UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Improving risk prediction for the malignant transformation of low-grade oral dysplasia - clinicopathological features and loss of heterozygosity Rock, Leigha Duree


A major barrier to oral cancer prevention is the lack of risk predictors for the malignant progression of oral potentially malignant lesions (OPML). OPML with evidence of dysplasia are at risk of progressing to oral cancer. However, not all will progress and predicting which low- grade dysplasia (LGD; mild/moderate dysplasia) are at risk of progression is challenging. The overall goal of this thesis was to advance risk stratification and to improve the prediction of malignant progression in LGD. Three research projects were developed to accomplish this goal. Each identified important insights into the phenotypic changes associated with malignant transformation and advanced risk prediction by exploring the association between histological, clinical and molecular biomarkers and malignant progression. The first project revealed that dysplasia with or without lichenoid mucositis (LM) had similar cancer risk and that pathologists and clinicians should not discount dysplasia in the presence of LM. The second project compared the clinical and molecular features of LGD in smokers in contrast to those of non-smokers (NS) and confirmed that NS possess an increased risk of progression, and progressed more quickly, than smokers. These findings emphasize the need for clinicians to consider smoking history (or the lack thereof) and molecular profiles in the triage and management of LGD. The final project aimed to advance a risk prediction model using microsatellite analysis for loss of heterozygosity (LOH) and repeated measures of clinicopathological features. Multivariable analysis showed that after LOH risk category, temporal repeated measures of toluidine blue status was the most significant predictor of progression. Two risk prediction models are presented and provide a systematic decision-making process for these very heterogeneous group of lesions. Patients at higher risk could be offered intensified surveillance or targeted interventions based on their predicted risk of disease, while patients at low risk would be spared from excessive screening and treatment. This body of work has advanced the risk stratification of LGD and presents an important framework to give scientists and clinicians a better view into the natural history of the disease and a novel approach to integrate repeated measurements of change over time into risk models.

Item Media

Item Citations and Data


Attribution-NonCommercial-NoDerivatives 4.0 International