PI Perspective #18
April 1996 View PDF En
español
The New Era in AIDS Treatment …
One Chance to Use it Right or Lose it
Recent announcements of interim results from studies of the new
protease inhibitors ritonavir (Abbott) and indinavir (Merck) signal
the most potent assault yet on HIV. One study of ritonavir showed
a 50% reduction in the death rate over 5 months in people with advanced
disease. A small study of indinavir showed that when it was used
as part of a 3-drug combination with AZT and 3TC, 90% of the people
saw their virus levels drop below detectability. All is not perfect
in this new era, however. A careful look at the data thus far, along
with accumulating knowledge of how protease resistance (and multi-drug
cross-resistance) develops, leads to a sobering conclusion. However
potent these drugs may be, they must be used with great care and
deliberation. Without carefully chosen strategies, their benefits
will be short-lived and quickly followed by the development of multi-drug
cross-resistance to most of the field of protease inhibitors. This
multi-drug resistant virus could be transmitted to others. This
situation closely resembles that of tuberculosis, where improper
use of therapies will actually make things worse by inducing development
of multi-drug resistant strains of the organism.
Unlike nucleoside analogues like AZT or ddI, resistance to one
protease inhibitor often confers resistance to many others. For
example, when a person becomes resistant to Abbott’s ritonavir,
Merck’s drug will also become relatively impotent. The same
cross-resistance is then likely to apply to Agouron’s nelfinavir.
The only slight escape comes for people who might first develop
resistance to Roche’s saquinavir; its resistance confers only
one of four mutation steps needed to develop cross-resistance to
the other protease drugs. Still, even this first step sets the necessary
conditions for, and hastens, the eventual development of multi-drug
resistance. Similarly, resistance to ritonavir or indinavir may
not automatically confer later resistance to saquinavir, but does
take some initial steps in that direction.
The risks of failing to address resistance are readily seen in
the Abbott “salvage” study which tested ritonavir in
people with advanced disease (CD4+ counts under 100). Everyone applauds
the idea of testing new drugs in this group, but the protocol design
created as many problems as it solved. The study randomized volunteers
to receive either ritonavir or a placebo, which was added to whatever
other therapy the person happened to use (or to no other therapy).
There was no strategy for dealing with the resistance question.
Some saw this as a welcome step toward a “real world”
model of drug testing, arguing that this is just how people are
likely to use the new drug. Others, however, viewed it as a sanctioned
way of using a new drug improperly or at least not optimally. Lack
of optimal use might not be much of a problem if the only impact
were on the reputation of the drug, but in this case, it had clear
health implications for the patient volunteers.
Five months into the trial, both supporters and critics saw what
they wanted. On the positive side, the group receiving the drug
had about a 50% reduction in the death rate and a similar reduction
in new AIDS-defining infections compared to those who received a
placebo. But this was only part of the story. Less discussed was
the fact that people receiving the drug saw only an average 1.6
log drop in their viral load early in the study (for more information
about understanding logs, call Project Inform at 1-800-822-7422).
After five months, their viral load was rising quickly, showing
only a 0.6 log reduction, signaling the onset of viral resistance.
The size and duration of the viral load drops were lower than those
in some other protease studies. Very likely, the study volunteers
were quickly becoming resistant to ritonavir. With longer follow-up,
the clinical and survival benefits would likely have been shown
to diminish. While people saw a short term reduction in death and
new infections, the sub-optimal “real world” protocol
design also reduced the overall potential impact of the protease
inhibitor (compared to that seen in more structured studies). Used
in this fashion, the initial benefits from a protease inhibitor
may be over in six months, with little or no prospect for future
benefit from this class of drugs. Some argue that if that’s
the “real world” outcome, then so be it since that’s
the way people will use these drugs. They maintain that at least
there was some benefit, and that this trial tells us what we can
realistically expect from a single protease inhibitor used in this
fashion, in this population. Such reasoning seems based on very
low expectations, both about the drugs and peoples’ ability
to act rationally.
Other studies suggest that far more potent and lasting results
are possible if the new drugs are used strategically in combinations.
