Clinical significance of inflammation-coagulation-autoantibody triad in systemic sclerosis-associated interstitial lung disease: a retrospective study Authors: Ronghong Guo, Xianzhe Zhou, Jiaxin Chen, Jinfang Gao, Yanli Yang, Ke Xu Published: 2025 ABSTRACT BackgroundInterstitial lung disease (ILD) is highly prevalent among systemic sclerosis (SSc) patients and is the leading cause of mortality. This study aims to evaluate the independent and combined value of inflammatory markers, coagulation parameters, and autoantibodies for the presence of SSc-ILD.MethodsA total of 177 patients with SSc were enrolled in this study, including 116 patients with SSc-ILD and 61 without ILD. Multivariate logistic regression was performed to identify independently associated factors for SSc-ILD. The identification efficiency of individual and combined biomarkers was assessed using receiver operating characteristic curves, with the DeLong test applied to compare differences in area under the curve.ResultsThe incidence of dcSSc, Raynaud’s phenomenon, and digestive system involvement was significantly higher in the SSc-ILD group. Four stepwise models were constructed, and the results of the full-factor model indicated that elevated NLR (neutrophil to lymphocyte ratio), prolonged PT (prothrombin time), and anti-SCL-70 antibody positivity were independent associated factors for SSc-ILD. The combined model of NLR, PT, and anti-SCL-70 yielded an AUC of 0.879 (95% CI: 0.828–0.930), which was significantly greater than any individual marker (P < 0.001), with a sensitivity of 82.6% and a specificity of 85.0%.ConclusionNLR, PT, and anti-SCL-70 antibody together form a triad of independent associated factors for SSc-ILD, demonstrating substantial discriminatory potential when considered as a composite indicator. Background Systemic sclerosis (SSc) is a connective tissue disease characterized by skin and visceral fibrosis, microvascular damage, and autoimmune dysfunction. It can involve multiple organs, leading to substantial impairment in patients’ health, quality of life, and life expectancy [1]. Interstitial lung disease (ILD) is the most common internal organ complication of SSc, with an incidence of 65% to 80%, and has become the leading cause of mortality in SSc patients, with a 10-year mortality rate reaching up to 40% [2]. The natural course of SSc-ILD is characterized by significant heterogeneity, and it often progresses insidiously in the early stage, posing considerable challenges to clinical diagnosis and management. Although high-resolution computed tomography (HRCT) remains the current gold-standard imaging modality for detecting ILD, there is an urgent need for more optimal serological detection tools in clinical practice, particularly serum biomarkers that combine convenience, non-invasiveness, and reliability [3]. Therefore, identifying and applying reliable biomarkers is essential to optimize personalized treatment strategies and improve patient survival rates and quality of life. The onset of SSc-ILD begins with environmental triggers causing damage to endothelial and epithelial cells in susceptible individuals. This initiates the release of active mediators that promote the infiltration and activation of inflammatory cells, ultimately resulting in fibroblast activation and abnormal deposition of the extracellular matrix [4]. Anti-nuclear antibodies (ANA) can be detected in over 95% of SSc patients [5]. Anti-centromere antibody (ACA), anti-Scl-70 (anti-Scleroderma 70 antibody, also known as anti-topoisomerase I antibody), and anti-RNA polymerase III antibody (ARA) were present in more than 50% of patients and serve as core predictors of organ involvement. Anti-Scl-70-positive patients have a significantly higher risk of developing clinically significant ILD compared to those positive for ACA or ARA [6]. This association may be mediated through antibody-induced fibroblast activation and alveolar epithelial cell injury. Emerging inflammatory markers such as the neutrophil-to-lymphocyte ratio (NLR) and the neutrophil-to-platelet ratio (NPR), derived from complete blood counts, have proven effective in identifying the existence and severity of multiple autoimmune diseases, including rheumatoid arthritis (RA) [7], psoriasis [8], and systemic lupus erythematosus (SLE) [9]. In SSc, neutrophils, lymphocytes, and platelets have been shown to be involved in disease pathophysiology, with associations between elevated levels and more severe disease progression as well as increased mortality [10]. NLR has been particularly closely linked to the progression of pulmonary diseases [11]. The generation of thrombin through activation of the coagulation cascade is a well-established mechanism of lung injury, which has been widely documented in SSc-ILD [12, 13]. Abnormalities in coagulation may contribute to both vascular lesions and fibrotic processes in SSc. However, a single marker cannot encompass the multiple inflammatory-coagulation-autoimmune