HIV-positive women being treated with a non-nucleoside reverse transcriptase inhibitor (NNRTI)-based drug regimen are more likely to have detectable amounts of HIV in their genital tracts compared to those being treated with a protease inhibitor (PI)-based regimen. These new data, published in the January 1 issue of the Journal of Acquired Immune Deficiency Syndromes by the U.S. Women’s Interagency HIV Study (WIHS), may explain why some HIV-positive women with undetectable viral loads may still face a small risk of transmitting HIV to their sexual partners and babies.

Previous studies have demonstrated an association between blood viral loads and viral “shedding” in the genital tract. In other words, HIV-positive women who have undetectable viral loads while receiving antiretroviral therapy are less likely to have detectable HIV in their genital tracts. As a result, they are significantly less likely to transmit the virus to their sexual partners or babies during pregnancy and delivery.

The association between blood viral load and genital HIV shedding is not guaranteed, however. In a previous study conducted by WIHS, 27 (33%) of 83 women with undetectable or low blood viral loads (less than 500) had detectable cervical HIV secretions, indicating a potential increased risk of viral transmission to their sexual partners and babies.

To identify factors associated with cervical HIV shedding in women with low and undetectable viral loads, WIHS researchers recently conducted a study involving a larger cohort of women. Based on research suggesting that HIV drugs are not equal with respect to their ability to penetrate genital tissues, the researchers hypothesized that the risk of cervical HIV shedding in women with undetectable viral loads might be tied to the use of either NNRTI-based or PI-based regimens.

The study involved measuring HIV-RNA levels in cervical swab specimens obtained from 290 female WIHS participants who were on antiretroviral treatment and had viral loads below 500.

Overall, 44 (15%) of the women had detectable HIV-RNA in their cervical specimens: 23 (13%) of the 179 PI users and 21 (19%) of the 111 NNRTI users. Among the 44 women with evidence of cervical HIV shedding, six (14%) were considered “high shedders,” with cervical HIV levels in excess of 1,000.

Shedders were more likely to have had more than one male sex partner in the previous six months, to have used crack cocaine or injection drugs, to have smoked tobacco, and to have blood viral loads between 50 and 500 at the time of their cervical swab collection.

Shedders were also more likely to have a history of syphilis – a known HIV transmission factor. However, there was evidence of sexually transmitted infection (STI)-related genital inflammation in both the shedders and non-shedders, suggesting little association between STI status and the risk of cervical HIV shedding.

In the analysis of the data, shedding was independently associated with NNRTI use (compared to PI use) and illicit drug use.

Other studies have concluded that the penetration of antiretrovirals into genital secretions is dependent on the specific drug used. For example, Viramune® (nevirapine), Retrovir® (zidovudine), Epivir® (lamivudine), and Emtriva® (emtricitabine) have demonstrated greater penetration than Sustiva® (efavirenz) or any of the protease inhibitors.

While the WIHS study was not designed to detect differences in cervical shedding related to specific drugs, it did find increased shedding among women treated with Sustiva. This, the researchers noted in their report, is consistent with published data.

In contrast, the study team’s finding of decreased cervical HIV shedding in women treated with PIs was unexpected, given the decreased penetration of PIs into female secretions. This paradox, the researchers added, will be explored by WIHS in future studies.

“Shedding in this population was common,” wrote the study team, “and NNRTI-based highly active antiretroviral therapy (HAART) (vs. PI-based HAART) was associated with genital HIV shedding. Further study is required to determine the impact of these findings on transmission of HIV from mother to child or to sexual partners.”