In earlier studies of indinavir monotherapy, sub-optimal use diminished
its potency and hastened resistance. More recent studies have sought
to optimize its impact by using it in combination with a second
and third drug. Used in three-drug combinations, both the size and
duration of antiviral effect were greatly enhanced compared to the
“do whatever you want” trial of ritonavir. People receiving
two different three-drug combinations including indinavir fared
far better than those using single agents or two-drug combinations.
The same studies showed that the choice of the other drugs used
in the combinations also mattered. But in all the indinavir three-drug
combinations, extent and duration of antiviral effect were enhanced
and drug resistance was not in evidence at 6 months. Other studies
of ritonavir are also looking at specific combinations and there
is reason to expect excellent results from these.
Studies of high-dose saquinavir at Stanford University add yet
another wrinkle to the story. New data from these studies indicate
that if people are careless or sloppy in their use of these drugs;
for example, by skipping doses occasionally or reducing dosage,
the almost certain outcome is far more rapid development of resistance.
The clear conclusion is that how these drugs are used will impact
enormously on how well and long they work. Furthermore, no other
entirely new class of drugs is remotely close in the pipeline so
it is not realistic to expect, as in the days of the nucleoside
drugs, that “something else” will soon become available
when the protease drugs fail.
In short, people and their physicians must learn to use these drugs
correctly, or they will quickly be rendered useless as a class of
agents. The comparison to tuberculosis is obvious. Effective treatment
of that disease requires a carefully chosen combination of drugs,
employed with almost religious fervor, for the duration of treatment.
Inadequate combinations, low doses, or sloppy, undisciplined use
hastens the development of multi-drug resistant strains of tuberculosis.
These strains become largely untreatable in the patient, and they
can be passed on to others in all the ways TB is normally transmitted.
Health officials and scientists wouldn’t dream of sanctioning
sub-optimal use of TB therapies by shrugging and declaring it the
“real world.” Instead, they have created aggressive
education campaigns teaching patients and physicians alike about
the critical importance of using these therapies properly. Nothing
less is called for in AIDS. In this regard, it is critical that
the FDA makes sure that labeling and standard-of-care recommendations
address this issue.
Making Decisions: Optimal Use of Protease Inhibitors
While final answers on the best use of protease inhibitors will
require years of study, some hints can be seen for people who cannot,
or will not, wait. To start, there are a few things which help distinguish
“worst case” from “best case” use.
Worst case scenarios (these scenarios are least desirable and most
likely to lead to rapid multi-drug resistance):
- Using a protease inhibitor alone, especially in advanced disease.
The only time it would be reasonable to use a protease inhibitor
alone is when a person simply cannot tolerate or access other
drugs for use in combination.
- Using an inadequate dose. If adequate dosing is impractical,
as with the early formulation of saquinavir, this can be minimized
by using the best other available agents along with it in a three-drug
combination.
- Reducing the dose below the prescribed level, even temporarily.
If intolerance or side effects develop, it will be better to temporarily
stop the drug rather than reduce its dose. This runs counter to
common clinical practice with the older nucleoside drugs, in which
physicians routinely lowered dose as a way of dealing with side
effects. The fastest way to develop resistance is to use the protease
inhibitor at an inadequate or inconsistent dose level.
Sub-optimum use (these scenarios induce some risk of diminished
effectiveness and more rapid resistance):
- Adding a protease inhibitor on top of whatever else a person
is doing, without consideration of drug resistance.
- Adding a protease inhibitor to only one other drug.
- Adding a protease inhibitor to another drug or a two-drug combination
without careful selection or knowing the current effectiveness
of the other drugs used.
Best case scenarios:
- Beginning a protease inhibitor at about the same time as two
other new drugs for combination use or at least with two drugs
which haven’t been used for a long time. If all or most
other drugs have been used previously, it would be important to
use the protease inhibitors with a combination which has proven
synergy, such as AZT plus 3TC.
- Using a protease inhibitor with the most potent two or three
other drugs, all of them for the first time (prior to any opportunity
for resistance). This, however, leads to the unanswered question,
“When to Start?”
- “Pretesting for drug sensitivity” before beginning
a new combination. (a technology not commonly available, to test
drugs against the virus to insure they have activity prior to
initiating therapy).
