Pharmacological Treatment of Idiopathic Pulmonary Fibrosis (Update) and Progressive Pulmonary Fibroses: S2k Guideline of the German Respiratory Society Open Access

1. Preamble: Definitions: IPF; PPF (PF-ILD);

2. Discussion and Appraisal of Different Pharmacological Approaches

2.1. Shall IPF Patients Be Treated with an Antifibrotic Drug?

2.1.1. Shall IPF Patients Be Treated with Nintedanib?

2.1.2. Shall IPF Patients Be Treated with Pirfenidone?

2.2. Shall IPF Patients Be Treated with Antacid Medication?

2.3. Shall IPF Patients Be Treated with N-Acetylcysteine?

2.4. Shall IPF Patients Be Treated with Dual Endothelin Receptor Antagonists (ET-A and ET-B) Bosentan or Macitentan?

2.5. Shall IPF Patients Be Treated with the PDE5 Inhibitor Sildenafil?

2.6. Shall PPF Patients Be Treated with Anti-inflammatory Agents?

2.6.1. Non-Autoimmune PPF

2.6.2. Autoimmune PPF

2.6.2.1. Introduction

2.6.2.2. Rheumatoid Arthritis-Associated ILD

2.6.2.3. Sjögren’s-Associated ILD

2.6.2.4. Myositis-Associated ILD

2.6.2.5. Systemic Sclerosis-Associated Interstitial Lung Disease

2.6.2.6. Mixed Connective Tissue Disease-Associated ILD

2.7. Shall PPF Patients Be Treated with an Antifibrotic Drug?

2.7.1. Introduction

2.7.2. Shall PPF Patients Be Treated with Nintedanib?

2.7.3. Shall PPF Patients Receive (Off-Label) Treatment with Pirfenidone?

2.7.4. Treatment of Pulmonary Fibrosis Related to Occupational Exposure to Fibrogenic Noxae

2.8. Shall Patients with Fibrotic ILD (Including IPF) and Pulmonary Hypertension be Treated with Inhaled Prostanoids?

3. Basis and Objectives of the General Recommendations

4. Discussion and General Considerations for Antifibrotic Therapy in IPF and in PPF

4.1. When Should Antifibrotic Therapy Be Started?

4.1.1. In IPF

4.1.2. In PPF

4.2. When Should Antifibrotic Therapy Be Discontinued?

4.2.1. In IPF

4.2.2. In PPF

4.3. When Should Switching Antifibrotic Drugs Be Recommended?

4.3.1. In IPF

4.3.2. In PPF

4.4. Is Antifibrotic Combination Therapy an Option?

4.5. What Are the Implications of the Differences in the EMA Approvals of Pirfenidone and Nintedanib in IPF?

4.6. When Should Anti-inflammatory Therapy Be Discontinued?

Interstitial lung diseases (ILD) are characterized by a variable degree of inflammatory and fibrotic changes within the alveolar space and distal airways (bronchioles) [1]. An inverse correlation exists between the extent of fibrosis and the possibility that an ILD is reversible. While the acute (inflammatory) type of extrinsic allergic alveolitis may resolve without sequelae (restitutio ad integrum), idiopathic pulmonary fibrosis (IPF) is the prototypic fibrotic ILD with a progressive course, leading to irreversible and progressive fibrosis of the lung parenchyma. Figure 1 provides an overview of the major disease types and the proportion of progressive fibrotic ILD (PF-ILD) [2].

Diagnostic guidelines are available for the accurate diagnosis of IPF [3, 4]. In addition, it has been demonstrated multiple times that the diagnosis of IPF per se is associated with a progressive fibrotic process and high mortality in the majority of cases, without such progression necessarily being recognizable at the time of diagnosis. The progression of IPF manifests as a worsening of respiratory symptoms, concomitant decline in lung function, and progressive remodeling of healthy lung tissue to fibrotic scar tissue, as seen in high-resolution computed tomography (HRCT) imaging. These findings have led to the approval of two drugs – nintedanib and pirfenidone – for the treatment of IPF because they reduce progression, as measured by a decline in forced vital capacity (FVC).

The situation is less obvious in other ILDs with a relevant inflammatory component, where a progressive fibrotic phenotype often only develops during the course of the disease. Observational studies demonstrated that, on average, 18–32% of ILDs transition to a PF-ILD [5‒10]. The fibrotic component of PF-ILD has pathobiological similarities with IPF and also responds favorably to antifibrotic therapy. This has so far been confirmed by studies and a meta-analysis of the antifibrotic drug nintedanib, which was approved for the treatment of PF-ILD on this basis [11, 12]. Although two randomized, controlled trials and their meta-analysis show that the antifibrotic effect of pirfenidone slows disease progression, these studies were flawed due to technical problems or insufficient sample size [1, 13, 14]. Pirfenidone is currently not approved for use in patients with PF-ILD or progressive pulmonary fibrosis (PPF), respectively. Specific clinical criteria were used in the respective pivotal studies to define disease progression and determine the presence of PF-ILD. In the studies of nintedanib, at least one of the following criteria had to be met at the time of inclusion in the study [11]:

• 1.

  FVC relative decline ≥10% predicted within 24 months

• 2.

  FVC relative decline ≥5% to <10% predicted and worsening respiratory symptoms within 24 months

• 3.

  FVC relative decline ≥5% to <10% predicted and increased signs of fibrosis in HRCT within 24 months

• 4.

  Increase in signs of fibrosis in HRCT and worsening respiratory symptoms within 24 months

The following criteria were used in the pirfenidone studies to determine the PF-ILD phenotype [13, 14]:

• 1.

  FVC absolute decline ≥5% predicted per year, extrapolated based on at least three measurements within 6–24 months [13]

• 2.

  FVC absolute decline >5% predicted within 6 months [14]

• 3.

  Significant worsening of respiratory symptoms within 6 months not explained by cardiovascular or other causes [14]

Other criteria for determining progression and, thus, the PF-ILD phenotype were used in some publications:

• FVC or CO diffusion capacity (DLco) absolute decline ≥10% predicted (FVC) or ≥15% predicted (DLco) within 24 months [15]

• Relative FVC deterioration ≥5% and DLco decline ≥15% within 24 months [16]

In addition to the criteria described above, the following findings and clinical events may also indicate progression of the fibrotic ILD, provided that other causes have been excluded [17]:

• 1.

  Deterioration of oxygenation and initiation of long-term oxygen supplementation during exercise or at rest according to the current Guideline for Long-Term Oxygen Therapy [18]

• 2.

  Hospitalization for worsening respiratory symptoms or acute ILD exacerbation

• 3.

  Deterioration in the distance walked in the 6-min walk test (6MWT)

• 4.

  Decrease in minimum oxygen saturation during 6MWT

While a minimum observation period of 6 months was required in the RELIEF study, no other study or recommendation has addressed the question of the minimum observation time required to detect progression [13]. Pretreatment according to the standard of care was recommended but there was no strict requirement for any specific anti-inflammatory therapy for inclusion in these studies.

The current international ATS/ERS/JRS/ALAT clinical practice guideline introduces the term “progressive pulmonary fibrosis” (PPF) to replace the previously used term “PF-ILD” and defines progression as follows [2]:

At least two of the following three key criteria occurring within the past year:

• 1.

  Worsening respiratory symptoms

• 2.

  Physiological evidence of disease progression based on lung function (either of the following):

  • Absolute decline in FVC ≥5% predicted

  • Absolute decline in DLco (corrected for Hb) ≥10% predicted

• 3.

  Radiological evidence of disease progression (one or more of the following):

  • Increased extent or severity of traction bronchiectasis and bronchiolectasis

  • New ground-glass opacity with traction bronchiectasis

  • New fine reticulation

  • Increased extent or coarseness of reticular abnormality

  • New or increased honeycombing

  • Increased lobar volume loss

It is critical to exclude alternative causes of worsening of respiratory symptoms, lung function, or radiological findings before the above criteria can be applied as evidence of disease progression in terms of a PPF. This is particularly true for the DLco, which is influenced by a variety of factors, including non-pulmonary ones [2].

The German Guideline Committee supports the proposed term “progressive pulmonary fibrosis” (PPF), highlighting that it should be seen as synonymous with the previously used term “PF-ILD,” which was also used in the EMA approval document. The German Guideline Committee considers the definition of progression proposed by the international guideline as too narrow, in particular, because important clinical indicators of progression are ignored and some patients may not be able to undergo lung function testing to satisfy the functional criteria, e.g., following an exacerbation, even though a progression is obvious in this case. This might lead to the exclusion of relevant patient groups from beneficial antifibrotic treatment as they will no longer be able to meet the progression criteria.

From the Guideline Committee’s point of view, the following requirements shall be met to initiate an antifibrotic therapy in PPF patients:

• 1.

  Fibrotic ILD affecting at least 10% of the lung parenchyma on HRCT imaging.

• 2.

  Signs of progression – clinical, functional or radiomorphological deterioration – must be present for a reasonable period of time within a 24-month period. Adopted from the INBUILD pivotal trial [11], the following criteria apply as a general guideline:

Either

• an FVC relative decline ≥10%

Or evidence of at least two of the following criteria:

• 1.

  Worsening respiratory symptoms

• 2.

  FVC relative decline ≥5% predicted

• 3.

  Increase in signs of fibrosis on HRCT*

• 4.

  DLco absolute decline ≥15% predicted (DLco)

• 5.

