Interstitial lung disease in rheumatoid arthritis

INTRODUCTION — Interstitial lung disease (ILD) is the most common manifestation of rheumatoid lung disease [1,2]. However, rheumatoid arthritis-associated ILD (RA-ILD) is not a single pathologic diagnosis, but rather presents with a variety of clinical, radiographic, and pathologic features that impact the natural history, response to treatment, and long-term prognosis.

ILD complicating rheumatoid arthritis (RA) will be reviewed here. Other pulmonary complications associated with RA and their management are discussed separately (table 1 and table 2).

●(See "Overview of pleuropulmonary diseases associated with rheumatoid arthritis".)

●(See "Drug-induced lung disease in rheumatoid arthritis".)

EPIDEMIOLOGY — As RA-ILD is often asymptomatic, the reported frequency depends upon the type of investigation used and the severity of RA in the population studied [3-10]. The following observations illustrate the range of findings:

●In a database study that examined insurance claims from 2003 to 2014, the prevalence of RA-ILD ranged from 3.2 to 6.0 cases per 100,000, and the incidence ranged from 2.7 to 3.8 per 100,000 [11]. The prevalence increased over time, while the incidence remained stable. The median survival was 7.8 years (95% CI 7.1-8.3).

●Clinically significant RA-ILD occurs in nearly 10 to 15 percent of patients with RA [9,12,13] and is generally more common among men with RA than women [9,14].

●Among patients with early rheumatoid disease (joint symptoms <2 years) who were surveyed with a range of investigations including pulmonary function tests (PFTs), chest radiographs, and bronchoalveolar lavage (BAL), 21 of 36 patients (58 percent) had abnormal findings on at least one modality that were consistent with ILD [5]. Among these patients, pulmonary involvement was clinically apparent in 14 percent and clinically silent in 44 percent.

●In another series, the prevalence of changes characteristic of idiopathic pulmonary fibrosis (IPF)on high-resolution computed tomography (HRCT), regardless of pulmonary symptoms, abnormal lung function, or the duration of the rheumatoid disease, was nearly 20 percent [6].

●An autopsy study of 81 patients with longstanding RA noted that 28 (35 percent) had evidence of ILD, and in seven (9 percent), the cause of death was respiratory failure related to ILD [3].

Disease-modifying antirheumatic drugs (DMARDs) and biologic therapies have reduced the extra-articular manifestations of RA, but their impact on RA-ILD is not clear. Case reports, series, and data from registries demonstrate a spectrum of pulmonary effects, including development of new ILD, worsening of preexisting ILD, and resolution of ILD.

In an analysis of data from the British Society for Rheumatology Biologics Register, outcomes of anti-tumor necrosis factor-alpha (TNF-alpha) therapy were compared with DMARDs in patients with known RA-ILD [15]. After adjustment for age, sex, and other potential confounders, the adjusted mortality rate ratio was 0.81 (95% CI 0.38-1.73) for the anti-TNF-alpha cohort compared with the DMARD cohort, suggesting that anti-TNF-alpha agents did not increase mortality. However, RA-ILD was a more common cause of death in the anti-TNF cohort. At this stage, it is not clear what the overall impact of biologic agents is on morbidity or mortality from ILD [15-17].

RISK FACTORS AND GENETIC PREDISPOSITION — Risk factors for RA-ILD include more severe RA, high C-reactive protein, male sex, older age, obesity, cigarette smoking, severe extra-articular disease, and exposure to fine particulate matter [13,18-22]. The principal preventable risk factor for ILD is cigarette smoking [7,23]. One study of 336 patients with RA found that those with a >25 pack-year smoking history were significantly more likely to have radiographic evidence of ILD (odds ratio [OR] 3.76, 95% CI 1.59-8.88) [23].

●Serologic features – Serologic studies, such as a high titer of rheumatoid factor (RF)(eg, ≥90 international units/mL), may identify patients at higher risk of ILD [10,20]. While a higher titer of anti-cyclic citrullinated peptide (anti-CCP) antibodies is also a risk factor for ILD, the presence of these antibodies more strongly predicts RA-associated airway disease [24]. Certain subtypes of anti-CCP antibodies (eg, Hsp90) may be more specific markers for ILD. (See 'Experimental serum markers' below.)

●Cigarette smoking and environmental exposures – Lung injury from cigarette smoking and other stimuli may contribute to the posttranslational modification (citrullination) of proteins, followed by an autoimmune response [25]. Higher levels of a variety of specific anti-citrullinated peptide antibodies (ACPAs) and an expanded repertoire of these antibodies were present in patients with RA-ILD and lung function abnormalities [26]. RA-associated antibodies have also been found in the sputum of patients at risk of RA well before joint disease developed [27,28]. This is an area of active research. (See "Pathogenesis of rheumatoid arthritis".)

●Genetic risk factors – The gain-of-function MUC5B promoter variant rs35705950 is the strongest risk factor for development of idiopathic pulmonary fibrosis (IPF); among patients with RA, this variant is associated with an increased risk for RA-ILD (adjusted OR 3.1; 95% CI 1.8-5.4), a higher lifetime risk of RA-ILD (17 versus 6 percent) and increased risk of RA-ILD development within two years of RA diagnosis (adjusted OR 3.3; 95% CI 2.0-5.6) [29-31]. Additionally, in patients with RA-ILD, the MUC5B variant is associated with features of usual interstitial pneumonia (UIP) on high-resolution CT (HRCT). The exact role of MUC5B protein in the pathogenesis of fibrotic lung disease is not known, but accumulation of MUC5B may disrupt repair mechanisms or interfere with ciliary clearance. This MUC5B variant has not been associated with ILD in systemic sclerosis or inflammatory myositis, underscoring the similarity of RA-ILD to IPF. In IPF, presence of the MUC5B variant has been associated with significantly better survival [32]. By contrast, the MUC5B promoter variant did not affect transplant-free survival in 261 patients with RA-ILD [33]. (See "Pathogenesis of idiopathic pulmonary fibrosis", section on 'MUC5B'.)

Genome-wide association studies have also identified other variants associated with ILD development in patients with RA, such as PPFIBP2 [34].

PATHOLOGIC PATTERNS — A spectrum of lung pathology is seen in RA-ILD, with the lung histopathologic features generally resembling those of idiopathic interstitial pneumonias (table 3). However, the presence of RA as an underlying driver of pulmonary pathology alters the epidemiology, natural history, and treatment of patients with these pathologic patterns compared with those of patients with idiopathic disease. (See 'Treatment' below.)

As practice has evolved in the direction of less frequent lung biopsy, pathologic patterns are now more frequently inferred from radiographic features. (See 'Imaging' below.)

The most common histopathologic patterns seen in RA-ILD are usual interstitial pneumonia (UIP) and nonspecific interstitial pneumonia (NSIP) [18,35-41]. Differences in the pathological picture of the idiopathic and RA-associated UIP picture have been noted, with fewer fibroblastic foci in RA [42]. (See "Idiopathic interstitial pneumonias: Classification and pathology".)

The full spectrum of histopathologic patterns of ILD associated with RA include [35-41,43,44]:

●NSIP (picture 1 and picture 2). (See "Causes, clinical manifestations, evaluation, and diagnosis of nonspecific interstitial pneumonia".)

●UIP (picture 3), which is the pattern associated with idiopathic pulmonary fibrosis (IPF) in patients without underlying connective tissue disease. (See "Idiopathic interstitial pneumonias: Classification and pathology", section on 'Usual interstitial pneumonia'.)

●Organizing pneumonia (picture 4 and picture 5). (See "Cryptogenic organizing pneumonia".)

●Lymphoid interstitial pneumonia (picture 6). (See "Lymphoid interstitial pneumonia".)

●Desquamative interstitial pneumonia (DIP) (picture 7). DIP is most commonly associated with cigarette smoking, but rare cases have been reported in RA in the absence of smoking [39,45,46]. (See "Idiopathic interstitial pneumonias: Classification and pathology", section on 'Desquamative interstitial pneumonia'.)

