Objective: Colposcopy has been used to detect epithelial damage with vaginal microbicides. In animal models, optical coherence tomography provided increased sensitivity over colposcopy in detecting epithelial injury. This randomized, double-blinded, clinical study compared optical coherence tomography to colposcopy for the evaluation of epithelial injury in women using placebo or nonoxynol-9.
Methods: Thirty women aged 18-45 were randomized to use hydroxyethyl cellulose placebo or nonoxynol-9 vaginal gel twice daily for 5.5 days. Imaging with colposcopy and optical coherence tomography was performed before product use, after the last dose, and 1 week later. Colposcopy was graded using standard criteria. Optical coherence tomography images were scored for epithelial integrity based on a published scoring system and were measured for epithelial thickness.
Results: Colposcopy findings, optical coherence tomography scores, and epithelial thicknesses were similar between treatment groups at baseline. After treatment, there were significant differences between the nonoxynol-9 (1.37) and control group (1.15) optical coherence tomography scores (P<.001), indicating epithelial injury, and there was epithelial thinning in the nonoxynol-9 group (237 micrometers) compared with the control group (292 micrometers; P=.008). There were no significant posttreatment colposcopic differences in epithelial disruption between treatment groups, with only increased erythema noted after nonoxynol-9 use (P=.02).
Conclusion: Optical coherence tomography detected epithelial disruption and thinning not identified by colposcopy. Vaginal epithelial thickness, a measure previously available only through biopsy, decreased after nonoxynol-9 use, a finding that may contribute to increased susceptibility to human immunodeficiency virus after frequent use. Optical coherence tomography shows promise for the noninvasive clinical assessment of vaginal epithelial damage.
Clinical trial registration: UMIN Clinical Trials Registry, www.umin.ac.jp/ctr/index.htm, R000006186.
Level of evidence: I.