ddC: AZT Combination Approval Recommended

On April 21 the Antiviral Drug Products Advisory Committee
of the U. S. Food and Drug Administration voted to recommend
approval of anti-HIV treatment with a combination of ddC
(dideoxycytidine, also called zalcitabine; brand name HIVID) and
AZT for certain patients. It did not recommend approval of ddC
by itself. The decision was a difficult one because of the
limited information now available about this combination
treatment; the split vote, eight to three, reflected this
difficulty, since decisions of FDA advisory committees are
generally unanimous.

The FDA is not required to accept this committee's decision.
But it is likely to do so and to formally approve ddC plus AZT
combination treatment, after review of additional data, probably
during the next few months.

The Antiviral Advisory Committee is one of a number of such
committees set up by the FDA but composed of outside experts (not
FDA employees). Most of the work of these committees is public.

Background

ddC, now being developed by Hoffmann-La Roche, was one of
the earliest potential anti-HIV agents tested. It is very active
against the virus in the laboratory, and also has anti-HIV
activity in people. But because of its toxicity -- especially
peripheral neuropathy, resulting in pain or numbness in the hands
or feet -- it must be used in very small doses, less than one
one-hundredth the dose of AZT or of ddI (the two anti-HIV
treatments now approved in the U. S.). The main question with
ddC as a single-drug therapy has been whether there is a dose
which can be effective without unacceptable toxicity.

ddC was first tested in combination with AZT in a small
study called ACTG 106, funded by the AIDS Clinical Trials Group
(ACTG) of the U. S. National Institute of Allergy and Infectious
Diseases. This study was designed to try various doses to make
sure there was no unexpected toxicity when the drugs were used in
combination. No such problem was found. But the combination
treatment unexpectedly showed much greater improvement in T-
helper cell counts than that seen with any known antiviral alone
-- especially surprising since the volunteers in this study
started with very low T-helper counts (a median of about 70).

Because of the small number of patients in this study,
certain problems in how the study was designed, and the fact that
most researchers were not expecting any major benefit from this
particular drug combination since both drugs have the same
general mechanism of action, these results became a subject of
controversy. Most of the debate took place behind the scenes.
The study was not formally published for over a year; it finally
appeared in the Annals of Internal Medicine, January 1, 1992 (it
had been presented at scientific meetings during the
International Conference on AIDS in Florence, Italy, during June
1991). But the results had leaked out before then, and been
published by Project Inform and by AIDS TREATMENT NEWS in
November, 1990. As the results became known in the medical
community, several thousand people started using the combination
treatment -- mostly on the recommendation of their physicians --
obtaining the ddC from buyers' clubs, since it was not approved
and was not otherwise available for combination use.

New Data at Committee Meeting

Before the April 20-21 meeting of the Antiviral Advisory
Committee, the main issue had been whether the T-helper results
seen in this study (ACTG 106) were real. Few expected that the
committee could recommend approval of the drug based only on that
single small study. And almost no one expected ddC to be
recommended for approval as a single drug, since a major study
had shown that AZT alone led to a much greater reduction in AIDS
deaths than ddC alone.

But at the meeting new data was presented which supported
the main conclusion of ACTG 106 -- that combination treatment
with ddC plus AZT could produce larger and longer-lasting T-
helper cell rises than treatment with AZT by itself. The most
important new data was from a Burroughs-Wellcome trial which is
primarily studying drug resistance. About 50 patients are
receiving AZT alone, with a similar group getting AZT in
combination with ddC, and a third in combination with ddI. This
trial started recently; the early data now available shows a
larger T-helper rise from AZT plus ddC than from AZT alone. (The
ddI combination data was not released, as it was not relevant to
the committee's decision on what to recommend for ddC. Also, no
data was released on the traditional "endpoints" of survival or
disease progression; this trial is probably too small to show
such differences, even when it has been completed.)

