CATIE News

15 March 2012 

Pre-exposure prophylaxis—adherence is key and may explain disappointing trial results

Pre-exposure prophylaxis, or PrEP, is currently being studied as a potential method for a person who is at risk of HIV infection to reduce their risk of becoming infected. It involves taking anti-HIV medications on a regular basis.

There are several types of PrEP being investigated. PrEP drugs may be available in a variety of forms, including pills taken orally or gels inserted into the vagina or rectum. The drugs may need to be taken every day, or coitally (before and after sex), or intermittently (once or twice a week or month).

Most types of PrEP being studied use the antiretroviral drug tenofovir (called Viread when used in pill form) or combinations of tenofovir and FTC (sold as a fixed-dose co-formulation pill called Truvada).

While some PrEP clinical trials have had promising results, others have not. This has raised questions about how well PrEP will work, particularly among women.

Differences in adherence to PrEP among study participants has been suggested as one explanation for the wide range of effectiveness levels in completed clinical trials.

Measuring adherence in clinical trials

Regular adherence to PrEP is likely needed to maintain high levels of drug in the body and maximize the protection provided by PrEP.

Therefore, an accurate measure of adherence among study participants is needed to correctly interpret study results and understand how well PrEP really works.

Adherence among study participants in PrEP clinical trials is normally measured in the following two ways:

  • Self-report – every month participants are asked by study investigators how well they have been adhering
  • Pill or applicator counts – every month participants are asked to return any unused pills or gel applicators to study investigators

Unfortunately, these two methods of measuring adherence are not very accurate and tend to overestimate how well participants are adhering to PrEP. Participants often self-report that they are adhering more frequently than they actually are and participants may not return all of their unused pills or gel applicators.

Overestimation of adherence among study participants through self-report or pill/applicator counts can lead to underestimation of the effectiveness of PrEP.

Drug levels as a measure of adherence

Recently, some completed PrEP clinical trials have also measured adherence by looking at whether study participants who were supposed to be taking PrEP had detectable levels of anti-HIV drugs in their blood or other body fluids. This method of measuring adherence is more accurate than self-report or pill/applicator counts because it is a direct measure of exposure to the drug.

Using the levels of anti-HIV drugs in participants to estimate adherence has a few limitations. A participant’s drug levels are generally only measured once a month during the clinical trial. Therefore, having detectable drug levels suggests that a participant was adhering to PrEP at the time the drug levels were measured; it does not necessarily reflect how good their overall adherence was in the past month since their last visit. Also, anti-HIV drug levels can remain detectable for a few days after taking PrEP pills, therefore having a detectable level of drug at a study visit means that a participant had taken PrEP within the past few days.

Recently released information on drug levels suggests that poor adherence may explain the low levels of PrEP effectiveness in some studies and why some study participants became infected with HIV while taking PrEP.

PrEP clinical trials – effectiveness, adherence and drug levels

iPrEx trial

Study population: 2499 HIV-negative men who have sex with men

Study arms:

  1. Daily Truvada pill – 36 HIV infections
  2. Daily placebo pill – 64 HIV infections

Effectiveness:

  • Overall, a daily Truvada pill was 44% effective at reducing the risk of HIV transmission compared to a placebo.

Estimated Adherence:

  • Self-report: 95%
  • Drug levels: In an analysis of 43 participants who were supposed to be taking a Truvada pill daily (and were not infected with HIV), only 51% had detectable levels of drug in their blood

Additional information on anti-HIV drug levels:

  • Among men who became infected with HIV in the daily Truvada group, only 9% had detectable drug in their blood at the study visit closest to the time of their HIV infection.
  • Men with detectable drug in their blood were 92% less likely to become infected with HIV compared to men who did not have detectable drug in their blood.

FEM-PrEP Trial

Study population: 2,120 HIV-negative heterosexual women

Study arms:

  1. Daily Truvada pill – 33 HIV infections
  2. Daily placebo pill – 35 HIV infections

Effectiveness:

  • Overall, a daily Truvada pill was not effective at reducing the risk of HIV transmission compared to a placebo.

Estimated Adherence:

  • Self-report: 95%
  • Pill counts: 87%
  • Drug levels: In an analysis of 99 participants who were supposed to be taking a Truvada pill daily (and were not infected with HIV), only 38% had detectable levels of drug in their blood

Additional information on anti-HIV drug levels:

  • Among women who became infected with HIV in the daily Truvada group, only 21% had detectable drug in their blood at the study visit closest to the time of their HIV infection.

