Several presentations at last month’s Conference on Retroviruses and Opportunistic Infections (CROI 2026) in Denver, US provided hope for the development of therapeutic vaccinations as well as T-cell and antibody therapies that could lead to long-term control or even a cure for HIV.
It had been thought that HIV infection, especially long-term infection, exhausts the immune cells that normally fight infections, leading to a state of immune ‘senescence’ (premature ageing) and long-term, low-level inflammation.
It’s been assumed that even in people who maintained viral suppression for decades, this state of ongoing immune dysregulation was the primary driver of the relatively higher levels of cardiovascular disease, non-AIDS-related cancers, and neurodegenerative disorders seen in people with HIV.
Long-term HIV need not rule out a cure
In the last few years, however, it has become apparent that at least some of HIV’s long-term survivors – people who have been on antiretroviral therapy (ART) for 20 years or more – have an immune system that’s better placed to continue fighting HIV than was thought.
At the IAS 2025 Conference in Kigali last year, a plenary talk by Professor Xu Yu of the Ragon Institute at Harvard showed that a large proportion of people on long-term ART had all the HIV DNA in their cellular ‘reservoir’ of potentially replicable virus stored in locations where it was unlikely ever to reactivate.
At least some of these people, if their ART was stopped, might prove to be post-treatment controllers (PTCs), and this has been known since the Ragon group published some of their findings in 2023. What was unknown was which immune mechanism selectively inactivated the reservoir DNA, although there were signs it might involve the natural killer cells of the innate immune system in people treated soon after acquiring HIV.
Long-term HIV, but CD8 cells look brand-new
At CROI, however, Dr Victor Appay of the University of Bordeaux in France demonstrated that many long-term virally suppressed people have a population of HIV-sensitised CD8 cells – which include cytotoxic T-lymphocytes, in other words cells that kill cells, in this case ones harbouring HIV.
Dr Appay was able to assess CD8 function in blood samples from the IMMUNOCO cohort, a group recruited in the 1990s with samples dating from as early as 1988, before any ART was available, and compare them with samples taken from participants receiving antiretroviral monotherapy in the early 1990s. In some cases, he was able to compare these with more recent samples taken from the same individuals, taken some 20 years later, by which time they were receiving combination ART.
He told the audience: “When I started this project, I thought that over so many years of treated infection as well as ageing in the donors, I would be characterising old, senescent CD8 cells.”
He did find that over time, as the immune signal from active HIV faded, the number of CD8 cells that recognised HIV also decreased and were only detectable in a third of the long-term treated participants, at a rate of only 0.1 cells per million.
He found that the cells from long-term treated participants had much lower levels of CD38 and CD57, markers of immune activation and cellular ageing, suggesting the cells had not been worn down by years of fighting the virus.
But what was surprising was that they were much more likely to carry a marker called CD28. This is characteristic of ‘new’ CD8 cells – ones that are early in the process of differentiation, as if they’ve only recently ‘learned’ about HIV. Tests showed that the cells could not have re-acquired CD28 or somehow reversed their development process, so they must be genuinely new cells with properties akin to stem cells.
They also carried a transcription factor called TCF-1, which regulates the self-renewal and generation of memory T-cells. It was found to be more frequent in people in cure studies who had responded well to broadly neutralising antibody (bnAb) therapy by being able to stay off ART for long periods without HIV bouncing back.
Dr Appay said it was also even more frequent in people given a T-cell therapeutic vaccine as well as bnAbs. Although only 0.1% of the T-cells in the long-term treated samples recognised HIV compared with 0.4% pre-ART, 20% of them carried the TCF-1 marker compared to just 5% pre-ART.
The real importance of this was revealed when the CD8 cells were exposed to HIV in a lab dish and given a vaccine-like boost of HIV antigens. Fifty per cent of the CD8 cells from long-term treated participants proliferated in response to HIV compared with only 0.5% of cells from people before ART. A large majority of these cells were also able to secrete perforin and granzyme B, the two cell-killing proteins which CD8 cells deploy to kill their targets.