●Diffuse alveolar damage, which is the pathologic correlate of acute interstitial pneumonia (picture 8 and picture 9). (See "Acute interstitial pneumonia (Hamman-Rich syndrome)".)

●Pleuroparenchymal fibroelastosis (PPFE). PPFE is characterized by bilateral subpleural fibrosis and pleural thickening, predominantly affecting the upper lobes, and is rare in RA [47-49]. The diagnosis is often a clinical one, made on the basis of definite or consistent features on high-resolution CT (HRCT). (See "Idiopathic interstitial pneumonias: Classification and pathology", section on 'Idiopathic pleuroparenchymal fibroelastosis'.)

In some patients, a combination of histopathologic types may be present, in which case a determination of the dominant type may be made to assist in treatment decisions.

CLINICAL FEATURES — In RA-ILD, the onset of symptoms is typically around 50 to 60 years of age [14,35,50]. Males with RA are two to three times more likely to acquire ILD than females. While RA-ILD is often associated with erosive joint disease and postdates the onset of joint symptoms by up to five years, it can occasionally precede joint disease [18]. Subclinical or preclinical interstitial changes may also be seen on incidental chest imaging, and have been associated with male sex, later onset of RA, MUC5B promoter variant rs35705950, and increased disease activity [51,52].

The exact clinical presentation of RA-associated ILD depends in part on the underlying lung pathology. Most often, symptoms develop insidiously and include dyspnea on exertion and a nonproductive cough. Recognition of exertional dyspnea may be delayed due to the exercise limitation associated with joint disease. Less common manifestations include fever and chest pain.

Patients with the pathologic pattern of usual interstitial pneumonia (UIP) typically become symptomatic late in its course when widespread fibrosis is present. By contrast, a fulminant onset has been described in a few cases of rapidly fatal Hamman-Rich type syndrome, which has the pathologic pattern of acute interstitial pneumonia. (See "Idiopathic interstitial pneumonias: Classification and pathology", section on 'Classification' and "Acute interstitial pneumonia (Hamman-Rich syndrome)".)

Physical signs may be absent in early RA-ILD disease. Bibasilar crackles are present in over 75 percent, while signs of pulmonary hypertension and respiratory failure may develop late in the course of disease [35,50]. Clubbing is frequently noted in patients with the UIP pattern of RA-ILD (>75 percent), but is much less common among those with other patterns of RA-ILD [53].

EVALUATION — RA-ILD is usually suspected when a patient with RA develops dyspnea, cough, auscultatory crackles, or abnormalities on pulmonary function tests (PFTs) or chest imaging. The evaluation of suspected RA-ILD typically includes a combination of laboratory testing, PFTs, imaging, and sometimes bronchoalveolar lavage (BAL), but uncommonly a lung biopsy. These tests are designed to characterize the presence, pattern, and severity of ILD, and also to exclude alternative diagnoses. (See "Approach to the adult with interstitial lung disease: Diagnostic testing".)

General approach — Patients with suspected RA-ILD based on clinical evaluation should receive chest imaging for further evaluation. All of the radiographic patterns seen in idiopathic ILD can occur in the context of RA (table 3). For patients with a known diagnosis of RA, if there is a compatible high-resolution CT (HRCT) scan showing fibrotic ILD, further identification of usual interstitial pneumonia (UIP) or non-UIP lung disease is not pursued as the therapeutic approach for UIP and non-UIP fibrotic RA-ILD is the same. In addition, patterns suggesting organizing pneumonia, nonfibrotic nonspecific interstitial pneumonia (NSIP), or lymphoid interstitial pneumonia are typically treated without pathologic confirmation.

In a minority of cases, when the clinical and HRCT features are discordant or atypical, the patient is symptomatic and fit for surgery, and the biopsy would change the therapeutic approach (eg, treatment of an underlying infection or use of glucocorticoids for organizing pneumonia), characterization of the lung disease by lung biopsy may be appropriate [54].

It is important to determine whether the patient is experiencing a first presentation of new RA-ILD, an exacerbation of previously unknown ILD (usually a predominantly fibrotic pattern), or one of these possibilities combined with a superimposed comorbid disease not directly due to RA. Investigations evaluating the various pulmonary manifestations of RA [2] are designed to exclude the possibility that another lung disease or extrapulmonary process is etiologic or coexistent, such as:

●Infection (especially in an immunosuppressed host)

●Drug-induced lung disease [55] (eg, methotrexate, rituximab, anti-tumor necrosis factor-alpha [TNF-alpha]) (see "Methotrexate-induced lung injury" and "Drug-induced lung disease in rheumatoid arthritis")

●Hypersensitivity pneumonitis due to an inhalational exposure (see "Hypersensitivity pneumonitis (extrinsic allergic alveolitis): Clinical manifestations and diagnosis")

●A new or intercurrent ILD, such as acute interstitial pneumonitis or vasculitis, if symptoms are rapidly progressive

●Heart failure, pulmonary embolism, cancer, or recurrent gastroesophageal aspiration

Imaging — In patients with RA, a chest radiograph is often obtained initially to assess complaints of dyspnea or abnormal findings on lung examination. However, patients with ongoing suspicion of RA-ILD due to symptoms, examination findings, or pulmonary function abnormalities require cross-sectional imaging with HRCT to assess for milder disease.

●Chest radiograph – The chest radiograph may be normal in patients with early or mild RA-ILD. When abnormal, potential chest radiograph findings include bibasilar ground-glass opacities, reticular and nodular opacities, patchy opacifications, and honeycombing. Late in the course of the disease, changes suggestive of pulmonary hypertension (eg, enlargement of central pulmonary arteries, attenuation of peripheral vessels) may be detectable. (See "Evaluation of diffuse lung disease by conventional chest radiography" and "Clinical features and diagnosis of pulmonary hypertension of unclear etiology in adults", section on 'Imaging'.)

●High-resolution computed tomography – HRCT should be obtained in patients with symptoms, PFT findings, or chest radiograph abnormalities suggestive of diffuse parenchymal disease. Both prone and supine views are preferably obtained to avoid misinterpretation of gravity-induced opacities in dependent areas. HRCT detects abnormalities earlier than chest radiography and may reveal a range of parenchymal findings [39,56]. In one study of 20 nonsmoking patients with RA and normal chest radiographs, five patients had basilar bronchiectasis and one had mild ILD by HRCT [57]. In a review of 84 patients with longstanding RA, 29 percent of asymptomatic and 69 percent of symptomatic patients had abnormalities on HRCT [58]. These findings included bronchiectasis or bronchiolectasis in the absence of fibrosis (19 percent); ground-glass attenuation (14 percent); nonseptal linear attenuation (18 percent); and honeycombing (10 percent). In general, the HRCT findings accurately predict pathologic patterns. (See "High resolution computed tomography of the lungs".)

HRCT can distinguish a predominantly ground-glass pattern from reticular changes and honeycombing, which is helpful in differentiating among the various types of ILD. As examples:

•Ground-glass opacification is consistent with NSIP, acute interstitial pneumonia, and desquamative interstitial pneumonia.

•Reticular changes, traction bronchiectasis, and honeycombing are more typical of UIP (image 1) [8,59-61]. Infrequently, however, the HRCT may suggest UIP, but NSIP will be identified by biopsy [39].

•Persistent areas of subpleural consolidation are more suggestive of organizing pneumonia [36].

Review of previously performed CT images, including abdominal CTs with views that include the lung bases, may identify a preexisting ILD. In addition, review of older images can help determine the rate of progression of ILD and whether the timing of changes in CT findings over time correlates with symptoms or medication usage.

