Clinical Course of Interstitial Lung Disease in Patients With Rheumatoid Arthritis
Editorial support was provided by Elizabeth Ng and Wendy Morris of FleishmanHillard, London, UK, which was contracted and funded by Boehringer Ingelheim Pharmaceuticals, Inc. Boehringer Ingelheim was given the opportunity to review the article for medical and scientific accuracy as well as intellectual property considerations. Boehringer Ingelheim had no role in the writing of the manuscript or the decision to submit the manuscript for publication. Publication of this article was not contingent upon approval by Boehringer Ingelheim.
1Eric L. Matteson, MD: Mayo Clinic College of Medicine and Science, Rochester, Minnesota; 2Elisabeth Bendstrup, MD: Aarhus University Hospital and Aarhus University, Aarhus, Denmark; 3Mary E. Strek, MD: University of Chicago, Chicago, Illinois; 4Philippe Dieudé, MD, PhD: Assistance Publique-Hôpitaux de Paris Cité, Bichat-Claude Bernard University Hospital, INSERM UMR1152, University of Paris, Paris, France.
Author disclosures are available at https://onlinelibrary.wiley.com/doi/10.1002/acr2.11736.
Abstract
Interstitial lung disease (ILD) is a frequent manifestation of rheumatoid arthritis (RA) that is associated with high mortality. RA-ILD may initially be asymptomatic, and lung function may be markedly impaired by the time it is diagnosed. The course of RA-ILD is highly variable, with some patients experiencing no discernable progression or a slow decline, whereas others experience more rapid deterioration. Some patients develop progressive pulmonary fibrosis, which is associated with high mortality. Although risk factors for the progression of RA-ILD have been identified, including older age, worse lung function, and a usual interstitial pneumonia pattern on high-resolution computed tomography, it is not possible to predict the course of RA-ILD in an individual patient. The association between RA disease activity and progression of RA-ILD remains unclear. Regular monitoring is important to enable the prompt identification of progression and early intervention to preserve lung function. The management of RA-ILD requires a multidisciplinary and individualized approach, taking account of the severity and progression of articular and lung disease, risk factors for the progression of RA-ILD, and the patient's preferences, and may include immunosuppression, antifibrotic therapy, and supportive care.
Introduction
Interstitial lung disease (ILD) is a frequent manifestation of rheumatoid arthritis (RA).1-7 The reported prevalence of ILD in patients with RA varies widely depending on the population studied and the definition used, which may only require evidence of ILD on high-resolution computed tomography (HRCT) and not be restricted to clinically significant disease. For example, among 137 patients newly diagnosed with RA at a Danish rheumatology/pulmonology center, 6% had evidence of ILD on HRCT,8 whereas among 151 patients with a mean RA duration of 7.5 years, 19% had evidence of ILD on HRCT.4 RA-ILD has a variable clinical course but is associated with significant morbidity and mortality. In this narrative review, we discuss the clinical course of RA-ILD and its implications for monitoring and management.
Methods
We performed a PubMed search for English language papers published between January 1, 2000 and October 23, 2023 using the following search terms: (arthritis, rheumatoid) AND (interstitial lung disease OR pulmonary fibrosis) AND (natural history OR disease course OR disease progression OR clinical deterioration OR forced vital capacity OR respiratory function tests OR mortality OR survival OR survival rate). The papers identified were hand-searched for relevance. We identified 115 papers on retrospective observational studies and 22 papers on prospective observational studies that provided data on the clinical course of RA-ILD.
