Acute exacerbation in fibrotic interstitial lung disease: an international working group report

Abstract

Acute exacerbations (AEs) occur both in patients with idiopathic pulmonary fibrosis (IPF) and those with non-IPF fibrotic interstitial lung disease (fILD). These events confer high morbidity and mortality, with a lack of proven effective therapeutic interventions. The objective of this state-of-the-art document is to summarize latest evidence since the 2016 International Working Group Report on AE-IPF, expanding it across the spectrum of all fILDs. A comprehensive literature review on the epidemiology, associated and risk factors, prognosis, and management of AE-fILD is summarized. In addition to revising the AE definition and diagnostic criteria for broad application across different fILDs, a conceptual framework for acute respiratory worsening (ARW) has been proposed to encompass a variety of acute respiratory deteriorations, both related and unrelated to AEs. This allows structured evaluation in both clinical and research settings. The proposed revised definition for AE-fILD is an acute respiratory event characterized by increased respiratory symptoms or signs and associated with radiologic or histologic features consistent with diffuse alveolar damage (with or without superimposed organizing pneumonia) in a patient with known or newly diagnosed fILD. On the other hand, ARW refers to a heterogeneous group of clinical events with acute symptom worsening not attributable to diffuse alveolar damage in patients with fILD, such as pulmonary edema, bronchitis, and pneumonia, although severe pneumonia can trigger AE-fILD. Additionally, we discuss considerations for inclusion of AE as a clinical trial endpoint, as well as research priorities for advancing knowledge on the pathogenic mechanisms, event prediction, risk stratification, and development of drugs and supportive treatments.

Introduction

In 2016, an international working group defined acute exacerbation of idiopathic pulmonary fibrosis (AE-IPF) as a clinically significant respiratory deterioration developing within typically less than 1 month, accompanied by new bilateral ground glass opacity (GGO) and/or consolidation on high-resolution computed tomography (HRCT) that is not fully explained by cardiac failure or fluid overload.¹ Subsequent studies demonstrate that such acute events are not limited to idiopathic pulmonary fibrosis (IPF), but may also complicate the course of other fibrotic interstitial lung diseases (fILDs).^2-4 The 2016 definition of AE-IPF has been adopted as a secondary endpoint in clinical trials of therapies for progressive pulmonary fibrosis (PPF) despite the lack of a standardized approach for its application in non-IPF fILDs.^5,6 This state-of-the-art document aims to (1) revise the proposed definition of acute exacerbation (AE) to extend it to fILDs beyond IPF; (2) provide a diagnostic approach for evaluation of patients with fILD presenting with acute respiratory worsening (ARW); (3) review currently available therapeutic options for AE-fILD; (4) explore the utility of AE as an endpoint for clinical trials conducted in fILD; and (5) discuss future research priorities for AE-fILD.

Methods

An international multidisciplinary group of 41 experts, including pulmonology, rheumatology, radiology, pathology, and pharmacology, was assembled (Table S1). Working group members met in a series of video conferences to discuss the objectives, scope, and specific content of this state-of-the-art document. An evidence-based medicine subcommittee (Y.H.K., A.A., B.F.C., E.F., S.M., C.A.N., and A.J.P.) conducted a systematic search of Ovid Medline, Embase, and All EBM Reviews databases for literature related to AE-IPF (from February 2016) and non-IPF fILD (from inception) up to March 17, 2025 (Table S2). The reference list of the 2016 AE-IPF International Working Group Report was cross-checked.¹  Figure 1 summarizes the study selection process for different sections, with 922 unique citations included. Data were primarily synthesized narratively. Random-effects meta-analyses were performed to estimate the cumulative incidence, clinical features, mortality rates, and treatment effects for AE, with harvest plots being used to summarize evaluation of association/risk and prognostic factors. Details of literature review methodology are described in the Supplementary Appendix and Tables S3 through S5, with characteristics of included studies presented in Tables S6 through S9. For document development, the working group was divided into writing groups for each section, with support from the evidence-based medicine subcommittee lead (Y.H.K.). Sections were collated into the full manuscript draft for iterative revision and approval by all working group members.

Figure 1.

Systemic literature review flowchart and study inclusion. *Reference lists of these 42 systematic reviews (as listed in Table S5) were screened for relevant studies. Abbreviations: AE, acute exacerbation; ILD, interstitial lung disease; IPF, idiopathic pulmonary fibrosis; SRs, systematic reviews.

Pathogenesis

AE-IPF arises from a complex pathogenic cascade (Figure 2), which culminates in one or a combination of histopathologic features of diffuse alveolar damage (DAD) with intra-alveolar fibrin hyaline membrane deposition, with or without organizing pneumonia. Our understanding of the pathogenesis of AE-fILD is based on extrapolations from limited studies in AE-IPF and acute respiratory distress syndrome, both sharing pathology in the final common pathway of dysfunctional repair and/or widespread alveolar damage.⁷

Figure 2.

Pathogenesis of acute exacerbation in IPF. The pathogenesis of IPF during slowly progressive disease is characterized by repetitive microinjuries of bronchiolar and alveolar epithelial cells, including aberrant basaloid cells and mucus abnormalities (1, 2). This leads to fibroblast proliferation, accumulation of extracellular matrix (3), and vascular remodeling (4). Acute exacerbations can develop following unspecific triggers (5) that can be viral/bacterial infection or microbiome dysbiosis, deposition of antibody complex, mechanical stress, or other injuries that lead to oxidative stress. This leads to increased production of inflammatory cytokines whereas protective mechanisms are reduced. Various inflammatory cells enter the interstitial and alveolar space through the increasingly leaking basal membranes and protective cell layers on the alveolar and capillary end of the barrier (6). Formation of hyaline membranes, edema of the interstitial space, accelerated matrix deposition, and disruptions of the coagulation cascades lead to worsening acute lung injury (7). Abbreviations: IPF, idiopathic pulmonary fibrosis; ROS, reactive oxygen species.

There are predisposing factors affecting the lungs of individuals with pulmonary fibrosis that exacerbate the response to known triggers of AE-IPF. Alveolar epithelial cells and alveolar capillary endothelial cells, which are required for structural integrity and barrier function in the lung, are reduced in IPF.⁸ In addition, there are downregulated expressions of zonula adherens molecules with structural alterations, including E-cadherin in airway epithelial cells and VE-cadherin in capillary endothelium. The loss of this barrier is further impaired by the replacement of alveolar epithelium and endothelium by aberrant basaloid cells and COLl15⁺ endothelial cells lacking regenerative and homeostatic capacity. Shortened telomeres in IPF also contribute to local progenitor dysfunction. Furthermore, fibrotic lungs exhibit augmented response to inflammatory stimuli, such as a heightened NLRP3-inflammasone caspase-1 pathway, accumulation of SPP1⁺ monocyte-derived alveolar macrophages, and transforming growth factor-β–producing regulatory T cells.⁹ Immune dysfunction, characterized by diminished antigen recognition by reduced Toll-like receptors and impaired macrophage function and neutrophil activity, leads to impaired innate responses to infection.¹⁰ Increased mucin 5B in the lung may impair mucosal clearance of pathogens or inhaled particles.¹¹ In addition, dysbiosis is frequently seen in IPF and may result in subclinical structural damage from viruses or other microorganism,¹² making patients more susceptible to infection and DAD.

