Background: Early detection of esophageal cancer in patients with head and neck cancers may alter treatment planning and improve survival. However, standard endoscopic screening is not feasible for some patients with tumor-related airway compromise or postirradiation trismus.
Objective: To evaluate a novel, sequential approach by integrating ultrathin endoscopy with narrow-band imaging and Lugol chromoendoscopy.
Design: Cross-sectional study.
Setting: Single center in Taiwan.
Patients: Forty-four consecutive patients with transoral difficulty screened for synchronous or metachronous esophageal cancer.
Main outcome measurements: Sensitivity, specificity, and accuracy in the detection of mucosal high-grade neoplasia or invasive cancer.
Results: Fifty-four endoscopic interpretations were obtained, and 11 mucosal high-grade neoplasia and 7 invasive cancers were confirmed by histology. The mean examination time was 19.4 minutes (range 7.9-35.2 minutes), and all patients tolerated the procedure well. Sensitivity, specificity, and accuracy (with 95% CI) were 55.6% (95% CI, 33.5%-75.6%), 97.2% (95% CI, 85.8%-99.3%), and 83.3% (95% CI, 71.2%-90.9%), respectively, for standard endoscopy; 88.9% (95% CI, 66.9%-96.6%), 97.2% (95% CI, 85.8%-99.3%), and 94.4% (95% CI, 84.9%-97.9%), respectively, with the adjunct of narrow-band imaging; and 88.9% (95% CI, 66.9%-96.6%), 72.2% (95% CI, 55.9%-84.1%), and 77.8% (95% CI, 64.9%-86.8%), respectively, with the adjunct of Lugol chromoendoscopy. When we integrated all interpretations on the basis of the sequential approach, the estimated probability of false-negative findings was 1.2% (95% CI, 0.1%-4.6%).
Limitations: Inherent shortcomings of ultrathin endoscopy, such as its resolution, light source, and lack of magnification.
Conclusions: The use of ultrathin endoscopy in a sequential approach for multimodal detection is feasible in patients with transoral difficulty and substantially increases the detection rate of synchronous or metachronous neoplasms.