Developments and Challenges in Treating Rheumatoid Arthritis-Related Interstitial Lung Disease: From Pathogenesis to Treatment Opportunities

Abstract

Interstitial lung disease (ILD) is the most severe extra-articular manifestation of rheumatoid arthritis (RA), representing one the most frequent causes of death for patients with RA. The treatment of RA-ILD is still debated and challenging for both rheumatologist and pulmonologist. Ideally, it should aim to control the underlying joint disease activity, to prevent ILD, or to reduce the progression of lung damage, in particular fibrotic changes. Disease-modifying antirheumatic drugs (DMARDS) are used in daily practice for the treatment of joint involvement but are not demonstrated to be effective in ILD, although good control of the systemic disease might improve patients’ prognosis. However, immunosuppressants, usually suggested for the treatment of ILD related to autoimmune rheumatic diseases, often have low efficacy in regard to inflammatory joint manifestations of RA. Finally, the awareness of potential pulmonary toxicity related to disease-modifying antirheumatic drugs further complicates this scenario. Therefore, a multidisciplinary discussion, including at least a rheumatologist, pulmonologist, pathologist, and thoracic radiologist is generally requested to decide the best therapeutic strategy for an individual patient. In this paper, we will review the current available options for the treatment of RA-ILD, focusing on their possible use according to the current knowledge on pathogenesis and clinical evolution of RA-ILD.

Key Points

Treatment of rheumatoid arthritis interstitial lung disease should aim to control joint disease activity and to reduce the progression of lung damage.

A multidisciplinary discussion, including at least a rheumatologist, pulmonologist, and thoracic radiologist is required to decide the best therapeutic strategy for a single patient.

Introduction

Rheumatoid arthritis (RA) is a chronic inflammatory disease affecting about 0.4% of the adult population worldwide [1]. The disease is characterized by inflammatory joint involvement and several possible extra-articular manifestations, clinically observed in more than 20% of patients [1–3]. Among them, interstitial lung disease (ILD) is the most severe owing to its morbidity and mortality; in fact, about 10–20% of mortality in RA can be directly addressed to lung involvement [4, 5].

Despite the improvement in the management of RA in the last years, the all-cause mortality risk for patients with RA-ILD still remains 3–10 times higher than in patients with RA without lung involvement [6, 7].

Noteworthy, the underlying causes of death in patients with RA-ILD is directly related to the systemic or pulmonary manifestations of RA in more than 75% of cases, confirming the difficulty in the treatment and management of patients with RA-ILD [5, 6].

In fact, in patients with RA-ILD, therapeutic approach should be guided not only by articular disease activity [8] but also by a comprehensive evaluation of the extent and pattern of ILD on high-resolution computed tomography (HRCT), along with monitoring changes in lung function and symptoms over time [9–12].

Two main radiological and histopathological patterns have been identified in RA-ILD. Usual interstitial pneumonia (UIP) is the most common, followed by nonspecific interstitial pneumonia (NSIP), while organizing pneumonia (OP) and lymphocytic interstitial pneumonia are reported less frequently [2, 13, 14].

The treatment of RA-ILD is still debated and challenging for both rheumatologist and pulmonologist. This challenge is not only owing to the lack of randomized controlled trials involving DMARDs and immunosuppressants, but it is also influenced by factors such as the heterogeneity of RA-ILD presentations, variability in disease progression, and differences in patient comorbidities and responses to therapy. Consequently, management strategies are often individualized and based on limited evidence, expert opinion, and clinical judgment. Ideally, the treatment should aim to control the underlying joint disease activity and to reduce the progression of lung damage, in particular fibrotic changes [9, 10, 15, 16]. To complicate the therapeutic management of these patients, the treatment of joint involvement is not demonstrated to be effective on ILD, although some studies suggest that the control of systemic disease activity might contribute to improve ILD survival [14, 17, 18]. However, immunosuppressants usually proposed for the therapy of ILD related to autoimmune rheumatic diseases (ARDs), such as cyclophosphamide, azathioprine, and mycophenolate mofetil, are not effective on arthritis in many cases. Some evidence has been reported supporting the efficacy of cyclophosphamide and azathioprine on RA, but they are no longer recommended as first-line agents owing to their unfavourable risk–benefit profile compared with other conventional synthetic disease-modifying antirheumatic drugs (DMARDs) [8, 19, 20]. Finally, the awareness of potential pulmonary toxicity related to disease-modifying antirheumatic drugs (DMARDs), such as methotrexate, TNF inhibitors (TNFi), or leflunomide, further complicates this scenario [10, 16, 21, 22].