Partners PrEP trial

Study population: 4758 heterosexual men and women (serodiscordant couples)

Study arms:

  1. Daily Viread pill – 17 HIV infections
  2. Daily Truvada pill – 13 HIV infections
  3. Daily placebo pill – 52 HIV infections

Effectiveness:

  • Overall, a daily Viread pill was 67% effective and a daily Truvada pill was 75% effective in reducing the risk of HIV transmission compared to a placebo.

Estimated Adherence:

  • Pill counts: 97%
  • Drug levels: In an analysis of 198 men and women who were supposed to be taking a Viread or Truvada pill daily (and did not become infected with HIV in the study), drug was detected in 82% of their study visits.

Additional information on anti-HIV drug levels:

  • Among the men and women who became infected with HIV in the daily Viread or Truvada group, only 35% (Viread) and 25% (Truvada) had detectable drug in their blood at the study visit closest to the time of their HIV infection.
  • Men and women who had detectable drug in their blood were 86% to 90% less likely to become infected with HIV than men and women who did not have detectable drug in their blood.

Implications for interpreting PrEP trial results

Recently released information on drug levels has provided additional insight into PrEP clinical trial results.

Poor adherence may explain the lower levels of effectiveness in the iPrEx trial and the lack of any effectiveness for women in the FEM-PrEP trial. More consistent adherence among couples in the Partners PrEP trial may explain the higher level of effectiveness in this study.

Poor adherence may also explain the disappointing results from an ongoing PrEP study called VOICE. The VOICE study found that the use of a daily Viread pill and a daily tenofovir gel did not reduce the risk of HIV infection among women in the study. However, information on drug levels in the women in this study has not been released. The trial is still ongoing and investigating the effectiveness of a daily Truvada pill.

It is possible that the disappointing results among women in the FEM-PrEP and VOICE trials is only partly be due to adherence. Additional research is exploring other factors that might have also contributed to these results.

Among men and women in the clinical trials who did take PrEP regularly, PrEP seems to have provided a high level of protection against the sexual transmission of HIV. The majority of study participants who became infected with HIV while taking PrEP did not have detectable drug in their blood, suggesting they were not adhering consistently. However, these drug levels came from the study visit closest to the time of a participant’s HIV infection and therefore may not reflect the drug level at the time of the exposure to HIV that led to infection. It’s also important to note that in all of the completed PrEP studies, HIV infections still occurred among study participants who were taking PrEP regularly and had detectable drug in their blood.

The lack of toxicity, side effects and drug resistance observed in completed PrEP studies might also be due to poor adherence. If study participants were not taking PrEP regularly, it’s unlikely that they would experience toxicity or side effects or develop drug resistance if they became infected with HIV.

Implications for implementation and further research

The information on drug levels suggests that many study participants had trouble adhering to PrEP. This means that the implementation of PrEP will need to focus on supporting PrEP users when it comes to regular adherence.

Research is needed to identify the barriers that prevented study participants from adhering regularly to PrEP. At the same time, more research is needed to explore other types of PrEP that would not require regular adherence, such as vaginal rings or injections that only need to be used once a month.

James Wilton

Resources

CATIE News story on contradictory PrEP results among women

CATIE News story on VOICE study

CATIE Prevention in Focus article on PrEP

CATIE Fact Sheet on PrEP

CATIE News story on iPrEx trial

CATIE News story on FEM-PrEP trial

CATIE News story on TDF2 and Partners PrEP

REFERENCES:

  1. Baeten J, et al. ARV PrEP for HIV-1 prevention among heterosexual men and women. 19th Conference on Retroviruses and Opportunistic Infections, Seattle, Abstract 29, 2012.
  2. Van Damme L, et al. The FEM-PrEP Trial of Emtricitabine/Tenofovir Disoproxil Fumarate (Truvada) among African Women. 19th Conference on Retroviruses and Opportunistic Infections, Seattle, Abstract 32LB, 2012.
  3. Donnell D, et al. Tenofovir disoproxil fumarate drug levels indicate PrEP use is strongly correlated with HIV-1 protective effects: Kenya and Uganda. 19th Conference on Retroviruses and Opportunistic Infections, Seattle, Abstract 30, 2012.