[Imaging diagnosis of esophageal carcinoma by computed tomography and magnetic resonance imaging]

Nihon Geka Gakkai Zasshi. 2002 Apr;103(4):331-6.
[Article in Japanese]

Abstract

The staging diagnosis of esophageal carcinoma is important to determine therapeutic modalities and to predict prognosis. The current status of imaging diagnosis of tumor invasion to the adjacent organs and lymph node metastasis is described. The diagnostic criteria used to determine tumor invasion to the adjacent or gans by computed tomography (CT) and magnetic resonance imaging(MRI) are displacement and compression deformity of the tracheobronchial tree and obliteration of the periaortic fat plane over more than 90 degrees of the aortic circumference. Detection of the fat plane between the esophagus and the aorta supported by density profile analyzing software on CT may enable the diagnosis of invasion. Cine-MRI imaging is also useful to obtain dynamic information on the tumor and aorta. Tumor invasion to the aortic wall can be excluded when a low-intensity stripe is recognized between the tumor and the aortic wall. Although the criterion for lymph node metastasis on CT is 10 mm or more in long transverse diameter, the diagnostic accuracy is poor. The accuracy improves when imaging patterns such as heterogeneous internal structures in the enhanced lymph nodes and/or hyperenhancement in the lymph nodes in the early phase by dynamic study are added to the diagnostic criteria. However, small metastatic lymph node remain undetected and it is difficult to diagnose negative lymph node metastasis properly on CT and MRI. It is important to have full knowledge of the advantages and limitations of each imaging modality and to obtain objective information form them.

Publication types

  • English Abstract

MeSH terms

  • Esophageal Neoplasms / diagnosis*
  • Esophageal Neoplasms / pathology
  • Humans
  • Lymph Nodes / pathology*
  • Lymphatic Metastasis / diagnosis
  • Magnetic Resonance Imaging*
  • Neoplasm Invasiveness
  • Tomography, X-Ray Computed*