Tat Drug Development: Current Status by John S. James

Many AIDS experts agree that one of the most promising treatment
research areas at this time is the development of drugs to inhibit the
HIV regulatory protein called tat. At least until recently, however,
research has moved slowly. This article will outline the current status
of development of tat inhibitors, including results of the only human
trials at this time, and plans for future trials.
Tat inhibitors are promising for several reasons:
(1) The tat protein, which is only produced by HIV, greatly
increases the replication of HIV. It is essential for its replication;
without tat, HIV cannot reproduce.
(2) Unlike current anti-HIV drugs, tat inhibitors are likely to be
effective in chronically infected cells. All the antiretrovirals in use
now only work to prevent infection of new cells; in cells that are
already infected, they do nothing. Chronically infected cells, which are
not killed by the infection, can continue to produce toxic substances,
which may be a major cause of the immune dysregulation found in AIDS.
(The approved drug alpha interferon, and "alternative" antioxidant
treatments such as NAC, might have some effect on chronically infected
cells. But as AIDS treatments, these are limited at best.) Tat
inhibitors are likely to be effective in chronically infected as well as
other cells.
(3) Viral resistance is less likely to occur with tat inhibitors
than with the anti-HIV drugs now in use. AIDS TREATMENT NEWS previously
reported that no viral resistance to a tat inhibitor was found after a
year's laboratory test; by contrast, resistance developed rapidly to
other drugs tested (issue #141, December 20, 1991).
There is also a theoretical reason to suspect that resistance may
be much less of a problem with tat inhibitors than with other anti-HIV
drugs in use. Chronically infected cells can continue to produce new
viruses, despite the AZT, ddI, or ddC in the body. Those drugs may
prevent most of these viruses from infecting new cells. But uninhibited
production of new virus is likely to create excellent opportunities for
mutations to produce new viral strains which are resistant to AZT or the
other drugs. These new strains would have a selective advantage because
only they could reproduce if the drug treatment is continued. Therefore,
viral resistance to AZT and to other reverse- transcriptase inhibitors
is likely to be a problemQthe very problem which destroyed most of the
hopes last year for a class of non-nucleoside reverse-transcriptase
inhibitors such as Merck's L-661.
A tat inhibitor shuts down HIV replication, as well as other HIV
activity, in these chronically infected cells. With few new viral
particles being produced, there is less chance for resistant mutants to
arise.
(4) A practical reason for interest in tat inhibitors is that one
has already been tested in people (see below), in two small trials with
good toxicity and blood-level results so far (no efficacy indicators
were measured, so there is still no data on whether this drug will
actually work in people).
(5) Another practical consideration is that tat inhibitors work
against the virus in a completely different way than any other drug now
in use. Therefore, the first tat inhibitor which proves useful as a
human drug is likely to revolutionize treatment by greatly increasing
the number of therapeutic options available. Combination treatments,
for example, usually work best when the drugs being combined have
different mechanisms of action.
(6) The tat inhibitor currently in human studies is taken orally,
making it more practical to administer on a long-term basis than drugs
which need to be injected.
(7) Besides its anti-HIV use, a tat inhibitor might possibly be
useful against PML (progressive multifocal leukoencephalopathy). The
reason for this is that the tat protein might increase the replication
of the JC virus, which is suspected to cause PML.
Tat-Inhibitor Development
The technology which made tat inhibitors possibleQa fairly simple
assay which could be used to screen thousands of drugs or other
chemicals for anti-tat activityQwas developed independently by a number
of pharmaceutical companies.
We have heard that several companies have now found or synthesized
anti-tat substances for possible development as drugs. But
pharmaceutical companies are traditionally secretive about pre-clinical
drug development, and the only company which we know for sure is
developing a tat drug is Hoffmann-La Roche, which has completed two
small human trials at Johns Hopkins University.
We will discuss this drug below. But note that what is important
is this class of treatments, not necessarily any one particular drug.
Different tat inhibitors can be entirely unrelated to each
otherQtoxicity, bioavailability, dosage and scheduling, chemical
structure, etc. may have nothing in common. What is important is the
technology for finding these drugs. It is likely that anyone with
appropriate laboratory facilities could search for new anti-tat drugs.
