Evaluation of the prognostic value of visual and quantitative CT analysis in myositis-associated interstitial lung...

To evaluate chest high-resolution CT (HRCT) images in idiopathic inflammatory myositis (IIM) using both quantitative CT and visual assessment.

HRCT images from a single-centre longitudinal IIM cohort were reviewed both visually and by quantitative imaging analysis (QIA), which quantified the extent of ground glass (QGG), fibrosis (QLF) and total interstitial lung disease burden (QILD). Visual review identified ILD presence and pattern and excluded other comorbid disease processes. Linear regression assessed correlations between QIA measures and pulmonary function tests (PFTs). Trends of QIA and PFTs were analysed in patients with ≥2 measurements using mixed-effects models. Associations with all-cause mortality were evaluated using Cox regression.

Among 277 patients (mean age 51 years; 74% female; 77% dermatomyositis; 22% had antisynthetase syndrome), 168 had HRCT evaluable by both methods. ILD was identified in 51% by visual review. QLF and QILD had strong correlations with forced vital capacity (β=-0.7 and -0.59) and diffusing capacity of the lung for carbon monoxide (β=-0.75 and -0.64) across all visits (p<0.0001 for all). Higher baseline QLF and QILD scores (HR 1.04, 95% CI 1.00 to 1.08) and HR 1.08, 95% CI 1.01 to 1.16) and worsening QLF and QILD trends in the first 12 months (HR 1.04, 95% CI 1.01 to 1.07 and HR 1.06, 95% CI 1.01 to 1.11) were associated with greater all-cause mortality (p<0.05 for all). Survival did not differ by the presence of visual ILD or by ILD pattern.

In this large, real-world IIM cohort, quantitative HRCT measures-particularly QLF and QILD-were strongly associated with lung function and predicted mortality. Higher baseline and early worsening of QLF and QILD were associated with increased mortality, whereas visual ILD presence and pattern did not predict survival.