Idiopathic inflammatory myopathies (IIMs), a group of autoimmune systemic diseases with diverse clinical manifestations, can affect multiple parts of the body, including the skin, muscles, joints, and lungs (1); however, the lung is one of the most commonly involved extramuscular organs, with the reported prevalence ranging between 20% and 86% (2), associated with poor prognosis and increased mortality (3, 4). Most dermatomyositis (DM) (90%) and anti-synthetase syndrome (ASS) (70%–85%) cases with myositis-specific autoantibodies (MSAs) are usually associated with lung injury (5). Interstitial lung disease (ILD) is one of the most common pulmonary manifestations primarily characterized by inflammation and fibrosis of the lung tissue, which can occur before, after, or concurrently with cutaneous or muscular manifestations but typically occurs early in the disease course (6, 7). Despite conventional treatment targeting the underlying condition, a subset of patients with ILD continue to experience a progressive fibrotic phenotype characterized by inexorable deterioration of lung function, respiratory symptoms, signs on high-resolution computed tomography (HRCT), as well as higher early mortality rate, which can be described as progressive fibrosing interstitial lung disease (PF-ILD) (8–10). The introduction of the PF-ILD concept was associated with the clinical trial of an antifibrotic drug (11), and was progressively integrated into subsequent research, which suggested that individuals with ILDs other than IPF, including those with IIM-ILD and other connective tissue disease associated ILD (CTD-ILD) (12–14), are also at risk of developing PF-ILD. Although various clinical drug trials (11, 15) and experts (16) have proposed different criteria, the assessment predominantly relies on changes in lung function, radiological findings, and clinical symptomatology. The prevalence of PF-ILD in patients with CTDs is challenging to ascertain, which has been reported a range of 18%–44% in current studies with limited scope at present, including the groups of IIMs, rheumatoid arthritis, systemic sclerosis (SSc), and other CTDs (17–20). Further subdivision of IIMs according to few studies revealed that about 50% of patients with anti-MDA5⁺ DM developed PF-ILD (20), whereas 35% of those with ASS developed PF-ILD (12). As the prognosis of patients with PF-ILD is worse than that of those with stable and reversible ILD, early identification of IIM with comorbid progressive pulmonary fibrosis is crucial for clinical management and improving prognosis (7).

Despite generalizable risk factors, various cohort studies have also found predictive risk factors for progressive fibrosis specific to autoimmune ILDs. It has been proposed that patients with CTD with diabetes mellitus, steroid treatment, and a pattern of fibrosis on HRCT have a higher risk of developing PF-ILD in 2-year follow-up (21). Other laboratory and clinical indicators identified to predict rapid progression or poor prognosis include serum ferritin, C-reactive protein (CRP), lactate dehydrogenase (LDH), and Krebs Von den Lungen-6 (KL-6), as well as anti-melanoma differentiation-associated protein 5 (anti-MDA5) antibody titres, anti-Ro-52 levels, white blood cell (WBC) count, disease duration, fever, and age, although these studies mainly focused on patients with DM, particularly those with anti-MDA5⁺ (22–24).

Previous studies have found numerical or functional dysregulation of circulating natural killer (NK) cell plays a pivotal role in IIM-ILD (25, 26). One of our recent studies discovered that an amyopathic dermatomyositis-associated ILD group with a cluster of activated CD45RA⁺HLA-DR⁺CD8⁺ T cells and a reduced proportion of CD56^dim NK cells showed a high prevalence of rapidly progressive ILD and increased mortality rates (27), indicating that peripheral lymphocyte immunological profiles have predictable potential for the development of ILD fibrotic progression. Lin et al. also elucidated that in anti-MDA5⁺ DM patients, the significantly reduced levels and an inhibitory phenotype of peripheral NK cells are correlated with heightened disease activity and adverse prognosis (28).

Given the important role of NK cells in IIM-ILD, we aimed to explore the prognostic value of lymphocyte subsets, particularly NK cells, on ILD progression in a broader range of patients with IIM. Early identification of patients with PF-ILD is essential for clinical management. Therefore, we investigated the peripheral blood immunological characteristics and related variables in a group of patients with IIM, mainly containing two subgroups, anti-MDA5⁺ DM and ASS, to assess the risk of PF-ILD by exploring new immunological indices. This will help to identify early high-risk populations with fibrosing progression of ILD, guide more precise individualized therapy, and provide new insights for clinical treatment and prognostic assessment.