Key findings
Early onset of RA-ILD
ILD may be an early complication of RA or even the first presentation of the disease. Prospectively collected data from 679 patients with RA-ILD in a nationwide Danish registry showed that 15% of patients were diagnosed with ILD in the year before the RA diagnosis and 19% in the year after the RA diagnosis.2 In a retrospective analysis of 137 patients with RA-ILD at a center in the United States, the ILD diagnosis predated the RA diagnosis in 10% of patients and was made at the same time as the RA diagnosis in 17% of patients.9
Although patients with RA-ILD often present with respiratory symptoms such as cough and dyspnea, RA-ILD may be asymptomatic. In a French cohort of 163 patients with RA (median duration of 14 years) who did not have a history of ILD or respiratory symptoms, 19% had evidence of ILD on HRCT.10 Among 184 patients at a US center with RA (median duration of 8.5 years), 21% had evidence of ILD on HRCT, and there was no difference in shortness of breath scores between patients who did and did not have ILD.11 Lung function may be markedly impaired by the time that RA-ILD is diagnosed. In a retrospective study on the medical records of 167 patients at a US tertiary care center, median forced vital capacity (FVC) at diagnosis of RA-ILD was 71% predicted, and median diffusing capacity for carbon monoxide (DLco) was 55% predicted.12 Among 102 patients newly diagnosed with RA-ILD at a referral center in Denmark, FVC was relatively well-preserved (mean 90% predicted), but DLco was markedly impaired (mean 55% predicted).13 These data illustrate the importance of screening high-risk patients with RA for ILD to enable RA-ILD to be detected promptly and treated before substantial lung function is lost.
RA-ILD may be associated with pulmonary fibrosis. Fibrotic features, such as reticulation, traction bronchiectasis, and honeycombing, which are radiographic features of the usual interstitial pneumonia (UIP) pattern, are commonly seen on HRCT scans from patients with RA-ILD (Figure 1).12-16 Inflammatory features, such as ground glass opacities, are also commonly observed.12-14, 16 In contrast to other autoimmune disease-related ILDs, the most frequently observed pattern on CT in patients with RA-ILD is a UIP pattern. In a pooled analysis of data from 86 studies that included 4,897 patients with RA-ILD, 46% had a UIP pattern on CT, whereas 35% had a nonspecific interstitial pneumonia (NSIP) pattern.17
Progression of RA-ILD
The course of RA-ILD is highly variable, with some patients showing no discernable progression or a slow decline and others experiencing more rapid deterioration.13, 18-21 In a prospective multicenter study in Korea, group-based modeling of data from 140 patients with RA-ILD identified four distinctive trajectories of change in predicted FVC.21 Over three years, 8% of patients were classified as “improving,” 38% as “stable,” 49% as “slowly declining,” and 5% as “rapidly declining.” The mean annual changes in FVC in these groups were +26.8 mL, −23.4 mL, −65.7 mL, and −139.6 mL.21 Different courses of RA-ILD progression are also observed in terms of increases in the extent of ILD on HRCT. In a prospective cohort of 34 patients with RA-ILD, changes in automated quantitative ILD scores showed four types of trajectories over 24 months: improvement, worsening, improvement then worsening (convex-like), and worsening then improvement (concave-like).19
Although it is clear that RA-ILD may be progressive,12, 13, 19-29 the criteria used to define progression vary across studies. In an analysis of medical records of 102 patients with RA-ILD at a referral center, progression defined as death from respiratory cause, relative decline in FVC ≥10%, relative decline in FVC ≥5% to <10% with worsening symptoms or worsening radiographic findings, relative decline in DLco ≥15%, or clinical characteristics of progression was observed in 52% of patients in the two years after RA-ILD diagnosis.13 In an international retrospective analysis of data from 256 patients with RA-ILD, 13% had progression defined as an absolute or relative decline in predicted FVC ≥10% in the two years after RA-ILD diagnosis.24 In a prospective multicenter study of 116 patients with RA-ILD receiving disease-modifying antirheumatic drugs (DMARDs), 20% had progression defined as a decline in predicted FVC ≥10%, or a decline in predicted DLCO ≥15% with radiologic progression, over a follow-up of 60 months (Figure 2).20
The INBUILD trial defined progressive pulmonary fibrosis (PPF) in patients with fibrosing ILDs other than idiopathic pulmonary fibrosis (IPF) as a relative decline in predicted FVC ≥10% or at least two of the following: a relative decline in predicted FVC ≥5% to <10%, worsened respiratory symptoms, and increased extent of fibrosis on HRCT within the previous two years despite management that was deemed appropriate in clinical practice.30 In a retrospective review of the medical records of 168 patients at a single center in Korea, 34.5% fulfilled the INBUILD definition of PPF.25 A retrospective analysis of data from 89 patients with RA-ILD at a Belgian center found that 53.9% had PPF according to the INBUILD trial definition.28 Among 189 patients with fibrotic RA-ILD in the Canadian Registry for Pulmonary Fibrosis, 46% had PPF as defined by the INBUILD trial, had a lung transplant, or died.31
In May 2022, international respiratory societies proposed criteria to define PPF in patients with fibrosing ILDs other than IPF based on worsening of respiratory symptoms, decline in predicted FVC ≥5% or decline in predicted DLCO ≥10%, and/or radiologic evidence of progression in the past year.32 To date, there are no published data on the proportion of patients with RA-ILD who meet these criteria.
