Abstract
Patients with systemic autoimmune rheumatic diseases-interstitial lung disease (SARD-ILD) exhibit diverse clinical courses, highlighting the importance of prognostic prediction for effective management. This study aimed to validate the gender-age-physiology (GAP) model in patients with SARD-ILD and identify additional prognostic factors. Clinical data of patients diagnosed with SARD-ILD at a tertiary center in South Korea were retrospectively analyzed. Using variables from the GAP model, along with exercise capacity, chest computed tomography (CT) patterns, and clinical course factors such as progressive pulmonary fibrosis (PPF), multivariate and survival analyses were performed to evaluate prognostic factors for 3-year mortality. Among 142 patients with SARD-ILD, 27 died and one underwent lung transplantation over a median follow-up period of 32.8 months. In the multivariate analysis, higher GAP stages, the combination of radiologic usual interstitial pneumonia (UIP) patterns and exertional hypoxia during the 6-min walk test (hazard ratio [HR]: 2.506, P = 0.049), and PPF (HR 3.481, P = 0.001) were independently associated with 3-year mortality. Additionally, among patients in advanced GAP groups (GAP stage II or III, n = 45), those with both exertional hypoxia and a UIP pattern had significantly worse prognosis (median survival: 18.0 vs. 31.2 months, P = 0.042). Additional prognostic factors beyond the GAP model were identified, including exercise capacity, CT imaging patterns, and clinical course. These findings underscore the importance of incorporating multiple prognostic indicators for improved stratification and management of patients with SARD-ILD.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-025-08484-3.
Keywords: Interstitial lung disease, Collagen vascular diseases, Lung fibrosis
Subject terms: Respiratory tract diseases, Connective tissue diseases, Respiratory signs and symptoms
Introduction
Interstitial lung diseases (ILD) involve inflammation and/or fibrosis in the lung parenchyma, leading to significant morbidity and mortality1. In patients with systemic autoimmune rheumatic diseases (SARDs), clinically evident SARD has a variable prevalence but occurs in up to 85% of patients and might be the only manifestation of SARD2. SARD-ILD has a diverse clinical course, from incidental radiographic image findings to rapid progression to respiratory failure and death3.
Although staging systems have been proposed for ILD and specific types of SARD-ILD, such as ILD-gender-age-physiology (ILD-GAP) model4, systemic sclerosis (SSc)-ILD5, and rheumatoid arthritis (RA)-ILD6, few studies include all forms of SARD-ILD because of their heterogeneity. The GAP index and staging system were originally developed and validated to predict mortality in patients with idiopathic pulmonary fibrosis (IPF)7,8. Previous reports have also suggested that the GAP model can be used as a mortality risk prediction tool in patients with all ILD subtypes4, and RA-ILD9–11. However, the GAP model does not consider exercise capacity12,13 or high resolution computed tomography (HRCT) patterns, such as the usual interstitial pneumonia (UIP) pattern14,15, both of which recognized as prognostic factors in patients with ILD. Therefore, this study aimed to validate the GAP model in patients with SARD-ILD and identify additional prognostic factors for this population.
Methods
Study population and clinical data
We retrospectively collected data from patients with SARD-ILD between April 1, 2007, and October 27, 2022, from a tertiary referral hospital in South Korea. A total of 197 patients were gathered in the study cohort. Among them, 55 patients who did not have baseline data, such as forced vital capacity (FVC), diffusing capacity for carbon monoxide (DLco), 6-min walk distance (6MWD), and exertional hypoxia during 6-min walk test (6MWT) at that time of ILD diagnosis, were excluded. Specific SARD diagnoses were made by rheumatologists following established guidelines16–23. This study was conducted following the Declaration of Helsinki. The Institutional Review Board of Asan Medical Center endorsed this study protocol (IRB no. 2022-1564, 2023-1394). The Institutional Review Board of Asan Medical Center waived the need for informed consent because of anonymous clinical data and retrospective study design.
Data collection and definition
All clinical information was collected during routine patient visits, and the medical records of included patients were retrospectively analyzed on May 31, 2024. Survival and clinical data were primarily obtained from the medical records of our institution, with additional support for survival data from the database of the National Health Insurance Service in South Korea. The primary endpoint was defined as death or lung transplantation within 3 years.
The GAP index and staging system consists of the following variables: gender (male, female), age (years > 65, 61–65, ≤ 60), FVC (% predicted > 75, 50–75, < 50), and DLco (% predicted > 55, 36–55, ≤ 35, unable to perform). Points were assigned to each variable category as described in a previous study, and the total score was used to determine the stage: 0–3 points correspond to stage I, 4–5 points to stage II, and 6–8 points to stage III7. HRCT images were independently reviewed by thoracic radiologists blinded to clinical and pathological information. UIP on CT was defined as either a UIP or probable UIP pattern, whereas non-UIP was defined as patterns indeterminate for UIP or suggestive of an alternative diagnosis24. Progressive pulmonary fibrosis (PPF) was defined as the presence of at least two of the following three criteria within 1 year: 1) physiological criteria of disease progression (absolute decline in FVC ≥ 5% predicted or absolute decline in DLco (corrected for Hb) ≥ 10% predicted), 2) worsening respiratory symptoms, and 3) radiological evidence of disease progression25. Because of the small number of patients in GAP stage III (n = 4), we grouped advanced GAP stage patients (GAP stages II or III) for subgroup analysis.