The optimal method for using CT scans to identify the underlying histopathology and monitor for progression of RA-ILD has not been determined. One method may be to combine scoring systems used in other forms of ILD. In one combined case series, 157 patients with RA-ILD were characterized using: 1) a visual scoring system for systemic sclerosis ILD; 2) the Fleischner Society guidelines for idiopathic pulmonary fibrosis (IPF); and 3) an automated computer-based CT analysis (CALIPER) and followed for approximately three years [62-64]. The combination of the two visual CT-based scoring systems identified a progressive fibrotic subset of the original cohort with poor prognosis similar to that of IPF (c-statistic of 0.71); the automated CT analysis further improved the predictive accuracy. Automated scoring systems are currently a research tool but have promise for the future [65]. (See "Prognosis and monitoring of idiopathic pulmonary fibrosis", section on 'Imaging'.)

Laboratory studies

Routine evaluation — For patients with (or without) RA who present with diffuse lung disease, we generally obtain a complete cell count and differential to look for leukocytosis (infection), leukopenia (immune suppression due to medication), or eosinophilia (possible drug reaction). A serum natriuretic peptide level is useful to evaluate for heart failure or pulmonary hypertension.

Most patients have already had serologic testing for rheumatoid factor (RF) and anti-cyclic citrullinated peptide (anti-CCP) antibodies, but full assessment of other autoantibodies should be performed, including antinuclear antibodies and anti-double-stranded DNA antibodies, and also cryoglobulins to assess for coexistent rheumatic disease that may be contributory in the appropriate clinical setting, such as when purpura, Raynaud phenomenon, skin ulcers, or renal disease are present. RF may be present in high titer in patients with ILD [20]. Anti-citrullinated peptide antibody (ACPA) positivity also correlates with the presence of RA-ILD and higher titers may be associated with more severe ILD [20,26,66,67].

While the sedimentation rate and C-reactive protein correlate with activity of RA joint disease, their role in the evaluation of lung disease is unclear.

Experimental serum markers — No serum markers have demonstrated clinical utility for the diagnosis of RA-associated ILD, although some may be promising.

●Krebs von den Lungen-6 (KL-6) – Increased serum concentrations of KL-6, a glycoprotein found predominantly on type II pneumocytes and alveolar macrophages, have been reported in patients with UIP [68]. As an example, one study assessed the potential role of serum KL-6 for the diagnosis of ILD associated with systemic inflammatory disorders in 57 patients, 22 of whom had known ILD [69]. Patients with ILD had significantly higher KL-6 values than those without lung disease, with the sensitivity and specificity ILD estimated at 61 and 99 percent, respectively, in this selected population. Measurement of serum KL-6 remains a research tool at present but may become clinically useful in the future if the high specificity of the test is confirmed.

●Cytokines and matrix metalloproteinases – Another report noted that serum anti-interleukin-1-alpha antibody titers were significantly higher in patients with RA and ILD, in comparison to patients with RA but not ILD, and to controls. Higher titers were associated with higher serum lactate dehydrogenase concentrations and larger alveolar to arterial oxygen gradients [70]. In addition, a matrix metalloproteinase-7 (MMP-7) and interferon-gamma-inducible protein-10 (IP-10/CXCL10) were found to be elevated in all patients with RA-ILD and a dose-response relationship was seen between the levels of these markers and radiographic severity [71]. A different matrix metalloproteinase and C-X-C motif cytokine pair (MMP-13 and CXCL11) were associated with RA-ILD progression over five years [72].

●Autoantibodies – In a case series (58 patients with RA-ILD; 27 with RA but no ILD), serum antibodies to citrullinated Hsp90 appeared specific (>95 percent), although not sensitive, for RA-ILD [24]. Anti-citrullinated Hsp90 antibodies were not found in 41 patients with mixed connective tissue disease or 33 patients with IPF, further suggesting specificity. The role of these autoantibodies to citrullinated-Hsp90 in identifying patients with RA-associated ILD needs validation in other groups of patients with RA. In a separate study, a stronger association was observed between the number of ACPAs and radiographic UIP than with NSIP [26]. In a meta-analysis, serum ACPA positivity was associated with an increased odds ratio (OR) of ILD (OR 3.39, 95% CI 1.67-6.88) [67]. If confirmed, the serum ACPA would be a very useful test to help predict RA-ILD among patients with RA.

Pulmonary function testing — We obtain spirometry, lung volumes, diffusing capacity and pulse oximetry in all patients with suspected ILD in order to assess the pattern, severity, and progression of respiratory impairment.

Abnormalities associated with ILD include reductions in vital capacity, lung volumes, and diffusing capacity for carbon monoxide (DLCO), oxygen desaturation during exercise, and (in late disease) resting hypoxemia. For example, in a study of 81 patients with recent-onset RA, 33 percent had a DLCO <80 percent of predicted, while only 14 percent had symptoms [5]. When assessing changes over time, clinically important differences include a decrease in forced vital capacity(FVC) of ≥10 percent predicted or a change in DLCO of ≥15 percent [54]. (See "Overview of pulmonary function testing in adults" and "Approach to the adult with interstitial lung disease: Diagnostic testing", section on 'Pulmonary function testing'.)

Among patients with RA, restrictive abnormalities on PFTs are common even in the absence of symptoms and may reflect poor muscle strength or kyphosis due to osteoporosis rather than ILD. The association of restrictive abnormalities and evidence of abnormal gas exchange (eg, reduced DLCO, low pulse oxygen saturation) favor the diagnosis of ILD.

Breathlessness and hypoxemia are usually more evident on exertion than at rest. Thus, measuring oximetry during exertion or a six-minute walk test provides more information regarding abnormal gas exchange than resting values. Arterial blood gases and cardiopulmonary exercise testing are occasionally required to corroborate abnormal pulse oxygen saturation or elucidate the etiology of dyspnea.

Invasive testing, in selected patients — The primary role for invasive testing is to rule out infections or malignancies or to obtain pathologic information that would alter management. Because most patients with RA-ILD are treated with anti-inflammatory therapies and/or antifibrotics based on clinical and radiographic criteria, lung biopsy is uncommon in current practice.

Bronchoalveolar lavage — For patients presenting with interstitial abnormalities after acute onset of fever or respiratory symptoms, BAL is helpful to exclude acute eosinophilic pneumonia, alveolar hemorrhage, malignancy, or opportunistic or atypical infections [54]. BAL is frequently abnormal in patients with RA-ILD, but the findings are nonspecific and not helpful for management. (See "Basic principles and technique of bronchoalveolar lavage" and "Role of bronchoalveolar lavage in diagnosis of interstitial lung disease".)

Abnormalities in cellular constituents and mediators found on BAL are not generally useful for diagnosis or management. The following BAL findings have been reported from research studies:

●In patients with clinical evidence of RA-ILD, total cells, neutrophils, and occasionally eosinophils are elevated [4].

●In the absence of symptoms, lymphocytosis is more common [73,74]. This finding may be associated with a better prognosis, as evidenced by the subclinical nature of the lung disease.

●Macrophages isolated from BAL demonstrated increases in the production of TNF-alpha, superoxide anion, fibronectin, and collagenase activity [75,76].

Lung biopsy — As HRCT patterns have been found to correlate reasonably closely with ILD histopathologic patterns, lung biopsy is rarely required in patients with RA-ILD [61]. We typically consider lung biopsy on a case-by-case basis when clinical and radiographic findings are ambiguous after multidisciplinary discussion and the underlying lung pathology is likely to alter management decisions. A transbronchial biopsy obtained via flexible bronchoscopy is usually inadequate for ILD histopathologic diagnosis, so lung biopsy is typically performed by either video-assisted thoracoscopy or open thoracotomy. The decision about whether a lung biopsy should be performed should take into account the patient’s clinical condition and the potential for complications, including exacerbation of fibrotic ILD. The role of transbronchial cryobiopsy remains unclear and is described separately. (See "Approach to the adult with interstitial lung disease: Diagnostic testing", section on 'Role of lung biopsy' and "Role of lung biopsy in the diagnosis of interstitial lung disease" and "Interpretation of lung biopsy results in interstitial lung disease", section on 'Interstitial pneumonias'.)

