Screening for Rheumatoid Arthritis–Associated Interstitial Lung Disease: Current Evidence and Next Steps Needed for Routine Clinical Use

Prevalence and clinical features of RA-ILD

The first screening principle is “epidemiology of the disease or condition,” so it is important to establish the significance and prevalence of RA-ILD to justify screening.²⁰ ILD is a serious and common extraarticular manifestation of RA, leading to increased morbidity and mortality.^21,22 Although RA typically occurs more often in female individuals, male sex is a risk factor for RA-ILD; this is known as the sex paradox of RA-ILD.²³ Estimating the exact prevalence of RA-ILD has been challenging, as the case definition of the disease depends on the study population, study design, and the criteria used to define ILD.²⁴ Many patients with RA and lung involvement may not exhibit symptoms despite having some radiographic or physiologic abnormalities, which may affect the estimation of the disease prevalence.²⁵ For example, some patients with RA may not be active enough to experience dyspnea due to joint involvement. They may also attribute dyspnea to other etiologies, such as deconditioning, older age, frailty, obesity, or other comorbidities, leading to underascertainment. Conversely, retrospective review of clinically indicated chest imaging may reveal high prevalence since this is a population enriched for lung disease. One limitation of prospective studies is that mild lung abnormalities may be misclassified as ILD.²⁶ For example, expert radiologists are primed to inspect carefully for ILD and may inadvertently misclassify a subtle radiologic abnormality as ILD. Additionally, participants with pulmonary symptoms may self-select to enroll in prospective studies.²⁷ Physicians may also refer patients who they suspect have ILD to participate in screening studies. Thus, the potential for misclassification of ILD and enrichment of prospective cohorts for participants with pulmonary symptoms may lead to an overestimation of ILD prevalence. Given this heterogeneity, studies have a wide estimate of the prevalence of lung involvement in patients with RA, ranging from 2% to 60%.^9,25,28-31 However, a recent metaanalysis reported the pooled prevalence of clinically apparent ILD as 11%.³⁰ Finally, people with interstitial lung abnormalities (ILAs) may be incorrectly interpreted to have ILD. ILAs may have a different natural history, with some never progressing to clinical ILD.

In 2020, the Fleischner Society suggested that ILA in a predisposing condition such as RA may be called “preclinical ILD.”³² However, updated guidelines published in 2025 by the American Thoracic Society suggest that ILA is defined as nondependent bilateral parenchymal abnormalities found on chest high-resolution computed tomography (HRCT).³³ These findings can include ground glass opacities, reticulations, lung distortion, traction bronchiectasis, or honeycombing.³³ ILD is differentiated from ILA by the presence of symptoms, fibrotic or progressive imaging abnormalities, or presence of major fibrotic ILD patterns, which are further discussed below.³³ For consistency with the literature, we will refer to the presence of ILA in a patient with RA as “subclinical ILD.”³⁴

According to screening principle 2, understanding the “natural history of disease” is necessary for developing a screening strategy.²⁰ Subclinical ILD is present in about 19% to 40% of individuals with RA based on lung abnormalities observed on HRCT imaging.^22,27,35-37 Clinically significant ILD occurs in up to 15% of individuals with RA throughout their disease course, and most patients experience respiratory symptoms within 5 years of RA-ILD diagnosis.^4,23,29 Clinical manifestations of RA-ILD are nonspecific. Upon physical examination, crackles are almost always heard when there is clinically significant disease, and patients often present with chest pain, fatigue, cough, and progressive breathlessness.^29,38,39 Although usually a feature of severe disease, digital clubbing may rarely be present on physical examination.³⁹ Patients who have abnormal pulmonary function tests (PFTs) are more likely to report respiratory symptoms.⁴⁰