Statistical analysis
All data are expressed as median [interquartile ranges (IQR)] for continuous variables, and as percentages for categorical variables. Continuous variables were tested using the Mann–Whitney U test, whereas categorical variables were tested using the Chi-square test or Fisher’s exact test. Patients with SARD-ILD were categorized into two groups based on 3-year mortality; non-survivors and survivors. Risk factors for all-cause mortality over 3 years were analyzed using the Cox proportional hazards analysis. Variables with P < 0.1 in the unadjusted analysis were entered into multivariable models. The Kaplan–Meier (KM) estimates and the log-rank test were used for survival analysis. We employed a minimum saturation of peripheral oxygen (SpO2) cutoff value of 88% during 6MWT and 6-min walk distance (6MWD) cutoff value of 250 m to analyze clinical outcomes based on its importance in the prognosis of patients with ILD26–28. Statistical analysis was performed using IBM SPSS version 24.0 (IBM, Armonk, NY, USA) and R software version 4.2.1 (the R Foundation, Vienna, Austria).
Results
Baseline characteristics
A total of 197 patients with SARD-ILD who were screened between April 2007 and October 2022 were initially identified. Of these, 142 patients had sufficient data for inclusion in the final analysis (Fig. 1), and the median follow-up was 32.8 months (IQR, 22.5–44.9 months). During follow-up, 27 (19.0%) patients died, and one patient underwent lung transplantation within 3 years. The baseline demographic characteristics are presented in Table 1. Non-survivors were older (67 vs. 59 years, P = 0.005), had a higher proportion of ever-smokers (57.1 vs. 31.6%, P = 0.022), and a higher prevalence of the UIP pattern on HRCT (53.6 vs. 20.2%, P = 0.001). For the types of SARDs, there were no significant differences between the two groups. Pulmonary function parameters were significantly lower in non-survivors (FVC, predicted %, 58.0 vs. 76.0%, P < 0.001; DLco, predicted %, 40.5 vs. 58.0%, P < 0.001). Exercise capacity was also lower in non-survivors, as indicated by a shorter 6MWD (390.0 vs. 460.0 m, P = 0.005) and lower SpO2 during the 6MWT (90.5 vs. 95.0%, P = 0.001). The comparisons of variables used in the GAP model between the two groups are presented in Table S1. Clinical characteristics of patients with SARD-ILD according to sex are presented in Table S2.
Fig. 1.
Flow diagram of study protocol. SARD-ILD, systemic autoimmune rheumatic diseases-interstitial lung disease; FVC, forced vital capacity; DLco, diffusing capacity of the lungs for carbon monoxide; 6MWD (m), 6-min walk distance; RA, rheumatoid arthritis; DM/PM, dermatomyositis and polymyositis; MCTD, mixed connective tissue disease; SjS, Sjogren’s syndrome; SLE, systemic lupus erythematosus; SSc, systemic sclerosis; UCTD, undifferentiated connective tissue disease.
Table 1.
Clinical characteristics of patients with systemic autoimmune rheumatic diseases interstitial lung disease according to the 3-year survival.
| Characteristics | Total (n = 142) | Non-survivor (n = 28) | Survivor (n = 114) | p-value |
|---|---|---|---|---|
| Age | 61 (52.0–68.0) | 67 (63.5–70.5) | 59 (50.0–67.0) | 0.005 |
| Male sex | 52 (36.6%) | 15 (53.6%) | 37 (32.5%) | 0.063 |
| Ever smoker | 52 (36.6%) | 16 (57.1%) | 36 (31.6%) | 0.022 |
| BMI | 23.1 (20.9–25.3) | 23.4 (19.9–26.2) | 23.1 (21.1–25.2) | 0.967 |
| HRCT pattern | 0.001 | |||
| UIP pattern | 38 (26.8%) | 15 (53.6%) | 23 (20.2%) | |
| Non-UIP pattern | 104 (73.2%) | 13 (46.4%) | 91 (79.8%) | |
| Underlying SARD | ||||
| RA | 62 (43.7%) | 15 (53.6%) | 47 (41.2%) | 0.333 |
| DM/PM | 22 (15.5%) | 4 (14.3%) | 18 (15.8%) | > 0.999 |
| MCTD | 2 (1.4%) | 0 (0.0%) | 2 (1.8%) | > 0.999 |
| Sjogren’s syndrome | 13 (9.2%) | 2 (7.1%) | 11 (9.6%) | 0.963 |
| SLE | 17 (12.0%) | 2 (7.1%) | 15 (13.2%) | 0.580 |
| SSc | 22 (15.5%) | 5 (17.9%) | 17 (14.9%) | 0.925 |
| UCTD | 4 (2.8%) | 0 (0.0%) | 4 (3.5%) | 0.713 |
| Baseline PFT | ||||
| FVC, predicted % | 73.5 (59.0–84.0) | 58.0 (50.0–68.0) | 76.0 (64.0–86.0) | < .001 |
| DLco, predicted % | 52.5 (41.0–65.0) | 40.5 (34.5–51.5) | 58.0 (46.0–68.0) | < .001 |
| 6MWD (m) | 451.5 (360.0–505.0) | 390.0 (257.5–483.0) | 460.0 (390.0–510.0) | 0.005 |
| Minimum SpO2 (%), 6MWT | 94.0 (90.0–97.0) | 90.5 (84.0–95.0) | 95.0 (92.0–97.0) | 0.001 |
Significant values are in italics.