Therefore, the therapeutic choice should originate from a multidisciplinary discussion including at least a rheumatologist, pulmonologist, and a radiologist. The decision on the optimal treatment is mainly based on the severity and progression of ILD, on radiologic patterns, on patients’ characteristics, including age and comorbidities, as well as arthritis features. Pharmacologic treatment should be supported by complementary strategies, including a smoking cessation support program, infection prevention, pulmonary rehabilitation, oxygen supplementation and lung transplant when appropriate [2, 23].

Currently, only Spanish and Italian guidelines [10, 11] have proposed recommendations for treatment of both articular and lung involvement in patients with RA-ILD, while the European Alliance of Associations for Rheumatology (EULAR)–European Respiratory Society (ERS) [12] and the 2023 American College of Rheumatology/American College of Chest Physicians (CHEST) guideline for the management of ILD-related rheumatic diseases [9] suggest recommendations only for the treatment of ILD in patients with ARDs. In particular, this last guideline highlights warnings about the potential safety risks associated with the use of conventional synthetic and biologic DMARDs for the treatment of patients with RA-ILD. However, they do not address the use of these drugs, either alone or in combination with other immunosuppressants or antifibrotics, for treating concurrent joint involvement in RA [9].

In this narrative review, we aimed to describe therapeutic options for RA-ILD according to the current available knowledge on RA-ILD pathogenesis.

Pathogenesis of Rheumatoid Arthritis-Related Interstitial Lung Disease (RA-ILD)

Pathogenesis of RA-ILD is still largely unknown; a combination of genetic, immune, and environmental factors has been suggested to be involved [24]. Unlike idiopathic pulmonary fibrosis (IPF), RA-ILD appears in a context of systemic inflammation and autoimmunity, indicating a plausible role for the immune system in the development of lung remodeling and fibrosis [25]. A better understanding of RA-related ILD could facilitate the development of more targeted therapeutic strategies (Table 1).

Table 1.

Possible pathogenetic mechanism of RA-ILD suggesting treatment with biologic and targeted synthetic DMARDS

TNFi	TNF-alpha upregulates TGF-β1 in vitro and in vivo models [35]	Associated to a possible increased risk of acute exacerbation of ILD [9–12]
	TNF-alpha improves bleomycin-induced lung fibrosis [34]
Abatacept	It reduces fibrogenic T-cell proliferation in murine models of ILD [40]	Suggested for the treatment of arthritis. Possible effect on ILD, also in fibrotic pattern [9–12]
	It reduces activation of M2 macrophages in the lungs [41]
	It reduces the iBALT activity and proliferation [20]
Rituximab	It reduces the iBALT activity and proliferation [20]	Suggested for the treatment of arthritis. Suggested for the treatment of inflammatory patterns of ILD [9–12]
	B cell depletion inhibited fibrosis in mice models [44]
IL-6 inhibitors	IL-6 induces fibroblast to myofibroblast transition [73]	Alternative to rituximab and abatacept for the treatment of arthritis [9–12]
	IL-6 is increased in lung samples from RA-ILD [72]
	IL-6 induces epithelial-mesenchymal transition of alveolar epithelial cells [63]
	IL-6 induces JAK-STAT pathways [61]
JAK inhibitors	JAK1 is overexpressed in inflammatory and epithelial lung cells [58]	Alternative to rituximab and abatacept for the treatment of arthritis. Some data suggest their role in the treatment of ILD [9–12]
	JAK2 increases IL-17-induced fibroblast response [60]
	STAT3/JAK2 increases TGF-β1 and IL-6/IL-13-induced FMT [59]
	JAK/STAT pathway induces the epithelial to mesenchymal transition [63]

RA-ILD rheumatoid arthritis-related interstitial lung disease, DMARDs disease modifying antirheumatic drugs, TNF tumour necrosis factor, TGF Transforming growth factor, IL interleukin, iBALT inducible bronchus associated lymphoid tissue, JAK Janus kinases, FMT fibroblast to myofibroblast transition

The current model of fibrosing ILD is characterized, in genetically susceptible individuals, by repeated alveolar and airway epithelium damage that can lead to the chronic activation of profibrotic pathways, myofibroblast formation, stem cell exhaustion, and cell senescence [24, 26]. Compared with IPF, more CD4⁺ lymphocytes, B lymphocytes, and dendritic cells have been detected at higher degree in RA-UIP. Moreover, inducible bronchus-associated lymphoid tissue has been found overexpressed in the lung of patients with RA-ILD compared with IPF [27].

