Diagnostic efficiency of differentiating small cancerous and precancerous lesions using mucosal brush smears of the oral cavity--a prospective and blinded study.

The aim of this study was to evaluate the diagnostic accuracy of oral brush biopsy to identify early malignancy. One hundred and eighty-six brush biopsies of suspicious mucosal lesions were obtained, haematoxilin and eosin (H&E)-stained and compared with the histology of conventional excision biopsies of the same site performed concomitantly. The sensitivity for identifying squamous cell carcinoma (SCC) was 88.5%. High-risk lesions including squamous intraepithelial neoplasia (SIN II, SIN III) and SCC were identified with a sensitivity of 86.4%, using a pap-analogous classification, which is considered to be carcinomatous, as well as moderate and severe dysplastic cells positive. Depending on the cytopathologic definition for malignancy and the tumour size, the test accuracy varied: Extending the cytopathologic criteria for malignancy by defining all dysplastic or malignant cytopathologic findings as positive, the sensitivity was increased to 95.2% at the expense of the specificity, which was reduced from 94.9% to 82.3%. Separately analysing SCCs of less than 20 mm, the sensitivity was reduced by 9.5% to 78%. Although small malignant lesions seem to be less reliable by the conventional oral brush biopsy, it is a useful screening instrument for early diagnosis of suspicious, epithelial lesions and could therefore contribute to improved cancer prognosis.