Topical application of a mucoadhesive freeze-dried black raspberry gel induces clinical and histologic regression and reduces loss of heterozygosity events in premalignant oral intraepithelial lesions: results from a multicentered, placebo-controlled clinical trial

Abstract

Purpose: Approximately 30% higher grade premalignant oral intraepithelial neoplasia (OIN) lesions will progress to oral cancer. Although surgery is the OIN treatment mainstay, many OIN lesions recur, which is highly problematic for both surgeons and patients. This clinical trial assessed the chemopreventive efficacy of a natural product-based bioadhesive gel on OIN lesions.

Experimental design: This placebo-controlled multicenter study investigated the effects of topical application of bioadhesive gels that contained either 10% w/w freeze-dried black raspberries (BRB) or an identical formulation devoid of BRB placebo to biopsy-confirmed OIN lesions (0.5 g × q.i.d., 12 weeks). Baseline evaluative parameters (size, histologic grade, LOH events) were comparable in the randomly assigned BRB (n = 22) and placebo (n = 18) gel cohorts. Evaluative parameters were: histologic grade, clinical size, and LOH.

Results: Topical application of the BRB gel to OIN lesions resulted in statistically significant reductions in lesional sizes, histologic grades, and LOH events. In contrast, placebo gel lesions demonstrated a significant increase in lesional size and no significant effects on histologic grade or LOH events. Collectively, these data strongly support BRB's chemopreventive impact. A cohort of very BRB-responsive patients, as demonstrated by high therapeutic efficacy, was identified. Corresponding protein profiling studies, which demonstrated higher pretreatment levels of BRB metabolic and keratinocyte differentiation enzymes in BRB-responsive lesions, reinforce the importance of local metabolism and differentiation competency.

Conclusions: Results from this trial substantiate the LOH reductions identified in the pilot BRB gel study and extend therapeutic effects to significant improvements in histologic grade and lesional size.

Conflict of interest statement

Disclosure of potential conflicts of interest: No potential conflicts of interest were identified.

©2014 AACR.

Figures

Figure 1

Effect of gel treatment on lesional size. Histograms A. (BRB gel) and B.(Placebo gel) depict pre- and post-treatment lesional size (mm2) of individual subjects’ OIN lesions. The lesional sizes of subject A4 in BRB group and subject P11 in Placebo group were not included as the extensive distribution and confluent nature of their dysplastic lesions made accurate measurement impossible. C. The mean pretreatment lesional sizes of both groups were statistically comparable. Intragroup pre/post lesional sizes significantly decreased in BRB gel treatment group while significantly increased in Placebo group. D. Intergroup comparison reviewed a significant difference between BRB and Placebo groups. E. This clinical photograph of subject A6 was taken prior to the initial biopsy of the crisply defined rhomboid white plaque close to commissure. The area of the residual lesion remained after initial biopsy was measured as the pretreatment biopsy size. F. Clinical photograph at the same site after 3 months BRB gel treatment and immediately prior to the final biopsy. The remaining lesion had completely regressed and the buccal mucosa at the treatment site had a normal clinical appearance.

Figure 2

Effect of gel treatment on histologic grade. Histograms A. (BRB gel) and B. (Placebo gel) depict pre- and post-treatment histologic grades of individual subjects’ OIN lesions using the following scale: 0=normal, 1=atypia, 2=mild dysplasia, 3=mild to moderate dysplasia, 4=moderate dysplasia, 5=moderate to severe dysplasia, 6=severe dysplasia. None of the pre- and post- treatment specimens was diagnosed as carcinoma in situ (7) or OSCC (8). C. The baseline histologic grades of both pretreatment groups were statistically comparable. Intragroup pre/post histologic grades significantly decreased in BRB gel treatment group while changes in Placebo group were statistically insignificant. D. Categorization of subjects as per histopathologic responsiveness. E. and F. Photomicrographs of pre- and post- treatment specimens of subject A6 (10x) demonstrated complete histopathologic regression from mild dysplasia (E, pretreatment) to normal epithelium (F, posttreatment). G. and H. pre- and post- treatment photomicrographs of subject A7 (10x) showed partial regression from severe dysplasia (G, pretreatment) to mild to moderate dysplasia (H, posttreatment).

Figure 3

Effect of gel treatment on loss of heterozygosity (LOH). A. and B. pre- and post-treatment detected LOH events of individual subjects in BRB gel and Placebo gel groups, respectively. C. Intragroup and intergroup statistical analyses. Mean LOH events significantly decreased in BRB gel treatment group while changes in Placebo group were statistically insignificant. The baseline LOH events of both pretreatment groups were statistically comparable. D. classification of subjects as per treatment effects on LOH events. E. and F. Representative genotyping data depict an LOH event occurred on marker D9S171 in subject A19’s pretreatment samples [loss of one allele (al 154) in the epithelium tissue (E), compared to the two alleles (al 154 and al 162) in the patient’s matched normal connective tissue (F)]. G. and H. The lost allele (al 154) of D9S171 was recovered in the epithelial tissue (G) of the same patient after 3 months of BRB gel treatment, and the ratio of peak heights in epithelium (G) was comparable to the connective tissue sample (H).

Figure 4

Overall therapeutic responsiveness. A. and B. cumulative responsiveness score of individual subjects in BRB (A) and Placebo (B) groups. The cumulative responsiveness score is defined as the sum of lesional size score (−3 to 3 as per the percent quartile of pre-/post- lesional size changes i.e. 75–100% reduction=3, 50–74% reduction=2, 25–49% reduction=1, 24–0% reduction=0, 0–24% increase=0, 25–49% increase= −1, 50–74% increase= −2, > 75% increase= −3), histologic grade score (pretreatment histologic grade – posttreatment histologic grade) and LOH score (pretreatment LOH events – posttreatment LOH events). C. Comparison of the mean cumulative responsiveness scores between BRB and Placebo groups. D, Categorization of subjects according to their cumulative responsiveness score, where high responder ≥4, intermediate responder=3, low responder=2 or 1, and non-responder ≤0. E. A significant correlation was demonstrated between BRB gel treatment effects on histologic effects and lesional size. No comparable correlation was detected in the placebo gel group.

Figure 5

Densitometry analyses of Protein Immunoprecipitation studies. Five proteins associated with keratinocyte terminal differentiation (TGase 1, high molecular weight involucrin, low molecular weight involucrin, loricrin, cytokeratin 10/13) and two with BRB metabolism (UDP-Glc dehydrogenase and UGT1A) were evaluated in pre- and post- treatment biopsy samples of those lesions that had adequate tissue. All densitometry results were normalized to the percentage of sample-matched pancytokeratin level, which reflected the proportion of epithelium tissue in each sample. Previous studies have confirmed the epithelial distribution of UDP-Glc dehydrogenase and UGT1(19). Responder and non-responder were defined according to the cumulative responsiveness score (responder≥1, non-responder ≤0). A. and B. pre- and post- treatment protein levels in BRB group. Responders demonstrated a trend of possessing higher level of associated proteins relative to the non-responders. C and D, pre- and post- treatment protein levels in Placebo group. No statistical significance was identified.