A still controversial hypothesis of RA-ILD pathogenesis suggests that a NSIP-like inflammatory pattern could precede and eventually evolve to UIP-like fibrosis [28]. In this model, inflammatory cytokines increase the expression of adhesion molecules on pulmonary vascular endothelial cells, leading to the recruitment in the lung of circulating leukocytes, and follicle formation similar to cellular NSIP [28]. At this point, a second hit may occur, triggering local antigen presentation by macrophages and dendritic cells, inducing the activation of adaptive immune response, tissue injury, and release of profibrotic factors, particularly transforming growth factor (TGF)-β, which in turn activates fibroblasts [28, 29].

The second hit can be represented by external environmental factors (e.g., smoking) and genetic/epigenetic factors that lead to neoantigen production, cellular senescence, and constitutive production of profibrotic mediators. Aberrantly and chronically activated fibroblasts produce extracellular matrix proteins in a deregulated way, leading to deposition of collagen and development of fibrosis in the lung, until irreversible fibrotic lung disease develops, such as UIP [28, 29].

A cell type with a key role in fibrosing ILD is the macrophage. Following lung injury, macrophages shift toward a proinflammatory phenotype that induces differentiation and activation of myofibroblasts. Classically, in IPF, activated M1 macrophages are thought to have antifibrotic properties. Instead, alternatively activated M2 macrophages are thought to have profibrotic and anti-inflammatory properties, being involved in angiogenesis, tissue remodeling, and wound healing [30, 31]. Limited information is still available on macrophages in RA-ILD [32].

Finally, many studies are investigating the role of genetics in risk of developing RA-ILD. The functional MUC5B rs35705950 promoter variant, the major genetic risk factor for IPF [33], has been demonstrated as a risk factor for RA-UIP, while it is not associated with ILD other than UIP [34]. However, the presence of rs35705950 promoter variant did not influence the survival or decline in lung function in patients with RA-ILD in a large multicenter study [35].

Pathogenic variants of telomere-related genes have been associated to an increased penetrance of pulmonary fibrosis, in particular in familial pulmonary fibrosis [35]. An excess of rare deleterious variants of telomere-related genes (TERT, RTEL1, and PARN) has been observed in patients with RA-ILD compared with healthy controls [36].

Moreover, patients with RA-ILD show shorter telomeres than patients with RA without lung involvement, and this has been associated with greater baseline lung disease severity and accelerated decline in lung function over 12 months. Nevertheless, a recent study was not able to demonstrate an association between telomere-related genes variants and an increased RA risk [37]. In Fig. 1, we reported the proposed model of RA-ILD and the possible mechanism of action of the current available treatments.

Fig. 1.

Suggested mechanisms of action of DMARDs and antifibrotic drugs according to the pathogenetic model of rheumatoid arthritis-related interstitial lung disease. ACPA anticyclic citrullinated peptide antibodies, APC antigen-presenting cell, IL interleukin, NETs neutrophil extracellular traps, TNF tumour necrosis factor

Glucocorticoids

Glucocorticoid (GC) therapy remains the cornerstone of the treatment strategy in many cases, in particular in the presence of a suspected inflammatory ILD pattern on chest HRCT, namely NSIP and OP [9, 10, 16]. GCs have a symptomatic effect in many cases, also in advanced cases of lung disease, but their role in long-term treatment of ARD-related ILD is still debated. According to the last recommendations for the treatment of RA, international and national guidelines suggest limiting the use of GCs to the lowest dose and for the shortest possible time [8–10, 38], while the use of high-dose intravenous GCs are nowadays limited to the treatment of acute exacerbation and rapidly progressive ILD. However, treatment duration and the time and modality of tapering are unknown and change according to physician’s experience [9, 12, 39].

Very recently, a study from Denmark analyzed the lung function trajectories in 101 patients with RA-ILD treated or not with GCs. Although the study did not allow a definitive conclusion, the authors observed that GCs were unable to modify the decline in lung function in patients with RA-ILD, regardless of radiologic pattern of ILD [40].