There are different subtypes of RA-ILD classified based on histopathological and radiologic patterns. The 2 primarily observed patterns are usual interstitial pneumonia (UIP), which is the most common pattern observed in patients with RA-ILD, and nonspecific interstitial pneumonia (NSIP). Although not common, organizing pneumonia is also observed in about 5% of patients with RA-ILD.³⁸ Other less common subtypes include diffuse alveolar damage, lymphocytic interstitial pneumonia, desquamative interstitial pneumonia, and respiratory bronchiolitis–associated ILD, each with varying prevalence in the literature (Table 2).^21,41 Studies have reported the UIP pattern in 37% to 56% of patients with RA-ILD and NSIP in 33% to 40%.^21,30,38 Identifying these patterns is important as they are associated with varied courses of disease progression and have prognostic implications. RA-ILD associated with UIP pattern has worse survival compared to other subtypes.^38,42 Of note, compared to other systemic autoimmune rheumatic diseases (SARDs), RA has a stronger predilection for UIP.³⁰ The ILD subtype in RA is typically determined by chest HRCT imaging and recognition of typical imaging patterns. Sometimes, bronchoalveolar lavage is pursued to assess for possible infection. Less commonly, lung biopsy is pursued, typically to rule out lung cancer or due to ambiguous clinical scenarios. Transbronchial biopsies are not typically diagnostic, so wedge resection is often needed.⁴³

Outcomes of RA-ILD

RA-ILD is associated with increased morbidity, mortality, and excess healthcare costs. RA-ILD is one of the deadliest outcomes of RA, second only to cardiovascular-related deaths.⁴⁴ In a large US study, 6.6% of deaths in RA were attributed to ILD.²² A metaanalysis of 78 studies of patients with RA-ILD estimated a median survival of 2-14 years after RA-ILD diagnosis, with 9% mortality within 1 year of follow-up, and 49% between 5 and 10 years of follow-up.⁶ Patients with RA-ILD have 2- to 10-fold higher mortality compared to patients with RA without lung disease.^4,22,23 Patients with delayed diagnosis of RA-ILD have even higher mortality, highlighting the potential need for screening and earlier detection.⁴⁵

Mortality is driven by several factors, primarily respiratory-related complications but also encompassing nonrespiratory complications. In current practice, RA-ILD is typically diagnosed by the time pulmonary function has already been compromised. For example, in a retrospective study, 29% of patients had diffusing capacity for carbon monoxide (DLCO) < 40% predicted, and 14% required supplemental oxygen at the time of ILD diagnosis.⁴⁶ More than half of patients with RA-ILD experience disease progression after diagnosis, as measured by PFTs, HRCT imaging, and supplemental oxygen use.^46,47 This decline in respiratory function may ultimately progress to lung transplantation or death.^47,48 Respiratory complications, including acute exacerbations of RA-ILD, drug-induced pneumonitis, pneumonia, pulmonary embolism, and progressive fibrosis and respiratory failure, are major contributors to mortality.^49-52 In particular, acute exacerbations of RA-ILD occur often, leading to increased mortality within 30 days, according to a metaanalysis of 21 studies.⁴⁹

RA-ILD is also associated with significant comorbidities and healthcare costs. The association between RA disease activity and the severity of RA-ILD is currently unknown.^31,53 There is evidence to suggest that the severity of RA disease activity is greater in patients with RA-ILD.⁵⁴ However, there is also evidence suggesting that some measures of disease activity, including joint counts and patient global assessments, are not significantly associated with RA-ILD development.⁵⁵ A separate observation that patients with RA-ILD are slightly less likely than patients with RA to receive DMARDs (84% vs 98%) may highlight practice differences, perhaps due to fears about drug-induced pneumonitis or effects of methotrexate on lung fibrosis.^52,56-59 These observations present an opportunity to better control RA disease activity in patients with RA-ILD. In addition, patients with RA-ILD are at increased risk for serious infections and cancer.^9,60-62 Compared to the general population, patients with RA-ILD are 3-fold more likely to be diagnosed with or die from lung cancer, even in never smokers.⁶² Finally, RA-ILD is associated with increased healthcare costs.⁵ Overall, the multitude of adverse outcomes associated with RA-ILD motivate consideration of screening and early detection strategies.

Quantifying risks and benefits of screening for RA-ILD

Aligned with the “screening program benefits and harms” principle, it is necessary to establish that the benefits of screening outweigh the risks.²⁰ Screening for RA-ILD has the potential to reduce mortality and morbidity by enabling early detection and intervention, but this is not yet proven. A retrospective study estimated that 22% of patients with RA had evidence of ILAs/ILD and 28% experienced radiologic progression.⁶³ Considering the rates of ILAs and ILD progression, early detection that could be enabled by screening may be beneficial to some patients. Although there are some proposed screening strategies, at the moment, there is no agreed-upon consensus on optimal screening protocols for ILD in patients with RA.⁶⁴ Early identification of lung involvement through screening using HRCT or PFTs may allow for closer monitoring and early intervention and the possibility of slowing disease progression and preventing lung damage.^64,65 Thus, there is the potential rationale for the benefits of screening, at least in some subgroups.

