Kazuhiro Imai 1 , Yoshihiro Minamiya , Kouichi Ishiyama , Manabu Hashimoto , Hajime Saito , Satoru Motoyama , Yusuke Sato , Jun-ichi Ogawa Show Affiliations »
Abstract
PURPOSE: To develop a simple noninvasive technique for evaluating pleural invasion by using routine preoperative computed tomography (CT). MATERIALS AND METHODS: The institutional review board approved this retrospective study, and written informed consent was obtained for performing the initial and follow-up CT studies. Preoperative CT findings (169 patients with possible pleural invasion) and pathologic diagnoses after surgical resection were evaluated. The length of the interface between the primary tumor and neighboring structures (arch distance) and the maximum tumor diameter were measured on CT images, after which arch distance-to-maximum tumor diameter ratios were calculated. Receiver operating characteristic (ROC) curves were used to analyze the ratios. RESULTS: Median arch distance-to-maximum tumor diameter ratios for pleural invasion categories (pl1, pl2, pl3) assessed by using the Union Internationale Contre le Cancer TNM staging system were as follows: pl1, 0.206 (25th-75th percentile, 0-0.486); pl2, 0.638 (25th-75th percentile, 0.385-0.830); and pl3, 1.092 (25th-75th percentile, 1.045-1.214) (P < .001 between groups). On the basis of the ROC curves, the cut-off value for invasion was an arch distance-to-maximum tumor diameter ratio of 0.9. When the ratio was greater than 0.9, the sensitivity and specificity for thoracic invasion and area under the ROC curve were 89.7%, 96.0%, and 0.976, respectively, which represents an improvement over values obtained by using conventional criteria (radiologists A and B: 46.7% and 74.2% and 91.3% and 84.8%, respectively). CONCLUSION: When diagnosing T3 or T4 lung cancer based on arch distance-to-maximum tumor diameter ratios, a higher performance level was achieved than that with use of conventional criteria. Measurement of the ratios is a simple noninvasive technique for evaluating pleural invasion at CT. © RSNA, 2013.
PURPOSE: To develop a simple noninvasive technique for evaluating pleural invasion by using routine preoperative computed tomography (CT). MATERIALS AND METHODS: The institutional review board approved this retrospective study, and written informed consent was obtained for performing the initial and follow-up CT studies. Preoperative CT findings (169 patients with possible pleural invasion ) and pathologic diagnoses after surgical resection were evaluated. The length of the interface between the primary tumor and neighboring structures (arch distance) and the maximum tumor diameter were measured on CT images, after which arch distance-to-maximum tumor diameter ratios were calculated. Receiver operating characteristic (ROC) curves were used to analyze the ratios. RESULTS: Median arch distance-to-maximum tumor diameter ratios for pleural invasion categories (pl1, pl2, pl3) assessed by using the Union Internationale Contre le Cancer TNM staging system were as follows: pl1, 0.206 (25th-75th percentile, 0-0.486); pl2, 0.638 (25th-75th percentile, 0.385-0.830); and pl3, 1.092 (25th-75th percentile, 1.045-1.214) (P < .001 between groups). On the basis of the ROC curves, the cut-off value for invasion was an arch distance-to-maximum tumor diameter ratio of 0.9. When the ratio was greater than 0.9, the sensitivity and specificity for thoracic invasion and area under the ROC curve were 89.7%, 96.0%, and 0.976, respectively, which represents an improvement over values obtained by using conventional criteria (radiologists A and B: 46.7% and 74.2% and 91.3% and 84.8%, respectively). CONCLUSION: When diagnosing T3 or T4 lung cancer based on arch distance-to-maximum tumor diameter ratios, a higher performance level was achieved than that with use of conventional criteria. Measurement of the ratios is a simple noninvasive technique for evaluating pleural invasion at CT. © RSNA, 2013.
Entities: Disease
Species
Mesh： See more »
Year: 2013
PMID： 23329658 DOI： 10.1148/radiol.12120864
Source DB: PubMed Journal: Radiology ISSN： 0033-8419 Impact factor: 11.105