Feasibility of automated quantitative CT imaging metrics in pulmonary Langerhans cell histiocytosis: a pilot study

Visual scoring of pulmonary Langerhans cell histiocytosis (PLCH) lesions on computed tomography (CT) is reader-dependent, and automated quantitative imaging biomarkers have not previously been applied to systematically assess structure–function relationships in this rare cystic lung disease. The objective of this study was to evaluate the feasibility and agreement of automated quantitative CT biomarkers, including densitometry, parametric response mapping (PRM), and to explore their correlations with pulmonary function. As a secondary objective, we assessed pulmonary vascular involvement using Computer-Aided Lung Informatics for Pathology Evaluation and Rating (CALIPER)-derived pulmonary vascular volume. In this prospective pilot study, patients with cystic PLCH underwent paired inspiratory-expiratory CT and concomitant pulmonary function tests (PFT). Imaging was analysed via conventional densitometry, PRM and CALIPER-derived vascular volume assessment. PRM classified lung voxels into normal density (PRM Norm ), persistent low-density areas (PRM Pers ), and functional small airways disease (PRM fSAD ). Visual scoring was performed by two expert radiologists. The agreement among these methods was evaluated using intraclass correlation coefficients (ICC) and the Bland–Altman approach. Associations between CT-derived metrics and PFT parameters were examined via Pearson correlations and multiple linear regression adjusted for smoking history. Eighteen patients were evaluated. Densitometry and PRM quantification of cystic lesions correlated with expert visual scoring (all r ≥ 0.80) and with PFT parameters, including D LCO ( p < 0.05) with good agreement (ICC 0.96; 95% CI, 0.91–0.99). Multivariate analysis adjusted for smoking pack-years showed that PRM Pers was significantly associated with all PFT parameters ( p < 0.05), independent of PRM fSAD . However, PRM fSAD values were variable among patients with similar FEV₁ % predicted values, suggesting heterogeneity of small airway disease in this population. The regional cystic extent assessed by PRM and pulmonary vascular volume assessed by CALIPER exhibited a significant inverse association ( p = 0.002). Automated quantitative CT imaging provides objective metrics of structural lung abnormalities in PLCH that correlate with functional impairment. PRM allowed an exploratory assessment of the relative contributions of bronchiolar and cystic abnormalities to airflow obstruction, offering complementary insights beyond conventional global measures. These findings support the feasibility of quantitative CT biomarkers in PLCH and warrant further evaluation of their potential value for future disease phenotyping and longitudinal assessment.