Q&A on the PARTNER Study: How to Interpret the Zero Transmission Results
The following Q&A are linked to the PARTNER study that was published in the Journal of the American Medical Association (JAMA) on 12th July 2016 with full free access. The technical report of this study in HTB is here.
What is the PARTNER study?
The PARTNER study is a large international study. It looked at the risk of HIV transmission when viral load is undetectable on HIV treatment (ART).
Who was enrolled?
The study enrolled 1166 couples where one partner was HIV positive and on ART, and the other was HIV negative. In order to join the study, couples had to already be having sex without condoms.
The study enrolled both gay and straight couples and the average age was about 40 years old.
How long had the couples been having sex without condoms?
There was an interesting range of time that couples had already not been using condoms.
On average this was roughly:
- 1.5 years for gay couples (IQR 0.5 to 3.5)
- 2.8 years for straight couples where the woman was HIV+ (IQR: 0.6 to 7.5)
- 3.6 years (IQR: 0.7 to 11.4) for straight couples where the man was HIV+.
The IQR in brackets refers to the range for the middle 50% of each group.
So 25% of gay couples had been having condomless sex for less than 6 months.
Similarly, 25% of straight couples where the man was positive had been having condomless sex from more than 11 years.
How long had the positive partners been on ART?
This time was thought to be important because the longer someone is on treatment, the lower viral load is likely to be.
Also, the longer someone has had an undetectable viral load, the lower the risk of viral rebound.
Again there were differences with each group and between groups.
HIV positive partners had been on ART for approximately:
- 4.8 years (IQR: 1.9 to 11.4) for gay men.
- 7.5 years (IQR: 3.3 to 14.2) for straight women.
- 10.6 years for straight men (IQR: 4.3 to 15.6).
When did the study run?
Enrolment took place between September 2010 and May 2014.
Early results were presented in February 2014 before the end of the study.
These final results were published in July 2016.
An extension of the PARTNER study called PARTNER 2 is continuing in gay men. This is to get an equivalent amount of data for gay men as for straight couples.
Where did the study take place?
This was a European study at 75 centres in 14 countries. These countries and centres are listed on the PARTNER website.
PARTNER 2 is still enrolling gay couples.
What were the main PARTNER study results?
The main results were that there were no linked HIV transmissions from the HIV positive to the negative partner. This was after about 58,000 times when couples had sex without using condoms.
To be included in the results, viral load had to be undetectable at the most recent test. Undetectable in this study was defined as being less than 200 copies/mL.
The results were not affected even with other STIs, They were not affected by likely viral load blips between viral load tests.
How does the study know about these risks?
Everyone in the study also provided detailed diaries about their sex lives.
This meant that the study could look at different types of risk.
For example, we have a good idea of how many times this included anal sex and vaginal sex with ejaculation by the positive partner. These are the highest sexual risks.
What is new about the published results?
Early results from PARTNER were presented at a conference in 2014.
The new results include more follow up -- now with approximately 58,000 times that participants in the study had sex without condoms (compared to 45,000 in the 2014 report).
The results have also now been extensively reviewed by other experts. The results have been published in one of the world's leading medical journals.
The study still reports zero linked transmissions.
What is special about the PARTNER study compared to other transmission studies?
Most other studies only reported relative risks -- for example comparing the risk from being on ART compared to not being on ART.
PARTNER was able to estimate actual risks for different types of sex
Importantly, all risks are estimates to be zero -- even for highest risks when the positive partner comes inside the negative partner.
This is after approximately 58,000 when couples had sex without condoms.
How many people became positive from their partner?
Zero. There were no transmissions within couples when viral load was undetectable.
Were there other HIV transmissions?
Yes, 11 people did become HIV positive during the study (ten gay men and one heterosexual). In nearly all these cases, the person said that they had sex with another person other than their main partner.
Approximately 33% of gay couples and 4% of straight couples were in open relationships.
Importantly, none of these transmissions were linked to their HIV positive partner.
How can the study prove that transmission didn't come from the positive partner?
Research tests are able to compare different viruses from different people.
This uses special laboratory testing -- called phylogenetic analysis. If the genetic structure of the new virus is very different to the partner's virus, this shows the infections are not linked. It shows that HIV came from outside the main relationship.
These results were reviewed by independent experts and are an essential part of the study.