Data are presented as the median (interquartile range) or number (percentage).
BMI, body mass Index; HRCT, high-resolution computed tomography; UIP, usual interstitial pneumonia; RA, rheumatoid arthritis; DM/PM, dermatomyositis and polymyositis; MCTD, mixed connective tissue disease; SARD, systemic autoimmune rheumatic disease; SLE, systemic lupus erythematosus; SSc, systemic sclerosis; UCTD, undifferentiated connective tissue disease; PFT, pulmonary function test; FVC, forced vital capacity; DLco, diffusing capacity of the lungs for carbon monoxide; 6MWD (m), 6-min walk distance; SpO2, saturation of peripheral oxygen.
Treatment and clinical outcomes
Table 2 presents the treatment and clinical outcomes of non-survivors and survivors over the 3-year follow-up. Among the individuals included in this study, 117 (82.4%) received steroid treatment, 4 (2.8%) were treated with pirfenidone, and 3 (2.1%) received nintedanib. One patient (0.7%) received pirfenidone followed by nintedanib. Although there was no significant difference in the proportion of steroid use or the maximal dose between the two groups, treatment duration was significantly shorter in the non-survival group than in the survival group (10.5 vs. 16.0 months, P = 0.021). The proportion of PPF was significantly higher in the non-survival group than in the survival group (39.3 vs. 8.8%, P < 0.001) (Table 2).
Table 2.
Treatment and clinical course of patients with systemic autoimmune rheumatic diseases according to the 3-year survival.
| Characteristics | Total (n = 142) | Non-survivor (n = 28) | Survivor (n = 114) | p-value |
|---|---|---|---|---|
| Treatment | ||||
| Steroid | 117 (82.4%) | 26 (92.9%) | 91 (79.8%) | 0.178 |
| Maximal dose (PD, mg) | 30.0 (7.5–50.0) | 30.0 (15.0–40.0) | 30.0 (7.5–50.0) | 0.525 |
| Duration (months) | 12.0 (6.0–30.0) | 10.5 (4.0–16.0) | 16.0 (8.0–36.5) | 0.021 |
| Immunosuppressant | 88 (62.0%) | 19 (67.9%) | 69 (60.5%) | 0.618 |
| Mycophenolate mofetil | 36 (25.4%) | 10 (35.7%) | 26 (22.8%) | 0.244 |
| Azathioprine | 34 (23.9%) | 6 (21.4%) | 28 (24.6%) | 0.920 |
| Cyclosporine | 6 (4.2%) | 2 (7.1%) | 4 (3.5%) | 0.740 |
| Tacrolimus | 14 (9.9%) | 3 (10.7%) | 11 (9.6%) | > 0.999 |
| Antifibrotics* | 8 (5.6%) | 3 (10.7%) | 5 (4.4%) | 0.399 |
| Pirfenidone | 5 (3.5%) | 1 (3.6%) | 4 (3.5%) | 1.000 |
| Nintedanib | 4 (2.8%) | 2 (7.1%) | 2 (1.8%) | 0.365 |
| Clinical outcomes | ||||
| PPF | 21 (14.8%) | 11 (39.3%) | 10 (8.8%) | < .001 |
Significant values are in italics.
PD, prednisolone; PPF, progressive pulmonary fibrosis.
*One patient received pirfenidone followed by nintedanib.
Factors affecting mortality
In the univariate Cox proportional hazards analysis, older age, male sex, lower lung function, poorer exercise capacity, and UIP pattern were significantly associated with 3-year mortality in patients with SARD-ILD (Table 3). Moreover, advanced GAP stage and PPF were also related to 3-year mortality. In the multivariate analysis, advanced GAP stage (stage I vs. II, hazard ratio [HR], 10.547; 95% confidence interval (CI) 3.792–29.340, P < 0.001 and stage I vs. III, HR 17.594; 95% CI 4.351–71.145, P < 0.001), exertional hypoxia and UIP pattern (HR 2.506; 95% CI 1.004–6.259, P = 0.049), and PPF (HR 3.481; 95% CI 1.618–7.490, P = 0.001) were independently associated with 3-year mortality (Table 3).
