AZT, Antiviral Combinations, and Resistance

Most of the clinically useful information from the Sixth
International Conference on AIDS concerned either fine-tuning of
the use of drugs that are already approved and in standard use
(e.g., AZT, alpha-interferon, and various prophylaxes for
pneumocystis), or preliminary suggestions on how to use drugs
that are not yet FDA-approved but are likely to become more
broadly available in the coming months and years (e.g., the
anti-HIV drugs ddC and ddI, and the anti-CMV drug foscarnet).
Much of the remaining information was more useful to researchers,
doctors who are also involved in designing clinical trials, and
policy makers and activists who are trying to implement
increasingly efficient and attractive ways to test potential
treatments.

The major focus of this article will be on information which
people can use today in making treatment choices. We will also
discuss some of the important information on the use of
treatments and combinations of treatments which may be available
in the foreseeable future.



AZT: When and How Much?

Much discussion concerned data already thoroughly critiqued
in the past several months about when to start AZT and how much
to use. The latest official recommendations suggest 600 mg per
day for persons with fewer that 500 T-helper cells. Most
physicians now use either 500 or 600 mg.

What about AZT for people with more than 500 T-helper cells?
Margaret Fischl, M. D., a leading researcher on AZT, reported
that the large government-sponsored study with asymptomatic
patients has not at this time found a statistically significant
difference in disease progression between patients with more than
500 T-helper cells who took AZT and those who took a placebo.
However she also noted that this data is inconclusive, and does
not show whether or not AZT would be helpful for these people.

At this time very few early studies are designed to test the
effects of drugs in asymptomatic patients. Pharmaceutical
companies may avoid studies in groups of people where the effect
may not be big enough to earn FDA approval of their drug. But
answers for these patients will become increasingly important to
those who are wondering when they should start using an anti-
retroviral, or prophylaxis for specific opportunistic infections.

One very small study suggested that 300 mg of AZT per day
may be effective. This study was designed to test the effect of
combining high dose acyclovir with AZT. Only 22 patients were
enrolled; approximately half took 600 mg of AZT and the others
took 300 mg. Although the addition of acyclovir in half of the
patients on each dose did not produce differences in tests used
to assess viral activity (p24 antigen and plasma viremia), the
300 mg dose appeared as effective in lowering these measures, and
perhaps even more effective in sustaining an increase in T-
helper cell counts, as the 600 mg dose.

This small study is far from conclusive, but it does suggest
that people who have to lower their AZT dosage because of side
effects or the use of other drugs which depress blood cell counts
may still benefit. It also supports the growing suspicion that
we have not yet found the minimum effective dose of AZT.

This confusion about dosage has implications for how early
tests of anti-viral drugs in humans are designed and interpreted.
Currently, the standard approach is to look for what is called
the "maximum tolerated dose." As we have seen with the example
of AZT, this dose was quite toxic. Yet early studies continue to
look for the maximum dose, and then later studies use that dose,
because it is the best studied. Treatment activists have long
pointed out this flaw in drug development, and have suggested
using the infectious-disease approach of seeking the minimum
effective dose, instead of the cancer-chemotherapy approach of
maximum dose. But pharmaceutical companies have an incentive to
use larger doses, because if their drug fails to show efficacy in
a controlled trial, they may not get another chance to test it,
and their drug will be dead; yet they can reduce the dose later
if necessary to manage toxicity.



AZT in Combination with Approved Drugs

* Acyclovir: It has long been believed that high doses of
acyclovir may increase the effectiveness of AZT. One small study
reported on AZT plus high dose acyclovir; it did not show any
increased anti-viral benefit from the addition of acyclovir. It
is important to remember that acyclovir is still useful in
treating herpes infections and often in preventing their
recurrence.

* Alpha Interferon: Alpha interferon is currently approved
for the treatment of Kaposi's sarcoma in some people with AIDS.
Some scientists and doctors believe that it might increase the
effectiveness of AZT. Preliminary data from an ongoing study on
the use of alpha interferon with AZT (600 mg) in patients with
more than 500 T-helper cells does not yet answer this question.
Unfortunately, more subjective and laboratory toxicities have
occurred in the combination group than in patients on either drug
alone. These side effects included fatigue, loss of appetite,
nausea, elevated liver enzymes, and decreased red and white blood
cell counts. However, so far the numbers of people who have
withdrawn from the three parts of the study are the same
(suggesting that there was no great difference in side effects
severe enough to require stopping treatment), and dose
modifications were required in all three parts to reach safe and
comfortable doses. Therefore, if this combination is found to be
effective, it may be possible to tailor doses to individual
patients to manage side effects.

