Electronic Thesis and Dissertation Repository

Thesis Format



Master of Science




Darling, Mark

2nd Supervisor

Armstrong, Jerrold

Joint Supervisor


Five-year survival of oral cancer has remained relatively unchanged despite advancements in treatment, mostly because diagnosis is often made at an advanced stage of disease. The progression of dysplasia to oral cancer often follows a stepwise progression. Histopathology is considered the ‘gold standard’ for diagnosing dysplasia and lesions at a high risk of progression to oral cancer, but lends itself to subjectivity. The protein biomarker, S100A7, in oral dysplasia and squamous cell carcinoma has shown some predictive value for the transformation of dysplasia to cancer. Straticyte, a diagnostic test utilizing S100A7 to predict the probability of progression of oral dysplasia to malignancy, has recently been developed. Straticyte has never been used to predict progression of oral dysplasia alone.

The objective of this study is to determine if S100A7 is a valuable biomarker in predicting the progression of oral dysplasia. We also evaluated if Straticyte is a useful tool to predict oral dysplasia progression.

Formalin-fixed paraffin-embedded (FFPE) specimens were obtained from the Tissue Archives of the Division of Oral Pathology at Western University. This study included 29 cases of progressing oral dysplastic lesions, 17 cases of non-progressing oral dysplastic lesions and 25 control cases of normal tissue. FFPE sections were stained for S100A7, cell cycle-related protein cyclin D1, and B-Catenin using standard immunohistochemistry. Immunoreactivity of S100A7 was evaluated semi-quantitatively, using an intensity and proportion scale, as well as quantitatively using an automated scoring method (Straticyte) by image analysis. The data was analyzed to compare the manual and automated scoring methods and to look for a correlation between S100A7 expression and progression of disease. Cyclin D1 and B-catenin, other protein biomarkers, were also analyzed qualitatively.

Mean manual score for the initial biopsy of the progressing, non-progressing and control groups was 4.93, 4.83 and 3.52 respectively. The mean Straticyte score for the initial biopsy of the progressing, non-progressing and control groups was 25.93, 34.91 and 30.65. Stepwise regression analysis showed the manual scoring method to be the best predictor of lesions being non-progressing compared to controls (p=0.016). The same analysis also showed the automated scoring method to be the best predictor of lesions being progressing compared to non-progressing (p=0.078).

Neither the manual or the automated scoring methods proved to be significantly superior to the other in predicting progressing of oral dysplasia. S100A7 did not prove to be a useful biomarker in predicting progressing of oral dysplasia. More studies are needed to determine both the usefulness of S100A7 and Straticyte for predicting progression of oral dysplasia.

Summary for Lay Audience

Oral cancer is a disease that can lead to suffering and pre-mature death. The early identification of oral cancer can be difficult and for this reason when oral cancer is detected it is often large and has sometimes already spread to other parts of the body. In this study, we evaluated if a protein, S100A7, can assist with detecting oral lesions (abnormal oral tissue) that have a high likelihood of developing into oral cancer. To explore this, we used a staining technique called immunohistochemistry (IHC), that allows for the stain to be attached to the protein, S100A7, so that it can be visualized under a microscope. We measured the amount of S100A7 both manually (manual method) and using a commercial test called, Straticyte (automated method), that uses machine learning. We compared the manual and automated methods to determine if either method was superior at measuring the amount of S100A7 and predicting which lesions were high risk.

We determined that both methods showed some value; however, neither the manual nor the automated method proved to be clearly superior at measuring S100A7 and predicting high risk lesions.

We concluded that further studies are warranted to determine the value of S100A7 in identifying high risk lesions and in validating the usefulness of Straticyte.