Methotrexate and RA-ILD

Methotrexate (MTX) remains the first-line treatment for RA and the anchor drug for many combination therapies [8]. For many years, MTX has been suggested to be potentially responsible for the occurrence of ILD in patients with RA, for its progression, and for the occurrence of toxic pneumonia [16, 21, 41, 42]. However, a meta-analysis evaluating more than 8000 patients with RA treated with MTX has questioned the frequency and the consistency of MTX-induced acute pneumonia [43]. More interestingly, no episodes of MTX-related acute pneumonia have been recorded in controlled trials since 2001 [44].

In the Cardiovascular Inflammation Reduction Trial, the potential efficacy of low-dose oral MTX in preventing cardiovascular events was evaluated in patients with prior ischemic heart disease, diabetes mellitus, or metabolic syndrome, showing a low incidence of pneumonia of only 0.25% [45].

Unlike to what was previously generally accepted, MTX has been associated with improved survival in patients with RA-ILD [18]. Furthermore, two different studies demonstrated that MTX was able both to reduce and delay the development of RA-ILD [46, 47].

Conflicting suggestions arise from the available guidelines. While the Italian, Spanish, and ACR guidelines conditionally recommend continuing MTX for RA-ILD unless pulmonary toxicity is suspected [9–12], the 2023 ACR/CHEST guidelines conditionally recommend against the use of this drug for ILD treatment [9]. Although clinicians should be aware of the risk of MTX-induced pneumonia, the treatment of joint inflammation should be considered a primary outcome, even in patients with RA-ILD [10, 11, 42]. MTX should not be contraindicated in patients with stable ILD or when ILD is occasionally found in a patient with well-controlled RA on stable MTX therapy. The temporal relationship between the introduction of MTX in therapy and the diagnosis or the progression of ILD is a crucial factor for the therapeutic decision of whether to continue or not continue the drug. Therefore, in presence of lung disease progression or other pulmonary complication, the multidisciplinary team should be involved to evaluate the possible causative role of MTX and the opportunity of discontinuing the drug [2, 9–11].

Biologic DMARDs and RA-ILD

Tumor Necrosis Factor Alpha Inhibitors

Tumor necrosis factor alpha inhibitors (TNFi) were the first class of biologic DMARDs approved for the treatment of RA, and they usually represent the first biologic drug after the failure of MTX, in many cases in combination with MTX itself [8, 38, 48]. In experimental models, TNF showed both antifibrotic and profibrotic effects, with possible opposite manifestations in predisposed patients, possibly triggering fibrosis or stabilizing ILD [49, 50]. In transgenic mice, an overexpression of TNF-alpha was associated with the development of interstitial pneumonia similar to IPF. TNF-alpha upregulates the synthesis of TGF-β1 in vitro and in vivo models, resulting in chronic inflammation and lung fibrosis [50]. However, supplementation of TNF-alpha improves the lung function and structure of lung parenchyma in bleomycin-induced lung fibrosis in TNF-alpha-depleted mice [49].

Since the introduction of TNFi for the treatment of RA in the early 2000s, no study, whether randomized, observational, or retrospective, has demonstrated the efficacy of these drugs in treating pulmonary involvement in patients with RA-ILD. Conversely, many case reports and case series reported a high number of acute ILD worsening in patients with RA treated with TNFi. In particular, the BIOGEAS project, developed in the early 2000s for monitoring the safety of biologic drugs, reported more than 100 cases of RA with suspected lung toxicity and acute exacerbation (AE) of ILD related to TNFi therapy [51]. In recent years, a large retrospective study showed no difference in the incidence of new ILD diagnoses according to the biologic DMARD used for the treatment of RA, including TNFi [52]. Another retrospective study from the British Society for Rheumatology Biologics Register showed no difference in mortality and cause of death in patients with RA-ILD treated with TNFi or rituximab [53].

Nevertheless, since the availability of contradictory data, many concerns remain on the use of TNFi in patients with RA-ILD [42]. They should be prescribed cautiously in patients with RA with ILD, regardless of concurrent therapy with MTX [54, 55]. Currently, there is no agreement on discontinuation of TNFi in patients with a new diagnosis of ILD. Therefore, recommendations do not suggest discontinuation in patients with RA with good control of joint symptoms and an incidental diagnosis of ILD. However, withdrawal should be suggested in patients with symptomatic or worsening ILD [9–11, 56].

Abatacept

Abatacept and rituximab are currently thought of as the safest options for the treatment of RA-ILD and, in case of contraindications or inadequate response, IL-6 antagonists or targeted synthetic DMARDs can be considered [9–12, 16, 56].

