Autofluorescence Visualization Detection for Oral Epithelial Dysplasia

To detect oral epithelial dysplasia, iodine staining method is generally used.  However, iodine staining has weak points of mucosal irritation and no adaptation in the keratinized mucosa.  Recently, autofluorescence visualization (AV) method has been noticed instead of conventional iodine staining (IS) method.  To carry out the method, an autofluorescence visualization device, which emits a cone of light in the blue spectrum between 400-460 nm wavelength into the oral mucosa, is used.  Normal epithelium emits a green autofluorescence, but oral epithelial dysplasia reveals dark due to fluorescence visualization loss (FVL).  AV method detects oral epithelial dysplasia, according to the relative luminance difference.  The purpose of this study is to evaluate the clinical usefulness of AV method with VELscope^®to detect oral epithelial dysplasia.  

The subjects were 31 oral mucosal lesions in the 19 patients (12 men and 7 women, mean age 64.1 y/o) referred to the department of Oral and Maxillofacial Surgery, Tsurumi University Dental Hospital, between February and September in 2014.  All subjective lesions were evaluated by AV method, iodine staining method and biopsy.  The digital photo documents of VELscope^® were analyzed to quantize the relative luminance difference between the blackest point of FVL and the average of three whitest points from the surrounding areas by using Image J^®software.  The cut-off value of the relative luminance difference was calculated area under receiver operating characteristic curve.  One-way analysis of variance and Mann-Whitney U test were used for statistical analysis, and probability value less than 0.05 was considered as statistical significance.  Subsequently, the cut-off value of the relative luminance difference has been set to 43.9 for determination of a possibility of oral epithelial dysplasia.  The accuracy of AV method was assessed and compared to that of IS method.  Histopathological diagnosis was set as the gold standard of this study.

 Histopathologically, the 13 lesions were diagnosed as squamous cell carcinoma (SCC), 11 lesions as leukoplakia with dysplasia (LD), 5 lesions as leukoplakia without dysplasia (LND), and 2 lesions as lichen planus (LP).  The mean relative luminance differences of SCC, LD, LND and LP were 87.0, 64.1, 11.7 and 61.4, respectively.  In the examination with AV method,epithelial dysplasia was detected in the 23 of 31 lesions, and remaining 8 lesions had no dysplasia.  The accuracy of AV method was 90.3%, sensitivity was 91.7%, and specificity was 85.7%.  Meanwhile, 24 lesions had dysplasia and 7 lesions had no dysplasia when IS method.  Consequently, the accuracy of iodine staining method was 74.2%, sensitivity was 83.3%, and specificity was 42.9%.  Moreover, the grade of epithelial dysplasia is approximately proportional to the level of relative luminance differences. 

Based on the mentioned results, the accuracy of AV method was higher than that of IS method, and it is suggested that AV method is more appropriate to detect oral epithelial dysplasia.  Although the sensitivity of AV method was almost as same as that of IS method, its specificity was much higher than that of IS method.  Therefore, AV method was suggested to have a low risk for overdiagnosis, comparing to IS method, and it has possibility to expect the grade of dysplasia by the relative luminance difference. 

In conclusion, AV method seems to be useful modality for detecting oral epithelial dysplasia instead of conventional IS method.

Farah CS. Head Neck. 2012;34(6):856-62.

Awan KH. Oral Oncol. 2011;47(4):274-7.