  Initiation of long-term oxygen supplementation during exercise or at rest or permanent increase in the oxygen flow of an ongoing long-term oxygen supplementation therapy by at least 1 L/min according to the current Guideline for Long-Term Oxygen Therapy [18]

• 6.

  Hospitalization in case of worsening respiratory difficulties or acute ILD exacerbation

• 7.

  Decrease in the distance walked in the 6MWT by ≥50 m or 20% and/or drop in minimum oxygen saturation during the 6MWT by >5% and below 88% absolute on room air, or ongoing oxygen supplementation at a steady flow rate

*Hallmarks of the radiologic pattern of fibrosis are

• Increase in the extent or severity of traction bronchiectasis and bronchiolectasis

• New ground-glass opacity with traction bronchiectasis

• New fine reticulation

• Increased coarseness of reticular abnormality

• New or increased honeycombing

• Increased lobar volume loss

From the Guideline Committee’s point of view, this proposal amalgamates the criteria successfully used in studies and clinically reasonable criteria, without excluding a patient group. In addition, the requirement that at least 10% of the lung parenchyma is affected by fibrotic remodeling increases the probability of progression. The Guideline Committee also emphasizes that each of these criteria is only valid and can therefore be used as an indicator of progressive fibrosis if other causes for the deterioration have been rigorously excluded. Causes of clinical deterioration to be ruled out include, in particular, left heart failure with pulmonary vascular congestion, developing pulmonary hypertension (PH), overhydration due to heart or kidney failure, and infections.

While the approval of nintedanib and pirfenidone for the treatment of IPF is a regular, diagnosis-based approval, the approval of nintedanib for the treatment of PPF is not an approval for the treatment of a specific disease but rather for a specific clinical progressive behavior that manifests with a variety of underlying diseases. It is therefore important to note that, in addition to the progression criteria, the presence of a fibrotic ILD needs to be confirmed based on CT morphology using an adequate thin-slice HRCT scan or based on histological evidence (if available). Reliable radiological criteria of pulmonary fibrosis are honeycombing and traction bronchiectasis or bronchiolectasis; reticulation is also an indicator of fibrosis, especially if quite extensive and subpleural-predominant [3, 4].

All statements regarding the medical therapy made in this guideline refer either to patients diagnosed with IPF based on current guideline criteria in the context of a multidisciplinary team discussion (MDD) [4] or, in the case of PPF, to patients with a fibrotic ILD affecting at least 10% of the lung parenchyma, with signs of progression as per the criteria specified above. An ILD MDD (ILD Board) is the gold standard for diagnosing PPF.

The ILD MDD (ILD Board) therefore plays a central role both in diagnosing the condition and in determining progression and, thus, the therapy indication. This requires, at a very minimum, the consultation of a pulmonologist experienced in ILD and of a radiologist with appropriate ILD expertise. If histological findings are available, a pulmonary pathologist shall be consulted as well [3]. Other optional members of an ILD multidisciplinary team are rheumatologists, thoracic surgeons, physiotherapists, dietitians, and palliative care physicians.

Interstitial lung abnormalities, defined as incidental findings of abnormalities affecting >5% of the lung parenchyma in the upper, mid, and lower lung zones in otherwise asymptomatic individuals, deserve specific consideration and are not included in this guideline [19]. Please refer to the relevant position paper from the Fleischner Society for information on their management [19].

Nintedanib is a tyrosine kinase inhibitor blocking in particular the signaling cascade of the vascular endothelial growth factor, the fibroblast growth factor and the platelet-derived growth factor receptors.

The recommendation provided in the latest 2017 update to the S2k guideline on the pharmacological treatment of IPF states as follows: “IPF patients shall be treated with nintedanib” [20]. In the meantime, antifibrotic treatment has become the “standard of care” for IPF patients. The efficacy of nintedanib therapy was demonstrated in several studies and further confirmed in real-world populations.

Before receiving approval, nintedanib was investigated in three clinical trials: one phase II trial (TOMORROW) and two phase III twin trials (INPULSIS-1 and -2) [21, 22]. All three studies consistently demonstrated that 150 mg of nintedanib b.i.d. can reduce the annual FVC decline and, thus, slow disease progression (TOMORROW: placebo arm −190 mL/year, nintedanib arm −60 mL/year; INPULSIS-1: placebo arm −239.9 mL/year, nintedanib arm −114.7 mL/year; INPULSIS-2: placebo arm −207.3 mL/year, nintedanib arm −113.6 mL/year; each p < 0.001). In addition, the time to the first acute exacerbation was longer in INPULSIS-2 [22]. In the pooled analysis of the two INPULSIS studies, a significant positive effect was found in terms of adjudicated acute exacerbations [22]. In the pooled analysis of TOMORROW, INPULSIS-1 and INPLUSIS-2, the time to first investigator-reported acute exacerbation was also significantly extended [22], and there was a trend in favor of nintedanib in terms of quality of life as assessed by the SGRQ (Saint George Respiratory Questionnaire) [22].

The most common side effects reported in these studies were diarrhea and nausea. Other adverse events, particularly serious adverse events, occurred at a similar rate in the active and in the placebo arm [22].

Data from the TOMORROW extension studies and the INPULSIS studies (INPULSIS-ON) confirmed the positive effect of nintedanib on disease progression in IPF beyond 52 weeks (median treatment time 45 months, maximum treatment time 68 months) with consistent side effect profiles [23, 24].

Various post hoc analyses of subgroups of the INPULSIS studies showed no differences in the effect of nintedanib with regard to age, gender, ethnicity, smoking status, baseline FVC, baseline SGRQ, steroid therapy, bronchodilator therapy, antacid co-medication, CT pattern, concomitant emphysema, and comorbidities [25‒30]. In a pooled analysis of five randomized trials, the positive effect of nintedanib on FVC decline was comparable for patients <75 and ≥75 years of age as well as for patients with <5 and ≥5 comorbidities [31]. Additional post hoc analyses of the INPULSIS studies revealed that more patients in the nintedanib arm than in the placebo arm experienced an improvement or no decline in FVC over the study period (25% vs. 9%) [32] and that nintedanib reduced the risk of occurrence of acute exacerbations [33, 34].

The INPULSIS studies excluded patients with severely impaired lung function (FVC ≤50% or DLco ≤30%), while patients with such conditions were included in the INPULSIS-ON study [24, 35] and in the INSTAGE study investigating the effect of nintedanib versus a nintedanib/sildenafil combination therapy over a 24-week period in patients with IPF and severely impaired gas exchange (DLco ≤35%) [36]. In these studies, the effect of nintedanib on FVC decline in patients with advanced lung function impairment was comparable to that seen in the INPULSIS pivotal studies for mild to moderate lung function impairment [36]. The safety and tolerability of nintedanib were similar in both patient groups [37].

Clinical insights from IPF centers in Germany support the positive effects of nintedanib in real-world patients [38]. So-called real-world experience reports from various other countries also confirm similar clinical outcomes with the already known side effect profile in patients with more advanced disease stages than in those investigated in the clinical trials [39‒42]. Retrospective studies showed that treatment with nintedanib before lung transplantation is safe in terms of bleeding complications, wound healing and short-term post-procedural survival [43‒45].

With regard to all-cause mortality, the pooled analysis of the INPULSIS studies found a beneficial effect, but statistical significance was only achieved in the on-treatment analysis [11]. Extrapolated pooled data from six clinical trials meanwhile suggest that nintedanib improves the life expectancy of IPF patients [46]. This is also supported by data from Medicare beneficiaries in the USA and from the German INSIGHTS-IPF registry as well as the European IPF registry, which indicate that antifibrotic medication may confer a survival benefit [47‒49]. However, the proportion of patients treated with nintedanib or pirfenidone was different, and no difference was seen where comparisons between the two treatment options were made.

The analysis of global pharmacovigilance data sets from 2014 to 2018 showed that the safety profile of nintedanib in IPF patients was no different from that already observed in the clinical trials and as described in the package leaflet [50]. No safety concerns were raised. From the Guideline Committee’s point of view, the recent post hoc analyses, the long-term data, as well as insights from real-world observations, speak for a beneficial and clinically relevant effect of nintedanib on the progression of IPF, with a generally acceptable side effect profile.

Pirfenidone is an antifibrotic agent and can block the release of pro-fibrotic and pro-inflammatory cytokines [51] as well as reduce collagen synthesis and fibroblast proliferation [52]. The latest 2017 update of the S2k guideline on the pharmacological treatment of IPF provided the same recommendation for pirfenidone as for nintedanib: “IPF patients shall be treated with pirfenidone” [20]. The efficacy of pirfenidone was confirmed in several studies as well as in real-world settings.