Although early detection of RA-ILD may be clinically helpful, it also poses several important risks. A key concern is overdiagnosis, where mild or subclinical ILD may be identified in patients who may never develop progressive disease. For example, in the previously mentioned retrospective study, only 29% of people with RA and ILAs had radiologic progression, whereas 44% of people with RA and ILD progressed.⁶³ It is possible that identification of nonprogressive ILAs/ILD may lead to unnecessary anxiety, follow-up, and treatment.⁶⁶ This can result in unnecessary exposure of patients to medications with significant side effects. It also may cause clinicians to unnecessarily alter the DMARD regimen controlling joint disease due to fears of exacerbating ILD or for infection risk.

To address the “economic evaluation of screening” principle, it is important to assess the economic burden of RA-ILD, including the costs of screening and increased costs of diagnosing.²⁰ The mean annual healthcare cost of RA-ILD in the US is $173,505 per patient.⁵ Although implementing screening programs can be costly, it is possible that screening may ultimately reduce the healthcare costs of RA-ILD by enabling early diagnosis and preventing patients from developing more severe disease. However, screening may paradoxically increase the cost because earlier diagnosis may result in additional follow-up HRCT and PFTs, which can have a significant burden on healthcare costs.⁵ Medications to treat ILD may also be costly, with unclear benefit. Additionally, in order to satisfy the “screening program infrastructure” and “screening program coordination and integration” principles, it may be important to consider that, at this time, not all healthcare systems will have adequate resources to provide ILD screening to patients with RA.^20,67

RA-ILD screening also presents increased risks due to radiation and incidental findings. HRCT is expensive and has radiation, which may pose both financial and health consequences. Further, HRCT can be prone to incidental findings, which lead to repeat testing, additional radiation exposure, referrals to specialists, and lung biopsies, some of which may be unnecessary and costly. Last, there are psychological and quality of life (QOL) implications. Similar to screening for lung cancer, patients with incidental findings upon screening for ILD may experience anxiety due to prolonged diagnostic uncertainty.^68,69 This distress may be unnecessary if the subclinical ILD never progresses to a more clinically severe form. Considering the risks associated with screening, it is important to emphasize the “screening program acceptability and ethics” principle, which highlights the importance of patient autonomy and decision making regarding screening participation.²⁰

Although a formal cost-benefit analysis has not been performed, patients seem to show a preference for screening. As part of the process for the 2023 ACR/CHEST ILD guideline, a patient panel convened and gave qualitative feedback about the pros and cons of screening for ILD.⁷⁰ The participants identified 10 key themes across the following 4 clusters: communication, screening and monitoring, treatment goals, and treatment adverse effects. Notably, patients emphasized the importance of recognizing ILD symptoms, the necessity of screening and close monitoring, the importance of prioritizing survival and QOL, and a willingness to accept treatment risks, provided there is effective communication with healthcare providers (Figure 1). These patient-identified themes significantly influenced the ACR/CHEST voting panel’s discussions, leading to recommendations that align more closely with patient values and preferences. This approach recognizes the importance of incorporating patient engagement in guideline development, ensuring that clinical decisions reflect the outcomes most important to those affected by the disease.⁷⁰

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Weighing the pros (right side of balance scale) and cons (left side of balance scale) of screening for rheumatoid arthritis–associated interstitial lung disease.

Screening for ILD in SSc

Use of lung screening in RA may derive insights from SSc, another autoimmune condition commonly associated with ILD. SSc is a multiorgan autoimmune disease characterized by immune dysfunction, inflammation, fibrosis, and vasculopathy.^71,72 Due to the high prevalence of progressive ILD in patients with SSc, this is one of the few SARDs where it is clinically accepted to screen nearly all patients for ILD.⁷³ Thus, this may provide lessons for what may be needed to justify screening for ILD in RA.