DIAGNOSIS — The diagnosis of RA-ILD is generally based on the combination of compatible clinical features; pulmonary function testing (PFT; eg, restrictive changes and a gas transfer abnormality); high-resolution CT (HRCT) findings (eg, reticular, ground-glass, or consolidative changes); and also exclusion of other processes, such as infection, drug-induced pulmonary toxicity, and malignancy. The use of a multidisciplinary team meeting, including pulmonologists, radiologists, and rheumatologists/immunologists, is encouraged [77,78]. (See 'Differential diagnosis' below.)

Determination of the underlying pattern of RA-ILD is based on a typical HRCT pattern or, less commonly, on lung biopsy findings. (See 'Imaging' above and 'Pathologic patterns' above.)

DIFFERENTIAL DIAGNOSIS — In patients with RA, the differential diagnosis of diffuse lung disease includes drug-induced lung toxicity, opportunistic infection, heart failure, recurrent aspiration, malignancy, and other inflammatory causes of ILD. In addition, patients presenting with new respiratory symptoms with evidence of ILD may have an exacerbation of previously unknown ILD. In the latter situation, obtaining old CT images, even if performed for an abdominal problem, may provide clues to preexisting disease.

●Drug-induced lung toxicity – Drug-induced lung toxicity has been associated with most of the medications used to treat RA, including nonsteroidal anti-inflammatory drugs (NSAIDs), methotrexate, leflunomide, gold, penicillamine, and biologic agents (eg, tumor necrosis factor [TNF] inhibitors, rituximab, tocilizumab) [79]. Toxicity has rarely been reported with anakinra and not with abatacept. An essential step in the evaluation of possible drug-induced lung toxicity is to stop any implicated medication(s) and observe for improvement over the next few days to weeks. (See "Drug-induced lung disease in rheumatoid arthritis".)

Development of a sarcoid-like reaction in the lungs has been reported with infliximab, etanercept, and adalimumab and appears to be a class effect of anti-TNF-alpha agents [80-84]. Patients may present with dry cough, night sweats, and weight loss [85]. Onset of disease ranges from 1 to 50 months after initiation of the anti-TNF agent [85]. Sarcoid-like granulomas have also been reported in the skin, lymph nodes, and bone marrow in association with anti-TNF-alpha agents. In one series, the serum angiotensin-converting enzyme level was elevated in 48 percent of patients [86]. Cessation of the anti-TNF-alpha agent generally leads to resolution of the granulomas over several months [87]. Recurrences have been reported when the same anti-TNF-alpha agent is resumed but appears less common when an alternate agent is used [85,87].

●Opportunistic infection – Opportunistic infections are well-known complications of immunosuppressive therapies used to treat RA. The diagnosis of opportunistic infections typically requires special stains and culture of induced sputum and/or bronchoalveolar lavage (BAL) specimens.

Pneumocystis (jirovecii) pneumonia (PCP) is associated with all of the immunosuppressive agents, particularly when the patient is receiving a glucocorticoid dose equivalent to ≥20 mg of prednisone daily for one month or longer in addition to a second immunosuppressive agent, or taking an anti-TNF-alpha agent in combination with other intensive immunosuppression. PCP should be in the differential of new, recent-onset dyspnea, fever, and diffuse of patchy radiographic disease. (See "Epidemiology, clinical manifestations, and diagnosis of Pneumocystis pneumonia in patients without HIV".)

Anti-TNF-alpha agents also increase the risk for new and reactivated latent fungal infections, such as histoplasmosis, coccidioidomycosis, cryptococcosis, and other invasive fungal infections. (See "Tumor necrosis factor-alpha inhibitors: Bacterial, viral, and fungal infections".)

Mycobacterial disease (both tuberculous and nontuberculous) is a well-described complication of anti-TNF-alpha agents. (See "Risk of mycobacterial infection associated with biologic agents and JAK inhibitors".)

●Hypersensitivity pneumonitis – The clinical, imaging, and histopathologic characteristics of chronic hypersensitivity pneumonitis are similar to those of the usual interstitial pneumonia (UIP) pattern of RA-ILD. The radiographic findings typical of subacute hypersensitivity pneumonitis (eg, diffuse micronodules, ground-glass attenuation) are also seen in some patients with RA and organizing pneumonia. (See "Hypersensitivity pneumonitis (extrinsic allergic alveolitis): Clinical manifestations and diagnosis".)

●Other causes – Heart failure is generally excluded based on physical examination, natriuretic peptide measurement, and echocardiogram.

Recurrent aspiration typically affects the lower lobes and may demonstrate tree-in-bud abnormalities or an airway-centric distribution of inflammation; swallowing difficulties provide a clue to the diagnosis, although they are not always present.

The additional differential diagnosis of diffuse lung disease is discussed separately. (See "Approach to the adult with interstitial lung disease: Clinical evaluation" and "Approach to the adult with interstitial lung disease: Diagnostic testing".)

TREATMENT — The optimal treatment for RA-ILD has not been determined. Asymptomatic and mildly affected patients may be monitored clinically and with serial pulmonary function testing (PFT). For patients with more advanced or progressive disease, treatment with immunosuppressive agents may target both the pulmonary pathology and the underlying autoimmune process. For lung disease alone, our contributors prefer steroid-sparing agents (eg, mycophenolate, rituximab) except for those presenting with likely organizing pneumonia, in which case corticosteroids are preferred. However, patients may be managed with a combination of agents (eg, with corticosteroids to control joint disease as well as steroid-sparing agents for their lung disease). Nintedanib has been shown to slow progression in those with progressive fibrosing disease [88] and may be used in addition to other therapy.

General measures — Supportive measures, including evaluation for supplemental oxygen, patient education, pulmonary rehabilitation, smoking cessation, vaccination for the prevention of pulmonary infections, and palliation of pulmonary symptoms, are similar in all patients with interstitial lung disease. (See "Treatment of idiopathic pulmonary fibrosis", section on 'Supportive care'.)

Patients with advanced disease are appropriate for referral for palliative care and/or lung transplantation. (See 'Lung transplantation' below and "Palliative care for adults with nonmalignant chronic lung disease".)

Deciding whom to treat — The decision to treat RA-ILD needs to weigh prognosis, likelihood of response to therapy, and potential benefits of early therapy (ie, before fibrosis is established) against the potentially significant adverse effects of treatment (eg, risk of infection). Abnormalities in any single PFT are common in patients with RA. Thus, the diagnosis of clinically significant disease that warrants close monitoring or treatment is based upon the severity of impairment, rate of progression, and pattern of abnormalities identified by the investigations described above, rather than results of a single test.

As a way to guide treatment and monitoring strategies, a newer approach used in guidelines for idiopathic interstitial pneumonias has been to categorize the disease behavior as self-limited, reversible, stable, progressive, or irreversible, with or without the potential for long-term stabilization with therapy (table 4) [2,41,89]. At present, predictors of poor prognosis (eg, low diffusing capacity or extensive fibrosis on the high-resolution CT [HRCT]), combined with observed rate of progression offer the best guide to treatment decisions in the face of infrequent pathologic confirmation and heterogeneous outcomes.

Patients with mild ILD or stable fibrotic disease — For asymptomatic patients and those with mild RA-ILD (eg, mild dyspnea or cough, vital capacity, lung volumes, and diffusing capacity ≥70 percent predicted), we suggest clinical follow-up and PFT surveillance, rather than anti-inflammatory or antifibrotic therapy. Chest imaging (typically HRCT) should be obtained in the setting of new symptoms or PFT decline.

Similarly, for patients with a usual interstitial pneumonia (UIP) pattern and stable disease by symptoms, PFTs, and HRCT, we suggest monitoring without specific therapy (other than treatment of their articular disease) rather than other treatments. However, close clinical and PFT monitoring are appropriate to allow for early treatment of disease progression. (See 'Antifibrotic therapy, for progressive fibrosis' below.)

