A group of researchers have for the first time used infusions
of antibodies to produce prolonged suppression of HIV viral load without antiretroviral
therapy (ART) in the majority of a group given them.

In the small study, nine out of eleven people with HIV, given
three infusions of two so-called broadly neutralising antibodies (bNAbs) were
able to stop their ART for an average of five months. Two of those people were
still off ART and maintaining viral undetectability (one with a few ‘blips’) at
the end of the trial, more than seven months after the last antibody infusion.

The other seven maintained undetectable viral loads for
periods of between 10 and 21 weeks (average 14) and stayed off ART for an
average of 16 weeks.

Two people, however, had viral loads that rapidly reappeared
after the last antibody infusion and almost immediately restarted ART. It
turned out that both had pre-existing resistance to one of the two antibodies
used and that this was not detected when they were originally screened for the

In a second study, the same antibodies were given to seven
people who had not been taking HIV treatment. Four of the seven responded, two to just
one dose of the antibodies. Their viral loads reduced by 2.2 logs on average
(about a 200-fold reduction) for an average of about three months. This was not
enough to produce persistent viral undetectability in three of the four but the
person who started the study with the lowest viral load (under 1000 copies/ml) maintained
a viral load under 20 copies/ml for over eight weeks.

These studies are the most successful yet in a series using bNAbs
as candidate antiretroviral drugs. The two antibodies chosen for these studies
acted as entry inhibitors, attaching themselves to and blocking different part
of the gp120 protein which HIV uses to attach itself to cells.

Such antibodies develop naturally in some people with
chronic HIV infection, but HIV easily develops resistance to them individually
and studies using single antibodies have had no effect. Nor, despite promising
studies in monkeys, have dual combinations previously worked in humans. The two antibodies studied – called 3BNC117 and 10-1074 – were
picked for their superior resistance profiles and persistence in the body.

The result is the first demonstration that dual antibody therapy
can work as a form of ART.

There are a couple of possible advantages to using bNAbs.
One is that a single dose lasts a long time. After three doses, the half-life (the
time taken for levels in the blood to fall by 50%) of 3BNC117 and 10-1074 in people
already on ART was 23.2 and 17.6 days respectively – as the effective concentration is actually less than half that of the peak concentration, this suggests that one dose
would be enough for several months. In the people not on ART and with
detectable viral loads, the half-lives were 12.7 and 12.3 days respectively. This
is shorter probably because there are more viral particles around to ‘soak up’
the antibodies.

The other possible advantage is that bNAbs may, in some
cases, act as a therapeutic vaccine. Two people in the first study were still
virally suppressed seven months after their last dose, long after effective
levels of antibody had disappeared from their blood. This raises the possibility
that in these cases the antibody infusions did not just act passively, as long-lasting
ART, but actively stimulated a systemic anti-HIV response. Or it could be that
these two people are just among the lucky minority who are able to suppress HIV for
prolonged periods of time off ART anyway. It has not yet been determined
whether antibody-enhanced immune responses were seen in these two, or any other

It is clear from the studies that we are at the beginning of
a process of discovering bNAb combinations that are potent and broad-spectrum
enough to be practically used as drugs. What made the dual combination in these
two studies fail in some people was the old enemy that made dual-combination
ART fail in the 1990s – HIV drug resistance.

The detailed writeup of this study shows that resistance to
these two antibodies was so widespread that they would be of limited benefit in a
clinical setting.

In the first study, of people already undetectable on ART, of
108 people who were screened for suitability, 54% had HIV with resistance
mutations to one or both antibodies used. In addition, four of the 15 people finally selected to be in the study had a viral load over 20 copies/ml on the
first day they received antibodies, and were excluded from the final analysis.
These four were also found to have pre-existing resistance to one or both
antibodies. This suggests up to two-thirds of people with HIV might have
pre-existing resistance to these two antibodies.

In the study of seven people not on ART, the one person who
had no response at all to the antibodies was found to have HIV resistant to
both antibodies and the two people who only responded for two or three weeks
had small amounts of pre-existing resistance.

These resistance mutations strengthened during the course of
the study, and resistance to the antibodies also arose in the course of treatment
in the on-ART study. All the seven people in this study who were able to stay
off ART for several months, but eventually had to resume it, developed some
degree of resistance to one of the two antibodies. This was because levels of 3BNC117
fell faster than those of 10-1074, so for a while people were effectively on 10-1074
monotherapy – to which they developed resistance.

These studies therefore should be seen as a
proof-of-concept: bNAbs certainly can work as long-acting antiretroviral drugs
when used in combination though not, as with other ARVs, singly. But we may
need other bNAbs to which HIV finds it more difficult to develop resistance,
and – as with antiretroviral drugs – they may need to be given in combinations of three or
more to work.