Expiratory thin-section CT in evaluation of air-trapping and predicting bronchiolitis obliterans syndrome for patients after allogeneic hematopoietic stem cell transplantation

Jian-xin HE,Jie CHEN,An-hui ZHU,Xiao-jin ZHANG
DOI: https://doi.org/10.13929/j.1672-8475.2013.03.002
2013-01-01
Abstract:Objective: To explore the optimal assessment method of expiratory thin-section CT in assessment of air-trapping and predicting bronchiolitis obliterans syndrome (BOS) in patients after hematopoietic stem cell transplantation (HSCT). Methods: Five-point scale method and visual method were retrospectively used to analyze expiratory thin-section CT radiological findings in 29 patients after HSCT and diagnosed as BOS. These two methods were evaluated according to air trapping, and the predictive value of 5-point scale method, visual method and pulmonary function tests (PFTs) were compared. Results: PFTs showed BOS0 in 17, BOS-p in 7, BOS1 in 2, BOS2 in 1 and BOS3 in 2 patients, with the long-term BOS predictive sensitivity of 41.38% (12/29). The area under the ROC curve of five-points scale method (0.566) was larger than that of visual method (0.485). When 5-point scale method was used, the sensitivity was 25.00% (3/12), specificity was 82.35% (14/17), PPV was 50.00% (3/6) and NPV was 60.87% (14/23), while the sensitivity of prediction to BOS was 55.17% (16/29). When the visual evaluation method was used, the sensitivity was 41.67% (5/12), specificity was 58.82% (10/17), PPV was 41.67% (5/12) and NPV was 58.82% (10/17), while the sensitivity of prediction to BOS was 79.31% (22/29). Conclusion: Five-point scale method has a higher diagnosis value than visual method. While visual evaluation method has the strongest prediction ability, visual method medium and PFTs have weak prediction ability for early BOS. Copyright © 2013 by the Press of Chinese Journal of Medical Imaging Technology.
What problem does this paper attempt to address?