Pirfenidone has been studied in IPF patients in the context of six clinical trials: two phase II trials, of which one was an open-label [53] and the other one a randomized trial [54], as well as four phase III trials [55‒57]. While the patients in the phase III CAPACITY 1 trial were randomized into three groups (high-dose pirfenidone: 2,403 mg/day; low-dose pirfenidone: 1,197 mg/day; placebo), the otherwise identically designed CAPACITY 2 study had only two groups (high-dose pirfenidone: 2,403 mg/day; placebo) [56]. In CAPACITY 1, a significant decline in FVC was seen in the high-dose group versus placebo over a 72-week-period (mean change vs. baseline of −8.0% absolute vs. −12% absolute; p = 0.01), while no significant difference was seen in CAPACITY 2 (−9.0% vs. −9.6%; p = 0.50). The pooled analysis of the two CAPACITY studies (high-dose vs. placebo) revealed significant benefits of pirfenidone in terms of FVC decline, categorical deterioration of FVC >10%, progression-free survival, and distance walked in the 6MWT, but not with regard to all-cause mortality [56]. Patients in the ASCEND study (A Randomized, Double-Blind, Placebo-Controlled Trial of Pirfenidone in Patients with Idiopathic Pulmonary Fibrosis) were given either pirfenidone (2,403 mg/day) or placebo [57]. After 52 weeks of treatment with pirfenidone, the decline in FVC (% predicted) was 41.5% less and the percentage of patients who had died or experienced a decline FVC ≥10% was 47.9% less (p = 0.001) than in the placebo group. Treatment with pirfenidone resulted in significant benefits in terms of distance walked in the 6MWT, progression-free survival and reduced mortality numerically, although this difference was not significant (HR 0.55; CI: 0.26–1.15; p = 0.10) [57]. The pooled analysis of the CAPACITY 1 + 2 and ASCEND studies, including all patients under 2,403 mg of pirfenidone versus placebo, demonstrated significant benefits in terms of FVC decline, FVC decline ≥10% or mortality, progression-free survival, 6MWT distance as well as patient-reported dyspnea [58]. The pooled analyses of all three studies revealed a significant benefit of pirfenidone therapy in terms of all-cause mortality, IPF-related all-cause mortality, as well as all-cause and IPF-related mortality [57, 59]. The open-label extension studies of ASCEND and CAPACITY (RECAP) revealed no new aspects in terms of the long-term tolerability and safety profile of pirfenidone [60].

The pooled data indicate that nausea, skin rush (including phototoxicity), and anorexia, in particular, are more frequent under treatment with pirfenidone [58, 61]. Various post hoc subgroup analyses were published for the CAPACITY and ASCEND studies. There were no differences in the effect of the treatment with regard to FVC and the GAP (gender, age, physiology) index at baseline [62]. Patients with severely impaired lung function (FVC <50% and/or DLco <35%) also benefited from the treatment with pirfenidone in terms of FVC decline, all-cause mortality and hospitalization, without a significantly increased risk of adverse events [63]. In addition, the analyses showed that pirfenidone reduces the annualized risk of respiratory-related hospitalization [64]. Another post hoc analysis of the CAPACITY and ASCEND data found that pirfenidone reduced the multiple occurrences of disease progression events (defined as relative FVC decline ≥10%, absolute decline in distance walked in the 6MWT ≥50 m, respiratory-related hospitalization or all-cause death) and death in the year following such progression event [65]. In addition, patients with more advanced disease stages (FVC <80% and/or GAP stage II/III) who were treated with pirfenidone experienced positive effects with regard to patient-reported dyspnea under pirfenidone [66]. A post hoc analysis of the RECAP extension study revealed that, even in patients with severely impaired lung function (FVC <50% and/or DLco <35%), the FVC decline takes a similar course over an extended period as in patients with less impaired lung function, and that the safety profile is similar [67]. In addition, concomitant antacid therapy was comparable to pirfenidone alone, according to data from the CAPACITY and ASCEND studies [68]. The observations made in the PASSPORT multicenter study, in which real-world IPF patients under pirfenidone were followed for 2 years, were consistent with the safety profile observed in the clinical trials [69]. Real-world data from Germany also confirm the therapeutic benefit of pirfenidone without the occurrence of new adverse events [70‒72]. Treatment with pirfenidone also appears to be safe before lung transplantation and is not associated with an increased risk of bleeding, delayed wound healing, or poorer short-term post-procedural survival [43‒45].

Overall, the treatment with one of the two currently approved antifibrotic drugs seems to improve survival for IPF patients [47‒49]. Data from a small-scale study on 43 patients also demonstrate the positive effects of pirfenidone on cough symptoms in patients with IPF [73]. From the Guideline Committee’s point of view, these studies continue to highlight the beneficial and clinically relevant effect of pirfenidone on disease progression in IPF with a generally acceptable side effect profile.

An increased prevalence of gastroesophageal reflux (GERD) [74] is seen in IPF patients, and gastroesophageal reflux-associated (micro-)aspiration may damage the lung parenchyma [75]. A study on a small number of IPF patients suggests that surgical fundoplication may slow the progression of IPF [76] in this context. The extent to which antacid therapy might have a similar effect on IPF progression has not yet been investigated in a prospective, placebo-controlled study. Various post hoc analyses did not reveal any apparent beneficial effects of antacid therapy in terms of survival or the slowing down of the decline in lung function in patients under antacid therapy [77]. In the pirfenidone and nintedanib studies, neither the placebo nor the active treatment arms provided a positive effect in terms of overall survival or lung function decline in patients treated with additional proton pump inhibitors (PPI) [29, 78, 79]. These post hoc observations are consistent with analyses from the British healthcare system, where – using a so-called “new user cohort study design” – the effect of ongoing PPI therapy was investigated in over 1,800 patients already diagnosed with IPF, and compared to matched IPF patients not treated with PPI [80]. There was no difference in terms of all-cause mortality or hospitalization rate associated with the use of PPI [81].

Based on the study data available to date, there is no value in antacid therapy (such as PPI) for the treatment of IPF. Irrespective thereof, a guideline-directed medical therapy for gastroesophageal reflux disease is clearly indicated for IPF patients.

Since the first studies on the effect of N-acetylcysteine in combination with azathioprine (AZA) and prednisone or as monotherapy in patients with IPF were conducted [82, 83], in which no convincing effect of N-acetylcysteine on FVC was seen, further studies have investigated the effectiveness of N-acetylcysteine in IPF. While a 2019 meta-analysis describes a positive effect of N-acetylcysteine on FVC [84], reservations remain as, among other factors, the majority of the studies reviewed was insufficiently powered due to a small sample size and investigated a combination therapy of N-acetylcysteine and corticosteroids. The combination of N-acetylcysteine with pirfenidone as one of the drugs currently recommended for the treatment of IPF shows no additional benefit in terms of quality of life or delaying disease progression [85, 86]. The administration of inhaled N-acetylcysteine as add-on to pirfenidone may even be potentially harmful [87]. Of interest are post hoc analyses of patient cohorts treated with N-acetylcysteine in a study context (including those mentioned above). These analyses show differences in the effects of N-acetylcysteine on lung function depending on TOLLIP genotype [88]. A prospective study, in which this aspect will be investigated, is currently recruiting IPF patients for genotype-stratified treatment with N-acetylcysteine (clinicaltrials.gov; NCT04300920). In addition, another retrospective study suggests that N-acetylcysteine may have a positive effect on the transplant-free survival of patients with ANA-positive pulmonary fibrosis [89].

Based on currently available data, IPF patients shall not be treated with N-acetylcysteine outside controlled clinical trials. It may be possible, however, that IPF-associated genetic polymorphisms and/or phenotypes will be identified in the future, where N-acetylcysteine can have a beneficial effect.

A number of in vitro and animal study data suggest that the blocking of endothelin receptors may have a positive effect on the progression of pulmonary fibrosis [90, 91]. The BUILD-1 study of bosentan showed a statistically non-significant reduction in disease progression in the bosentan group, although the primary endpoint (distance walked in the 6MWT) was not met [92]. A consecutive phase-3 trial of bosentan versus placebo (BUILD-3) in a larger IPF patient population showed no effect on a combined primary endpoint including lung function and acute exacerbation [93]. Similar negative results were obtained in the MUSIC study, which investigated the effect of macitentan primarily on clinical deterioration and lung function in a population of 178 IPF patients. Treatment with macitentan showed no significant beneficial effect with regard to lung function or clinical deterioration of IPF [94].

A study on the effect of bosentan on the pulmonary hemodynamics of IPF patients could not demonstrate a therapeutic hemodynamic or clinical effect [95]. Based on currently available data, IPF patients shall not be treated with bosentan or macitentan.

Treatment of IPF with sildenafil was initially tested in two prospective studies with 180 IPF patients who were treated with 3 × 25 mg of sildenafil in the larger of the two studies (STEP-IPF study) [96, 97]. The primary endpoint (improvement in the distance walked in the 6MWT) was not met in either study, although some of the secondary endpoints, e.g., oxygenation and patient-reported dyspnea, improved significantly under sildenafil in the STEP-IPF study [96]. A significant effect on distance walking and quality of life was evident in particular in the sub-study of patients with echocardiographic findings of right-ventricular dysfunction [98]. In view of these potential benefits of sildenafil, two large studies investigated the extent to which sildenafil as add-on to an ongoing antifibrotic treatment with nintedanib [36] or pirfenidone [99] might have a positive effect on quality-of-life parameters, distance walked in the 6MWT, as well as lung function. Both studies included mainly patients with severely impaired diffusing capacity (DLco ≤35% and ≤40%, respectively). In addition, only patients with confirmed precapillary hypertension or with echocardiographic signs suggestive of PH were included in the pirfenidone/sildenafil study [100]. Both study results suggest that the addition of sildenafil to an ongoing antifibrotic treatment is safe. However, the primary endpoint was not reached in either study. The combination of sildenafil and nintedanib did not significantly improve patient-reported dyspnea (measured using specific questionnaires) at week 12 or 24, and there was also no significant difference in the lung function parameters. The only significant difference was that brain natriuretic peptide (BNP) plasma concentrations were lower in the active arm, which may be indicative of lesser myocardial (presumably right-ventricular) strain. A subgroup analysis of patients with or without echocardiographic signs of right-ventricular strain also found no significant difference in terms of quality of life and lung function, although the effect on BNP was stronger in patients with pre-existing right-ventricular strain [101]. The pirfenidone-sildenafil combination therapy over 1 year also had no significant positive effect on disease progression compared to a pirfenidone monotherapy [99], with disease progression being defined as a significant decline in the distance walked in the 6MWT, respiratory hospitalization, or death.