Even with the new data confirming better T-helper count
improvements with the ddC plus AZT combination than with AZT
alone, the committee was left with the difficult decision of
whether to recommend approval of a somewhat toxic treatment based
only on laboratory-test data (T-helper counts), with no data to
show whether or not there was clinical improvement in patients'
survival or health. The committee was clearly willing to use T-
helper counts as part of the approval decision. But it also
wanted some showing of direct benefit to patients.

"Accelerated Approval"

Shortly before the meeting, the FDA announced a new option,
called accelerated approval, "to expedite marketing approval of
new drugs and biologics for patients with serious and life-
threatening illnesses when a drug provides a meaningful
therapeutic benefit over existing therapy." [Quotes are from an
outline distributed at a meeting of the AIDS Clinical Trials
Group a week before the Antiviral Advisory Committee meeting.]
Accelerated approval can be used "when there is evidence of a
drug or biological product's effect on a surrogate endpoint, such
as a laboratory measurement or physical sign, which reasonably
suggests that the drug is effective OR that the drug has an
effect on a clinical endpoint other than survival or irreversible
morbidity.... For drugs approved on the basis of a surrogate
endpoint, the sponsor may be required to conduct any clinical
studies necessary to determine the actual clinical benefit of the
drug on survival, disease complications, or long-term symptoms."
The drug may be withdrawn from the market if clinical benefit is
not shown, if the sponsor fails to carry out the agreed studies,
if the sponsor promotes the drug in a misleading way, or if other
evidence shows that the drug is not safe and effective. It has
been estimated that this accelerated approval system could speed
the overall development of critically important drugs by one to
three years.

Hoffmann-La Roche asked the Antiviral Advisory Committee to
consider the following language [our explanatory comments are in
brackets]:

"HIVID (zalcitabine) [other names for ddC] in combination
with Retrovir (zidovudine, Burroughs Wellcome) [i.e., AZT] is
indicated for the management of HIV-infected adult patients with
AIDS or advanced HIV disease (CD4 cell count <= 300/mm3). In two
independent studies (see description of studies), increases in
CD4 cell counts were higher and more sustained in patients
treated with the combination of HIVID and Retrovir than the
clinical experience observed with Retrovir alone."

The committee did not recommend any changes to this
statement, not because it necessarily agreed with it, but because
no data was available to support any particular change. Note,
for example, that this language does not specify whether the
combination should be approved only if other treatments have
failed, or as a first-line treatment, or both. The dilemma here
is that a new treatment is most needed when other alternatives
have failed; physicians are more likely to try the better-known
alternative of AZT alone as first-line treatment. Yet the
available data on the combination would not support such a
restriction, since it is from patients who had little or no prior
experience with AZT.

The Antiviral Advisory Committee supported accelerated
approval of the ddC plus AZT combination with the condition that
further testing be completed to prove that this treatment does
have concrete value to patients. Exactly what is to be required
of Hoffmann-La Roche will be determined by the FDA before
approval is granted.

Comment

The issues raised by ddC combination treatment have been
difficult for everyone involved. Scientists, physicians, and
regulators were naturally reluctant to approve a drug for tens of
thousands of people on the basis of one small trial (ACTG 106)
which was never designed to test the drug's efficacy. But many
patients and physicians, when they learned how good the results
looked, were unwilling to wait for another trial to be designed,
implemented, conducted, and analyzed. As a result, buyers' clubs
provided ddC to thousands of people (who obtained their AZT for
the combination treatment from conventional sources).

This situation, of a de facto standard of care accepted by
many leading AIDS physicians even though it included an
unapproved drug, appears at first glance to be a problem. In
fact it may have been the best possible solution to a deeper
structural problem, a confusion about what we as a society use
FDA approval for.

On one hand, FDA approval provides legal access to
treatments. In theory the FDA does not regulate the practice of
medicine, but rather the marketing of pharmaceuticals and other
healthcare products; in practice this distinction tends to become
moot, since if a drug cannot be sold or otherwise distributed, it
usually cannot be obtained by patients or physicians.

