A US collaboration that pooled data from 14 scientific studies containing between them more than 600 HIV-positive people has found that 67 of them were able to maintain or re-establish a viral load below 400 copies/ml for at least 48 weeks after their antiretroviral therapy (ART) was stopped.
This is the first large study to estimate the proportion of people who might be ‘post-treatment controllers’ since the VISCONTI cohort in 2013. This found 14 such controllers, suggesting that nearly one in six people who started ART less than six months after infection might be able to control their viral loads for at least a year after being taken off ART.
The present study, called CHAMP (Control of HIV after Antiretroviral Medication Pause), looked at ten randomised controlled trials and four cohort studies conducted in the US and Canada with results published from 2000 to 2017.
Most of these were studies of treatment interruption, either in its own right, or alongside a therapeutic vaccination or other immune treatment. Three of the four cohort studies were of people treated in early HIV infection while the fourth was of people already known to be viral controllers.
The average age of the participants was 41 with no difference between those treated in early and chronic infection – 19% were women, 69% were white and 25% black. The definition used by CHAMP of a post-treatment controller was that the person had to have a viral load below 400 copies/ml at least two-thirds of the time for at least 48 weeks after stopping ART.
The researchers at Brigham and Women’s Hospital and Harvard Medical School, found that 13% of people who started ART soon after infection were able to achieve post-treatment viral control by this definition. This is very close to the VISCONTI researchers’ estimate that 15% of people treated in early infection might be post-treatment controllers.
CHAMP in addition looked for post-treatment controllers who did not start ART until they were in chronic infection. It found 25 of these, representing 4% of people. This is the first time a study has come up with an estimate for the proportion of people who start treatment in chronic HIV infection who may be post-treatment controllers.
The average time people had been on ART before they stopped was two years, four months in people treated early and five years, seven months for those who didn't start. The average duration of post-treatment viral suppression was 89 weeks – about a year and nine months. Three-quarters of people maintained viral control for more than a year, and 22% of them for at least five years. Although few people were followed for longer than this, the researchers say that two people, both treated in early infection, still have viral suppression more than ten years after stopping ART.
Post-treatment controllers generally maintained their CD4 counts, with only a slight decline of 32 cells/mm3 a year, compared with 221 cells/ mm3 in non-controllers.
CHAMP used a relatively broad definition of viral suppression. The researchers point out that if having a viral load under 400 copies/ml for at least 90% of the time had been the criterion, the proportion qualifying as post-treatment controllers would only be 8% of early-treated and 2% of later-treated people.
This relaxed criterion turned out to be important. The CHAMP researchers found that 40% of post-treatment controllers had at least one viral load measurement over 400 copies/ml after stopping ART and 31% over 1000 copies/ml. However, these viral load peaks happened soon after ART was withdrawn (on average eight weeks, and never more than 24 weeks, after) and they then re-suppressed their HIV.
The researchers also looked at the minority of people who had had their viral loads monitored very intensively – once a week on average. In this minority, transient high viral load peaks were found more often. Two-thirds had at least one viral load higher than 400 copies/ml after ART withdrawal, 45% over 1000 copies/ml, 33% over 10,000 copies/ml and 11% over 100,000 copies/ml.
The researchers speculate that these short bursts of viral replication may actually contribute to the later viral suppression, at least in some people, as they may stimulate a suppressive anti-HIV immune response.
They also note that treatment interruption studies that have strict criteria about whether to re-start ART – typically after just one or two viral loads over a few hundred copies/ml – may miss a lot of people who would have gone on to re-suppress their HIV.
This pooling of studies lacks certain data, such as pre-treatment viral loads for those treated in chronic infection, but it does find that pre-treatment viral load in those treated in early infection was only slightly lower in post-treatment controllers than non-controllers – about 50,000 versus 80,000 copies/ml – and their CD4 counts at ART interruption also very similar – 882 cells/mm3 for controllers and 825 cells/mm3 for non-controllers.
There is one interesting detail: 42% of post-treatment controllers who started ART in chronic infection came from a single study, ACTG 5068. Ten per cent of people in this study ended up being post-treatment controllers compared with 3% in all other studies of people starting treatment in chronic infection.
The distinguishing feature of this study was that it used so-called “pulsed treatment interruptions”. This means that after initially stopping ART, people were put on two short treatment interruptions of 4-6 weeks, with a period of ART in between, before ART was discontinued indefinitely. This was a randomised trial and in the people who received the pulsed treatment interruptions, 15% became post-treatment controllers – as high a proportion as those starting ART in early infection.
This is a suggestive detail as the pulsed interruptions are about long enough to allow the short viral load bursts that may lead, researchers hypothesise, to what they call ‘auto-immunisation’ against HIV.
However, this still only led to 15% of people becoming post-treatment controllers and the combination of factors that lead to a minority of people being able to control their HIV off-treatment, while the majority do not, remain mysterious.
As long as they do remain mysterious, only a small proportion of people will undergo the structured treatment interruptions that might shed light on the science. But at least this US study confirms that post-treatment control is not an especially rare phenomenon.
Background: HIV post-treatment controllers are rare individuals who start antiretroviral therapy (ART), but maintain HIV suppression after treatment interruption. The frequency of post-treatment control and post-treatment interruption viral dynamics have not been well-characterized.
Methods: Post-treatment controllers were identified from 14 studies and defined as individuals who underwent treatment interruption with viral loads ≤400 copies/mL at ≥2/3 of time points for ≥24 weeks. Viral load and CD4+ cell dynamics were compared between post-treatment controllers and non-controllers.
Results: Of the 67 post-treatment controllers identified, 38 initiated ART during early HIV infection. Post-treatment controllers were more frequently identified in those treated during early vs. chronic infection (13% vs. 4%, P<0.001). In post-treatment controllers with weekly viral load monitoring, 45% had a peak post-treatment interruption viral load of ≥1,000 copies/mL and 33% had a peak viral load ≥10,000 copies/mL. 55% of post-treatment controllers maintained HIV control for 2 years, with approximately 20% maintaining control for ≥5 years.
Conclusions: Post-treatment control was more commonly identified amongst early-treated individuals, frequently characterized by early transient viral rebound and heterogeneous durability of HIV remission. These results may provide mechanistic insights and has implications for the design of trials aimed at achieving HIV remission.
Golnaz Namazi, Jesse M Fajnzylber, Evgenia Aga, Ron Bosch, Edward P Acosta, Radwa Sharaf, Wendy Hartogensis, Jeffrey M Jacobson, Elizabeth Connick, Paul Volberding, Daniel Skiest, David Margolis, Michael C Sneller, Susan J Little, Sara Gianella, Davey Smith, Daniel R Kuritzkes, Roy M Gulick, John W Mellors, Vikram Mehraj, Rajesh T Gandhi, Ronald Mitsuyasu, Robert T Schooley, Keith Henry, Pablo Tebas, Steve Deeks, Tae-Wook Chun, Ann C Collier, Jean-Pierre Routy, Frederick M Hecht, Bruce D Walker, Jonathan Z Li