pathological processes of SSc-ILD; thus, identifying panel biomarkers holds unique value. This study aims to systematically evaluate the associations of inflammatory markers, coagulation parameters, and autoantibodies with SSc-ILD as well as their clinical significance by means of retrospective analysis, and to further explore the evaluative value of their combined application. It is expected that the study results will provide clinical references for elucidating the pathogenesis and pathophysiological processes of SSc-ILD, and yield research enlightenments for promoting the optimization of clinical diagnosis and treatment strategies. Materials and methods Study participants A retrospective cohort study was conducted, involving 177 patients diagnosed with SSc who were consecutively enrolled between January 2016 and April 2025 at the Department of Rheumatology and Immunology, Shanxi Bethune Hospital. The inclusion criteria included: (1) individuals aged between 18 and 80 years; (2) diagnosis based on the 2013 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) classification criteria [14]. The exclusion criteria were: (1) presence of other connective tissue diseases (e.g., overlap syndrome, mixed connective tissue disease); (2) patients with malignant tumors, active infections, hematological disorders, severe hepatic or renal insufficiency, pregnancy, or lactation; (3) patients who had received drugs within 3 months before enrollment that could interfere with the study results: glucocorticoids (e.g., prednisone, methylprednisolone); Immunosuppressive agents (e.g., cyclophosphamide, mycophenolate mofetil); Antifibrotic drugs (e.g., nintedanib, pirfenidone); Drugs that affect coagulation parameters (e.g., aspirin). The research was conducted in strict compliance with the ethical principles of the Declaration of Helsinki, and the study protocol received approval from the Ethics Committee of Shanxi Bethune Hospital (Approval No.: LYLL-2024-003/PJ03). Collection and evaluation of clinical data Two rheumatologists independently collected data on demographic characteristics, clinical manifestations, and organ involvement assessments of all participants through a systematic review of electronic medical records and paper files. Specific collection indicators and diagnostic criteria are outlined below. Basic information encompassed gender, age, body mass index (BMI), and disease duration, defined as the time from the first occurrence of non-Raynaud phenomenon. The extent of skin involvement was evaluated using the modified Rodnan Skin Score (mRSS) [15]. Raynaud’s phenomenon is defined as a paroxysmal spasm of the small arteries in the extremities triggered by cold exposure or emotional stress, resulting in a characteristic triphasic color change in the fingers (and toes), including pallor, cyanosis, and erythema, often accompanied by numbness and pain [16]. Cardiovascular diseases in SSc were assessed using a combination of echocardiography (to detect left ventricular systolic or diastolic dysfunction), electrocardiography (commonly showing left bundle branch block or atrioventricular block), cardiac magnetic resonance imaging, and biomarkers such as B-type natriuretic peptide. Diagnoses were established by experienced clinicians according to recognized diagnostic criteria. Additionally, the existence of ILD was evaluated through high-resolution computed tomography (HRCT) scans of the thorax (GE DiscoveryRT). HRCT observations encompassed reticular opacities, ground-glass opacities, traction bronchiectasis, and honeycombing, primarily observed in the lower regions of both lungs. All HRCT images were independently reviewed by at least two rheumatologists and cardiothoracic radiologists who were blinded to the group assignments. Potential differential diagnoses, including infection, pulmonary embolism, and lung cancer, were systematically excluded. The clinical diagnosis of ILD was established based on a comprehensive evaluation of imaging findings, respiratory symptoms and physical signs, as well as laboratory results, with auxiliary assessments performed when appropriate [17]. Gastrointestinal involvement was characterized by the occurrence of symptoms like acid reflux or heartburn, or the confirmation of reflux esophagitis via gastrointestinal endoscopy. Hematologic involvement was identified through routine blood tests, often revealing abnormalities in red blood cell, white blood cell, or platelet counts or function. Musculoskeletal (MSK) involvement was defined and documented by the attending rheumatologist based on clinical assessment, including findings such as joint swelling, deformity, arthralgia, contracture, or tendon friction rubs [18]. Disease activity was measured using the European Scleroderma Study Group (EScSG) disease activity index [19], which incorporates seven weighted parameters, including skin progression, inflammatory markers, and worsening cardiopulmonary function. A cumulative score of ≥ 3 indicates