In a murine model of ILD, abatacept demonstrated the ability to significantly reduce fibrogenic marker levels, T cell proliferation, and M1/M2 macrophage infiltration in the lungs [57, 58]. Concurrently, it improved the fibrosis score on histology and lung density on HRCT [58].

In the last years, many retrospective studies have demonstrated the safety of abatacept on RA-ILD [59–61]. A meta-analysis, including nine studies, showed that abatacept resulted in significantly lower rates of ILD worsening than TNFi, and it was associated with a 90% reduction in the relative risk of ILD deterioration at 24 months of follow-up compared with TNFi and conventional synthetic (cs)DMARDs [62].

Stability or improvement of ILD were globally reported in more than 85% of cases, regardless of the radiologic pattern of ILD. According to this meta-analysis, combination therapy with MTX allowed for the reduction of the glucocorticoid (GC) dose but did not change the rate of progression of ILD [62].

Rituximab

Rituximab is a chimeric monoclonal antibody directed against the B cell surface antigen CD20, and it is indicated for the treatment of RA in combination with MTX [63]. Contrary to abatacept, only a few small studies have investigated the use of rituximab specifically for treating RA-ILD [56, 64, 65]. Therefore, many suggestions have been derived from studies on ILD related to ARDs, including RA-ILD, systemic sclerosis, and inflammatory idiopathic myopathies [16, 66, 67]. The possible role of B cell depletion in the treatment of ILD has been demonstrated in mice with bleomycin-induced systemic sclerosis, where the depletion of B cell inhibited fibrosis development, suggesting a possible role of B cell depletion in other ARDs as well [68–70].

In a recent meta-analysis, including 314 patients from 15 studies, stability or improvement of RA-ILD was reported in a large proportion of patients (mean 88%, 95% confidence interval [CI] 76–96%) [71].

Moreover, a registry-based study enrolling 290 patients from UK demonstrated a 48% reduction in all-cause mortality for patients treated with rituximab compared with patients with RA-ILD receiving TNFi [7]. In 31 patients with RA-ILD with a decline higher than 10% in the last 2 years, rituximab allowed for the reversal of the decline of pulmonary function tests in a large proportion of patients, regardless of radiologic ILD pattern [72]. However, some concerns remain about the risk of side effects of rituximab, mainly infections [73].

Janus Kinase Inhibitors

Janus kinases (JAK) are a group of intracellular tyrosine kinases involved in the transduction of signal induced by many membrane receptors [74]. In lung tissues from murine bleomycin-induced fibrosis models, JAK1 is overexpressed in inflammatory and epithelial cells [75]. Moreover, histological analysis of samples from patients with IPF revealed that JAK2 is primarily found in hyperplastic alveolar epithelial type II cells, fibroblasts, and the intima, as well as in the middle layer of small pulmonary arteries [76, 77]. Among the different JAK/STAT isoforms, it appears that JAK2/STAT3 are predominant, promoting the cellular changes observed in ILDs.

The JAK/STAT pathway can be activated by a large number of profibrotic/proinflammatory cytokines, such as IL-6, IL-11, and IL-13, which are increased in different ILD models [24, 78]. JAK2 inhibits IL-17-induced fibroblast response, which increases cell proliferation, fibroblast to myofibroblast differentiation, and collagen type I and fibronectin production in human lung fibroblast models [24, 77, 79, 80].

Available data on the use of JAK inhibitors (JAKi) in RA-ILD are limited; baricitinib and tofacitinib, the first two drugs approved for this class, currently have the most evidence supporting their use. A recent retrospective study from Spain on 72 patients with RA-ILD treated with baricitinib confirmed the results of two previous Italian studies, including 31 and 43 patients, respectively, treated with different JAKi [81–83]. Globally, about 10–20% of patients showed a deterioration of ILD on HRCT or forced vital capacity (FVC), while lung function remained stable in the majority of them. Finally, an improvement was detected in approximately 10% of cases [81, 84, 85].

A meta-analysis, including 318 patients with RA-ILD from ten studies, confirmed these results, even reporting a comparable beneficial effect of JAKi and abatacept on RA-ILD [85]. Although a large part of the available data refers to baricitinib and tofacitinib, small case series suggest that upadacitinib and filgotinib might provide similar results for both safety and effectiveness in RA-ILD [86–88].

