Interstitial lung disease (ILD) comprise a heterogeneous group of pulmonary diseases that affect the lung interstitium and can present with different radiological and histopathological patterns. Main causes of ILD are idiopathic pulmonary fibrosis (IPF), environmental exposures and connective tissue diseases (CTD).1 An important proportion of patients with ILD are positive to highly specific autoantibodies, but many of these patients may not be classified as specific CTD according current classification criteria. For example, those with positive myositis-specific autoantibodies (MSA) such as the anti-aminoacyl tRNA synthetases (anti tRNAs: anti Jo-1, anti PL-7, anti PL-12, anti Ej, anti Oj, anti KS, anti Zo, anti Ha), and anti MDA-5 or myositis-associated antibodies (MAA) like anti PM/SCL, anti Ku, or anti Ro-52, among others.2

The myositis or idiopathic inflammatory myopathies (IIM) are a group of autoimmune diseases that share some clinical manifestations such as muscle inflammation, skin rash, arthritis, and high prevalence of ILD. Different phenotypes among the IIM are closely related to the subtype of MSA and MAA,2 and although it is well recognized that ILD is a very important manifestation of some IIM in the context of MSA/MAA, ILD is not included in the EULAR/ACR classification criteria for IIM.3 So, a great proportion of ILD patients positive to MSA/MAA may not be classified as IIM, but may be classified as interstitial pneumonia with autoimmune features (IPAF) according to the proposal of the American Thoracic Society/European Respiratory Society (ATS/ERS).4

At present, there is limited knowledge about the differences in clinical manifestations, response to treatment, pulmonary function and prognosis of ILD patients positive to MSA/MAA. These are diverse autoantibodies, and as we mentioned before, they are associated to different phenotypes and perhaps different prognosis.

With this background, we designed this study with the aim to describe the clinical, radiological and pulmonary function of Latin-American patients with ILD positive to MSA/MAA and evaluate risk factors associated to pulmonary function at onset and long term follow up.

We conducted a multicentric retrospective cohort study. We included patients with ILD and MSA/MAA evaluated between 2016 and 2018 in 3 multidisciplinary ILD clinics in Argentina (Hospital María Ferrer), Chile (Instituto Nacional del Tórax) and Mexico (Instituto Nacional de Enfermedades Respiratorias). Every patient was confirmed to have ILD by thoracic high-resolution computed tomography (HRCT) and their medical records were reviewed. All patients were evaluated by a multidisciplinary team of pulmonologists, immunologists/rheumatologists and radiologists.

The following variables were described: age, gender, tobacco exposure, gastroesophageal reflux disease (GERD) symptoms, dyspnea severity (assessed with the mMRC scale) and treatment schemes. We registered the following data related with the presence of an underlying CTD: arthritis, muscle weakness, mechanic's hands (distal digital hyperkeratosis and fissuring), fever, Raynaud's phenomenon, sclerodactyly, puffy hands, telangiectasias, skin ulcers, sicca symptoms (xerostomia and/or xerophthalmia) and dermatomyositis rash (heliotrope rash, Gottron's papules, shawl sign). These were called systemic symptoms.

We included 211 patients. Chile 119 (56.4%), Mexico 50 (23.7%) and Argentina 42 (19.9%). The majority of patients were women (77.4%), with a mean age of 57 years (SD: 12), and 34% were past or current smokers (Table 1).

In 146 patients (70%) ILD was diagnosed first, in 43 patients (18%) ILD and CTD were diagnosed simultaneously and CTD was diagnosed first in 25 patients (12%). The mean interval between ILD and CTD diagnosis was 7.6 months.

Patients with positive ANA were 135 (65.2%), with speckled and homogeneous patterns being the most prevalent (22.1% and 17.9%, respectively). The most frequent MSA/MAA were: Ro-52 (54.2%), Jo-1 (25%), PL-12 (18.8%), PM-Scl 75-100 (15.5%), PL-7 (13.9%) and Ku (12%) (Table 1).