active disease. Specimens were collected and tested Concurrently with the gathering of clinical data, venous blood samples were collected from all enrolled participants on the morning following admission, after a fasting period. The samples were collected in accordance with a standardized protocol to ensure the consistency of baseline data. These samples were stored in EDTA-K2 and heparin anticoagulant tubes. Each sample was labeled with essential information, including sample ID, collection time, patient number, clinical diagnosis, and intended use. Following centrifugation, serum and plasma samples were promptly packaged and stored at −80℃ in the research refrigerator to prevent cross-contamination. Hematological parameters, including white blood cell count, neutrophil count, lymphocyte count, platelet count (PLT), and erythrocyte sedimentation rate (ESR), were measured using an automated hematology analyzer along with its compatible reagents. Derived indicators, including the NLR and NPR, were calculated using the formula: NLR = neutrophil count (×10⁹/L)/lymphocyte count (×10⁹/L); NPR = neutrophil count (×10⁹/L)/platelet count (×10⁹/L). C-reactive protein (CRP) levels were measured using the Beckman AU5800 automatic biochemical analyzer. Coagulation parameters, including prothrombin time (PT) and D-dimer levels, were assessed using an automatic coagulation analyzer. ANA, Anti-Scl-70 antibodies, and ACA were detected using an automatic chemiluminescence immunoassay analyzer based on magnetic particle chemiluminescence immunoassay technology. The original calibrators were applied for single/multi-point calibration after daily startup, with deionized water serving as the blank control. Quality control was based on values established by our hospital’s clinical laboratory. During the use of respective quality control substances, the obtained measurements were expected to remain within the predefined acceptable range. Throughout the testing procedure, all equipment, assay kits, and reagents were kept at ambient temperature (20–25 °C), and the experiments were carried out under these controlled conditions. Statistical analysis All statistical analyses were conducted using SPSS 26.0 (IBM, USA), with the significance level set at α = 0.05 (two-tailed test). Measurement data were assessed for normality; variables with a normal distribution were presented as mean ± standard deviation (x̄ ± SD), and group comparisons were performed using the independent samples t-test. Data that did not follow a normal distribution were summarized as median (interquartile range) [M (P25, P75)], and the Mann-Whitney U test was applied for group comparisons. Categorical data were expressed as frequency and percentage [n (%)], and comparisons between groups were conducted using the chi-square test or Fisher’s exact test, as appropriate. To evaluate relationships between variables, Spearman’s rank correlation was used. Variables with p < 0.05 in univariate analysis were included in a multivariate logistic regression model based on the Likelihood Ratio method to identify independent predictors of the outcome variables “SSc” and “SSc-ILD”. The diagnostic performance was evaluated using receiver operating characteristic (ROC) curves and the corresponding area under the curve (AUC), with AUC comparisons performed using the DeLong test. All statistical tests were two-sided, and a p-value < 0.05 was considered statistically significant. Results Baseline characteristics of SSc patients A total of 177 SSc individuals were enrolled in this study. Detailed demographic and baseline clinical characteristics of the patients are summarized in Table 1. ILD is a common complication in SSc, with an incidence rate of 65.54%. The presence of ANA is observed in the majority of SSc patients.Table 1Basic characteristics of the SSc patients (N = 177)CharacteristicsSSc patients (N = 177)Demographic characteristicsAge, years55.47 ± 11.96Sex-female, n (%)146 (82.49%)BMI, kg/m221.60(19.12, 24.25)Clinical CharacteristicsDisease duration, years4.33(1.25, 10.00)Diffuse cutaneous SSc, n (%)159 (89.83%)Modified Rodnan Skin Score, units8.00(5.00, 15.00)Active patients with disease activity index > 3 EScSG, n (%)96 (54.24%)Raynaud’s phenomenon, n (%)144 (81.36%)Cardiovascular diseases, n (%)31 (17.51%)Interstitial lung disease, n (%)116(65.54%)Gastrointestinal involvement, n (%)80 (45.20%)MSK involvement n (%)58 (32.77%)Hematologic involvement, n (%)22 (12.43%)Antinuclear antibody, n (%)163 (92.09%)Anti-Scl-70, n (%)69 (38.93%)ACA, n (%)43 (24.29%)Abbreviation: BMI body mass index, EScSG European Scleroderma Study Group, SSc systemic sclerosis, Anti-Scl-70 Anti-Scleroderma 70 antibody, ACA Anti-Centromere Protein AntibodyContinuous variable with a normal or skewed distribution was presented as mean ± SD or median (25th-75th percentile), and categorical variables were presented as numbers (percentages) Table 1Basic characteristics of the SSc patients (N = 177)CharacteristicsSSc