Viral load in vaginal fluid

There have been fewer studies in women, compared to men, examining the association between viral loads in the blood plasma with that in the genital tract. However, most studies that have looked at this suggest that viral load in blood and vaginal secretions is usually strongly correlated, and that it is rare for a woman who has an undetectable blood plasma viral load to have detectable HIV in vaginal secretions.

A 2001 article in The Lancet¹ looked at 271 HIV-positive women in the US Women’s Interagency HIV Study (WIHS) cohort study. Three-quarters of the women were on HAART and half of them on PI-based regimens. Both blood plasma and lymphocyte RNA measurements were performed on the majority. Detectable viral RNA was found in the vaginal secretions of 130 out of 163 women (80%) with a blood plasma viral load of over 500 copies/ml, and in 27 out of 83 women (33%) with a blood plasma viral load under that figure. However, the researchers concluded that only one in ten of all the women with detectable viral RNA in vaginal secretions had infectious virus.

A study in Italy² divided 122 women into four categories according to their experience of antiretroviral therapy: 55 women were naive to therapy; 21 women were taking a break from antiretroviral treatment due to treatment failure; 19 women were on treatment with one or two NRTIs alone; and 28 were on highly active antiretroviral therapy (HAART).

This study found that almost three-quarters of the women on non-HAART regimens had detectable plasma viral load (73%) and detectable (78%) cervicovaginal viral load, 50% of women taking HAART had a detectable plasma viral load, and 40% on HAART had detectable HIV in cervicovaginal secretions.

Looking only at the women who were on HAART or who were on mono or dual therapy, 50% of women on HAART and 63% on mono or dual therapy had undetectable viral loads (less than 80 copies/ml) in both blood and vaginal secretions, and 29% of women on HAART and 11% on mono or dual therapy had detectable viral loads in both. This left eleven women who were on some kind of HIV therapy that had discordant viral loads. Twenty-one per cent of women on HAART and 10% of women on mono or dual therapy had detectable HIV in their blood, but none in their vaginal secretions: conversely 16% of women on mono or dual therapy, but none on HAART, had detectable HIV in their vaginal secretions, but none in their blood.

HAART was thus significantly associated with having a higher viral load in blood than vaginal secretions, whereas suboptimal mono/dual therapy was significantly associated with the opposite situation. The explanation for this, and for low undetectability rates in general, was probably that here was a group of women with quite advanced HIV disease (35% had an AIDS diagnosis) and possible adherence issues (37% were injecting drug users) who were taking therapy in the early days of HAART (1995-2001) before the widespread use of boosted protease inhibitors.

Even in this situation, we note that no woman on HAART with an undetectable viral load in blood had detectable virus in vaginal secretions. This was not the case, however, in women on mono/dual therapy, or on no therapy. In the latter case, amongst the 75 women taking no HIV drugs, one in 15 (6.7%) of women had no detectable viral load at all, one in twelve (8%) had detectable HIV in blood but not in vaginal secretions, and just over one in 20 (5.3%) had detectable HIV in vaginal secretions but not in blood.

This has led researchers to tentatively conclude that while it is probably rare for  women with an undetectable blood plasma viral load on HAART to have detectable vaginal HIV, the same may not be the case for women who are undetectable off HAART.

Similarly, a later US longitudinal study in 97 women³ found that all of the women with an undetectable viral load in the blood, and who were on HAART, also had undetectable cervicovaginal viral loads. In women not on HAART, although some did achieve undetectable plasma viral loads, some still had evidence of cervicovaginal shedding: 5% of the time there were greater levels of virus in the vagina than in the blood. When women achieved undetectable virus levels in both the blood and vagina, rebound of virus occurred in the blood first or at about the same time.

Successful antiretroviral therapy, however, cannot guarantee that there will be no cervicovaginal shedding. A four-week study of 20 sex workers in Mombasa, Kenya,⁴ found that seven women (35%) continued to have detectable viral load in their genital secretions at the end of the study. Similarly, although antiretroviral treatment significantly reduced the frequency of genital shedding of HIV in a study conducted in Burkina Faso,⁵ HIV remained detectable in the genital tract of a significant proportion of women even when they had an undetectable viral load in their blood. However, all the women in the study were infected with the genital herpes virus HSV-2, and it is known that this can increase genital shedding of HIV.

Other studies have found that viral load in the female genital tract can vary during the course of a menstrual cycle, even among women on antiretroviral treatment. A 2004 study⁶ of viral load changes during the menstrual cycle found that viral load levels in vaginal fluid tended to peak at the time of menstruation and fell to the lowest level just prior to ovulation.