The most frequent extra-thoracic symptoms were sicca symptoms (59.2%), arthritis (53.6%), mechanic's hands (48.3%) and Raynaud's (45%). Less than 50% had muscle weakness (Table 1). Almost one third of patients (27.1%) had diagnosis of a CTD different from IIMs, mainly Sjögren's syndrome and systemic sclerosis (SSc). The most frequent radiological patterns were: NSIP/OP (45.5%) and NSIP (34.5%). UIP pattern was found in only 9.6% of patients (Table 1).

We describe the clinical characteristics and variables associated with functional improvement in a cohort of patients with interstitial lung disease and MSA/MAA from three centers specialized in ILD from Latin America.

In our cohort we observed a greater proportion of middle-aged female patients, concordant with previously published literature. Among MSA, anti tRNAS were the most frequently identified, especially anti Jo-1. These autoantibodies are associated with variable presence of ILD, myositis, arthritis, Raynaud's phenomenon, mechanic's hands and fever, configuring the so-called “Anti-synthetase syndrome”, despite the fact that this clinical syndrome has not validated classification criteria yet.9

The Jo-1 antibody is the most frequently identified in our cohort and in other published data. Unlike the rest of the anti tRNAS, it is more often associated with concomitant myositis at debut of the disease. This is important because although the anti tRNAS are closely associated with the spectrum of the IIM, the presence of muscle involvement is not universal. Anti PL-12, PL-7, KS, OJ and, in smaller proportion, anti EJ are associated with isolated ILD at the beginning of the disease and do not usually develop muscle involvement during follow-up.10

The most frequent extrapulmonary manifestations were sicca symptoms, arthritis, mechanic's hands and Raynaud's phenomenon; whereas muscle weakness and elevated muscle enzymes were not a constant finding, similar to those described in other cohorts.11,12

The pattern of NSIP isolated or associated with OP was the most frequently HRCT pattern encountered, similar to the study published by Waseda et al.,13 where the majority of patients with ILD associated with anti tRNAS had an inconsistent pattern with UIP, corresponding to the NSIP and OP patterns. This tomographic pattern is typically described in patients with MSA/MAA, whereas UIP pattern is less frequent.14 ANA was negative in almost a third of the cohort. In the scenario of MSA/MAA, particularly anti tRNAS, this finding is not unusual due to the cytoplasmic localization of the autoantigens. This is important because many patients could erroneously be classified as having an idiopathic interstitial disease if ANA is negative and no extrapulmonary manifestation is present at the time of the evaluation, delaying the correct diagnosis and an adequate treatment.14,15 In the case of our cohort, 70% of the patients received the initial diagnosis of isolated ILD, with a time interval until the final diagnosis of 7.6 months with a maximum interval of 280 months (23 years).

Regarding MAA, anti Ro-52 was identified in almost half of the cohort. This association is described in the literature not only for the spectrum of IIM but also has been identified in other CTDs like SSc and systemic lupus erythematosus. The presence of this autoantibody has been linked to ILD, Raynaud's phenomenon, mechanic's hands and Jo-1. The relationship of anti Ro-52 with prognosis and course of the disease is controversial.16,17 In our cohort the positivity of PM/SCL was identified in 15% of patients. This antibody is mainly found in the superposition of systemic sclerosis with myositis, and is associated with a higher frequency of sclerodactyly, Raynaud's phenomenon and skin manifestations associated with dermatomyositis. It has been suggested that anti tRNAS and PM/SCL antibodies could present a similar spectrum of clinical manifestations and share certain human leukocyte antigens (HLA).12,18 Finally, anti MDA-5 was observed in 8% of the patients in our cohort. This autoantibody test is not universally available in all our ILD clinics, so we do not know if its frequency could be higher in our MSA/MAA-ILD patients. MDA-5 is described to be associated with clinically amyopathic dermatomyositis, digital ulcers and rapidly progressive ILD with poor response to immunosuppressive treatment.19

Only a small percentage of patients received treatment with corticosteroids (CS) as monotherapy; in the majority of cases CS were used in association with mycophenolate, azathioprine or cyclophosphamide. At the moment, no randomized clinical trials have been published to evaluate the effectiveness of these drugs in the treatment of ILD associated with the presence MSA/MAA, and most of the information comes from observational studies.20–22 Therefore, the indication and choice of the immunosuppressive treatment is based on the available information, expert opinion, local experience with a particular drug, tolerance profile and access to certain drugs. Despite differences in treatment strategies among the three centers, most of the patients remained stable or improved after immunosuppressive therapy.

