[18F]FDG-PET/MRI Performance in Head-and-Neck Cancers at Initial Staging, Restaging and Follow Up.

BACKGROUND AND PURPOSE: Management of head-and-neck cancer (HNC) requires a precise imaging pretreatment assessment. The complementarity of MRI and PET in a single examination permit a comprehensive multimodality tumor characterization with matching spatiotemporal data. We aimed to assess the performance of simultaneous [18F]FDG-PET/MR for TNM-staging of HNC.

METHODS: Consecutive HNC patients who underwent [18F]FDG-PET/MR examination were included in this single-institution retrospective study. The protocol included simultaneous [18F]FDG-PET/MR acquisitions, first head-and-neck, then of the whole-body. We assessed T and N-staging on [18F]FDG-PET/MR based on the AJCC manual 7th-edition. Results from histopathology served as the reference standard. Sensitivity, specificity, positive and negative predictive values and accuracy were calculated. Detection rates of metastases and synchronous diseases were analyzed according to patients' follow-up.

RESULTS: We included 117 patients (62 men, 59.2±17.4) with 124 examinations: 77 (62%) initial stagings, 32 (26%) restagings and 15 (12%) therapeutic evaluations. Histopathological data of resected tumors was available in 69 cases. [18F]FDG-PET/MR accuracy for defining T-stage was 0.89 at initial staging, 0.83 at restaging and 1 at follow-up, resulting in an overall accuracy of 89%. Among the 59 patients with cervical-node resection, 51 (86%) were correctly staged by [18F]FDG-PET/MR. Its accuracy for lymph-node characterization was 0.86 at initial staging, 0.92 at restaging and 0.82 at follow-up.

CONCLUSIONS: [18F]FDG-PET/MR has excellent diagnostic performances for both T and N-staging in HNC, showing a good accuracy whether it was at initial staging, restaging or during the follow-up.