Use, accuracy, and implications for patient management of [18F]-2-fluorodeoxyglucose-positron emission/computerized tomography for head and neck tumors

Objectives/Hypothesis. Positron emission tomography (PET) has shown promise for early detection and accurate staging of cancer patients. A limited number of studies suggest PET/computed tomography (CT) may improve these variables; however, no published study has specifically evaluated clinical outcomes with PET/CT for head and neck (HN) tumors. The current study evaluates the use, accuracy, and implications for patient management of PET/CT scans in patients with HN tumors. Study Design: Retrospective cohort outcomes study at a tertiary care center. Methods. The authors identified 795 consecutive PET/CT at our institution. A total of 113 were obtained for HN tumors; 97 were used in the final analysis. Accuracy, use, and implications for patient care management decisions were correlated with each PET/CT scan. Multiple regression analysis was performed. Results. Accuracy, sensitivity, and specificity were measured by comparing the PET/CT results at the primary tumor site, cervical node sites, and distant sites with either pathologic or definitive clinical diagnoses. PET/CT had an overall per scan accuracy of 72% and a per patient accuracy of 69%. When stratification for rationale of obtaining the scan was performed, accuracy was 80% for staging distant disease, 67% for primary tumor evaluation, 72% for evaluation for recurrence, and 60% for unknown primary tumor evaluation. Conclusions: PET/CT imaging is a promising tool for evaluating HN tumors; however, in clinical practice, the proper use of such technology is not well studied. In our study, PET/CT had an overall accuracy of 72% in evaluating HN tumors, and PET/CT had the most accuracy in the detection of distant metastasis.