However, some concerns remain regarding the safety of JAKi. The ORAL surveillance study reported an excess of cancer and cardiovascular events in patients with RA treated with tofacitinib compared with TNFi [89]. While waiting for controlled studies evaluating possible differences between the single molecules, a warning has been addressed by the European Medicines Agency (EMA) and Food and Drug Administration (FDA) for all JAKi in elderly patients and smokers, and in patients with increased cardiovascular risk regardless of the coexistence of ILD [8].

This result was recently confirmed by Shih and colleagues in a population of patients with RA-ILD. They retrospectively analyzed mortality data of 1624 patients with RA-ILD treated with JAKi or TNFi, finding that the all-cause mortality risk in the JAKi cohort was 50% higher than in the TNFi cohort [90]. However, mortality increased in patients with a history of cardiovascular issues but not in the subgroup without cardiovascular risk. The sensitivity analyses showed a significantly higher risk of all-cause mortality for the JAKi cohort compared with the TNFi cohort among patients aged 65 years and older [90]. Therefore, caution should be used when prescribing JAKi to patients with RA-ILD who are older than 65 years or have a high cardiovascular risk until more consistent data are available.

Interleukin-6 Inhibitors

Interleukin-6 (IL-6) is increased in lung samples from patients with RA-ILD [80] and in experimental models, and IL-6 inhibitors (IL-6i) might be useful for the treatment of inflammatory alterations preceding fibrosis [91, 92]. Experience from clinical trials on systemic sclerosis seem to be consistent with this hypothesis [93, 94].

While only a case series evaluated the effect of sarilumab on RA-ILD [95], some retrospective studies are available for tocilizumab [96, 97]. For the limited number of available data, IL-6 inhibitors (IL-6i) are usually considered as second-line drug for the treatment of RA-ILD [9–12]. In an Italian study on 28 patients, tocilizumab was effective in about 75% of RA-ILD cases, while in a Japanese longitudinal study, a decrease of Krebs von den Lungen-6 and matrix metalloproteinase-3 was reported during the study period [96, 97]. However, an increased risk of mortality and AE-ILD was observed over time, with a mortality of 32.4% at 3 years [96].

Recently, Frideres and colleagues emulated three trials comparing abatacept, tocilizumab, and tofacitinib with rituximab using the Target Trial Emulation framework, observing no significant differences between the drugs in a composite outcome, including, among others, death and hospitalizations [56].

Immunosuppressants

Immunosuppressive drugs have been proposed as first-line treatment by ACR/CHEST guidelines [9]. Data derive primarily from the experience on connective tissue disease (CTD)-ILD, mainly systemic sclerosis [9, 67, 98]. Although some immunosuppressants, namely cyclophosphamide and azathioprine, have shown a little efficacy in the treatment of joint involvement of RA, it is important to remember that they are no longer recommended as first-line agents owing to their unfavorable risk–benefit profile compared with other conventional synthetic DMARDs [8, 19, 20]. Therefore, immunosuppressants should only be suggested in patients with well-controlled arthritis or in combination with a csDMARD or a biologic DMARD [9–12].

In a real-world retrospective study on 212 patients with RA-ILD, an improved trajectory in forced vital capacity (FVC) and diffusion lung for carbon monoxide (DLCO) was observed in patients with RA-ILD treated with immunosuppressants, including rituximab, when compared with pretreatment lung function behavior [64]. However, in the same cohort, a combination therapy with csDMARDS and/or biologic DMARDs was ongoing in about 70% of patients for the treatment of arthritis, while 67.9% of the entire cohort was also receiving prednisone with a mean dose of more than 10 mg daily. No data were provided regarding the outcome of arthritis [64].

Regardless of the kind of immunomodulant treatment, ILD, bronchiectasis, and airway disease contribute to increasing patients’ infectious risk [24]. In detail, a combination of GCs and biologic and conventional synthetic DMARDs was associated with the highest risk of infection in individuals with RA-ILD [99, 100]. Therefore, vaccination and prevention strategies should be suggested to all patients before starting an immunosuppressive treatment [2, 16].

Antifibrotics

The INBUILD trial demonstrated the efficacy of nintedanib in reducing the decline of FVC in patients with progressive pulmonary fibrosis (PPF) other than IPF [101], including a large number of patients with ARDs. Although the drug was more effective in the group of patients with a UIP-like radiologic pattern, nintedanib was able to reduce the decline in lung function in patients with other fibrotic patterns as well [101]. After the INBUILD study, an increasing number of patients with fibrosing-progressive ARDs have been treated with antifibrotics, including RA [102–108].

