Diagnostic accuracy of diffusion weighted MRI in cervical lymphadenopathy cases correlated with pathology results

Introduction: Diffusion weighted magnetic resonance imaging (DW MRI) is an imaging technique showing molecular diffusion. Cell size, density and integrity influence the signal intensity seen on diffusion-weighted images. This technique is a helpful complementary tool to distinguish tumoral from non tumoral tissue. The aim of this prospective study is to define the diagnostic accuracy of DWI to differentiate benign from malignant cervical lymph nodes. Patients and methods: Twenty six patients who presented with 32 nodes were included in this study, 9 males (35.6%) and 17 females (65.4%). Their age ranged from: 6 to 76 years, mean age 45 +/- 18.8 years referred to the radiology department of NCI, complaining of neck swelling, ultrasound showed cervical nodes. Results: According to histopathological analysis we divided the examined lymph nodes (n=32) into 2 categories: malignant lymph nodes 75% (n=24) benign lymph nodes 25% (n=8). DWI and ADC (apparent diffusion coefficient) values revealed 27 malignant lesions (84%), 5 benign (16%). The accuracy of the DWMRI was 89%. A significant difference between benign and malignant cervical nodes on DWI and on ADC maps is reported. The results obtained were 24 true positive, 3 false positive, 5 true negative. No false negative cases were identified, yielding a sensitivity of 100%, specificity of 62.5%, NPV=100% and PPV=89%. The difference between the mean ADC values between benign and malignant lesions was statistically significant (P < 0.0001). Conclusion: MR diffusion imaging could be an important supportive tool in differentiation between benign and malignant lymph nodes, can to a lesser extent differentiate between the types of malignant lymphadenopathy and can be used as an indicator for improvement and recurrence post chemo and radiotherapy. (C) 2014 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by Elsevier B.V. Open access under CC BY-NC-ND license.