The evidence from available data is not sufficiently robust to support that sildenafil (as monotherapy or as add-on to antifibrotic therapy) would significantly improve relevant endpoints in IPF patients. IPF patients – with or without pre-existing PH – should not be treated with sildenafil monotherapy or with sildenafil in combination with antifibrotic drugs.

In most cases of chronic progressive fibrosing ILD, progression is triggered by a chronic inflammatory immune response. Prolonged inflammation eventually leads to scarring. Based on these considerations, anti-inflammatory therapy is routinely used to treat PPF. The strategy is complicated, however, by the fact that PPF can be caused by different types of inflammatory responses, e.g., by B- or T-cell-driven processes. Different anti-inflammatory drugs or drug combinations are therefore used based on a patient’s specific condition. The underlying inflammatory activity of the PPF varies greatly, also from patient to patient within a specific entity. In addition, inflammatory activity may change as the disease progresses [102] and often tends to get less as the fibrosis progresses. A retrospective study of patients with chronic fibrotic ILDs (37 iNSIP, 16 iPPFE, 133 unclassifiable IIP) in Japan showed that BAL lymphocytosis was associated with a better prognosis only in patients treated with anti-inflammatory drugs but not when no anti-inflammatory agents were used [103].

Evidence regarding the use of anti-inflammatory therapy in ILDs other than those caused by rheumatoid arthritis (RA) and collagen vascular systemic diseases is limited because there are no randomized controlled trials. Nevertheless, anti-inflammatory therapies are routinely used to treat progressive fibrotic ILDs. Even in the INBUILD study, patients with a UIP-like pattern were treated with glucocorticoids (GC; <20 mg/day) at baseline in approximately 50% of cases and with other immunomodulators in 18% of cases [102]. The same goes for the RELIEF study on PF-ILD, in which 81% of patients were treated with GC and/or other immunomodulators [13]. In a recent international online survey, specialists stated that 25–70% [104] of their PPF patients were not treated with anti-inflammatory drugs, especially considering those patients with idiopathic NSIP and unclassifiable ILD [105], where controversial data on the effectiveness of anti-inflammatory therapy exist. In patients with chronic progressive idiopathic NSIP and hypersensitivity pneumonitis (HP) who progress despite steroid monotherapy, AZA, mycophenolate mofetil (MMF), and cyclophosphamide (CYC) are mostly used as add-on to the steroid therapy, usually in a steroid-sparing intention [106‒112].

The successful use of rituximab (RTX) was reported in a small number of patients with progressive chronic HP (cHP) [113]. Additional data from clinical trials on the role of RTX for the treatment of ILD are to become available in the near future (NCT02990286).

When should anti-inflammatory therapy be started in non-autoimmune ILD? For patients with cHP, desquamative interstitial pneumonia, silicosis, and asbestosis, it is necessary to first rule out that the progression of pulmonary fibrosis is related to continued exposure (e.g., triggering antigen, smoking, occupational exposure). With regard to cHP, the risk of progression is higher and the prognosis is poorer if antigen avoidance is not implemented or if the triggering antigen is not known and antigen avoidance therefore not possible [114, 115]. The authors of this guideline use the following parameters to decide for or against the initiation/increase of anti-inflammatory therapy for PPF patients:

• 1.

  clinical parameters (e.g., previous response to therapy),

• 2.

  HRCT morphology (inflammatory vs. fibrotic),

• 3.

  clinical chemistry (e.g., CRP, IL-6, sIL2R),

• 4.

  differential cytology of BAL, as applicable, and

• 5.

  histology findings.

The decision shall be taken in the context of an ILD MDD (ILD Board). In the study by Yamagata et al. [103], BAL lymphocytosis <15% was an independent negative predictor for response to anti-inflammatory therapy. In a retrospective study of 91 patients with fibrotic HP, BAL lymphocytosis <20% in conjunction with honeycombing on HRCT was found to be corticosteroid refractory [116]. Consequently, a BAL lymphocytosis greater than 20–30% in the absence of a UIP pattern on HRCT may foster an anti-inflammatory treatment strategy.

Autoimmune ILD manifests as a highly heterogeneous group of conditions. Due to their heterogeneity and the limited data available, the following general rule applies: as with non-autoimmune PF-ILDs, a multidisciplinary approach with expert input from a rheumatologist is required.

The type of underlying rheumatic disease shall be considered in the decision on the anti-inflammatory therapy of choice for the autoimmune ILD, with due consideration of any additional organ involvement. The therapy shall treat the autoimmune ILD as well as any other organ manifestations. Below is an overview of the evidence available for the various disease types. The recommendations are specified, where possible.

RA-ILD affects approximately 10% of RA patients, primarily those who are APCA positive. Smoking, exposure to dust and male sex are risk factors for RA-ILD. In addition, the MUC5b polymorphism described as the main risk factor for IPF is also highly relevant for RA-ILD [117, 118]. The risk factors mentioned are associated in particular with the occurrence of a UIP pattern, which accounts for approximately half of RA-ILD cases and has a significantly higher rate of progression [117].

The evidence base for anti-inflammatory therapy in RA-ILD is generally weak. There are no randomized controlled trials, and the available studies rarely consider the underlying HRCT pattern.

Several retrospective studies report a better outcome and response for this therapy in RA-ILD types with no UIP pattern [119]. A large observational study with 2,701 patients [119] and one meta-analysis [120] both conclude that methotrexate therapy is not associated with the occurrence of ILD; on the contrary, it may delay the manifestation of ILD in RA patients. Numerous other drugs are used to treat RA-ILD, e.g., abatacept, tocilizumab, RTX, CYC, mycophenolate, and more recently, Jak inhibitors [121, 122]. A meta-analysis led to the impression that treatment with leflunomide, TNF inhibitors, RTX, or tocilizumab might increase the risk of developing RA-ILD [123]. In the meantime, however, investigators tend to assume that the impression was created because these drugs were often chosen for the treatment of severe cases with pulmonary involvement. A study by the British Society for Rheumatology demonstrated that RTX is superior to TNF inhibitors for the treatment of patients with RA-ILD [124].

ILD can be a rare manifestation of Sjögren’s syndrome [125]. However, given the high incidence of the condition and its usually quite slow progression, Sjögren’s-associated ILD is not a rare disease for pulmonologists. Women are most commonly affected by Sjögren’s syndrome. The ILD manifestations of Sjögren’s syndrome vary greatly and often go along with bronchial involvement [126]. The histology is often characterized by a follicular component, but lymphocytic interstitial pneumonia and, more rarely, a definitive UIP pattern and amyloidosis are also seen. No prospective therapy studies focusing on Sjögren’s-associated ILD were found [125]. Patients with Sjögren’s-associated ILD are commonly subsumed under collagen vascular disease-associated ILDs.

The spectrum of myositis disorders includes a rare group of heterogeneous conditions with varied prevalence of ILD. Approximately 20–40% of myositis patients develop ILD, and in many cases, this is the first manifestation of the disease [127]. Some types of MA-ILD do not go along with clinically relevant myositis. Specific autoantibodies are often associated with ILD, in particular anti-MDA5, anti-SSA and anti-aminoacyl-tRNA synthetase (“anti-synthetase antibodies”: Jo-1, PL-7, PL-12, EJ, OJ, KS, YRS [HA], Zo). MDA-5-associated ILD is a rare disease and has a significantly faster progression and higher mortality rate than the other MA-ILDs [128]. In MA-ILD, the inflammatory component seems to be more significant than the fibrotic component, and anti-inflammatory therapy therefore plays an important role.

No randomized controlled trials are currently available on the management of ILD in the context of immune-mediated Myositis. The therapy is based on case studies, case-control studies and (retrospective) cohort studies, and can vary across institutions. An important cornerstone in the management of MA-ILD is the close cooperation between rheumatologists and respiratory physicians. GC is often used as monotherapy or in combination with CYC, AZA, MMF, calcineurin inhibitors (tacrolimus [TAC] as well as cyclosporin A [CSA]) and RTX.

In an open-label phase II study, intravenous CYC was associated with improvement in many patients with MA-ILD [129]. A retrospective study by Johns Hopkins University investigated MMF and AZA in 66 patients. Both agents had a positive effect on lung function [130]. In a US cohort, most of the 54 patients with MA-ILD stabilized under GC + conventional DMARDS (AZA, MTX or MMF). Non-responders were treated with TAC, and most of them responded to the therapy [131]. A prospective, multicenter study demonstrated the superiority of initial GC + TAC combination therapy over (historical) GC monotherapy in terms of early mortality at 52 weeks and FVC in patients with dermatomyositis/polymyositis [131]. A randomized, prospective, controlled study of 85 MA-ILD patients investigated GC + TAC versus GC + CSA therapy. The primary endpoint was progression-free survival, where the TAC regimen proved to be slightly superior to CSA. FVC improved in both groups [132]. A combination therapy of GC + TAC + i.v. CYC had a positive effect on lung function in 29 patients with MDA5-associated MA-ILD in a prospective, multicenter trial [133]. Likewise, the JAK inhibitor tofacitinib had a positive effect in 18 patients with MDA5-positive ILD [134]. In 5 patients with MDA-5-associated ILD who were refractory with GC+CYC+CSA treatment, tofacitinib was used as add-on therapy. Improved survival was seen, with a higher incidence of infections than in the historical controls [135]. It is important to note that the number of subjects was small. RTX also showed promise for the treatment of MDA5-positive ILD in a meta-analysis [135]. Moreover, RTX led to an improvement or stabilization of ILD in 9 out of 10 patients with anti-synthetase auto-antibody-positive ILD in a prospective setting [136].

