Improving Adherence to Post-Cervical Biopsy Sexual Abstinence in Kenyan Female Sex Workers

Abstract

Problem: Cervical biopsies offer a unique opportunity for studying local immune response. To investigate hormonally induced immune fluctuations in cervical tissues of Kenyan female sex workers, we improved biopsy sampling protocol safety. Here, we report on steps taken to minimize exposure to HIV following two cervical biopsies.

Methods of study: Women were asked to abstain from vaginal intercourse to limit HIV exposure during wound healing with financial compensation. A comprehension tool for informed consent, on-site detection of prostate-specific antigens indicating unprotected intercourse within 48 hr, and bi-weekly text message reminders were implemented.

Results: The implemented methods improved compliance with post-procedure abstinence by two times (P = 0.013). Fourteen days following a cervical biopsy, no sign of genital inflammation or change in HIV T-cell target proportion were observed.

Conclusions: This study provides new tools for limiting HIV exposure in studies requiring biopsy sampling among women at risk of acquiring HIV.

Keywords: Female genital tract; HIV; immunology; mucosal; prostate-specific antigens; safety.

Figure 1

Schematic representation of the study visits. Participants were followed for 5 visits (1 month). The first biopsy sample was collected at visit 1, and a follow‐up visit to monitor the healing process took place 3–5 days post‐procedure (visit 2). The second biopsy (visit 3) was taken 2 weeks later, and healing was monitored 3–5 days later (visit 4). After the 31 days of sexual abstinence, participants came to the clinic for a final exam of the biopsy sites and a final follow‐up (visit 5).

Figure 2

PSA‐semiquant cassette test serial dilutions to indicate PSA concentrations. Semens were diluted in PBS. A total of 120 ul of the semen dilutions were loaded on the immunochromatographic test strip cassette (Seratec PSA Semiquant, Germany). Negative test results were defined by the presence of bands in the control (C) and internal standard columns, but not in the test result column (T). A positive test showed bands in the control (C), the internal standard, and the test result columns (T). PSA concentration >1 ng/mL were detectable by the kit, which corresponded to a 10⁻⁶ semen dilution as tested by us.

Figure 3

Concentration of pro‐inflammatory mediators in HIV‐uninfected and HIV‐infected FSW before and after biopsy sampling. Cervical concentrations of IL‐8, MIP1β, MIP3α, MIP1α, MCP‐1, MIG, IP‐10, IL‐1α, IL‐1β, and TNF‐α were measured by Milliplex (Millipore, Merck KGaA) in the cervicovaginal lavage collected at visit 1 (sampling), visit 3 (sampling), and visit 5 (final follow‐up). The difference in the cervical concentrations of inflammatory mediators at different visits was calculated using the nonparametric Friedman test for repeated measures. When significant, posthoc Dunn's multiple comparisons test was applied to compare between visits.

Figure 4

Proportion of cervical HIV T‐cell targets (CD4⁺CCR5⁺ T cells) in HIV‐uninfected and HIV‐infected FSWs before and after biopsy sampling. Cervical mononuclear cells were collected by endocervical cytobrush and ectocervical spatula scraping in PBS at visit 1 (sampling), visit 3 (sampling), and visit 5 (final follow‐up). Cells were stained for live cell discrimination and identification of HIV T‐cell targets (CD3, CD4, CCR5) by flow cytometry. The difference in the proportion of HIV T‐cell targets at different visits was calculated using the ordinary two‐way anova and the Sidak's multiple comparisons test.