patients (N = 177)Demographic characteristicsAge, years55.47 ± 11.96Sex-female, n (%)146 (82.49%)BMI, kg/m221.60(19.12, 24.25)Clinical CharacteristicsDisease duration, years4.33(1.25, 10.00)Diffuse cutaneous SSc, n (%)159 (89.83%)Modified Rodnan Skin Score, units8.00(5.00, 15.00)Active patients with disease activity index > 3 EScSG, n (%)96 (54.24%)Raynaud’s phenomenon, n (%)144 (81.36%)Cardiovascular diseases, n (%)31 (17.51%)Interstitial lung disease, n (%)116(65.54%)Gastrointestinal involvement, n (%)80 (45.20%)MSK involvement n (%)58 (32.77%)Hematologic involvement, n (%)22 (12.43%)Antinuclear antibody, n (%)163 (92.09%)Anti-Scl-70, n (%)69 (38.93%)ACA, n (%)43 (24.29%)Abbreviation: BMI body mass index, EScSG European Scleroderma Study Group, SSc systemic sclerosis, Anti-Scl-70 Anti-Scleroderma 70 antibody, ACA Anti-Centromere Protein AntibodyContinuous variable with a normal or skewed distribution was presented as mean ± SD or median (25th-75th percentile), and categorical variables were presented as numbers (percentages) Basic characteristics of the SSc patients (N = 177) Abbreviation: BMI body mass index, EScSG European Scleroderma Study Group, SSc systemic sclerosis, Anti-Scl-70 Anti-Scleroderma 70 antibody, ACA Anti-Centromere Protein Antibody Continuous variable with a normal or skewed distribution was presented as mean ± SD or median (25th-75th percentile), and categorical variables were presented as numbers (percentages) The differences in inflammatory markers, coagulation parameters, and autoantibody profiles between the SSc-ILD and SSc-non-ILD As presented in Table 2, there was no significant difference in age or gender composition between the SSc-ILD group and the SSc-non-ILD group. Regarding clinical characteristics, the proportion of dcSSc was markedly higher in the SSc-ILD. Additionally, the mRSS was also higher in the SSc-ILD group, as was the disease duration. Importantly, the SSc-ILD group demonstrated significantly higher frequencies of Raynaud’s phenomenon and gastrointestinal involvement compared to the SSc-non-ILD group. No significant differences were observed in the prevalence of cardiovascular diseases, MSK involvement, or other complications between the two groups. In terms of inflammatory markers, the NLR in the SSc-ILD group was significantly higher than that in the control group [3.34 (2.12, 5.29) vs. 2.85 (2.37, 4.41), p < 0.001]. In terms of inflammatory indicators, the NLR in the SSc-ILD group was significantly higher [3.34 (2.12, 5.29) vs. 2.85 (2.37, 4.41), p < 0.001]. Additionally, levels of NPR [0.020 (0.015, 0.029) vs. 0.016 (0.012, 0.020), p < 0.001] and ESR [24.00 (14.00, 39.00) vs. 12.50 (8.00, 31.50) mm/h, p < 0.001] were also elevated.Table 2Comparison of characteristics between SSc-ILD and SSc-non-ILD cohortsCharacteristicsSSc-ILD(N = 116)SSc-non-ILD(N = 61)p ValuesAge, years56.33 ± 10.8953.84 ± 13.710.280Sex-female, n (%)100 (86.21%)46(75.41%)0.072BMI, kg/m221.60(18.78, 24.22)22.23(19.20, 24.62)0.395Disease duration, years6.50(3.00, 12.00)2.00(1.00, 7.00)0.001Diffuse cutaneous SSc, n (%)114 (98.28%)45 (73.77%)< 0.001Limited cutaneous SSc, n (%)2 (1.72%)16 (26.23%)< 0.001Modified Rodnan Skin Score, units9.00(6.00, 16.00)6.00(2.25, 12.00)0.020EScSG3.50(3.00, 5.00)3.50(2.00, 5.00)0.069Raynaud’s phenomenon, n (%)102 (87.93%)42 (68.85%)0.002Cardiovascular diseases, n (%)25 (21.55%)6 (9.84%)0.051Gastrointestinal involvement, n (%)60 (51.72%)20 (32.79%)0.016MSK involvement n (%)38 (32.76%)16 (26.23%)0.179Hematologic involvement, n (%)17 (14.66%)5 (8.20%)0.209Antinuclear antibody, n (%)111 (95.69%)52(85.26%)0.014Anti-Scl-70 antibodies, n (%)66 (56.90%)3 (4.92%)< 0.001Anti-centromere antibodies, n (%)13 (11.21%)30 (49.18%)< 0.001NLR3.34(2.12, 5.29)2.85(2.37, 4.41)< 0.001NPR0.020(0.015,0.029)0.016(0.012, 0.020)< 0.001PT, s11.85(11.20, 14.18)11.20(10.73, 12.28)< 0.001D-dimer, ng/mL161.00(96.25, 275.25)122.00(64.50, 202.50)0.012ESR, mm/h24.00(14.00, 39.00)12.50(8.00, 31.50)0.002CRP, mg/L6.95(2.90, 13.68)5.87(3.14, 11.20)0.649Abbreviation: BMI body mass index, SSc systemic sclerosis, EScSG European Scleroderma Study Group, NLR neutrophil to lymphocyte ratio, NPR neutrophil to platelet ratio, PT Prothrombin time, ESR Erythrocyte sedimentation rate, CRP C-reactive proteinContinuous variable with a normal or skewed distribution was presented as mean ± SD or median (25th-75th percentile), and categorical variables were presented as numbers (percentages). The Bolded value represents statistical significance Table 2Comparison of characteristics between SSc-ILD and SSc-non-ILD cohortsCharacteristicsSSc-ILD(N = 116)SSc-non-ILD(N = 61)p ValuesAge, years56.33 ± 10.8953.84 ± 13.710.280Sex-female, n (%)100 (86.21%)46(75.41%)0.072BMI, kg/m221.60(18.78, 24.22)22.23(19.20, 24.62)0.395Disease duration, years6.50(3.00, 12.00)2.00(1.00, 7.00)0.001Diffuse cutaneous SSc, n (%)114 (98.28%)45 (73.77%)< 0.001Limited cutaneous SSc, n (%)2 (1.72%)16 (26.23%)< 0.001Modified Rodnan Skin Score, units9.00(6.00, 