No prospective randomized controlled trials on the treatment of MA-ILD were found. Nevertheless, some groups report positive effects of various agents on MA-ILD, suggesting that anti-inflammatory therapy should be used to treat these patients. Treatment typically involves a combination of several drugs. Currently available data are insufficient to serve as basis for a definite recommendation regarding the choice of drug (or drug class).

Pulmonary involvement is common in systemic sclerosis (SSc), where it is the leading cause of death [137, 138]. Excellent data are therefore available on anti-inflammatory therapy for the treatment of SSc-ILD, without, however, considering the PF-ILD phenotype. In 2006, Hoyles et al. [139] demonstrated improved lung function under intravenous CYC pulse therapy followed by AZA in 45 patients with SSc-ILD versus placebo (treatment period: 1 year), just short of being statistically significant due to insufficient power.

In 2006, another randomized placebo-controlled multicenter study (scleroderma lung study SLS I) demonstrated the efficacy of oral CYC versus placebo in 158 SSc patients. CYC showed a moderate improvement in FVC and TLC over a 1-year treatment period, in addition to other improvements such as improved skin scores and chest CT findings [140]. After discontinuing the CYC therapy after 1 year, the improvement in FVC was lost during follow-up by the end of the second year [138]. In the subsequent SLS II study, the same group compared 12 months of oral CYC versus MMF (target dose 2 × 1.5 g/day) over a 24-month period in 142 patients [141]. A comparable positive effect of both therapies was seen in this study; the therapy improved the FVC in almost 70% of patients, while MMF was better tolerated. It is important to note that only patients with confirmed ground-glass opacities were randomized in both studies (SLS I and SLS II). The SENSCIS study, in which approximately half of the patients enrolled were treated with MMF, also suggests the efficacy of MMF in the treatment of SSc-ILD [142].

The double-blind, placebo-controlled phase II + III studies of the interleukin-6 antagonist tocilizumab [143, 144] could not show an effect of tocilizumab with regard to the primary endpoint (skin). However, the differences versus placebo in the secondary endpoints regarding the lung of the phase III trial were so convincing that tocilizumab was approved by the FDA for the treatment of severe SSc-ILD in 2021.

In 2021, Ebata et al. [145] presented the data of a randomized, double-blind, placebo-controlled study of RTX (4 × 375 mg/m²) in 56 SSc patients in Japan. Even though the primary endpoint was skin-related, the secondary endpoints showed a significant improvement in FVC versus placebo. As early as 2018, a phase II study from India compared i.v. RTX with i.v. CYC in 60 SSc patients with skin and lung involvement, demonstrating significant superiority of RTX with regard to FsVC after 6 months [146]. Data from the RECITAL trial (NCT01862926), in which patients with CTD-ILD were randomized to RTX versus CYC, are pending.

For SSc patients with a particularly rapid progression in the first 5 years of the disease, a total of 3 randomized, controlled studies showed significant superiority of autologous stem cell transplantation (HSCT) versus CYC pulse therapy in terms of overall survival, although CYC was only given for 12 months [147‒149]. The positive effect on the lungs was confirmed in a prospective non-interventional real-life observational study [150].

ILD in mixed connective tissue disease is similar to SSc-ILD and therefore not discussed separately.

This recommendation is based on the latest international guideline on “Idiopathic Pulmonary Fibrosis (an Update) and Progressive Pulmonary Fibrosis in Adults: An Official ATS/ERS/JRS/ALAT Clinical Practice Guideline” developed by international experts and approved via consensus voting [2]. The German Guideline Committee supplemented the systematic literature search on which the ATS/ERS/JRS/ALAT is based on literature published since the guideline was drafted. Hence, no other original treatment studies were published. A systematic review including meta-analysis covered controlled randomized trials on adults with IPF or non-IPF-ILDs. These trials evaluated pirfenidone or nintedanib and included mortality and lung function data (FVC) [151]. The analysis of 13 studies showed similar, significant effects of antifibrotic therapy in IPF and non-IPF PF-ILD on lung function; pooled data also showed a significant effect on mortality, which was, however, not significant in the analysis of the non-IPF studies (risk ratio 0.908 [0.547–1.508], p = 0.71). A small (n = 21) retrospective analysis supported the use of nintedanib in progressive pleuroparenchymal fibroelastosis [152]. In addition, several published sub-analyses of the INBUILD study support the previous data [153‒155]. These include an investigation into the combination of immunomodulators (prednisone ≤20 mg/day) and nintedanib versus placebo in connection with the INBUILD study [102] and showed that the effect of nintedanib on FVC decline and the side effect profile were not affected by the concomitant anti-inflammatory therapy.

Nintedanib is an antifibrotic agent, which – much like pirfenidone – delays disease progression in IPF patients. Nintedanib is an oral tyrosine kinase inhibitor and blocks key pathways of fibrogenesis. This therapy has been recommended for the treatment of IPF in previous guidelines [156].

The question “Shall PPF patients be treated with nintedanib?” was also examined for 8 different types of ILD with a PPF phenotype. Critical outcome measures included mortality and disease progression (change in FVC); important outcome measures included symptoms (changes in the King’s Brief Interstitial Lung Disease Questionnaire [K-BILD]) as well as acute exacerbations.

The systematic review that informed the recommendation provided in the international guideline was published separately [12]. In summary, the systematic review identified one randomized trial [11] and one post hoc analysis of the study [157]. In this trial (INBUILD), 663 PPF patients were randomized to receive nintedanib or placebo for 52 weeks, while the post hoc analysis compared the effects of nintedanib versus placebo in different ILD subtypes versus PPF. The ILD subtype was determined by the study sites with no prespecified diagnostic criteria having been provided to investigators and without a central review process. It is therefore possible that the diagnostic process varied across institutions.

Across all PPF patients, FVC declined in both the nintedanib and the placebo arm of the INBUILD trial, but the mean annual decline was significantly less (107 mL) in the nintedanib arm. The trial described “progression of ILD” as an adverse event, without defining it in this context; however, nintedanib lowered the risk of such progression by a factor of 2.4. The difference in the annual decline in FVC between the nintedanib and the placebo arm was 128 mL/year in patients with a radiological UIP pattern, and 75.3 mL/year in patients with no radiological UIP pattern [11]. Nintedanib lowered the risk of progression, measured as an adverse event, by a factor of 2.3 in patients with a radiological UIP pattern, while no significant difference was seen in patients without a radiological UIP pattern [11]. The effect of nintedanib versus placebo on reducing the rate of FVC decline (mL/year) was consistent across the five subgroups by ILD diagnosis in the overall population (HP 73.1 [95% CI: −8.6–154.8]; autoimmune-associated ILD 104.0 [21.1–186.9]; idiopathic NSIP 141.6 [46.0–237.2]; unclassifiable idiopathic interstitial pneumonia 68.3 [−31.4–168.1]; and other ILDs 197.1 [77.6–316.7]; p = 0.41 for treatment by subgroup by time interaction). There was also no significant difference in terms of progression in the context of an adverse event for any type of ILD. It should be noted that the estimates are based on small subgroup sizes: autoimmune ILD (n = 147), idiopathic NSIP (n = 125), fibrotic HP (n = 173), other exposure-related ILD (n = 39), sarcoidosis (n = 12), non-classifiable ILD (n = 114), others (n = 53) [157].

The INBUILD trial found no significant difference in all-cause mortality with regard to the radiological pattern of the PPF [11].

Observed adverse events were consistent with the data reported in the IPF studies.

The general quality of evidence was rated as low. The rating considers the lowest evidence quality rating of the two critical outcomes; the quality of evidence was moderate for disease progression but low for mortality because it was a single randomized trial and the number of events was small. The decision of the international of the ATS/ERS/JRS/ALAT Guideline 2022 Committee in favor of a conditional recommendation for nintedanib in PPF patients was based on two main reasons:

• 1.

  a statistically significant delay of disease progression, measured as the annual FVC decline, and

• 2.

  reversibility of AEs after discontinuation of the therapy.

Further data from the INBUILD study that have since been published also show a significant reduction in acute exacerbations [154], and systematic reviews including meta-analysis are available, which consistently reached a favorable assessment [12, 151]. From the German Guideline Committee’s point of view, a strong recommendation is therefore justified.

It should be noted, however, that the effects of the therapy may vary depending on the type of underlying ILD and that the management approach may be based on the underlying ILD in the future. Currently available data are, however, insufficient to support a targeted approach. The International Guideline Committee therefore recommends research into the efficacy, effectiveness, and adverse effects of nintedanib in patients with PF-ILD for the different ILD subtypes.

