HT arrays (Olink, Uppsala, Sweden) were used to generate proteomic data in the discovery and validation cohorts, respectively. These arrays use proximity extension assays to estimate the relative abundance of circulating plasma proteins.\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e Quantile normalization was performed to harmonize proteomic data generated across different batches, with each protein categorized according to decile of relative abundance. To increase confidence in biologically plausible results, the analysis was restricted to proteins previously linked to ILD survival in human-based studies and organ fibrosis in mechanistic studies.\u003c/p\u003e\u003ch3\u003eCausal Mediation Analysis\u003c/h3\u003e\u003cp\u003eBased on the causal framework depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, there is no direct causal pathway from a circulating protein to death in those with ILD without an intermediate process. Instead, a protein likely influences this outcome by contributing to ILD progression (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA) or an unmeasured condition (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). A non-causal association between protein and outcome could also exist due to unmeasured confounding (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). To discriminate these pathways, causal mediation analysis was performed using the \u003cem\u003emediate\u003c/em\u003e package in STATA (version 18, College Station, TX).\u003c/p\u003e\u003cp\u003e Exposure was defined as decile of relative protein abundance and modeled as a continuous variable. Mediator was defined as degree of lung function decline and modeled as a continuous variable. To capture the prognostic significance of declining FVC and DLCO,\u003csup\u003e3,12\u003c/sup\u003e a composite measure of annualized relative decline for both was developed (\u003cb\u003esupplementary methods\u003c/b\u003e). Because missing DLCO measures can result from the inability to perform the maneuver, which has prognostic significance,\u003csup\u003e60\u003c/sup\u003e imputation was performed to estimate the expected rate of DLCO decline when missing for those in the discovery (6.3%; 123/1963) and validation (9.9%; 116/1172) cohorts (\u003cb\u003esupplementary methods\u003c/b\u003e). Outcome was defined as three-year restricted mean transplant-free survival time (RMST), which converts time-to-event data to a continuous measure for generalized linear modeling.\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e RMST was estimated using the \u003cem\u003estpmean\u003c/em\u003e package in STATA, with transplant-free survival defined as the time from blood draw to death, lung transplant or censoring at 36-months or sooner if lost-to-follow-up.\u003c/p\u003e\u003cp\u003eThe mediation model framework is depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD. To derive causal interpretations, mediation analysis assumes that there exists no confounding of the 1) exposure-outcome relationship, 2) the exposure-mediator relationship, 3) the mediator-outcome relationship, and 4) the mediator-outcome relationship caused by the exposure.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e To satisfy assumption two, the mediator model was adjusted for center, proteomic batch, age, sex, race, ILD diagnosis, smoking history, baseline percent predicted FVC and DLCO, pulmonary hypertension risk and exposure to anti-fibrotic (nintedanib or pirfenidone) and immunosuppressant (mycophenolate mofetil, azathioprine, rituximab or cyclophosphamide) therapy at the time of blood draw. To satisfy assumptions one and three, the outcome model was adjusted for these covariates plus new anti-fibrotic and immunosuppressant exposure following blood draw. To address assumption four, we utilized relatively short windows between exposure, mediator and outcome,\u003csup\u003e10\u003c/sup\u003e which reduced the likelihood of death due to a competing condition.\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eWhen reporting results, the total effect represents the RMST difference in months between groups in the first and tenth deciles of protein relative abundance. The natural indirect effect (NIE), also referred to as the mediated effect, represents the difference in RMST between these groups due to declining lung function (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). The natural direct effect (NDE) represents the difference in RMST between these groups due to an unmeasured pathway (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB-C). Exposure-mediator interaction was allowed in all analyses, and robust standard errors were used when estimating effect estimates.