Discrimination of metastatic cervical lymph nodes with diffusion-weighted MR imaging in patients with head and neck cancer.

BACKGROUND AND PURPOSE Metastasis to the regional cervical lymph nodes may be associated with alterations in water diffusivity and microcirculation of the node. We tested whether diffusion-weighted MR imaging could discriminate metastatic nodes. METHODS Diffusion-weighted echo-planar and T1- and T2-weighted MR imaging sequences were performed on histologically proved metastatic cervical lymph nodes (25 nodes), benign lymphadenopathy (25 nodes), and nodal lymphomas (five nodes). The apparent diffusion coefficient (ADC) was calculated by using two b factors (500 and 1000 s/mm(2)). RESULTS The ADC was significantly greater in metastatic lymph nodes (0.410 +/- 0.105 x 10(-3) mm(2)/s, P <.01) than in benign lymphadenopathy (0.302 +/- 0.062 x 10(-3) mm(2)/s). Nodal lymphomas showed even lower levels of the ADC (0.223 +/- 0.056 x 10(-3) mm(2)/s). ADC criteria for metastatic nodes (>/= 0.400 x 10(-3) mm(2)/s) yielded a moderate negative predictive value (71%) and high positive predictive value (93%). Receiver operating characteristic analysis demonstrated that the criteria of abnormal signal intensity on T1- or T2-weighted images (A(z) = 0.8437 +/- 0.0230) and ADC (A(z) = 0.8440 +/- 0.0538) provided similar levels of diagnostic ability in differentiating metastatic nodes. The ADC from metastatic nodes from highly or moderately differentiated cancers (0.440 +/- 0.020 x 10(-3) mm(2)/s, P <.01) was significantly greater than that from poorly differentiated cancers (0.356 +/- 0.042 x 10(-3) mm(2)/s). CONCLUSION Diffusion-weighted imaging is useful in discriminating metastatic nodes.