Management of Connective Tissue Disease–associated Interstitial Lung Disease

A thorough, often multidisciplinary assessment to determine extrathoracic versus intrathoracic disease activity and degrees of impairment is needed to optimize the management of connective tissue disease (CTD)–associated interstitial lung disease (ILD). Pharmacologic intervention with immunosuppression is the mainstay of therapy for all forms of CTD-ILD, but should be reserved for those that show clinically significant and/or progressive disease. The management of CTD-ILD is not yet evidence based and there is a need for controlled trials across the spectrum of CTD-ILD. Nonpharmacologic management strategies and addressing comorbidities or aggravating factors should be included in the comprehensive treatment plan for CTD-ILD.

Key points

  • Connective tissue disease (CTD)–associated interstitial lung disease (ILD) reflects a heterogeneous spectrum of diverse CTDs and a variety of patterns of interstitial pneumonia. Other than a few controlled trials in scleroderma-ILD, there are few studies to reliably inform an evidence-based approach to managing CTD-ILD, and, in general, clinicians are left with experience-based approaches.

  • The management of CTD-ILD is limited to cases with progressive and/or clinically significant disease.

  • Immunosuppression with corticosteroids and cytotoxic medications are the mainstay of pharmacologic treatment.

  • Extrathoracic manifestations of the CTD need to be assessed and may also affect choice and intensity of immunosuppressive therapies.

  • Nonpharmacologic approaches to treatment should be considered for each patient with CTD-ILD.


Although connective tissue disease (CTD)–associated interstitial lung diseases (ILDs) are sometimes considered a homogeneous entity (such as in this and many other articles), the spectrum of CTD-ILD reflects a heterogeneous category of diseases comprising the different CTDs along with the various interstitial pneumonia (IP) patterns. It remains to be determined whether the approach to management of one type of CTD-ILD (eg, systemic sclerosis [SSc]–nonspecific IP [NSIP]) can be applied to other forms of CTD-ILD (eg, rheumatoid arthritis [RA]–usual IP [UIP] or idiopathic inflammatory myositis [IIM]–organizing pneumonia [OP]). This article discusses our approach to the pharmacologic and nonpharmacologic therapeutic strategies for the spectrum of CTD-ILD.


Although connective tissue disease (CTD)–associated interstitial lung diseases (ILDs) are sometimes considered a homogeneous entity (such as in this and many other articles), the spectrum of CTD-ILD reflects a heterogeneous category of diseases comprising the different CTDs along with the various interstitial pneumonia (IP) patterns. It remains to be determined whether the approach to management of one type of CTD-ILD (eg, systemic sclerosis [SSc]–nonspecific IP [NSIP]) can be applied to other forms of CTD-ILD (eg, rheumatoid arthritis [RA]–usual IP [UIP] or idiopathic inflammatory myositis [IIM]–organizing pneumonia [OP]). This article discusses our approach to the pharmacologic and nonpharmacologic therapeutic strategies for the spectrum of CTD-ILD.

General principles

A fundamental principle pertinent to CTD-ILD is that many patients do not require immunosuppressive therapy targeting the ILD. In many cases, immunosuppressive treatment is needed for the extrathoracic inflammatory disease features (eg, synovitis or myositis) but not the ILD. Given the high prevalence of subclinical ILD in CTDs, it is crucial to determine the degree of respiratory impairment in all patients with CTD-ILD (see article by Ryerson elsewhere in this issue). The decision to treat CTD-ILD is often based on whether the patient is clinically impaired by the ILD; whether the ILD is progressive by symptoms, physiology, and/or imaging; and what extrathoracic features require therapy.

