Introduction

Lung involvement in the ANCA-associated vasculitides (granulomatosis with polyangiitis-GPA, microscopic polyangiitis-MPA, and eosinophilic granulomatosis with polyangiitis-EGPA) is frequent. Its manifestations often arise at diagnosis or subsequently throughout the disease course. Their treatment is mostly according to the presence of more common disease features (e.g., renal, cutaneous, general, neurological). Most recently, interstitial lung disease (ILD), which was thought to be a late complication, is observed not only at the time of diagnosis, but even as the culprit of precedent complaints. This disease pattern is being increasingly recognized, and there are diverse aspects which are not yet understood. Relevant to this article, important treatment gaps exist. Overall, ILD occurs in up to 85% of patients with MPA, with fibrosis occurring in up to a quarter of GPA patients, and in almost half of those affected by MPA. The pattern reflecting fibrosis is, for the most part, that known as usual interstitial pneumonia (UIP). As for nonspecific interstitial pneumonia (NSIP), there are less precise figures which probably reveal a lower frequency. In that case, the fibrotic NSIP subtype clearly prevails over the inflammatory one [1•].

This review will touch on some pathogenetic hypothesis of how ILD—specifically the two most common types: UIP and NSIP—may develop within the AAV, pinpoint on specific diagnostic items in the context of ILD, provide some of the imaging features of ILD with differences found between the two most common AAV (GPA and MPA), discuss the different scenarios regarding the interplay between antineutrophil cytoplasm autoantibodies (ANCA), especially those which target myeloperoxidase (MPO-ANCA) and ILD, and finally, provide some of the scarce data concerning treatment of this complication aside the most usual approach for the overall diseases.

ILD constitutes a complex condition bearing important unsolved needs, which go from unrevealing its pathogenesis to treatment challenges, as currently, no proven treatments are available for this condition in the context of an AAV.

The physiopathology of ILD in the AAV: what is known and some hypotheses

As in the majority of autoimmune diseases which affect multiple organs, complex mechanisms are known to operate. In the general physiopathology known to operate in the AAV, which leads to the majority of the clinical manifestations (renal, cutaneous, secondary to pulmonary capillaritis, or peripheral neuropathy), some of them are well-established with important roles being played by the genetic background of the host and the interplay with environmental aspects, among them those of infectious origin and some related to inorganic agents. This interaction leads to immune dysregulation, where loss of self-tolerance is key. Many actors may then operate to the physiopathological events which lead to disease.

The main cellular culprits are both B and T lymphocytes, antigen-presenting cells such as monocytes/macrophages and dendritic cells, endothelial cells, and a very relevant culprit: the neutrophils. Also, the complement system is known now to play an important and probably, very early role in the disease process although the levels of circulating components of this system do not reflect, as in other diseases, an apparent significant role as interpreted due to its apparent lack of activation.

In particular, regarding the development of ILD in the AAV and especially lung fibrosis (LF), the main hypothesis is that the latter may develop after repeated episodes of intra-alveolar hemorrhage [2, 3]. This has found support in histological evidence of acute or chronic hemorrhage in more than half of lung biopsies of ANCA patients in some previous series. Markers of chronic alveolar bleeding (hemosiderin-laden macrophages) were found increased in bronchoalveolar lavage (BAL) samples of patients with lung involvement associated with AAV, as opposed to other autoimmune diseases such as scleroderma or systemic lupus erythematosus [4, 5]. Nonetheless, many individuals did not have overt manifestations of diffuse pulmonary bleeding, indicating that subclinical episodes are overlooked [4, 6, 7]. In addition, some of the initial computed tomography (CT) images of consolidation (corresponding histologically to alveolar bleeding) were shown to correlate with honeycombing areas in the long-term follow-up of some of those patients [8].

In contrast, the physiopathology of the ILD of idiopathic origin involves actors not usually known to participate in the genesis of the AAV. Among them, very important are the epithelial cells lining the alveoli, and mesenchymal cells, such as fibrocytes, myofibrocytes, and fibroblasts. In fact, a process known as epithelial to mesenchymal transition (EMT) is essential for the development of one of the main categories of ILD, LF, which has a morphologic (both imaging and histopathological) pattern called UIP [9, 10].