16.00)6.00(2.25, 12.00)0.020EScSG3.50(3.00, 5.00)3.50(2.00, 5.00)0.069Raynaud’s phenomenon, n (%)102 (87.93%)42 (68.85%)0.002Cardiovascular diseases, n (%)25 (21.55%)6 (9.84%)0.051Gastrointestinal involvement, n (%)60 (51.72%)20 (32.79%)0.016MSK involvement n (%)38 (32.76%)16 (26.23%)0.179Hematologic involvement, n (%)17 (14.66%)5 (8.20%)0.209Antinuclear antibody, n (%)111 (95.69%)52(85.26%)0.014Anti-Scl-70 antibodies, n (%)66 (56.90%)3 (4.92%)< 0.001Anti-centromere antibodies, n (%)13 (11.21%)30 (49.18%)< 0.001NLR3.34(2.12, 5.29)2.85(2.37, 4.41)< 0.001NPR0.020(0.015,0.029)0.016(0.012, 0.020)< 0.001PT, s11.85(11.20, 14.18)11.20(10.73, 12.28)< 0.001D-dimer, ng/mL161.00(96.25, 275.25)122.00(64.50, 202.50)0.012ESR, mm/h24.00(14.00, 39.00)12.50(8.00, 31.50)0.002CRP, mg/L6.95(2.90, 13.68)5.87(3.14, 11.20)0.649Abbreviation: BMI body mass index, SSc systemic sclerosis, EScSG European Scleroderma Study Group, NLR neutrophil to lymphocyte ratio, NPR neutrophil to platelet ratio, PT Prothrombin time, ESR Erythrocyte sedimentation rate, CRP C-reactive proteinContinuous variable with a normal or skewed distribution was presented as mean ± SD or median (25th-75th percentile), and categorical variables were presented as numbers (percentages). The Bolded value represents statistical significance Comparison of characteristics between SSc-ILD and SSc-non-ILD cohorts Abbreviation: BMI body mass index, SSc systemic sclerosis, EScSG European Scleroderma Study Group, NLR neutrophil to lymphocyte ratio, NPR neutrophil to platelet ratio, PT Prothrombin time, ESR Erythrocyte sedimentation rate, CRP C-reactive protein Continuous variable with a normal or skewed distribution was presented as mean ± SD or median (25th-75th percentile), and categorical variables were presented as numbers (percentages). The Bolded value represents statistical significance The PT was significantly prolonged in the SSc-ILD group [11.85 (11.20, 14.18) vs. 11.20 (10.73, 12.28) seconds, p < 0.001], and D-dimer levels were significantly increased [161.00 (96.25, 275.25) vs. 122.00 (64.50, 202.50) ng/mL, p = 0.012], indicating activation of the coagulation system. Analysis of autoantibody profiles showed distinct variations in autoantibody presence between the groups. Specifically, the positive rate of Anti-SCL-70 antibody in the SSc-ILD group was 56.90%, which markedly exceeded that in the non-ILD group (4.92%, p < 0.001), whereas the positive rate of ACA was markedly lower (11.21% vs. 49.18%, p < 0.001). Correlation analysis of inflammatory markers, coagulation parameters, and autoantibody profiles in SSc-ILD patients The following significant association patterns were observed in SSc-ILD patients: (1) Inflammation/coagulation network activation association: NLR demonstrated a positive correlation with NPR (r = 0.583, p< 0.001), D-dimer (r = 0.243, p= 0.010), and ESR (r = 0.238, p= 0.016). Furthermore, a positive relationship was also found between D-dimer and ESR (r = 0.329, p= 0.001). (2) The pattern of autoantibody antagonism. Anti-Scl-70 was negatively correlated with ACA (r = −0.408, p< 0.001). Multivariate associated factor analysis of SSc-ILD Based on the aforementioned results, we further analyzed the independent associated factors for SSc-ILD using multivariate logistic regression and constructed four sequential models (Table 3). Model 1 was established as the inflammation-based risk model. After adjusting for ESR, NLR demonstrated a statistically significant association with SSc-ILD (OR = 1.417, 95% CI: 1.120–1.792, p= 0.004). Model 2, the inflammation-coagulation model, further incorporated coagulation parameters into Model 1. The results indicated that both NLR (OR = 1.375, 95% CI: 1.084–1.743, p= 0.009) and PT (OR = 1.507, 95% CI: 1.142–1.988, p= 0.004) were significantly associated with SSc-ILD and served as independent associated factors. Building upon Model 2, Model 3 (the inflammation-coagulation-organ involvement model) incorporated organ involvement features such as Raynaud's phenomenon and gastrointestinal involvement. In the model 3, both NLR (OR = 1.415, 95% CI 1.100–1.820, p= 0.007) and PT (OR = 1.554, 95% CI 1.155–2.091, p= 0.004) were independently associated. Model 4, the full-variable model, encompassed all clinical, laboratory, and autoantibody-related variables. The findings revealed that NLR (OR = 1.569, 95% CI 1.158–2.127, p= 0.004), PT (OR = 1.484, 95% CI 1.070–2.057, p= 0.018), and Anti-Scl-70 positivity (OR = 24.620, 95% CI 5.951–101.849, p< 0.001) were independent factors for SSc-ILD.Table 3Multivariate associated factor analysis of SSc-ILDCharacteristicsBetaOR(95% CI)p ValuesModel 1Constant−0.961NLR0.3491.417(1.120, 1.792)0.004Model 2Constant−5.808NLR0.3181.375(1.084, 1.743)0.009PT0.4101.507(1.142, 1.988)0.004Model 3Constant−7.356NLR0.3471.415(1.100, 1.820)0.007PT0.4411.554(1.155, 2.091)0.004Raynaud’s phenomenon1.0092.744(1.075, 7.004)0.035Gastrointestinal involvement0.9182.505(1.130, 5.555)0.024Model 4Constant−7.539NLR0.4511.569(1.158, 