Table 3.
Prognostic factors for 3-year mortality in patients with SARD-ILD.
| Univariate | Multivariate | |||
|---|---|---|---|---|
| Hazard ratio (95% CI) | p-value | Hazard ratio (95% CI) | p-value | |
| Age† | 1.055 (1.014–1.097) | 0.007 | ||
| Male sex | 2.193 (1.043–4.612) | 0.038 | ||
| BMI† | 0.994 (0.889–1.099) | 0.900 | ||
| *GAP stage | ||||
| Stage I | Ref | Ref | ||
| Stage II | 13.411 (4.984–36.090) | < .001 | 10.547 (3.792–29.340) | < .001 |
| Stage III | 30.114 (7.841–115.653) | < .001 | 17.594 (4.351–71.145) | < .001 |
| FVC, predicted %† | 0.949 (0.928–0.970) | < .001 | ||
| DLco, predicted %† | 0.943 (0.920–0.968) | < .001 | ||
| 6MWD (< 250 m) | 3.402 (1.290–8.969) | 0.013 | ||
| Exertional hypoxia (SpO2 < 88%) | 4.513 (2.109–9.661) | < .001 | ||
| UIP pattern on HRCT | 3.808 (1.807–8.027) | < .001 | ||
| *Exertional hypoxia and UIP pattern | 8.604 (3.580–20.677) | < .001 | 2.506 (1.004–6.259) | 0.049 |
| *PPF | 4.264 (1.995–9.111) | < .001 | 3.481 (1.618–7.490) | 0.001 |
Significant values are in italics.
SARD-ILD, systemic autoimmune rheumatic diseases interstitial lung disease; BMI, body mass index; GAP model, gender, age, physiology model; FVC, forced vital capacity; DLco, diffusing capacity of the lungs for carbon monoxide; 6MWD (m), 6-min walk distance; SpO2, saturation of peripheral oxygen; UIP, usual interstitial pneumonia; HRCT, high-resolution computed tomography; PPF, progressive pulmonary fibrosis.
*Included in the multivariate analysis.
†Indicated continuous variables.
Survival analysis
The KM survival curves, stratified by GAP stage, are presented in Fig. 2A. As the GAP stage increased, survival outcomes progressively worsened, with median survival times as follows (P < 0.001): GAP stage I (not reached), stage II (31.2 months), and stage III (19.8 months). KM analysis stratifying patients into non-advanced (GAP stage I) and advanced GAP (GAP stage II or III) groups are presented in Fig. 2B. The results indicated that the non-advanced GAP group exhibited better survival rates than the advanced GAP group (median survival: not reached vs. 28.1 months, P < 0.001). In addition to GAP staging, we examined other prognostic factors. Patients with a UIP pattern on HRCT demonstrated a higher mortality rate than those without this pattern (median survival: not reached vs. not reached, P < 0.001, Fig. 3A). Additionally, patients with exertional hypoxia on the 6MWT had significantly shorter survival than those without (median survival: 32.0 months vs. not reached, P < 0.001, Fig. 3B). Patients with both exertional hypoxia and a UIP pattern showed poorer prognosis (median survival: 18.0 months vs. not reached, P < 0.001, Fig. 3C).
Fig. 2.
3-Year survival curves in patients with SARD-ILD according to the GAP model. (A) Comparison of survival curves according to the GAP model. (B) Comparison of survival curves between patients with GAP stage I and GAP stages II or III. SARD-ILD, systemic autoimmune rheumatic diseases interstitial lung disease; GAP model, gender, age, physiology model.
Fig. 3.
3-Year survival curves in patients with SARD-ILD according to other prognostic factors. (A) Comparison of survival curves according to the CT pattern. (B) Comparison of survival curves according to the exertional hypoxia during 6MWT. (C) Comparison of survival curves based on exertional hypoxia and the UIP pattern. SARD-ILD, systemic autoimmune rheumatic diseases- interstitial lung disease; UIP, usual interstitial pneumonia; 6MWT, 6-min walk test.
Subgroup analysis of patients with GAP stage II or III
Baseline characteristics, categorized as advanced and non-advanced, are presented in Table S3. Patients in the advanced GAP group were older, had a higher proportion of male and smoking history, and had a higher proportion of the UIP pattern. Additionally, physiological variables, including FVC, DLco, 6MWD, and minimum SpO₂ during the 6MWT, were lower in the advanced GAP group (Table S3).
Among patients with advanced GAP (n = 45), 23 (51%) died within 3 years. Table 4 presents the prognostic factors for 3-year mortality in patients with advanced GAP stage. In the multivariable analysis, the combination of exertional hypoxia and UIP pattern was identified as independent prognostic indicators (HR 2.539; 95% CI 1.024–6.295, P = 0.044) along with PPF (HR 2.190; 95% CI 0.897–5.348, P = 0.085). Figure 4 shows that among patients with advanced GAP stage, those with both UIP pattern and exertional hypoxia had significantly worse prognosis compared to those without. (median survival: 18.0 vs. 31.2 months, P = 0.042).
