Diagnostic value of diffusion-weighted imaging and 18F-FDG-PET/CT for the detection of unknown primary head and neck cancer in patients presenting with cervical metastasis

Background and purpose: Head and neck squamous cell carcinoma (HNSCC) may present with cervical metastases without an apparent primary tumor. Detecting the primary tumor results in more targeted treatment. Acquisition of DWI is improving with less artifacts and image distortion. We assessed the diagnostic value of DWI and 18F-FDG-PET/CT for detecting primary tumors in patients presenting with nodal metastasis of an unknown primary HNSCC. Materials and methods: For this retrospective study we included 31 patients (male/female ratio = 23/8, median age = 66 years, age range = 40–80 years) who presented with a pathologically proven cervical nodal metastasis from HNSCC without overt primary tumor location between January 2013 and November 2016 and underwent both DWI and 18F-FDG-PET/CT. Both modalities were assessed qualitatively and quantitatively. With ROC analysis we determined the optimal cut-off for imaging parameters in separating occult malignancy from benign tissue. Results: Qualitative analysis of MRI including DWI resulted in a sensitivity of 81.3% (95%CI) = 53.7–95.0) and specificity of 73.3% (95%CI = 44.8–91.1). With qualitative scoring of 18F-FDG-PET/CT a sensitivity and specificity of 93.8% (95%CI = 67.8–99.7) and 73.3% (95%CI = 44.8–91.1) were found. With quantitative analysis sensitivity and specificity of SUVmax were 81.3% (95%CI = 53.6–95.0) and 93.3% (95%CI = 66.0–99.7), respectively. Combining DWI and 18F-FDG-PET/CT resulted in a sensitivity of 93.8% (95%CI = 67.7–99.7%) and specificity of 60.0% (95%CI = 32.9–82.5%). Conclusion: In this study on HNSCC patients presenting with clinically UP lesions the diagnostic accuracy of qualitative analysis with DWI and 18F-FDG-PET/CT and quantitative analysis of 18F-FDG-PET/CT using SUVmax were high. Adding DWI did not improve the accuracy of 18F-FDG-PET/CT