BackgroundSecond primary cancer of the esophagus is frequent in head and neck patients, especially in high-risk populations, and has a great impact on the prognosis. Although Positron emission tomography (PET)/computed tomography (CT) scan is commonly conducted in head and neck patients, its ability to detect early esophageal cancer is limited. Narrow-band imaging endoscopy is an accurate and convenient technique for esophageal examination. We aimed to compare PET/CT scan and narrow-band imaging endoscopy for the detection of esophageal cancer in head and neck cancer patients.MethodsFrom November 2015 to November 2018, all head and neck cancer patients who underwent both PET/CT scan and narrow-band imaging endoscopy at Changhua Christian Hospital were retrospectively enrolled. Descriptive statistics, receiver operating characteristic curve analysis, logistic regression analysis, independent Student’s t-test, and Kaplan–Meier survival analysis were conducted with MedCalc Statistical Software.ResultsA total of 147 subjects were included in the analysis; suspicious esophageal lesions were identified by PET/CT scan in 8 (5.44%) and by narrow-band imaging in 35 (23.81%). The final pathologic diagnoses were esophageal squamous cell carcinoma in 10 and high-grade dysplasia in 5. The respective sensitivity, specificity, and area under the curve for detecting suspicious esophageal lesions were 33.33, 97.73%, and 0.655 for PET/CT scan, and 100.0, 84.85%, and 0.924 for narrow-band imaging endoscopy. Hypopharyngeal or laryngeal location of the primary head and neck cancer was the only risk factor for developing second primary esophageal cancer.ConclusionsPET/CT scan was inferior to narrow-band imaging endoscopy in detecting second primary esophageal cancer in head and neck cancer patients. In addition to PET/CT scan, narrow-band imaging endoscopy should be considered in head and neck patients at high risk for developing second primary esophageal cancer.