The 2016 HIV surveillance figures released by Public Health
England (PHE) today reveal that what has already been reported for
specific clinics
and at
is true nationally: HIV diagnoses in gay men last year fell significantly
for the first time throughout the UK since antiretroviral therapy
started becoming available 20 years ago. This forms part of an overall picture
of an across-the-board 18% decline in diagnoses in 2016.

The PHE figures reveal that HIV diagnoses are in retreat in
virtually all groups, even in ones where until recently there appeared to be a
continuing upward trend.

For instance, there has been a 25% drop since 2014 in diagnoses in gay
men born in the UK, and this started showing up in last year’s PHE
figures, with a drop of 12% between 2014 and 2015. However the figures for gay
men born elsewhere in Europe appeared at that point still to be on the rise, with an increase of 11.5% in a 2014-2015. In 2016, however, diagnoses in this group also went into reverse, with a 21% decrease in the last year.

Decreases have been particularly pronounced in London, with
a 23% decline among all populations and a 29% decrease in gay men in the last
year. However there have been declines in all areas of the UK in the last year,
though they vary from area to area: 31.5% (among all groups) in Wales, 17% in the
East Midlands, 16% in Scotland, but only 8% in Yorkshire and the North of England.
The overall decline in infections in gay men outside London was 11%.

Heterosexual infections have been declining in London and
elsewhere since 2005. This has been largely due to the impact of lower
immigration from high-prevalence countries, with less ongoing transmission amongst
first-generation members of the UK-born African community also important.

UK diagnoses among women born in the rest of the EU, however
– most of them not of African ethnicity – increased 31% between 2005 and 2015. However
they too fell by 9.5% last year. Diagnoses actually rose among men born in
southern Africa and women from Asia – but these formed a tiny proportion of the
total and the increases look like insignificant variations in a largely flat trajectory of diagnoses.

The decrease in diagnoses among gay men varies by age.
Diagnoses in men aged 35-49 peaked in 2007 and have declined by 28% since then,
though some of these diagnoses are among heterosexual Africans. However infections
among men younger than 35 appeared to be on a remorselessly upward curve until 2015,
but fell by 25% last year – and most of these will be infections in young gay

One trend that is less welcome, however, is that the improvement
in CD4 counts in people being diagnosed appears to have stalled. Until 2015,
the average CD4 count at diagnosis increased from 314 in 2008 to 428 in 2015,
but fell back to 407 in 2016. Similarly, the proportion of people diagnosed
with a CD4 count below 350 cells/mm3 fell from 55% in 2008 to 39% in
2015 – but increased to 43% in 2016.

CD4 counts at diagnosis have always been especially low in
heterosexual men. In them, the average CD4 count at diagnosis rose from 232 in
2008 to 313 in 2015 but fell to 280 last year, and the proportion diagnosed
with a CD4 count below 350 cells/mm3 is 60%, not a lot better than the 67% seen in 2008, though in 2015 it was

This appears to be at odds with the assumption that we are catching more early infections and have diagnosed most chronic ones. Certainly
the trend in London gay men is different, with an increase in average CD4 count
at diagnosis from 396 cells/mm3 in 2015 to 501 cells/mm3
in 2016. However there is still a high number of men, many of them older heterosexual
men, getting diagnosed with low CD4 counts: in men over 50, 60% were diagnosed
with CD4 counts below 350 cells/mm3.

This has a real impact on mortality. The death rate in the
first year after infection in people diagnosed with a CD4 count below 350 cells/mm3
was 7% in people aged 50-64, over ten times the rate in people in that age group diagnosed
at higher CD4 counts, and also nearly ten times the mortality rate in 25-34
year olds diagnosed with low CD4 counts. In fact, diagnosis with a CD4 count
under 350 cells/mm3 increased the risk of dying in the
first year after diagnosis tenfold in all groups but the over-65s.

In contrast, three figures are higher than ever: the number
of gay men testing and therefore diagnosed, the proportion of diagnosed people in care, and the
proportion virally suppressed.

Last year, out of a total of about 106,500 people with HIV
in the UK, 91,987 or 86.4% were diagnosed (a higher proportion than ever
before); of these, 88,089 were on antiretroviral therapy (95.7%); and of these,
93% had a viral load below 50 copies/ml, or 96% if we use the threshold of 200
copies/ml, as the US CDC did in its recent report.

This 86/96/96 treatment cascade implies that 79.5% of all people
with HIV in the UK are virally-suppressed and therefore non-infectious. This is
probably a sufficient explanation for why HIV diagnosis rates are falling in
all areas and nearly all groups, though it should be noted that it has taken a
big increase in consistent testing in gay men, undoubtedly due in part to the availability of PrEP,
as well as the maintenance of extraordinarily high levels of successful
treatment, that have led to this very welcome situation.

PHE comments: “Combination prevention is working: the
decline is driven by large increases in HIV tests among gay and bisexual men
attending sexual health clinics (from 37,224 in 2007 to 143,560 in 2016), including
repeat testing in higher risk men, as well as improvements in the uptake of ART
following HIV diagnosis.”