After one year there were no significant differences in
changes in T-helper cell counts between the group taking 600 mg
AZT with alpha interferon, and the group taking 1200 mg of AZT
alone. About 10 percent of the patients were p24 antigen
positive when they entered the study and all experienced a
decrease in this lab value, which may be a marker of disease
progression. Although the numbers are very small (a total of
only 10 patients), there appeared to be a more sustained decrease
in p24 antigen levels in the combination group than in the AZT-
only group. It has been difficult to complete enrollment in the
study; therefore the data is incomplete and not entirely useful.

[Note: readers should be careful in interpreting statements
that "no statistically significant difference" has been found in
an incomplete study; this does not mean no difference has been
found. There may be a big difference -- but just not enough
patients yet to provide statistical proof. Also, the criteria
for claiming statistical significance for an incomplete study --
and terminating the study as a result -- are extremely
conservative, often several times as severe as the criteria for
claiming statistical significance at the end of the study; for
background on this statistical quirk, see AIDS TREATMENT NEWS
#81, June 16, 1989. In addition, researchers may be reluctant to
give out any early information about a trial which is continuing
(unless those results are so extreme that they force the trial to
be stopped), because early information could cause patients to
drop out of the arms which seem less effective, which could bias
or otherwise damage the study.]



AZT in Combination with Experimental Drugs

The only information on the combination of AZT with an
experimental drug that was presented at the Conference was on
ddC, which works in the same way as AZT as an inhibitor of the
enzyme reverse transcriptase, but is different enough that it has
different side effects. The idea of combining antiretroviral
treatments has three main attractions. First, combining or
alternating treatments may make it possible to manage the toxic
side effects of each treatment better than with a single
treatment. Second, the problem of drug-resistant strains of HIV
may be reduced by using more than one drug with a similar
mechanism, like the nucleoside analogs AZT, ddC and ddI (more on
resistance below). Third, by combining drugs that block
different parts of the viral life cycle, it should be possible to
limit HIV replication more completely than we have been able to
accomplish with a single anti-viral drug alone.

The consensus at the Conference was that drug combination
(either using the drugs at the same time, or in alternation) is
the wave of the future. As more antivirals become available to
patients through expanded access programs and eventually through
FDA approval, the questions which need to be answered today are
how to use these drugs together safely and effectively.

The ddC plus AZT studies presented at the Conference show
us more about how complex the problem of interpreting and
comparing studies has become than about how best to use these
drugs together. None of the three studies used the same dose
combinations, thus the different ways of combining these drugs
cannot be compared across the studies. Two of the studies tested
alternating the two drugs every week or month, or using each one
intermittently (with a period of no treatment), one in patients
who were intolerant to AZT and one in patients who were not
intolerant to AZT when they entered the study. The third tested
combinations of the drug used concurrently. Each study compared
various dose combinations.

The worst side effect of ddC is peripheral neuropathy, while
the primary side effects of AZT are on the bone marrow, leading
to depressed blood cell counts. So the challenge of studying
combinations or alternations of these drugs is to minimize
peripheral neuropathy and red and white blood cell count
depression while maximizing increases in T-helper cell counts and
other markers of anti-viral activity.

Unfortunately, the major study showed that the most
effective dose of ddC (high) was also the most toxic. This
toxicity seems to be decreased by alternating AZT and ddC on a
monthly rather than weekly schedule. Thus, the conclusion of the
researchers who presented this data is that the optimum dosing
schedule is monthly alternating high dose ddC with AZT. Although
that is one conclusion which could be reached, some people may
prefer to continue to look for more desirable combinations of
these two drugs. It is important to note that this study used
1200 mg of AZT.

In patients who were intolerant of AZT, there was more
depression of blood cell counts in the monthly than weekly
alternating dose group; the least blood-count toxicity was in the
group that used ddC on an intermittent basis without AZT at all.
The researchers who conducted this study concluded that monthly
alternating low dose ddC with AZT may be an acceptable "salvage
therapy" for patients who have failed AZT. Again, some people
would probably rather try other combinations, including longer
periods on ddC than AZT, and lower doses of AZT, before reaching
a conclusion about optimum dosing of these drugs in people who
are intolerant of AZT.