These mechanisms are not known yet to operate in the ILD seen in AAV. However, ANCA, and especially MPO-ANCA, may play a relevant role in the pathogenesis of LF [7]. It has been shown that the activation of neutrophils by MPO-ANCA results in the production of oxidative species, one of them is hypochlorous acid, which was able to trigger fibroblast proliferation in vitro. In addition, MPO-ANCA may contribute to lung tissue harm after the local release of proteolytic enzymes from activated neutrophils [11]. Indeed, elastase, also contained in neutrophils, is able to induce pulmonary fibrosis in mice [12, 13]. Damage induced by eosinophils and neutrophil extracellular traps (NETs) could also play a role. In specimens of ANCA lung tissue bearing extensive interstitial fibrosis, tissue eosinophilia has been reported [4]. It is known that lung fibroblasts might influence the function and survival of eosinophils [14]. Interestingly, BAL eosinophilia has been suggested as a marker of LF [15], and it is known that an important role in tissue repair, which may lead to fibrosis, comes from relevant cytokines produced mainly by eosinophils, like interleukin-4 (IL-4) and especially, interleukin-5 (IL-5). The main source of the latter are eosinophils, which also are able to secrete transforming growth factor beta-1, the main actor in fibrosis [16]. This is of special interest in EGPA, where the frequency of LF is in fact, low, and in where undoubtedly, long-standing damage has occurred throughout many patients history, without the development of such complication.

Neutrophil extracellular traps (NETs) are peculiar structures which play an important physiological role in the defense against microbes. In essence, they arise from neutrophils which undergo a different death pathway and are an important part of the innate immunity. They are composed of deoxyribonucleic acid (DNA), histones, and enzymes which are released once the death process is set. Regarding their link with AAV, they have been shown in glomeruli of AAV patients, and they can be released by ANCA-activated neutrophils [17•]. Additionally, they have the ability to activate lung fibroblasts and promote their differentiation into myofibroblasts [18], which are cells importantly recruited towards fibrotic areas in some ANCA patient specimens [19]. It has been shown too that NETs can activate the alternative pathway of complement. The relevance of this route in AAV has been extensively demonstrated since the past decade.

As “not-to-forget” cofactors that may be of relevance, some lifestyles such as smoking, or additional common conditions such as the role of gastroesophageal reflux (still controversial) in the processes that may lead to parenchymal lung lesions, are important [20,21,22,23]. Also, some drugs can also elicit NETs formation, among them, and very relevant to the AAV is propylthiouracil [24].

It is interesting that in comparison with what is known in idiopathic pulmonary fibrosis (IPF), some key actors have not been demonstrated in the AAV. For example, epithelial apoptosis which arises from continued damage of the epithelium can lead to EMT in IPF. This has not been shown in AAV-related LF. In contrast, the role of neutrophils and the complement system in IPF seems negligible as opposed to what have been recently developed in the pathogenesis of the AAV [25].

From the abovementioned, it seems probable that a conjunction of factors known to operate in early stages of AAV leads to repeated episodes of endothelial injury and necrosis, with severe lung inflammation, and thereafter this severe inflammation may trigger a reactive fibrotic state, characteristic of LF [19, 26, 27]. It is interesting that the best-known animal models of AAV show practically no evidence of LF. This can be naturally due to a longer lapse needed for the initial lesions to evolve into a chronic scarring phase. This assumption cannot be easily transposed into humans, and into a clinical scenario where ILD appears early in the course of the AAV.

Figure 1 depicts schematically the mechanisms that may operate regarding LF development in the AAV. It considers what may occur from lung capillaritis in an acute phase to a chronic scarring process leading to LF. However, different physiopathological mechanisms could exist between IPF and the LF in AAV. It is then essential to pursue more knowledge in this area and to dissect such differences.

