Subgroups of Oral Epithelial Dysplasia Identified by Immunostaining

Objectives: To determine if staining for p53, p16^INK4a, and ki67 coupled with specific criteria for categorizing tissues as positive or negative for each marker could be used to 1) distinguish between cases of severe epithelial dysplasia or carcinoma in situ (very high grade dysplasia, VHGD) and controls with normal or reactive, non-dysplastic epithelium; and to 2) identify subgroups of VHGD with different staining patterns or clinical parameters.

Methods: Paraffin-embedded tissue from 35 consecutive intraoral cases diagnosed histopathologically as VHGD, and 35 non-neoplastic controls, were stained by immunoperoxidase with primary antibodies to p53 (DakoCytomation), p16^INK4a (Ab-4 Neomarkers), or ki67 (Vector Laboratories). Tissues were categorized by specific criteria as positive or negative for each marker. Ki67 scores from 1 to 5 were assigned according to the most superficial epithelial layer staining positively for ki67.

Results: None of the 35 non-dysplastic controls stained positively for p53 or p16. Based on p53 and p16 staining, 3 major groups of VHGD cases were identified. Group 1) p53⁺/p16^- (10 cases, p<0.01 vs. controls); group 2) p16⁺/p53^- (14 cases, p<0.01 vs. controls); group 3) p53^-/p16^- (9 cases). Two additional cases were p53⁺/p16⁺. Mean ki67 scores for controls and groups 1, 2 and 3 were 1.31, 2.50, 4.86 and 3.33, respectively (each group greater than controls, p<0.01). Group 2 (p16⁺/p53^-) had a higher mean ki67 score than groups 1 and 3 (p<0.02), a lower mean age than group 3 (50 vs. 66, p<0.01), and a trend toward a higher male to female ratio than group 1 (3.7 vs. 0.4, p=0.035). Thirty-four of 35 VGHD were positive for one of the markers (p53⁺, p16⁺, or ki67 score >2) versus none of the controls (p<<0.001).

Conclusions: Using specific criteria, positive staining for p53, p16, or ki67 differentiated between non-dysplastic controls and VHGD, and identified 3 subgroups of VHGD.