Up to 35% of deaths in SSc are attributed to ILD, making it a leading cause of mortality.⁷⁴ Due to the high frequency and lethality of SSc-associated ILD (SSc-ILD), early screening and detection of lung involvement in patients with SSc is crucial and generally accepted by rheumatologists. This is typically done at diagnosis, and there is now debate about whether or when to rescreen patients with SSc who initially had a negative HRCT.⁷⁵

It is recommended that patients newly diagnosed with SSc should obtain baseline HRCT, PFTs, and assessment of respiratory symptoms.⁷¹ Unlike in RA, the most common imaging pattern observed in SSc-ILD is NSIP.⁷¹ HRCT is considered the gold standard to screen for ILD in SSc.⁷⁶ Whereas PFTs are not sensitive enough to reliably detect early ILD, they are recommended at baseline and annually for the first 5 years.^76,77 In the experience of SSc, high ILD prevalence and associations with progressive ILD and mortality were the most important factors justifying screening. Thus, it may be reasonable to screen for ILD in RA subgroups where the expected prevalence of ILD approaches 50%. Also, identifying patients with RA-ILD who are more likely to progress, such as those with UIP, may be crucial to justify screening.^78,79 The positive predictive value (PPV) of an HRCT diagnosis of UIP is between 90% and 100%.^80,81 The PPV and the reliability of HRCT in SSc screening satisfies the “screening test performance characteristics” principle of screening.^20,77

Current state of screening for ILD in RA

Although it is common to screen for SSc-ILD, the clinical screening practices for RA-ILD are less well defined. Screening for ILD among people with RA who are at higher risk should be prioritized, but diagnosis should be pursued for people with symptoms, regardless of ILD risk factors.⁷³ Currently, subclinical ILD is identified as an incidental finding on imaging performed for other reasons.⁸² Because smoking is highly associated with increased risk for lung cancer, the US Preventive Services Task Force recommends screening for lung cancer with low-dose CT (LDCT) in adults with a 30 pack-year smoking history and currently smoke or have quit within the past 15 years.^83-85 Prior studies also show that cigarette smoking is associated with increased risk for RA, indicating that the population of adults with RA is enriched for smokers.^86-90 Thus, people with RA who are also smokers are eligible to receive screening for lung cancer with LDCT. During the screening process, ILD may be found incidentally, often classified as subclinical.⁸² Our group recently found that, among 15,033 consecutive patients screened with LDCT at our center, those with RA were more likely to have fibrotic lung disease and bronchiectasis than those without RA.⁹¹

In 2023, the ACR and CHEST released the first clinical practice guideline for screening and monitoring of ILD in people with SARDs, including those with RA (Figure 2).⁷³ The guideline recommends consideration of screening for people with RA who have risk factors for the development of ILD, including high-titer rheumatoid factor (RF), high-titer anticyclic citrullinated peptide (anti-CCP), cigarette smoking, older age at RA onset, high RA disease activity, male sex, and higher BMI.^23,92-96 For people with RA at risk for developing ILD, the ACR/CHEST guideline conditionally recommends screening with PFTs—which include spirometry, lung volumes, and DLCO—and chest HRCT imaging.⁹⁷ For this same population, the guideline conditionally recommends against screening with chest radiography, the 6-minute walk test distance (6MWD), ambulatory desaturation testing, bronchoscopy, and surgical lung biopsy.^98,99

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Summary of the 2023 ACR/CHEST guideline for screening and monitoring of ILD in people with RA (adapted with permission from Johnson et al⁷³). 6MWD: 6-minute walk distance; ACR: American College of Rheumatology; anti-CCP: anticyclic citrullinated peptide; CHEST: American College of Chest Physicians; HRCT: high-resolution computed tomography; ILD: interstitial lung disease; PFT: pulmonary function test; RA: rheumatoid arthritis; RA-ILD: RA-associated ILD.

Also in 2023, a panel of expert pulmonologists and rheumatologists proposed RA-ILD screening criteria based on the Delphi methodology.¹⁰⁰ Like the ACR/CHEST guideline, screening was based on the following RA-ILD risk factors: male sex, advanced age and late onset of disease, smoking, and positive RF or anti-CCP antibodies.¹⁰⁰ Though not included in the ACR/CHEST guideline, the Delphi risk factors included moderate or high sustained RA activity according to Disease Activity Score in 28 joints (DAS28) scores, presence of some biomarkers, MUC5B promoter variant, and telomerase genetic variants.¹⁰⁰ There was consensus among experts that it is appropriate to screen patients with respiratory symptoms lasting for > 3 months or in patients who have “Velcro-like” crackles detected on respiratory auscultation, regardless of symptoms.¹⁰⁰ Though there was less consensus, it was also suggested that it may be appropriate to screen patients without respiratory symptoms and with normal respiratory auscultation if enough risk factors are present.¹⁰⁰