What might make the most sense would be to test thousands of existing
drugs, natural products, and other chemicals already in human use, to
see if there is a tat inhibitor which is already available, which might
then be used almost immediately, after minimal human testing. This is
unlikely to be done, however, because pharmaceutical companies are
unlikely to test other companies' drugs; and academic, government, and
community- based groups have not stepped forward.
Hoffmann-La Roche RO-24-7429
The first tat inhibitor in human testing happens to be a
benzodiazepineQa member of a class of chemicals which includes certain
tranquilizers, including Valium. (Valium and other tranquilizers were
tested and found not to have anti- tat activity.) The reason a
benzodiazepine was found first is that Hoffmann-La Roche, which
developed one of the first assays for tat inhibitors, applied that
technology to substances available in the Roche "library" of
chemicalsQmany of them leftovers from the development of Valium. The
first tat inhibitor, RO-5-3335, was found to be too toxic in animal
tests, due to kidney toxicity; therefore a variant of it, code-named RO-
24-7429, is now being developed.
The first human study tested two doses in a handful of HIV-
positive volunteers. Only a single dose was given to any one patient in
this study, in order to measure levels of the drug in the blood. The
purpose was to design a regimen to achieve a target blood level of one
micromolar, which was found to inhibit HIV at least 90 percent in
laboratory tests. (What matters is the level of the drug inside the
cells, rather than in the blood; however, the researchers had no way to
measure the level in cells.) The target level was easily reached and
maintained for at least four hours with an oral dose of 200 mg. No
serious toxicity was found.
The second study was placebo controlled, with one third of the
volunteers getting a placebo. The drug was given every six hours for
five and a half days. With a total dose of 200 mg per day (divided into
the four smaller daily doses), the blood concentration was maintained
above the 1 micromolar target in most of the six patients. With a dose
of 600 mg per day, most of the six volunteers were ten times the target
level or more (this is the "trough" level, the lowest drug level reached
between doses). There were no side effects that were different in those
receiving the drug than in those receiving the placebo.
RO-24-7429 was found to have complex pharmacokinetics (the body's
absorption, distribution, etc. of drugs). For example, the trough
level gradually increased over time, which was not expected. And
increasing the dose does not increase the blood level proportionately,
but more than that. These findings complicate the dosing of the drug,
and also raise questions about long-term effects.
The most notable side effect of the drug is that it turns the urine
bright yellow. Chemists have discovered what causes this coloring. In
the bloodstream, three different chemicals are found after someone has
taken RO-24-7429: the drug itself, and two metabolites (one of which,
found in only small amounts, happens to be RO-5-3335). In the urine,
there are three different substances; these are hydroxylated derivatives
of the substances in the blood. Apparently it is these derivatives
which give the urine its color.
In animals, high doses of the drug cause kidney toxicity. No one
knows if this will be a problem in humans at doses which might be used
therapeutically. Kidney function will be carefully monitored in these
studies
In the first trials there was controversy over exclusion of women
of childbearing potential. Future studies will probably allow women
providing they are not pregnant or breast feeding, and will not become
pregnant in the study. (For an example of the problems resulting from
the earlier study, see "He for God Only: Why Sexism in Clinical Trials
Is Killing Women," by Teresa McGovern, QW, May 31, 1992, pages 44-49.
QW, formerly named NYQ, is a new lesbian and gay magazine published in
New York.)
Future Trial Plans
Hoffmann-La Roche recently held a meeting of a newly-formed
community advisory board, and outlined plans for future trials.
The next trial of RO-24-7429, a pharmacokinetic study on safety and
tolerance of the drug for 14 days, will be done at Johns Hopkins; it has
currently been submitted to the IRB (institutional review board) for
approval. It could be recruiting in about a month. Because of the
buildup of this drug in the body, researchers want to study it over a
longer period than tested before, in order to accurately administer it
to humans. While primarily a pharmacokinetic study, this trial will
measure at least one indicator of viral activity. Volunteers will need
an acid-dissociated p24 of over 70 in order to enter; the reason for
this requirement is so that the study can test whether the drug has
anti-HIV activity in people.
