Lung ultrasound B-line quantification in CTD-ILD: a cross-sectional single-center observational study
摘要
To evaluate the diagnostic and clinical relevance of quantitative lung ultrasound (LUS) score in patients with connective tissue disease-associated interstitial lung disease (CTD-ILD), and to determine its correlation with high-resolution computed tomography (HRCT) severity, pulmonary function, and laboratory parameters. The cross-sectional, single-center observational study enrolled patients with connective tissue diseases who had confirmed interstitial lung disease and underwent both HRCT and LUS within 7 days of evaluation. Lung ultrasound was performed using a 12-zone scanning protocol, and B-lines were quantified to derive a semi-quantitative LUS-B score representing subpleural interstitial involvement. ILD severity categories were defined based on HRCT Warrick scores as mild (0–7), moderate (8–15), and severe (> 15). Pulmonary function tests and laboratory parameters, including inflammatory and immune-related markers, were collected. Correlations between LUS-B score, HRCT severity, pulmonary function, and laboratory variables were analyzed, and multinomial logistic regression was used to evaluate the association between LUS-B score and ILD severity categories. Diagnostic performance was assessed using ROC analysis. 117 patients with CTD-ILD (mean age 57 ± 12 years; 74% female) were enrolled. The cohort included patients with Sjögren’s syndrome (SS) (n = 32), systemic sclerosis (SSc) (n = 18), rheumatoid arthritis (RA) (n = 16), idiopathic inflammatory myopathy (IIM)/anti-synthetase syndrome (ASS) (n = 27), as well as other CTDs (n = 24), including systemic lupus erythematosus (n = 6), ANCA-associated vasculitis (n = 5), and undifferentiated connective tissue disease (n = 13). LUS scores showed a negative correlation with force expiratory volume in one second (FEV1) (rs = − 0.214, P = 0.03) and a positive correlation with HRCT Warrick scores (rs = 0.304, P = 0.001). An exploratory cut-off value of 7 for the LUS score was determined using the Youden index. Group comparisons based on this threshold showed patients with higher LUS scores showed greater HRCT severity (Z = − 2.503, P = 0.012), lower FEV1 (Z = − 3.128, P = 0.002) and FVC (Z = − 2.335, P = 0.020) ratios, and higher CD4⁺ T cell levels (Z = − 2.144, P = 0.032). Subgroup analysis demonstrated that the correlation between LUS and HRCT score was strongest in Sjögren’s-syndrome-associated ILD (rs = 0.417, P = 0.018), whereas other CTD subtypes showed weaker or nonsignificant associations. In multivariable logistic regression, parameters entered a priori (age, sex, LUS score, DLco, hemoglobin, and baseline pulmonary disease) identified LUS (OR = 0.87, P = 0.016) and DLco (OR = 1.12, P < 0.001) as independent correlates of ILD severity. ROC analysis based on HRCT-defined severity demonstrated that the LUS score had moderate discriminatory ability for severe ILD (AUC = 0.642, sensitivity = 86%, specificity = 42%); however, combining LUS with DLco significantly improved severity stratification. Quantitative LUS score provides a reproducible, non-invasive assessment of ILD burden and correlates moderately with HRCT severity and pulmonary function impairment in CTD-ILD. LUS demonstrated its greatest value in identifying severe ILD, particularly in Sjögren’s syndrome, whereas its ability to distinguish intermediate severity was limited. When combined with DLco, its stratification performance improved. Thus, LUS serves as a complementary bedside tool to HRCT for evaluating disease burden and risk in CTD-ILD.