UK cohort of elite controllers finds 32 people with undetectable viral loads who have never taken treatment

Since 2017, UK researchers have identified a group of 32 ‘elite controllers’ – rare individuals who are living with HIV and are able to maintain undetectable viral load without taking antiretroviral therapy (ART) – and they expect to find more controllers over time. They have also been able to reassure another ten individuals who had ambiguous HIV test results that they don’t have HIV after all.

The IDRIS (Indeterminate Retrovirus Infection Service) collaboration is a referral service based at Imperial College London that has two aims. The first is to confirm or rule out HIV infection in people with no detectable viral load off therapy and who may also have ambiguous HIV antibody test results. The second is to study how these ‘elite controllers’ manage to keep their HIV suppressed without having to take treatment. This is in the hope that scientists can replicate this feat in others and even achieve a functional cure – one that does not necessarily eliminate all HIV from the body but keeps it permanently suppressed.

In IDRIS’ first four years, between 2017 and 2021, 42 people with indeterminate HIV antibody status and viral loads persistently under 20 copies/ml were referred from their HIV clinics. During these four years, 17,904 people tested positive for HIV in the UK, so this figure would represent only 0.18% or one in 560 of all those positive tests. But it's important to emphasise that these are unlikely to be the total number of people in the UK who could be elite controllers – or post-treatment controllers if they came off ART.

IDRIS co-coordinator Professor Sarah Fidler told aidsmap: “We only see referrals where there are people who chose not to start ART despite recommendations. So there may well be many more spontaneous controllers who started ART. And we only see those referred to our specialist service from across big centres or via the UK Health Security Agency. Some clinics manage such cases on an ad hoc basis on their own so I’m sure we are not seeing even all the controllers who are not on ART.”

Also, IDRIS uses a strict definition of ‘controller’: someone not taking antiretroviral therapy (ART) who has at least two viral loads under 20 spaced at least six months apart.

Although there is no standard accepted definition, an ‘HIV controller’ is generally someone who maintains a high CD4 count and has a detectable but low, or intermittently undetectable, viral load without taking ART. An ‘elite controller’ maintains a high CD4 count and has a persistently undetectable viral load (often under 40, though definitions vary). An ‘exceptional elite controller’ is someone who has a viral load persistently undetectable even by ultrasensitive tests and may also have no detectable proviral HIV DNA in the central-memory T-cells that comprise the HIV reservoir. An even rarer elite, such as  Californian Loreen Willenberg, appear to have achieved the ultimate by eliminating replication-competent HIV from their bodies altogether, thus achieving a self-cure: but no more than nine of these people have so far been found.

Most of the current members of the IDRIS cohort had first tested HIV positive longer ago than 2017. The average time between their first positive (or ambiguous) HIV test and confirmation (or retraction) of their HIV status by IDRIS was six to seven years, and some had been waiting with an indeterminate HIV status for 20 years or more.

Ambiguous test results

‘Indeterminate’ may mean several things. The initial HIV test used at sexual health clinics is now an antibody/antigen test, meaning it detects not only the immune system’s reaction to an HIV infection, but also a specific viral protein that appears before antibodies do. If someone’s result is positive to either of them, they are given a second ELISA test, usually by a different manufacturer. If this second test gives a negative result, then there is an ‘indeterminate’ or ‘ambiguous’ result.

This most likely means the first result was a false-positive, which can arise for several reasons from simple lab error to having certain infections such as hepatitis B or C or having an auto-immune condition like rheumatoid arthritis. But occasionally it signifies a very low-level infection, such as is occasionally seen in people who are infected with HIV despite taking PrEP.

The next step in confirming an HIV result used to be the Western blot test. This is more specific than an ELISA as it detects single HIV proteins – usually nine in all. If more than three proteins are detected in a Western blot test, the result is regarded as proof of infection. It can take up to 90 days after infection before a Western blot can give a definite result, which is why they are no longer used as the standard confirmatory test. But they are still useful if ELISAs are giving ambiguous results. 

Most of the current members of the IDRIS cohort had first tested HIV positive longer ago than 2017. The average time between their first positive (or ambiguous) HIV test and confirmation (or retraction) of their HIV status by IDRIS was six to seven years, and some had been waiting with an indeterminate HIV status for 20 years or more.

Western blot results conducted by the Health Security Agency were able to confirm or rule out HIV infection in all but two of the 42 samples. In Imperial College’s own labs, the patients’ samples were also tested with ultrasensitive in-house assays able to detect as little as one copy/ml of HIV RNA in blood plasma and one copy of DNA in 200,000 white blood cells.

Controller, HIV-negative or indeterminate?

Of the 42, 10 had HIV infection ruled out; they were found not to have HIV antibodies by Western blot test, and also had no HIV RNA or DNA detectable. For many of these individuals, this ended years of uncertainty about their HIV status.

In their paper, the IDRIS team comment that ceasing to use the Western blot as the standard confirmatory test may have left a gap in the protocol for resolving an indeterminate HIV test.

Sarah Fidler says: “If people have an initially indeterminate HIV antibody test, clinics will test for many other infections and autoimmune diseases that have been reported to impact HIV antibody tests. There’s many, and it can take as long as two to three years. Often these two to three years of testing have happened before they are seen in our clinic, and if the Western blot and HIV DNA have remained negative we can usually confirm status for most people.”

Another 20 were classic elite controllers. Their confirmatory western Blot tests were positive but they maintained a viral load below 20. But they were “molecular positive”, which means HIV RNA and DNA could be detected by PCR testing.  

And there are 10 more who might fit into the category of “exceptional elite controllers”. They maintained virological undetectability but also (at least, initially – see below) did not have any HIV RNA or DNA detected in their blood plasma or cells. But their Western blots were still positive.