\u003c/p\u003e\u003cp\u003eBecause mediation analysis requires an exposure-outcome association and a plausible biological relationship between exposure, mediator and outcome, only proteins with total effect p \u0026lt; 0.05 after false discovery rate (FDR) adjustment using the Benjamini Hochberg procedure\u003csup\u003e\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e and previously linked to organ fibrosis in mechanistic studies were considered. Proteins associated with RMST through the lung function decline pathway (NIE FDR p \u0026lt; 0.05) in the discovery cohort were advanced for validation cohort testing. Those with sustained mediation by declining lung function in the validation cohort at NIE FDR p \u0026lt; 0.05 were considered potentially causal of progressive ILD and classified as candidate therapeutic targets. Discovery and validation cohorts were then pooled and effect plots generated to visualize mediated effects over the full range of protein values. Subgroup analyses were performed after stratification by key demographic, physiological, and diagnostic subgroups. Sensitivity analyses were performed to evaluate the effect of different quantile normalization strategies and exclusion of those with imputed DLCO decline values. Confounding sensitivity analysis was performed to estimate the mediational E-value for each protein, which estimates amount of residual confounding that would be required to attenuate results.\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eAbstract word count\u003c/h2\u003e \u003cp\u003e158\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eManuscript word count\u003c/strong\u003e \u003cp\u003e3085\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eData Sharing Statement\u003c/h2\u003e \u003cp\u003eIndividual level data and summary statistics for this study will be made available within 6 months of publication through BioLINCC (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://biolincc.nhlbi.nih.gov/home/\u003c/span\u003e\u003cspan address=\"https://biolincc.nhlbi.nih.gov/home/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Investigators interested in accessing individual-level data prior to BioLINCC release should contact Dr. Justin Oldham ([email protected]).\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eDeclaration of Interest\u003c/h2\u003e \u003cp\u003eJMO reports grant funding from the national heart, lung and blood institute related to the submitted work, unrelated grant funding from Boehringer Ingelheim and personal fees from Boehringer Ingelheim, Genentech, Roche, Mediar Therapeutics, Oorja Bio, BMS, Insmed and GSK. PLM reports grant funding from AstraZeneca, GSK, Asthma and Lung UK and Action for Pulmonary Fibrosis and personal fees from Roche, Boehringer Ingelheim, AstraZeneca, Trevi, Qureight, Endeavor BioMedicines, United Therapeutics and Redx. CAN reports personal fees from Boehringer Ingelheim, BMS and Medpace, Inc. AA reports consulting fees from Boehringer Ingelheim, Genentech, Roche, Ingen, Medscape, Abbvie and Patient Mpower. LVW holds a GSK/British Lung Foundation Chair in Respiratory research. RGJ are funded by a UK National Institute for Health and Care Research (NIHR) Research Professorship. MES reports personal fees from Boehringer Ingelheim, BMS, Fibrogen and the Pulmonary Fibrosis Foundation. WAF is an employee of Glaxo Smith Kline. DK reports grant funding from the National Institutes of Health, Department of Defense, Boehringer Ingelheim and Merck and personal fees from Abbvie, Amgen, Argenx, Boehringer Ingelheim, BMS, Cabeletta, Merck, NKarta, Novartis and Zura Bio. CJR reports grant funding from Boehringer Ingelheim and personal fees from Boehringer Ingelheim, Pliant, AstraZeneca, Trevi, Avalyn, Abbvie, Veracyte, Merck, BMS and JAMP. KRF reports grant funding from Boehringer Ingelheim and personal fees from Fibrogen, Pliant, United Therapeutics, PureTech, CSL Behring, Dawoong, Insilico, Vicore, GSK, Avalyn, Boehringer Ingelheim, Trevi and Tempus. AMHV reports personal fees from Abbvie, Avalyn, Boehringer Ingelheim, Calluna, Roche, Genentech, Janssen Biotech, Medscape, Merck, MSD, Novartis, Pliant and Werfen. TMM reports personal fees from Amgen, AstraZeneca, Boehringer Ingelheim, BMS, Celgene, FibroGen, Genentech, GSK, Merck, PureTech, Sanofi, Trevi, United Therapeutics. CKG reports an unrelated grant from Boehringer Ingelheim and prior advisory board service for Pliant Therapeutics, unrelated to the submitted work. PJW reports grants from Sanofi, grants and personal fees from Boehringer Ingelheim and Roche/Genentech, personal fees from Gossamer Bio, Blade Therapeutics, and Pliant, unrelated to the submitted work. FJM reports personal fees from Altos Labs, AstraZeneca, Biogen, Boehringer Ingelheim, BMS, Chiesi, DevPro, Endeavor, Excalibur, GSK, Lung Therapeutics, Medtronic, Nitto, Novartis, Regeneron, Respivant, Roche, RS BioTherapeutics, Sanofi, Teva, Tvardi, Vicore and Zambon. IN reports personal fees from Veractye, Boehringer Ingelheim and Sanofi. All remaining authors report no disclosures. The PROFILE study was funded by the Medical Research Council (grant number G0901226), the NIHR (grant numberRP-2017-08-ST2-014), and GSK R\u0026amp;D (grant number CRT114316), and was sponsored by the University of Nottingham and the Royal Brompton and Harefield NHS Foundation Trust.