In 2024, a panel of experts using the Delphi methodology reported 91% agreement that patients with RA should undergo auscultation regularly.⁶⁴ There was also consensus that symptomatic patients should be screened with PFTs and noncontrast HRCT.⁶⁴ They agreed that the known risk factors for RA-ILD are smoking, high-titer anti-CCP antibodies and RF, male sex, high disease activity, advanced age, and family history of RA-ILD.⁶⁴ For asymptomatic RA patients, the panel suggests a baseline PFT and HRCT and further management only if results indicate RA-ILD.⁶⁴ For patients referred for more frequent RA-ILD screening, the panel recommends annual follow-up PFTs rather than HRCT.⁶⁴

In 2025, the European Respiratory Society (ERS) and European Alliance of Associations for Rheumatology (EULAR) released a clinical practice guideline for connective tissue disease–associated ILD.¹⁰¹ Like the ACR/CHEST guideline, the ERS/EULAR guideline recommends screening for people with RA if they have ILD risk factors, including older age, smoking history, elevated RF or anti-CCP antibodies, male sex, and high articular disease activity.¹⁰¹ Notably, the ERS/EULAR guideline risk factors do not include higher BMI.¹⁰¹ The guideline strongly recommends against HRCT over PFTs and conditionally recommends against HRCT over lung ultrasound for ILD screening in patients with RA.¹⁰¹ Aligned with the third screening principle, the ACR/CHEST guideline, Delphi consensus statements, and the ERS/EULAR guideline have been important for defining the target population for RA-ILD screening as patients with risk factors.²⁰

The primary clinical symptoms resulting from the development of RA-ILD are dyspnea, especially during exertion, and a chronic dry cough.¹⁰² Since these are not specific to RA-ILD, symptoms alone are not diagnostic.¹⁰² Thus, it may be beneficial for clinicians to consider HRCT in people with RA who have certain signs and symptoms of ILD that may allow for early identification of ILD. However, the benefits of early identification include earlier implementation of management strategies and treatment. Once an RA-ILD diagnosis is determined, continued monitoring is conditionally recommended to determine progression and need for treatment.⁷³ The recommendations for monitoring proposed by the ACR/CHEST and ERS/EULAR guidelines allow for screening participants to be identified for further follow-up, diagnostic testing, treatment, or intervention.^73,101 Therefore, these guidelines support the screening principles of “interpretation of screening test results” and “post-screening test options.”²⁰

Biomarkers for RA-ILD

In recent years, many biomarkers associated with RA-ILD have been identified.⁷³ However relatively few have been externally validated, and many may be markers for other types of lung diseases, which limits their use in routine clinical care.⁷³ Most biomarkers are for research purposes and are not yet clinically available; however, they offer the opportunity for mechanistic insights and drug targets. In the future, they may also identify subgroups amenable for screening.

The MUC5B promoter variant is the strongest genetic risk factor for RA-ILD, particularly for the UIP subtype.¹⁰³ This was previously known to be a genetic risk factor for idiopathic pulmonary fibrosis (IPF) and pointed toward shared pathogenesis.¹⁰⁴ A Japanese genome-wide association study (GWAS) discovered the association of RPA3-UMAD1 with RA-ILD.¹⁰⁵ However, this was not confirmed in a European cohort.¹⁰⁶ A recent study applied 12 IPF genetic risk factors to an RA-ILD cohort to develop a genetic risk score that may be useful to risk-stratify patients.¹⁰⁷ Another study examined rare genetic variants related to familial pulmonary fibrosis for RA-ILD risk.¹⁰⁸ A large GWAS for RA-ILD would be helpful to identify drug targets and find additional subgroups for screening. However, genetic testing for ILD risk is not routinely performed in clinic.

Several peripheral blood protein biomarkers have been identified as being potentially involved in RA-ILD progression and possibly effective in monitoring. Some proteins associated with RA-ILD, including matrix metalloproteinases, Krebs von den Lungen 6, surfactant proteins, and inflammatory cytokines, are involved in processes ranging from apoptosis, alveolar function, aberrant immunity, extracellular matrix deposition, and cellular signaling.^79,109-113 These may be useful to detect RA-ILD and monitor for progression and treatment response.

