Objectives: The role of fused modality [F]-2-fluorodeoxyglucose-positron emission tomography/computed tomography (PET/CT) in diagnosing and accurately staging patients with primary, metastatic, and recurrent head and neck (HN) cancer is evolving, and the clinical implications need to be further defined. A few retrospective studies have been performed, but adequate sample sizes are lacking because the number of HN cancer patients is relatively small. This study evaluates the positive predictive value (PPV), sensitivity, specificity, and accuracy of PET/CT in previously untreated HN cancer patients at a single tertiary care institution. The purpose of this study is to evaluate the role of this new technology in the management of previously untreated HN cancer patients.
Study design: Retrospective cohort outcomes study at a tertiary National Cancer Institute Comprehensive Cancer Center.
Materials and methods: Institutional review board exemption #4 (45 CFR 46.101 [4]) criteria were applied for and accepted by the office of responsible research practices at the Ohio State University College of Medicine. The authors identified 268 consecutive PET/CT examinations between March 2005 and January 2006 for HN cancer ordered by the two senior authors at the James Cancer Hospital and Solove Research Institute of the Ohio State University Medical Center. PET/CT examinations were interpreted by one of three neuroradiologists. PPV, sensitivity, specificity, accuracy, diagnostic upstaging, and treatment management changes were determined from subset analysis of 123 previously untreated patients with HN cancer. Synchronous lesions were detected in 10 patients with use of this modality. PET/CT was also used to help manage 22 patients with unknown primary HN cancer. The statistics were verified by comparing PET/CT results with surgical specimen histopathology.
Results: : PET/CT was true-positive in 82.9% (102/123), with a per patient PPV of 87.2% and a per lesion PPV of 89.4%. PET/CT was false-positive in 12.2% (15/123) of patients and had a false-positive rate of 8.3% when calculated per lesion. In 67 patients who underwent neck dissection, PET/CT had a PPV of 92.7%. The accuracy was 89.7% in 20 patients who had bilateral neck dissections. The unknown primary site was found in 72.7% (16/22) of patients with unknown primary HN cancer. Synchronous lesions were found in 8.1% of patients by PET/CT, with a PPV of 66.6%. Distant metastases were detected in 15.4% (19/123) of patients. Treatment was altered in 30.9% (38/123) of patients as a result of this imaging modality.
Conclusions: The benefit of the PET/CT imaging modality resides in its fusion of anatomic detail of the HN region with the sensitivity of detecting tumors with increased metabolic activity at distant sites. Treatment was altered in 30.9% of our previously untreated HN cancer patients because of this imaging technique, with altered treatment including upstaging, diagnosing distant and unresectable disease, and working-up second primary malignancies. The false-positive findings did not result in additional morbidity to these patients. Although PET/CT is sensitive in detecting occult cervical nodal metastases, it does not yet have the ability to replace neck dissection as the diagnostic standard of care. This study supports the use of PET/CT in patients with newly diagnosed HN cancer because of its high PPV and superiority of detecting distant metastases and synchronous lesions.