In addition to the pivotal trials on the effectiveness of pirfenidone in IPF patients [56, 57], pirfenidone has also been studied in PPF patients in two prospective randomized trials. The uILD study [14] included 253 patients with unclassifiable PPF who were treated with placebo or pirfenidone (2,403 mg/day) over a 24-week period. The primary study endpoint was the difference in FVC as measured by home spirometry, the secondary endpoints were the differences in FVC, DLco, and in the 6MWT as measured at the study site. The primary endpoint was negative as a meaningful analysis was not possible due to considerable variability and implausibility of patient-reported measured data, which were attributable, at least in part, to technical reasons. By contrast, a significantly smaller decline in FVC was seen when measured at the study site, along with a similarly smaller drop in DLco and the distance walked in the 6MWT. In the RELIEF study [13], 127 patients with PPF associated with fibrotic NSIP, cHP, autoimmune ILD or asbestos-related pulmonary fibrosis were included and randomized to receive pirfenidone or placebo for a period of 1 year. The study was terminated early due to slow recruitment, and study results should therefore be interpreted with caution. With regard to the primary study endpoint, the FVC decline seen in patients treated with pirfenidone was still significantly less than in those who received the placebo, and a consistently lesser decline in DLco and 6MWT distance was seen in the active arm. Another randomized prospective study to evaluate the safety and effectiveness of pirfenidone in patients with RA-associated ILD (TRAIL-1 study, NCT02808871) was completed, but only the study design has been published to date [158].

Two recent meta-analyses on the efficacy of antifibrotic therapies in IPF and PPF [151, 159] included studies with pirfenidone in addition to those with nintedanib. They show a significantly smaller FVC decline with an almost identical effect size (approximately −31%) for both IPF and PPF, as well as significantly reduced mortality.

Due to insufficient data, no application has yet been submitted for the approval of pirfenidone to treat PPF. However, data available to date suggest that the efficacy of pirfenidone for the treatment of PPF may be similar to that of nintedanib, although both pirfenidone studies have obvious flaws (endpoint, low recruitment). PPF patients should therefore preferably be treated with the approved standard of care (nintedanib). However, if this therapy is not sufficiently effective (e.g., unchanged FVC decline under treatment or >10% per year) or not tolerated, the Guideline Committee considers the available data on pirfenidone for the treatment of PPF as sufficiently positive to initiate a second-line, off-label treatment attempt with pirfenidone in these patients, after obtaining health insurance approval for such off-label use.

Potential occupational causes of ILD need to be carefully identified based on the patient’s occupational history, in particular with a view to eliminating ongoing harmful occupational factors and submitting the mandatory report in respect of an alleged occupational disease, which can help avoid risks for other potentially exposed persons.

The surprisingly high percentage of potential occupational causes of a seemingly “idiopathic” pulmonary fibrosis (26%) [160, 161] serves as a reminder for us to systematically investigate a patient’s occupational history. The “idiopathic” pulmonary fibrosis studies compiled by Blanc et al. [161] excluded asbestosis and silicosis populations, but other occupational causes of ILD had not led to differentiated diagnostic classifications other than IPF.

Given the high rate of occupational risk factors in IPF, two mechanisms are possible.

• 1.

  Failure to identify manifestations of classical pneumoconioses to a relevant extent, or/and

• 2.

  Occupational exposure to vapors/gas/dust/smoke can trigger the development of IPF without classic pneumoconiosis.

This raises the question of the extent to which this may have been caused by long-term “low-level” exposure or potential exposure peaks. The latter has not yet been addressed by the German Occupational Safety and Health Act. Action in this regard is necessary.

Park et al. [160] conducted a systematic review including a meta-analysis of case-control studies for occupational and environmental risk factors, looking to identify occupational risk factors of IPF [162]. Twelve studies out of 2,490 references were included. The following significantly increased odds ratios were identified:

• metal dust: OR = 1.83 (95% CI: 1.15–2.91)

• wood dust: OR = 1.62 (95% CI: 1.04–2.53)

• pesticide use: OR = 2.07 (95% CI: 1.24–3.45)

• working in agriculture: OR = 1.88 (95% CI: 1.17–3.04)

An additional factor is “smoker status”; OR = 1.39 (95% CI: 1.01–1.91). The systematic review with meta-analysis by Park et al. [160] did not include a major 2020 Australian case-control study: Abramson et al. [162] recruited 503 patients with IPF and 902 controls. They identified the following risk factors [163]:

• Positive family history for IPF: OR = 12.6 (95% CI: 6.52–24.2)

• Occupational exposure to second-hand smoke: OR = 2.1 (95% CI: 1.2–3.7)

• Occupational exposure to dust: OR = 1.38 (95% CI: 1.04–1.82)

• Occupational exposure to asbestos: OR = 1.57 (95% CI: 1.15–2.15)

“Have you ever smoked?” was included as a factor with an OR = 2.2 (95% CI: 1.74–2.70). The authors concluded that the burden of IPF could be reduced by intensified tobacco control, occupational dust control measures and elimination of asbestos at work [163].

(Please note: It is well-known that asbestosis can also exhibit a UIP pattern. In the absence of bridging phenomena such as pleural plaques and a history of relatively low asbestos exposure, it often presents a problem for the expert opinion if the level of cumulative asbestos exposure is not deemed sufficient for asbestosis while the presentation of a UIP pattern is not regarded as “typical of asbestosis”).

From a pragmatic point of view, nothing speaks against treating types of IPF manifesting after occupational exposure to dust that is not yet reflected in the German list of occupational diseases as IPF is not related to occupational exposure. These pathologies are also covered by the indications of pirfenidone (“treatment of mild to moderate IPF”) and nintedanib (“treatment of IPF,” “treatment of other progressive fibrotic ILDs”).

Nintedanib slows disease progression in patients with PPF other than IPF [11]. For pirfenidone, two phase-2 studies of unclassifiable pulmonary fibrosis and progressive non-IPF pulmonary fibrosis demonstrated that pirfenidone slows the rate of FVC decline [13, 14]. The study by Behr et al. [13] included 3 asbestosis patients.

In a 24-week phase-1 prospective observational study of 10 asbestosis patients with progressive lung function decline, Miedema et al. [163] investigated the safety and tolerability of pirfenidone 801 mg 3×/day. While home spirometry showed a decline in FVC in the 12 weeks before starting pirfenidone, it did not decline further during the 24-week treatment period, but the differences were not statistically significant [164].

In the INBUILD study, 332 patients with fibrotic non-IPF lung disease were treated with nintedanib over 16 ± 7 months and compared to a control group of 331 patients [11]. Both in the overall cohort and in the UIP pattern subgroup (n = 206 patients treated with nintedanib and n = 206 patients treated with placebo), the rate of FVC decline was significantly lower in those treated with nintedanib. The rate of pneumoconioses is not explicitly stated.

Bonella et al. [164] analyzed the effects of nintedanib in fibrotic lung disease in a meta-analysis of four placebo-controlled phase III trials: INPULSIS-1 and -2 for IPF, SENSCIS for SSc-ILD, and the aforementioned INBUILD study. The authors showed that treatment effects did not vary significantly across disease types. Even if this meta-analysis does not explicitly report or separately analyze data for occupational fibrotic lung diseases, it seems reasonable to apply the same conclusion also to “occupational IPF” and pneumoconiosis [165].

Zeng et al. [165] analyzed eight controlled studies with a total of 292 pneumoconiosis patients who were followed for at least 1 year with regard to functional capacity after whole-lung lavage [166]. Despite a search across all languages, only studies written in Chinese were identified. Most of the pneumoconioses were coal miners’ pneumoconioses, i.e., mixed pneumoconioses related to exposure to dust. Only two studies were ranked as being of good quality. The description of the quality criteria with regard to lung function findings was largely inadequate. The mean decline in FEV1 and FVC across all studies was less in the intervention group than in the control group after two and 4 years and in 1 study after 6 years [165]. The longer the follow-up, the greater the heterogeneity across studies. Taken together, the problematic quality of the studies makes it difficult to draw valid conclusions. The extent to which “sub-acute” silicosis in the form of alveolar proteinosis may have been present is not evident.

The additional development of precapillary PH has a serious impact on the prognosis of IPF patients. Several studies have shown that the development of PH significantly reduces the exercise tolerance and survival rates of IPF patients [167, 168]. Moreover, previous studies showed that inhaled prostaglandin I₂ (PGI₂), an established and approved treatment option for PAH, plays a role in lowering pulmonary vascular pressure or resistance in ILD patients with PH without triggering or amplifying a ventilation-to-perfusion mismatch. There was no further deterioration in the already impaired gas exchange in this patient group [169]. On the contrary, inhalation therapy may even improve the ventilation-to-perfusion ratio as it primarily reaches the well-ventilated areas of the lungs where it improves perfusion.

