UK data suggest that treatment isn’t curbing the gay HIV epidemic – because of undiagnosed infections

In the UK, the majority of the HIV-positive gay and bisexual men who are likely to transmit HIV remain undiagnosed, Valerie Delpech of the Health Protection Agency said on Monday. Moreover, modelling work suggests that 48% of all new infections in gay and bisexual men are acquired from men with undiagnosed primary HIV infection, and 34% from men with undiagnosed long-term infection.

Speaking to the IAPAC meeting Controlling the HIV epidemic with antiretrovirals, she warned that good access to antiretrovirals in the UK has not led to a reduction in new infections. “This is due to the undiagnosed fraction and particularly men in primary infection,” she said.

Dr Delpech suggested that the situation in the UK provides useful insights into what is possible when other countries expand their access to antiretroviral treatment, perhaps motivated by considerations of ‘treatment as prevention’ (TasP).

In the context of the UK's free, open-access healthcare system, over 95% of people newly diagnosed with HIV are connected with specialist care within three months, and over 95% of people who attend during one year are retained in care the following year. Moreover, 87% of people with a CD4 cell count below 350 cells/mm³ are taking antiretroviral treatment. Few other countries have comparable results.

The area in which the UK performs poorly is testing and diagnosis.

The area in which the UK performs more poorly is testing and diagnosis. One quarter of people who have HIV are undiagnosed and one half of people with HIV are diagnosed late: in other words, when they already need HIV treatment (CD4 cell count below 350 cells/mm³). These figures have not changed much over the past decade.

It is thought that only 15 to 25% of gay and bisexual men take an HIV test in any given year.

Modelling work by Paul Birrell of the Medical Research Council (unpublished at present) suggests that the annual number of new HIV infections in gay and bisexual men (incidence) changed very little between 2001 and 2010, with between 2000 and 3000 new infections each year.

Dr Delpech commented: “Despite substantial progress of ‘test and treat’ prevention policies over the past decade in the UK, there is no evidence of a reduction in the incidence of HIV infection in men who have sex with men.”

In order to begin to explain this, she showed analyses prepared by her Health Protection Agency colleague Alison Brown, who wished to estimate the number of men who are ‘infective’; in other words, who have a viral load above 1500 copies/ml, and so would be more likely to transmit the virus.

Dr Brown found that, of approximately 40,000 HIV-positive gay and bisexual men living in the UK in 2010, one third (35%) were infective. But, importantly, 62% of those who were infective remained undiagnosed.

Of those remaining, just 5% had a CD4 cell count below 350 cells/mm³ but weren’t taking treatment. A further 12% had CD4 cell counts between 350 and 500 cells/mm³ and 16% had a CD4 cell count above 500 cells/mm^3. Finally, 5% were taking treatment but were not yet virally suppressed.

So, although the number of people taking treatment could be increased, this wouldn’t make a substantial difference to the epidemic. Dr Brown found that changing guidelines so that treatment was recommended for all with a CD4 cell count below 500 cells/mm³ would only reduce the proportion of infective men from 35 to 29%.

Dr Delpech then presented further unpublished modelling work, this time prepared by Andrew Phillips of University College London. This individual-based stochastic computer simulation model incorporates an extensive range of behavioural and public health surveillance data collected over three decades.

The model aims to estimate new infections, disease progression and the effect of antiretroviral therapy in gay and bisexual men in the UK.

Professor Phillips found that the high incidence of the early 1980s declined following widespread condom use. Incidence then rose by 26% after the introduction of antiretroviral treatment, due to a modest reduction in condom use, to reach approximately 0.5 per 100 person-years. 

However, the finding of the modelling exercise that most surprised delegates was that, in 2010, an estimated 48% of new infections were acquired from undiagnosed men who themselves were in the phase of primary infection (i.e. they were infected in the last six months and so would have had an exceptionally high viral load).

A further 34% of new infections were acquired from other undiagnosed men, those with long-term infection.

“How are we going to make TasP work if people are infecting one another very early on when they are unaware of their status?” asked Dr Delpech.

“The relative contribution of infections from those who are either diagnosed or on treatment is actually very small in comparison,” she said. Just 10% of new infections were thought to have originated in men who were diagnosed but not on treatment, while 7% came from men on treatment.

Professor Phillips’ model also examined a series of 'what if' situations, to see what might have happened to the UK epidemic in a range of alternative scenarios. Incidence would have steadily and further increased had HIV treatment never been introduced. Much more dramatically, it would have quadrupled if all gay and bisexual men had stopped using condoms from the year 2000 onwards.

Dr Delpech underlined the continued importance of condom use in limiting the spread of the epidemic, even when antiretroviral treatment is widely used.

Another 'what if' scenario was to ask what would have happened if, from the year 2000, treatment had been recommended to all as soon as they were diagnosed. This only brought incidence down by about 20%.

A much more positive response would have been a combination of much higher testing rates (two-thirds of men testing annually) and treatment for everyone diagnosed. This would have brought incidence down by 62%.

Reducing the number of men with undiagnosed primary HIV infection will be extremely challenging.

Looking at all these data together, Valerie Delpech concluded that the problem of undiagnosed infection – especially undiagnosed primary infection – poses a substantial challenge to the concept of ‘test and treat’. She stressed that recommending treatment at a higher CD4 cell count would not in itself make a significant difference.

“We have excellent care but unless we do something about that undiagnosed fraction, as well as not disinvesting in primary prevention, I don’t think we’re going to completely eliminate HIV from the UK,” she said.

Reducing the number of men with undiagnosed primary HIV infection will be extremely challenging without far more frequent HIV testing, further improvements in testing technologies, widespread behaviour change, targeted partner-notification services and better awareness by healthcare staff of seroconversion symptoms.

Commenting on the presentation, Kevin Fenton of the Centers for Disease Control and Prevention said that there were very different patterns in the United States and the United Kingdom. Testing rates are somewhat higher in the US, whereas retention in care and access to antiretroviral treatment are considerably poorer. He said he believed that in the US, they would have a “much bigger bang for the buck” by focusing on "prevention with positives" – in other words, improving linkage to and retention in care, access to HIV treatment, risk reduction interventions and treatment of sexually transmitted infections in people with diagnosed HIV.

The slides for Valerie Delpech's presentation are available on the IAPAC website.