Nodal assessment and extranodal extension in head and neck squamous cell cancer: insights from computed tomography and magnetic resonance imaging

The aim of the study was to evaluate the concordance between computed tomography (CT)/magnetic resonance imaging (MRI) and histological examination in the evaluation of lymph nodal involvement (N+) and extranodal extension (ENE+) in patients with squamous cell cancer of the head–neck district (HNSCC). The second end point was to evaluate which of the imaging features was more sensitive and specific in establishing N+ and ENE+, and comparing the diagnostic accuracy between CT and MRI. 58 patients with HNSCC undergoing surgical treatment with lymph node dissection at the Policlinico of Bari were enrolled in this retrospective study. The criteria used for identifying radiological N+ (rN+) included six characteristics; the presence of any two or more criteria out of these six was considered rN+. For each rN+, the radiological extranodal extension (rENE+) was evaluated analyzing three characteristics; rENE was considered positive if any one criterion was present. Of 167 lymph nodes, 45/167 (27%) had rN+, out of which 20/45 (44%) were rENE+. On pathological examination, 38/45 (84%) nodes were N+ and 11/20 (55%) were ENE+. The agreement between imaging and histology in the evaluation of N was (149/167) 89% with a good concordance (k = 0.7). The agreement between imaging and histology in the evaluation of ENE was (35/45) 78% with a moderate concordance (k = 0.5). Loss of fatty hilum was found to be the most sensitive (84%) imaging finding in N+ evaluation. Capsular irregularity with fat stranding had the highest sensitivity (82%) for the determination of ENE+. Comparing CT and MRI imaging in the evaluation of N+ and ENE+, we found no statistically significant differences (p-value of 0.3 for N+ and p-value of 0.3 for ENE +). Imaging has good confidence in detecting rN+ but modest in assessing rENE+. Further research could improve the imaging specificity for the determination of rENE.