Management of extrathoracic manifestations

Traditional and biologic disease modifying antirheumatic drugs (DMARDs) are commonly used to treat the extrathoracic manifestations of the CTD, particularly synovitis and myositis. However, there is also evidence that some of these therapies, methotrexate (MTX) in particular, have the potential for causing pneumonitis. Those with prior lung involvement are possibly more susceptible. In general, because of its potential for lung toxicity, and because it is often difficult to distinguish MTX-pneumonitis from a flare of underlying ILD, we tend to avoid or remove MTX in our patients with CTD-ILD. The biologic DMARDs have become a mainstay in rheumatology practice, and have shown a high degree of efficacy for synovitis, myositis, ocular aspects, and cutaneous aspects of RA and other CTDs. There has been some evidence based on case reports, retrospective studies, and postmarketing surveillance that biologic DMARDs, particularly the anti–tumor necrosis factor (TNF) class, may be associated with an increased risk of pneumonitis. However, because no prospective controlled data exist and there is an appreciation that multiple variables, including the underlying CTD, are associated with ILD development, progression, or exacerbation, we suggest that these agents be used with caution in patients with CTD-ILD and other treatment options prioritized when possible. We frequently use the spectrum of biologic DMARDs in our patients with CTD-ILD to manage the extrathoracic manifestations and find that these agents often have minimal, if any, impact on the ILD. When both ILD and extrathoracic manifestations require treatment, we combine an agent to target the ILD (eg, azathioprine [AZA] or mycophenolate mofetil [MMF]) with a biologic DMARD (eg, etanercept or rituximab [RTX]) to target the synovitis or myositis.

Pharmacologic therapy

For those individuals with CTD-ILD in whom the ILD has been deemed to be clinically significant and progressive, pharmacologic treatment with immunosuppression is judged an appropriate step in management ( Fig. 1 ). There are few data to adequately inform the discussion on management strategies for CTD-ILD, the only controlled data being limited to 2 clinical trials in SSc-ILD. As such, much of the management of the spectrum of CTD-ILD is left to experience-based rather than evidence-based practice. When considering the general approach to the management of CTD-ILD, we consider the concept of induction followed by maintenance therapy, similar to the approach to systemic vasculitis. Induction therapy may require higher dosing of corticosteroids (CSs) along with short-term use of a more potent (and potentially more toxic) agent such as cyclophosphamide (CYC) followed by a maintenance regimen with a less toxic agent (such as AZA or MMF) and CS tapering. In addition, because of the often-poor prognosis associated with ILD, the ability to stabilize disease is considered a successful outcome, and, as such, setting expectations with patients and having informed discussions about goals of therapy can be helpful.

Fig. 1

Suggested management algorithm for CTD-ILD. CS, corticosteroids; CYC, cyclophosphamide; GERD, gastroesophageal reflux disease; IV, intravenous; PH, pulmonary hypertension; PjP, Pneumocystis jiroveci pneumonia.


CS therapy is the cornerstone of induction therapy in most forms of CTD-ILD in which immunosuppressive therapy is deemed to be necessary. Considering that long-term immunosuppression is typically indicated, CS therapy is usually combined with a CS-sparing agent. We tend to initiate CSs at a dose of 0.5 to 1.0 mg/kg/d of prednisone equivalent. Depending on the clinical response and tolerability of the CS and the secondary agent, we often slowly taper the CS to attain a daily dose of 10 mg of prednisone equivalent between the fourth and sixth months of therapy, and hope for tapering off altogether when clinically feasible. No tapering regimen has been studied or proved to be more effective. A notable exception to CS use is in SSc-ILD, in which moderate to high doses of CSs are traditionally considered a risk for SSc renal crisis. In SSc-ILD, we aim to keep the prednisone dose less than or equal to 15 mg/d. In our experience, there are clinical scenarios for which more intense use of CSs should be considered: acute IP, cellular NSIP or OP that may be more reversible with more intense up-front dosing of CS followed by a prolonged taper, or the NSIP encountered in patients with antisynthetase syndrome. In these cases, a pulse-dose course of intravenous (IV) methylprednisolone (500–1000 mg IV for 3 days followed by weekly pulses of 250–1000 mg IV for several weeks) concomitantly with daily oral CS at 1 mg/kg of prednisone equivalent should be considered.