Fig. 1
figure 1

A. The development of the lung manifestations of an AAV most surely depends on several factors (genetic, environmental, infectious, related to immune dysregulation) which are different for each individual and determine the final pulmonary disease. B. Once under such influx, ANCA production is favored, which depends on the interaction of various T cell populations (Tregs, TH17) and also B cells (Bregs, CD19+ B cells). Their production is also influenced by initial neutrophil activation. These cells release mediators which increase B cells proliferation and inhibit their apoptosis, which leads to mature plasma cell development and subsequent ANCA production. ANCA also induce neutrophils to secrete chemoattractants which ultimately lead to the activation of the alternative pathway of complement, with the anaphylotoxin C5a as an important actor. An amplification loop is then established, with more neutrophils being led to the sites of vascular inflammation. It has also been shown that ANCA lead to neutrophil extracellular traps (NETs) formation which may have in such instances, deleterious effects. When these events occur in the pulmonary capillaries, the necrosis and destruction of them lead to the breakdown of the vessels, with red blood cells extravasation into the alveolar spaces. This translates into the clinical manifestation of diffuse alveolar hemorrhage (DAH) (depicted on the left side of the scheme). C. It has been suggested, on the other hand (right side of the picture) that lung fibrosis (LF) in AAV can be the result of various factors: (1) repeated and/or chronic bouts of lung hemorrhage. This may not have been evident historically though, as some patients develop LF before other AAV manifestations, either pulmonary or extrapulmonary; (2) MPO-ANCA induce liberation of oxidation products (such as hypochlorous acid) which may stimulate fibroblast proliferation; and (3) presence of continuous neutrophil activation with local release of proteolytic enzymes (e.g., elastase) that could enhance a fibrotic response.

Full size image

A defining feature of the AAV is of course, ANCA. This biomarker is undoubtedly helpful for the diagnosis of these conditions, but has not proven useful for prognosis or follow-up. In this context, the knowledge of the pathophysiology of the diseases can also aid in the search of new biomarkers aside from the usual ones (erythrocyte sedimentation rate (ESR), C-reactive protein). Within this objective, also the validation of the significance of others could be sought. For example, it has been found that circulating tenascin C levels correlated with the presence of lung infiltrates in AAV patients [28•]. Relevant to LF, CC-chemokine ligand 18 (CCL-18) in peripheral blood mononuclear cells were highly expressed in active AAV, as opposed to patients with lupus nephritis, IgA vasculitis, or infections. CCL-18 has been shown to increase the production of type I collagen from lung fibroblasts, which might be important for the generation of LF in AAV [29].

ILD: what to consider when the suspicion arises?

From a clinical standpoint, the signs and symptoms observed in ILD are nonspecific. A broad range of diseases with different etiologies needs consideration. They go from hereditary to infectious in origin, and often, as many are more common than an AAV, lead to delay in suspecting the presence of the latter. This hampers an expeditious diagnostic route.

Patients with ILD usually follow an indolent course where nonspecific symptoms and signs are reported. Major symptoms of patients with ANCA positive pulmonary fibrosis are progressive dyspnea (50–73%) and nonproductive cough (21–60%) [30,31,32, 33••]. Severe and more acute respiratory symptoms as pulmonary hemorrhage and hemoptysis are less frequent (5%). Fever may occur in up to a third of patients, and it is important to discard concurrent infections. If patients present with previous or concurrent nonrespiratory symptoms, reflecting that the AAV started prior to the development of ILD, then general symptoms can be present in up to 80%, and other accompanying extrapulmonary manifestations have been reported from 70 to 100% of cases [1•]. Some authors have reported that in the context of MPA, patients with LF had a less severe systemic inflammatory response, manifested as lower erythrocyte sedimentation rates (ESR), higher hemoglobin levels and a reduced frequency of clinically evident diffuse alveolar hemorrhage, peripheral nerve derangement, and kidney involvement [6, 34].

In the context of ILD, Table 1 shows which diseases are relevant to consider when one faces this broad pulmonary disease category, and a patient is suspected to have an AAV as responsible.

Table 1 Differential diagnosis of interstitial lung disease
Full size table

Being clinicopathological entities, the role of the lung biopsy in the AAV is certainly relevant. The best results regarding histopathological confirmation of an AAV are for nodular disease in GPA though, not for ILD. Many practical issues limit their obtainment. One of the most important is the time since disease manifestations started, mainly dyspnea and cough. By that time, patients can be already so ill due to a much compromised respiratory function, that the procedure, especially if obtained through open surgery, may represent more hazard than a benefit. Regarding other modes of tissue obtainment, like transbronchial or guided by CT, the amount of the samples may be often scarce. Usually, transbronchial biopsies do not yield enough material for adequate processing and interpretation. The role of cryobiopsies in ILD is still under evaluation [35, 36], but a recent study has found, in 21 patients, poor concordance between sequential transbronchial lung cryobiopsies and those obtained surgically [37••]. A surgical lung biopsy is at present, and analogous to what occurs in ILF, not essential if a definite UIP pattern is observed in HRCT [38]. Video-assisted thoracic surgery may at times be an alternative to open thoracotomy, but this relies on the surgical team expertise. Although the yield of the samples obtained is similar, complications can still occur in similar numbers. In conclusion, the decision to perform an open lung biopsy must rely on a patient-to-patient basis and take into consideration if its performance surpasses its risk.

