AIDS Pathogenesis -- New Understanding

While there was new information at the Tenth International
Conference on AIDS, August 7-12 in Yokohama, that meeting
mainly served to summarize the current state of scientific
knowledge about many aspects of HIV disease -- areas of
evolving consensus, and also continuing disagreement among
experts. In pathogenesis -- the study of the mechanisms by
which disease develops -- there are still vast unknowns, gaps
in knowledge filled only by guesses. For example, it is known
that relatively few T-helper cells are killed directly by HIV
-- and no one knows what is causing the disappearance of most
of the cells, although there are several theories.

Despite the unknowns, there is more clarity today about HIV
pathogenesis than is commonly realized. The developing
knowledge will be useful for understanding the limitations of
current treatments, and the requirements for better ones.
This article looks at part of the evolving picture of HIV
disease, focusing on aspects highlighted at the conference.

** Latent Infection Surprisingly Large

In a major plenary address, Dr. Ashley T. Haase of the
University of Minnesota reported finding an "extraordinarily
large" number of cells in lymph nodes and some other tissues
infected by HIV early in the disease process. For example, 25
percent or more of CD4 cells in the lymph nodes were found to
be infected; but only about one percent of those were
actively producing virus at any one time. The others were
latently infected; HIV had entered the cell and become part
of its genetic inheritance, but because the cell was not
activated, the HIV was not reproducing or creating abnormal
proteins. These latently infected cells are undetectable by
the immune system, since the HIV genetic information remains
quietly inside the nucleus of the cell and does not affect
the cell surface; for the same reason, this latent infection
is undetectable by most laboratory methods. Dr. Haase's team
used a special technique called in situ PCR; this causes PCR
(polymerase chain reaction, which detects a particular DNA
sequence by causing it to successively duplicate itself) to
occur within each cell, allowing latently infected cells to
be distinguished from uninfected ones in the laboratory.

This latent infection, a "Trojan horse," may not cause damage
while it remains latent; but at any time some of the cells
will become activated. Then more virus is produced, spreading
the infection and/or killing the cell. The latently infected
cells form a reservoir of virus which is not affected by the
immune system, or by AZT or similar drugs; this reservoir
enables the virus to persist, and may also be involved in
transmission of HIV from person to person.

This view of HIV infection was not new at Yokohama, but grew
out of several years of work at the Department of
Microbiology of the University of Minnesota Medical School in
Minneapolis; a major paper was published in 1993. (1)

[Comment: This information does suggest approaches to
treatment, even though there are no known drugs which can
selectively attack the latently infected cells, or the HIV
genetic material within them. AZT, etc., and also the
protease inhibitors, or other drugs which target various
enzymes in the HIV life cycle, can help to prevent the
infection from spreading into new cells; protease inhibitors
might be better than AZT-type drugs for doing this, because
they cause the newly-produced virus to be defective and non-
infectious, while AZT, etc., only help block infection of new
cells after it is already underway.

These antivirals can be combined with other treatment
approaches which may reduce the amount of activation.
Aggressive diagnosis and treatment of opportunistic or other
infections can help, since infections increase the activation
of immune cells.

Activation could also be reduced by various kinds of drugs.
One example is immune suppressive therapies, which can be
dangerous, but might be used in some cases, for example in
early treatment, before serious immune deficiency has
developed; or treatments might be used to reduce particular
immune responses which are overactive. Various drugs are
being tried to reduce abnormally high levels of TNF (tumor
necrosis factor), which increase activation of HIV. Still
another approach is to screen for drugs which inhibit the LTR
(long terminal repeat) of HIV; it is now believed that
certain CD8 cells inhibit HIV quite effectively by producing
a substance (so far unidentified) which inhibits the LTR.
Drugs to reduce activation are not likely to provide the
whole answer, since some activation of immune-system cells is
necessary for normal functioning; but reducing activation can
be one of a number of approaches toward helping the immune
system maintain or re-establish control of the HIV
infection.]

