Introduction: Pre-treatment evaluation of the presence of cervical nodal metastases is important in head and neck cancers and has major prognostic implications. In this study, we aim to determine the accuracy of CT/PET as a tool for identifying such metastases.

Methods: All patients from Royal Prince Alfred and Liverpool Hospitals, who underwent CT/PET for any cancer arising from the head and neck, and who underwent subsequent surgery (which included a neck dissection) within 8 weeks of the CT/PET were included. Nodal staging was undertaken by utilising imaging-based nodal classification, and comparison with pathologic data from the surgical specimen was made. PET was considered positive if the SUV was greater than 2.

Results: We identified 111 patients from the above criteria. 80 of such patients were treated for squamous cell carcinoma (SCC). CT/PET identified unsuspected metastatic disease in 6 patients. Correlation of CT/PET findings and the presence of disease at the primary site: sensitivity: 98%, specificity: 93%, positive predictive value (PPV): 98% and negative predictive value (NPV): 93%. Correlating CT/PET findings with the presence of nodal disease at any level: sensitivity: 95%, specificity: 88%, PPV: 95% and NPV: 88%. CT/PET was anatomically accurate in predicting the site of metastases in 62/74 (84%).

Conclusion: PET is accurate in predicting both presence of nodal metastases and the level of involvement. CT/PET should be undertaken as a pre-operative tool to assist in planning the extent of surgery required in head and neck cancers.