Data on cellular and molecular markers identified during AE-fILD are limited. Lungs from deceased individuals with AE-IPF reveal widespread apoptosis and hyperplasia of the alveolar epithelium with deposition of Cyclin A2, a cell cycle regulator.¹³ The potential contribution of antibody complex deposition precipitating immune damage of structural cells requires investigation, given preliminary findings of positive response to plasma exchange in selected cases of AE-IPF.¹⁴ Elevated angiopoietin-2 is a marker and mediator of compromised capillary function,¹⁵ and increased colony-stimulating factor-3 impairs alveolar epithelial barrier function. Profibrotic molecules present during AE include growth differentiation factor-15 causing senescence in airway epithelial cells, reduced stimulator of interferon genes, and lower let-7d-5p microRNA that promotes fibrosis and collagen deposition.^16,17

Genome-wide transcriptomic studies suggest increased activity of mucosal innate immunity in AE-IPF.¹³ Analysis of bronchoalveolar lavage (BAL) fluid demonstrates macrophage activation, and neutrophil recruitment with increased interleukin (IL)-8, chemokine C-X-C motif ligand 1 (CXCL1), and IL-1β.^18,19 Multiple studies have shown increased inflammatory cytokines such as IL-6, IL-8, interferon-γ, soluble CD163, chemokine C–C motif ligand 2, and CXCL10,^20-22 along with altered expression of other mediators including raised S100A, leptin, and Toll-like receptor 4, with reduced soluble receptor for advanced glycation end products.^23-28 Reduced innate immune responses to pathogens are demonstrated by low interferon, antigen recognition, antigen-presenting cells, and CD4 T cells leading to an increased microbial load and infection.²⁹

Epidemiology

Incidence estimates of AE-fILD vary considerably depending on population, geographic region, season, study design, and diagnostic criteria. In IPF, clinical trials have reported AE rates ranging from 4 to 8 events per 100 patient-years.¹ Real-world registry and cohort studies consistently yield higher estimates, with 1-year risks ranging from 8% to 14% and cumulative 3-year risks approaching 20% to 28%.¹ Recent data indicate that incidence and prevalence are broadly comparable between IPF and non-IPF fILDs, though certain subgroups, particularly those with radiological usual interstitial pneumonia (UIP) pattern, experience disproportionately higher rates (Figure 3, Tables S10 and S11). Importantly, adoption of the 2016 definition has led to greater capture of clinically relevant events, producing higher incidence estimates compared with the 2007 definition (Figure 3).

Figure 3.

Pooled estimates of incidence rates of acute exacerbation in fibrotic interstitial lung disease (AE-fILD). The pooled estimated incidence rate for AE-fILD was 91 per 1000 person-years. The pooled estimated incidence rates for AE were higher with using the 2016 definition compared to the 2007 definition, and in those with IPF compared to non-IPF fILD. For calculation of AE events per 1000 person-years for each included study, the reported median, mean, or maximum follow-up durations were used. AE definition: subgroup analyses based on the AE definitions from 2007 and 2016 statements. ILD subgroups: subgroup analyses according to the ILD subgroups. Abbreviations: AE, acute exacerbation; CI, confidence interval; ILD, interstitial lung disease; IPF, idiopathic pulmonary fibrosis; n, number.

Etiology

The etiology of AE-fILD has long been debated, with mounting evidence supporting a multifactorial paradigm. The synthesis of case reports and observational studies highlights the wide spectrum of infectious triggers (Table S12). Documented bacterial precipitants include common respiratory pathogens such as Streptococcus pneumoniae, Haemophilus influenzae, and Pseudomonas aeruginosa.^30-45 Viral triggers are also diverse, spanning influenza, respiratory syncytial virus, rhinovirus, and coronaviruses including SARS-CoV-2.^{31,39,42,44-64} Fungal organisms, particularly Aspergillus spp. and Pneumocystis jirovecii, have also been implicated.^{35,39,40,43,45,65-67} Notably, the reported detection rates of viral and bacterial RNA or DNA vary, with some being identified in only a minority of cases, suggesting either underdetection related to testing timing and techniques or that intrinsic lung vulnerability permits small insults to trigger catastrophic decompensation.¹

Noninfectious precipitants are increasingly recognized. Surgical interventions, particularly thoracic procedures such as lung cancer resection or surgical lung biopsy, carry increased risks of AE-fILD (Table S12). This is likely due to ventilation-related barotrauma with resultant lung stretch, with AE occurrence more frequently observed in the contralateral lung in patients undergoing lung biopsy. Bronchoscopic procedures have occasionally triggered AE-fILD, including BAL^68-70 and transbronchial forceps biopsy^69,71,72 or cryobiopsy.^73-75 Pharmacologic agents, including chemotherapy, immune checkpoint inhibitors, and other systemic therapies, have been associated with AE across case reports and series (Table S12). Other reported contributors include aspiration, air travel, vaccination, and even less invasive diagnostic procedures, though causality remains difficult to establish with confidence. Nevertheless, it is possible that these noninfectious precipitants may result in subtle alveolar epithelial cell injury that triggers AE-fILD.

Risk factors

A variety of patient-, disease-, and treatment-related factors have been evaluated for association with AE-fILD (Table S13). Figure 4 summarizes commonly evaluated factors for their associations with AE in patients with fILD, predominantly based on studies of hospitalized cohorts with possible biases. Low FVC and low Dl_CO are consistently associated with increased risk for AE-fILD. In addition, other markers of increased interstitial lung disease (ILD) severity, including higher gender, age, and physiology (GAP) or ILD-GAP index,^76-80 presence of hypoxemia as indicated by exercise testing or use of home oxygen therapy,^81-83 and worse dyspnea,^77,84-87 are found to be associated with AE-fILD. Other less commonly studied risk factors with potential significance for AE-fILD include reduced 6-minute walk distance^39,82 and disease progression measured by FVC or Dl_CO decline.^81,85,88-91 Recent studies evaluating features on CT identify UIP pattern and extent of honeycomb change as the factors that most consistently predict risk of subsequent AE in both IPF and non-IPF fILD.^87,92-96 Recent studies show positive associations between increased air pollutants, in particular levels of particulate matter <2.5 μm, nitrogen dioxide, and ozone, with risk of AE-fILD.^97-101

Figure 4.

Harvest plots of commonly evaluated factors for associations with acute exacerbation in fibrotic ILD. Each bar represents a single study that investigated the factor of interest, with different colors representing the ILD subgroups studied (IPF, non-IPF fibrotic ILD, and mixed fibrotic ILD). The height of the bar indicates the number of study participants, with the taller bars representing larger sample sizes. Detailed data of individual studies for each factor of interest are provided in Table S13. Abbreviations: BMI, body mass index; CT, computed tomography; GAP, gender, age, and physiology; ILD, interstitial lung disease; ILD-GAP, interstitial lung disease subtypes and gender, age, and physiology; IPF, idiopathic pulmonary fibrosis; PH, pulmonary hypertension; UIP, usual interstitial pneumonia.