Table 4.
Prognostic factors for 3-year mortality in patients with systemic autoimmune rheumatic diseases interstitial lung disease in advanced GAP (GAP stage II or III).
| Univariate | Multivariate | |||
|---|---|---|---|---|
| Hazard Ratio (95% CI) | p-value | Hazard Ratio (95% CI) | p-value | |
| Age | ||||
| ≤ 60 | Ref | |||
| 61–65 | 4.340 (0.866–21.755) | 0.074 | ||
| ≥ 65 | 2.792 (0.633–12.305) | 0.175 | ||
| Male sex | 0.984 (0.425–2.278) | 0.969 | ||
| FVC, predicted % | ||||
| > 75 | Ref | |||
| 50–75 | 2.169 (0.288–16.324) | 0.452 | ||
| < 50 | 1.389 (0.162–11.904) | 0.764 | ||
| DLco, predicted % | ||||
| > 55 | Ref | |||
| 36–55 | 3.430 (0.447–26.289) | 0.236 | ||
| ≤ 35 | 3.516 (0.443–27.877) | 0.234 | ||
| *PPF | 2.131 (0.875–5.191) | 0.096 | 2.190 (0.897–5.348) | 0.085 |
| *Exertional hypoxia and UIP pattern | 2.473 (1.002–6.102) | 0.049 | 2.539 (1.024–6.295) | 0.044 |
Significant values are in italics.
FVC, forced vital capacity; DLco, diffusing capacity of the lungs for carbon monoxide; PPF, progressive pulmonary fibrosis; UIP, usual interstitial pneumonia.
*Included in the multivariate analysis.
Fig. 4.
Comparison of overall survival curves in SARD-ILD patients with advanced GAP stage (II or III) according to the exertional hypoxia and UIP pattern. SARD-ILD, systemic autoimmune rheumatic diseases interstitial lung disease; GAP index, gender, age, physiology index; UIP, usual interstitial pneumonia.
Discussion
In this study, the GAP model, originally validated in IPF, has shown some utility in predicting prognosis in patients with SARD-ILD; however, other prognostic factors, such as exercise capacity, UIP pattern, and PPF, have also been identified. Notably, even among patients with advanced GAP stages (II or III), those with the UIP pattern and exertional hypoxia demonstrated worse prognosis. These findings indicate that predicting prognosis in patients with SARD-ILD requires a comprehensive approach, incorporating imaging features, exercise capacity, and clinical course assessment, in addition to the variables included in the original GAP model.
Previous studies have demonstrated the utility of the GAP model in SARD-ILD. Ryerson et al. demonstrated the utility of the ILD-GAP model in a cohort of 1,012 patients with various ILDs, including 281 patients with SARD-ILD4. Morisset et al. demonstrated that the GAP model could be used as a risk prediction tool in a cohort of 309 patients with RA-ILD across four academic centers (with a C-index of 0.753 for the original GAP model)9. However, SARD-ILD has different characteristics compared to IPF3, such as a higher proportion of female patients and a relatively younger age, suggesting that the GAP model may be less accurate in this population29. In the present study, only 4 of 142 patients met the criteria for GAP stage III, indicating that differences in demographics may make it challenging for patients with SARD-ILD to receive higher scores in the GAP model. Likewise in a study with 675 Chinese patients with SARD-ILD, the ILD-GAP model demonstrated a C-index of 0.652, whereas a predictive nomogram that incorporated various variables such as specific diagnosis, treatment, laboratory data, and CT patterns showed superior performance with a C-index of 0.76730. In our study, the presence of a UIP pattern on HRCT and exertional hypoxia during the 6MWT were prognostic factors in patients with SARD-ILD, particularly in those with advanced GAP stages. Patients with SARD-ILD can exhibit various radiologic patterns, with the NSIP pattern being the most common except in cases of RA-ILD31. Walsh et al. reported that patients with concordant UIP (radiologic and histologic UIP) had the highest mortality, whereas those with concordant NSIP showed the lowest mortality rates in a study of 168 patients with SARD-ILD32. Additionally, the UIP pattern is more common in RA-ILD than in other types of SARD-ILD33 and is associated with a poorer prognosis compared to those with a non-UIP pattern14,34. The valuable role of the 6MWT in assessing exercise capacity and providing prognostic insights has been demonstrated not only for patients with IPF35 but also for those with SARD-ILD12,13. The distance-oxygen-GAP (DO-GAP) index, which incorporates the 6MWD and exertional hypoxia into the original GAP index, has recently been proposed and validated in patients with IPF, highlighting the importance of the 6MWT28,36. Consistent with these previous studies, our study identified UIP patterns and poorer exercise capacity as prognostic factors in patients with SARD-ILD. Notably, in patients with advanced GAP stages, the presence of both UIP and exertional hypoxia might serve as an additional tool for prognostic classification. Further large-scale studies are needed to validate these findings. In our current study, the proportion of RA was higher in male patients compared to females. Although there were no significant differences in baseline pulmonary function between sexes, male patients were more likely to fall into higher GAP stages, which may partly reflect the influence of sex itself, as the GAP model includes male sex as a prognostic factor. Notably, RA is known to be more commonly associated with a UIP pattern compared to other types of SARD-ILD, which may be linked to a worse prognosis14,37. Therefore, the overrepresentation of RA in male patients may have contributed to the observed sex imbalance in mortality and disease severity. Given the retrospective nature of this study, our findings are not sufficient to determine sex-based prognostic differences in SARD-ILD, and further prospective research is warranted to explore this issue in more detail.