Although evidence suggests the potential value of lymphocytes in the progression and prognosis of autoimmune disease-associated ILD, related studies are limited (27, 35). In this prospective study, we systematically described the clinical and immune characteristics of patients with IIM-ILD, with a focus on lung involvement, and separately discussed the two subpopulations, anti-MDA5⁺ DM and ASS, which are most susceptible to ILD; the relationships among peripheral blood lymphocyte characteristics, clinical parameters, and disease progression were also discussed. We found that immune cell depletion was strongly associated with pulmonary fibrosis progression in patients with IIM-ILD, and the lymphocyte characteristics of different IIM subtypes were partly different. The depletion of NK and Th cells and older age of onset were associated with PF-ILD in patients with anti-MDA5⁺ DM, whereas lymphopenia, NK cell depletion, and high levels of ferritin in the peripheral blood of patients with ASS were associated with PF-ILD. To the best of our knowledge, this is the first prospective study to focus on the immunological subtypes as risk factors of PF-ILD in a larger group of patients with IIM, particularly the two different IIM subtypes, and we identified the potential value of lymphocytes (especially NK cells) in predicting PF-ILD.

The ‘PF-ILD’ concept is more appropriate for the broader IIM population than the short-term assessed ‘rapidly progressive ILD (RP-ILD)’, favoring the exploration of more meaningful indicators by reducing bias from critically ill patients. In our study, PF-ILD occurred in approximately 40% of patients with IIM, which is slightly higher than other related studies. First, similar CTD-ILD studies associated with PF-ILD were notably limited, most of which did not explicitly analyze each disease subtype; second, most patients in our cohort were those with anti-ARS⁺ and anti-MDA5⁺, which can be supported by studies that found approximately a 50% probability of PF-ILD in anti-MDA5⁺ DM (20).

Currently, it is well known that risk factors for developing ILD during IIM are mainly associated with MSAs (such as MDA5) and ARS (36), and each IIM subgroup has its pathophysiology. Thus, we separately analyzed the immune profiles of anti-MDA5⁺ DM and ASS. Based on our findings, individuals with anti-MDA5⁺ DM had a significantly higher probability of developing PF-ILD than those with ASS; additionally, they have a higher likelihood of having classic rashes, such as heliotrope rash and Gottron sign and a tendency to develop fever and arthritis. Moreover, individuals with anti-MDA5⁺ DM have more severe lymphocyte depletion than those with ASS, as well as higher levels of liver enzymes and ferritin, which are similar to the clinical characteristics of viral infection (37). These findings reflect differences in the pathogenesis of the two IIM-ILD groups with different antibodies; however, the risk factors for PF-ILD in anti-MDA5⁺ DM and ASS were somewhat similar.

Our results suggest that NK cell depletion was the significant risk factor for PF-ILD within 1 year in both patients with anti-MDA5⁺ DM and ASS, providing strong evidence for the association between NK cells and IIM-ILD and revealing the potential role of NK cell count as a predictor of pulmonary fibrotic progression. Nevertheless, the optimal cutoffs are different, with a lower cutoff in the anti-MDA5⁺ DM group, whereas the cutoff for ASS is within the normal range of NK cells, possibly because of differences in sample sizes that require further expansion. NK cells are natural effector cells involved in innate immunity and are currently considered to be involved in regulating adaptive immunity, playing an important role in various immunological diseases (38). However, the potential of NK cells in IIM appears to be underestimated, and related research is limited. Several studies have found that NK cell depletion is more noticeable in patients with ASS with severe lung involvement (25) and in those with anti-MDA5⁺ DM than in those with other IIM subtypes (28). Our results not only supported that patients with anti-MDA5⁺ DM show more pronounced depletion of NK cells but also indicated that the number of NK cells can be used to identify patient populations more prone to progressive exacerbations of ILD in anti-MDA5⁺ DM and ASS. Additionally, comparing the characteristics of the lower NK cell counts group with the higher NK cell counts group within the overall IIM cohort, it was observed that patients with NK cell counts below the median not only exhibited a significant increase in the incidence of PF-ILD but also showed a heightened risk of RP-ILD. This aligned with the findings of Ye’s study (27), which reported a higher proportion of RP-ILD in patients with a reduction in CD56^dim NK cells accompanied by a cluster of activated CD45RA⁺HLA^-DR⁺CD8⁺ T cells. Our findings, therefore, reinforced the significant role of NK cells in evaluating and predicting the progression of ILD.

It is suggested that NK cells in the peripheral circulation play an inhibitory role in disease progression by coordinating the extent of the inflammatory response (39), whereas a reduction in circulating NK cells may weaken the suppression of pathogenic immune cells. Nevertheless, the exact mechanism of the NK cell depletion in IIM remains unknown. One hypothesis is that the depletion of circulating NK cells may be due to their migration from the peripheral circulation to the local site or apoptosis, reported by related clinical studies about ASS (26), SSc (40, 41), and sarcoidosis (42). Notably, the migration of NK cells to the lungs in anti-MDA5⁺ DM patients may be driven by the local production of chemokines induced by IFN-γ, such as CXCL10 (43, 44). Once recruited to the lungs, NK cells can further secrete IFN-γ, creating a positive feedback loop that amplifies the inflammatory response and exacerbates tissue damage (45). NK cell dysregulation has been reported in studies related to the COVID-19 infection, which inferred that direct viral attack, hyperinflammatory responses, activation-induced cell death, and the mobilization/homing of NK cells to affected tissues may be responsible for the NK cell depletion (46). However, these speculative mechanisms require further studies to elucidate the precise role of NK cells and IFN-γ in the pathogenesis of IIM-ILD.

