Two weeks ago, the announcement that a baby had been ‘functionally cured’ of HIV disease with the use of very early antiretroviral therapy (ART) caused great excitement at the start of the Conference on Retroviruses and Opportunistic Infections (CROI 2013). Now a study from France has found 14 adult patients who also started a course of ART soon after infection, who subsequently stopped it, and have not had to re-start because they have largely – and in eight cases completely – maintained undetectable viral loads for at least four years after stopping therapy (the baby has, so far, only managed a year off therapy).
Furthermore, the researchers suggest that such cases are only not more common simply because, once having started ART, few people stop. They estimate that 15% of people with HIV, if ART is started within six months of HIV infection and maintained for at least a year, could subsequently become so-called “post-treatment controllers”.
Their estimate is a stark contrast to findings from studies conducted between 1996 and 2000, soon after the introduction of highly active antiretroviral therapy, which found no evidence that people who began treatment in primary infection could control HIV after stopping treatment. The key difference between those studies, and the French patients described this week, is that earlier studies looked at HIV control in people who had only received treatment for 12 to 18 months. The French patients had been on treatment for an average of three years before stopping, and all started treatment within ten weeks of infection, compared to within six months in previous studies.
The French patients
The 14 patients with durable viral control were found at a number of French clinics and became the subjects of a study called VISCONTI (Virological and Immunological Studies in CONtrollers after Treatment Interruption).
There were ten men and four women who had been infected with HIV between 1996 and 2002. All had started ART within ten weeks of primary HIV infection being detected and in all but two cases PHI was symptomatic. Viral load when first measured was high (average, 100,000 copies/ml), in contrast to spontaneous HIV controllers, who maintain low viral loads from the start. Four started ART within four weeks of infection, the others within ten weeks.
They took ART for at least one year and for an average of three years, with the earliest-diagnosed taking regimens that would now be considered suboptimal, such as dual or triple-NRTI therapies. They then stopped ART and have stayed off it for a minimum of four years and an average of 7.5 years (longest so far, 9.5 years).
During the time off treatment, eight people have not had a single viral load result above 50 copies/ml and only two had any viral load over 400 copies/ml. Four have had a pattern of periods of low but detectable viral load and in three of these, their last viral load was detectable (91, 224 and 289 copies/ml respectively).
Median CD4 count at ART initiation was 502 and at ART discontinuation was 927 cells/mm3; since then median CD4 count has declined slightly to 837 cells/mm3 but in five cases has actually increased.
HIV within reservoir cells
Integrated HIV DNA within cells was measured in eight people during the time off treatment. In these eight, DNA levels in two remained stable, and in one – one of the intermittently detectable ‘blippers’ – increased, but in the other five it has continued to decrease, even off ART. Median intracellular DNA levels among the eight decreased from 2389 copies per million leukocytes during primary HIV infection, to 116 just before ending ART to 39 copies per million cells at the last measurement, an average of six years off ART.
HIV could be induced from patients’ resting cells if they were activated by immune stimulation, showing that they still met the definition of infection, and in two cases from cells where no viral DNA had been detected. However, viral production was very sluggish.
The scientists found specific differences in the distribution of cells that contained HIV DNA.
In the rare group of spontaneous HIV controllers (HICs) who maintain undetectable viral loads despite never having been on treatment, there are similar amounts of HIV DNA in the long-lived central-memory reservoir cells and in partly activated transitional-memory cells, but not in fully activated effector-memory cells. This suggests that in these patients there is ongoing DNA activation but that it is stopped by immune-system surveillance before actual viruses are produced.
In the case of the post-treatment controllers (PTCs), there was a significantly lower proportion of HIV DNA in central-memory cells, suggesting that the long-lived reservoir of potential new HIV replication had indeed shrunk.
How these patients control HIV without treatment
The researchers also compared the immune characteristics of post-treatment controllers and spontaneous HIV controllers.
They found that the viral control seen in these two groups had almost opposite explanations. The immune cells of HICs tended to have human leukocyte antigens (HLAs, cell surface molecules that stimulate the body’s response to viral invasion) that responded only sluggishly to HIV, and which have been associated with slow disease progression in population studies. This was combined this with cytotoxic T-lymphocytes – CTLs or CD8 cells – that were particularly alert to HIV and efficient at destroying HIV-infected cells.
The PTCs, on the other hand, tended to have HLA types that were highly responsive to HIV, strongly associated with fast disease progression, and tended to have had primary HIV infection (PHI) characterised by high viral loads and severe symptoms (which is why many started early treatment). Post-treatment, however, their immune response was characterised by a lack of CD8+ cell responses to HIV – and a lack of activated CD8+ T-cells.
In spontaneous viral controllers, one might say, the body erects a strong defence against HIV, while in post-treatment controllers, its immune defence is so quiet that it lets HIV stay asleep.
Post-treatment controllers may be more common than we might think. Spontaneous HIV controllers are very rare – less than 1% of the HIV positive population. In contrast, the researchers found that 15% of patients on the French patient database who had initiated ART within six months of infection, taken ART for at least a year, and then interrupted it, maintained an undetectable viral load for at least two years afterwards, and 9% for at least 4.5 years. However, they could only find 70 patients who had stopped ART this way, because most continue once they start.
In short, as the researchers comment, “off-therapy viral control for several years may be associated with very early and prolonged viral treatment. These findings argue in favour of very early ART initiation.”
The challenge now is threefold – to identify whether this pattern holds true in other groups of patients treated from very early in HIV infection, to increase testing rates so that more people are diagnosed within the first two months after infection and so have the opportunity to choose early treatment, and to find ways to extend the immune profile seen in post-treatment controllers to other people with HIV so that those who start treatment in chronic infection can benefit too.
Sáez-Cirión A et al. Post-treatment HIV-1 controllers with a long-term virological remission after the interruption of early initiated antiretroviral therapy: ANRS VISCONTI study. PLOS Pathogens, 9(3): doi:10.1371/journal.ppat.1003211, 2013.