Difficult Questions / Few Answers
A critical question, still unanswered, is when to begin using a
protease inhibitor at all. Underneath this lies the issue of when
to begin therapy of any kind. The answer might be fairly obvious
in some situations, less clear in others. Most researchers agree
that use of a protease inhibitor would be clearly indicated when:
- a person has shown evidence of disease progression;
- there is evidence that existing therapies aren’t working,
as measured by viral load, CD4+ cell decline, or clinical condition
and events;
- a person has a “high” viral load, regardless of
clinical condition or CD4+ counts (one new study reports that
even viral loads above 10,000 predict a higher risk of progression,
so the definition of “high” may need to be revised).
Increasingly, researchers are also interested in testing three-,
or four-drug combinations (including protease inhibitors), or administering
protease inhibitors in people who have recently been exposed to
HIV, such as health care workers stuck by an infected syringe. A
similar idea of “hit it hard and hit it early” is being
suggested for people known to be in or just beyond the “acute
infection” stage—a period of flu-like illness which
often occurs a month or so after initial infection with HIV. The
hope of aggressive treatment is to either stop the infection altogether,
or at least minimize its potential for long-term damage by lowering
viral load.
Lack of data makes it unclear if a similar “hit hard and
hit early” philosophy is always right for everyone:
- Should all HIV-infected people, regardless of CD4+ cell levels
or viral load, be treated immediately with the most potent multi-drug
combinations? Some scientists believe this is appropriate, at
least as a research question. Others fear even the best therapies
will not work indefinitely and using them in relatively healthy
people, or people with low viral loads might waste our best weapons
in many people before they are needed.
- In some people, will therapy with drugs like the AZT/3TC combination
be sufficient to sustain low viral load? In this case, does adding
a protease inhibitor help in any way?
- Do people with low viral levels, say under 5,000 on the PCR
test, require therapy at all? If viral load drives disease progression,
why use therapy when viral load is already low? Does using the
drug simply create the opportunity to develop viral resistance,
without adding to clinical health? Or, at the other extreme, might
using a protease combination early on make it possible to suppress
the disease for very long periods or even permanently?
- In general, should the decision to initiate therapy be based
on viral load instead of CD4+ counts, as some recent studies suggest,
or should both be taken into account?
These are difficult questions. There are many opinions but very
little hard data. There are data, however, that indicate that at
least a small subset of HIV-infected people fare very well for long
periods without therapy. Some researchers argue science must not
upset this balance in such people. But, it may be equally valid
to test therapy in such people to see whether, with a little more
help, the body might be able to rid itself permanently of any risk
of symptomatic disease. But for now, both views are equally valid
and in need of research.
One thing is certain: how these drugs are used will be critical.
Test centers are already seeing newly diagnosed patients with AZT-resistant
virus. If protease drugs are misused, it won’t be long before
we see newly infected people who begin their HIV-infection with
multi-drug, cross-resistant virus. A new wave of clinical trials
must be initiated immediately to address treatment strategy. A massive
educational campaign must begin to teach physicians and patients
all that is known about the proper use of these drugs. Approval
of these drugs should not be structured in a way that permits pharmaceutical
companies to say their work is finished and wash their hands of
further research. Nor can we permit the federal apparatus, led by
the Office of AIDS Research, to turn away in pursuit of easier,
less complex work in the field of basic science, and shy away from
the difficult task which lies ahead.
Similarly, the advent of protease inhibitors and their impact on
viral load make it more necessary than ever that viral load testing
be a routine part of the standard of care for HIV-infected people.
It is time for the debates to end and for the third-party payers
to stop whining about costs. The expense of mismanaging the treatment
of AIDS is the real cost with which they should be concerned. Viral
load testing will tell physicians when it is time to start therapy,
time to change therapy, and time to withhold it. HIV medicine cannot
be intelligently practiced without this critical tool.
Protease inhibitors may meet their potential as the largest leap
yet in AIDS treatment, or they may become only a footnote on the
way to an era of a new form of multi-drug resistant HIV infection.
The difference is up to us.
NEXT