Preoperative evaluation of pleural adhesions with dynamic chest radiography: a retrospective study of 146 patients with lung cancer.

ABSTRACT

AIM: To assess the utility of dynamic chest radiography (DCR) during the preoperative evaluation of pleural adhesions. MATERIALS AND METHODS: Sequential chest radiographs of 146 patients with lung cancer were acquired during forced respiration using a DCR system. The presence of pleural adhesions and their grades were determined by retrospective surgery video assessment (absent 121, present 25). The maximum inspiration to expiration lung area ratio was used as an index for air intake volume. A ratio of ≥0.65 was regarded as insufficient respiration. Two radiologists assessed the images for pleural adhesions based on motion findings. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were compared for each adhesion grade and patient group (patients with sufficient/insufficient respiration). Pearson's chi-squared test compared the group. Statistical significance was set at p<0.05. RESULTS: DCR correctly identified 22/25 patients with pleural adhesions, with 20 false-positive results (sensitivity, 88%; specificity, 83.5%; PPV, 52.4%; NPV, 97.12%). Although the diagnostic performances for the various adhesion grades were similar, specificity in patients with sufficient respiration increased to 93.9% (31/33), identifying all cases except for those with loose adhesions. CONCLUSIONS: DCR images revealed restricted and/or distorted motions in lung structures and structural tension in patients with pleural adhesions. DCR could be a useful technique for routine preoperative evaluation of pleural adhesions. Further development of computerised methods can assist in the quantitative assessment of abnormal motion findings.