This leaves just two who, on the basis of the definition used by the IDRIS researchers, could not be categorised as HIV positive or negative because their antibody tests were persistently indeterminate – they gave ambiguous results.

“Essentially they didn’t have more than three bands on a Western blot and no detectable plasma RNA (one later tested RNA-positive),” Sarah Fidler comments. “The presence of HIV DNA does make excluding infection difficult.”

Does anything distinguish elite controllers?

Excluding these two, the characteristics of the other 40 cohort members revealed clear differences between the controllers and HIV-negative members but not between the controllers who had detectable HIV RNA/DNA and ones who did not. Over half the controllers were women and there were more people of Black ethnicity in the whole cohort and among the controllers (57% in both cases) than White (22.5% of the whole cohort, 20% of the controllers).

This is a little different to the demographics of other groups of people with HIV in the UK, but the reasons can only be guessed at. It is an established fact that women tend to have lower viral loads than men, as oestrogen seems to act as a viral suppressant, and the majority of the handful of functional cures so far have been reported in women. The larger number of Black people in the cohort could, the authors suggest, be due to natural variations in the virulence of non-B subtypes.

The HIV-negative group however were very different from the controllers in other ways. They were a lot younger, with a median age of 29, compared with 54 for the 20 HIV-positive people with detectable HIV proviral DNA, and 40 for the 10 undetectable. And the median time from their initial HIV positive antibody test was just three years compared with a median of six to seven years for the controllers.

This makes it clearer that IDRIS were having two quite separate groups of people referred to them: a majority of HIV controllers with unusually well-suppressed infections, and a minority of HIV-negative people who for one reason or another had an initial positive antibody test.

Both the HIV controllers and the HIV-negative people had high CD4 counts and normal CD4:CD8 ratios. In the case of the controllers, this is a clue to what may be going on in their immune systems. People without HIV normally have 50% to 70% more CD4 cells (which direct other parts of the immune system) than CD8 cells (which directly kill virus-infected cells). In most people with HIV this ratio is reversed, and even if their CD4 counts reach normal levels, they may still have fewer CD4 than CD8 cells. But the controllers had the same CD4:CD8 ratios as the HIV negative people.

In some of the controllers, there was evidence of immune over-activation. The normal CD4 count range is wide, from 300-1400. But three of the controllers with detectable proviral DNA had very high CD4 counts: 1594, 2446 and 2466 respectively. The normal CD8 count range is 200 to 900 and the two people with the highest CD4 counts also had very high CD8 counts (both, by coincidence, the same, at 2531). Three others had CD4 and especially CD8 counts at the high end of normal, including one of the DNA-negative people.

This may add to the evidence that in some cases, controllers manage their HIV with a particularly strong and specific CD8 response. This is not an entirely good thing: elite controllers often develop other conditions characteristic of inflammation and immune over-activation. The controllers also had somewhat higher levels on average of CD25 and HLA receptor activity in their CD4 and CD8 cells, though this was not statistically significant.

Counter-intuitively, however, the controllers also had lower levels than the HIV-negative people of three of the cytokines that indicate inflammation: TNF-a, MIP-1b and especially CRP. In the case of CRP, the average levels were only 30% in the DNA-negative viral controllers of what they were in the HIV-negative people. None of these differences were statistically significant because only 23 people had their cytokines measured, but they do point to differences in immune response among controllers that might be quite subtle.

 “I think the numbers are too small to make any assumptions of mechanism or significance at the moment,” Sarah Fidler said. She added that IDRIS has seen another 20 or so participants in 2022 and 2023, so she expects to get more data over time.

In the case of people who originally tested negative for plasma HIV RNA and proviral DNA, there is always the hope that they might go one stage further and become one of the rare people who achieve permanent remission, i.e. a functional cure. However, this hasn’t happened to anyone yet.

Of the 30 initial viral controllers, two, both DNA positive, lost virological control; they developed viral loads of 7405 and 802 and started ART, respectively two months and one year after their original IDRIS assessment. Another two (one of them DNA-negative) chose to start ART anyway, despite undetectable viral loads.

Of the remaining 26, 20 attended follow-up appointments over the next four years and continued to have their plasma RNA and proviral DNA levels checked. Eleven switched their ‘molecular status’.  Two switched from being RNA- and DNA-positive to negative, thus becoming ‘exceptional elite controllers’. They did so quickly, with no measurable DNA two months after it was originally found (though it is worth remembering that there is always a margin of error and the chance of false-positive and false-negative results in these complex tests).

But nine of the ten initially RNA- and DNA-negative switched to being positive: proviral DNA was found in all of them, and plasma RNA at low levels (from two to 11 copies) in all but one of them. The interval between the initial negative assay result and the positive one ranged from one month to 27 months. This leaves one person who has been negative for RNA and DNA throughout.    

Sarah Fidler comments: “The assays used to measure proviral DNA are research tests that are probably functioning around the limit of detection. So the fluctuations of positive to negative and vice versa may not be very significant. I agree detection of proviral DNA indicates that none of the study participants are ‘cured’ of HIV but these individuals are representatives of excellent and long lasting spontaneous viral control. Also, we only measure whole DNA: our assay can’t distinguish between defective and intact DNA.”

Investigations into these patients have not stopped. Amongst other tests planned, IDRIS is collaborating with the group at Harvard University that performed the whole-genome sequencing that identified that controllers might have HIV DNA integrated into different areas. The significance of IDRIS’ work is not that they have discovered another functional cure for HIV, but that a research cohort of elite controllers is now finally underway in the UK.