Typically, these latter patients are older, and their RA-ILD is less likely to respond to glucocorticoids or immunosuppressive therapies. Therapy of their joint disease continues as indicated, although any drugs that may have contributed to lung toxicity are discontinued. (See "Drug-induced lung disease in rheumatoid arthritis".)

Initial therapy for more severe or progressive disease — For patients with more severe or progressive interstitial lung disease, we suggest initiation of anti-inflammatory therapy, typically with steroid-sparing agents (eg, mycophenolate or rituximab), rather than upfront antifibrotic therapy or expectant management. Features that suggest treatment of RA-ILD with anti-inflammatory therapies is likely to be beneficial include younger age and worsening of symptoms, PFTs, or HRCT over the preceding three to six months (ie, reversible or progressive disease (table 4)). The decision to commence therapy is also influenced by the presence of comorbid diseases that might increase the risk of adverse effects from immunosuppression or toxicities of individual agents. There are only sparse data to guide anti-inflammatory treatment decisions; randomized controlled studies are needed to determine the patient subgroups most likely to benefit from a given regimen.

In one observational and retrospective study of 212 patients with RA-ILD treated with immunosuppressant agents (mycophenolate, azathioprine, or rituximab), initiation of immunosuppression correlated with stabilization in lung function trajectory [90]. Comparing lung function trends before and after immunosuppressant initiation for all three agents, forced vital capacity (FVC) percent predicted improved by 3.9 percent (95% CI 2.0-5.8) and diffusing capacity for carbon monoxide (DLCO) by 4.5 percent (95% CI 2.1-6.9) over 12 months compared with their trajectory prior to treatment. There were no significant differences in outcomes between the different immunosuppressive agents. The drop-out rate due to side effects was slightly higher with azathioprine than mycophenolate or rituximab.

Similarities between RA-ILD with a UIP pattern and idiopathic pulmonary fibrosis (IPF) have led to concern about potential harm from glucocorticoids and other anti-inflammatory therapies in these patients based on data from IPF [91]. However, evidence from observational studies suggests a trial of immunosuppression in suitable patients is warranted, even for those with a UIP pattern [90,92].

●In the observational trial of immunosuppressant therapy described above [90], 38 percent of patients had a UIP pattern on CT, but the presence of possible or probable UIP did not impact positive treatment response significantly at 12 months.

●Similarly, in a retrospective study of 59 patients with RA-associated UIP (RA-UIP) receiving immunosuppressive therapy for ILD progression or severe disease, 50 percent of those treated improved or remained stable over a median 33-month follow-up despite the expectation that most would progress [92]. Furthermore, the treated patients had similar survival compared with untreated patients despite worse initial lung function and disease trajectory.

Our contributors typically avoid glucocorticoids and azathioprine in patients with RA-UIP in the absence of prospective data verifying benefit and evidence of harm in IPF patients with UIP [91]. In patients with RA-ILD and UIP treated with immunosuppressants, it is appropriate to monitor closely for progressive fibrosing disease, at which point we suggest treatment with antifibrotic therapy. (See 'Antifibrotic therapy, for progressive fibrosis' below.)

In contrast, patients with RA-ILD accompanied by radiographic or pathologic evidence of organizing pneumonia are considered likely to respond to glucocorticoid therapy based on experience with cryptogenic organizing pneumonia and on clinical reports [93]. We treat these patients like those with cryptogenic disease. (See "Cryptogenic organizing pneumonia", section on 'Treatment'.)

For other patient groups, our contributors also typically avoid use of glucocorticoids for ILD alone (as opposed to extrapulmonary manifestations) in the absence of rapid deterioration (ie, fulminant disease), or suspicion for an acute exacerbation of fibrotic ILD. As glucocorticoids act more quickly than most other immunosuppressive regimens, they are used routinely in these settings. (See 'Treatment of fulminant disease' below and "Acute exacerbations of idiopathic pulmonary fibrosis", section on 'Treatment'.)

Dosing and administration of individual agents — Data regarding the efficacy of individual agents for RA-ILD are highly limited, so recommendations are based primarily on case series and clinical experience. British and Spanish guidelines favor abatacept and rituximab as they appear to have a lower likelihood of worsening preexisting ILD or development of new-onset ILD compared with other biologic disease-modifying anti-rheumatic drugs (DMARDs) [94,95]. Decisions are made on a case-by-case basis, considering patient experience with other agents, the treatment requirements of extrapulmonary disease, and adverse effects of individual agents in the context of patient comorbidities. Dosing and administration of DMARDs are generally similar to their use in other settings, except where otherwise noted.

●Mycophenolate – When used for ILD, mycophenolate mofetil is initiated at 500 mg given twice a day and is increased as tolerated to a target dose of 2 to 3 g/day. Clinical experience suggests lower daily doses are often not effective for this indication. In addition, mycophenolate is not effective in the treatment of articular manifestations of RA [96], so may need to be added to other DMARDs. The administration of mycophenolate is described separately. (See "Mycophenolate: Overview of use and adverse effects in the treatment of rheumatic diseases".)

Evidence in favor of mycophenolate for RA-ILD comes from a series of 125 patients with connective tissue disease-associated ILD (CTD-ILD), including 18 with RA [97]. Mycophenolate mofetil was associated with modest improvements in FVC and diffusing capacity and reductions in the prednisone dose (mean decrease among RA patients 20 mg). The discontinuation rate for adverse effects (gastrointestinal intolerance, hepatic transaminase elevation, cytopenia, and nonspecific symptoms) was under 10 percent. One study using registry data did not suggest an increased risk for malignancy with mycophenolate treatment for CTD-ILD [98].

●Rituximab – Rituximab is a B lymphocyte-depleting biologic agent effective for RA-ILD and also used in patients whose RA is inadequately controlled by other biologic DMARDs. After preparative immunizations (eg, pneumococcal, seasonal influenza, severe acute respiratory syndrome coronavirus 2 [SARS-CoV-2]), serologic testing (eg, hepatitis B, hepatitis C, human immunodeficiency virus [HIV]), checking immunoglobulin levels (immunoglobulin G [IgG] and immunoglobulin M [IgM]), and screening for tuberculosis, rituximab is administered as two intravenous infusions of 1000 mg, 14 days apart. In general, if low-dose glucocorticoids (eg, 10 mg daily) or a DMARD have been used for extrapulmonary disease, they are continued. Details of the treatment protocols are provided separately. (See "Rituximab: Principles of use and adverse effects in rheumatoid arthritis".)

Evidence in favor of rituximab comes from a retrospective cohort study of 44 patients with moderate to severe RA-ILD, in which addition of rituximab was associated with stabilization or improvement in lung function in 30 patients (68 percent) [99]. However, 14 patients (32 percent) experienced progressive lung disease and, of these, seven died. In two smaller case series, rituximab was associated with stabilization of lung disease or modest improvement in a portion of the patients [100,101], although adverse effects included infusion reactions, heart failure, and possible pneumonia.

●Abatacept – Abatacept is a fusion protein (CTLA-Ig) that inhibits activation of T lymphocytes by blocking interaction between CD28 and CD80. Screening for infection and preparation with vaccination, as for many of these agents, should occur before treatment is commenced. Abatacept is administered by intravenous infusion every four weeks or subcutaneous injection that can be self-administered weekly. In general, patients should remain on either low-dose glucocorticoids (eg, 10 mg) or a DMARD if these have been used for extrapulmonary disease. Additional details of dosing and administration are described separately. (See "Treatment of rheumatoid arthritis in adults resistant to initial biologic DMARD therapy", section on 'Abatacept'.)

Retrospective data suggest a beneficial effect of abatacept in RA-ILD [96]. A multicenter, open-label study in 63 patients with RA-ILD revealed that abatacept was associated with stabilization in two-thirds and improvement in one-fourth of patients; 11 patients (17 percent) discontinued therapy due to adverse effects [102].