RA-ILD is associated with high mortality,1, 2, 5, 6, 28, 33-37 with several studies suggesting a two- to three-fold greater risk of death in patients with RA-ILD compared with patients with RA without ILD.1, 2, 6, 28, 35 Prospectively collected data from a nationwide Danish registry found that five-year mortality was 39% in patients with RA-ILD compared with 18% in a cohort of patients with RA without ILD matched by birth year, sex, and age at RA diagnosis (Figure 3).2 Acute worsenings of RA-ILD, known as acute exacerbations, are associated with very high mortality.38-42 A retrospective study of 149 patients with RA-ILD who visited a Japanese hospital found that 18% had an acute exacerbation and that this was the most frequent cause of death.40 An analysis of medical records from 310 patients with RA-ILD from a center in Korea suggested that the one-, three-, and five-year cumulative incidences of acute exacerbation were 9.2%, 19.8%, and 29.4%, respectively.42 In this study, the 30- and 90-day mortality rates after an acute exacerbation were 12.6% and 29.9%, respectively.42
Predicting progression of RA-ILD
Several studies have shown that older age, lower FVC or DLco, a greater extent of ILD on HRCT, and a UIP pattern on HRCT are associated with a greater risk of death in patients with RA-ILD (Table 1). In a prospective cohort of 227 patients, compared with patients with predicted DLCO ≥80% at baseline, hazard ratios (HRs) for death over a mean follow-up of 4.7 years were 2.1 (95% confidence interval [CI] 0.9–4.8) in patients with predicted DLCO 50% to 80% and 3.8 (95% CI 1.5–9.5) in patients with predicted DLCO 30% to 50%. For patients with predicted FVC ≥80% at baseline, HRs for death were 1.8 (95% CI 1.1–3.0) in patients with predicted FVC50% to 80% and 1.8 (95% CI 0.6–4.9) in patients with predicted FVC 30% to 50% (Figure 4).36 In a retrospective analysis of 266 patients, the HR for death over a median follow-up of 51 months in patients with >30% compared with ≤30% lung involvement was 2.2 (95% CI 1.2–3.9).45 A retrospective analysis of data from patients with RA-ILD at a single US center showed that patients with a definite or possible UIP patten on HRCT (n = 123) had worse survival than those with NSIP on HRCT (n = 35) over a median follow-up of three years (Figure 5).54 In a prospective study of 116 patients with RA-ILD receiving DMARDs, the mean survival after inclusion in the study was 56 months in patients with a UIP pattern on HRCT compared with 70 months in patients with an NSIP pattern.20 As established in other ILDs,58 decline in FVC is a predictor of death in patients with RA-ILD.18, 50 Modeling of data from 137 patients with RA-ILD suggested that the mortality rate was 2.5-fold greater in patients who had a decline in FVC ≥10% during follow-up than in patients who did not.50 Smoking has been identified as a risk factor for progression of RA-ILD.20, 23 Recent data suggest that high exposure to small particulate matter (air pollution) may also increase the risk of hospitalizations and death in patients with RA-ILD.59, 60
