ABSTRACT
The overarching aim of this grant proposal is to define and understand the conditions that create the best
chance for cure or remission in HIV infection. There is a strong rationale for believing that in utero (IU)
infection may provide the most promising starting point, because of the potential to initiate antiretroviral
therapy (ART) within hours of birth, and because the immunotolerant environment in early life may reduce
the ability of HIV to establish large viral reservoirs early in the course of infection. Aim 1 focuses on
mechanisms of minimising the size of the viral reservoir in paediatric infection. Aim 2 addresses the ability of
the effector arm of the paediatric immune system to eradicate HIV-infected cells as part of shock-and-kill
strategies. Finally, to tackle the issue of how to identify children who will and who will not be suitable for
ART treatment interruption, Aim 3 seeks to identify predictors of outcome post-treatment interruption.
To approach these aims, we will study the cohorts of HIV-infected children we have generated over the past
two decades of collaborations in South Africa. This includes two cohorts of HIV-infected children in whom
ART was initiated in the neonatal period, one a historical cohort from 2002-2005, the other a current cohort
in whom ART is initiated within 48hrs of birth. In addition we have identified in South Africa ~300 ART-naïve
children aged >5yrs who have maintained normal health and normal-for-age CD4 counts (>750 cells/mm3),
whom we have termed `paediatric non-progressors' or `PNP'. These PNP represent approximately 5-10% of
HIV-infected children. Some of these ART-naïve children we have followed throughout childhood (ie from 0-
10yrs of age) and beyond, and these children still attend our clinics with their mothers today. These cohorts
of children and mother-child pairs together provide us with unique opportunities to address our study aims
that are fundamental to understanding how to achieve the best chance of HIV cure or remission.
In Aim 1, Mechanisms of minimising the size of the viral reservoir in paediatric infection, we will
determine the impact on the size of viral reservoir of ART initiation at 1-2 weeks of age – the current
standard of care in South Africa – versus initiation within the first 1-2 days of life. In this sub-aim, we will
build upon the infrastructure already established through our Wellcome Trust funded study started in 2015 in
South Africa designed to establish the feasibility of point-of-care testing to diagnose in utero infection and
initiated ART within the first 48hrs of life. We will also examine the infuence of viral factors such as viral
replicative capacity - shown to be important in adults - on reservoir size, taking advantage of the opportunity
to study transmitted virus within hours of birth. Finally we will test the hypothesis that paediatric non-
progressors (defined above), in whom we have shown low levels of HIV infection in the long-lived Tscm and
Tcm CD4 T-cell subsets compared to the Tem compartment, will show rapid and substantial decline in size
of viral reservoirs on ART, as a result of the cellular localisation of HIV infection.
In Aim 2, Mechanisms of optimal effector function for reservoir eradication, we argue that HIV-specific
cytotoxic T lymphocytes (CTL) are likely to play an important role in eliminating HIV-infected cells that
comprise the viral reservoir; however, control of viraemia is highly unusual in ART-naïve paediatric infection.
Among ~300 ART-naïve paediatric non-progressors, we have identified <10% who have either reached
undetectable viral loads (<20 copies/ml), or stable viral loads of <1000 c/ml. In contrast to adult elite
controllers, control of viraemia in paediatric infection arises only after several years of infection, it is usually
transient, and it appears unrelated to expression of protective HLA molecule such as HLA-
B*57/58:01/81:01. Finally, unlike adults, HIV-infected children typically can generate CTL responses against
CTL escape variants: potentilly this is the ultimate solution to HIV Study of these viraemic controllers
longitudinally, in some cases from infancy, and comparison with viraemic non-controller children, will provide
critical insights into the mechanisms of viraemic control in HIV-infected children.
In Aim 3, Biomarkers predicting outcome post-treatment interruption, we propose the hypothesis that
factors contributing to low viral reservoirs in the key, long-lived T cell subsets (Tscm and Tcm) also mediate
HIV non-progression. In this aim we will first seek to identify features distinguishing progression from non-
progression in longitudinally tracked children. In the second sub-aim, we will test the ability of these markers
to predict outcome in a historical treatment interruption cohort of infants in whom ART was initiated at 4
weeks' age and interrupted after 12 months. Post-treatment interruption, 40% of infants progressed rapidly
to restart ART within a median of 0.23yrs; in contrast, another 40% initiated ART >5yrs later.