Routine RA-specific autoantibodies may also be associated with ILD. Anticitrullinated protein antibodies are associated with lung disease and predicted incident ILD.^114,115 Elevated RF may be a risk factor for RA-ILD and associated with mortality.^115,116 Antimalondialdehyde-acetaldehyde (anti-MAA) antibodies have been shown to have an association with RA-ILD when compared to RA and chronic obstructive pulmonary disease controls.¹¹⁷ However, anti-MAA may be limited in utility and specificity due to MAA’s potential role in injury of other organs and its association with smoking.¹¹⁸ Other autoantibodies are also being investigated for RA-ILD risk.

There are other factors that may also carry risk of RA-ILD or even play some role in pathogenesis. Telomere shortening, which is known to disrupt highly proliferative processes like hematopoiesis, also affects organs with slower-cycling tissues, such as the lungs, in IPF.¹¹⁹ This effect is unlikely limited to IPF, as a small study found telomere length to be associated with the presence of RA-ILD.¹²⁰ Specific cell types may be associated with RA-ILD pathogenesis and have some potential utility for risk models. In particular, CD14+ monocytes were shown to exhibit upregulated expression of genes associated with the regulation of inflammation and fibrosis in a small analysis of cells isolated from patients with RA-ILD.¹²¹ Transcriptomics may be useful to identify pathways and novel drug targets. Lung tissue and bronchoalveolar lavage may be particularly useful to identify cellular signatures of RA-ILD. Induced sputum or breath biomarkers may be less invasive ways that can be implemented routinely in the clinic. The microbiome is also being pursued for RA-ILD risk and progression.¹²²

Potential screening strategies for RA-ILD

Given the lower prevalence of ILD in RA, screening strategies would likely need to focus on groups at higher risk. Several risk factors for RA-ILD have been identified across retrospective and prospective studies. In a recent metaanalysis, the following risk factors associated with RA-ILD were identified: male sex, older age at RA onset, longer duration of RA, positive autoantibodies, elevated inflammatory markers, and smoking history.³⁰ Moreover, another recent metaanalysis found older age, lung complications, rheumatoid nodules, and leflunomide usage to be risk factors for RA-ILD, in addition to older RA onset age and male sex.¹²³ The effect of specific DMARDs for ILD risk is controversial, but methotrexate was not associated with incident ILD in another recent metaanalysis.⁵⁶

There have been several proposed RA-ILD screening strategies that target patients with risk factors (Table 3).¹²⁴ The ANCHOR-RA study uses a targeted screening approach, enrolling patients with ≥ 2 of the following 6 ILD risk factors: high-titer anti-CCP/RF autoantibodies, presence of extraarticular RA manifestations, male sex, smoking, RA onset after age 60 years, and moderate/high disease activity.¹²⁵ The ESPOIR (Étude et Suivi des Polyarthrites Indifférenciées Récentes) and Study of Inflammatory Arthritis and Interstitial Lung Disease in Early RA (SAIL-RA) studies aimed to identify prevalence and risk factors of RA-ILD among the general RA population and patients with early RA, respectively.^36,125 Both studies identified male sex, older RA onset age, and moderate/high RA disease activity as strong risk factors for RA-ILD.^30,123,125 The ESPOIR study additionally included the MUC5B promoter variant.³⁶ This is in line with the 2022 Spanish Society of Rheumatology and Spanish Society of Pneumology and Thoracic Surgery (SER-SEPAR) recommendation for RA-ILD, which also considers family history and RA-related autoantibodies.¹²⁶ The Four Factor Score is a point-based weighted RA-ILD risk score composed of older age, RF elevation, anti-CCP elevation, and smoking.¹²⁷ The 2023 ACR/CHEST guideline to diagnose RA-ILD recommends PFTs and HRCT to screen for RA-ILD in asymptomatic RA patients with ILD risk factors.^36,128

Several promising screening modalities are also being investigated through the SAIL-RA, ANCHOR-RA, and ESPOIR studies. SAIL-RA uses quantitative CT as an objective measure to identify and quantify parenchymal lung damage through machine learning of the images. The ANCHOR-RA study performs lung ultrasound, a potential modality that could be more accessible and affordable as an initial screen than the current HRCT gold standard.^73,128 Lung magnetic resonance imaging and positron emission tomography scans are promising imaging modalities that may improve screening for ILD but require further research.^129,130 As screening programs begin to be implemented, future work should focus on ensuring quality controls and performance targets are being met, as described by the “screening program quality and performance management” principle.²⁰

