FDG PET/CT for metastatic squamous cell carcinoma of unknown primary of the head and neck.

OBJECTIVE

The role of FDG PET/CT is uncertain in squamous cell carcinoma of unknown primary (SCCUP). Published data are limited by heterogeneity of pathology, varied diagnostic criteria, and small sample size.

MATERIALS/

METHODS

This retrospective study was conducted in 40 patients who had FDG PET/CT for SCCUP. Prior to PET/CT, patients had underwent standard clinical evaluations including flexible rhinolaryngoscopy. The majority of the patients had anatomic imaging. All subjects had histopathological investigations, and follow-up data after PET/CT.

RESULTS

FDG PET/CT detected the primary in 16 of 40 patients (40%), and the most common locations of the primary were in the base of the tongue, palatine tonsil, and hypopharynx. There were 10 false positive and 6 false negative PET/CT scans. There was no difference in the mean SUVs of the suspected primary lesions between true and false positive groups. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall accuracy of FDG PET/CT for SCCUP in this patient cohort were 72.7% (16/22), 44.4% (8/18), 61.5% (16/26), 57.1% (8/14), and 60% (24/40), respectively. The tonsil was the site with the most false positive and false negative findings.

CONCLUSIONS

FDG PET/CT might be an effective single shot of whole-body imaging for detection of the primary in SCCUP. Low specificity remained the most notable weakness of FDG-PET in work-up for SCCUP. In addition to varied physiologic uptake or inflammatory related uptake in the common locations of the primary, the image indication bias might be another significant contributor of high false positive rate.