Specific patient demographics have not been consistently associated with AE-fILD, and very few studies evaluated the relationships with comorbidities. Multimorbidity may be associated with increased risk of AE-IPF in those individuals not treated with antifibrotic therapies.¹⁰² Comorbidities that have been reported and warrant further investigation include cardiovascular disease,⁷⁶ diabetes mellitus,¹⁰³ and lung cancer, particularly in the presence of coexisting emphysema.⁷⁸ Increased BAL pepsin level was observed in patients with AE-IPF compared to those with stable IPF, suggestive of microaspiration and gastroesophageal reflux being potential risk factors.¹⁰⁴ The use of immunomodulatory medications, including systemic corticosteroids, is associated with increased risk of AE in patients with IPF,^77,105 but not in those with non-IPF fILD who predominantly had connective tissue disease–associated ILD (CTD-ILD).^{39,92,106-109} Recent data from an open-label phase of a clinical trial suggest potential beneficial effect of inhaled treprostinil in reducing AE risk in patients with pulmonary hypertension due to fILD,^110,111 despite conflicting data on the relationship of AE-fILD in patients with coexisting pulmonary hypertension (Table S13).

Few studies have examined the relationships between blood-based or BAL biomarkers, measured either at baseline or throughout the disease, and incidence of AE-fILD (Table S14). Higher monocyte counts^27,79 and serum KL-6 levels^77,112-116 are associated with AE-fILD, with most other candidates having only been investigated in individual studies.

Prognosis

The occurrence of AE-fILD has important prognostic implications, although most studies reporting mortality are retrospective and at risk of selection and misclassification bias. Short-term survival in AE-fILD is reduced, with an estimated 30-day mortality of 30% (95% confidence interval [CI], 25%-35%) and 1-year mortality of 56% (95% CI, 47%-63%) (Figure 5, Figure S1). AE-IPF generally has higher mortality compared to AE of non-IPF fILD. Pooled mortality rates for AE-fILD are also higher using the 2007 definition compared to the 2016 definition, which may reflect prognostic differences in classification criteria or differences in overall mortality over time in the setting of changing prescriber patterns for ILD-directed therapies (Figure 5, Figure S1). In-hospital mortality rates are as high as 80% for patients with AE-fILD requiring ICU admission.¹¹⁷ One study of fILD reported a lower mortality rate with infection-triggered AE, compared to idiopathic AE or other triggers⁵²; however, conflicting results are reported in other studies assessing mortality in idiopathic versus triggered AE-IPF.^43,81,88,118

Figure 5.

Pooled cumulative 30-day (A) and 1-year (B) mortality in patients with acute exacerbation of fibrotic interstitial lung disease (AE-fILD). The pooled cumulative mortality rates for AE-fILD were 30% in 30 days and 56% in 1 year. The pooled cumulative mortality rates were higher in patients with IPF compared to those with non-IPF fILD, as well as using the 2007 definition compared to the 2016 definition. Abbreviations: AE, acute exacerbation; CI, confidence interval; ILD, interstitial lung disease; IPF, idiopathic pulmonary fibrosis; n, number.

A variety of clinical parameters have been evaluated for prognostication in patients with AE-fILD (Figure 6A, Tables S15 and S16). Increased age and preexisting hypoxemia are associated with increased mortality in AE-fILD. The presence of comorbidities also has prognostic implications, with increased mortality risk in presence of coexisting pulmonary hypertension,^119,120 chronic kidney disease,^121,122 and lung cancer.^121,123 While prognostic values of CT findings for underlying ILD vary across studies in AE-fILD, more extensive lung parenchymal involvement and diffuse pattern of acute changes at AE presentations are consistently associated with increased mortality. Reduced PaO₂/FiO₂ ratio at AE presentation,^88,124-127 as well as increased severity measured by predictive scoring systems used in ICUs and sepsis, such as Acute Physiology and Chronic Health Evaluation II (APACHE II) and Sequential Organ Failure Assessment (SOFA),^128-131 are associated with short-term mortality in AE-fILD. Extrapulmonary organ failure is also associated with increased risk of in-hospital death in AE-IPF.¹²⁷

Figure 6.

Harvest plots of prognostic factors for acute exacerbation in fibrotic ILD. Each bar represents a single study that investigated the factor of interest, with different colors representing the ILD subgroups studied (IPF, non-IPF fibrotic ILD, and mixed fibrotic ILD). Studies assessing short-term survival outcomes are represented by solid bars, while studies assessing long-term survival outcomes are represented by patterned bars. The height of the bar indicates the number of study participants, with the taller bars representing larger sample sizes. Detailed data of individual studies for each factor of interest are provided in Tables S15-S17. (A) Clinical parameters. (B) Biomarkers measured during AE presentation. Abbreviations: APACHE, Acute Physiology and Chronic Health Evaluation; BMI, body mass index; CRP, C-reactive protein; CT, computed tomography; GAP, gender, age, and physiology; ILD, interstitial lung disease; ILD-GAP, interstitial lung disease subtypes and gender, age, and physiology; IPF, idiopathic pulmonary fibrosis; KL-6, Krebs von den Lungen-6; LDH, lactate dehydrogenase; PPF, progressive pulmonary fibrosis; SP-D, surfactant protein D; WBC, white blood count.

Different clinically available and novel blood-based biomarkers have been evaluated for mortality risk associations in AE-fILD, and a small number of studies have investigated BAL samples (Table S17, Figure 6B). Increased lactate dehydrogenase levels,^{36,42,119,132-136} as well as higher neutrophil count in blood and BAL fluid,^{26,38,119,137} are associated with poor prognosis in AE for both IPF and non-IPF fILD. Several multivariable models have been developed to predict outcomes of AE-IPF, though none have been validated prospectively.^126,138

Proposed definitions

Based on current evidence on clinical and radiologic characteristics, prevalence, and outcomes, a revised definition of AE-fILD was proposed to establish a framework that is broadly applicable across both IPF and non-IPF fibrotic ILDs, and useful in both clinical practice and research. A key concept underpinning the revised definition is that the term “acute exacerbation” is used differently in fILD than in other chronic respiratory diseases (eg, asthma, chronic obstructive respiratory disease [COPD], and bronchiectasis). In those conditions, AE refers to any episode of acute symptomatic deterioration, regardless of etiology or severity. In contrast, in fILD, AE is reserved for a distinct clinicopathologic syndrome characterized by confirmed or presumed DAD that is associated with high morbidity and mortality. To reduce misclassification and explicitly acknowledge that patients with fILD frequently experience acute symptom worsening not attributable to DAD, the term “acute respiratory worsening” has been proposed to describe these broader, often transient events and to distinguish them from AE-fILD.

Acute exacerbation of fibrotic interstitial lung disease

An AE-fILD is proposed to be defined as an acute respiratory event characterized by increased respiratory symptoms or signs and associated with radiologic or histologic features consistent with DAD (with or without superimposed organizing pneumonia) in a patient with known or newly diagnosed fILD (Figure 7, Table S18). Although lung biopsy is rarely feasible in patients with AE-fILD due to the acute nature and severity of illness, studies in which tissue was obtained have identified DAD as the predominant histopathologic pattern, albeit most data are from patients with IPF or a UIP pattern of lung fibrosis (Table S19). The histopathologic anchor is considered essential to maintain diagnostic specificity, prognostic relevance, and conceptual clarity, but it is emphasized that biopsy should not be pursued in most cases to confirm the diagnosis of AE-fILD.

Figure 7.

Proposed definitions and diagnostic criteria for acute respiratory worsening and acute exacerbation in fibrotic ILD. Abbreviations: AE-fILD, acute exacerbation of fibrotic interstitial lung disease; ARW, acute respiratory worsening; CT, computed tomography; DAD, diffuse alveolar damage; GGO, ground glass opacity; ILD, interstitial lung disease; OP, organizing pneumonia.

