We examined patients admitted to Osaka Medical and Pharmaceutical University Hospital between October 2014 and August 2022. The diagnosis of PM, ADM, or CADM was made based on the Bohan and Peter criteria (1, 2) or criteria outlined by Sontheimer and Gerami et al. (3, 4) The diagnosis of interstitial pneumonia with autoimmune-features (IPAF) was made using the criteria of the ERS/ATS task force (22). ACR/EULAR classification criteria were used to check whether they were classified as “definite IIM” or “probable IIM” (6). Patients with malignancies were excluded from this study. The presence of ILD was assessed using HRCT. ILD can be divided into acute/subacute ILD (A/S-ILD) and chronic ILD (C-ILD). A/S-ILD was defined as rapid worsening of the respiratory condition, laboratory pulmonary function tests, arterial blood gas findings, and chest HRCT images within 3 months of onset (23). On the contrary, C-ILD does not fulfil the definition of A/S-ILD. Disease duration was defined as the period between the onset of symptoms associated with anti-ARS-ILD and hospital admission. Clinical data were obtained from the medical records. This study was conducted in accordance with the Declaration of Helsinki and its amendments and it was approved by the Ethics Committee of Osaka Medical and Pharmaceutical University (approval no. 1529). Informed consent was obtained from each patient.

Arterial blood gas analysis, including PaO2, PaCO2, and alveolar-arterial oxygen difference (AaDO2), was performed at admission. Respiratory function, including forced vital capacity (FVC) and diffusion capacity of the lungs for carbon monoxide (DLco), was measured using spirometry (SYSTEM21; Minato Medical Science, Osaka, Japan), as previously described (24). Respiratory function test results are expressed as percentages of the predicted value.

HRCT was performed using a 64-detector row CT Aquilon multiscanner (Toshiba Medical Systems Corporation, Tokyo, Japan). The slice thickness was 1.0–1.5 mm every 10 mm, and the scan area included the entire lung. All patients underwent chest HRCT before treatment, and the images were reviewed independently by three observers (TK, TSu, and TSh) who were blinded to the patients’ clinical information. Inter-observer disagreements were resolved by consensus. Ground-glass opacity (GGO) and fibrosis were scored to assess the HRCT findings as previously described (25). Each patient’s lobe was scored by the same observer, and the average value was used. The scores were then summed to obtain the total CT score.

Nonspecific interstitial pneumonia (NSIP) and organizing pneumonia (OP) are typical characteristics in anti-ARS-ILD, so we classified the HRCT pattern into NSIP, OP, and NSIP with OP pattern, as previously reported (16, 26). All patients underwent chest HRCT prior to treatment, and images were reviewed independently by 2 pulmonologists (TK, and TS) blinded to the patients’ clinical information.

The levels of creatine kinase (CK), ALD, lactic acid dehydrogenase (LDH), C-reactive protein (CRP), KL-6, SP-D, and ferritin were measured. Anti-ARS antibodies were examined using ELISA (MESACUP; MBL, Nagoya, Japan) and a blot assay (Myositis Profile Euroline Blot Test Kit; EUROIMMUN, Lübeck, Germany). Autoantibodies (antigen including Jo-1, PL-12, PL-7, EJ, Ro-52) were measured by this immunoblotting.

PSL (0.5–1.0 mg/kg/day) was administered to all the patients. CSA or TAC were used as a combination therapy at the physician’s discretion. CSA was started at 4 mg/kg/day, once daily, before breakfast, and the concentration 2 h after administration was adjusted to 1,000 ng/mL or higher. TAC was started at 0.1 mg/kg/day twice a day before breakfast and dinner, and the trough concentration was adjusted to 5–10 ng/mL. Additional treatments, such as methylprednisolone (MPDN) pulse therapy, intravenous cyclophosphamide pulse therapy (IVCY), intravenous immunoglobulin (IVIG), plasma exchange (PE), and mycophenolate mofetil (MMF), were administered based on each patient’s condition at the physician’s discretion.

PSL was progressively tapered according to the clinical effectiveness after remission induction therapy. In all cases, clinical and biochemical evaluation was performed every 1-2 months. Relapse of ILD was defined when both of the following conditions were fulfilled after the remission induction therapy: the appearance of a new GGO on chest HRCT and the requirement of intensified treatment (15, 27, 28).. We excluded other conditions which could cause ground glass opacities on HRCT, including cardiac failure, fluid overload, and pulmonary embolism (27). A concrete example of a relapse in chest HRCT is shown in Supplementary Figure 1 .