Pirfenidone, the other antifibrotic drug currently available, was evaluated in the TRAIL1 trial, a randomized, double-blind, placebo-controlled, phase 2 trial enrolling patients with RA with a fibrotic pattern of ILD, regardless of progression [109]. When about half of the expected patients were enrolled, the study was prematurely concluded because of the coronavirus disease 2019 (COVID-19) pandemic. Although the primary endpoint was not reached, the drug was shown to slow the decline of FVC in the treated group, particularly in patients with a UIP pattern; for these reasons, pirfenidone has been approved by the FDA for the treatment of fibrotic RA-ILD, but it is not currently available in Europe [9, 109].

Since only 89 patients with RA-ILD were enrolled in the INBUILD study, the 2023 ACR/CHEST guideline conditionally recommended nintedanib for the treatment of PPF related to RA, while pirfenidone was suggested only in selected patients with a UIP pattern [9].

Neither nintedanib nor pirfenidone has known immunomodulatory activity; therefore, combination therapy with DMARDs is often required in patients with RA-ILD for a holistic approach to the disease [9, 10, 24]. Since only a small number of patients from the controlled trial were treated in combination with DMARDs [101, 109], some concerns remained about the safety of antifibrotics in patients with comorbidities and in combination with other antirheumatic drugs. In this regard, recent real-world experiences have shown that the retention rate of nintedanib does not change when combined with DMARDs [103, 105]. In two retrospective observational studies on 74 and 65 patients with RA, 1-year retention rate of nintedanib was very similar (78.4 and 76.7%, respectively) and not influenced by a combination therapy with csDMARDS or biologic DMARDs in about 80% of cases [103, 105].

Gastrointestinal side effects were the most frequently observed, and the proportion of patients who discontinued the drug within the first year was below 20%, with most discontinuations occurring during the first 6 months of therapy [105]. Currently, there are no specific studies on pirfenidone. Therefore, it should be proposed only for patients who have not tolerated nintedanib.

Recently, a new antifibrotic drug, nerandomilast, has been investigated in two phase 3, double-blind, randomized trials investigating its efficacy in patients with IPF and PPF [110, 111]. Among 325 patients with ARDs enrolled in the study, 119 had RA-ILD [110]. Interestingly, enrolled patients could continue background treatment with nintedanib and/or immunosuppressants, with the exclusion of cyclophosphamide, tocilizumab, mycophenolate, and rituximab. In a 52-week period, nerandomilast slowed down the progression of pulmonary fibrosis in patients with PPF, with a low rate of adverse events. In contrast to nintedanib and pirfenidone, nerandomilast showed a significant effect on time to death, in particular at a dose of 18 mg twice a day (bid) [110, 111].

As reported above, the role of TGF-β1 in the pathogenesis of human IPF is well established [24]. The increased activity of TGF-β1 not only influences the activated fibroblasts but also affects other cell types that directly contribute to the progression of lung fibrosis [112]. Nerandomilast may have the capacity to reduce the defective signaling pathways by modulating the activity of growth factors and cytokines, thereby inducing a stabilization of lung function in pulmonary fibrosis [113]. The inhibition of phosphodiesterase 4B by this drug potentially induces antifibrotic and immunomodulatory effects, while the supposed role on vascular endothelial cells requires further investigation [113]. Therefore, the mechanism of action of nerandomilast appears of particular interest in the treatment of ILD in patients with ARDs. As previously described, in these cases, an excess of inflammatory cells and tertiary lymphoid organs is also observed in lung tissue from UIP, suggesting a possible dual effect of nerandomilast on both immune-mediated and fibrosing pathways [24, 25]. where both fibrosis and inflammation may be involved in the pathogenesis of the lung disease.

Treatment of Acute Exacerbation

AE-ILD represents the most severe complication of RA-ILD and one of the most frequent causes of death for these patients, together with cancer and infections [4–6, 114]. AE is an acute, clinically significant respiratory deterioration characterized by evidence of new widespread alveolar abnormality, firstly described in IPF, but occurring also in patients with rheumatic diseases with a fibrotic pattern of ILD [114, 115]. Even though there is an absence of specific longitudinal studies, the incidence of AE-ILD in patients with ARDs ranges from 3.19 and 5.77 per 100 patient-years [2, 114, 116].

