OBJECTIVES: We developed an objective endoscopic score of abnormality of the nasopharynx to predict the likelihood of malignancy. METHODS: A score sheet with 44 variables was developed to objectively quantify the bilateral endoscopic assessment of the nasopharynx. Patients scheduled to undergo nasopharyngeal biopsies were recruited. The nasopharynx was assessed endoscopically, photographed, and scored on 44 variables. The scores were compared to the biopsy results, and predictors of malignancy were modeled with regression analysis. The sensitivity and specificity of the novel scoring system were examined. RESULTS: Seventeen patients had carcinoma, and 60 had a benign lesion or no disease. Patients with a nasopharyngeal malignancy scored significantly higher than did patients with a benign lesion or no disease. No patient with a malignant lesion had a score of less than 12. With a receiver operating characteristic curve area of 0.917, the score demonstrated an excellent ability to discriminate between nasopharynges that were likely or unlikely to contain malignant disease. Independent predictors for both malignant disease and a score greater than 12 were modeled. CONCLUSIONS: A cutoff score above 12 on the novel objective endoscopic assessment of the nasopharynx measure was highly predictive of possible malignancy.
OBJECTIVES: We developed an objective endoscopic score of abnormality of the nasopharynx to predict the likelihood of malignancy. METHODS: A score sheet with 44 variables was developed to objectively quantify the bilateral endoscopic assessment of the nasopharynx. Patients scheduled to undergo nasopharyngeal biopsies were recruited. The nasopharynx was assessed endoscopically, photographed, and scored on 44 variables. The scores were compared to the biopsy results, and predictors of malignancy were modeled with regression analysis. The sensitivity and specificity of the novel scoring system were examined. RESULTS: Seventeen patients had carcinoma, and 60 had a benign lesion or no disease. Patients with a nasopharyngeal malignancy scored significantly higher than did patients with a benign lesion or no disease. No patient with a malignant lesion had a score of less than 12. With a receiver operating characteristic curve area of 0.917, the score demonstrated an excellent ability to discriminate between nasopharynges that were likely or unlikely to contain malignant disease. Independent predictors for both malignant disease and a score greater than 12 were modeled. CONCLUSIONS: A cutoff score above 12 on the novel objective endoscopic assessment of the nasopharynx measure was highly predictive of possible malignancy.