2.127)0.004PT0.3951.484(1.070, 2.057)0.018Raynaud’s phenomenon0.8802.411(0.797, 7.290)0.119Gastrointestinal involvement0.5951.814(0.676, 4.864)0.237Anti-Scl-703.20424.620(5.951, 101.849)< 0.001ACA−0.4290.651(0.228, 1.860)0.423OR Odds Ratio, CI Confidence Interval, ESR Erythrocyte sedimentation rate, NLR neutrophil to lymphocyte ratio, PT Prothrombin time, Anti-Scl-70 Anti-Scleroderma 70 antibody, ACA Anti-Centromere Protein AntibodyThe Bolded value represents statistical significanceModel 1 was adjusted according to NLR, ESR; Model 2 was further adjusted for PT, D-dimer; Model 3 was developed based on Model 2, with adjustments made to account for Raynaud's phenomenon and gastrointestinal involvement, Model 4 was further refined based on Model 3, with adjustments made for Anti-Scl-70 and ACA Table 3Multivariate associated factor analysis of SSc-ILDCharacteristicsBetaOR(95% CI)p ValuesModel 1Constant−0.961NLR0.3491.417(1.120, 1.792)0.004Model 2Constant−5.808NLR0.3181.375(1.084, 1.743)0.009PT0.4101.507(1.142, 1.988)0.004Model 3Constant−7.356NLR0.3471.415(1.100, 1.820)0.007PT0.4411.554(1.155, 2.091)0.004Raynaud’s phenomenon1.0092.744(1.075, 7.004)0.035Gastrointestinal involvement0.9182.505(1.130, 5.555)0.024Model 4Constant−7.539NLR0.4511.569(1.158, 2.127)0.004PT0.3951.484(1.070, 2.057)0.018Raynaud’s phenomenon0.8802.411(0.797, 7.290)0.119Gastrointestinal involvement0.5951.814(0.676, 4.864)0.237Anti-Scl-703.20424.620(5.951, 101.849)< 0.001ACA−0.4290.651(0.228, 1.860)0.423OR Odds Ratio, CI Confidence Interval, ESR Erythrocyte sedimentation rate, NLR neutrophil to lymphocyte ratio, PT Prothrombin time, Anti-Scl-70 Anti-Scleroderma 70 antibody, ACA Anti-Centromere Protein AntibodyThe Bolded value represents statistical significanceModel 1 was adjusted according to NLR, ESR; Model 2 was further adjusted for PT, D-dimer; Model 3 was developed based on Model 2, with adjustments made to account for Raynaud's phenomenon and gastrointestinal involvement, Model 4 was further refined based on Model 3, with adjustments made for Anti-Scl-70 and ACA Multivariate associated factor analysis of SSc-ILD OR Odds Ratio, CI Confidence Interval, ESR Erythrocyte sedimentation rate, NLR neutrophil to lymphocyte ratio, PT Prothrombin time, Anti-Scl-70 Anti-Scleroderma 70 antibody, ACA Anti-Centromere Protein Antibody The Bolded value represents statistical significance Model 1 was adjusted according to NLR, ESR; Model 2 was further adjusted for PT, D-dimer; Model 3 was developed based on Model 2, with adjustments made to account for Raynaud's phenomenon and gastrointestinal involvement, Model 4 was further refined based on Model 3, with adjustments made for Anti-Scl-70 and ACA The predictive performance of the integrated model for SSc-ILD Although the Anti-Scl-70 exhibited high specificity (95.08%), its sensitivity was relatively low at 56.90%, posing a considerable risk of missed diagnosis and limiting its effectiveness for early screening. To address this limitation and improve diagnostic accuracy, we constructed a multi-biomarker model that integrates NLR, PT, and the Anti-Scl-70, and assessed its discriminatory power was evaluated using ROC curve analysis (Figure 1). The AUC for NLR was 0.685 (95% CI: 0.604–0.766), with a sensitivity of 71.30% and specificity of 60.66%. For PT, the AUC was 0.691 (95% CI: 0.600–0.763), with a sensitivity of 41.38% and specificity of 88.33%. The combined detection of all three markers significantly improved predictive accuracy, yielding an AUC of 0.879 (95% CI: 0.828–0.930), with a sensitivity of 82.61% and specificity of 85.00%. This performance was significantly superior to that of any individual marker (DeLong test, all p< 0.001).Fig. 1ROC curves of individual indicators for the presence of SSc-ILD. The single-index ROC curves for NLR (red) and PT (green) are shown. The ROC curve for the combined model of NLR, PT, and Anti-Scl-70 (blue) demonstrated significantly improved performance compared to individual indicators, as determined by the Delong test (all p< 0.001). The dotted line represents the reference line. NLR, neutrophil to lymphocyte ratio; PT, Prothrombin time; Anti-Scl-70, Anti-Scleroderma 70 antibody ROC curves of individual indicators for the presence of SSc-ILD. The single-index ROC curves for NLR (red) and PT (green) are shown. The ROC curve for the combined model of NLR, PT, and Anti-Scl-70 (blue) demonstrated significantly improved performance compared to individual indicators, as determined by the Delong test (all p< 0.001). The dotted line represents the reference line. NLR, neutrophil to lymphocyte ratio; PT, Prothrombin time; Anti-Scl-70, Anti-Scleroderma 70 antibody Discussion The pathogenesis of SSc-ILD involves vascular injury, inflammation, autoimmunity, and ultimately tissue fibrosis [3]. Here, we present a large single-center cohort of SSc patients to investigate whether three distinct pathological dimensions, inflammatory markers, coagulation parameters, and autoantibodies