The role of CSs (or any immunosuppressive agent) in patients with the UIP pattern of lung injury may be controversial. For those with idiopathic UIP (ie, the clinical scenario of idiopathic pulmonary fibrosis [IPF]), immunosuppression with the combination of CS and CYC has been proved to be ineffective. Furthermore, recent data suggest that the combination of AZA, CS, and N -acetylcysteine for the treatment of IPF may lead to an increase risk of hospitalization and mortality. However, such data are limited to IPF and, in our opinion, should not be extended to UIP associated with CTD. We tend to treat our patients with CTD-UIP with immunosuppression and this usually includes an initial course of CS. Further studies are needed to better guide the role of CSs in CTD-ILD in general and with UIP in particular.


CYC, one of the most potent CS-sparing immunosuppressive medications, is often used to treat a variety of organ-threatening CTD manifestations. Small prospective and retrospective studies have suggested that the use of CYC in CTD-ILD, and SSc-ILD in particular, may lead to stabilization or improvement in lung function. In practice, CYC is often considered the first line of therapy for the most severe forms of CTD-ILD. CYC is the only agent for which there are controlled clinical trial data in CTD-ILD, but those data are limited to SSc-ILD and these studies only show modest effects of CYC.

In the Scleroderma Lung Study (SLS), 158 subjects were randomized to oral CYC less than or equal to 2 mg/kg/d or placebo for 12 months. The primary end point was change in forced vital capacity (FVC). The cohort comprised subjects with SSc with evidence of active ILD by bronchoalveolar lavage or thoracic high-resolution computed tomography (HRCT) scan, early disease (first non-Raynaud symptom within 7 years), an FVC of 45% to 85% of predicted, and at least moderate exertional dyspnea. Subjects with a lung diffusion capacity for carbon monoxide (DL co ) less than 30% of predicted, tobacco use in the prior 6 months, or any other significant pulmonary disease were excluded. The subjects were mostly women (70.3%) with a mean age of 47.9 ± 1.0 years and 59.5% had diffuse cutaneous SSc (dcSSc). The baseline mean FVC was 68.1% ± 1.0% of predicted and DL co was 47.2% ± 1.1% of predicted. The FVC difference (ΔFVC) at 12 months was +2.53% of predicted ( P <.03) in favor of the CYC group, remaining significant at 18 months from time of CYC initiation. However, by 24 months from CYC initiation, the treated group had regressed to FVC measures similar to the placebo arm. Subjects with more restrictive disease (FVC, <70% of predicted), higher fibrosis scores on thoracic HRCT scan, or more skin thickening had a more robust response to CYC (FVC at 18 months from baseline, +5.10% with CYC vs −4.71% with placebo; ΔFVC, +9.81%; P <.001).

The Fibrosing Alveolitis in Scleroderma Trial (FAST) randomized 22 subjects to IV CYC 600 mg/m 2 monthly for the first 6 months followed by AZA 2.5 mg/kg/d with background oral prednisolone 20 mg on alternate days compared with placebo (n = 23). Subjects were mostly women with limited cutaneous SSc. The ΔFVC in the active treatment group (FVC 0 , 80.1% ± 10.3% of predicted and FVC 12 , 82.5% ± 11.3% of predicted) compared with the placebo group (FVC 0 , 81.0% ± 18.8% of predicted and FVC 12 , 78.0% ± 21.6% of predicted) showed a trend toward statistical significance (ΔFVC, +4.19% between groups; P = .08). The ΔFVC was more favorable in FAST than in SLS (+4.2% vs +2.5%, respectively) but the smaller number of subjects in FAST (n = 45) compared with SLS (n = 158) affected the ability to achieve statistical significance.

In our opinion, the results from SLS and FAST have diminished enthusiasm for the use of CYC. Although modest improvements were noted in FVC, with its substantial toxicity (bone marrow suppression, infection risk, malignancy risk), its short-term and long-term use is tempered in SSc-ILD and in all forms of CTD-ILD.

