 |
Vol.51 No.6 contents | Japanese/English |
 | Full Text of PDF (377K) Article in Japanese |
- Original Article -
Retrospective Study Comparing Differentiation Methods in Noninvasive Pulmonary Small Adenocarcinomas Using Mediastinal and Lung Window Images Acquired by Thin-section Computed Tomography
Tetsuro Kondo1, Shuji Murakami1, Haruhiro Saito1, Fumihiro Oshita1, Hiroyuki Ito2, Masahiro Tsuboi2, Haruhiko Nakayama2, Tomoyuki Yokose3, Youichi Kameda3, Kouzo Yamada11Department of Thoracic Oncology, 2Department of Thoracic Surgery, 3Department of Pathology, Kanagawa Cancer Center, Japan
Objective. We sought to compare two methods of distinguishing invasive and noninvasive cancers in peripheral small lung adenocarcinoma. Method 1 consisted of using the ratio of two major tumor dimensions obtained from thin-section computed tomography (TS-CT) lung window image and mediastinal window image. Method 2 consisted of using just one major tumor dimension obtained from the TS-CT mediastinal window image. Methods. Enrolled in the study were 430 patients with small peripheral lung adenocarcinomas (189 men, 241 women; 441 lesions) who underwent surgical resection at our institute from July 1997 to May 2008. The tumor shadow diameter was pathologically shown to be 20 mm or less. The cases free from the following were classified as noninvasive, whereas cases with any of the following were classified as invasive: lymph node metastasis, pleural invasion, vascular invasion, lymphatic permeation, and recurrence. Using preoperative TS-CT images, the tumor shadow disappearance rate (%) was calculated by Method 1 using the following equation: (1-major tumor dimension on mediastinal image/major tumor dimension on lung image)×100. The tumor shadow disappearance rate was determined by Method 2 using the major tumor dimension with only the mediastinal image. Receiver operating characteristic (ROC) and logistic-regression analyses were used to evaluate the effectiveness of these methods. Results. We found 441 lesions that comprised 328 noninvasive and 113 invasive lesions. ROC analysis revealed that Method 1 was more effective than Method 2. The multivariate logistic-regression analysis revealed that the factors "female" and "high tumor disappearance rate" were independently related to noninvasiveness. Conclusion. When differentiating the invasiveness and noninvasiveness of peripheral small lung adenocarcinoma, the ratio of the major tumor dimensions obtained from the TS-CT lung window image and mediastinal window image (Method 1) provided a more effective index than the major tumor dimension obtained from only the TS-CT mediastinal window image (Method 2).
key words: Adenocarcinoma, Tumor shadow disappearance rate, Thin-section computed tomography, Noninvasive cancer
Received: February 16, 2011
Accepted: July 14, 2011
JJLC 51 (6): 694-700, 2011