Imaging

Imaging plays an essential role in the diagnosis of ILD. In fact, many entities of ILD are defined according to findings in high-resolution computed tomography (HRCT) of the lungs, and this is especially important in IPF [39•], and probably by extension to other forms of LF. It is well established there are histopathological correlations of ILD with tomographic patterns [10]. This is avoided in many patients’ confirmatory biopsy obtainment.

In accordance with the American Thoracic Society/European Respiratory Society (ATS/ERS) International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias criteria [10, 40], UIP is the most frequent radiological pattern (50–57%), followed by NSIP (7–31%), and finally, desquamative interstitial pneumonia (14%) [41,42,43,44]. Regarding the two most common, in UIP (mainly observed in MPA with up to 60% of patients with this AAV having developed ILD in some series), some frequent HRCT features are reticular opacities (40–77%), often accompanied by bronchiectases (32–38%) [42, 45•]. Honeycombing, an essential component of the UIP definition, is present in 25–52% of patients. As in IPF, the distribution of the lesions is mostly in the periphery of the lower lobes [8, 41, 43, 46]. Ground glass opacities (GGO) are also present in UIP, from a quarter to even more than 90% of patients, and they correspond histopathologically with interstitial chronic inflammation, alveolar hemorrhage, and areas of fibrosis. Regarding NSIP, reticular shadowing (41–77%), interlobular septal thickening (41–71%), consolidations (23–78%), and honeycombing (23–52%) can be seen [8, 42,43,44, 45•, 46].

With respect to distribution, the interstitial involvement was reported to be symmetrical in 50–100% of patients, affecting predominantly the lung periphery and lower areas. In one study, the interstitial disease affected > 40% of pulmonary parenchyma in more than 60% of patients [42]. Airway abnormalities are reported: bronchiolitis (55%), bronchial wall thickening (44%), or bronchiectases (32–38%). Combined pulmonary fibrosis with emphysema has been recently reported in MPA patients [44, 45•, 47, 48]. Importantly, 4–40% of studied cases will not fit any specific CT-pattern [6, 41,42,43,44]. During follow-up of treated patients, some of the interstitial changes may resolve, i.e., GGO, reticular pattern, interlobular septal thickening, and consolidations [45•]. Honeycombing, which is usually considered a late-stage finding indicating interstitial fibrosis, often progresses or at least, does not change [8, 42, 49, 50]. These data suggest that in certain subjects, both processes, inflammation and fibrosis, may coexist in different proportions [45•].

In cases with MPO-ANCA positive ILD but without generalized involvement, UIP is also the most common abnormal pattern [30, 50,51,52,53,54]. Regarding anti-proteinase-3 (PR3-ANCA) positive LF, a small series reported UIP in 38% and NSIP in 31% [55•]. In these patients, CT can show, in decreasing order, honeycombing (37–100%), GGO (25–100%), consolidations (0–78%), nodules (0–45%), cyst (27%), traction bronchiectasis (80–100%), thickening of bronchovascular bundles (51%), or interlobular septal thickening (45%) [8, 30,31,32, 54]. Fibrotic changes were usually extensive, subpleural, peripheral, and located in the lower lung fields [30, 31]. It must be said that in these articles, no significant differences in CT images were found between ANCA-positive and ANCA-negative LF [30, 33••, 54].

Recently, Mohammad et al. [56•] studied the imaging findings in accordance with the serotype (MPO-ANCA vs. PR3-ANCA). Significantly, features of ILD and, once again, UIP were more frequent in MPO-ANCA positive. Adding to data that may help to find distinct features between patients with MPA and ILD versus those exclusively MPO-ANCA positive with ILD, we have preliminary data that found some differences (unpublished observations). Lymph node enlargement, septal thickening, and bronchiectases predominated in MPO-ANCA-positive patients preliminarily.

Treating ILD in AAV

In contrast to other lung manifestations, such as nodular disease, or alveolar hemorrhage, where treatment with glucocorticoids (GC), cyclophosphamide (CYC), rituximab (RTX), or methotrexate has been the subject of defined studies, there are no proven therapies for ILD in AAV. Current recommendations about the treatment of the AAV do not issue this patients subset [57•]. Present approaches arise from the overall experience with patients of AAV and this complication, and from retrospective series. There are no prospective studies examining this issue.