The role of CYC in other forms of CTD-ILD is based on retrospective studies. In a study of 46 patients with IIM-ILD resistant to CSs, 24 were subsequently treated with oral CYC. At 6 months, the ΔFVC was +8.0%, the DL co difference (ΔDL co ) was +5.0%, and the prednisone dose was reduced from 40 mg/d to 10 mg/d. In another report by Yamasaki and colleagues, 8 of 17 patients had greater than 10% improvement of vital capacity (VC) and 9 of 17 had greater than 10-point reduction in their thoracic HRCT scan scores after 6 months of treatment with CYC.

We tend to use CYC for the spectrum of CTD-ILD when the disease is severe or rapidly progressive. Because of the better safety profile of the monthly IV form compared with daily oral administration, we tend to infuse CYC monthly, generally for at least 6 months but rarely for more than 12 months. After the course of CYC, we tend to switch to a less toxic agent (eg, AZA or MMF) for longer-term maintenance therapy.


AZA is a commonly used medication for the treatment of CTD-ILD. However, other than as used in FAST, the data for AZA are limited to small and retrospective series. Studies using AZA as a maintenance therapy following 6 months of IV CYC in SSc-ILD report contradictory data. Less well studied as an induction agent, AZA in SSc-ILD has been retrospectively assessed in 14 patients, 8 treated for 12 months and 7 for 18 months. With a baseline FVC of 54.3% ± 3.5% of predicted, 5 patients had an increase of greater than 10% and 3 stayed within 10% of their baseline at 12 months. Another recent series in SSc-ILD compared 15 patients receiving AZA (1.5–2 mg/kg/d) as an induction agent with 21 patients with oral CYC (up to 2 mg/kg/d) for 6 months with low-dose prednisolone (≤10 mg). The FVCs at baseline and 12 months after treatment in AZA-treated patients were 62.8% ± 9.8% of predicted and 71.1% ± 20.9% of predicted, respectively (ΔFVC, +7.6% ± 13.1%; P = .05), and in CYC-treated patients were 59.5% ± 10.7% of predicted and 63.1% ± 16.2% of predicted (ΔFVC, +2.9% ± 11.5%; P = .19). In the retrospective study by Mira-Avendano and colleagues of 46 patients with IIM-ILD resistant to CSs, 13 were treated with AZA. At 6 months, the ΔFVC was −4.0%, the ΔDL co was +1.0%, and the prednisone dose was reduced from 40 mg/d to 13 mg/d. In a cohort of Sjögren syndrome–ILD with multiple IP patterns, 11 patients were treated with AZA for 6 months: 7 had an increase of their FVC by greater than 10% compared with none in the 9 nontreated patients ( P <.05).

In general, we find AZA to be a well-tolerated therapy and plausibly effective CS-sparing agent suitable for the long-term treatment of CTD-ILD. We particularly like to use AZA in RA-ILD because this agent can help control both the synovitic and ILD components, and, when needed for more effective synovitis control, we combine AZA with a biologic DMARD.

Mycophenolate Mofetil

MMF has become an increasingly popular treatment in CTD-ILD. The first series advocating MMF in CTD-ILD comprised 28 subjects and showed that MMF was well tolerated and associated with preservation of lung function among a diverse spectrum of CTD-ILD. Several other small series have suggested a beneficial role of MMF for SSc-ILD. The largest study of MMF use for CTD-ILD included a heterogeneous cohort of 125 patients with CTD-ILD (44 SSc-ILD, 32 IIM-ILD, 18 RA-ILD), with a mean age of 60.4 ± 11.6 years; 42% were women and most were treated with MMF 3 g/d over a 3-year period. In this large and diverse CTD-ILD retrospective cohort, MMF treatment was associated with effective CS dose tapering (from a median of 20 mg/d to 5 mg/d of prednisone at 12 months from MMF initiation [ P <.0001]). MMF was also associated with longitudinal improvements in FVC and DL co and was a well-tolerated therapy (∼90% adherence rate).

