One hundred twelve patients with esophageal carcinoma who had undergone CT and MRI before surgery were studied retrospectively. The longest transverse diameter (LTD) and shortest transverse diameter (STD) of 4, 018 resected lymph nodes were measured. Only 29 (1%) of the STD were 10mm or longer among 3, 685 non-metastatic lymph nodes, while 128 (38%) were shorter than 5mm among the 333 metastatic lymph nodes.
To differentiate metastatic lymph nodes from non-metastatic nodes based on size, the regional maximum standards of the LTD and STD of non-metastatic nodes were calculated.
The STD showed a smaller variation and more useful parameter than the LTD. A standard maximum size for normal STD of nodes in each region was proposed as follows: 12mm for nodes in # 106tb-R; 10mm for nodes in # 107, # 109-L, # 5, and # 9; and 8mm for nodes in other regions.
A guideline based on an STD of 10mm for the diagnosis of metastatic nodes was considered appropriate. The accuracy of the CT diagnosis of lymph node metastasis was 75%, sensitivity 55% and specificity 85%, according to this criterion. Even using STD/LTD ratio criteria in addition to the STD size criterion, there was no remarkable change in the accuracy of detection of lymph node metastasis.
Helical CT was used in the other 12 patients with esophageal carcinoma. Helical CT can detect smaller lymph nodes compared to conventional CT. Therefore, the accuracy rate was increased from 43% by conventional CT to 51% by helical CT.
The accuracy of a MRI diagnosis of lymph node metastasis was 69%, sensitivity 59% and specificity 74%, using similar CT size criteria. In addition to these criteria, with heterogeneous pattern criteria for the lymph nodes the results were 71%, 43% and 86%, respectively.
Other new criteria and techniques should be considered to improve the evaluation of lymph node metastasis from esophageal carcinoma.
