Women experience rapid fall in genital secretion of HIV after starting antiretovirals, but 50% still potentially infectious

Genital shedding of HIV decreases rapidly in women in the month after starting antiretroviral therapy, according to a prospective study conducted amongst sex workers in Kenya and published in the February 19^th edition of AIDS. However, the team of American and Kenyan investigators who conducted the study found that 50% of women continued to have either HIV-infected cells or detectable levels of HIV in their cervical and vaginal secretions, indicating an ongoing risk of HIV transmission.

Potent anti-HIV therapy has brought a longer, healthier life for hundreds of thousands of individuals worldwide. It is also hoped that expanding use of antiretroviral therapy will reduce sexual transmission of HIV, particularly in areas of high prevalence. Although there is good evidence that the use of anti-HIV drugs can significantly reduce the risk of mother-to-child transmission of HIV, there are limited data regarding the impact of antiretrovirals on sexual transmission of HIV, and the exciting data come from cross-sectional studies.

Investigators therefore designed a prospective study involving women commencing antiretroviral therapy. They hypothesised that initiating anti-HIV therapy would lead to a rapid fall in HIV viral load in genital secretions during the first month of therapy, but that the speed and magnitude of this fall might differ from that seen in plasma.

The study involved 20 sex workers in Mombasa, Kenya. All had a CD4 cell count below 200 cells/mm³ and were provided with an antiretroviral regimen consisting of 3TC (lamivudine, Epivir), d4T (stavudine, Zerit) and nevirapine (Viramune). Before entry to the study at regular intervals during follow-up, the women were screened for sexually transmitted diseases. Cervical and vaginal samples were obtained to monitor levels of HIV secretion at baseline and then at days 1, 2, 4, 7, 14, 21 and 28. The investigators also modelled the decline of viral load in plasma during the first month of HIV therapy and compared it to that observed in vaginal secretions. In addition, the investigators monitored the number of HIV-infected cells in cervical secretions.

All the women were highly adherent to their anti-HIV therapy and none had gonorrhoea, chlamydia or non-specific cervicitis at baseline, although eight did have bacterial vaginosis, three a yeast infection, and one trichomoniasis (for which appropriate therapy was provided). A genital ulcer was also observed in one woman on entry to the study, and in two others during follow-up.

HIV was detectable in both the cervical and vaginal secretions of all the women at baseline. However, two days after starting HIV therapy cervical viral load had fallen from a median of 5,000 copies/ml to 650 copies/ml, and then fell to a median of below 50 copies/ml where it remained for the duration of the study. Median vaginal viral load was 6,000 copies/ml at baseline and fell to a median of 250 copies/ml by day 4 and by day 14 median vaginal shedding of HIV was undetectable.

Both cervical and vaginal viral load were significantly lower than seen in plasma, the median viral load in blood at baseline being over 300,000 copies/ml, and 1,300 copies/ml at day 28.

However, the investigators noted that seven (35%) women continued to have detectable viral load in their genital secretions at day 28, indicating an ongoing risk of HIV transmission.

A model constructed by the investigators showed that viral load in vaginal secretions fell significantly faster than in blood (p = 0.02).

HIV-infected cells were detected in the cervical secretions of 53% of women at baseline and 26% by day 28. This decline was not statistically significant. However, a statistically significant fall was seen in the number of women with HIV-infected cells in their vaginal secretions from 47% at baseline to 5% at day 28 (p

“This is the first prospective cohort study to offer detailed quantitative assessment of the time course and magnitude of cervical and vaginal HIV-1 suppression among women initiating antiretroviral therapy”, write the investigators, adding, “the study adds substantially to previous cross-sectional analyses, which demonstrated an association between antiretroviral therapy and lower genital levels but did not assess the temporal impact of antiretroviral therapy on this outcome”.

They go on to comment, “steep declines in genital viral levels are likely to represent a rapid decline in infectivity. However, suppression of genital tract HIV-1 shedding was not complete, which may indicate an ongoing risk of transmission to sexual partners.”