International AIDS Society Newsletter, no. 21

gence of the virus and cannot maintain the
exponential curve of virus production. Up to
now such a strategy has not demonstrated
major benefit in chronic infection. Treatment
interruptions are followed, usually within
about three weeks, by virus relapses and
HAART reintroduction is required in most
cases within a month or two, depending on the
guidelines for therapy. Lack of restoration of a
solid immunity against HIV is currently the
major drawback limiting the efficacy of the STI
strategies and raises the need for additive therapeutic strategies.
Restoring Immune Control of HIV with
Therapeutic Immunizations
This need has prompted the development of
new therapeutic strategies capable of restoring
in treated patients solid defenses against HIV
and prolonging time off therapy. One strategy,
linking anti-HIV vaccines to HAART before
interrupting the treatments, appears to be a
safer strategy than the hazardous STI [15,161.
However, a number of questions are raised by
those immune-based strategies. First, which
immune correlates of protection against HIV
should be used to evaluate efficacy of therapeutic immunizations against HIV? Lessons from
the Long-Term Non Progressors, who spontaneously control HIV for prolonged periods of
time, and recent progresses in vaccine development in macaques firmly suggest that a strong
T-cell-mediated immunity to HIV, as mediated by cytotoxic T-lymphocytes and CD4 Thelper-1 cells, can limit virus replication and
protect against disease progression in the
absence of treatment. Whether increasing
HIV-specific antibodies will be useful in such
strategies remains unknown. The main goal of
these strategies is therefore to boost immunity
to HIV with vaccines both during HAART
and before re-exposure to HIV. This is in order
to restore strong and durable CD4 T-helper-1
cells specific for HIV and broad, intense CD8
T-cell responses to HIV that would provide a
barrier to HIV before rather than after virus
relapses. As a consequence, the virus set point
after relapse would be easier to control at lower

levels, which would allow one to prolong the
time off therapy. Such a goal is theoretically
achievable with anti-HIV vaccines.
A second question is then raised: which
immunogens are safe, allow multiple injections for chronic usage, and can offer strong
immunogenicity for CD4 and CD8 T-cells
against HIV? It is relatively easy to elicit CD4
T-cell help by using inactivated viruses, proteins, peptides or poorly replicating viruses,
since any of these compounds will be endocytosed and processed by the antigen-presenting
cells to be introduced to specific CD4-T cells.
Eliciting virus-specific CD8 T-cells requires,
however, a more stringent procedure since only
very small peptides or proteins encoded and
synthetized within the antigen-presenting
cells or the target cells will be loaded into the
major histocompatibility complex (HLA) class
I molecules and will be capable to stimulate
CD8 T-cells. Some candidate vaccines are
already available, have been tested in many
seronegative individuals and have proven to be
safe but are usually weak immunogens.
- Recombinant viral proteins such as gp,
p24, tat, rev, and nef can induce specific antibodies and CD4 T-cells. They are, however,
unable to induce CTL unless they are combined with potent adjuvants.
- Peptides (or lipopeptides) provide limited
numbers of small size antigens (usually 10 to
40 amino-acids long) that have been carefully
chosen from the most immunogenic parts of
the virus. They are potent at eliciting both specific CTL and CD4 T-cells against such a repertoire and are currently being evaluated in therapeutic immunization strategies.
- Inactivated virus particles provide the
whole set of HIV proteins (except for the envelope in Remune~) and induce specific antibodies and CD4 T-cells as described above, but
no, or very little CTL. Such a strategy has been
extensively studied as a therapeutic vaccine
during the past years and might be a potent
additive strategy to other vaccines due to its
ability to restore the help.
- Naked DNA can also provide large
regions of HIV and induce both CD4 and CD8

specific T-cells. This strategy is currently
being evaluated in Phase I and II trials in HIVpositive patients.
- Recombinant viral vectors that have been
genetically modified to incorporate various
HIV genes are today considered the firstchoice strategy for vaccines. They can induce
both CD4 and CD8 specific T-cells. A first
series belong to the family of highly attenuated
recombinant pox viruses. The pox virus infecting canaries is able to infect human cells but is
not replicative and is highly safe. This vector
has been tested so far as a prophylactic vaccine
in hundreds of seronegative individuals and as
a therapeutic vaccine in many countries. New
promising, and maybe more potent, highly
attenuated members of the cow pox family,
such as MVA or NYVAC can probably induce
stronger CD4 and CD8-mediated defenses and
will soon be tested in HIV-infected patients. A
new generation of recombinant vaccines have
also been produced by recombining defective
adenoviruses for HIV genes and appear to be
highly immunogenic in macaques and even in
humans. A strong pre-existing immunity due
to natural infection by adenoviruses in humans
might, however, somewhat limit the efficacy of
these vaccines.
The careful evaluation of these new
vaccines in an international effort of clinical
trials will provide unique information on
the efficacy of these vaccines but also raise
a third series of questions: what will be the
ideal design for clinical trials evaluating therapeutic vaccine efficacy? The schemes and
schedules of immunization will have to be
defined according to each candidate vaccine's
properties. New endpoints for therapeutic vaccine efficacy will also have to be defined that
will use, virus kinetics. CD4 counts, or simply
time off therapy. Such endpoints will have to
be adapted to the current standards of care.
(cont'd)