Role of Narrow-Band Imaging in Detection of Head and Neck Unknown Primary Squamous Cell Carcinoma

Objectives/Hypothesis: There is no general consensus on what kind of examination to include in an optimal detection strategy for head and neck squamous cell carcinomas of unknown primary (SCCUPs). This study investigates the role of narrow-band imaging (NBI) in their identification. Study Design: Case series. Methods: Twenty-nine consecutive patients affected by SCCUPs were referred at two academic institutions. Selection criteria were: 1) lymph node cytology positive for squamous cell carcinoma (SCC); 2) no evidence of any primary at white light (WL); 3) negative contrast-enhanced computed tomography (CT) and/or magnetic resonance imaging, and positron emission tomography-CT; and 4) no contraindication to general anesthesia. Each patient underwent office-based NBI panendoscopy. If a suspicious area was identified, a biopsy was performed for histological confirmation. When no suspicious area was detected in the office, patients underwent WL and NBI under general anesthesia. If this examination was still negative, bilateral tonsillectomy and base of the tongue (BOT) mucosectomy were performed. Results: Office-based NBI identified 10 SCCs (34.5%), one in the nasopharynx, three in the tonsil, four in the BOT, and two in the supraglottis. In only one (3.5%) NBI-negative patient was a primary found in the BOT. In one (3.5%) we found an NBI suspicious area during panendoscopy under general anesthesia, but histology did not confirm this finding after BOT mucosectomy. Seventeen (58.6%) patients remained as having true SCCUPs. NBI sensitivity, specificity, positive, negative predictive values, and accuracy were 91%, 95%, 91%, 95%, and 90%, respectively. Conclusions: In the present study, office-based NBI increased the detection rate of head and neck SCCUPs by 34.5% and should be strongly recommended as an adjunctive tool in their diagnostic workup.