Because biopsy is rarely performed, the diagnosis of AE-fILD relies primarily on clinical and radiologic surrogates of DAD (Figure 7). Clinical features of AE-fILD include new or worsening dyspnea, exercise intolerance, and/or worsening oxygenation status, with an onset of typically less than a month in a patient with a previous or new diagnosis of fILD. The proposed clinical criteria were expanded from the 2016 definition to include hypoxemia, given its prevalence of over 90% in published literature (Figure S2). Incorporating cough was considered to more fully capture the spectrum of clinical presentations. However, given the broad differential diagnosis of new or worsening cough in fILD, it was concluded that this symptom lacks adequate specificity and would rarely be the sole defining clinical feature. In clinical practice, the emergence or progression of cough should prompt an evaluation for other signs and symptoms of AE-fILD.

Symptom onset of typically less than 1 month was retained from the 2016 definition, informed by evidence synthesis demonstrating a mean symptom duration of 14 days in recent studies (95% CI, 5-22 days; Table S20). Among studies that included patients with mixed fILD and employed varying definitions of AE-fILD, the mean symptom duration was 11 days (95% CI, 8-14 days), indicating that most patients present within approximately 2 weeks of symptom onset. Although a shorter threshold of 14 days was considered based on these data, it was ultimately determined that a <1-month window was more appropriate to prioritize sensitivity for more indolent clinical presentations.

The proposed radiologic criteria consist of new and/or worsening bilateral GGO with or without consolidation on CT, superimposed on a background pattern of fILD (Figure 8, Figures S3 and S4). The 2016 criterion that findings not solely be explained by cardiac failure or fluid overload has been integrated into the radiological criterion. It was acknowledged that the specificity of GGO for DAD in this context is uncertain and is likely lower in non-IPF fILD than in IPF, given the greater frequency of comorbid conditions and treatments that can mimic AE-fILD. Differential diagnoses for bilateral GGO in this setting commonly include pulmonary edema, pneumonia due to bacterial, viral, or opportunistic pathogens (particularly in patients treated with immunomodulatory medications), drug-induced pneumonitis, procedure-related lung injury, and other causes. While radiation pneumonitis typically presents with changes confined to the radiated area, bilateral GGOs occur in rare cases.^139,140 These alternative etiologies should be systematically considered and worked up as described in the “Diagnostic approach” section below. However, these events can also precipitate DAD, and distinguishing them from an AE-fILD at presentation is often impossible with the currently available diagnostic tools. The response to treatment and subsequent clinical course may help clarify whether DAD was the underlying process, but such information is generally unavailable at the time of initial presentation. Accordingly, the 2016 categorization is retained for triggered AE, when a specific inciting factor is identified, and idiopathic AE, when no cause can be determined despite a comprehensive evaluation (Figure 9). Literature review suggested that consolidation on CT is observed more frequently in AE-fILD than in AE-IPF, consistent with a higher prevalence of histopathologic organizing pneumonia in fILD and AE-fILD (Figure S3). However, isolated consolidation was considered too nonspecific to indicate underlying DAD in the absence of concomitant increase in GGO.

Figure 8.

Radiological features of acute exacerbation in fibrotic interstitial lung disease (ILD). At baseline, representative axial images of computed tomography (CT) show background fibrotic nonspecific interstitial pneumonia pattern in a patient with systemic sclerosis–associated ILD (Left). A month later, the patient was admitted to the hospital after experiencing worsening exertional dyspnea for 5 days followed by 2 days of persistent fever. The repeat CT showed superimposed bilateral ground glass opacities (Right). No triggers were identified, with negative microbiological assessments on BAL. The patient was diagnosed with idiopathic acute exacerbation.

Figure 9.

Classification of acute respiratory worsening and acute exacerbation in fibrotic interstitial lung disease. *In patients with coexisting COPD or CPFE. Abbreviations: AE, acute exacerbation; ARW, acute respiratory worsening; COPD, chronic obstructive pulmonary disease; CPFE, combined pulmonary fibrosis and emphysema; DAD, diffuse alveolar damage; ILD, interstitial lung disease.

The histopathologic criterion was included to improve diagnostic accuracy when tissue is available but is not required to establish AE-fILD clinically. Histologically, DAD is characterized by thickened alveolar septa due to interstitial edema and inflammation, pneumocyte hyperplasia, and surface hyaline membrane formation in the acute phase (Figure 10). The acute phase is followed by an evolution to abundant, loosely organizing connective tissue within the interstitium and filling alveolar spaces in the organizing phase. Importantly, the inclusion of a histopathologic criterion does not imply that biopsy should be pursued for diagnosis of AE-fILD, but rather that its presence, when available, provides confirmatory evidence. This proposed framework therefore anchors the definition of AE-fILD in its underlying pathobiology, DAD, while recognizing that diagnosis must rely on clinical and imaging features consistent with this process in most cases.

Figure 10.

Histopathology of acute exacerbation in fibrotic interstitial lung disease. The background lung shows a pattern of usual interstitial pneumonia, with destructive subpleural fibrosis, embedded honeycombing, and fibroblastic foci (Left, hematoxylin and eosin [H&E] stain, 40×). There is superimposed diffuse alveolar damage in the acute phase, with thickened alveolar septa due to interstitial edema and inflammation, pneumocyte hyperplasia and atypia, and notable surface hyaline membranes lining the alveolar walls (Right, H&E stain, 200×).

Acute respiratory worsening

It was recognized that acute respiratory deteriorations are very common in patients with fILD and encompass a heterogeneous group of clinical events with distinct mechanisms and outcomes. Patients with fILDs often receive immunomodulatory medications,¹⁴¹ which increase susceptibility to typical and atypical infections. Many fILDs are associated with systemic conditions that involve pleural, gastroesophageal, cardiac, and neuromuscular comorbidities, each of which may contribute to acute respiratory deterioration. To capture these heterogenous conditions, facilitate timely differential diagnosis, and streamline the diagnostic approach, the broader construct of ARW was proposed, defined as an acute respiratory event characterized by increased respiratory symptoms or signs in a patient with fILD, encompassing both AE and non-AE-related events (Figures 7 and 9, Figure S5). This proposed definition was developed to be broadly applicable across clinical (both hospitalized and outpatient) and research settings and to identify a recognizable clinical syndrome, analogous to approaches used in other chronic respiratory conditions.

ARW criteria include new or worsening dyspnea, exercise intolerance, and/or worsening oxygenation status, with an onset of typically less than month in a patient with a previous or new diagnosis of fILD. These criteria are identical to the clinical features of AE-fILD, but do not require the specific imaging or histopathologic pattern suggestive of DAD. ARW has a broad spectrum of causes that includes non-AE-fILD causes such as airway infection, COPD exacerbation (in patients with coexisting COPD or combined pulmonary fibrosis and emphysema), pleural effusion, pneumothorax, pulmonary edema, pulmonary embolism, and respiratory muscle weakness. ARW also includes conditions that mimic an AE-ILD but generally have a more favorable prognosis when promptly recognized and treated, such as pneumonia, drug- and radiation-induced pneumonitis, procedure-related complications, and aspiration pneumonitis. These conditions are usually responsive to treatment with antibiotics and/or immunosuppressive therapies, without the development of DAD. They are often transient with good recovery and do not carry the poor prognostic implications associated with AE-fILD.