Future research

As suggested by the 2023 ACR/CHEST guideline for the screening and monitoring of ILD in people with SARDs and the baseline results of the SAIL-RA study, enhanced screening strategies may allow for early ILD detection and intervention.^73,125 Thus, one of the primary goals for improved screening in patients with RA at risk for developing ILD is the potential for early intervention with antifibrotics. Nintedanib and pirfenidone are both antifibrotic drugs that have been approved by the US Food and Drug Administration for slowing the progression of IPF.^131,132 They are also both conditionally recommended in patients with progression of RA-ILD after first-line treatment according to the 2023 ACR/CHEST guideline for treatment of ILD in people with SARDs.^17,18 The effectiveness of nintedanib and pirfenidone are respectively supported by results of the INBUILD and the Treatment for Rheumatoid Arthritis and Interstitial Lung Disease 1 (TRAIL1) trials, which both showed slower rates of forced vital capacity decline for treatment groups compared to placebo groups.^14,15,133 However, INBUILD included progressive ILDs, and the primary outcome for TRAIL1 was negative, which complicates the interpretation of these results for RA-ILD. A recent phase III trial of nerandomilast showed promising efficacy and potential lower mortality in a basket trial, including in some patients with RA-ILD.¹⁶ A trial establishing that early intervention for patients with subclinical RA-ILD slows the progression of ILD would help justify the need for a screening program (Table 4).

In addition to evaluating if improved screening can result in early and beneficial treatment interventions, future studies should aim to evaluate the risks and benefits of screening. As noted in the 2023 ACR/CHEST guideline for screening, increased screening in people with RA may result in identification of some cases of nonprogressive subclinical ILD.⁷³ If screening is increased, there may be incidental findings that require follow-up testing, which may result in patient anxiety, increased exposure to radiation, and unnecessary healthcare spending.⁶⁷ The goal is to determine whether the benefits of increased screening for RA-ILD outweigh the associated risks, leading to a meaningful improvement in the overall quantity and QOL for people with RA-ILD.

Among the limitations presented in the 2023 ACR/CHEST guideline is the low quality of evidence.⁷³ The recommended frequency of PFTs and HRCT scans in people with RA at risk for developing ILD is based on expert opinions and patient preferences.⁷³ The evidence cited was based on small-scale observational studies, indirect evidence, or expert opinion.⁷³ Thus, there is a need for larger observational studies and trials focused on assessing the frequency with which people with RA should be screened and monitored for ILD. It is currently uncertain and should be investigated whether monitoring should differ based on radiologic patterns, such as UIP, found on chest HRCTs.⁷³ Additionally, visual assessment of the chest HRCTs is time- and resource-intensive and requires readings from expert thoracic radiologists, and even then, there may be ambiguous cases.

Conclusion

Over the last decade, there have been incredible advances in identifying risk factors and describing the natural history of RA-ILD. Given the devastating nature of RA-ILD, screening strategies have been proposed. These seem to perform with reasonable accuracy to identify subgroups at high risk for ILD. Thus, a simple model could currently risk-stratify who should be considered for ILD screening. However, many barriers remain. Most importantly, it is not clear that an early intervention exists to modify the natural history. Whereas early pulmonary referral, closer monitoring, and avoidance of inhalant toxins may partly justify a screening program, an early intervention that clearly improves QOL or mortality would best justify a screening program to practicing clinicians. This intervention may involve antiinflammatories and/or antifibrotics. Clinical trials are urgently needed to settle the question on how these medications may treat ILD. Using a subclinical RA-ILD population could be useful, since a placebo group may be justified since the current standard of care for most is observation. However, a low rate of progression may mean that the trial would have to be large and/or have lengthy follow-up.

Clearly, other research is needed to identify novel biomarkers and imaging techniques that may enhance RA-ILD detection and prognostication. Unlike SSc, if a screening program is implemented in RA, it would likely be in smaller subgroups rather than the entire disease population, given the lower prevalence of ILD. We are tantalizingly close to the implementation of RA-ILD screening in clinic once the barriers of enhanced stratification strategies, quantification of risks and benefits, and early intervention are addressed. Considering that RA is the most prevalent systemic rheumatic disease and ILD is one of the most severe complications, there is an urgent need to fill these research gaps to improve patient outcomes.