Abatacept was also associated with stability or improvement in lung function and HRCT in the majority of 44 patients with RA-ILD followed for six months [103]. In a separate series, abatacept appeared protective against the emergence of ILD in a small number of patients with RA [104]. However, concomitant use of abatacept and methotrexate in patients with RA was associated with an increased risk of ILD in one case series [105]. (See "Treatment of rheumatoid arthritis in adults resistant to initial conventional synthetic (nonbiologic) DMARD therapy", section on 'Methotrexate plus abatacept' and "Treatment of rheumatoid arthritis in adults resistant to initial biologic DMARD therapy", section on 'Abatacept'.)

●Systemic glucocorticoid therapy – If a trial of corticosteroids is used as part of initial therapy (generally in combination with another immunosuppressant), it is typically given at a low dose, such as 0.25 mg/kg ideal body weight per day (calculator 1), as a single morning dose [54]. If a response is going to occur, it is usually seen within one to three months. The prednisone dose should be slowly reduced to a maintenance dose of 10 mg/day once a response occurs, using symptomatic response and PFTs to monitor disease activity. Most patients are transitioned to a separate DMARD (or begun on one upfront) as a steroid-sparing agent to facilitate tapering and minimization of prednisone. Use of systemic glucocorticoids for organizing pneumonia or fulminant disease are covered separately and below. (See "Cryptogenic organizing pneumonia", section on 'Initial approach' and 'Treatment of fulminant disease' below.)

●Agents that are less commonly used – Accumulating evidence suggests that methotrexate does not typically cause worsening of lung disease and is unlikely to be a culprit agent for development of RA-ILD [106]. This makes it acceptable as a maintenance therapy for joint disease in patients with stable RA-ILD, but our authors do not use it for treatment of worsening lung disease. Methotrexate-related acute (hypersensitivity) pneumonitis is exceedingly rare, but methotrexate discontinuation is appropriate in those with appropriate symptoms and radiographic findings without other likely potential antigens. (See "Use of methotrexate in the treatment of rheumatoid arthritis" and "Drug-induced lung disease in rheumatoid arthritis", section on 'Methotrexate'.)

Tumor necrosis factor-alpha (TNF-alpha) inhibitors are less often used in patients with RA-ILD because of multiple, although infrequent, case reports of lung toxicity with these agents [94-96]. Conflicting data have been presented for azathioprine in RA-ILD [96,107]. Hydroxychloroquine was used successfully in combination with mycophenolate in a small case series, but it is almost never used as a single agent for RA-ILD [108].

Treatment of fulminant disease — For the minority of patients who develop rapidly progressive acute ILD or organizing pneumonia as a complication of RA, after excluding infection and drug-induced lung toxicity, we follow treatment regimens for the particular type of ILD (eg, acute interstitial pneumonitis, organizing pneumonia). As these patients typically have impending or actual respiratory failure, we suggest high-dose systemic glucocorticoids (eg, methylprednisolone 500 to 1000 mg per day given intravenously as a pulse or in divided doses for three to five days) [96,109]. An immunosuppressive agent may be added at the same time, such as cyclophosphamide [110], rituximab, or azathioprine, although evidence in favor of this practice is limited. Concomitant empiric antibiotics are prudent pending results of microbiologic studies. (See "Acute interstitial pneumonia (Hamman-Rich syndrome)", section on 'Treatment' and "Cryptogenic organizing pneumonia", section on 'Patients with rapidly progressive disease or respiratory failure'.)

Data are limited regarding the role of cyclophosphamide in fulminant RA-ILD [110]. A systematic review of cyclophosphamide in CTD-ILD found that cyclophosphamide has, at best, a modest benefit in preserving lung function [111]. Cyclophosphamide can be administered as intravenous infusions monthly or taken orally every day. Given the toxicity of cyclophosphamide, use of this drug is generally reserved for more severe or refractory disease, and the duration of treatment is limited to six months. The administration of cyclophosphamide is described separately. (See "General principles of the use of cyclophosphamide in rheumatic diseases".)

Prognosis in fulminant disease is poor, and depending on the patient's age, other comorbidities, and patient wishes, a symptom-based palliative approach may be appropriate. (See "Palliative care for adults with nonmalignant chronic lung disease".)

Disease progression despite initial treatment — For patients with RA-ILD, disease progression is frequent despite optimal medical management. The symptoms of disease progression, such as worsening cough, dyspnea, or hypoxemia, commonly overlap with other conditions such as respiratory infection, acute decompensated heart failure, pulmonary hypertension, pulmonary embolism, or pneumothorax. Additionally, worsening pulmonary symptoms can be triggered by acute exacerbation of fibrotic ILD, drug toxicity, malignancy, anemia, coronary artery disease, and deconditioning, among others.

For patients with worsening pulmonary symptoms, initial evaluation to the cause of worsening often includes PFTs and imaging, such as a chest radiograph or CT, tailored based on the presenting symptoms.

For patients who have suspected worsening of ILD after work-up for alternative causes, therapeutic options differ depending on whether the progression is characterized by worsening fibrosis or other processes more likely to be amenable to anti-inflammatory therapies. In either case, referral for lung transplantation evaluation is reasonable for patients who are otherwise appropriate candidates.

Adjunctive immunosuppression, for worsening nonfibrotic ILD — For patients experiencing clinical worsening of their RA-ILD without evidence of worsening or exacerbating fibrotic ILD, our typical approach is to add or change disease-modifying antirheumatic therapies. For patients on mycophenolate, this typically involves adding rituximab or abatacept; for those on one of these therapies, we would usually add mycophenolate or switch from one to the other. Depending on the pace of clinical worsening, glucocorticoids may also be used as a bridge to therapeutic control with these other agents. Dosing and administration are as noted above. (See 'Dosing and administration of individual agents' above.)

Antifibrotic therapy, for progressive fibrosis — For patients with RA, HRCT features of fibrosing ILD, and clinical evidence of declining lung function despite immunosuppressive therapies, we suggest treatment with the antifibrotic nintedanib to slow disease progression. Based on clinical trials and accumulating evidence of clinical efficacy of nintedanib across studies, patients most likely to benefit include those with reticular abnormalities with traction bronchiectasis (with or without honeycombing) on HRCT and a fall in FVC of 10 percent predicted despite initial treatment. Nintedanib has also been shown to slow decline in lung function in other predominantly fibrotic lung diseases, including IPF and systemic sclerosis-associated ILD. (See "Treatment of idiopathic pulmonary fibrosis", section on 'Nintedanib' and "Treatment and prognosis of interstitial lung disease in systemic sclerosis (scleroderma)", section on 'Nintedanib'.)

Nintedanib was evaluated in a trial (INBUILD) that included 663 patients with various progressive fibrosing ILDs that had worsened despite standard of care and who had at least 10 percent of the lung affected as seen on HRCT [88]. Thirteen percent of patients included had RA-ILD. Evidence of disease progression required for enrollment included any of the three following criteria: relative decline in the FVC of at least 10 percent of the predicted value; a relative decline in the FVC of 5 to 10 percent of predicted and worsening of respiratory symptoms or increased extent of fibrosis on HRCT; or both worsening of respiratory symptoms and an increased extent of fibrosis. After 52 weeks:

●In the overall population, nintedanib slowed the overall adjusted rate of decline in FVC with a between-group difference of 107.0 mL/year (95% CI 65.4-148.5), which is comparable to the benefit seen in IPF. The results of this trial lend support to the concept that fibrotic lung disease reflects a final common pathway that is initiated in a spectrum of ILDs and that this phase of disease may respond to antifibrotic agents that are not disease specific. Additional details about the trial are described separately. (See "Treatment of idiopathic pulmonary fibrosis", section on 'Nintedanib' and "Treatment and prognosis of interstitial lung disease in systemic sclerosis (scleroderma)", section on 'Nintedanib' and "Hypersensitivity pneumonitis (extrinsic allergic alveolitis): Treatment, prognosis, and prevention", section on 'Antifibrotic agents'.)