Risk factor | Key studies | |||
---|---|---|---|---|
Reference | Single or multicenter | Design | Number of patients with RA-ILD | |
Older age | Song et al18 (2013) | Two centers | Retrospective | 84 (UIP) |
Nurmi et al43 (2017) | Single center | Retrospective | 59 | |
Zamora-Legoff et al44 (2017) | Single center | Retrospective | 181 | |
Fu et al45 (2019) | Single center | Retrospective | 266 | |
Yamakawa et al46 (2019) | Single center | Retrospective | 96 | |
Li et al47 (2020) | Single center | Retrospective | 278 | |
Kim et al48 (2020) | Single center | Retrospective | 170 | |
Oh et al49 (2022) | Single center | Retrospective | 144 | |
Lower FVC | Brooks et al36 (2022) | Multicenter | Prospective | 227 |
Juge et al24 (2021) | Multicenter | Retrospective | 261 | |
Solomon et al50 (2016) | Single center | Retrospective | 137 | |
Song et al18 (2013) | Two centers | Retrospective | 84 (UIP) | |
Lower DLco | Brooks et al36 (2022) | Multicenter | Prospective | 227 |
Juge et al24 (2021) | Multicenter | Retrospective | 261 | |
Zamora-Legoff et al44 (2017) | Single center | Retrospective | 181 | |
Hyldgaard et al13 (2019) | Single center | Retrospective | 102 | |
Nurmi et al86 (2017) | Single center | Retrospective | 59 | |
Greater extent of ILD on HRCT | Fu et al45 (2019) | Single center | Retrospective | 266 |
Ito et al51 (2019) | Single center | Retrospective | 97 | |
Li et al47 (2020) | Single center | Retrospective | 278 | |
Kim et al48 (2020) | Single center | Retrospective | 170 | |
Oh et al49 (2022) | Single center | Retrospective | 144 | |
Chai et al33 (2023) | Single center | Retrospective | 120 | |
UIP pattern on HRCT | Nieto et al52 (2021) | Two centers | Retrospective | 47 |
Nurmi et al53 (2016) | Single center | Retrospective | 59 | |
Solomon et al50 (2016) | Single center | Retrospective | 137 | |
Yunt et al54 (2017) | Single center | Retrospective | 15 | |
Kim et al48 (2020) | Single center | Retrospective | 170 | |
Ekici et al55 (2021) | Single center | Retrospective | 156 | |
High RA disease activity (based on DAS28-ESR or CDAI) | Brooks et al36 (2022) | Multicenter | Prospective | 227 |
Rojas-Serrano et al56 (2022) | Single center | Retrospective | 37 | |
Liu et al57 (2022) | Single center | Retrospective | 201 | |
Chai et al23 (2023) | Single center | Retrospective | 120 |
- * CDAI, clinical disease activity index; DAS28-ESR, Disease Activity Score using 28 joints with erythrocyte sedimentation rate; DLco, diffusing capacity for carbon monoxide; FVC, forced vital capacity; HRCT, high-resolution computed tomography; ILD, interstitial lung disease; RA, rheumatoid arthritis; UIP, usual interstitial pneumonia.
