New research points the way for future treatment of women with HIV

Polly Clayden, HIV i-base

The 9th Conference on Retroviruses and Opportunistic Infections appeared to devote more space to women-specific issues than any previous gathering of this meeting.

A whole poster session was focused on the ‘Effect of gender/sex on viral load, pharmacokinetics and response to antiretroviral therapy’, which unfortunately delivered little new, and overall there was little reported with immediate practical, clinical utility. There was however some interesting research that may have important implications for future practical application. (We covered pregnancy and mother to child transmission in HTB 3,4).

Genital tract and HIV viral load

Several studies looked at the presence of HIV RNA in the female genital tract.

A better understanding of these levels, their evaluation and associated factors will, it is hoped, lead to a better understanding of both mother to child and sexual transmission of HIV.

A poster from Dr Cu Uvin and colleagues from Brown and Emory Universities entitled ‘Factors associated with HIV RNA shedding in the female genital tract’ looked at the association between cervicovaginal lavage (CVL) HIV-1 RNA and antiretroviral therapy. [1] This study evaluated plasma viral load, lower genital tract infections and the presence of semen among 97 HIV positive women over a 24-month period.

The women studied were stratified into three groups: A) antiretroviral naïve starting HAART (n=36), B) nucleoside experienced starting HAART (n=26) and C) no past or present antiretroviral therapy. The investigators found a highly significant association between plasma viral load and CVL HIV-1 RNA detection (p=<0.0001). They reported that for each log10 increase in plasma viral load there was a 4-fold increase in detecting CVL viral load. Neither antiretroviral use nor CD4 cell count made any significant difference in HIV-1 RNA in the female genital tract in this study.

In an oral presentation, Dr C Critchlow from the University of Washington reported findings from a study in which it was sought to evaluate whether lower rates of transmission of HIV-2 compared to HIV-1 could be explained by differences in levels of HIV in the genital tract at a similar disease stage [2].

Two hundred and twenty-one women were recruited from the University of Dakar Infectious Disease Clinic in Senegal, and blood and CVL samples were taken to determine HIV RNA levels in each. Out of this group, the women with HIV-1 (n=170) tended to be younger compared to women with HIV-2 (n=51) — 31 and 36 years respectively, and they also tended to present with lower CD4 counts (301 vs. 352 cells/mL) and more advanced disease.

Of the women in which HIV was detected in the genital tract they reported that >1000 HIV RNA copies/mL were more frequently found in women with HIV-1 (46%) than with HIV-2 (0%). They also found that HIV-1, lower CD4 count, higher plasma HIV RNA levels and vaginal pH were associated with higher vaginal HIV RNA levels. The investigators concluded that “Increased rates of vaginal shedding and most likely of transmission seen among women with HIV-1appears to be attributable to higher viral burden at a given CD4 count among women with HIV-1.”

A poster reporting a small longitudinal substudy of women with HIV-1 (n=11) and HIV-2 (n=8), from the same group found lower CD4 count was associated with the increased likelihood of detecting the presence of HIV RNA in the genital tract as was recent menes and amenorrhea [3]. The investigators also observed that over this short term (6-weeks) almost half (47%) of the HIV-1 and HIV-2 positive women studied had intermittent detectable levels of HIV in the genital tract. They suggest therefore that studies evaluating factors contributing to HIV transmission relying on a single sample may not reflect vaginal shedding over an extended period of time.

A late breaker presentation by Dr S Benki from the Fred Hutchinson Cancer Research Centre reported findings from a study in which 17 women were followed intensively over a period of one month to investigate the “Relationship between the menstrual cycle and the daily pattern of HIV-1 RNA shedding in the genital tract of HIV-1 positive women” [4].

Throughout one menstrual cycle endocervical swabs and luteinising hormone (LH) measurements were taken daily, HIV-1 RNA was then determined in the cervical swab. HIV-1 RNA was detected in 89% of the samples 402/451. The investigators reported an increase of 0.05 log10 copies per swab as the number of days from the LH surge increased (p=0.004).

This relationship between the menstrual cycle and HIV-1 RNA levels has been analysed previously in several studies but these have generated contradictory data. From this detailed analysis the authors summarise that their findings suggest that, ‘… in cervical mucosa, viral load reaches a nadir during the peri-ovulatory period, and then subsequently increases during the post-ovulatory phase.’

However, methods of collection of female genital secretions vary – the most frequently utilised are: ‘sno-strips’, aspiration, cytobrush and swabs. These differ in the subcompartment that they sample and different methodology may in turn result in significant variation in HIV RNA reported in the female genital tract.

Another poster from Dr Cu-Uvin ‘s group compared HIV-1 RNA levels using CVL and ‘sno-strip’ collection [5]. The investigators found that the sno-strip collection method detected more HIV-1 RNA in genital secretions in women with detectable plasma viral load than CVL. They suggest that this could be caused by the dilution effect of CVL or the different subcompartments being sampled.