Based on this preparatory work, the INCREASE study, a prospective randomized trial published in 2021, randomized 326 patients with ILD and FVC <70% predicted as well as precapillary PH, as documented by right heart catheterization, to receive either inhaled treprostinil, a synthetic PGI₂ analog, 72 µg 4×/day 7 or placebo over a period of 16 weeks [170]. The primary endpoint was the 6MWT distance; secondary endpoints were NT-proBNP levels, clinical deterioration rate, as well as lung function parameters such as FVC. Relevant parameters such as age and sex, percentage of idiopathic interstitial pneumonia or IPF, or background therapy with antifibrotics at baseline were comparable in both groups. Treatment with treprostinil was associated with a statistically highly significant increase in the 6MWT distance versus the placebo arm. The difference in walking distance between the active and the placebo arm was 31.12 m and hence above the minimal clinically important difference of 28 m for IPF patients [171]. A simultaneous drop in proBNP levels and fewer exacerbations of the underlying lung disease were also observed as secondary endpoints. Interestingly, a significant increase in FVC percent predicted versus placebo was seen in a post hoc analysis of the study. This effect was most pronounced in the IPF patient group. In view of the diverse, positive extravascular effects of prostacyclin on the epithelial [172] as well as on the mesenchymal [173, 174] compartment, it is hypothesized that inhaled prostacyclin may also unfold antifibrotic properties in patients with fibrotic lung diseases.

The study by Waxman [170] is the first to demonstrate a clinically relevant effect of a pulmonary vasoactive agent – in this case, the PGI₂ analog treprostinil – in PH patients with relevant ILD. For further guidance please refer to the recent PH guideline generated jointly by the European Society of Cardiology (ESC) and European Respiratory Society (ERS) [175].

The following recommendations address common clinically relevant questions for which sufficient scientific evidence is not available to date. The following statements reflect the opinion of the members of the Guideline Committee.

The consensus-based recommendations of the 2017 S2k guideline on “Pharmacological treatment of IPF (update)” advise prompt antifibrotic therapy of symptomatic patients with a definite IPF diagnosis from the time of diagnosis. This recommendation to start treatment at the earliest possible time remains unchanged. Therapy initiation at the time of IPF diagnosis aims to slow disease progression as early as possible and improve the prognosis. It is important to note that in the placebo arms of the clinical trials patients with “normal” baseline FVC exhibited a similar magnitude of FVC decline as compared to patients with reduced FVC at baseline. The FVC decline also occurs after only a short observation period. A study in which only IPF patients with FVC >80% participated saw a decline in FVC of 70 mL in the placebo group after just 12 weeks [176]. Hence, antifibrotic therapy should be initiated, even if the lung function is within reference range [25, 27, 62, 72, 177].

In special cases (e.g., incidental finding during CT examination or lung resection for other indication) of non-symptomatic patients with no or only minimal limitation of the pulmonary function. A wait-and-watch approach concerning the initiation of therapy can initially be adopted. Also, prognosis-limiting concomitant diseases (e.g., lung cancer) may be a reason not to initiate an antifibrotic therapy. In any event, the individual therapeutic approach has to be discussed with the patient in an open and understandable manner. Even if a decision is made against an immediate initiation of therapy, the patient should always be subjected to continuous clinical monitoring and control of pulmonary function (at least every 3 months).

The PPF phenotype includes a wide range of ILD diagnoses. Most patients with a non-IPF diagnosis do not progress and respond to anti-inflammatory therapy or avoidance of exposure, etc. [7, 178]).

Disease progression is therefore usually closely monitored and/or an anti-inflammatory therapy is initiated at the time of diagnosis, and, in the case of HP and other exposure-associated conditions (exposure to metal dust, medication, etc.), avoidance of the causative noxae is recommended. Only if, in spite of such measures, the development of the disease is classified as PPF according to the progression criteria defined in the preamble, antifibrotic therapy shall be initiated regardless of HRCT pattern [16, 178]. A retrospective analysis of the disease course may also be conducted.

The question of whether patients with prognostic factors for unfavorable outcomes, e.g., UIP pattern, should be treated with an antifibrotic drug from the time of diagnosis remains unresolved [179, 180]. In the INBUILD study, the FVC decline in patients with UIP pattern tended to be greater in the placebo group and the treatment effect of nintedanib better than in patients with other fibrotic HRCT patterns [11]. For more information on the selection of the antifibrotic therapy for PPF, see chapter 2.7 “Shall PPF Patients Be Treated with Antifibrotic Therapy?” and chapter 4.3 “When Should Switching Antifibrotic Drugs Be Recommended?”

In principle, treatment failure or non-tolerable side effects may justify a discontinuation of therapy. There is no standard definition for treatment failure. An acute exacerbation or other acute respiratory deterioration is no reason to discontinue the therapy as post hoc analyses have shown that continuing the antifibrotic treatment after an acute event is associated with a survival benefit [33, 64]. Even patients with severe IPF still benefit from antifibrotic therapy [35, 63, 67]. Therefore, an advanced disease stage is also no reason to discontinue the therapy.

There are only limited data on the efficacy of antifibrotic treatment over a period of more than 52–72 weeks, i.e., beyond the treatment period in the placebo-controlled pivotal studies. Additional tolerability and safety data from long-term observational studies and real-life cohorts are now available for both antifibrotic drugs. The open-label extension of the ASCEND and CAPACITY (RECAP) trials (mean treatment period 122 weeks, maximum treatment period 349 weeks) and of the phase IV study (PASSPORT) (real-world IPF patients) revealed no new aspects for pirfenidone in terms of long-term tolerability and safety profile [60, 69]. In addition, long-term analyses of 1,299 patients from 5 clinical trials indicate that pirfenidone is also well tolerated over an extended period of up to 9.9 years (median: 1.7 years) [61]. The open-label extension of TOMORROW (nintedanib) demonstrated that the efficacy of nintedanib and the tolerability was maintained over 52 weeks (median treatment period 28 months, maximum treatment period 86 months) [181]. The long-term observational study of nintedanib (INPULSIS-ON) confirmed the positive effect on disease progression (mean treatment period 45 weeks, maximum treatment period 68 months) with a consistent side effect profile [24]. Real-life data from Medicare patients in the USA, from the German INSIGHTS-IPF registry and the European IPF registry, suggest that antifibrotic treatment can improve survival [47‒49].

Unlike IPF, no robust data from long-term follow-up studies, registry studies, or real-world populations are available for the treatment of PPF with antifibrotic therapy. Unlike IPF, however, the rate of progression in PPF patients is often known before starting the therapy. This information can therefore be considered in deciding on the discontinuation of therapy due to treatment failure. If the treatment seems to have no beneficial effect on disease progression over a treatment period of 6–12 months, the termination of the therapy may be considered, after switching antifibrotic drugs, as applicable (see chapter 4.3 “When Should Switching Antifibrotic Drugs Be Recommended?”).

IPF patient cohorts who switched antifibrotic drugs were analyzed in several retrospective studies [38, 40, 182‒184]. A retrospective multicenter study analyzed 62 IPF patients treated with nintedanib, of which 48 patients had previously been treated with pirfenidone. The reasons for switching were pirfenidone intolerance (56%) or disease progression (44%). In this analysis, progression was defined as FVC decline (% pred) >5% within 6 months or FVC decline (% pred) <5% with simultaneous worsening of symptoms or radiological findings [38]. Suzuki et al. [182] analyzed a collective of 262 IPF patients treated with antifibrotics. Thirty-seven patients (14%) switched the therapy (pirfenidone to nintedanib n = 29; nintedanib to pirfenidone n = 8) (46% due to progression, 46% due to side effects of antifibrotics, 8% due to first diagnosis of lung cancer, vasospastic angina, or at the patient’s explicit request). In patients classified as exhibiting progression, the median FVC decline (% predicted) was 14.0% (5.1–22.2) within a median period of 29.9 (24.2–41.6) months before being started on the second-line antifibrotic therapy. Brunnemer et al. [40] studied 64 IPF patients treated with nintedanib. Thirty patients had previously been treated with pirfenidone; the reasons for switching to nintedanib were adverse events (70%) and disease progression of IPF under pirfenidone (30%). Progression criteria were FVC decline (% pred) ≥5% and/or deterioration of DLco ≥15% at any time. A smaller case series of 12 patients, who switched from pirfenidone to nintedanib, yielded similar results (side effects n = 9; progression n = 3) [183].

It is not entirely clear from the above-mentioned studies whether patients progressing on antifibrotic therapy might generally benefit from a switch to an alternative antifibrotic drug. A recently published Canadian registry study assessed 165 IPF patients who had experienced an FVC decline ≥10% within 6 months under antifibrotic therapy. There was no difference in terms of a further lung function decline between those patients whose antifibrotic treatment was switched and those whose antifibrotic treatment was continued. However, patients whose antifibrotic therapy was definitely terminated had a significantly poorer prognosis [185]. Post hoc subgroup analyses from the pivotal studies had already demonstrated for both antifibrotic drugs that a consistent continuation of the therapy improves the prognosis versus placebo, even if there is a further FVC decline after therapy initiation [186, 187].

There are no published data on switching antifibrotic therapy in PPF patients. It remains unclear if the criteria for disease progression in PPF (see preamble) might also constitute an indication to switch the antifibrotic therapy. If progression does not slowdown in spite of antifibrotic treatment, a therapy switch or the option of participating in a clinical treatment trial should be considered.

As with IPF, a change of therapy or the option to participate in a clinical trial should be considered, if side effects cannot be controlled despite symptomatic therapy, dose reductions or temporary treatment interruptions. The respective decision should be made in the context of an ILD MDD (ILD Board). For information on the optimized management of the side effects of antifibrotic therapy, see the relevant literature on this topic [188‒190].

