Abstract
OBJECTIVE: We wanted to investigate the accuracy of contrast-enhanced MR imaging for the detection of lymph node metastases in a head and neck cancer rabbit model. MATERIALS AND METHODS: The metastatic lymph node model we used was created by inoculating VX2 tumors into the auricles of six New Zealand White rabbits. T1-weighted MR images were obtained before and after injecting gadopentetate dimeglumine at three weeks after tumor cell inoculation. The sizes, signal intensity ratios (i.e., the postcontrast signal intensities of the affected nodes relative to the adjacent muscle) and the enhancement patterns of 36 regional lymph nodes (parotid and caudal mandibular nodes) were evaluated on MR images and then compared with the histopathologic findings. RESULTS: No statistical difference was found between the sizes of 12 metastatic (10.5+/-3.2 mm) and 24 hyperplastic (8.0+/-3.6 mm) lymph nodes (p > 0.05). On the contrast-enhanced T1-weighted MR images, nine metastatic and four hyperplastic lymph nodes had peripheral high and central low signal intensity, whereas three metastatic and 20 hyperplastic lymph nodes had homogeneous high signal intensity. Using a signal intensity ratio less than one as a diagnostic criterion for a metastatic lymph node, the sensitivity, specificity and positive and negative predictive values of the enhanced MR images were 75% (9/12), 83% (20/24), 69% (9/13) and 87% (20/23), respectively, with areas under receiver-operating-characteristic curve values of 0.81. CONCLUSION: This experimental study confirms that metastatic and hyperplastic lymph nodes can be differentiated using MR images on the basis of the contrast uptake patterns, but that they cannot be differentiated using any particular size criteria.