Dr Appay’s team also compared the proliferative ability and the production of granzyme B and perforin in cells from long-term treated donors with those from long-term HIV controllers, and found their profiles were almost identical.
It’s still important to remember that only about a third of people on long-term ART carry CD8 cells that recognise HIV at all, but Dr Appay said CD8 cells could be ‘primed’ to recognise HIV anew by exposing them to a combination of HIV proteins, an immunogenic peptide (protein fragment) called ELA, derived from a naturally occurring anti-cancer protein, and an adjuvant (a substance that boosts immune responses) called cGAMP.
In the case of people who still have HIV-sensitised CD8 cells, using a different immune booster called a TLR8 agonist produces significant increases in proliferation and granzyme B and perforin production. Dr Appay’s team at Bordeaux, in collaboration with Japan’s Laboratory of Precision Immunology, have conducted pre-clinical lab-dish research combining an mRNA therapeutic vaccine with a TLR8 agonist to produce CD8 cells with enhanced proliferative and cytotoxic abilities.
‘anAbs’ and CD8s in post-intervention controllers
Victor Appay was not the only researcher to present interesting findings on the immune response to HIV, especially in people who are long-term post-treatment controllers (PTCs) or, in the case of people who took part in a cure study and controlled the virus without ART, post-intervention controllers (PICs).
Dr Katie Fisher from Aarhus University in Denmark presented findings from three PICs who took part in three different cure studies – here, here and here – and looked at their antibody and CD8 responses. Two of these people have been off ART without viral rebound for 7.5 and 6.5 years respectively, while the third stayed off ART for 2.5 years before their HIV came back.
These three people all had potent levels and types of anAbs – not bnAbs, but autologous neutralising antibodies, which are antibodies the person generated themselves. These antibodies were not present in the one person who had supplied a blood sample after HIV infection but before they started ART. However, they were present in all three people while they were taking ART and before they started an analytic treatment interruption (ATI, in other words a treatment break), suggesting that they were poised to combat HIV before it even became detectable in viral load tests.
In the person who had to resume ART, it could be shown that this was directly due to a mutation in the virus that gave it immunity to their repertoire of anAbs.
All three people also had CD8 cells that had high levels of expression of a particular gene characteristic of new or ‘stemlike’ CD8 cells, which was not activated in non-PICs, and when mice infected with HIV were inoculated with components of one of the patient’ CD8 cells, which by themselves reduced the viral load in the mice by 2 logs (100-fold).
Active genes in late rebounders
Anna Farrell-Sherman from the Fred Hutchinson Cancer Center in Seattle, USA, compared the genetic profiles of 20 people who stopped ART without bnAbs or any other intervention. Thirteen were initial ‘non-controllers’ – meaning that their pre-ART HIV viral loads had consistently been over 3000 copies – and in all cases their HIV rebounded to above 200 copies within a median time of 17 days, and all within a month. There were also seven ‘controllers’ whose pre-ART viral loads had always been under 3000. For this group, the median time to viral rebound was longer at 40 days, with only two rebounding within a month, and one person not rebounding until 82 days.
All participants were monitored very closely and donated a whole battery of blood samples so that the activity of genes from different parts of their immune system could be measured.
It was found that genes that regulated antiviral activity started to become active in the controllers earlier than in the non-controllers – even in the week before HIV became detectable to an ultrasensitive viral load test with a detection limit of two copies.
But what was more impressive was that several families of genes that governed the activity of CD8 cells, antigen-presenting cells and Natural Killer (NK) cells were active right from the start of the controllers’ ATIs. By contrast, no activity was seen in any of these genes at any point during the non-controllers’ ATIs.
The fact that there was such a sharp distinction between controllers and non-controllers gives researchers a set of targets to aim for when devising gene therapies and therapeutic vaccines to control HIV.