All these reviewers agreed that the infections were not connected -- even after running the analyses in different ways.
What does the "95% confidence interval" mean?
Study results usually need to consider the possibility that the actual result might not have been seen in the study. For example, whether different results would be seen if the study continued for longer. Or if it had enrolled more people.
The potential range of results, when allowing for other factors is called the confidence interval or CI.
The 95% CI is the range of results that the researchers are comfortable for being a true result. The full range would be a 100% CI, so there is still a chance that the true result might be also be outside this range.
In the PARTNER study, the overall results -- and all the separate risks -- also reported an upper risk that might be possible. This upper risk does not represent a risk that was actually observed.
As the study accumulates larger numbers of couple-years in follow-up without transmissions this risk would become closer and closer to zero. Even if there were no transmissions with a study twice as large and for twice as long, the number never reaches zero.
Because the actual risk was zero, it doesn't mean this upper risk is an actual risk. It is a factor of the way research describes data. But the true value could be within this range,
The amount of follow-up time with these risks was used to calculate different "upper limits" for different risks.
As an example, the upper limit of the 95%CI for anal sex is much higher in this study for straight couples compared to gay couples. This doesn't mean that the actual risk is higher, just that because there are much fewer years of follow-up for straight couples having anal sex.
More than 20% of straight couples reported having anal sex.
What are the implications of the results?
The lack of HIV transmissions should challenge the wrongly held common assumption that there is always a risk just because someone is HIV positive.
The results actually go further. The lack of transmission challenges scientists to prove that transmission is actually possible when viral load is undetectable.
The PARTNER study -- as with other studies -- suggests that there is likely to be a level of viral load where HIV transmission does not take occur. PARTNER suggests this might be at 50 copies/mL, or at 200 copies/mL or perhaps even higher.
How will people benefit from these results?
There are many benefits from these results.
- HIV positive people can become less anxious and concerned that they are a risk to their partners whenever they have sex. This can still be a worry, even when using condoms.
- HIV negative people can be less anxious about risk. Even when using condoms, this residual risk can limit full enjoyment of sex.
- Less anxiety and fear can help with closer communication and better sex. For many people, a good sex life is an important and essential part of life.
- Some people might enjoy not using condoms in a way that wasn't possible when they still worried about HIV.
- Sero-different couples who want to have children can conceive from just having sex without the need for additional PrEP.
- Reducing fear about HIV transmission might reduce the stigma and rejection HIV positive people encounter when meeting new partners.
- Legal cases where HIV is used because of a theoretical rather than actual risk will hopefully become more rare. The results might enable some people to launch an appeal.
US activist Sean Strub from the SERO project (www.seroproject.com) said in response: "Hundreds of people living with HIV in the U.S. have been charged with criminal offences for the perceived or potential risk of HIV exposure or transmission. Some are serving or have served long prison sentences for spitting, scratching or biting and others for not being able to prove they had disclosed their HIV positive status before having sexual contact (even in the absence of any risk of HIV transmission). HIV criminalisation has created a viral underclass in the law, further burdening a disenfranchised community, putting a disproportionate share of the shared responsibility for preventing sexually-transmitted infections on one party, and discouraging people at risk from getting tested for HIV."
Which other studies have looked at HIV transmission with low viral load?
Other studies include:
- The ongoing Opposites Attract study in Australia in gay couples.
- The HPTN-052 study in heterosexual couples that reported results in 2010.
- The Rakai study in Uganda that first reported a clear connection between viral load and risk of transmission in heterosexual couples in 2000.
What is the PARTNER 2 study?
The PARTNER 2 study is a continuation of the PARTNER study that is continuing in gay men. Gay couples are conituning to be followed up until 2019.
PARTNER 2 is still enroling people until 2017. If you are interested in joining this research please see this link to the study website.
Where can I get further information?
A Q&A from the PARTNER study is also online as a PDF.
The full study is available free on the JAMA website.
Why did the results take so long to get published?
This is always difficult. The researchers in PARTNER were working hard to get this published as soon as possible.
However, now that the results are published in such a high profile journal, they can formally be included for guidelines, for information produced in CDC recommendations and for use in legal cases.
It is also significant that JAMA have made online access free for the full article. This reflects the importance of the results have the widest possible access.