In addition to parameters at the time of ILD diagnosis, information obtained during the clinical course may also be important for predicting prognosis. PPF is associated with a worse prognosis in patients with SARD-ILD38–40, and our study confirmed these findings. Given that some medications have shown efficacy in SARD-ILD41,42, close monitoring of the clinical course and timely decision-making regarding medication use may be essential for improving prognosis and patient outcomes. In our study, non-survivors were found to have received a shorter duration of steroid treatment compared to survivors. The median follow-up duration was markedly shorter in non-survivors (13.4 months) than in survivors (36.0 months), which may partly explain the shorter cumulative steroid exposure in this group. However, this study was not designed to evaluate treatment efficacy or the causal impact of steroid duration on outcomes. Given the retrospective nature of our analysis and the absence of standardized guidelines regarding steroid dosing or treatment duration in SARD-ILD, these findings should be interpreted with caution. Prospective studies or randomized controlled trials are needed to clarify the optimal role of corticosteroids in this population.
Our study has several limitations. First, it was a retrospective cohort study conducted at a single center, and thus its generalizability may be limited. Second, our study may have resulted in selection bias, as it only included patients with available PFT, 6MWT, and HRCT data at the time of initial diagnosis. In addition, a limitation of our study is the lack of detailed smoking history, including pack-year data, due to the retrospective nature of data collection from medical records. As a result, we were unable to apply a standardized or minimal pack-year–based definition of “ever smokers,” which may have introduced some misclassification bias in smoking status. Third, we did not account for comorbidities that could impact prognosis. In addition, the exact cause of death could not be determined in many cases. As a result, it was difficult to ascertain whether the deaths were related to respiratory causes, which may limit the accuracy of our interpretation. Fourth, a detailed analysis of each SARD-ILD disease entity was not performed, and grouping UIP and exertional hypoxia may be considered arbitrary. Moreover, although there is ongoing debate, bronchoalveolar lavage or lung biopsy findings might have influenced patient’s characteristics and prognosis43. However, none of the patients in our study underwent these procedures, making it difficult to assess their potential impact. Finally, the SARD-ILD group consisted of heterogeneous disease entities, with a high proportion of RA-ILD (43.7%), which may differ from other SARD-ILDs. However, only 26.8% of the entire cohort showed a UIP pattern, and among the 62 patients with RA, only 24 patients (38.7%) had a confirmed UIP pattern, suggesting that the potential impact of this issue may be limited. However, further studies with larger populations and specific SARD subtypes are warranted. Despite these limitations, a key strength of this study is its comprehensive consideration of physiologic variables, exercise capacity, imaging patterns, and clinical course.
Conclusion
Although the GAP model can be applied to prognostic prediction in SARD-ILD, other prognostic factors such as CT imaging patterns, exercise capacity, and clinical course were also identified. Notably, in patients with advanced GAP stages, the presence of both UIP pattern and exertional hypoxia provided additional prognostic stratification. These findings highlight the need for a comprehensive approach to the management of patients with SARD-ILD.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Abbreviations
- CI
Confidence interval
- GAP model
Gender, age, physiology model
- DLco
Diffusing capacity of the lung for carbon monoxide
- DO-GAP
Distance-oxygen-GAP
- FVC
Forced vital capacity
- HR
Hazard ratio
- HRCT
High resolution computed tomography
- IPF
Idiopathic pulmonary fibrosis
- IQR
Interquartile range
- ILD
Interstitial lung disease
- KM
Kaplan–Meier
- PPF
Progressive pulmonary fibrosis
- RA
Rheumatoid arthritis
- SARDs
Systemic autoimmune rheumatic diseases
- SpO2
Saturation of peripheral oxygen
- 6MWD
6-Minute walk distance
- 6MWT
6-Minute walk test
- SSc
Systemic sclerosis
- UIP
Usual interstitial pneumonia
Author contributions
HCK takes responsibility for the content of the manuscript, including the data and analysis. YJH and HCK contributed to the study conception and design. YJH, JKL, JHL, SML, SH, and HCK contributed to the acquisition of data. YJH, JKL, JHL, SML, SH, and HCK contributed to the analysis and interpretation of data. YJH and HCK drafted the manuscript. All authors read and approved the final manuscript.