It is important to distinguish ARW from rapidly progressive ILD, which is most often seen in myositis-associated ILD (eg, anti-melanoma differentiation-associated gene 5 [anti-MDA5] dermatomyositis) and typically requires high-dose immunosuppression.^142,143 Rapidly progressive ILD is characterized by a decline in lung function usually over weeks to months, which is typically associated with acute or subacute patterns of lung injury and may be the initial presentation of a new diagnosis of ILD,¹⁴⁴ whereas ARW typically evolves more acutely and requires preexisting fILD. Therefore, patients who initially present with rapidly progressive ILD would not meet the criteria for an AE-fILD, even though histological features of DAD and/or acute fibrinous organizing pneumonia may be observed. Conversely, some cases of CTD-ILD, including a minority of patients with idiopathic inflammatory myopathies,¹⁴⁵ may experience ARW with DAD (ie, AE-fILD), leading to progression of the underlying fILD and poor outcomes. Recent American College of Rheumatology/American College of Chest Physicians guidelines discuss the clinical trajectory and treatment pathway for rapidly progressive ILD in more detail.¹⁴⁶ As knowledge advances, the distinction between usual disease trajectory, ARW, and AE-fILD in patients with CTD-ILD may require further refinement in future multidisciplinary guidelines. On the other hand, clinical deteriorations in patients with PPF are insidious with gradual, persistent worsening of symptoms over months, with radiological features of fibrotic progression.¹⁴⁷ Thus, detailed assessments of patients’ clinical presentations with appropriate investigations are necessary.

This proposed integrative framework for classification of ARW emphasizes systematic evaluation of ARW to identify specific, potentially reversible causes, while recognizing that many triggers overlap with AE-fILD pathobiology. It aligns the approach to acute manifestations of fILD with the framework applied to chronic ILD,¹⁴⁸ providing a consistent foundation for clinical care, research classification, and future therapeutic trials.

Diagnostic approach

Among patients with fILD experiencing ARW, diagnostic evaluation necessitates targeted and parallel testing identifying characteristic features of AE-fILD and potential triggers, while excluding alternative causes of ARW such as pulmonary edema (Figure 11). A broad differential and clinical workup is needed given the absence of highly sensitive and specific tests for AE-fILD. Importantly, diagnostic evaluation of ARW should be guided by the patient’s preceding clinical trajectory and realistically available treatment options, while accounting for their preferences, including goals of care, and estimated prognosis. In patients who are candidates for lung transplantation, a more comprehensive workup may be warranted to identify reversible causes and guide aggressive management. Conversely, for patients who are ineligible for transplant or who prioritize comfort-focused care, the evaluation should be streamlined to avoid burdensome testing that is unlikely to alter their clinical course. Moreover, differences in the availability and access to diagnostic tests and management approaches across healthcare services and regions contribute to global variations in clinical care for patients with ARW and AE-fILD.

Figure 11.

Diagnostic algorithm for acute respiratory worsening in fILD. ^*Infection assessment is commonly performed for exclusion before the diagnosis of idiopathic episodes. Tests performed according to local guidance and suspected pathogens, including opportunistic infections. ^†An alternative cause may be identified with chest X-rays. ^‡BAL and endotracheal aspirate are considered in selected patients, when feasible (typically in mechanically ventilated cases). ^§Lung biopsy is not a standard evaluation and is only performed in exceptional cases following multidisciplinary discussion. Abbreviations: AE, acute exacerbation; AE-fILD, acute exacerbation of fibrotic interstitial lung disease; ARW, acute respiratory worsening; BNP, B-type natriuretic peptide; CBC, complete blood count; CT, computed tomography; GGO, ground glass opacity; ILD, interstitial lung disease; NAAT, nucleic acid amplification test; PCR, polymerase chain reaction.

Clinical history

A focused clinical history is necessary to clarify the tempo and character of deterioration and to identify clues that suggest specific causes of ARW. AE-fILDs typically evolve within 1 month or less, producing progressive and persistent respiratory symptoms and signs, as described previously. Rapid onset of symptoms over hours suggests alternative diagnoses to AE-fILD including bacterial or viral respiratory tract infections, pulmonary embolism, pneumothorax, or acute pulmonary edema. Opportunistic infections and tuberculosis, as well as treatment- and procedural-triggered ARWs, may develop over days or weeks. Productive cough and fever generally suggest infectious triggers. Other symptoms such as chest pain, vomiting, or choking often suggest alternative diagnoses, including pulmonary embolism, pneumothorax, or aspiration.

Patients with recent hospitalizations may develop nosocomial infections, and those who have recently undergone surgery may develop acute symptoms in the postoperative period. New medications, especially those with known pulmonary toxicity,¹⁴⁹ should prompt consideration of drug-induced pneumonitis. Environ­mental and occupational exposures that can trigger acute events, such as in fibrotic hypersensitivity pneumonitis, should be considered. Immunosuppressant exposure may increase the suspicion for opportunistic infections, including tuberculosis. In addition, arthralgias, myalgias, or rashes may suggest a flare of an underlying autoimmune condition that can occur in conjunction with ARW in CTD-ILD. Furthermore, clinical features of systemic manifestations of CTD, such as anemia and respiratory muscle weakness, can present as ARW unrelated to AE.

Physical examination

Patients presenting with AE-fILD often have physical signs of respiratory distress, including hypoxemia, tachypnea, and accessory muscle use, which are variably present in non-AE ARW depending on the severity of acute deterioration. Inspiratory crackles on chest auscultation are common but nonspecific and may simply be a manifestation of the underlying fILD. Patients with CTD-ILD may have systemic signs related to the underlying autoimmune disease, for example pathognomonic skin rash in dermatomyositis or paradoxical breathing with diaphragmatic weakness. Pertinent examination findings may yield diagnostic clues that point toward non-AE ARW, including asymmetric breath sounds in pneumothorax or pleural effusion, elevated jugular venous distention, cardiac gallops or murmurs, and peripheral edema in patients with heart failure or cor pulmonale, and unilateral leg swelling in patients with venous thromboembolism.

Laboratory evaluation

Given the lack of laboratory tests that are highly sensitive and specific for AE-fILD, noninvasive laboratory tests are primarily used to identify AE-fILD triggers and exclude alternative causes of ARW. Peripheral blood cell counts and differentials may reveal leukocytosis or anemia that could indicate infectious etiologies or hemorrhage. Abnormal chemistries, cardiac troponin, brain natriuretic peptide, and D-dimer may indicate renal failure, pulmonary edema, or pulmonary embolism. For investigation of infective-related ARW, sputum and blood cultures, urinary antigens, respiratory viral polymerase chain reaction tests, and nucleic acid ­amplification tests may confirm specific microbial etiologies. Tests for opportunistic and atypical infections should be considered in patients on long-term immunosuppression, including Pneumocystis jirovecii and mycobacteria. Elevated serum procalcitonin levels may suggest a higher likelihood of bacterial infection, though with marginal test performance.^150,151 Autoimmune serologic evaluation is warranted in patients with ARW as the initial presentation of ILD, including MDA5 antibody testing.^152,153