Moreover, we defined “responders” as patients who responded to remission induction therapy without disease exacerbation and were discharged, and defined “non responders” as those who did not respond to remission induction therapy with exacerbation of interstitial shadows on HRCT (29).

The data are presented as the median (interquartile range). Fisher’s exact test was used when appropriate, and the Mann–Whitney U test was used to compare median values. We used receiver operating characteristic (ROC) curve analysis to determine the most suitable cut-off level for predicting anti-ARS-ILD relapse. We used the Kaplan–Meier method to assess survival curves and the log-rank test to evaluate the significance of differences between the two groups. The correlations were evaluated using Spearman’s correlation coefficients. A P-value of <0.05 was considered significant. Data were analysed using JMP (version 17.0; SAS Institute Inc., Cary, NC, USA) and GraphPad Prism (version 8.0; GraphPad Software, La Jolla, CA, USA).

A flowchart illustrating the selection and prognosis of patients with anti-ARS-ILD is shown in Figure 1 . Sixty patients with anti-ARS-ILD were admitted to our hospital for the first remission induction therapy between October 2014 and August 2022, and 10 patients were excluded for the following reasons: coexistence of progressive cancer before the diagnosis of anti-ARS-ILD (1 case), not being treated with PSL and CNI (3 cases), discontinuation of IS within 1 year (1 case), and death within 1 year after remission induction therapy (5 cases). The clinical characteristics of 50 patients with anti-ARS-ILD are shown in Table 1 . The median patient age was 61 years, and 70% of the patients were women. Eighteen percent of patients were “definite IM” and 44% patients were “probable IIM” using ACR/EULAR criteria (6). Sixty percent (30 patients) had A/S-ILD, and 40% (20 patients) had C-ILD. All patients with C-ILD underwent therapeutic intervention because of the gradual progression of ILD from more than 3 months to several years. The median disease duration was 3.2 months. The median serum levels of various parameters were as follows: CK, 100 U/L (reference ranges: 41–153 U/L); ALD, 6.3 U/L (reference ranges: 2.1–6.1 U/L); CRP, 0.29 mg/dL (reference ranges <0.14 mg/dL); KL-6, 1,022 U/mL (reference ranges: 105–401 U/mL); SP-D, 167 ng/mL (reference ranges: <110 ng/mL); ferritin, 132 ng/mL (reference ranges: 6.2–138 ng/mL). In the PFT findings, the median percentage predicted for FVC and DLco was 81.6% and 51.2%, respectively. The HRCT pattern included NSIP in 54%, NSIP with OP overlap pattern in 42%, and OP in 4%. There were no patients who had classified as usual interstitial pneumonia pattern.

All patients were treated with prednisolone and calcineurin inhibitors as remission induction therapy and followed up for over a year with these combination therapies. ( Figure 1 ) All patients were responders in this study. During a median follow-up of 4.8 years, 19 patients relapsed after remission induction therapy. The median time from initiation of remission induction therapy to relapse of ILD was 2.5 years. The median dose of prednisolone was 6 mg/day, and that of CNI, including TAC and CSA, was 2.5 mg/day and 150 mg/day, respectively. We compared the clinical and laboratory findings between the relapsed (19 patients) and non-relapsed (31 patients) anti-ARS-ILD groups ( Table 2 ). No significant differences were observed in age, sex, or disease duration. The proportion of patients with A/S-ILD was higher in the relapsed group (79.0%) than that in the non-relapsed group (48.4%; P=0.04). Sixty-two percent of patients with anti-Jo-1 antibody, 56% of patients with anti-EJ antibody and 60% of patients with anti-PL-7/PL-12 antibodies were A/S-ILDs. There were no significant differences in the ratio of A/S-ILD for the type of anti-ARS antibody. Additionally, the initial serum levels of ALD and SP-D were higher in the relapsed group (8.6 U/L, 248 ng/mL, respectively) than those in the non-relapsed group (5.0 IU/L, 147 ng/mL) (P=0.004, 0.009, respectively). There were no significant differences in the initial serum levels of CK, CRP, KL-6, ferritin, or AaDO2 between the two groups. There were no significant differences in the ratio of the type of anti-ARS antibody between the two groups. Also, there were no significant differences in the anti-Ro52 positivity between the two groups. In the pulmonary function test results, the %FVC was significantly lower in the relapsed group (70%) than that in the non-relapsed group (84.9%; P=0.02) There were no significant differences in the total GGO and fibrosis score between the relapsed and non-relapsed groups. There were no significances in HRCT pattern between them.