Previous studies have found lymphocyte dysregulation as a risk factor for acute progression or death in anti-MDA5⁺ DM. For example, Zuo et al. demonstrated that reduced levels of CD3⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ T cells are predictors for the development of rapid progressive ILD in anti-MDA5⁺ DM (35), which was partially verified in our study. We also found a significant decrease in the levels of T, Th, Ts, and B cells and total lymphocytes in patients with PF-ILD; additionally, a decreased number of Th cells was found correlated with the risk of PF-ILD in patients with anti-MDA5⁺ DM. In contrast, lymphocyte depletion was also the significant risk factor for PF-ILD in patients with ASS, highlighting the important function and study prospect of lymphocytes in IIM-ILD.

Many studies have confirmed that laboratory markers, such as anti-MDA5 antibody, KL-6, ferritin, LDH, CK, CRP, and NLR, may serve as diagnostic and/or prognostic biomarkers of disease activity in IIM-ILD (47–53). Hyperferritinaemia has been recognized as an independent risk factor for poor prognosis in patients with anti-MDA5⁺ IIM-ILD (35, 49, 54) and in those with ASS-ILD (55). Ferritin is associated with macrophage activation (56). Our study revealed the predictive value of hyperferritinaemia for PF-ILD in patients with ASS. Given that our assessment criteria differ from death or acute exacerbation within 3 months, the results of multivariate regression analysis of anti-MDA5⁺ DM showed no statistical significance of high ferritin as an independent risk factor.

Furthermore, our study showed that patients with anti-MDA5⁺ or anti-ARS⁺ commonly exhibit comorbidity with anti-Ro52 antibodies because Ro52 forms Ro52/IgG/HLA-DR complexes on the cell surface, which are specifically recognized by autoantibodies in some patients with inflammatory myopathies (57).

While the vast majority of patients were treated with either glucocorticoids or immunosuppressive therapy at enrollment, our analysis revealed no substantial differences in the specifics of immunological therapy or antifibrotic treatment between the PF-ILD and non-PF-ILD groups at baseline. This finding suggests that the initial use of immunological therapy may not serve as a direct risk factor for PF-ILD. However, it is important to note that in the multivariate Cox regression analysis, immunosuppressant use appeared to emerge as an independent protective factor for prognosis in PF-ILD patients in ASS, despite showing no significance in the univariate analysis. However, due to the wide variety of immunosuppressants and the limited number of cases using each immunosuppressant, it is challenging to further analyze the impact of different immunosuppressants on prognosis.

This study had some limitations. Firstly, as a single-center study with a limited sample size, further multicenter studies with larger sample sizes and external validation are required to confirm these findings and explore the characteristics of more disease subtypes. Secondly, the heterogeneity of NK cells in different autoimmune diseases (58) and the potential impact of glucocorticoids and immunosuppressive treatments may have masked differences in clinical or laboratory test results. Furthermore, the functions and mechanisms of various lymphocyte subtypes in IIM-ILD were not thoroughly investigated.

Studies with larger sample sizes and more refined experimental designs are needed to dynamically record patients’ use of immunosuppressants during disease progression to perform a more accurate evaluation of the impact of different therapies on the prognosis of PF-ILD patients. Additionally, investigating the potential impact of immunosuppressants on lymphocyte subsets is crucial for comprehensively understanding the mechanisms of PF-ILD. Future research should also focus more on the functions of these immune subtypes to elucidate their significance in disease progression, particularly in different disease subtypes.

Patients with anti-MDA5⁺ DM and ASS have different immune profiles, and they have their independent risk factors for PF-ILD; however, both groups have components associated with lymphocyte exhaustion. The depletion of NK cells with different cutoff values is a risk factor for PF-ILD in these two disease subtypes, suggesting that NK cell count may provide valuable information for the prognostic assessment of patients with IIM-ILD, thereby improving our understanding of disease pathogenesis. Our findings indicate that patients at risk of PF-ILD can be identified earlier by testing the NK cell and other lymphocyte levels, which could help clinicians manage patients with IIM-ILD and have implications for clinical monitoring, individualized therapy, and pathogenesis studies. Future studies are needed to focus on the functions and mechanisms of NK cell depletion and other cell subsets in IIM-ILD.

The original contributions presented in the study are included in the article/ Supplementary Material , further inquiries can be directed to the corresponding author.

The studies involving humans were approved by Research Ethics Committee of Renji Hospital. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

CS: Conceptualization, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing, Data curation, Software, Visualization. NX: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. YZ: Data curation, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. XZ: Data curation, Investigation, Visualization, Writing – original draft, Writing – review & editing, Supervision. NY: Data curation, Investigation, Supervision, Writing – original draft, Writing – review & editing. QG: Investigation, Supervision, Writing – original draft, Writing – review & editing, Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Resources.