Women, HIV research and antiretrovirals

4th HIV and women workshop logo2

Polly Clayden, HIV i-Base

Several presentations at the 4th International Workshop on HIV and Women again highlighted the relative paucity of data to guide treatment decisions in women – particularly with newer drugs.

The inclusion (or exclusion) of women in HIV research

A collaboration between IAS, IAVI and AMFAR reviewing the inclusion of women in HIV research – trials of antiretrovirals, vaccines and cure strategies – exposed the unsurprising finding that women are underrepresented, particularly in antiretroviral trials and cure strategies. [1]

Notably publically funded antiretroviral clinical trials (including US National Institute of Health [NIH] sponsored trials) only included small proportions of women despite existing regulation intended to correct this.

Although the proportion of women in antiretroviral trials remains low, there has been an increase over time. The proportion of women is particularly low in trials conducted in high-income countries. Vaccine trials do better and include a higher proportion of women.

Shirin Heidari presented results from this literature review for which the investigators performed systematic searches in PubMed for antiretroviral, vaccine and cure trials. Antiretroviral included articles describing trials published during three time periods (1994-1997, 2001-2004 and 2008-2011). Vaccine included articles published 2000-2012 that reported results from vaccine trials. Cure included articles describing cure trials published through 2012.

The review excluded trials that only enrolled one sex. The investigators extracted data describing the number of women compared to the total number of participants (enrolled, completed the trial and/or reached an endpoint), date of publication, trial phase, countries in which the trial was conducted and funding sources.

The analysis included 387 antiretroviral, 53 vaccine and 113 cure trials. Women participants made up a median of 19.2%, 38.2% and 11.1% of the total study population in antiretroviral, vaccine and cure trials respectively.

The proportion of women included in antiretroviral trials increased over time, overall p=0.0001. But this was no greater than 28% in any time period.

Antiretroviral trials conducted in high-income countries included the least women, median percentages (excluding eight trials without country classification) were: 50%, 18% and 23.2% in low- and middle-income, high-income, and mixed income countries respectively, p<0.001.

There was a significant variation in the proportion of women in antiretroviral trials according to funding source. Median percentages were: 19%, 29.2%, 16.7%, 19.8% and 17.8% for private (commercial), private (non-commercial), public mixed and trials with no data respectively, p=0.05 (p=0.03 excluding “no data”).

NIH supported antiretroviral trials had a lower proportion of women compared to those sponsored by other sources, 15.3% (n=96) vs 22.3% (n=220), p=001.

The inclusion of women in vaccine trials also increased over time, p=0.03. No linear relationship was observed between the inclusion of women and time for cure trials. High-income countries were also associated with a lower proportion of women in cure trials, p=0.003, but a higher proportion in vaccine trials, p=0.02. Funding source did not have an effect on proportion of women in vaccine and cure trials.

The investigators noted that although federal policies have been established to address the gap between the proportion of men and women in trials, their analysis found that publically funded antiretroviral trials have even lower representation of women participants, suggesting that these policies are neither enforced nor monitored.

Sharon Warmsley illustrated the disparity between men and women in an invited lecture: State-of-the-ART new therapy options by showing an analysis of the proportion of women in pivotal clinical trials for more recently approved antiretrovirals. See Table 1.

Table 1: Proportion of women included in pivotal trials
Trial New drug Comparator % women
STARTMRK raltegravir efavirenz 19%
Single dolutegravir efavirenz 16%
Spring-2 dolutegravir raltegravir 15%
Flamingo dolutegravir darunavir/r 13%
Gilead 102 elvitegravir efavirenz 22%
Gilead 103 elvitegravir atazanavir/r 8%
ECHO rilpivirine efavirenz 23%
Thrive rilpivirine efavirenz 26%
STaR rilpivirine efavirenz 7%

Two presentations followed Dr Walmsley’s lecture with data from subgroup analyses women receiving rilpivirine/emtricitabine/tenofovir DF (RPV/FTC/TDF), and raltegravir (RAL) in the respective pivotal trials. For both analyses the numbers were so few that the confidence intervals around any findings are so wide that there is not much to guide treatment decisions for women from these trials.

Dr Walmsley also showed two examples of ongoing phase 3b trials designed to look at newer treatments in women and help to address this lack of meaningful data. See Table 2.