Due to the availability of two approved antifibrotic drugs, the question arises if, similar to other diseases (e.g., bronchial asthma, COPD, PAH, or lung carcinoma), a combination therapy might be an even more effective treatment approach. The multifactorial and heterogeneous pathogenesis of IPF and the resulting requirement for a drug therapy to target various pro-fibrotic signaling pathways is a rationale to consider combination therapy [191]. However, we do not yet know if the combination of two effective active drugs will result in a synergistic or additive effect or if drug-drug antagonism may emerge, in which the two drugs render each other less potent. A precedence case is the PANORAMA trial in which the combination of pirfenidone + N-acetylcysteine versus pirfenidone + placebo was investigated, in the context of which not only no positive effect but possibly even a negative effect of using acetylcysteine as add-on to pirfenidone was encountered [85]. Until this day, unexpected drug interactions with unacceptable side effects cannot be ruled out. In a Japanese phase-2 trial, the profile of side effects, tolerability and pharmacokinetics of nintedanib alone and in combination with pirfenidone was investigated in IPF patients. The study revealed a reduction of the maximum plasma levels and the AUC of nintedanib as well as more frequently occurring side effects in the case of co-medication with pirfenidone compared with nintedanib monotherapy [192]. The open-label, randomized INJOURNEY trial primarily investigated the tolerability of pirfenidone as add-on to stable nintedanib treatment [193]. The combination therapy was associated with an increased rate of gastrointestinal side effects, and the plasma concentration of nintedanib alone was similar to that of nintedanib in combination with pirfenidone. The exploratory efficacy analysis showed at least a signal in terms of a smaller FVC decline in the combination arm over the short 12-week study period (D FVC −13.3 ± 17.4 mL vs. −40.9 ± 31.4 mL). In an open-label, single-arm phase IV study over 24 weeks, the administration of nintedanib in addition to stable therapy with pirfenidone was assessed in terms of tolerability and side effect profile [194]. The combination therapy was tolerated by the majority of subjects (73/89) during the follow-up period, and the side effect profile was similar to that of the respective drugs when used as monotherapy. A small retrospective observational study in Japan analyzed the tolerability of nintedanib or pirfenidone as add-on to a stable antifibrotic therapy. 30.4% of patients discontinued treatment with either one of the two drugs or with both drugs [195]. Controlled trials will be needed to demonstrate that a combination therapy would be more effective than the respective monotherapy while having a similar safety profile, before a pirfenidone + nintedanib combination therapy can be recommended for IPF patients.

Based on the data of the INBUILD study [11], nintedanib is currently the only approved antifibrotic treatment for PPF. The results of the German RELIEF trial [13] also demonstrate the effectiveness of pirfenidone in the treatment of PPF, but the drug is not approved for this indication. There are no published data on the efficacy of nintedanib-pirfenidone combination therapy in PPF. As with IPF patients, an increased rate of side effects should be expected. Controlled trials will be necessary to demonstrate that a combination therapy would be more effective than the respective monotherapy while having a similar safety profile before a pirfenidone and nintedanib combination therapy can be recommended for PPF patients.

Keeping in mind the absence of a head-to-head comparison between nintedanib and pirfenidone, the currently available literature seems not to imply a relevant superiority of either of the two active drugs over the other, so that no recommendation can be made regarding a preferential use. A comparison of the two substances is complicated by the diverse patient cohorts in the respective pivotal trials resulting from the diverse inclusion and exclusion criteria in the CAPACITY, ASCEND, and INPULSIS trials [22, 56, 57]. In Germany, nintedanib has been approved for all IPF patients without consideration of the level of severity, whereas pirfenidone has been approved only for the mild and moderate IPF. As no objective definition exists for the level of severity, the judgment of severity is up to the treating physician – ideally the multidisciplinary team – in consideration of the clinical symptoms, functional limitations, radiomorphological picture, and comorbidities [196]. Meanwhile, a post hoc analysis of the CAPACITY and ASCEND phase III trial data for pirfenidone in patients with more severe functional pulmonary impairment (FVC <50% predicted or DLco <35% predicted) [63] has been published. Pirfenidone was significantly superior to placebo for the endpoints of all-cause mortality, FVC decline ≥10% predicted or all-cause mortality, and FVC decline ≥10% predicted or all-cause mortality or respiratory-related hospitalization. The premature discontinuation rate due to side effects was not increased. A Korean observational study prospectively investigated the efficacy and tolerability of pirfenidone in 219 patients, 18% of whom had advanced IPF, defined as FVC <50% predicted or DLco <35% predicted [197]. The incidence of side effects was similar, but the efficacy remained the same. In a retrospective analysis from Korea, pirfenidone was shown to have a similar efficacy and side effect profile in patients with advanced IPF (defined as FVC <50% predicted or DLco <30% predicted) as in those with non-advanced IPF [198]. In the prospective controlled SP-IPF study, patients with advanced IPF and risk of PH were treated with pirfenidone and randomized to receive add-on sildenafil 3 × 20 mg/day or placebo. This study confirmed the known side effect profile of pirfenidone, no new aspects were identified, in particular, no increased frequency or intensity [99]. Overall, the available study results suggest that pirfenidone is also effective for the treatment of advanced IPF and that its side effect profile is comparable to that observed in mild to moderate IPF. In the meantime based on pooled analysis of six randomized controlled trials pirfenidone has been approved by the European Medicinal Agency (EMA) for the treatment of IPF without severity limitations [200].

There is no definite answer to the question of when an anti-inflammatory therapy should be discontinued in non-autoimmune or autoimmune PPF. It is a complex decision. A variety of aspects need to be considered and the decision must be made on a case-by-case basis. Infection-related complications and/or increase in fibrotic changes, and the development of honeycombing and traction bronchiectasis in particular are important criteria for reducing or even discontinuing of anti-inflammatory therapy [178].

A placebo-controlled study in IPF patients demonstrated that anti-inflammatory combination therapy using prednisolone and AZA led to increased respiratory tract infections, acute exacerbations and poorer survival outcomes [199]. Especially, in patients with telomere lengths below the 10th percentile, the above immunosuppressive combination therapy was associated with a high infection-related complication rate and side effects [200]. The data from the INBUILD study show that in both treatment arms patients treated with anti-inflammatory drugs at baseline had more respiratory tract infections and bronchitis than those not receiving this therapy [102].

Treatment with anti-inflammatory agents should also be discontinued if side effects cannot be controlled despite symptomatic therapy, dose reductions, or temporary treatment interruptions. In addition, the activity of the underlying disease as well as extrapulmonary manifestations need to be considered in autoimmune PPF in order to control and, if necessary, terminate the anti-inflammatory therapy.

We thank Ms. Gunda Mundt for providing the first draft of the English translation of the guideline.

J.B. received honoraria for lectures, educational events, and consulting from AstraZeneca, Boehringer Ingelheim, BMS, Ferrer, Gossamer Bio, Novartis, Pulmovant, Sanofi-Genzyme, and United Therapeutics.

F.B. received consulting fees from Boehringer Ingelheim, Sanofi, Savara Pharma, and CSL Behring; lecture honoraria from Boehringer Ingelheim and Sanofi; travel support from Boehringer Ingelheim, AstraZeneca, and Atyr; and advisory board participation with Boehringer Ingelheim, Sanofi, and GSK, outside the submitted work.

B.C.F. received research support from Bristol Myer Squibb and Relief Therapeutics unrelated to the manuscript; received consulting and lecture fees from Advita Lifescience GmbH, Actelion, AstraZeneca, Boehringer Ingelheim, Novartis, Roche, and Vifor; received travel support from Boehringer Ingelheim; indicates the following intellectual property: WO2020225246A1; WO2021152119A1; and had shares of Relief Therapeutics in 2021.

A.G. reports grants, lecture payments, and/or consulting fees from Boehringer Ingelheim, Roche, Lung Therapeutics, and Pieris.

L.H. received honoraria for lectures from GSK and AstraZeneca and for consultation from Boehringer Ingelheim and Pfizer.

J.H. received honoraria for lectures and consultation from ABBVIE, BMS, Boehringer Ingelheim, GSK, Johnsson & Johnsson, Novartis, Roche, and UCB.

P.K. received honoraria for lectures and consultation from AbbVie, AstraZeneca, BMS, Biogen, Boehringer Ingelheim, Galapagos, Lilly, medac, Mylan, Novartis, Pfizer, UCB, and Janssen.

D.K. received honoraria for lectures and consultation from Boehringer Ingelheim and Roche.

M.K. received honoraria for lectures and consultation from Boehringer Ingelheim, Roche, Galapagos, GSK, AstraZeneca, and BMS.

G.L. received honoraria for lectures and consultation from Boehringer Ingelheim and honoraria for scientific articles from Roche.

D.N. received honoraria for lectures and consultation from AstraZeneca, Berlin Chemie, Boehringer Ingelheim, Bristol Myers Squibb, Chiesi, GSK, Mundipharma, Novartis, Pfizer, and Sanofi Aventis.

A.P. reports consulting fees from BI, CSL Behring, and Novartis; fees for speaking from BI, Euroimmun, Gilead, and Novartis; and support for travel from BI and holds a leadership or fiduciary role as a coordinator of the ILD group of the European Reference Network-lung.

B.Q. and H.S. report no conflicts of interest.

U.C. received honoraria for consultation from CSL Behring and Boehringer Ingelheim.

The German Respiratory Society (DGP) provided funding for the travel expenses of the Guideline Committee members.

Jürgen Behr served as guideline coordinator and Helmut Sitter as moderator during the formal consensus process. All authors contributed to the guideline development, were involved in the consensus process, and provided edits. All authors had access to the final document and provided their approval.