In our experience, MMF is often an effective agent in the management of CTD-ILD. In addition, similar to AZA, MMF is well tolerated and can be safely combined with biologic DMARDs in CTD-ILD when needed to control extrathoracic disease manifestations (eg, synovitis).

Calcineurin Antagonists

Several retrospective studies suggest beneficial roles of cyclosporine and tacrolimus in CTD-ILD. Harigai and colleagues reported the use of cyclosporine in 48 patients with CTD-ILD (26 IIM, 7 SSc, 7 RA, 8 other CTDs). Cyclosporine was effective for 72.2% and 33.3% in acute cases versus 50.0% and 50.0% in chronic cases in the IIM-ILD and other CTD-ILD, respectively. A more recent study retrospectively reviewed 17 patients with antisynthetase syndrome–ILD who failed prednisone within 12 months of ILD onset and subsequently started cyclosporine. At 2 years of follow-up, FVC increased from 65% to 76% of predicted and DL co from 55% to 63% of predicted. The results were substantially maintained, including at the last available follow-up (median follow-up time, 96 months). Tacrolimus can be an effective agent for the spectrum of CTD-ILD and, perhaps, in IIM-ILD in particular. In a retrospective study, 13 patients with both myositis and ILD were treated with tacrolimus for a mean period of 51.2 months, showing maintained improvement in myositis, FVC, and DL co during 150 weeks of follow-up. Other observations show potential efficacy in refractory cases previously treated with combination therapies including CYC and cyclosporine and addition of tacrolimus on conventional therapy (prednisolone, IV CYC, and/or cyclosporine) was shown to be associated with better survival.


Several recent reports suggest that RTX may be an effective agent for CTD-ILD. Keir and colleagues reported 8 cases of CTD-ILD (5 IIM-ILD; median FVC, 45% of predicted; median DL co , 25% of predicted) in which RTX was used as rescue therapy. Six of these patients had serial pulmonary function tests (PFTs): before RTX infusion, all had decline in FVC and DL co , and after RTX infusion, a median DL co improvement of 22% ( P = .04) and a median FVC improvement of 18% ( P = .03) were noted. The same group recently reported their experience with RTX infusions in 50 cases of severe and refractory ILD; 33 of these cases had CTD-ILD (10 IIM, 8 SSc, 9 undifferentiated CTD), 4 required mechanical ventilation, DL co was 24.5% of predicted, and FVC was 44.0% of predicted. In the CTD-ILD subgroup, 85% of the patients (most with IIM) were classified as responders. In the 6 to 12 months before RTX, a median decline in FVC of 13.3% and in DL co of 18.8% were noted compared with the 6 to 12 months after RTX therapy, in which an improvement of 8.9% of the FVC ( P <.01) and a stabilization of the DL co ( P <.01) were noted. Dodds and colleagues also reported their retrospective experience of RTX use in 22 patients with severe refractory CTD-ILD as rescue therapy (5 RA, 7 IIM, 2 SLE, 2 SSc, 3 other CTDs). Nine patients showed radiographic stability or improvement, whereas 4 patients showed deterioration under RTX. In those patients in whom RTX prevented disease progression (n = 8), there was improvement in FVC (+2.5%; range, 0%–7%; P = .08) over an average period of 19 months.