Genital tract and other infections

A small number of studies examined the presence of other viral infections in the female genital tract of HIV-positive women.

HIV-positive women are at an increased risk of HPV (human papillomavirus) associated cervical cancer. In an oral presentation, Dr L Conley from CDC Atlanta reported findings from a study conducted to assess the effect of HAART on HPV viral load in vaginal secretions [6].

Plasma HIV-1 RNA and CVL HPV DNA samples were obtained from 44 HIV-positive women over 345 visits – the changes in HPV loads were evaluated from initiation of HAART or from one visit to the next for women on no HAART (n=18), over a six month period.

In this study no significant effect of HAART was observed on HPV load in the female genital tract even among women using HAART who were below the level of detection. The investigators concluded that HPV and in turn the high incidence of cervical disease may persist in HIV-positive women even for those using HAART. They also noted that their findings may be limited by the small number of women in this study.

HIV/HCV coinfection is common among HIV-positive women especially those who are or have been injecting drug users. As with HIV, assessing for HCV presence in the female genital tract could lead to better understanding of sexual and mother to child transmission of this infection.

A poster from Dr M Nowicki and colleagues from the Women Interagency HIV Study (WIHS) reported a comparison of plasma and CVL HCV RNA in 9 HIV/HCV coinfected women [7]. HCV RNA was detected in plasma and CVL in 5/9 of the women evaluated. From four of these paired samples viral sequences from peripheral blood mononuclear cells (PBMC) were analysed. The single strand conformation polymorphism (SSCP) patterns for plasma and PBMC HCV sequences were identical in three women. In the other woman the PBMC-derived sequence pattern was identical to that from CVL, which the investigators believed suggested common origin.

It was concluded that these data suggest the presence of a separate HCV compartment in HIV-1/HCV co-infected women. These findings are interesting, but again it is hard to draw any conclusions from such a tiny study. The role this compartment plays in sexual and mother to child transmission warrants further investigation.

In addition the cytomegalovirus (CMV) was examined in female genital secretions. The role of CMV in sexual transmission of HIV-1 amongst HIV/CMV coinfected men, the subsequent disease progression of HIV, and the association with an increased risk of AIDS for those infected with multiple strains has been the focus of several studies. But less is known about HIV/CMV coinfection in women.

A poster from Dr N Lurain and colleagues reported results from a pilot study to detect and characterise CMV strains in a group of 45 women [8]. Plasma, CVL and oral samples for which the HIV-1 viral loads had already been established were analysed for the presence of CMV DNA. CMV was only detected in CVL samples from women with >2000 copies/mL HIV RNA. Infection with more than one strain was detected in samples from two women. The investigators suggested that after these preliminary results, future studies should focus on the genital tract as a site for interaction and transmission of both viruses.

Treatment for cervical disease

Women with HIV are at increased risk for squamous intraepithelial lesions (SIL) ie pre-invasive cervical cancer, as well as invasive cervical cancer. About 50% of HIV-positive women will develop SIL, but reports seem to demonstrate that its treatment is not as successful compared to that of HIV negative women.

A late breaker presented by Dr T Wright from Columbia University compared standard management approaches for SIL in a group of 122 HIV-positive and 257 HIV-negative women in a multi-arm trial conducted during 1994-2000 [9].

Women with satisfactory colposcopic examinations and low grade SIL were randomised to either observation or cryotherapy and women in later stage with high grade SIL were randomised to either cryotherapy or a loop electrosurgical excision procedure (LEEP). Any woman with an unsatisfactory examination (defined as an examination that failed to identify the extent of the lesion – usually because it extended to the cervical canal) had either a cold knife or LEEP diagnostic conization.

The women were interviewed and underwent an examination that included PAP smear, cervicovaginal lavage, colposcopy and a biopsy if indicated. These took place at entry and at four, eight and 12 months.

The investigators reported that all standard treatments for SIL were less effective in women with HIV compared to HIV-negative women. Among women with low-grade SIL randomised to observation alone, HIV-positive women were found to have a significantly lower rate of regression to normal compared to HIV-negative – 24% and 61% respectively (p<0.001) during the follow up period but the rate of progression (to biopsy confirmed hi-grade SIL) among both groups was low – 4% vs 9%. No treatment appeared to be more effective than others for HIV-positive women.

Although findings from this trial have limitations in terms of current HIV care — because when the study was conducted there were relatively few women using HAART — the investigators recommend that because of the high failure rate of treatment and the low rate of progression to high-grade SIL, “observation should be considered in managing biopsy confirmed low-grade SIL in HIV-positive women.”

Oral contraceptive use and viral diversity

The majority of HIV-positive women are infected with multiple HIV-1 viral variants early in their infection. It is unknown what factors predict whether multiple variants will be transmitted and what impact this will have on disease progression.

