The efficacy of oral brush biopsy with computer-assisted analysis in identifying precancerous and cancerous lesions

Abstract

Background: Cancer of the oral cavity is the sixth most common malignancy reported worldwide and one with the highest mortality rate among all malignancies. There is a paucity of reliable diagnostic methods to detect early malignancies. This study was performed to evaluate the sensitivity and specificity of brush biopsy in identifying oral premalignant and malignant lesions.

Methods: Oral brush and scalpel biopsies were performed on 85 consecutive patients presenting with an oral lesion deemed to be minimally suspicious by clinical examination and the results were compared.

Results: Of 79 patients with adequate brush biopsy samples with matching scalpel biopsies, 27 revealed histopathologic evidence of dysplasia or carcinoma, 26 of which were independently identified with the oral brush biopsy (sensitivity: 96.3%-95% CI, 87%-100%). 52 oral lesions did not reveal any histopathologic evidence of dysplasia or carcinoma and of these, brush biopsy reported 47 as "negative" and 5 as "atypical"(specificity of "positive" brush biopsy result is 100%- 95% CI, 93%-100%; specificity for "atypical" brush biopsy result is 90.4%-95% CI, 82%-97%. The positive predictive value of an abnormal oral brush biopsy was 84% and the negative predictive value was 98%.

Conclusion: Our study demonstrated that the oral brush biopsy is an accurate test in identifying oral premalignant and malignant lesions, even if minimally suspicious.

Figure 1

Clinical examples of minimally suspicious lesions sampled.

Figure 2

Histopathologic specimen demonstrating the oral biopsy defect sampling the entire thickness of the epithelium. (H &E x100).

Figure 3

Panorama of atypical and malignant cells identified from a brush biopsy specimen with the aid of a highly specialized neural network-based image-processing system. (Pap × 1000).