Fewer studies have investigated risk factors for deterioration in lung function, but the risk factors associated with a faster decline in FVC or DLco appear to be the same as those associated with death.12, 18, 21, 53, 57, 60 Male sex has been shown to be associated with a faster lung function decline or death in some studies,51, 57, 61 but other studies have found no association between sex and progression of RA-ILD.13, 44, 46 Few studies have investigated risk factors for acute exacerbation of RA-ILD, but there is some evidence that the risk of acute exacerbation is higher in patients with a UIP pattern on HRCT.38, 62, 63
There is some evidence of an association between RA disease activity and progression of RA-ILD. Higher RA disease activity based on the Disease Activity Score using 28 joints with erythrocyte sedimentation rate (DAS28-ESR) was associated with a greater risk of death in a prospective multicenter study of 227 patients36 and in two retrospective single-center studies.23, 57 Higher disease activity based on the clinical disease activity index score was associated with a greater risk of death in a retrospective single-center study of 37 patients.56 However, in a prospective study of 140 patients, no significant association was found between the trajectory of the DAS28 score and the trajectory of lung function.21 Higher levels of circulating rheumatoid factor have been associated with higher mortality in patients with RA-ILD.27, 64, 65 High anticyclic citrullinated peptide antibody titers have been associated with increased risk of RA-ILD progression.20, 45 Higher levels of circulating Krebs von den Lungen 6 and increases in C-X-X motif cemokine ligand 11/ITAC (interferon inducible T-cell alpha chemoattractant) and matrix metalloproteinase–13 over time have been associated with progressive versus stable RA-ILD.27, 66 Unlike IPF, the MUC5B rs35705950 T risk allele is not associated with progression of RA-ILD.24
Monitoring and management of RA-ILD
Although risk factors for the progression of RA-ILD have been identified, for an individual patient, the course of disease remains unpredictable. All patients with RA-ILD should be monitored, but patients with established risk factors for progression (such as a UIP pattern on HRCT) may warrant more frequent monitoring. In the absence of evidence-based guidelines, experts have recommended that pulmonary function tests be performed every three to six months, with repeat HRCT where clinically indicated.67-69 Prompt detection of PPF in patients with RA-ILD provides the opportunity for early intervention. However, decisions about when to start therapy and which therapies to use remain challenging. Therapeutic decision-making requires a multidisciplinary approach, involving, at minimum, input from a rheumatologist and pulmonologist, and should take account of the severity and progression of both articular and lung disease, risk factors for progression of ILD, and the patient's preferences.
In the randomized double-masked Treatment for Rheumatoid Arthritis and Interstitial Lung Disease 1 (TRAIL-1) trial in patients with RA-ILD and >10% extent of fibrosis on HRCT, treatment with pirfenidone (n = 63) reduced the rate of decline in FVC over 52 weeks by 55% compared with placebo (n = 60) (−66 vs −146 mL), but the trial was underpowered to detect a difference in the primary endpoint (decline in predicted FVC ≥10% or death over 52 weeks).70 An effect of pirfenidone was observed only among patients with a UIP-like pattern on HRCT, in whom the change in FVC over 52 weeks was −43 mL in the pirfenidone group versus −169 mL in the placebo group.70 Pirfenidone is not licensed to treat ILDs other than IPF. In the randomized double-blind INBUILD trial in patients with PPF, treatment with nintedanib (n = 332) reduced the rate of decline in FVC over 52 weeks by 57% compared with placebo (n = 331) (−80.8 vs −187.8 mL/year).30 Results from the subgroup of patients with RA-ILD (n = 89) (−82.6 vs −199.3 mL/year) were consistent with observations in the overall trial population.71 The effect of nintedanib on reducing the rate of decline in FVC was similar in patients who were taking DMARDs and/or glucocorticoids to the patients who were taking the same with RA-ILD.71 Nintedanib has been licensed by the Food and Drug Administration, European Medicines Agency, and other regulatory bodies for the treatment of chronic fibrosing ILDs with a progressive phenotype irrespective of etiology.