Radiologic imaging

Imaging is central to the diagnosis of AE-fILD, with CT findings of new or worsened GGOs with or without consolidation on a background of fibrotic changes such as reticulation, traction bronchiectasis, and honeycombing. The acute changes are usually diffuse but can also be multifocal or peripheral in some cases, particularly in early presentations.^154,155 While it is challenging to definitively exclude non-AE ARW with similar acute changes, the absence of such changes on CT excludes AE-fILD and indicates other causes of ARW. Chest CTs may point toward specific triggers of AE or other causes of ARW depending on the clinical context. Lobar consolidation may indicate pneumonia, upper-lobe predominant ground glass may suggest alveolar hemorrhage, perihilar ground glass and/or pleural effusions may indicate pulmonary edema, and focal or diffuse centrilobular nodules can occur in aspiration or inflammatory/infectious bronchiolitis. CT pulmonary angiography (CTPA) along with Doppler ultrasonography may be indicated when pulmonary embolism is on the differential diagnosis. For optimal assessment of patients with suspected AE-fILD, a concurrent inspiratory HRCT for lung parenchymal assessment immediately before contrast administration for CTPA can be performed. Transthoracic echocardiography is helpful in suspected cardiogenic pulmonary edema or as a screening test for pulmonary hypertension, allowing simultaneous evaluation of left and right ventricular function, valvular abnormalities, and estimation of pulmonary artery pressure.

Invasive testing

Invasive testing may be selectively applied, typically in patients who are mechanically ventilated, given the increased risk of adverse outcomes. Bronchoscopy with BAL can yield diagnostic information in immunosuppressed patients or when infection or diffuse alveolar hemorrhage is suspected but not confirmed noninvasively. If BAL can be performed safely, broad microbiologic testing for community/hospital-acquired and opportunistic infections should be considered, depending on the clinical context. Lung biopsy, whether bronchoscopic or surgical, is rarely performed in suspected AE-fILD because of high mortality risk and limited impact on management decisions.^156,157 Coronary artery and right heart catheterization may be considered in patients with clinical features concerning for cardiac and pulmonary vascular disease.

Management approach

Given that no therapeutic interventions are definitively proven to improve outcomes of AE-fILD, management of AE-fILD remains primarily supportive with a focus on symptom care, correction of hypoxemia, and treatment of potential triggers or associated conditions. Preventive measures include smoking cessation counseling, minimizing exposure to other inhalational triggers (eg, air pollution), optimization of comorbidities, infection prevention through vaccination when appropriate, measures for avoidance of infection (use of face mask in public indoor settings and regular hand washing), and application of lung-protective ventilation strategies where clinically indicated. Lung transplantation during AE-fILD may be performed in carefully selected patients.^31,158-161

Many patients with AE-fILD receive broad-spectrum antibiotics, guided by local antimicrobial recommendations and spectrum of potential pathogens, including opportunistic infections. Evidence regarding the impact of specific antibiotic regimens in AE-fILD is sparse and largely derived from small retrospective studies, which have suggested a possible survival benefit with azithromycin (Table S21, Figure S6).^121,128,162 Procalcitonin-guided antibiotic use has not been shown to impact health outcomes in AE-IPF.¹⁶³ Invasive mechanical ventilation and extracorporeal membrane oxygenation in AE-fILD are associated with very poor prognosis and are considered in select situations, such as a bridge to lung transplantation or when there is reasonable potential for meaningful recovery. Lung-protective ventilation strategies with low positive end-expiratory pressure are preferred.^164,165 There is increasing interest in the use of noninvasive ventilation (NIV) and high-flow nasal cannula oxygen therapy (HFOT) for treating acute respiratory failure, including AE-fILD.^62,166-173 While these non-invasive modalities have not been shown to convey survival benefit in AE-fILD, they may improve oxygenation and provide symptom relief. They represent alternative options to avoid intubation or for palliative support, warranting further evaluation. Careful patient selections for invasive ventilation and alternative strategies with NIV and HFOT are necessary, and should be guided by patients’ advanced care plan, anticipated trajectories for fILD and AE, the likelihood of lung transplantation, treating team expertise, and the availability of local resources.

AE-directed therapies

The role of immunomodulatory and antifibrotic therapies in AE-fILD continues to evolve, with a growing number of clinical trials and observational studies that were conducted primarily in AE-IPF (Figures 12-14). Most studies are of low quality with small sample size and retrospective data collection. The impact of antifibrotic and other emerging therapies targeting the underlying fILD on the occurrence of AE have also been studied.

Figure 12.

Summary of observational studies for systemic corticosteroids (CS) on mortality of acute exacerbation in fibrotic ILD. *Multicenter cohort. ^†Reduction in CS maintenance dose of >10% within 2 weeks of admission (excluding reduction after steroid pulse therapy). ^‡Administrative cohort. ^§CS use: at least methylprednisolone pulse ≥500 mg/day or high-dose prednisolone (≥0.5 mg/kg) for 2 days or more. ^llPulse CS: ≥250 mg/day of methylprednisolone or equivalent within 7 days of hospitalization; no pulse CS: usually 1 mg/kg/day of methylprednisolone. **High-dose CS: prednisolone >1 mg/kg. ^††High-dose CS: 500-1000 mg/day of IV methylprednisolone for 3 days; low-dose CS: 100-200 mg/day of IV methylprednisolone for at least 5 days; including lung transplantation as an event. ^‡‡CS: Prednisolone (≥0.5 mg/kg) for 3 days or more within 3 days of hospitalization; events included transplantation during hospitalization; non-IPF cohort receiving steroids had reduced risk of all-cause mortality compared to IPF cohort (P = .04). ^§§High-dose CS: prednisolone ≥0.6 mg/kg. Abbreviations: CS, corticosteroids; ILD, interstitial lung disease; IPF, idiopathic pulmonary fibrosis; IV, intravenous.

Figure 13.

Summary of antifibrotic and other therapies on the prevention of acute exacerbation in fibrotic interstitial lung disease (AE-fILD). Nintedanib and pirfenidone may reduce AE risk in IPF, with more heterogeneous results in non-IPF fILD. High-dose nerandomilast was associated with reduced AE risk among patients with progressive pulmonary fibrosis. Risk ratios were estimated using RevMan (including meta-analyses for ≥2 studies). X: Data extracted from individual studies; if more than one study, the direction of effects was determined based on the majority of studies. *During listing for lung transplantation. ^†For nerandomilast: low-dose = 9 mg twice daily, high-dose = 18 mg twice daily. Abbreviations: AE, acute exacerbation; HR, hazard ratio; ILD, interstitial lung disease; IPF, idiopathic pulmonary fibrosis; OR, odds ratio; RCT, randomized controlled trial; RR, risk ratio.

Figure 14.

Summary of antifibrotic and nonsteroidal immunomodulatory therapies on mortality of acute exacerbation in fibrotic interstitial lung disease (AE-fILD). Most studies of antifibrotic therapies showed no benefit on mortality in AE-fILD, with the majority being conducted in patients with IPF. Various nonsteroidal immunomodulatory therapies were investigated for treating AE-fILD, predominantly through observational studies, and most showed no mortality benefit. Studies of PMX-DHP showed inconsistent findings between short-term and long-term survival outcomes. Both RCTs of rhTM and intravenous cyclophosphamide did not demonstrate survival benefits in AE-IPF. Risk ratios were estimated using RevMan (including meta-analyses for ≥2 studies). X: Data extracted from individual studies; if more than one study, the direction of effects was determined based on the majority of studies. Abbreviations: HR, hazard ratio; ILD, interstitial lung disease; IPF, idiopathic pulmonary fibrosis; IVIG, intravenous immunoglobulin; OR, odds ratio; PEx, plasma exchange; PMX-DHP, polymyxin B–immobilized fiber column direct hemoperfusion; RCT, randomized controlled trial; rhTM, recombinant human soluble thrombomodulin; RR, risk ratio.