On the other hand, FDA marketing approval is often treated
as if it established an official standard of care. Insurers
often use the FDA-approved indications for a drug as an excuse to
deny reimbursement for uses outside of this official labeling,
even when the standard of care among leading physicians is to use
a drug for an off-label purpose. And -- more to the point on the
handling of ddC -- FDA approval is treated as a standard of care
in that it usually results in a drug being given routinely to
many thousands of people, often with little further thought or
independent evaluation.

The Antiviral Advisory Committee itself was confused on this
matter; at one point it asked for clarification on whether its
recommendation was to provide legal access to ddC for combination
treatment, or to recommend the therapy itself. No one could help
the Committee here, since this confusion is basic to the current
regulatory system. Legal access and quasi-official medical
recommendation should be separate decisions; but often they are
not.

We are reminded of one leading AIDS physician who was asked
if he prescribed combination treatment with ddI and AZT -- both
approved drugs, but not officially approved for use together. He
said that he was willing to prescribe the combination for an
informed patient who was participating in his or her own
treatment decisions, and clearly understood that the use of both
drugs together was experimental. But for those patients who
preferred not to be involved and to leave medical decisions
entirely to the doctor, he would not say, "Here, take this."

We suspect that the ddC "underground" may have been the best
solution available, because it separated access from
recommendation, which the official system is not set up to do.
With only one small study available, it would have been difficult
to say, "Here, take this" to tens of thousands of people. Yet it
would also be unacceptable to say, "You can't have this" to those
who had studied the matter and made an informed choice that, in
their situation, the combination treatment was their best option
available.

With the new data presented at the recent Antiviral Advisory
Committee hearing, the case for approval became stronger -- but
still not strong enough to make the committee comfortable, with
one member commenting that they were moving at lightning speed
(in changing how they operated), but at the same time dragging
their feet.

In the background during the Antiviral Advisory Committee's
deliberation was ACTG 155, a major trial comparing AZT alone, ddC
alone, and the combination treatment, in volunteers who have
already used AZT. This well-designed and very important trial,
with about 1,000 volunteers, will end later this year; it should
provide definitive information about the usefulness of the ddC
plus AZT combination for many patients. There was concern in the
Antiviral Advisory Committee that recommending approval of the
combination might damage this trial, by causing people to drop
out so that they could be sure of getting the combination. If
this were to happen it would cause a tragic loss of vital
information, as well as making it harder to get early approval of
drugs in the future. But the danger seems unlikely. ddC is
still not officially approved; the actual approval will probably
take weeks or months, by which time the trial will largely be
over. Also, people leave trials because of their physicians'
advice, and most physicians are reluctant to prescribe a
treatment until it is known to improve the condition of patients,
not just laboratory-test values. (A similar situation occurred
in July 1991 when ddI was recommended for approval; that
recommendation, and the subsequent approval of ddI in October
1991, did not damage the ongoing trial.)

The FDA's new procedure of accelerated approval will help to
relieve the pressure in the case of ddC, and in similar cases in
the future. It will be a limited relief, because it does not
address the fundamental problem of how to provide legal access
for those who are informed and prepared to make their own
decision, without also creating an official stamp of approval
which would send an experimental treatment to many others without
informed consent. Yet this problem will take a long time to
solve, because of the difficult issues around standard of care in
U. S. medicine -- as well as the philosophical issues about what
freedom means when growing complexity blurs the lines between
informed choice and manipulation.

The Antiviral Advisory Committee recommendation and expected
FDA approval of ddC/AZT combination treatment will put old issues
behind us, clearing the way for focusing on what is important
now. Approval will help to defuse potential conflict over
buyers' club access to ddC (as happened with other important
AIDS-related treatments which were subsequently approved,
especially clarithromycin and fluconazole, which were made
available by buyers' clubs before approval, but not afterwards).
It will let Hoffmann-La Roche focus its people and resources on
the important tat and protease drugs, instead of spending more
efforts on getting ddC approved. It will show the pharmaceutical
industry that there are advantages, not only drawbacks, to
developing urgently-needed drugs. And it will further the
development of new combination regimens, applying to AIDS the
multi-drug approaches which have proven successful against cancer
and other difficult diseases.