Detection of extranodal spread in head and neck cancer with [18F]FDG PET and MRI: improved accuracy?

Preoperative detection of extranodal spread (ENS) in head and neck cancer can have important consequences for patient management. The aim of this study was to determine whether 18-fluorodeoxyglucose positron emission tomography ([18F]FDG PET) or a combination with Magnetic Resonance Imaging (MRI) could more accurately predict ENS, especially with the near availability of fully integrated [18F]FDG PET/MRI scanners. In retrospective cohort design a total of twelve patients, with 18 lymphnode metastases were studied with [18F]FDG PET and MRI. Presence of ENS was scored on MRI, and [18F]FDG PET images using a SUV max cut-off point of 12. Histopathology results were used as reference standard. Sensitivity, specificity and accuracy were calculated. The sensitivity, specificity and accuracy of [18F]FDG PET for ENS reached 70%,100% and 83%, respectively. The mean SUVmax of ENS positive lymphnodes was 13.6 versus 8.7 for lymphnode metastases without ENS (P=0.03). The sensitivity, specificity and accuracy of MRI for ENS were 70%, 100% and 83%, respectively. When the [18F]FDG PET and MRI findings were combined sensitivity, specificity and accuracy were 80%, 100% and 89%, respectively. Thus, accuracy increased from 83% to 89%. When there is no ENS or doubt of ENS on MRI, [18F]FDG PET seems to have additional value since it improves sensitivity and resolves uncertainty in case of high FDG uptake. This benefit needs to be confirmed prospectively in a larger cohort.