Background Prevention of HIV and other STDs using vaginal microbicides must be safe. Colposcopy has not been shown to predict microbicide safety; therefore a more sensitive method is needed in safety evaluations of these vaginal products. The use of optical coherence tomography (OCT) has shown increased sensitivity to tissue injury over colposcopy in animal models. We describe a randomised double blind clinical trial using OCT to detect epithelial changes and injury related to vaginal microbicide use.

Methods 30 women aged 18–45 were randomised to use hydroxyethyl cellulose (HEC) placebo or nonoxynol-9 (N-9) vaginal gel twice daily for 5.5 days. Imaging with colposcopy and OCT was performed prior to product use, within 6 h of last dose of product, and 1 week after discontinuation of product. Colposcopy was graded based on vascular and epithelial disruption and erythema. OCT images were evaluated both by quantitative measurement of vaginal epithelial thickness and by use of a qualitative scoring system developed in the macaque to determine epithelial disruption and injury. Mixed model and significance of p=0.05 were used in data analysis.

Results Baseline colposcopy and OCT findings were similar between treatment groups. After treatment, there were no significant colposcopic differences in vascular or epithelial disruption between treatment groups, with only an increase in erythema noted after treatment in the N-9 group (p=0.01). OCT detected differences between groups in OCT scores (p<0.0001) and epithelial thickness, (p=0.008), both indicators of epithelial injury, after treatment with N-9. One week after discontinuation of treatment, OCT scores were similar between treatment groups (p=0.66) but epithelial thickness in the N-9 group was significantly thicker (p=0.0003).

Conclusions OCT can be used in a clinical setting to detect epithelial injury and is more sensitive than colposcopy to detect microscopic epithelial disruption. In addition, it gives information about the thickness of the epithelium, a measure previously available only through invasive biopsy. Vaginal epithelium thinned after treatment in the N-9 group, but then thickened after discontinuation of N-9, possibly due to post injury epithelial proliferation. OCT should be considered for inclusion in clinical trials for the detection of product-related toxicity to the vaginal and cervical epithelium.