Funding
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT, RS-2025-00556764).
Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Ethical approval
The Institutional Review Board of Asan Medical Center endorsed this study protocol (IRB no.2022-1564, 2023-1394). The Institutional Review Board of Asan Medical Center waived the need for informed consent because of anonymous clinical data and retrospective study design.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Wijsenbeek, M., Suzuki, A. & Maher, T. M. Interstitial lung diseases. Lancet400, 769–786 (2022). [DOI] [PubMed] [Google Scholar]
- 2.Jee, A. S. et al. Diagnosis and management of connective tissue disease-associated interstitial lung disease in Australia and New Zealand: A position statement from the Thoracic Society of Australia and New Zealand. Respirology26, 23–51 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Demoruelle, M. K., Mittoo, S. & Solomon, J. J. Connective tissue disease-related interstitial lung disease. Best Pract. Res. Clin. Rheumatol.30, 39–52 (2016). [DOI] [PubMed] [Google Scholar]
- 4.Ryerson, C. J. et al. Predicting survival across chronic interstitial lung disease: The ILD-GAP model. Chest145, 723–728 (2014). [DOI] [PubMed] [Google Scholar]
- 5.Goh, N. S. et al. Interstitial lung disease in systemic sclerosis: a simple staging system. Am. J. Respir. Crit. Care Med.177, 1248–1254 (2008). [DOI] [PubMed] [Google Scholar]
- 6.Jacob, J. et al. Predicting outcomes in rheumatoid arthritis related interstitial lung disease. Eur. Respir. J.53, 1800869 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ley, B. et al. A multidimensional index and staging system for idiopathic pulmonary fibrosis. Ann. Intern. Med.156, 684–691 (2012). [DOI] [PubMed] [Google Scholar]
- 8.Lee, S. H. et al. Predicting survival of patients with idiopathic pulmonary fibrosis using GAP score: A nationwide cohort study. Respir. Res.17, 131 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Morisset, J. et al. The performance of the GAP model in patients with rheumatoid arthritis associated interstitial lung disease. Respir. Med.127, 51–56 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zamora-Legoff, J. A., Krause, M. L., Crowson, C. S., Ryu, J. H. & Matteson, E. L. Patterns of interstitial lung disease and mortality in rheumatoid arthritis. Rheumatology (Oxford)56, 344–350 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Nurmi, H. M. et al. Are risk predicting models useful for estimating survival of patients with rheumatoid arthritis-associated interstitial lung disease?. BMC Pulm. Med.17, 16 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Harari, S. et al. The 6-min walk test as a primary end-point in interstitial lung disease. Eur. Respir. Rev.31, 220087 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Khor, Y. H. et al. Trajectories and prognostic significance of 6-minute walk test parameters in fibrotic interstitial lung disease: A multicenter study. Chest163, 345–357 (2023). [DOI] [PubMed] [Google Scholar]
- 14.Park, J. H. et al. Prognosis of fibrotic interstitial pneumonia: Idiopathic versus collagen vascular disease-related subtypes. Am. J. Respir. Crit. Care Med.175, 705–711 (2007). [DOI] [PubMed] [Google Scholar]
- 15.Chan, C., Ryerson, C. J., Dunne, J. V. & Wilcox, P. G. Demographic and clinical predictors of progression and mortality in connective tissue disease-associated interstitial lung disease: A retrospective cohort study. BMC Pulm. Med.19, 192 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.van den Hoogen, F. et al. 2013 classification criteria for systemic sclerosis: An American college of rheumatology/European league against rheumatism collaborative initiative. Ann. Rheum. Dis.72, 1747–1755 (2013). [DOI] [PubMed] [Google Scholar]
- 17.Aletaha, D. et al. 2010 Rheumatoid arthritis classification criteria: An American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum.62, 2569–2581 (2010). [DOI] [PubMed] [Google Scholar]
- 18.Radner, H., Neogi, T., Smolen, J. S. & Aletaha, D. Performance of the 2010 ACR/EULAR classification criteria for rheumatoid arthritis: A systematic literature review. Ann. Rheum. Dis.73, 114–123 (2014). [DOI] [PubMed] [Google Scholar]
- 19.Petri, M. et al. Derivation and validation of the Systemic Lupus International Collaborating Clinics classification criteria for systemic lupus erythematosus. Arthritis Rheum.64, 2677–2686 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.John, K. J. et al. Clinical and immunological profile of mixed connective tissue disease and a comparison of four diagnostic criteria. Int. J. Rheumatol.2020, 9692030 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bohan, A. & Peter, J. B. Polymyositis and dermatomyositis (first of two parts). N Engl J Med292, 344–347 (1975). [DOI] [PubMed] [Google Scholar]
- 22.Mosca, M., Neri, R. & Bombardieri, S. Undifferentiated connective tissue diseases (UCTD): A review of the literature and a proposal for preliminary classification criteria. Clin. Exp. Rheumatol.17, 615–620 (1999). [PubMed] [Google Scholar]
