Utilizing a chest CT interstitial lung disease scoring system for commonly encountered pediatric rheumatologic diseases with associated interstitial lung disease with clinical correlation

Background

Chest CT is useful for early detection and monitoring of rheumatologic-related interstitial lung disease (ILD) in children. However, interpretation is inconsistent and lacks standardization.

Objective

This study employs prior disease-specific CT scoring systems to systematically assess most rheumatologic diseases with ILD in pediatric patients and correlate scores with clinical findings.

Materials and methods

This retrospective study included patients (ages 0-21 years) with rheumatologic-related ILD and chest CT. Two pediatric radiologists determined CT quality and separated patients into disease and disease control groups. Three different blinded pediatric radiologists scored CT scans using this Pediatric Rheumatologic Interstitial Lung Disease (PRILD) scoring system, comprising of parenchymal, nodule/mass/cavitary/cyst, and pleural/pericardial components (total score, 125). Clinical data were extracted from medical records. Non-parametric tests were used for group comparisons, intraclass correlation coefficient (ICC) for interobserver agreement, and Spearman’s correlation for CT and pulmonary function relationships.

Results

Median age was 13.8 years and 84% were female. The disease group had statistically significant higher median total PRILD scores than the disease control group (7.3 vs. 1.7, P<0.001) and a higher percent active rheumatologic disease (93% vs 78%, P=0.039). There were no statistically significant differences between the two groups regarding age, sex, rheumatologic disease type, disease duration, or pulmonary symptoms. Interrater agreement was poor; ICC was 0.41. When both groups were combined, there were weak negative correlations with pulmonary function tests.

Conclusion

PRILD scoring system aids in systematic evaluation of thoracic involvement in rheumatologic-related ILD. However, further refinement and education are needed to improve interrater agreement, and given the rare nature of these diseases, a multicenter collaboration is necessary for validation of this CT scoring system.

Graphical abstract