In an oral presentation entitled ‘Correlates of viral diversity in primary HIV-1 infection in women’ several factors were outlined that could affect viral diversity near the time of seroconversion [10]. Presenting author Dr M Sagar’s group had previously shown that the initial HIV-1 proviral population from 60% of women tested is heterogeneous at or around the time of seroconversion.

They found that neither demographic nor behavioural factors had any effect on the complexity of the initial virus population. In addition they found no association with other sexually transmitted infections. They did however find a strong association with initial viral diversity with the use of oral contraceptive pills (p<0.006), Depo medroxyprogesterone (p<0.004) or any form of hormonal contraception (p<0.001) compared to women with an initial homogenous virus population.

They also reported that the median of the first viral load test – taken between four and 12 months post infection – was higher among women with initial viral heterogeneity 100,000 vs 45,709 copies/mL and this difference persisted in results five years post infection. The CD4 counts were also lower – median 394 vs 460. The investigators concluded that: “Hormonal contraception use near the time of infection was associated with acquisition of a genetically complex virus population. An initial heterogeneous virus population is associated with faster disease progression as measured by HIV-1 plasma viral load and CD4 count.”

Bone mineral density and older women

Finally a very welcome focus of research was presented as a poster from Dr Brown and colleagues from Montefiore medical centre [11]. Osteopenia has been recently identified as a side effect for HIV-positive people that may be associated with antiretroviral use. Decreased bone mineral density is also known to occur in menopause as part of the normal aging process of women.

This study analysed the bone mineral density (BMD) in 40 women (19 HIV-positive, 21 HIV-negative) of a median age of 48 years, 50% were defined as post-menopausal and 50% as perimenopausal. Among the HIV-positive women all had used antiretrovirals and 55% has used protease inhibitors. The investigators reported prevalence of decreased BMD to be 50% among HIV-positive postmenopausal women, 44% among HIV-negative postmenopausal women, 25% among HIV-positive perimenopausal women and 40% among HIV-negative perimenopausal women. Of the group of HIV-positive women 60% of PI using women had osteopenia compared to women who had not used PIs.

Although the investigators concluded that “protease inhibitor use is associated with significantly decreased bone mineral density in older women” this association did not reach statistical significance (p=0.04), and there was no mention of other possible cofactors.

Very little in the way of conclusion can be drawn from such a tiny study, and the mechanisms of decreased bone mineral density are not well understood in the HIV-positive, HAART-using population overall. However this study highlights a population that could be at extra risk for this adverse event, and as the average age of the HAART-using population increases it would seem that particularly close observation and monitoring of older women for osteopenia is warranted.


All from the Programme and abstracts on retroviruses and opportunistic infections, February 24-28, 2002, Seattle Washington.

  1. Cu-Uvin S, Caliendo A, Snyder B et al. Factors associated with HIV-1 RNA shedding in the female genital tract: 24 month follow up. Abs.784-w
  2. Critchlow C, Hawes S, Redman M et al. Detection of human immunodeficiency virus (HIV) type 1 and type 2 RNA and DNA in vaginal secretions among women in Senegal, West Africa. Abs.19
  3. Critchlow C, Hawes S, Redman M et al. A longtitudinal study of the detection of human immunodeficiency virus (HIV) type 1 and type 2 RNA in vaginal secretions among Senegalese women. Abs.785-w
  4. Benki S, Mostad SB, Richardson BA et al. Relationship between the menstrual cycle and the daily pattern of HIV-1 RNA shedding in the genital tract of HIV-1 infected women. Abs. LB2
  5. Cu-Uvin S, Snyder B, Donahue S et al. HIV-1 RNA levels in the female genital tract differ by collection method: analysis of cervicovaginal lavage and sno-strip collections. Abs.782-w
  6. Conley LJ, Bush TJ, Ellerbrock TV, et al. Lack of effect of HAART on HPV DNA levels in the female genital tract. Abs.119
  7. Nowicki M, Laskus T, Radkowski M et al. Presence of HCV RNA in cervicovaginal lavages obtained from HIV-1/HCV co-infected women. Abs. 648-m
  8. Lurain N, Robert E, Camarca M et al. Detection and characteristics of CMV in vaginal-cervical secretions of HVI-infected women. Abs 786-w
  9. Wright TC, Bush TJ, Sawo D et al. A clinical trial comparing standard treatment modalities for squamous intraepithelial lesions of the cervix in HIV-infected women Abs. LB 16
  10. Sagar M, Lavreys L, Baeten J et al. Correlates of viral diversity in primary HIV-1 infection in women. Abs 100
  11. Brown T, Timpone J, Ruppe M et al. Bone loss associated with abnormalities in glucose metabolism in HIV patients on protease inhibitors. Abs 716-T

Links to other websites are current at date of posting but not maintained.