●In a posthoc subgroup analysis of 89 patients with RA-ILD, the rate of FVC decline was -79 mL/year with nintedanib compared with -197 mL/year with placebo (absolute difference 118 mL/year, 95% CI 5.2-231 mL) [112]. Among a larger subgroup of 170 patients with progressive fibrosing ILD associated with autoimmune diseases, there was a trend towards decreased ILD exacerbation or death that did not reach statistical significance (10 versus 18 events).

The dosing of nintedanib for patients with RA and progressive fibrotic lung disease follows that for IPF and is described separately, as are potential adverse effects. Cost, side effects, and availability may limit use. (See "Treatment of idiopathic pulmonary fibrosis", section on 'Nintedanib'.)

For patients who do not tolerate or otherwise cannot take nintedanib, some of our contributors offer a trial of the alternative antifibrotic pirfenidone (off-label), although data are limited and lack of regulatory approval may impact affordability and availability. (See 'Future directions' below.)

Although patients with progressive lung fibrosis frequently require systemic glucocorticoids, oral immunosuppressants, or biological agents for extrapulmonary disease, there are limited data regarding the coadministration of these agents and antifibrotics. In our practice, we typically add nintedanib onto the patient's current disease-controlling therapies. The INBUILD trial excluded patients on azathioprine, cyclosporine, mycophenolate mofetil, tacrolimus, rituximab, cyclophosphamide, or oral glucocorticoids >20 mg/day. Ongoing trials are evaluating the safety and efficacy of combinations of agents.

Lung transplantation — Progressive RA-ILD despite anti-inflammatory therapy carries a poor prognosis, regardless of underlying pathology. Patients with ILD due to causes other than IPF make up 40 percent of the population transplanted for ILD, and criteria have been established for placing these patients on the transplant list [113]. Among 10 patients with RA-ILD who underwent lung transplantation, survival at one year was comparable to lung transplantation recipients with IPF, 67 and 69 percent, respectively [114]. A modest improvement in quality of life with respect to respiratory symptoms was also noted. Side effects of the therapy for RA (eg, osteoporosis) may be a contraindication; other extrapulmonary disease manifestations may also complicate transplantation. (See "Lung transplantation: General guidelines for recipient selection".)

MONITORING PATIENTS ON RA-ILD THERAPIES — For patients who are being treated with glucocorticoids, disease-modifying anti-rheumatic drugs (DMARDs), or antifibrotic agents, monitoring for an objective response to treatment is generally performed at three-month intervals with clinical assessment and pulmonary function testing (PFT), along with judicious use of high-resolution CT (HRCT) chest imaging based on clinical course.

Monitoring for adverse effects of therapies for RA-ILD is essential. As examples:

●Monitoring for nonpulmonary organ toxicity – Hematologic monitoring is needed with most immunosuppressive agents (eg, monthly initially and then every three months). In addition to hematologic monitoring, liver function tests are obtained monthly at first and then every three months for patients on mycophenolate, antifibrotic therapies, and methotrexate. Additional details about the administration and monitoring of these agents are provided in the table and separately (table 5). (See "Mycophenolate: Overview of use and adverse effects in the treatment of rheumatic diseases", section on 'Mycophenolate dose and administration' and "Treatment of idiopathic pulmonary fibrosis", section on 'Medical therapies'.)

●Drug-induced pulmonary toxicity – Almost all of the DMARDs and biologic therapies have been associated with lung toxicity (table 2), so clinicians should keep this possibility in mind should unexpected worsening of ILD occur during therapy. (See "Drug-induced lung disease in rheumatoid arthritis".)

An important clinical question is whether drugs known to cause lung toxicity should be avoided in patients with underlying lung abnormalities due to concern about potential exacerbation. A systematic review has shown the overall risk of a drug reaction is low (1 percent), although if a reaction occurs, it often has a high mortality [79]. Potentially life-benefiting antirheumatic medications should not necessarily be withheld for what appears to be an uncommon side effect, but such patients do require ongoing monitoring for worsening respiratory symptoms or function. (See "Drug-induced lung disease in rheumatoid arthritis", section on 'Evaluation and diagnosis'.)

●Infection – A variety of serious infections have been described with use of these immunosuppressive therapies. Prophylaxis against Pneumocystis (jirovecii) pneumonia (PCP)may be warranted for some of the above treatment regimens. While the low doses of prednisone and methotrexate typically used in RA do not warrant prophylaxis, the combination of a glucocorticoid dose equivalent to ≥20 mg of prednisone daily for one month or longer and a second immunosuppressive agent, or the combination of an anti-tumor necrosis factor-alpha (TNF-alpha) agent with other intensive immunosuppression, may warrant prophylaxis. (See "Treatment and prevention of Pneumocystis pneumonia in patients without HIV", section on 'Indications' and "Tumor necrosis factor-alpha inhibitors: Bacterial, viral, and fungal infections", section on 'Pneumocystis pneumonia'.)

Vaccination with the influenza vaccine should be provided annually to all patents with RA. Vaccination to prevent coronavirus disease 2019 (COVID-19) is recommended for all individuals 12 years and older. Administration of the polysaccharide pneumococcal vaccine is recommended in all adults with chronic lung disease. In addition, a second dose of the polysaccharide vaccine five years after the first is suggested in patients receiving immunosuppressive therapies. The pneumococcal conjugate vaccine is also suggested in such patients (table 6). (See "Seasonal influenza vaccination in adults" and "COVID-19: Vaccines" and "Pneumococcal vaccination in adults".)

●Malignancy – Patients on long-term therapy with cytotoxic medications are at risk of developing malignancy, particularly skin, cervical, and (with cyclophosphamide) bladder cancer. Thus, patients should be educated about avoidance of the sun and use of sunblock, and women should receive regular mammograms and cervical Papanicolaou smears.

PROGNOSIS — In patients with RA, the presence of RA-ILD is associated with greater mortality compared with the absence of ILD (five-year mortality 36 percent and 18 percent, respectively) [115]. Among patients with RA-ILD, the following factors affect prognosis:

●Demographics – The prognosis of RA-ILD depends on sex and age. Females contribute to the greatest number of deaths given the increased prevalence of RA in females, but males with RA are at increased risk of developing RA-ILD [14,116]. Older age has been significantly associated with increased mortality in several cohorts of patients with RA-ILD [14,92,117].

●Disease subtype – Overall, a radiographic or histologic pattern consistent with usual interstitial pneumonia (UIP) portends a greater likelihood of ILD progression and mortality [118-120]. For example, the effect of presumed histopathologic subtype was assessed in a retrospective review of 144 patients with RA-ILD using high-resolution CT (HRCT) and (in a smaller number of patients) pathology [118]. The poorest survival at five years was in those with diffuse alveolar damage (20 percent) and UIP (37 percent), while a better prognosis was found for organizing pneumonia (60 percent), bronchiectasis (87 percent), bronchiolitis (89 percent), and nonspecific interstitial pneumonia (NSIP; 94 percent).

•However, for some patients with RA-ILD and a UIP pattern, the pulmonary abnormalities do not progress and may remain subclinical. In a retrospective review of 84 patients with RA-associated UIP (RA-UIP) who were monitored for 33 months, respiratory abnormalities remained stable in approximately 50 percent (for a median of 45 months), progressed in 30 percent, and deteriorated rapidly in 17 percent [92].

●Respiratory impairment – Baseline pulmonary function tests (PFTs) and their course over time are correlated with survival. For example, in a cohort of 158 patients with RA-ILD, a lower forced vital capacity (FVC) or a 10 percent decline in FVC was independently associated with an increased risk of death regardless of baseline CT pattern [121]. Similarly, both decreased baseline diffusing capacity for carbon monoxide (DLCO) and a 10 to 15 percent decline in DLCO showed associations with progression of disease and mortality [119,121].

