Evaluation of proliferating cell nuclear antigen as a surrogate end point biomarker in actinic keratosis and adjacent, normal-appearing, and non-sun-exposed human skin samples.

The incidence of nonmelanoma skin cancer, including both squamous cell carcinoma and basal cell carcinoma, is a significant health problem in the United States. Actinic keratosis (AK), the precursor of cutaneous squamous cell carcinoma, is a major risk factor for nonmelanoma skin cancer. In addition, AKs are tissue targets for the identification of biomarkers for use in chemopreventive studies. The biomarker addressed in this study is epidermal cell proliferation, as quantitated by proliferating cell nuclear antigen (PCNA). Shave biopsies were obtained from AKs, tissue immediately adjacent to AKs, normal-appearing, upper-medial arm skin, and non-sun-exposed skin from 19 subjects. When any degree of PCNA staining was considered positive (semiquantitative 1-4 scale), there was a significant difference and a progressively increasing mean PCNA labeling index (LI) in the total epidermis (basal and suprabasal layers), beginning with non-sun-exposed buttock skin, with the lowest LI (2.5 + or - 1.6%), followed by upper-medial arm skin (12.3 + or - 7.4%; P = 0.0015), skin adjacent to AKs (19.2 + or - 12.2%; P = 0.0218), and finally, AKs with the highest LI (34.6 + or - 20.1%; P = 0.0017). This same pattern was observed when the epidermis was separated into basal and suprabasal layers, with the exception of a nonsignificant result for upper-medial arm skin compared with adjacent skin in the basal layer (P = 0.3981). PCNA LIs were also analyzed separately by staining intensity (i.e., scores of 1-4). The PCNA LI in skin with varying degrees of sun damage and/or histological atypia is a candidate surrogate end point biomarker for skin cancer chemoprevention studies.