Chemopreventive effects of a topically applied black raspberry gel on oral premalignant lesions

B35

Approximately 36% of untreated oral intraepithelial neoplasia (IEN) lesions undergo malignant transformation to oral squamous cell carcinoma (SCC). There is currently no means to accurately predict which IEN lesions will transform. Further, clinical management is complicated by recurrences of previously excised IEN tissues or development of new premalignant lesions. This study assessed the effects of topical application of a bioadhesive 10% (w/w) freeze dried black raspberry gel on oral IEN parameters that included histopathologic diagnosis, loss of heterozygosity (LOH) indices, gene expression profiles and levels of cyclooxygenase 2 (COX2) and inducible nitric oxide synthase (iNOS). Each participant with oral IEN lesions served as their own internal control. All lesions were initially photographed, and lesional tissue was hemisected to obtain a pre-treatment diagnosis and establish baseline biochemical and molecular parameters. Gel dosing to the remaining lesional tissue (0.5 grams applied four times daily for six weeks) was initiated one week after the initial biopsy. Genomic DNA was isolated from laser captured basilar and suprabasilar surface epithelial cells, followed by PCR amplification using primer sets that targeted known and presumed tumor suppressor gene loci associated with 9p21 (INK4a/ARF), 17p13 (p53), and 3p14 (FHIT). RNA was isolated from snap frozen tissues for microarray analyses, followed by QRT-PCR validation. Supplemental, epithelial gene specific RT-PCR analyses, and image analysis quantified immunohistochemistry (IHC) for lesional epithelial COX2 and iNOS were also conducted. None of the participants (10 normal controls, 20 IEN) developed any adverse side effects during treatment. Histopathologic results demonstrate appreciable interpatient variability, as 41% of our participants showed a post-treatment decrease in lesional grade, 23% showed an increase in lesional grade, and 35% of the patients showed stable disease. Treatment effects (Exact McNemar test) on LOH status demonstrated: 1) a decrease 24 events with comparison of all loci (p<0.0005), 2) site specific comparisons-decrease at initial biopsy site (10 events, p<0.05) and residual treated lesional site (14 events, p<0.005), 3) a decrease in 9p21-22 specific markers (16 events, p<0.005). Microarray data revealed a uniform suppression of genes associated with RNA processing and intracellular signaling. The relative fold changes obtained by QRT-PCR paralleled the trends observed in the gene chip. Supplemental, epithelial targeted QRT-PCR analyses demonstrated berry gel treatment markedly upregulated genes associated with keratinocyte differentiation and tumor suppression in a subset of patients. IHC analyses (Chi square) showed a post treatment reduction in lesional surface epithelial levels of the pro-inflammatory proteins COX2 (p<0.001) and iNOS (p<0.005). These data, which show topical berry gel application significantly reduced LOH indices, modulated gene expression profiles towards a less biologically active, more terminally differentiated tissue, and reduced epithelial COX2 and iNOS proteins, imply that further evaluation of berry gels for oral IEN chemoprevention is warranted.