Prevention of AE-fILD

Among patients with IPF, randomized controlled trials (RCTs) and post hoc analyses suggest that nintedanib and pirfenidone may reduce the risk of AE at 9 to 12 months (Table S22, Figure S7).^174-176 Delay in time to first adjudicated AE was observed with nintedanib compared to placebo in one of the 2 INPULSIS trials.¹⁷⁵ Observational studies of longer follow-up duration beyond 12 months have yielded mixed results for pirfenidone, with no effects in some studies^177-180 and benefit in others.^102,181 No significant reduction in AE-IPF has been observed in the RCT of low- or high-dose nerandomilast.¹⁸² Literature for effects of antifibrotic therapies in preventing AE of non-IPF fILD is even more heterogenous. A small RCT of patients with fibrotic hypersensitivity pneumonitis favored pirfenidone over control in prevention of AE³; however, no difference between pirfenidone and controls was observed among a pooled analysis of patients with fibrotic nonspecific interstitial pneumonia or mixed fILD over longer follow-up^179,183 (Table S22, Figure S7). High-dose nerandomilast was associated with reduced AE risk among patients with PPF⁶ (Table S23, Figure S8).

Treatment of AE-fILD

High-dose systemic corticosteroids remain the most widely used therapy in AE-fILD, despite the lack of RCT-based data (Figure 12). A recent systematic review suggests potential differential effects of systemic corticosteroids between patients with IPF and non-IPF fILD.¹⁸⁴ There are inconsistent findings for survival outcomes across studies of AE-IPF, ranging from no benefit to increased mortality and concern for potential harm. This challenges the previous guideline recommendation of using systemic corticosteroids in the majority of patients with AE-IPF.¹⁸⁵ The first placebo-controlled RCT of systemic corticosteroids for AE-IPF, EXAFIP2, is currently in progress to address these critical efficacy and safety uncertainties.¹⁸⁶ In contrast, high-dose systemic corticosteroids at >1.0 mg/kg prednisolone-equivalent in AE of non-IPF fILD have been associated with improved short-term survival,³⁵ with similar differential effects seen in subgroup analyses between IPF and non-IPF fILD in another study.⁵⁴ Thus, systemic corticosteroids should be used cautiously in AE-fILD, with a more favorable role in those with non-IPF fILD compared to IPF. When patients with AE-IPF are treated with systemic corticosteroids, a short course of high-dose therapy may be preferred, as this also minimizes the concern of increased mortality associated with immunosuppression in this population.

Evidence for other immunomodulatory medications as sole interventions or in addition to systemic corticosteroids in AE-fILD is limited, with the majority being small observational studies that were primarily conducted in Japan (Table S24, Figure S9). Confounding by indication remains a major concern, as nonsteroid immunosuppression is frequently used in patients with poor response to glucocorticoids. The only RCT examining immunomodulatory medications in AE-IPF evaluated intravenous cyclophosphamide.¹⁸⁷ All patients received standardized high-dose glucocorticoids, and the addition of intravenous cyclophosphamide conferred no survival benefit but rather demonstrated a trend toward higher 90-day mortality. Other small cohort studies and propensity-matched analyses of intravenous cyclophosphamide, cyclosporine, and tacrolimus have likewise failed to demonstrate any survival benefit. Large database analyses indicate that combined immunosuppression for AEs in both IPF and non-IPF fILD confers no survival advantage and may be associated with worse short-term outcomes.^188-191 Preliminary findings from case series and small pilot studies evaluating intravenous immunoglobulin and/or plasma exchange with and without other immunomodulatory medications suggest improved short-term survival.^14,33,192 A phase 2 RCT investigating combined plasma exchange, rituximab, and intravenous immunoglobulin has recently been completed, but findings are yet to be published.¹⁹³

Polymyxin B–immobilized fiber column direct hemoperfusion has been investigated as an adjuvant therapy for AE-fILD (Table S25, Figure S10). Early data suggest potential mortality benefit at 90 days and 1 year,^194-196 with inconsistent findings at 30 days.^131,195,197 Previous observational studies of recombinant human soluble thrombomodulin (rhTM) suggested potentially improved mortality in AE-IPF (Table S26, Figure S11),^198-201 but a more recent RCT demonstrated a trend toward harm of rhTM as add-on therapy to high-dose systemic corticosteroids with no survival benefits and increased bleeding events.²⁰² Thus, these interventions are currently not considered standard treatment options for AE-fILD.

Most studies found no benefit of antifibrotic therapies on mortality in AE-fILD^{121,125,126,181,203-206} (Table S27, Figure S12). However, most literature examining the role of antifibrotics in this setting comes from post hoc analyses of RCTs or small observational studies with varying cohort characteristics and event rates. While health database studies show potential beneficial effects of antifibrotic therapies with improved long-term survival following respiratory-related hospitalizations in patients with IPF, the admission diagnoses could not be ascertained.²⁰⁷ Future high-quality, prospective studies that use a standardized definition and severity description of AE are needed to clarify optimal timing, duration, and impact of antifibrotics in both IPF and non-IPF fILD subtypes.

Acute exacerbation as a clinical trial endpoint

Including AE as an endpoint for clinical trials of fILD may enhance clinical relevance and impact as a patient focused outcome. While clearly desirable, using AE as a trial endpoint is associated with statistical, diagnostic, and operational challenges.

One of the first clinical trials in IPF to utilize AE as a secondary endpoint was a double-blinded, placebo-controlled trial of pirfenidone conducted in Japan, which was terminated early when AEs were only observed in the placebo arm.¹⁷⁴ Subsequently, AE has been included either as a standalone secondary endpoint or as part of a composite secondary endpoint in several pivotal IPF and PPF RCTs.^{5,6,175,182,208-210} These trials typically used the 2007 or 2016 AE definition.^1,211 A recent report from a collaborative symposium involving ILD experts, patient representatives, statisticians, and representatives from the US Food and Drug Administration and the National Institutes of Health supported incorporating AE as a component of composite endpoints in IPF clinical trials, noting that it is a candidate endpoint that is feasible, efficient, and in alignment with how patients “feel, function, or survive.”²¹² The report noted the potential need for central adjudication of such events.

Although trialists generally consider AE as important, recent experience has revealed major challenges for its use as a clinical trial endpoint, raising doubts about feasibility and utility:

1. There has been no prior proposed definition of AE for patients with non-IPF fILDs. Consequently, AE-fILD has typically been defined using criteria developed for AE-IPF, consistent with the approach used to define PPF in the recent practice guideline.¹⁴⁷ This gap is now addressed by the proposed AE-fILD definition, which is applicable to both IPF and non-IPF fILD and is intended to support future research and facilitate refinement of the evidence base.

