HIV specialists and network security experts at UCL noticed that the
spread of HIV through the body using two methods – via the bloodstream and
directly between cells – was similar to how some computer worms spread through
both the internet and local networks respectively to infect as many computers
as possible. They worked together to create a model for this ‘hybrid
spreading’, which accurately predicted patients’ progression from HIV to AIDS in
a major clinical trial.
Detailed sample data from 17 HIV patients from London were used to verify
the model, showing that hybrid spreading provides the best explanation for HIV
progression and highlighting the benefits of early treatment.
HIV infects CD4+ T-cells, which play a vital role in the immune system and
protect us from diseases. As HIV progresses, it reduces the number of active
T-cells in the body until the immune system cannot function correctly, a state
known as ‘acquired immune deficiency syndrome’ or AIDS.
Current World Health Organisation guidelines, which the UK government
follows, recommend only beginning HIV treatment when the number of T-cells in
the bloodstream falls below a certain level. However, the new model predicts
that treatment should start as soon as possible after infection to prevent AIDS
from developing in the long term.
"With this new model, we should be able to assess the effectiveness of drugs against different modes of HIV spread in real patients" Dr Clare Jolly
“The number of HIV cells in the bloodstream is always relatively low, and
our model shows that HIV spread through the bloodstream alone would not be
enough to cause AIDS,” explains co-senior author Professor Benny Chain (UCL
Infection & Immunity). “It is likely that when HIV gains a foothold
somewhere with a high T-cell population, such as the gut, it uses a
cell-to-cell transfer mechanism to efficiently spread directly between them. As
such, if HIV has already spread to an area rich in T-cells by the time
treatment begins, preventing its spread through the bloodstream will not stop
AIDS. Our model suggests that completely blocking cell-to-cell transfer would
prevent progression to AIDS, highlighting the need to develop new treatments.”
The model was inspired by similarities between HIV and computer worms such
as the highly damaging ‘Conficker’ worm, first detected in 2008, which has
infected military and police computer networks across Europe and is still
active today.
“HIV and Conficker have a lot in common,” says lead author Changwang
Zhang (UCL Computer Science). “They both use hybrid spreading mechanisms,
persist for a very long time and are incredibly difficult to eradicate. Our
model enables us to explain these important properties and to predict the infection
process.”
Changwang’s supervisor, co-author Dr Shi Zhou (UCL Computer Science)
says: “Although the cybersecurity community organised an unprecedented
collaboration to tackle Conficker, they still failed to eliminate Conficker
from the Internet. HIV researchers face a similar problem. We hope that our new
understanding of hybrid epidemics will help us to fight against Conficker and
HIV.”
Laboratory research led by co-senior author Dr Clare Jolly (UCL Infection
& Immunity) has previously shown that some drugs are better than others at
stopping HIV from spreading directly between cells. However, it is not possible
to directly measure cell-to-cell spread in patients because it takes place
inside internal organs.
“With this new model, we should be able to assess the effectiveness of drugs
against different modes of HIV spread in real patients,” explains Dr Jolly. “This
could prove invaluable when interpreting the results of drug trials to
understand what works and why. Using computer models to understand processes
that we cannot directly observe is common in the physical sciences and supports
many fundamental theories. Our model provides strong evidence that cell-to-cell
spread is an important part of HIV spread, and we hope to show this directly in
future animal studies.”