Detection of mediastinal and hilar lymph nodes by 16-row MDCT: is contrast material needed?

To retrospectively evaluate whether contrast administration is necessary in the detection of mediastinal and hilar lymph nodes when thin slice axial and coronal MDCT images are used.

This study was approved by our Institutional Review Board, informed consent was not required. Thirty-five patients who needed a chest CT (0.75 mm x 16) for various reasons were included. Four different image sets were reconstructed for each patient: non-enhanced axial (N-Ax), non-enhanced coronal MPR (N-Co), enhanced axial (E-Ax) and enhanced coronal MPR (E-Co). All the images were 1mm thick and interval. Two board-certified chest radiologists independently evaluated whether a lymph node with a short diameter, larger than 5mm, existed in each nodal station of the mediastinum and hilum. Two different board-certified chest radiologists assessed all four image sets together and established a reference standard by consensus. Interobserver agreement between the two readers was assessed by kappa statistics. Accuracy was calculated on each image set and compared to each other by McNemar's test.

A total of 211 nodal stations, including 113 mediastinal and 98 hilar, were defined to be present and this was the reference standard. Except for N-Ax, the kappa values were within moderate to substantial (0.53-0.81). The accuracy for hilar nodes detection was significantly higher for the contrast enhanced images both in the axial (p<0.001) and coronal (p<0.01) data sets. The addition of contrast material did not significantly increase accuracy for the detection of mediastinal nodes (axial: p=0.542, coronal p=0.727).

Contrast administration is recommended in the detection of hilar lymph nodes both on axial and MPR views, however, for assessment of mediastinal lymph nodes its contribution is low.