** Major Lymph-Node Study Examines AZT, ddI Effects

DATRI-003 (run by the Division of AIDS Treatment Research
Initiative, of the U.S. National Institute of Allergy and
Infectious Diseases), is the largest study to date using
lymph-node biopsies; 32 patients had two biopsies, eight
weeks apart, and some of them started or changed treatment
after the first biopsy. This study is important because HIV
infection is especially active in lymph nodes. But some of
the preliminary results are confusing or hard to interpret;
we will know more when a full report is available.

All 32 patients had T-helper counts above 250, often much
higher. They were divided into four groups. In group 1,
patients had not had any antiretroviral therapy, and AZT was
started. In group 2, there was no prior therapy, and none was
started. In group 3, patients were already on AZT, and it was
continued. In group 4, they were already on AZT, and ddI was
added.

There was essentially no change in the percentage of infected
cells in the lymph nodes of the different groups. (This
result would be expected, as there is no reason to think that
AZT or ddI could eradicate infection in cells where it was
already established.)

Viral replication did decrease in the six patients who added
ddI to their treatment (group 4); this was seen by measuring
HIV RNA in both the lymph nodes and in the blood, although
the decrease in the lymph nodes, which was seen in four of
the six, was not statistically significant. Surprisingly,
however, no effect was seen on viral replication in the lymph
nodes in the group which started AZT -- although their T-
helper count did rise.

One possible explanation of this difference is that those who
started AZT began with a high T-helper count (mean 654),
while those who added ddI begin with a mean of 394; the
authors suggested that the antiviral effect of these drugs
may be less in early disease. However, we would like to see
the full report (which is currently being prepared for
publication), since viral levels can be hard to measure in
early HIV infection, even with the new tests. Is it clear
that AZT did not reduce viral replication, or could a
reduction have been missed because there is less to measure
at that early disease stage, and there were only a few
patients in this study arm? It would be a surprise if AZT
does not affect viral replication in the lymph nodes in early
HIV disease; therefore, we want to see more information
before reaching that conclusion.

There has also been some confusion about this study due to
the current lack of standardization of terminology. In the
Yokohama conference abstract, the authors used the term "HIV
burden" to mean the proportion of infected cells. The authors
plan to avoid this terminology in the future, because "viral
burden" is sometimes used to mean the number of copies of
viral RNA in blood plasma; "viral load", "viral activity,"
"viral reproduction" have also been used, more or less
interchangeably, for the latter meaning. No one knows how
this usage will standardize in the future; our guess is that
"proportion of infected cells" will refer to the former
meaning, while either "viral activity" (or perhaps "viral
load" or "viral burden") will be used for the latter. The two
meanings are quite different, since the proportion of
infected cells apparently becomes established early in HIV
disease and then tends changes only slowly, while the number
of copies of viral RNA in blood plasma can change greatly
within days, due to changes in drug treatment or for other
reasons. Because of this rapid response, and because the
proportion of infected cells is difficult to determine, while
tests to measure plasma HIV RNA have now become routinely
available (see AIDS Treatment News #204, August 5, 1994), and
because HIV RNA level indicates the number of viral particles
and correlates with the ability to grow the virus in culture,
the plasma RNA level is becoming the measurement of interest
for testing new drugs, and for individualizing therapy with
existing drugs.

[Note: There was much more information on pathogenesis at the
Yokohama conference. We will continue our coverage in future
articles.]

References

1. Embretson J, Zupancic M, Ribas JL, Burke A, Racz P,
Tenner-Racz K, and Haase AT. Massive covert infection of
helper T lymphocytes and macrophages by HIV during the
incubation period of AIDS. Nature. March 25, 1993; volume
362, number 6418, pages 359-362; comment on pages 292-293.

2. Cohen OJ, Pantaleo G, Graziosi C, Niu M, and Fauci AS.
Effect of antiretroviral therapy on HIV burden and
replication in lymphoid tissue. Tenth International
Conference on AIDS, Yokohama, August 7-12, 1994 [abstract
#001B).