We compared the contents of treatment between relapsed group and non-relapsed group ( Supplementary Table 1 ). The median initial PSL dose was significantly higher in the relapse group than in the non-relapse group. (P=0.02). Moreover, there were no differences observed in the doses of CSA or TAC or the frequency of MPDN, IVCY, PE, or MMF administration between the two groups.

A/S-ILD, higher serum ALD and SP-D levels, and lower %FVC were identified as risk factors in the univariate analysis. Next, we performed a Cox regression analysis using these risk factors. Univariate analysis using a Cox regression model showed that A/S-ILD, higher serum ALD and SP-D levels, and lower %FVC were predictors of ILD relapse in patients with anti-ARS-ILD (P=0.01, 0.0007, 0.01, and 0.0007, respectively; Table 3 ).

Previous reports have shown that ageing and disease duration from disease onset to therapeutic intervention are associated with poor prognosis in patients with PM/DM-ILD (30, 31). We performed multivariate analyses using Cox regression analysis to determine whether A/S-ILD, higher serum ALD and SP-D levels, and lower %FVC were independently associated with ILD relapse after adjusting for age and disease duration. After adjusting for these covariates, multivariate analysis using a Cox regression model also revealed that A/S-ILD, higher serum ALD and SP-D levels, and lower %FVC were independent risk factors for relapse in patients with anti-ARS-ILD (P=0.0007, 0.0004, 0.01, and 0.0005, respectively; Table 3 ).

To estimate the cut-off points for assessing factors related to the poor prognosis of anti-ARS-ILD, ROC curve analysis was performed using ALD, SP-D, and %FVC. The level that maximised the area under the ROC curve was 6.3 U/L for ALD (area under the curve [AUC]: 0.75, sensitivity: 83.3%, specificity: 66.7%), 207 ng/mL for SP-D (AUC: 0.75, sensitivity: 78.6%, specificity: 80%), and 76.8% for %FVC (AUC: 0.72, sensitivity: 62.5%, specificity: 80%). Thus, ALD ≥6.3 U/L, SP-D ≥207 ng/mL, and %FVC ≤76.8% were the best cut-off levels for indicating a poor prognosis. The ROC curves of ALD, SP-D, and %FVC for differentiating between the relapsed and non-relapsed groups are shown in Supplementary Figure 2 .

  The patients were then divided into two groups based on these cut-off levels, and Kaplan–Meier survival curves were plotted ( Figures 2A–D ). The 5-year relapse rate was significantly higher in patients with A/S-ILD (57.3%) than in those with C-ILD (23.7%; P=0.009) The 5-year relapse rate was significantly higher in patients with ALD ≥6.3 U/L (70.3%) than in those with ALD <6.3 U/L (11.8%; P=0.0005; Figure 2B ). The 5-year relapse rate was significantly higher in patients with SP-D ≥207 ng/mL (71.3%) than in those with SP-D <207 ng/mL (15.7%; P=0.0007; Figure 2C ). The 5-year relapse rate was significantly higher in patients with %FVC of ≤76.8% (78.2%) than in those with %FVC of ≻76.8% (28.7%). (P=0.0004; Figure 2D ).

A previous report showed that patients with ASS often present with fasciitis-dominant myopathy with elevated serum ALD levels, and myofascitis was more likely to be observed in patients with DM with rapid progressive ILD (RP-ILD) than in those without RP-ILD (32, 33). Based on these findings, we examined whether serum ALD levels were correlated with disease severity indicators, including CRP, ferritin, KL-6, SP-D, %FVC, %DLco, and HRCT scores, in patients with anti-ARS-ILD on admission, as shown in Table 4 (24). Serum ALD levels were significantly and positively correlated with CRP (R=0.51), ferritin (R=0.49), and total GGO scores (R=0.34; P=0.0002, 0.0005, and 0.02, respectively). Moreover, they were significantly negatively correlated with %FVC (R=-0.46) and %DLco (R=-0.41; P=0.003 and 0.009, respectively). In contrast, there were no significant correlations between serum CK levels and these biomarkers.