The following trial, which will be called ACTG 213, is planned for
four ACTG (AIDS Clinical Trials Group) sites: Johns Hopkins, University
of California San Diego, Case Western, and Harvard. These sites were
selected for their virology capabilities. This study will test the drug
at an earlier stage of HIV infection, with T-cell counts under 500.
Fortunately this trial will not necessarily be delayed until the new
Johns Hopkins trial is completed; these trials, and also the ones below,
can overlap.
This trial might be delayed, however, by the need to test a new
tablet form the drug to make sure it is properly absorbed. All the
previous trials have used RO-24-7429 formulated as an oral solution, but
oral solution is not a commercially viable way to deliver a drug. The
new Johns Hopkins trial will also use this solution, but then no more of
it will be made; the drug will be supplied as tablets instead. With the
complex planning required, no one knows whether the tablet or the
protocol will be ready first. Still, current plans call for much faster
development of this drug than is normally practiced, with faster
analysis, overlapping of trials, and pushing for rapid IRB approvals.
Meanwhile, or slightly later, Hoffmann-La Roche plans a separate
trial for persons with more clinically advanced disease. It will be
conducted outside the ACTG. It will not begin until about a month or
more of data from the above trials is analyzed, because of concern about
potential kidney toxicity. It, also, will use the drug in the tablet
form.
Later, a large combination trial will test RO-24-7429 with AZT,
probably within the ACTG. First, there will be a pharmacokinetic study
at Johns Hopkins to make sure there is no problem in how the drugs
interact. This sequence of trialsQearly disease, later disease, and
combinationQis modeled on the Hoffmann-La Roche trials, in Europe, of
the company's proteinase inhibitor (another class of anti-HIV drugs).
The community representatives at the Roche advisory-board meeting
were relieved that plans were to stagger the trials, instead of waiting
for each to finish before starting the next. But no time line for these
studies was given; also, there were no protocols at the meeting for
specific comments, although there are plans to let the board members
comment on the protocols later.
The main dissatisfaction of the advisory board was the lack of a
plan for a trial of RO-24-7429 for Kaposi's sarcoma. Apparently Roche
had plans for a seminar to look at this possibility, instead of for a
trial, which is what the community wants, of course. No final decision
has yet been made.
Comment and Suggestion
The plans outlined above suggest that RO-24-7429 could move more
rapidly than other drugs in the past. It is notable that a number of
trials have been thought throughQincluding coordination between ACTG and
non-ACTG trials.
We do have one suggestion, which might save much time in drug
development, as well as saving money for both Roche and the government.
There is an alternative design sometimes called the "N of one" trial, in
which a drug's effects are studied intensively in one patient, instead
of by statistical analysis of a large group receiving a standardized
protocol. The N of one design allows more flexibility in early testing
of how to use the drug in practice, since it does not require the
imposition of a rigid protocol in order to try to make different
patients comparable. Instead, the best medical and scientific judgment
of leading clinical researchers can be applied to modify the trial as it
proceeds.
Because AIDS is a public health emergency, several N of 1 trials
should be started simultaneously; they could be run by the same
researchers. This would differ from a standard statistical trial with
an N of several (too small to get useful statistics if there is much
variability in what is being measured) in that each course of treatment
could proceed individually, depending on the best interests of the
patient as well as the interests of scientific discovery. Clinical
information, not statistics, would be the goal.
There would be no shortage of volunteers, and such a trial would
pose no greater risk to them than the volunteers will be exposed to
anyway in the standard trials being planned. And consider some possible
outcomes. If most of the patients in the N of 1 trials had to stop the
drug due to toxicity, without showing signs of antiviral activity, then
it would be clear that the drug should be abandoned and another one (for
example, another tat inhibitor) tried in its place; months or years and
millions of dollars would have been saved. On the other hand, if the
drug looked good, then efforts to test it would be redoubled. The FDA
would still require standard statistical trials; but if a treatment were
promising enough, it might qualify for accelerated approval quickly,
with most of the large trials conducted post- marketing. Then the
treatment would be available to the physicians and patients who clearly
needed it, much sooner than under the current system.
And at the very least, the design of the standard trials could be
improved, based on the knowledge and experience gained from using the
drug to treat people, under the most rigorous scientific management and
observation available anywhere.