Table 2: Ongoing phase 3b trials of antiretrovirals in women
Trial Drug/regimen Comparator Sponsor Design/status Primary endpoint n
WAVES EVG/COBI/FTC/TDF ATV/RTV+FTC/TDF GileadSciences RCT, 1:1, blinded, placebo, trial in ART-naïve womenNCT01705574Enroling <50 copies/mL at 48 weeks 510255 per arm
ARIA DTG/ABC/3TC ATV/RTV+FTC/TDF ViiVHealthcare Randomised 1:1, open label trial in ART-naïve womenNCT01910402Enroling <50 copies/mL at 48 weeks 474237 per arm

Single-tablet regimen rilpivirine/emtricitabine/tenofovir DF

STaR compares the safety and efficacy of two once daily fixed dose combination (FDC) regimens: RPV/FTC/TDF vs efavirenz/emtricitabine/tenofovir DF (EFV/FTC/TDF).

It is an open-label, multicentre, randomised 1:1, 96-week study in treatment-naive participants. The primary endpoint was the proportion of participants with viral load <50 copies/mL at 48-weeks (12% non-inferiority margin; snapshot analysis).

Overall RPV/FTC/TDF (n=394) was non-inferior to EFV/FTC/TDF (n=392): 86% vs 82%, difference 4.1% (95% CI -1.1%, to 9.2%) at week 48; and 78% vs 72%, difference 5.5% (95% CI -0.6%, 11.5%) at week 96.

Beth Elbert presented the subgroup analysis of women in STaR. As the proportion of women in the trial was only 7% this represented 28 women in each arm at 48 weeks and 7 and 8 women had no 96-week data in the RPV and EFV arms respectively.

The proportion of women with viral load <50 copies/mL at 48 weeks was 79% vs 61% in the EFV/FTC/TDF vs RPV/FTC/TDF arms, difference 17.4% (95% CI -8.8% to 43.6%). At 96 weeks, the proportion was 68% vs 57%, difference 12% (95% CI-15.5% to 39.5%).

The rate of all grades treatment-associated adverse events of importance in >5% of women in the RPV/FTC/TDF and EFV/FTC/TDF arms respectively were: dizziness 3 (11%) vs 8 (29%), somnolence 1 (4%) vs 4 (14%), headache 3 (11%) vs 3 (11%), anxiety 2 (7%) vs 0, abnormal dreams 1 (4%) vs 2 (7%), insomnia 2 (7%) vs 5 (18%), depression 0 vs 2 (7%) and rash 1(4%) vs 8, (29%).

There were 2 adverse event-related discontinuations (1 cerebrovascular accident and 1 gout) in the RPV/FTC/TDF arm and 3 (1 depression, 1 pyrexia and 1 toxic skin eruption) in the EFV/FTC/TDF arm. All occurred before 48 weeks.

There were 5 discontinuations due to an adverse event among the women.


STARTMRK and QDMRK were double blind, randomised, controlled phase 3 trials of RAL in treatment-naive participants.

STARTMRK compared RAL 400mg (twice daily) to EFV (once daily) both given with TDF/FTC for up to 5 years. QDMARK compared RAL 400mg (twice daily) to RAL 800 mg (once daily), both with TDF/FTC for up to 48-weeks.

Kate Squires presented data at 48-weeks from a post hoc, pooled exploratory subgroup analysis by sex in the RAL 400 mg arms of these studies.

The proportion of participants with viral load <50 copies/mL and mean change from baseline in CD4 counts were summarised. Observed failure approach was used for missing data. Cinical and laboratory adverse events and changes in laboratory parameters were recorded.

Of 669 participants who received raltegravir 400 mg, 525 (78%) were men (mean age 38 years) and 144 (22%) were women (mean age 38.5 years).

There were a smaller proportion of white participants (37% vs 64%) and a greater proportion of black participants (29% vs 9%) and Asian participants (20% vs 8%) among women compared to men. Baseline viral load was >100,000 copies/mL in 40% of women vs 47% of men, and CD4 count was <200 cells/mm3 in 33% of women vs 39% of men. Overall, 89% of women and 91% of men completed 48 weeks of treatment.

Other than pregnancy, which occurred in 4 (3%) women, most reasons for discontinuation were similar between cohorts. At week 48, 93% of women (126/135) and 91% of men (458/505) had viral load <50 copies/mL, difference: -3.0, (95% CI (-7.4 to 3.0). The mean change in CD4 cell count from baseline to week 48 was 189 cells/mm3 in women and 194 cells/mm3 in men, difference: 5.3 (95% CI -21 to 31).

