Accuracy of FDG PET and CT Imaging for Nodal Metastasis

Abstract

Objective: After radiation treatment, head and neck cancer patients often undergo FDG PET and CT imaging to survey for nodal metastasis. Inconsistencies among these make it difficult to decide if surgery is necessary. Our goal is to determine the accuracy of these modalities using histopathology, both overall and among discrepant results. Method: We reviewed the records of 486 head and neck squamous cell carcinoma patients who underwent a postradiation FDG PET/CT scan between 2000 and 2011. PET and CT positivity were defined as a standardized uptake value (SUV) of ≥3.0 and a lymph node size of ≥1 cm, respectively. Results: Of the 64 patients with postradiation PET/CT scans, 16 underwent a neck dissection after imaging. When either PET or CT was positive, all 9 cases of malignancy were identified (sensitivity 100%), and the positive predictive value was 69.2% (9/13). Of the 7 with benign pathology, 3 were negative for both modalities (specificity 42.8%). In this setting, the negative predictive value was 100% (3/3). Eight cases had discrepant results between PET and CT imaging. For these, the positive and negative predictive values with PET imaging alone were 33% and 20%, while with only CT imaging, they were 80% and 66%, respectively. Conclusion: Corroborating with prior studies, our study observed a high negative predictive value when using the combination of PET/CT in the management of neck disease in postradiated patients. When PET and CT results conflicted, CT imaging had a higher predictive accuracy compared to PET scans.