Although no data from randomized controlled trials are available, there is some low-quality evidence from retrospective analyses that suggest that immunosuppression may slow the progression of RA-ILD. A retrospective analysis of data from 92 patients with RA-ILD who received azathioprine, mycophenolate mofetil, or rituximab found that FVC and DLco improved slightly in the 12 months after starting treatment compared with declines in FVC and DLco in the previous 12 months.72 However, in another retrospective study, an increased risk of respiratory mortality was observed in patients with RA-ILD taking azathioprine (n = 54) or cyclophosphamide (n = 21) compared with mycophenolate (n = 42).73 A retrospective study of 31 patients with progressive RA-ILD who received rituximab after glucocorticoids and conventional DMARDs showed cessation of lung function decline after rituximab was initiated.74 A prospective study of 57 patients with RA-ILD who received abatacept found that 72% showed stability or improvement in lung function over a median follow-up of 27 months.75 Similarly, in a retrospective study of 190 patients with RA-ILD, about 80% of patients started on abatacept showed stability or improvement in FVC after a median follow-up of 12 months.76
Results of studies on the effects of anti–tumor necrosis factor (anti-TNF) DMARDs in patients with RA-ILD are conflicting and should be interpreted with caution because these studies did not compare randomized groups. In a prospective analysis of 42 patients with RA-ILD assigned to anti-TNF treatment for one year, no changes in FVC or in the extent of ILD on HRCT were observed, but there were some improvements in small airways disease.77 In a retrospective single-center study, 30% of patients with RA-ILD taking anti-TNF agents (n = 46) experienced progression of ILD on HRCT over one year compared with no patients taking abatacept (n = 16).78 In a prospective study of 70 patients, no association was found between the use of anti-TNF DMARDs and the progression of RA-ILD over two years, whereas the use of non–anti-TNF DMARDs was associated with a reduced risk of progression.79 Prospective data from a registry suggested that patients with RA-ILD taking anti-TNF agents (n = 309) had a two-fold greater risk of five-year mortality than those taking rituximab (n = 43).80
Based on historical data, concerns were raised about a potential link between the use of methotrexate and the development or worsening of RA-ILD. However, recent studies indicate that methotrexate does not increase the risk of developing RA-ILD.81-83 Further, recent studies do not suggest an increased risk of RA-ILD progression in patients treated with methotrexate. In a prospective multicenter study of 143 patients, the time to RA-ILD progression (defined as a decline in predicted FVC ≥10%, decline in predicted DLCO ≥15%, or death due to ILD or pneumonia) was similar between patients taking and not taking methotrexate.84 Other studies have suggested that the use of methotrexate may slow the progression of RA-ILD.85, 86 A retrospective study of 170 patients with RA-ILD suggested that the use of methotrexate was associated with a lower risk of decline in predicted FVC ≥10% (odds ratio 0.27 [95% CI 0.09–0.77]) over a mean follow-up of 4.3 years.86 In a prospective study of patients with RA-ILD in Korea, the risk of rapidly declining predicted FVC or death over one year (observed in 15 patients) was greater in patients who had never used methotrexate than in ever-users (relative risk 8.07 [95% CI 1.82–35.77]).21
The management of RA-ILD should not be limited to pharmacological therapies. Targeting phenotypic traits that are clinically relevant and treatable, including both pulmonary and extrapulmonary manifestations of disease as well as comorbidities, has been proposed as an approach for delivering holistic and patient-centered care to patients with ILDs.87 Patients should be given advice on smoking cessation and reminded of the importance of vaccinations against respiratory infections such as influenza, pneumococcal pneumonia, and COVID-19. Pulmonary rehabilitation is recommended for patients with ILD to improve exercise capacity, symptoms, and quality of life.88 Ambulatory oxygen is recommended for patients with severe exertional hypoxemia.89 Patients with severe and progressive RA-ILD that has not responded to treatment who do not have extrapulmonary contraindications should be considered for lung transplantation.90 Education and psychological support should be offered to all patients.91
Conclusions
ILD is a frequent manifestation of RA that can have a significant impact on morbidity and mortality. The clinical course of RA-ILD is highly variable. Some patients develop PPF, which is associated with poor outcomes. Although risk factors for the progression of RA-ILD have been identified, it is not possible to predict the course of RA-ILD in an individual patient, so regular monitoring, including pulmonary function tests, is important. Prompt identification of PPF in patients with RA-ILD presents the opportunity for intervention to preserve lung function and improve outcomes. Management of RA-ILD requires a multidisciplinary and individualized approach that considers risk factors for progression.
AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. As corresponding author, Dr Matteson confirms that all authors have provided the final approval of the version to be published, and takes responsibility for the affirmations regarding article submission (eg, not under consideration by another journal), the integrity of the data presented, and the statements regarding compliance with institutional review board/Helsinki Declaration requirements.