- 23.Shiboski, C. H. et al. 2016 American College of Rheumatology/European league against rheumatism classification criteria for primary Sjogren’s syndrome: A consensus and data-driven methodology involving three international patient cohorts. Arthritis Rheumatol.69, 35–45 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Raghu, G. et al. Diagnosis of idiopathic pulmonary fibrosis. An official ATS/ERS/JRS/ALAT clinical practice guideline. Am. J. Respir. Crit. Care Med.198, e44–e68 (2018). [DOI] [PubMed] [Google Scholar]
- 25.Raghu, G. et al. Idiopathic pulmonary fibrosis (an update) and progressive pulmonary fibrosis in adults: An official ATS/ERS/JRS/ALAT clinical practice guideline. Am. J. Respir. Crit. Care Med.205, e18–e47 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.du Bois, R. M. et al. 6-Minute walk distance is an independent predictor of mortality in patients with idiopathic pulmonary fibrosis. Eur. Respir. J.43, 1421–1429 (2014). [DOI] [PubMed] [Google Scholar]
- 27.Jacobs, S. S. et al. Home oxygen therapy for adults with chronic lung disease. An official American thoracic society clinical practice guideline. Am. J. Respir. Crit. Care Med.202, e121–e141 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Chandel, A., Pastre, J., Valery, S., King, C. S. & Nathan, S. D. Derivation and validation of a simple multidimensional index incorporating exercise capacity parameters for survival prediction in idiopathic pulmonary fibrosis. Thorax78, 368–375 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kim, E. S. et al. Validation of the GAP score in Korean patients with idiopathic pulmonary fibrosis. Chest147, 430–437 (2015). [DOI] [PubMed] [Google Scholar]
- 30.Sun, D. et al. Prediction of long-term mortality by using machine learning models in Chinese patients with connective tissue disease-associated interstitial lung disease. Respir. Res.23, 4 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Joy, G. M. et al. Prevalence, imaging patterns and risk factors of interstitial lung disease in connective tissue disease: A systematic review and meta-analysis. Eur. Respir. Rev.32, 220210 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Walsh, S. L. et al. Connective tissue disease related fibrotic lung disease: High resolution computed tomographic and pulmonary function indices as prognostic determinants. Thorax69, 216–222 (2014). [DOI] [PubMed] [Google Scholar]
- 33.Assayag, D. et al. Rheumatoid arthritis-associated interstitial lung disease: Radiologic identification of usual interstitial pneumonia pattern. Radiology270, 583–588 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kim, E. J. et al. Usual interstitial pneumonia in rheumatoid arthritis-associated interstitial lung disease. Eur. Respir. J.35, 1322–1328 (2010). [DOI] [PubMed] [Google Scholar]
- 35.du Bois, R. M. et al. Six-minute-walk test in idiopathic pulmonary fibrosis: test validation and minimal clinically important difference. Am. J. Respir. Crit. Care Med.183, 1231–1237 (2011). [DOI] [PubMed] [Google Scholar]
- 36.Chandel, A. et al. External validation and longitudinal application of the DO-GAP index to individualise survival prediction in idiopathic pulmonary fibrosis. ERJ Open Res.9, 00124 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Kim, E. J., Collard, H. R. & King, T. E. Jr. Rheumatoid arthritis-associated interstitial lung disease: the relevance of histopathologic and radiographic pattern. Chest136, 1397–1405 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Chiu, Y. H. et al. Predictors for progressive fibrosis in patients with connective tissue disease associated interstitial lung diseases. Respir. Med.187, 106579 (2021). [DOI] [PubMed] [Google Scholar]
- 39.Khor, Y. H., Farooqi, M., Hambly, N., Kolb, M. & Ryerson, C. J. Patient characteristics and survival for progressive pulmonary fibrosis using different definitions. Am. J. Respir. Crit. Care Med.207, 102–105 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Spagnolo, P. et al. Mechanisms of progressive fibrosis in connective tissue disease (CTD)-associated interstitial lung diseases (ILDs). Ann. Rheum. Dis.80, 143–150 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Tashkin, D. P. et al. Mycophenolate mofetil versus oral cyclophosphamide in scleroderma-related interstitial lung disease (SLS II): A randomised controlled, double-blind, parallel group trial. Lancet Respir. Med.4, 708–719 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Distler, O. et al. Nintedanib for systemic sclerosis-associated interstitial lung disease. N. Engl. J. Med.380, 2518–2528 (2019). [DOI] [PubMed] [Google Scholar]
- 43.Tomassetti, S. et al. Bronchoalveolar lavage and lung biopsy in connective tissue diseases, to do or not to do?. Ther. Adv. Musculoskelet. Dis.13, 1759720x211059605 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.