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eNHLBI \u0026ndash; R01HL169166 (JMO), R01HL166290 (JMO), T15LM007033 (MVM), K23HL171871 (JVP), K23HL148498 (CAN), K23HL150301 (JSK), K23HL146942 (AA), R35HL176572 (BBM), R01HL093096 (CKG), R01HL139897 (PJW), UG3HL145266 (FJM, IN)\u003c/p\u003e\u003ch2\u003eAuthor Contributions\u003c/h2\u003e \u003cp\u003eJMO and JAS designed and supervised the study. JMO, PLM, CAN, JSK, JVP, GCG, AA, SRJ, RGJ, MES, DNO, DK, CJR, KRF, AMHV, TMM, CKG, PJW, FJM and IN recruited patients for the study. JMO, PLM, CAN, SK, JSK, JVP, GYL, GCG, AA, DB, RBH, RGJ, DK, CJR, KRF, AMHV, TMM, CKG and PJW collected clinical data. JMO, ALL, CHC, WAF, MVS and SFM processed plasma samples and generated proteomic data. JMO, MVM, SM, and JAS performed all analyses with scientific input from PLM, LVW, RGJ, GS, IS, RLZ, BBM, TMM and PJW. JMO and JAS wrote the manuscript with scientific input from PLM, RLZ, BBM, TMM and PJW. All authors interpreted results, reviewed the manuscript, and approved the final version of this manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eWe thank all the patients who contributed blood samples for this work, along with site investigators and research staff who helped recruit for these registries and cohort studies.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eWijsenbeek M, Cottin V (2020) Spectrum of Fibrotic Lung Diseases. N Engl J Med 383:958\u0026ndash;968. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1056/NEJMra2005230\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1056/NEJMra2005230\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCottin V et al (2018) Presentation, diagnosis and clinical course of the spectrum of progressive-fibrosing interstitial lung diseases. Eur Respir Rev 27. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1183/16000617.0076-2018\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1183/16000617.0076-2018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePugashetti JV et al (2023) Validation of Proposed Criteria for Progressive Pulmonary Fibrosis. Am J Respir Crit Care Med 207:69\u0026ndash;76. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202201-0124OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202201-0124OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOldham JM et al (2022) Lung function trajectory in progressive fibrosing interstitial lung disease. Eur Respir J 59. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1183/13993003.01396-2021\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1183/13993003.01396-2021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaher TM et al (2025) Nerandomilast in Patients with Progressive Pulmonary Fibrosis. N Engl J Med 392:2203\u0026ndash;2214. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1056/NEJMoa2503643\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1056/NEJMoa2503643\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFlaherty KR et al (2019) Nintedanib in Progressive Fibrosing Interstitial Lung Diseases. N Engl J Med. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1056/NEJMoa1908681\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1056/NEJMoa1908681\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRicheldi L et al (2025) Nerandomilast in Patients with Idiopathic Pulmonary Fibrosis. N Engl J Med 392:2193\u0026ndash;2202. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1056/NEJMoa2414108\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1056/NEJMoa2414108\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRicheldi L et al (2014) Efficacy and safety of nintedanib in idiopathic pulmonary fibrosis. N Engl J Med 370:2071\u0026ndash;2082. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1056/NEJMoa1402584\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1056/NEJMoa1402584\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKing TE Jr. et al (2014) A phase 3 trial of pirfenidone in patients with idiopathic pulmonary fibrosis. N Engl J Med 370:2083\u0026ndash;2092. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1056/NEJMoa1402582\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1056/NEJMoa1402582\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVanderWeele TJ, Mediation Analysis (2016) A Practitioner's Guide. Annu Rev Public Health 37:17\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1146/annurev-publhealth-032315-021402\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1146/annurev-publhealth-032315-021402\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePugashetti JV et al (2024) Biological Age, Chronological Age and Survival in Pulmonary Fibrosis: A Causal Mediation Analysis. Am J Respir Crit Care Med. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202310-1887OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202310-1887OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRaghu G et al (2022) 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 205:e18\u0026ndash;e47. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202202-0399ST\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202202-0399ST\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVanderWeele TJ, Ding P (2017) Sensitivity Analysis in Observational Research: Introducing the E-Value. Ann Intern Med 167:268\u0026ndash;274. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.7326/m16-2607\u003c/span\u003e\u003cspan address=\"https://doi.org:10.7326/m16-2607\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith LH, VanderWeele TJ (2019) Mediational E-values: Approximate Sensitivity Analysis for Unmeasured Mediator-Outcome Confounding. Epidemiology 30:835\u0026ndash;837. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1097/ede.0000000000001064\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1097/ede.0000000000001064\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao R et al (2024) Sustained amphiregulin expression in intermediate alveolar stem cells drives progressive fibrosis. Cell Stem Cell 31:1344\u0026ndash;1358e1346. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.stem.2024.07.004\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.stem.2024.07.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu T, De Los Santos FG, Ding L, Wu Z, Phan SH (2016) Amphiregulin Promotes Fibroblast Activation in Pulmonary Fibrosis. FASEB J 30 50.56\u0026ndash;50.56. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:https://doi.org/10.1096/fasebj.30.1_supplement.50.6\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1096/fasebj.30.1_supplement.50.6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDing L et al (2016) Bone Marrow CD11c+ Cell-Derived Amphiregulin Promotes Pulmonary Fibrosis. J Immunol 197:303\u0026ndash;312. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.4049/jimmunol.1502479\u003c/span\u003e\u003cspan address=\"https://doi.org:10.4049/jimmunol.1502479\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinutti CM et al (2019) A Macrophage-Pericyte Axis Directs Tissue Restoration via Amphiregulin-Induced Transforming Growth Factor Beta Activation. \u003cem\u003eImmunity\u003c/em\u003e 50, 645\u0026ndash;654.e646 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.immuni.2019.01.008\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.immuni.2019.01.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLindzen M et al (2021) Targeting autocrine amphiregulin robustly and reproducibly inhibits ovarian cancer in a syngeneic model: roles for wildtype p53. Oncogene 40:3665\u0026ndash;3679. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41388-021-01784-8\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41388-021-01784-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSon B et al (2022) SAMiRNA Targeting Amphiregulin Alleviate Total-Body-Irradiation-Induced Renal Fibrosis. Radiat Res 197:471\u0026ndash;479. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1667/rade-21-00220.1\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1667/rade-21-00220.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaher TM et al Late Breaking Abstract - PMG1015 demonstrates well-tolerated safety and favorable FVC changes after 12 weeks of treatment in IPF Patients. Eur Respir J 66, PA6092 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1183/13993003.congress-2025.PA6092\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1183/13993003.congress-2025.PA6092\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhmori T et al (2021) Molecular and Clinical Features of EGFR-TKI-Associated Lung Injury. Int J Mol Sci 22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.3390/ijms22020792\u003c/span\u003e\u003cspan address=\"https://doi.org:10.3390/ijms22020792\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaini G et al (2015) αvβ6 integrin may be a potential prognostic biomarker in interstitial lung disease. Eur Respir J 46:486\u0026ndash;494. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1183/09031936.00210414\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1183/09031936.00210414\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoran GS et al (2008) Partial inhibition of integrin alpha(v)beta6 prevents pulmonary fibrosis without exacerbating inflammation. Am J Respir Crit Care Med 177:56\u0026ndash;65. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.200706-805OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.200706-805OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLancaster L et al (2024) Bexotegrast in Patients with Idiopathic Pulmonary Fibrosis: The INTEGRIS-IPF Clinical Trial. Am J Respir Crit Care Med 210:424\u0026ndash;434. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202403-0636OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202403-0636OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRaghu G et al (2022) A Phase IIb Randomized Clinical Study of an Anti-α(v)β(6) Monoclonal Antibody in Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med 206:1128\u0026ndash;1139. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202112-2824OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202112-2824OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHsu JY et al (2017) Non-homeostatic body weight regulation through a brainstem-restricted receptor for GDF15. Nature 550:255\u0026ndash;259. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/nature24042\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/nature24042\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTsai VW et al (2013) TGF-b superfamily cytokine MIC-1/GDF15 is a physiological appetite and body weight regulator. PLoS ONE 8:e55174. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1371/journal.pone.0055174\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1371/journal.pone.0055174\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreit SN, Brown DA, Tsai VW (2021) The GDF15-GFRAL Pathway in Health and Metabolic Disease: Friend or Foe? Annu Rev Physiol 83:127\u0026ndash;151. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1146/annurev-physiol-022020-045449\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1146/annurev-physiol-022020-045449\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWelsh P et al (2022) Reference ranges for GDF-15, and risk factors associated with GDF-15, in a large general population cohort. Clin Chem Lab Med 60:1820\u0026ndash;1829. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1515/cclm-2022-0135\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1515/cclm-2022-0135\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerhamme FM, Freeman CM, Brusselle GG, Bracke KR, Curtis JL (2019) GDF-15 in Pulmonary and Critical Care Medicine. Am J Respir Cell Mol Biol 60:621\u0026ndash;628. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1165/rcmb.2018-0379TR\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1165/rcmb.2018-0379TR\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Y et al (2019) GDF15 is an epithelial-derived biomarker of idiopathic pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 317:L510\u0026ndash;l521. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1152/ajplung.00062.2019\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1152/ajplung.00062.2019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLehallier B et al (2019) Undulating changes in human plasma proteome profiles across the lifespan. Nat Med 25:1843\u0026ndash;1850. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41591-019-0673-2\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41591-019-0673-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRadwanska A et al (2022) Increased expression and accumulation of GDF15 in IPF extracellular matrix contribute to fibrosis. JCI Insight 7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1172/jci.insight.153058\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1172/jci.insight.153058\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Y et al (2019) GDF15 is an epithelial-derived biomarker of idiopathic pulmonary fibrosis. Am J Physiology-Lung Cell Mol Physiol 317:L510\u0026ndash;L521. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1152/ajplung.00062.2019\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1152/ajplung.00062.2019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAversa Z et al (2023) Biomarkers of cellular senescence in idiopathic pulmonary fibrosis. Respir Res 24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/s12931-023-02403-8\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/s12931-023-02403-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXing Z, Tremblay GM, Sime PJ, Gauldie J (1997) Overexpression of granulocyte-macrophage colony-stimulating factor induces pulmonary granulation tissue formation and fibrosis by induction of transforming growth factor-beta 1 and myofibroblast accumulation. Am J Pathol 150:59\u0026ndash;66\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChristensen PJ et al (2000) Role of diminished epithelial GM-CSF in the pathogenesis of bleomycin-induced pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 279:L487\u0026ndash;495. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1152/ajplung.2000.279.3.L487\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1152/ajplung.2000.279.3.L487\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYamamoto K, Hakoi H, Nomura S, Murakami M (2023) The Roles of sPLA(2)s in Skin Homeostasis and Disease. Biomolecules 13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.3390/biom13040668\u003c/span\u003e\u003cspan address=\"https://doi.org:10.3390/biom13040668\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaiswal A et al (2023) Cellular Distribution of Secreted Phospholipase A2 in Lungs of IPF Patients and Its Inhibition in Bleomycin-Induced Pulmonary Fibrosis in Mice. \u003cem\u003eCells\u003c/em\u003e 12 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.3390/cells12071044\u003c/span\u003e\u003cspan address=\"https://doi.org:10.3390/cells12071044\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang T et al (2024) Up-regulated PLA2G10 in cancer impairs T cell infiltration to dampen immunity. Sci Immunol 9:eadh2334. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1126/sciimmunol.adh2334\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1126/sciimmunol.adh2334\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurakami M et al (2016) The Roles of the Secreted Phospholipase A(2) Gene Family in Immunology. Adv Immunol 132:91\u0026ndash;134. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/bs.ai.2016.05.001\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/bs.ai.2016.05.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAoki J, Inoue A, Okudaira S (2008) Two pathways for lysophosphatidic acid production. Biochim Biophys Acta 1781:513\u0026ndash;518. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.bbalip.2008.06.005\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.bbalip.2008.06.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRicheldi L et al (2024) Zinpentraxin Alfa for Idiopathic Pulmonary Fibrosis: The Randomized Phase III STARSCAPE Trial. Am J Respir Crit Care Med 209:1132\u0026ndash;1140. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202401-0116OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202401-0116OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCorte TJ et al (2025) Efficacy and Safety of Admilparant, an LPA(1) Antagonist, in Pulmonary Fibrosis: A Phase 2 Randomized Clinical Trial. Am J Respir Crit Care Med 211:230\u0026ndash;238. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202405-0977OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202405-0977OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchlage P, Kockmann T, Sabino F, Kizhakkedathu JN (2015) Auf dem Keller, U. Matrix Metalloproteinase 10 Degradomics in Keratinocytes and Epidermal Tissue Identifies Bioactive Substrates With Pleiotropic Functions. Mol Cell Proteom 14:3234\u0026ndash;3246. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1074/mcp.M115.053520\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1074/mcp.M115.053520\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSokai A et al (2015) Matrix metalloproteinase-10: a novel biomarker for idiopathic pulmonary fibrosis. Respir Res 16:120. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/s12931-015-0280-9\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/s12931-015-0280-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIshimura T et al (2023) Matrix metalloproteinase-10 deficiency has protective effects against peritoneal inflammation and fibrosis via transcription factor NFκΒ pathway inhibition. Kidney Int 104:929\u0026ndash;942. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.kint.2023.08.010\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.kint.2023.08.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu Z et al (2025) A generative AI-discovered TNIK inhibitor for idiopathic pulmonary fibrosis: a randomized phase 2a trial. Nat Med 31:2602\u0026ndash;2610. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41591-025-03743-2\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41591-025-03743-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatucci A, Maggi E, Vultaggio A, Eosinophils (2019) the IL-5/IL-5Rα axis, and the biologic effects of benralizumab in severe asthma. Respir Med 160:105819. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.rmed.2019.105819\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.rmed.2019.105819\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlyth DI, Wharton TF, Pedrick MS, Savage TJ, Sanjar S (2000) Airway subepithelial fibrosis in a murine model of atopic asthma: suppression by dexamethasone or anti-interleukin-5 antibody. Am J Respir Cell Mol Biol 23:241\u0026ndash;246. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1165/ajrcmb.23.2.3999\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1165/ajrcmb.23.2.3999\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFlood-Page P et al (2003) Anti-IL-5 treatment reduces deposition of ECM proteins in the bronchial subepithelial basement membrane of mild atopic asthmatics. J Clin Invest 112:1029\u0026ndash;1036. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1172/jci17974\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1172/jci17974\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBajbouj K et al (2023) IL-5 receptor expression in lung fibroblasts: Potential role in airway remodeling in asthma. Allergy 78:882\u0026ndash;885. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1111/all.15627\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1111/all.15627\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen S, Zhao T, Xie S, Wan X (2024) Epithelial IL5RA promotes epithelial-mesenchymal transition in pulmonary fibrosis via Jak2/STAT3 cascade. Pulm Pharmacol Ther 84:102286. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.pupt.2024.102286\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.pupt.2024.102286\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNair P et al (2017) Oral Glucocorticoid\u0026ndash;Sparing Effect of Benralizumab in Severe Asthma. N Engl J Med 376:2448\u0026ndash;2458. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:doi:10.1056/NEJMoa1703501\u003c/span\u003e\u003cspan address=\"https://doi.org:doi:10.1056/NEJMoa1703501\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThenappan A et al (2024) Competing Causes of Death in Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med 210:938\u0026ndash;940. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202403-0595RL\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202403-0595RL\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOldham JM et al (2023) Proteomic Biomarkers of Survival in Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202301-0117OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202301-0117OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlqalyoobi S et al (2025) Proteomic Biomarkers of Survival in Non-IPF Interstitial Lung Disease. Am J Respir Crit Care Med. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.202407-1506OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.202407-1506OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLundberg M, Eriksson A, Tran B, Assarsson E, Fredriksson S (2011) Homogeneous antibody-based proximity extension assays provide sensitive and specific detection of low-abundant proteins in human blood. Nucleic Acids Res 39:e102. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1093/nar/gkr424\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1093/nar/gkr424\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLey B et al (2012) A multidimensional index and staging system for idiopathic pulmonary fibrosis. Ann Intern Med 156:684\u0026ndash;691. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.7326/0003-4819-156-10-201205150-00004\u003c/span\u003e\u003cspan address=\"https://doi.org:10.7326/0003-4819-156-10-201205150-00004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoyston P, Parmar MK (2013) B. Restricted mean survival time: an alternative to the hazard ratio for the design and analysis of randomized trials with a time-to-event outcome. BMC Med Res Methodol 13:152. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/1471-2288-13-152\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/1471-2288-13-152\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenjamini Y, Hochberg Y (1995) Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J Roy Stat Soc: Ser B (Methodol) 57:289\u0026ndash;300. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.2517-6161.1995.tb02031.x\u003c/span\u003e\u003cspan address=\"10.1111/j.2517-6161.1995.tb02031.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. https://doi.org:\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Interstitial Lung Disease, Progressive Pulmonary Fibrosis, Proteomics, Biomarker, Mediation","lastPublishedDoi":"10.21203/rs.3.rs-8714555/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8714555/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eProgressive interstitial lung disease (ILD) leads to declining lung function and death. New therapies to treat ILD are urgently needed. Here we performed a secondary analysis of proteomic data from ten ILD cohorts across the United States, Canada, and United Kingdom. Causal mediation analysis was used to estimate the effect of plasma proteins previously linked to organ fibrosis in mechanistic studies (exposure) on survival (outcome) through lung function decline (mediator). Of 102 proteins tested in a discovery cohort (n\u0026thinsp;=\u0026thinsp;1963), 47 were mediated by declining lung function. Of these 47 proteins, 7 showed sustained mediation in an independent validation cohort (n\u0026thinsp;=\u0026thinsp;1172). Proteins with the strongest mediated effect were amphiregulin and integrin beta six. Sensitivity analysis showed that results were robust to unmeasured confounding. Here we provide epidemiological evidence implicating seven proteins as potentially causal of progressive ILD. These findings build upon mechanistic studies showing a causal link between these proteins and organ fibrosis, supporting their prioritization for therapeutic consideration.\u003c/p\u003e","manuscriptTitle":"Prioritizing Therapeutic Targets for Interstitial Lung Disease: A Causal Mediation Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-05 06:10:42","doi":"10.21203/rs.3.rs-8714555/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2aa7c0e4-9b00-44eb-9df7-1cd6fbef4270","owner":[],"postedDate":"February 5th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":62328623,"name":"Health sciences/Biomarkers/Prognostic markers"},{"id":62328624,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2026-03-25T13:01:11+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-05 06:10:42","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8714555","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8714555","identity":"rs-8714555","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}