2. Currently, there is no established consensus on the optimal adjudication process for AE-fILD in clinical trials. While central adjudication was prespecified in some RCTs,^208-210 post hoc adjudication was performed in others, including the phase 3 RCTs of nintedanib.^5,175 Notably, central adjudication resulted in a 40% reduction in confirmed or suspected AEs compared to investigator-reported events, highlighting the challenges of accurate AE identification in global clinical trials when using stricter criteria.²¹³ In the recent phase 3 RCTs of nerandomilast,^6,182 only investigator-reported AEs were assessed.

3. The discrepancy between investigator-reported and centrally adjudicated events may be at least partially explained by variability in clinical practice in terms of how AEs are identified, managed, and reported.^214,215 In addition, seasonal variation and environmental exposures, including air pollution, may influence the incidence of AE, potentially contributing to geographic and temporal variability.^{100,101,216,217}

Analysis of the phase 3 RCT of nintedanib in patients with IPF indicated that investigator-reported AEs were associated with similarly poor outcomes compared to adjudicated confirmed or suspected AEs.^203,213 An expert-based practical algorithm for identifying respiratory-related hospitalizations, comprising definite or suspected AE as well as other respiratory-related causes, has recently been proposed, focusing on a broader concept of serious events requiring hospitalization that are meaningful to patients and the healthcare system.²¹⁸ While the proposed algorithm for respiratory-related hospitalizations might be a useful pragmatic approach, further validation is needed. In addition, inclusion of AE as an endpoint in clinical trials of fILD is crucial to detect these events as an adverse outcome of the investigational drugs, as seen with bexotegrast.²¹⁹

Future research

The proposed updated framework presented in this state-of-the-art document establishes standardized definitions to support multicenter collaborative research on AE-fILD. Nevertheless, there remains an urgent need for strategies that enable prediction, risk stratification, early detection, and prognostication of AE-fILD, in order to elucidate the drivers of these events and facilitate targeted and timely therapeutic trials.

No validated tools predict AE-fILD, despite advances in the development of clinical prediction models as well as many studies identifying predictive and prognostic biomarkers in fILD. Identifying biologic signatures that predict AE-fILD could transform clinical management. Priorities include the discovery and validation of molecular biomarkers that signal heightened AE-fILD risk. In parallel, digital and physiologic monitoring tools such as wearables and home spirometry offer a promising avenue for early detection. These modalities may capture subtle physiological changes that precede clinical recognition, enabling timely intervention. While interstitial lung abnormalities are reported to have increased risk of respiratory failure and drug toxicity, the relationship with acute exacerbation is yet to be elucidated and warrants future investigation.²²⁰

Risk stratification tools are also needed. AE-fILDs not requiring hospitalization have been reported in both clinical trials and observational studies.^4,221 While such events may fall within the proposed ARW framework, there remains a need for tools that can predict which events will progress to more severe AE-fILD events. The clinical impact of non-AE ARW and the associated mortality risk should also be better defined. A key objective is to develop integrative risk models that combine multimodal data to stratify patients by AE-fILD risk. This would support personalized prevention and monitoring strategies and also inform trial design. Fundamental questions remain: How can we stratify patients at risk of AE-fILD to target surveillance and prophylactic therapies? What physiological perturbations are detectable remotely before symptoms emerge? Can we define biomarker signatures that reliably predict AE-fILD severity?

Current theories suggest AE-fILDs occur due to either an intrinsic acceleration of fibrosis or increased epithelial susceptibility to an external insult that drives deterioration. Developing therapeutic interventions is challenging without a better understanding of the pathogenic mechanisms underpinning AE-fILD. It is therefore a priority to define the cellular and molecular events that precipitate lung injury in AE-fILD and clarify whether the mechanisms differ between IPF and non-IPF fILDs. By establishing the potential triggers, we may identify population-based interventions, such as vaccination, prophylactic antibiotics, or antiviral therapy, which could prevent or limit the severity of these episodes and ultimately reduce mortality associated with these events. Identification of specific pathways driving AE would also permit targeted therapies, which may differ across the spectrum of fILDs.

Existing treatment approaches for AE-fILDs are largely empirical and focused on supportive care, with limited evidence guiding practice. While systemic corticosteroids are widely used in the management of AE-fILD, their potential differential effects in IPF and non-IPF fILD warrant rigorous evaluation, as well as their optimal dosing, duration, and tapering strategies if proven beneficial. The use of immunomodulatory, antifibrotic, and novel agents in the acute setting remains speculative due to paucity of controlled studies. Nonpharmacologic interventions including ventilator strategies, high-flow nasal cannula, extracorporeal membrane oxygenation, and integration of palliative care are inconsistently applied, often driven by institutional capacity or clinician preference rather than evidence. Future research should prioritize well-designed multicenter pragmatic RCTs and adaptive platform trials to efficiently evaluate multiple interventions relevant to diverse care settings.

Efforts to improve AE-fILD outcomes are further hindered by a lack of validated surrogate endpoints and outcome measures, with heterogeneous reporting limiting comparability and generalizability. Defining a core set of outcomes, including physiological measurements, survival, healthcare utilization, functional recovery (eg, supplemental oxygen needs), future AE risks, and patient-reported markers, will promote consistency and enable cross-study synthesis. At the health systems level, research should address variability in recognition, triage, and management of AE-fILD, often stemming from fragmented care pathways or limited access to subspecialty centers. Once established, the scalability and real-world application of risk prediction models and remote monitoring for identification and triage of AE-fILD need to be explored. Implementation science can bridge the gap between evidence and practice by testing system-level interventions such as structured care pathways and digital triage tools, with the goal of reducing variability and improving the timeliness and appropriateness of care delivery for improved health outcomes in AE-fILD.

Conclusion

AE-fILDs have high mortality and burden on patients and families. The frequency and lethality of AEs across fILDs underscore the need for prevention, rapid recognition, and optimized clinical care. At the same time, the expanding list of identifiable triggers creates tangible opportunities for the development of interventions through infection control, perioperative risk reduction, and judicious use of systemic therapies. Advancing AE-fILD care requires aligning mechanistic, therapeutic, and systems-level research through global collaboration. Data sharing consortia, biobanks, interoperable platforms, and harmonized outcome measures are essential to enable high-quality, large-scale studies. Given the urgent need for a coordinated, collective research effort, the proposed updated definition of AE-fILD and conceptual framework of ARW in this state-of-the-art document provide a standardized approach for clinical and research application, enabling the generation of new knowledge with the potential to meaningfully transform patient care and improve outcomes.

Author contributions

Y.H.K., F.L., F.J.M., C.J.R., and A.J.P. were responsible for project conception and development. Y.H.K., A.A., B.F.C., E.F., S.B.M., C.A.N., and A.J.P. performed literature review. All authors were involved in the discussion and were divided into writing groups for drafting of each section, with Y.H.K. compiling the complete manuscript. All authors critically revised the manuscript and approved the final version for publication.

Supplementary material

Supplementary material is available at American Journal of Respiratory and Critical Care Medicine online.

Conflicts of interest

Please see the ICMJE disclosure forms, which have been provided as supplementary material. F.J.M., W.P.D., T.M.M., P.J.S., and A.J.P. are editorial board members of American Journal of Respiratory and Critical Care Medicine.

Funding

Y.H.K. is a National Health and Medical Research Council Emerging Leadership Fellow (ID: 2008255). B.K. reports support from a US Department of Veterans Affairs HSR Award 23-137 (1IK2HX003866-01A2) and the Oscar Auerbach Scholars Program.

Artificial intelligence disclaimer

No artificial intelligence tools were used in writing this manuscript.