There are several small studies of RTX in SSc-ILD. With secondary end points of PFT and thoracic HRCT scan, 15 patients with dcSSc, and FVC and DL co greater than or equal to 50% of predicted received the RA regimen. At 6 months, the FVC, the DL co , and the thoracic HRCT scan remained stable. Daoussis and colleagues reported the results of a small randomized controlled trial of 8 patients with SSc-ILD treated with the vasculitis protocol (375 mg/m 2 weekly for 4 weeks) and then again 6 months later compared with 6 subjects receiving standard treatment (including prednisone, MMF, CYC, and bosentan). At 1 year, the FVC in the RTX group increased by 10.3% (68.1% ± 19.7% to 75.6% ± 19.7%; P = .0018) compared with the control group (86.0% ± 19.6% to 81.7% ± 20.7%; P = .23) losing 5.0% ( P = .002). The DL co also improved by 9.7% (52.3% ± 20.7% to 62.0% ± 23.2%; P = .017) compared with a decrease of 7.5% ( P = .023) in the control group (65.3% ± 21.4% to 60.2% ± 23.7%; P = .25). Follow-up of the RTX-treated patients at 2 years from the first infusion showed stability for both FVC and DL co ( P <.0001 for both values).

Several case reports suggest that RTX may have a role in severe, refractory IIM-ILD. A small retrospective series of 11 patients with antisynthetase syndrome–ILD were treated with RTX as a rescue therapy. Comparing the 8 months’ pretreatment data and the 7 months’ posttreatment PFT data, 6 patients had an FVC improvement of greater than 10% and 3 had a DL co increase of greater than 15%. Three to 6 months following infusion, the thoracic HRCT scan showed a regression of the ground-glass opacities in 4 patients and progression in 1. In another study by Unger and colleagues, 8 patients with IIM-ILD were treated with RTX. The baseline FVC of 74% ± 19% of predicted increased to 83% ± 21% of predicted at 7 months, 91% ± 21% of predicted at 12 months, and 108% ± 15% of predicted at 21 months following the initial cycle (4 of these patients received more than 1 RTX cycle). The DL co was not significantly improved at 6 months from the initial RTX cycle.

In addition, the results of 10 patients with RA-ILD (4 UIP, 6 NSIP, baseline FVC 68% [range, 47%–89%], baseline DL co , 48% [range, 28%–73%]) treated with RTX was recently reported. Of the 7 subjects with data at baseline and 48 weeks, FVC and DL co worsened in 1 subject, stabilized in 4, and improved by greater than 10% in 2 (at 48 weeks, FVC 75% [range, 50%–102%] and DL co , 52% [range, 30%–75%]). Also, among 188 patients with RA-ILD in 16 centers across the United Kingdom during a 25-year period (65% UIP), 57 patients were treated with a biologic agent. No difference in all-cause or respiratory mortality was noted in patients treated with biologics versus other agents. However, there was a statistically significant difference in respiratory mortality between patients treated with anti-TNF (n = 30) versus RTX (n = 27) (15% vs 4%; P = .04) and in all-cause mortality in 31% of patients treated with anti-TNF versus 8% of patients treated with RTX ( P = .03) in the UIP subgroup.

The role of RTX for CTD-ILD remains to be defined. This agent may have a role in specific subsets of CTD-ILD, such as the antisynthetase syndrome and those cases in which lung biopsy suggests a role of B cells in the ILD pathogenesis. Further studies are needed to more precisely define its role in CTD-ILD.

Future Directions in the Pharmacologic Treatment of Connective Tissue Disease–associated Interstitial Lung Disease

There have been several novel antifibrotic therapies studied in ILD but these have almost exclusively been limited to clinical trials for patients with IPF. The only antifibrotic agent studied in CTD-ILD was bosentan for SSc-ILD in the BUILD-2 trial, which showed that bosentan is ineffective for the ILD in SSc. Recent studies of pirfenidone and nintedanib have shown a positive impact on disease progression in patients with IPF ; however, there are currently no data to support their use in CTD-ILD. There is an ongoing phase II study addressing safety and tolerability of pirfenidone for SSc-ILD (LOTUSS trial; NCT01933334 ).

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Sep 28, 2017 | Posted by in RHEUMATOLOGY | Comments Off on Management of Connective Tissue Disease–associated Interstitial Lung Disease

Full access? Get Clinical Tree

Get Clinical Tree app for offline access