Data is conflicting. For example, when patients with MPA and LF were treated with GC and CYC or RTX, their survival between 1 and 5 years was better than those who were treated only with GC as results of their basal disease [44]. Some studies have found that patients who received immunosuppressants were less prone to develop ILD [33••, 49, 58]. These studies are in contrast to what has been proven concerning the use of these therapies in IPF. A large, randomized study had to be prematurely stopped at the interim analysis of a planned 60-week study period as in IPF, GC, and azathioprine led to higher mortality and number of hospitalizations [59]. This different course might support the idea that IPF and the LF seen in either MPA or only-positive MPO-ANCA patients are driven by unequal mechanisms. Supporting this, also in LF associated with MPA or MPO-ANCA-positive subjects, there is evidence that treating this condition in AAV as it is established for other manifestations does not prolong survival and the disease progresses nonetheless [31, 41, 43]. In none of such studies, either patients with MPA and LF or patients with the latter and MPO-ANCA, treatment influenced positively the evolution.

As complex as ILD is, many aspects deserve to be evaluated when treatment concerning this complication is considered. It might be not only enough to have a diagnosis of ILD, but to start up with, to best define the type of it. This aspect is relevant and is something which has not been overtly clear in some of the published data. This is probably because the response of NSIP to immunosuppressants could be better than when a UIP pattern is observed. As it is very difficult to evaluate histopathologically the type of subsets of ILD prior to treatment, which could give an insight into the treatment possibilities, careful evaluation of imaging patterns is important. If GGO or an organizing pneumonia pattern is present on HRCT, which suggest inflammation, and therefore, a cellular subtype of NSIP if assessed cytologically, it might be beneficial to treat with GC and immunosuppressants, as shown in studies of idiopathic NSIP [60, 61], but if the imaging pattern has a predominant fibrotic component, then this choice might not be of favor to the patients, as previously stated for LF [62]. If viewed only as per the presence of LF, there are proponents that treatment with GC with or without immunosuppressants offers no benefit. However, as stated above, this approach is not to be discarded yet and formal studies, as were done for IPF in the past, could be considered [44].

On the other hand, the use of both approved anti-fibrotic treatments in IPF (pirfenidone and nintedanib) may help patients who have developed this complication in the context of an AAV. According to the 2015 ATS/ERS/Japanese Respiratory Society (JRS)/Asociación Latinoamericana de Tórax (ALAT) consensus, these drugs are recommended in patients with a mild to moderate LF (stated as having a > 50% forced vital capacity and a > 30% DLCO) [63, 64]. This issue could be analyzed in the near future in the patients we are discussing now. There is an ongoing recruitment seeking 15 patients with LF and MPO-ANCA that will evaluate as only treatment group, pirfenidone (Clinical Trials NCT NCT03385668).

It must be said that ILD patients who were included into the studies, which led to approval of both antifibrotic agents, were those with a mild to moderate impairment of lung function and that the main outcome was related to the decline of this item, not to a prolonged survival. Also, both drugs led to an important rate of adverse effects, mainly gastrointestinal (60–70%), the most frequent being diarrhea, followed by nausea, and vomiting [65,66,67]. Discontinuation of treatment was observed in up to a fifth of the patients. Therefore, in patients with a more widespread disease, these issues deserve careful consideration for the design of studies that will evaluate antifibrotic therapy in patients with AAV and ILD.

Although treating ILD in other autoimmune diseases could be taken as analogous to the AAV (e.g. systemic sclerosis or inflammatory myopathies) [68•, 69•], such premise is hard to sustain, as the physiopathological mechanisms might be entirely different.

Once LF is present, we shall not omit other nonpharmacological measures such as smoking cessation, and the prevention of frequent prevalent infections such as pneumococcal pneumonia and influenza with adequate vaccination in accordance with general established recommendations [70]. The efficacy of influenza vaccination has been evaluated and favorably tested in terms of safety in patients with AAV under remission [71].