are independently associated with the presence of SSc-ILD, and to assess their combined utility in identifying this complication. Our findings supported this objective: multivariate logistic regression analysis demonstrated that elevated NLR, PT, and Anti-SCL-70 positivity were each independently associated with SSc-ILD. Notably, the combined model exhibited significantly higher discriminative performance (AUC = 0.879) compared to any individual parameter. The majority of SSc-ILD patients enrolled in this study were diagnosed with dcSSc, a result that aligns with the results of multiple previous investigations. Compared to individuals with lcSSc, those with dcSSc demonstrated a notably higher incidence of ILD, and they were also more prone to experience progressive pulmonary fibrosis [20–22]. Moreover, the SSc-ILD cohort in this research exhibited an extended disease duration. While a shorter disease duration was previously considered predictive of rapid ILD progression, current evidence suggests that the apparent lung function plateau observed in patients with a longer disease duration may be attributed to survival bias [23]. Additionally, the high prevalence of Raynaud’s phenomenon and considerable gastrointestinal involvement in this cohort further supported the characteristic pattern of visceral involvement in SSc-ILD patients. This study confirmed that SSc-ILD patients exhibited features of systemic inflammatory activation. Multivariate regression analysis further demonstrated that NLR was an independent related factor for SSc-ILD. As a readily available inflammatory marker, NLR reflects the general inflammatory status of the body. Consistent with previous findings, NLR was notably elevated in SSc patients and positively correlated with disease severity [24]. Moreover, NLR levels in SSc-ILD patients were significantly higher than those in patients without ILD, highlighting its potential utility in the early identification of individuals at high risk for ILD [25]. Furthermore, NLR can serve as a biomarker to assess the effectiveness of treatment. In dcSSc patients who underwent autologous hematopoietic stem cell transplantation, the reduction of NLR was associated with improvements in both mRSS and FVC. Importantly, NLR levels increased again upon disease recurrence [26]. From a mechanistic perspective, neutrophils, as important effector cells of the innate immune system, contribute to fibrosis in the early stages of SSc-ILD through multiple pathways. Upon activation, neutrophils release neutrophil extracellular traps (NETs), which contain pro-inflammatory components such as histones and myeloperoxidase, and can directly damage alveolar epithelial cells [27, 28], while also promoting fibroblast activation. Furthermore, inflammation-activated neutrophils secrete cytokines and recruit additional inflammatory cells, thereby generating a positive feedback loop that amplifies the inflammatory response, enhances vascular permeability, and facilitates fibroblast migration [29]. The prolonged PT observed in SSc-ILD patients within our cohort indicated activation of the extrinsic coagulation pathway. Although no direct studies have investigated the relationship between PT and SSc specifically, clinical evidence indicates that thrombin activity in bronchoalveolar lavage fluid (BALF) is significantly increased in patients with SSc-ILD and idiopathic pulmonary fibrosis (IPF). Moreover, the pivotal role of thrombin has been substantiated in animal models of bleomycin-induced pulmonary fibrosis [30]. In both IPF and SSc-ILD, injured lung tissue displayed activation of the coagulation cascade and dysfunction of the fibrinolytic system [31, 32]. Thrombin, as the central effector molecule of this cascade, was generated in large quantities early after injury and exerted pathological effects beyond its procoagulant function. Initially, it contributed to vascular dysfunction by increasing vascular permeability and promoting endothelial cell angiogenesis [33]. Second, it exerted direct pro-fibrotic actions: the deposition of fibrin in the alveoli forms a provisional matrix that facilitates fibroblast attachment, differentiation, and migration. Finally, thrombin participated in modulating inflammation and immunity. It activated pathogenic Th2 lymphocytes in SSc, exacerbating the autoimmune response, and mediated neutrophil recruitment and endothelial cell proliferation [32, 34]. These observations support the hypothesis that anticoagulation therapy may delay the progression of ILD. Clinical trials have demonstrated that low molecular weight heparin or a thrombin inhibitor can attenuate the profibrotic effects of certain proteins in lung fibroblasts from SSc-ILD patients [13]. In our research, we observed a markedly higher prevalence of Anti-SCL-70. This result aligns with previous findings, as autoantibody profiles in SSc have been shown to correlate with disease activity [35]. Patients positive for Anti-Scl-70 exhibit a higher