Short-term mortality of AE-RA-ILD remains high, ranging between 30.0% and 58.3% [114, 116]. In a study from Japan, mortality for AE in patients with RA-ILD was similar to that observed in AE-IPF. In particular, a worse prognosis was reported in RA-ILD with a UIP pattern compared with non-UIP patterns [117].

The correct therapeutic strategy of AE-RA-ILD is still unknown and mainly based on the management of AE-IPF [98]. The main difference between patients with IPF and those with RA is that the latter generally are already on treatment with immunomodulatory therapy when AE occurs; for this reason, some aspects should be addressed before defining the therapeutic strategy. In particular, a causative role for DMARDs should always be excluded, as well as the possibility of an infection [98]. To avoid the risk of lung drug toxicity, the ongoing DMARD is often withdrawn after a diagnosis of AE-ILD. Similarly, broad-spectrum antibiotics are usually administered [98]. However, the more inflammatory pathophysiology in RA and other ARDs suggest the need for immunomodulant treatment [98, 114].

High-dose intravenous GC is generally used, despite the absence of recommendations on the most appropriate type, dose, or duration. A number of treatments, including cyclophosphamide, cyclosporine, and rituximab, combined or not with plasma exchange and intravenous immunoglobulin, have been proposed in patients with AE-ARD-ILD, without any demonstration of their efficacy [98].

In a large, retrospective Japanese study, combination therapy with intravenous cyclophosphamide and GCs did not show any advantage compared with GC therapy alone [118].

Recently, many authors have reported possible efficacy of rituximab for the treatment of AE-ILD different from IPF, but the experience in AE-RA-ILD remains very limited [119, 120].

Unmet Needs and Future Research Priorities

Many aspects of management of RA-ILD remain to be investigated. The natural history of the disease remains largely unknown owing to the lack of reliable biomarkers and shared indications for screening of ILD. Until now, nintedanib can be prescribed only according to the inclusion criteria of the INBUILD trial. Therefore, only patients with progressive and often advanced lung disease can benefit from antifibrotic treatment, while no studies have investigated the possible advantages of upfront therapy. Early identification of progressive ILD disease, by biomarkers, radiomics, or software based on machine learning might improve the therapeutic strategy, suggesting more stringent follow-up and more aggressive treatment.

Moreover, no alternative antifibrotic drugs are available in Europe for patients with RA-ILD, while the Food and Drug Administration has approved the use of pirfenidone in North America, even if only for patients with a UIP pattern. Nonetheless, no large real-world studies are available to confirm the effectiveness of this latter drug in clinical practice.

Finally, there is no conclusive evidence that the control of systemic inflammation with DMARDs or immunosuppressants could reduce the progression of ILD or that combination therapy with DMARDs and nintedanib might be more effective than nintedanib alone.

Other than new antifibrotics, other therapeutic strategies are currently under investigation. Among them, chimeric antigen receptor (CAR)-T cell therapy has represented a breakthrough in hematological cancer treatment [121]. Many studies are ongoing to evaluate the usefulness of CAR-T cell therapy in the treatment of ARDs, including severe RA and conditions such as systemic sclerosis and idiopathic inflammatory myopathies, which are characterized by severe ILD [122, 123]. If exploratory studies confirm preliminary data on the efficacy of T-CAR cell therapy for ARD-ILD, it may, in the future, represent a potential treatment option for severe and rapidly progressive RA-ILD.

Conclusions

Treatment of RA-ILD remains largely based on expert opinion, while evidence-based strategies are scarce or absent. The therapeutic approach is the result of a multidisciplinary discussion, involving at least rheumatology and pulmonology, that should take into account both the articular disease activity and the features of lung disease [2, 9, 10, 23].

While waiting for specific controlled trials in patients with RA-ILD, abatacept and rituximab, as well as JAK inhibitors in selected cases, remain the first choices for the treatment of arthritis; however, treatment of lung involvement needs to be evaluated case by case by a multidisciplinary team [2, 9, 10]. For this reason, referral of patients with RA-ILD to centers with experienced rheumatologists and pulmonologists should be evaluated for the management of these complex patients.

Funding

Open access funding provided by Aarhus University Hospital. The authors received no funding for the present work.

Declarations

Conflict of Interest

The authors declare no conflicts of interest.

Availability of data and material

Not applicable.

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Code availability

Not applicable.

Authors’ Contributions

MS, FL, and EB jointly determined the content of the manuscript. MS drafted the initial version, and FL and EB critically revised it. All authors have read and approved the final version of the manuscript and agree to be accountable for all aspects of the work.