Advances in the management of HIV over the past three decades have been spectacular, thanks to the development of antiretroviral drugs that prevent the virus from replicating. These drugs have very few side effects, prolong life and block sexual transmission. However, the virus is never eliminated — instead, it hides in immune cells called CD4+ T cells in a non-replicating, latent form. If treatment is stopped, the virus rapidly re-emerges from this latent reservoir1. Given the cost of antiretroviral drugs, the need for ongoing engagement in care and the persisting stigma for people living with HIV, there is intense focus on finding a way to target the latent virus so that treatment can be safely stopped without viral re-emergence. In a paper in Nature, Borducchi et al.2 report remarkable findings that may have achieved just that in a monkey model of HIV.
Disappointingly, no intervention has so far managed to eliminate the latent HIV reservoir in people3. Borducchi and colleagues set out to investigate whether a combination of two treatments could do so in monkeys. The first treatment, GS-9620 (vesatolimod), is an oral drug that activates the Toll-like receptor 7 (TLR7) protein. TLR7, in turn, activates immune cells — not only CD4+ T cells, but also CD8+ T cells and natural killer (NK) cells, both of which can hunt out and destroy virus-infected cells4. Activation of latent HIV contained in CD4+ T cells is thought to render them more susceptible to destruction by other immune cells5. The second treatment, PGT121, is an antibody, one end of which recognizes and binds to key HIV proteins on the surface of infected cells, with the opposite end triggering other immune cells to destroy the target cell6.
Borducchi and colleagues infected 44 monkeys with a hybrid of HIV and the simian immunodeficiency virus. Seven days later, they began to treat the animals with a potent combination of antiretrovirals, similar to that used in humans. HIV rapidly disappeared from the blood of all monkeys, as expected. After 96 weeks, the authors split the monkeys into 4 randomized groups of 11 — one group received no intervention, a second was given GS-9620, a third was injected with PGT121, and a fourth received both GS-9620 and PGT121. The monkeys received these treatments until week 114, then continued to receive antiretroviral therapy until week 130. The investigators then stopped antiretrovirals and waited to see whether the virus rebounded.
The researchers detected the virus in the blood of all 11 animals that received no intervention, within a median of 21 days after stopping antiretroviral treatment. Viral rebound was also seen in 10 and 9 animals in the groups given only GS-9620 and PGT121, respectively. In stark contrast, only 6 of the 11 monkeys treated with both GS-9620 and PGT121 showed signs of the virus rebounding by week 28 after antiretroviral treatment had ceased. The other 5 monkeys in this group remained completely clear of any detectable virus, even using sensitive assays.
Why was the approach so effective? Borducchi et al. found that CD4+ T cells and NK cells were activated in all monkeys that received GS-9620. But activating these cells clearly is not sufficient to destroy infected cells, because treatment with GS-9620 alone did not prevent viral rebound, consistent with a previous report4. GS-9620 has also been shown to activate latent virus, ‘shocking’ it out of its hiding place in monkeys to enable targeting by the immune system4. The authors did not find evidence for this in the current study, but that might be because they began treating their monkeys soon after infection. This meant that the animals had only a small reservoir of virus, which would be difficult to detect.
Although potent neutralizing antibodies such as PGT121 are being widely tested as a way to prevent HIV infection, it has been unclear whether these antibodies actually kill infected cells in the presence of antiretrovirals. There is an added layer of complexity if the virus is latent — can the antibody even recognize these cells? The authors propose that GS-9620 treatment activated the CD4+ T cells harbouring latent virus, rousing the virus and so allowing the antibody to target the infected cell, perhaps with assistance from activated NK cells that are directed to the cell by the antibody (Fig. 1). This type of approach is referred to as ‘shock and kill’.
Borducchi and colleagues’ findings are exciting, and offer hope of a cure for HIV, but there are a few reasons for tempered enthusiasm. First, because the authors began treating the monkeys with antiretrovirals extremely soon after infection, the pool of latently infected cells was small and potentially easier to clear than if the virus had had longer to replicate. Most people living with HIV are diagnosed months to years after infection. In the current study, the monkeys that did not rebound were those with the lowest pretreatment viral loads, supporting the idea that a lower burden of virus before treatment might make the reservoir easier to eliminate. Indeed, this has been shown recently in another monkey model7.
Second, the hybrid virus used here is potentially easier for the monkey immune system to control than are other monkey viruses8. Also, for people infected with HIV, control of virus is extremely rare, even if antiviral treatment is started within days of infection1. Third, the monkeys were followed for only about six months after antiretroviral treatment was stopped. In people with HIV who stop antiretroviral treatment, rebound of the virus can be delayed for as long as two years9,10, so longer follow-up of these monkeys is needed. Finally, and most importantly, we don’t yet know whether interventions in monkey models of HIV reflect what will happen in humans.
The biggest test will now be to see whether administration of GS-9620 and PGT121 (or a related antibody) can produce similar results in people. These clinical trials are being planned, and the results are eagerly awaited. In the meantime, antiretrovirals remain the best and only option for the long-term treatment of HIV infection.
Nature 563, 333-335 (2018)