Detection of head and neck squamous cell carcinoma with diffusion weighted MRI after (chemo)radiotherapy: correlation between radiologic and histopathologic findings.

Purpose: To investigate the value of diffusion weighted magnetic resonance imaging (DW-MRI) in differentiating persistent or recurrent head and neck squamous cell carcinoma (HNSCC) from nontumoral postradiotherapeutic alterations.
Methods and Materials: In 26 patients with suspicion of persistent or recurrent HNSCC, MRI of the head and neck was performed, including routine turbo spin-echo (TSE) sequences and an additional echo-planar DW-MRI sequence, using a large range of b-values (0-1000 s/mm(2)). Apparent diffusion coefficient (ADC) maps were calculated. In the suspect areas at the primary site and in the suspect lymph nodes, signal intensity was measured on the native b0 and b1000 images and ADC values were calculated for these tissues. The same was done for surrounding irradiated normal tissue. Imaging results were correlated to histopathology.
Results: Signal intensity on native b0 images was significantly lower for HNSCC than for nontumoral postradiotherapeutic tissue (p < 0.0001), resulting in a sensitivity of 66.2%, specificity of 60.8%, and accuracy of 62.4%. Signal intensity on native b1000 images was significantly higher for HNSCC than for nontumoral tissue (p < 0.0001), resulting in a sensitivity of 71.6%, specificity of 71.3%, and accuracy of 71.4%. ADC values were significantly lower for HNSCC than for nontumoral tissue (p < 0.0001), resulting in a sensitivity of 94.6%, specificity of 95.9%, and accuracy of 95.5%. When compared with computed tomography, TSE-MRI and fluorodeoxyglucose-positron emission tomography, DW-MRI yielded fewer false-positive results in persistent primary site abnormalities and in persistent adenopathies, and aided in the detection of subcentimetric nodal metastases.
Conclusions: Diffusion weighted-MRI accurately differentiates persistent or recurrent HNSCC from nontumoral tissue changes after (chemo)radiotherapy.