risk of ILD [36, 37], a condition frequently connected with pulmonary complications [38]. Anti-Scl-70-positive patients demonstrated the highest incidence of ILD and the lowest survival rates in both dcSSc and lcSSc. Immunization of mice with topoisomerase I and complete Freund's adjuvant has been shown to induce skin and pulmonary fibrosis [39]. Anti-Scl-70 can directly contribute to fibrosis, as the autoantigen topoisomerase I, released following endothelial cell injury, has bound to the surface of fibroblasts. This binding facilitated the recruitment of Scl-70 antibodies, leading to monocyte adhesion, the release of proinflammatory cytokines, and subsequent fibroblast activation [40]. Currently, the clinical diagnosis and monitoring of SSc-ILD primarily rely on HRCT and pulmonary function tests. HRCT is considered the gold standard for diagnosis; however, repeated use poses risks of cumulative radiation exposure and entails high costs. Pulmonary function tests are often limited by patient cooperation, such as restricted respiratory effort due to skin sclerosis, and by disease-related factors, including physical intolerance during acute exacerbations. Therefore, there remains a critical need for simple, cost-effective, non-invasive, and easily accessible serum biomarkers to complement the current evaluation system for SSc-ILD. Although the Anti-Scl-70 is a key serological marker, its sensitivity as a standalone test is suboptimal. This study demonstrated that integrating NLR and PT into a multi-parameter model enables a more comprehensive assessment of ILD presence in SSc patients at baseline. The combined model achieved a sensitivity of 82.61%, with ROC analysis yielding an AUC of 0.879 (95% CI: 0.828 – 0.930), significantly outperforming any single biomarker, and provided candidates for the serological evaluation of SSc-ILD. The theoretical foundation of this model was rooted in three primary mechanisms underlying SSc pathogenesis: microvascular injury, immune dysregulation, and multi-organ fibrosis [41]. Furthermore, a cascade mechanism known as the "inflammation-coagulation-autoimmunity axis" has been proposed. This axis provides a reference for understanding the pathophysiological process of SSc-ILD and provides enlightenment for the combined application of serum markers. This study has several limitations. First, as a retrospective study, its design may inherently lead to incomplete data collection and patient selection bias. All enrolled subjects were inpatients, whose disease status had specific clinical backgrounds: most were either newly diagnosed and untreated, or admitted due to disease exacerbation after long-term self-discontinuation of medication, and the majority had been in the disease duration of several years. Therefore, the study results may not apply to outpatients with early-stage or stable disease, and the identified biomarkers also cannot be used to predict early disease progression. Although we have made every effort to standardize blood collection time and assess disease activity, heterogeneity in disease stages remains a factor that requires cautious consideration when interpreting the results. In addition, the SSc-ILD group was predominantly composed of dcSSc patients, which is consistent with the clinical epidemiological characteristics of SSc; however, it remains difficult to fully disentangle the overlapping effects between the dcSSc phenotype and ILD in a cross-sectional design. Second, the cross-sectional nature of the study limits its ability to infer causality. We were unable to obtain data on the dynamic changes of indicators. Future prospective cohort studies that include both outpatients and inpatients are necessary, with systematic collection of medication use information and long-term follow-up, to validate the findings of this study. Third, although the sample size of this study (n=177) is considerable for a rare disease like SSc, it is still insufficient to analyze the unique clinical phenotypes of certain rare subtypes (e.g., patients positive for anti-RNA polymerase III antibodies). Finally, the proposed "inflammation-coagulation-autoimmunity" triple-marker synergy mechanism has not yet been fully elucidated, particularly lacking direct evidence at the tissue level. Conclusion This study has demonstrated that NLR, PT, and Anti-SCL-70 were independent factors for SSc-ILD. Moreover, the discriminative efficacy of the combined model of these three indicators is significantly superior to that of any single indicator, with a sensitivity of 82.6%. As the first attempt to integrate multi-dimensional indicators of inflammation, coagulation, and autoimmunity to explore SSc-ILD, this study still holds potential clinical reference value. The constructed combined biomarker model may assist in identifying SSc-ILD patients, provide a new perspective for understanding the multi-mechanism synergistic pathological process of SSc-ILD, and lay a foundation for subsequent prospective studies.