Adverse events were reported by 85% of women and 88% and led to treatment discontinuation in 2% of each cohort. Serious adverse events were less common in women (2%) than in men (9%). Laboratory adverse events occurred in 7% of women vs 9% of men and were considered drug-related in 3% of each cohort. Changes in laboratory parameters were similar between cohorts: grade 2/3 increase in LDL-cholesterol in 5% of women vs 6% of men; grade 2/3 increase in total cholesterol, 6% vs 5%; grade 2/3 increase in serum triglycerides, 0% vs 1%.

Gender disparities in treatment outcomes persist

Although data from women in trials frequently suggest that efficacy and safety of antiretrovirals and the occurrence of adverse events is similar in men and women, a presentation by Peter Saunders from the Royal Free Hospital, London reminded us that gender disparities in treatment outcomes persist in the era of modern antiretroviral therapy.

This study looked at all antiretroviral-naive HIV positive people who attended the Royal Free clinic and started ART from 1 January 2006 onwards. To be eligible for analysis they needed to have at least one documented viral load test after starting ART.

The proportion experiencing virological failure (1 of 2 consecutive viral loads >200 copies/mL >6 months post ART) and treatment modification were estimated using standard survival methods.

Of 1131 overall, 29% (327) were women, 58% (563) were men who have sex with men (MSM) and 19% (241) non-MSM men. Women and non-MSM men started ART at a more advanced stage, with a median CD4 at ART initiation of 219 and 218 cells/mm3 respectively compared to 298 in MSM. Women (60%) and non-MSM men (44%) were also more likely to be of black African ethnicity and to have a previous AIDS diagnosis.

Time to achieve virological suppression (viral load<50 copies/ml) was similar in all groups with 88.8% of MSM, 83.4% of non-MSM men and 84.7% of women achieving this by one year, p=0.19.

After 18 months of treatment a greater proportion of women had experienced virological failure: 2.6% MSM, 6.2% non-MSM men and 9.8%, p<0.0001. Non-MSM men had more than three times the rate of virological failure, AHR 3.69 (95% CI 1.76 to 7.74) and women more than 4 times the rate, AHR 4.63 (2.26 to 9.48), compared to MSM, p=0.0001.

After 12 months 42.6% of women had changed a component of their treatment regimen compared to 35.5% of non-MSM men and 26.9% of MSM, p<0.0001; 16.6% women changed their regimen due to adverse events compared 9.6% of MSM and 12.09% non-MSM men.

By 12 months, 5.0% of MSM, 12.0% of non-MSM men and 15.4% of women had completely discontinued ART for at least two weeks, p<0.0001.

The investigators found that non-MSM men had over twice, AHR 2.28 (95% CI 1.35 to 3.83) and women had more than three times the rate, AHR 3.45 (95% CI 2.20 to 5.40) of complete ART discontinuation compared to MSM, p<0.0001.

Dr Saunders concluded that women in this cohort are still more likely than both MSM men and non-MSM men to change or discontinue their ART regimen and also to experience virological failure. Further research is urgently needed to address the reasons for these differences.


Underrepresention of women in HIV clinical trials is no big surprise and much of the discussion following the presentations focused on the difficulties of enrollment. Strategies to increase this must continue to be improved.

Generally, data suggests that there are few differences in ARV efficacy between women and men. But because women often present very late, there can be implications for drugs that are not recommended at high viral loads such as rilpivirine. Any differences in toxicities by sex, tend to emerge in post marketing studies and will only be evaluated in a more systematic way if research is designed to look at how drugs perform in women – such as the ARIA and WAVES studies.

Disparities in treatment outcomes in real-life situations persist and need to be addressed.


All references from 4th International Workshop on HIV & Women, 13-14 January 2014, Washington DC unless indicated otherwise.

  1. Heidari S et al. The inclusion (or exclusion) of women in HIV research: from clinical studies of ARVs and vaccines to cure strategies. Oral abstract_3.
  2. Walmsley S. Invited Lecture: State-of-the art lecture on new therapy options.
  3. Creticos C et al. Single-tablet regimen rilpivirine / emtricitabine / tenofovir DF is safe and well-tolerated to efavirenz /emtricitabine/tenofovir DF in ART-naïve females. Oral abstract_16.
  4. Squires K et al. The efficacy and safety of raltegravir in women. Oral abstract_17.
  5. Saunders P et al. Gender disparities in treatment based outcomes persist in the era of modern ART. Oral Abstract_20.

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