[Analysis of risk factors for carcinogenesis of oral leukoplakia]

Abstract

Objective: To investigate the risk factors for malignant transformation of oral leukoplakia.

Methods: A total of 409 cases with oral leukoplakia was retrospectively analyzed. Single factor test was first performed to examine the associations between oral leukoplakia's histopathological classification and each of risk factors including sex, age, systemic diseases, course of disease, clinical classification, site, size, numbers of lesion, alcohol and tobacco consumption, and symptom. Then the association of these selected factors with oral leukoplakia's histopathological classification was evaluated using multiple logistic regression analysis.

Results: Fifty-two cases of all 409 patients with oral leukoplakia (including 9 severe dysplasia) developed oral cancer. The ratio of malignant transformation was 12.7%. Sex, age, clinical type, site and symptom were chosen as risk factors incorporated into the multiple logistic regression models. The risk of mild-moderate dysplasia in female was 2.40 times as high as that in male. The risk of mild-moderate dysplasia of speckled leukoplakia was 2.81 times as high as that of homogeneous leukoplakia. The risk of mild-moderate dysplasia of dangerous site was 1. 98 times as high as that non-dangerous site. The risk of mild-moderate dysplasia with symptom was 1.84 times as high as that without symptom. The risk of severe dysplasia and oral cancer in female was 3.11 times as high as that in male. The risk of severe dysplasia and oral cancer of speckled (4.50 times), ulcerative (5.63 times), verrucous leukoplakia (4.09 times) were much higher than that of homogeneous leukoplakia. The risk of severe dysplasia and oral cancer in dangerous site was 2.79 times as high as in non-dangerous site. The risk of severe dysplasia and oral cancer in leukoplakia with symptom was 4.38 times as high as without symptom.

Conclusions: The malignant transformation of oral leukoplakia is correlated to sex, clinical type, site and symptom.