ANCA, vasculitis, and ILD: different scenarios

From the previous section, treatment may differ according to the following different scenarios concerning ILD, ANCA—mainly MPO-ANCA—, and AAV. The first comprises patients who have LF and are MPO-ANCA positive, either at time of the ILD diagnosis or thereafter. Approximately a quarter of them will develop frank MPA with other organ involvement [72••]. It is unclear why some go on to develop MPA and some do not. In one study, a UIP pattern was associated with evolution into MPA. In both temporal instances, this is if MPO-ANCA are present at the time of LF diagnosis or later, those left untreated had a higher risk to develop MPA. The majority were treated with GC, although some also with immunosuppressants (either CYC or cyclosporine). Linking this evidence with the hypothesis concerning physiopathology of ILD in AAV, we believe that this subset of patients may represent a limited form of MPA, analogous to the well-established concept of limited or localized GPA [25]. Certainly, this issue deserves proof of concept.

For this patients’ subset, we want to recall how interesting was the concept of interstitial pneumonia with autoimmune features that did not include this group of patients [73], although all proposed requisites (clinical, morphological and serological) are present.

The second scenario concerns patients with MPA who have a precedent LF. In a French study, almost half of the patients (n = 49) had preceding lung disease prior to other MPA manifestations [44], and more recently, in the study from the cohort of Bad Bramstedt of patients with MPA, followed from 1991 to 2013, 15% of patients had interstitial fibrotic lung disease [74••].

A third one is of those patients with MPA who later on develop LF. Some do so with a previous history of lung hemorrhage, but this is not a constant feature, and there are others who do not have any history of pulmonary complaints and present with LF. This can be found incidentally when for other reasons, lung imaging is performed, as mentioned regarding the clinical presentation of the disease, or it can present indolently. As an increasing number of patients with MPA might present at any time point an ILD—mainly LF—, our current conduct is to follow them up with regular imaging, at least every 2 years, but preferably, in an annual fashion.

Table 2 summarizes these scenarios and relates them to potential therapeutic preferences. It must be recalled that, as yet, only case series (already referenced) with their implicit limitations support what is presented.

Table 2 Clinical scenarios and according feasible treatment options
Full size table

Some questions

From a critical standpoint, we think the following aspects would be desirable to consider when designing studies that will solve the unmet need of useful therapies for these patients. Some have been implicitly set on the table above, according to existent data, but some are personal opinions.

  1. a.

    Does the patient with ILD presents at the time of the initial MPA diagnosis or does he have an ILD bearing exclusively MPO-ANCA (other diseases of autoantibodies associations sufficiently excluded)?

  2. b.

    What kind of ILD does the patient have? Is it a definite UIP translating LF, or is it inflammatory or fibrotic NSIP? Or is it an unclassifiable ILD?

  3. c.

    What is the lung functional status of the patient? Is the disease advanced and a serious functional impairment already exist?

  4. d.

    In cases with NSIP, would bronchoalveolar lavage and/or biopsy be useful to better define if the patients might have an inflammatory or a fibrotic NSIP? It is recognized that lymphocytosis is shown on BAL or biopsy in the first, something absent in the second type stated.

  5. e.

    Could a combined approach of usual therapy for AAV be considered in conjunction with antifibrotics?

  6. f.

    Which would be the outcome measures? What is to be monitored in such trials?

  7. g.

    Which other comparison groups of ILD patients would be included?

Certainly, trials that could take all these points into consideration would require enormous efforts and might not be, for one or another reason, absolutely feasible, would need probably a multicenter approach and considerable time to be completed, but would provide with answers to which currently there is no firm evidence.

Prognosis

Overall, once ILD is present in the AAV, the prognosis is somber. Surely, the lack of proven efficacious therapies contributes to this. The development of ILD shortens survival in the majority of reports which have examined this matter [6, 34, 56•, 74••, 75••].

Finally, continuous monitoring of the ILD with current tools (imaging, lung function testing), and very importantly, follow-up of patients with LF and MPO-ANCA positivity is essential, as to promptly detect additional organ involvement which would require the addition of standard therapy.

Conclusions

ILD in the AAV is an important subject for which many questions remain unanswered, especially regarding physiopathology, evolution, and treatment. Regarding the latter, no proven approach is evidenced-based. In the context of full-blown AAV, the general treatment guidelines, mainly due to other organ involvement, are to be followed, but there exists difficulty in the treatment of patients who present since the beginning with ILD, or in those in which this complication antedates other AAV organ involvement. ANCA positive patients, especially MPO-ANCA ones with ILD, mainly UIP, are still an enigma from every perspective. The dilemma of treating them or not, especially with drugs with potentially important side effects, will be resolved with appropriate studies.