●Acute exacerbations – Acute episodes of pulmonary decompensation are a common feature of several interstitial lung diseases and are associated with a poor prognosis, including in patients with RA-ILD [122-124]. In one study of 310 patients with RA-ILD followed for a median of 48 months, 20 percent of patients experienced an acute exacerbation within three years [123]. Even after accounting for age, sex, smoking status, lung function, exercise capacity, and CT scan pattern, acute exacerbations were associated with decreased survival. Thirty-day and 90-day mortality after an acute exacerbation were 13 and 30 percent, respectively.

FUTURE DIRECTIONS — Newer therapies for RA (eg, tocilizumab, janus kinase [JAK] inhibitors) and the antifibrotic agent pirfenidone, which is used in idiopathic pulmonary fibrosis (IPF), may also have a role in treating RA-ILD [1,125-130]. Research examining their effects is awaited with interest.

●A small retrospective study of tocilizumab in 28 patients with RA-ILD followed for a mean duration of 30 months showed stabilization, improvement, or decline in forced vital capacity (FVC) and diffusing capacity for carbon monoxide (DLCO) in 56 percent, 20 percent, and 6 percent of participants, respectively [128].

●In one case series of 43 patients with RA-ILD treated with JAK inhibitors, 89 percent of patients demonstrated stability or improvement over a median follow-up of 19 months [130].

●The antifibrotic agent nintedanib has shown benefit in progressive fibrotic ILD as described above (see 'Antifibrotic therapy, for progressive fibrosis' above). Pirfenidone, an alternative antifibrotic, slows disease progression in IPF and is being assessed in other progressive fibrotic ILDs. In an underpowered trial ended early due to the COVID pandemic, patients with RA and fibrotic ILD of any type were randomly assigned to pirfenidone (2403 mg per day) or placebo [129]. Use of pirfenidone did not significantly change the proportion of patients with a ≥10 percent decline in FVC or death (7 [11 percent] of 63 patients in the pirfenidone group versus 9 [15 percent] of 60 patients in the placebo group). However, patients receiving pirfenidone did demonstrate a slower rate of annual FVC decline (-66 mL versus -146 mL). Additional study of pirfenidone treatment in fibrotic RA-ILD is warranted based on these results. (See "Treatment of idiopathic pulmonary fibrosis", section on 'Pirfenidone'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Rheumatoid arthritis" and "Society guideline links: Interstitial lung disease".)

SUMMARY AND RECOMMENDATIONS

●Epidemiology and risk factors – The prevalence of interstitial lung disease (ILD) in rheumatoid arthritis (RA) is somewhere between 10 and 50 percent. Risk factors for RA-associated ILD (RA-ILD) include more severe RA, high C-reactive protein, male sex, older age, obesity, cigarette smoking, exposure to fine particulate matter, anti-cyclic citrullinated peptide (anti-CCP) antibodies, and the presence of the MUC5B promoter variant. (See 'Epidemiology' above and 'Risk factors and genetic predisposition' above.)

●Histopathologic patterns in RA-ILD – The most common histopathologic patterns among patients with RA-ILD are those of usual interstitial pneumonia (UIP) and nonspecific interstitial pneumonia (NSIP). Other pathologic patterns include organizing pneumonitis, lymphoid interstitial pneumonitis, and diffuse alveolar damage (table 3). (See 'Pathologic patterns' above.)

●Clinical features – Symptoms of RA-ILD typically include insidious-onset dyspnea and a nonproductive cough. Fever, chest pain, and acute symptoms are less common. (See 'Clinical features' above and 'Evaluation' above.)

●Imaging – High-resolution computed tomography (HRCT) is essential to determine the radiographic features and extent of disease. (See 'Imaging' above.)

●Pulmonary function tests – Pulmonary function tests (PFTs) should be obtained in all patients with suspected ILD to assess the pattern, severity, and progression of respiratory impairment. Typical findings include a restrictive pattern with abnormal gas exchange, often accompanied by exertional hypoxemia. (See 'Pulmonary function testing' above.)

●Differential diagnosis – Drug-induced lung disease, infection, heart failure, recurrent gastroesophageal aspiration, and malignancy may either mimic ILD or lead to apparent deterioration in stable ILD. (See 'Evaluation' above and 'Differential diagnosis' above.)

●Diagnosis – The diagnosis of RA-ILD is generally based on the combination of clinical presentation, PFTs, HRCT, and, in selected cases, lung biopsy. Lung biopsy is reserved for patients with atypical findings (eg, rapid progression, fever, predominance of ground-glass opacities over reticular) after bronchoscopic sampling has ruled out infection. (See 'Evaluation' above and "Approach to the adult with interstitial lung disease: Clinical evaluation" and "Approach to the adult with interstitial lung disease: Diagnostic testing".)

●Approach to treatment – Clinically significant disease that warrants treatment is determined based upon the severity of impairment, rate of progression, and radiographic or histologic patterns (table 4). (See 'Deciding whom to treat' above.)

•Mild ILD or stable fibrotic disease – For asymptomatic patients and those with mild RA-ILD (eg, mild dyspnea or cough with vital capacity, lung volumes, and diffusing capacity ≥70 percent predicted), we suggest clinical follow-up and PFT surveillance, rather than anti-inflammatory or antifibrotic therapy (Grade 2C). Similarly, for patients with a UIP pattern and stable disease by symptoms, PFTs, and HRCT, we suggest monitoring clinically and with serial PFTs rather than specific treatments for ILD (Grade 2C).

•Initial therapy for more severe or progressive disease – For most patients with more severe or progressive interstitial lung disease, we suggest initiation of anti-inflammatory therapy, typically with steroid-sparing agents (eg, mycophenolate or rituximab), rather than upfront antifibrotic therapy or expectant management (Grade 2C). There are only sparse data to guide anti-inflammatory treatment decisions. (See 'Initial therapy for more severe or progressive disease' above.)

For patients with RA-ILD accompanied by radiographic or pathologic evidence of organizing pneumonia, we initiate treatment with glucocorticoids as in cryptogenic organizing pneumonia. (See "Cryptogenic organizing pneumonia", section on 'Treatment'.)

•Initial therapy for fulminant disease – For patients who present with fulminant ILD, after excluding infection and drug-induced lung toxicity, we suggest initial high-dose systemic glucocorticoids (Grade 2C). Adjunctive agents (eg, cyclophosphamide, rituximab) may be used in addition to glucocorticoids on a case-by-case basis. (See 'Treatment of fulminant disease' above.)

•Worsening nonfibrotic ILD despite initial therapy – For patients on initial therapy experiencing clinical worsening of their RA-ILD without evidence of progressive or exacerbating pulmonary fibrosis, our typical approach is to add or change disease-modifying antirheumatic therapies. (See 'Adjunctive immunosuppression, for worsening nonfibrotic ILD' above.)

•Progressive fibrosing ILD despite initial therapy – For patients with predominant fibrotic changes on HRCT and clinical progression of disease (eg, decline in forced vital capacity [FVC] by ≥10 percent predicted) despite initial immunosuppressant treatment, we suggest the addition of the antifibrotic nintedanib to maintenance agents for extrapulmonary RA rather than either therapy alone (Grade 2C). Cost, side effects, and availability may limit the use of nintedanib. (See 'Antifibrotic therapy, for progressive fibrosis' above.)

●Monitoring and prognosis – Monitoring disease activity includes clinical assessment, HRCT imaging, PFTs, and surveillance for adverse effects of immunosuppressant therapies. Older age, a UIP pattern on histology or chest imaging, decreased pulmonary function, and acute exacerbations of ILD are significant risk factors for disease progression and mortality. (See 'Monitoring patients on RA-ILD therapies' above and 'Prognosis' above.)