We observed that OP pattern was associated with functional improvement during follow-up, like other studies. This pattern is related to the presence of inflammation in the lung parenchyma and is described within the classification of the idiopathic ILD as reversible, with good response to treatment and a subacute-acute course.23–25 Although not reaching statically significance, the absence of sclerodactyly was a factor associated with FVC improvement. This could be explained because there is a subgroup of patients with MSA/MAA and SSc features overlap.26 In SSc one of the main pathogenic mechanism is vascular damage involving microvascular endothelial cells leading to fibroproliferative vasculopathy and capillary rarefication and the activation of fibroblasts to myofibroblasts leading to excessive extracellular matrix deposition in skin and blood vessels. It has been described that patients with loss of nailfold capillary density and progressive skin fibrosis within one year of the diagnosis had a FVC decline ≥10% and worse survival during follow up.27–29

The presence of systemic manifestations along with ILD was also related to functional improvement. This could be due to an early diagnosis of the disease and timely treatment, unlike when ILD is the only dominant manifestation.11,12,30 The latter association is in accordance with the fact that Jo-1 positivity was associated with functional improvement in the univariate analysis, as it is described by other authors; but the influence of Jo-1 disappeared in the multivariate analysis. One of the reasons might be that systemic manifestations are more common in anti Jo-1 patients.

Our study has some limitations. First, the inherent limitations of a retrospective study, such as the possible presence of a selection bias. Second, as it is a multicentric cohort study, patients could have been evaluated differently, not all the aforementioned antibodies were evaluated in all patients due to the access of each center to the determinations, so the prevalence of the antibodies may vary. Third, not all patients had baseline pulmonary function (maximum 164/211) or follow-up pulmonary function (maximum 121/211). Finally, the selection criteria for immunosuppressive treatment may have varied according to the preferences of each center and/or access to different drugs.

Despite these aspects, this study emphasizes the importance of searching for signs and symptoms related to the presence of an autoimmune disease in patients with ILD; also highlights the importance of testing for antibodies related to myositis in patients with ILD who do not present overt muscle or skin involvement, especially those with NSIP or OP pattern. Finally, we identify factors that would be related to functional improvement in those who receive immunosuppressive treatment in the context of MSA/MAA-ILD in an ethnically Latin American cohort.

In this article we describe the clinical, radiological and pulmonary function of a cohort of Latin-American patients with ILD who are positive for myositis specific and associated antibodies; and the risk factors associated to pulmonary function at onset and long term follow up.

Alberti María Laura: Conceptualization, Data curation, Investigation, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft (e) and Writing – review & editing (f).

Wolff Verónica: Conceptualization, Data curation, Investigation, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft (e) and Writing – review & editing (f).

Reyes Felipe: Conceptualization, Data curation, Formal analysis, Investigation, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft (e) and Writing – review & editing (f).

Juárez-León Ernesto: Conceptualization, Data curation and Writing – original draft (e).

Fassola Leandro: Data curation and Resources.

Carballo Gabriel: Resources and Writing – review & editing (f).

Buendía-Roldán Ivette: Data curation, Visualization and Writing – review & editing (f).

Rojas-Serrano Jorge: Visualization and Writing – review & editing (f).

Caro Fabián: Conceptualization, Data curation, Resources, Visualization and Writing – review & editing (f).

Florenzano Matías: Conceptualization, Data curation-Project administration, Visualization and Writing – review & editing (f).

Paulín Francisco: Conceptualization, Data curation, Formal analysis, Methodology, Visualization and Writing – review & editing (f).

This investigation has not received specific aid from public sector agencies, the commercial sector or non-profit entities.

All authors declare they have no conflicts of interest.