Finally, the study which tested the use of AZT and ddC at
the same time found no statistically significant differences in
either safety or effectiveness between the various combinations
tested -- at least in the partial data available at this time.
It used lower dosages of both AZT and ddC than did the two
studies discussed above. Because this study is not completed,
the data did not yet provide much useful information. It will be
impossible to compare this study directly with the other two
because of differences in the dosages used.

How can drug combinations be tested better in the future?
One approach, called response-surface methodology, was developed
in chemical engineering, where it is often necessary to vary the
concentrations of two or more chemicals to find the specific
combination which works best. In this kind of trial, a matrix of
many different dosage combinations of two drugs is tested
simultaneously, but very few patients are needed for each "cell"
-- each specific dose combination -- perhaps only three or four
patients, far too few for an arm of a standard drug trial. The
reason so few are needed is that the overall pattern which
emerges shows the best combination, although each individual cell
may be too small to provide statistical information by itself. As
far as we know, response-surface methodology has not yet been
used in AIDS research, although it is being discussed. It offers
the advantage of testing all reasonable dosage combinations at
the same time and for about the same expense and number of
patients as a standard trial which only compares a few dosage
combinations.



HIV Resistance to Anti-Viral Treatments

Antiviral drug resistance was a significant topic at the
Conference. Because one of the key enzymes involved in the
replication of HIV, reverse transcriptase, is quite error prone,
HIV mutates frequently. Some of the strains that are formed as a
result of the mutations are resistant to antiviral drugs like
AZT. AZT-resistant strains of the virus can be found in patients
who have never taken AZT, but are most frequently isolated from
patients who have been taking AZT for more than 6 months. AZT
resistance appears to increase with the amount of time one has
taken the drug, although some scientists believe that resistance
may develop more slowly when antiretrovirals are started earlier
in HIV infection. Fortunately, cross resistance does not seem to
be a problem with the nucleoside analogs that are being tested
today (AZT, ddC, ddI). Therefore, even if AZT has stopped
working because viral resistance has developed, ddI or ddC should
still work.

The clinical importance of AZT resistance has not been
proven. But some scientists believe that all future clinical
trials of antiretroviral drugs should include laboratory tests of
resistance. Managing the ability of HIV to mutate into resistant
forms is one of the reasons that many researchers believe that
combination anti-retroviral therapies will be so important.

Other viruses that affect people with HIV infection also
form drug resistant strains. For example, herpes viruses can
become resistant to acyclovir; the good news is that they often
still respond to the experimental anti-viral drug, foscarnet.
Also, CMV (cytomegalovirus), which often affects the retina of
the eye but can also cause problems in the colon, spleen and
other organs, has been shown to develop resistance to
ganciclovir, the only approved therapy for its treatment. The
resistance issue is one of the crucial reasons that test tube and
human trials of new antivirals for virally induced opportunistic
infections are in great need of research attention and money.



For More Information

More information is available in the following Conference
audio tapes. See "Obtaining Conference Information" in
Announcements, below, for ordering information. Note that each
of these sessions is recorded on two cassettes.

Talk by Margaret Fischl, Thursday plenary session. Tapes
#90ICA-10 and 90ICA-11.

New Concepts Regarding the Use of Zidovudine (AZT). Tapes
#90ICA-46 and 90ICA-47.

Clinical Trials with New Antiretrovirals. Tapes #90ICA-50 and
90ICA-51.

CMV and Other Viral Opportunistic Infections. Tapes #90ICA-60
and 90ICA-61.

Resistance to Antiretrovirals. Tapes #90ICA-66 and 90ICA-67.

In addition, information from the following abstracts was used in
this article:

On viral resistance: S. B. 81 through S. B. 88.

Foscarnet for acyclovir-resistant herpes: Th.B. 446 and Th.B.
447

Ganciclovir resistance in CMV retinitis: F. B. 93.

AZT plus alpha interferon: Th.B. 22.

ddC studies: Th.B. 23, S. B. 425, S. B. 426.

AZT plus acyclovir: Th.B. 24.