GMHC Treatment Issues 1993 May 1; 7(4): 1
Preliminary results from the long-awaited European trial of
"immediate" vs. "deferred" AZT treatment were released in a
letter to The Lancet on April 3, 1993. The study, dubbed
Concorde, was conducted in England, France, and Ireland. And
like the jet plane that shares its name, Concorde set off sonic
booms across the Atlantic.
Concorde began in October 1988 and was organized by the British
Medical Research Council and the French National AIDS Research
Agency. The trial was the largest randomized, double-blinded,
and placebo controlled AZT study ever conducted. Neither the
clinicians nor the participants knew who was given AZT and who
was given placebo.
In the study, 1,749 patients were followed for an average of
three years. At enrollment, all patients were HIV-positive and
asymptomatic (no AIDS or ARC) with a broad range of CD4 counts
(59 percent had less than 500, 6 percent had less than 200). A
total of 877 individuals were randomized to receive 1000mg per
day of AZT (the "immediate treatment" arm) and 872 were given
placebo (the "deferred treatment" arm). The immediate treatment
arm commenced AZT whatever their CD4 levels were, while the
deferred treatment arm was given placebo until they developed
symptomatic disease (AIDS or ARC). The characteristics of the
two groups were similar in age, sex (15 percent were women), and
Since all patients began AZT once they became symptomatic, the
trial was designed only to show whether it is beneficial to
begin AZT treatment before symptoms appear. The trial was not
designed to provide information on the usefulness of taking AZT
for symptomatic HIV disease.
In 1989, a U.S. study, ACTG 019, was halted after one year when
it found that AZT seemed to delay progression of disease in
people with less than 500 CD4 cells. The Concorde study was
modified on ethical grounds to allow participants with less than
500 CD4 cells to begin taking AZT open label. Those who chose to
begin AZT once their CD4 cells fell below 500, along with those
who began AZT because they developed symptomatic disease, were
included in the preliminary analysis published in The Lancet.
After an average follow-up of three years, the research team
reports that survival rates did not differ significantly. The
death rate was 8 percent in the immediate treatment group and 7
percent in the deferred treatment group. In addition, there was
no significant difference in disease progression. After three
years, 18 percent in both groups progressed to AIDS or death.
However, there was a statistically significant difference in CD4
levels between the two arms. After three months, the immediate
group had, on average, about 30 more CD4 cells than the deferred
group. This difference was sustained up to three years.
No new or unexpected toxicities were seen. Of those who had to
stop the treatment because of anemia or a drop in white blood
cells, sixteen were in the immediate group and three were in the
deferred group. An additional 82 in the immediate and 37 in the
deferred withdrew because of other events, mostly nausea and
The authors of the study conclude that "Concorde has not shown
any significant benefit from the immediate use of zidovudine
(AZT) compared with deferred therapy in symptom-free individuals
in terms of survival or disease progression, irrespective of
their initial CD4 counts. The discrepancy between this result
and the significant effect of immediate zidovudine on CD4 cell
counts casts doubt on the value of using changes over time in
CD4 count as a predictive measure for effects of antiviral
therapy on disease progression and survival."
CRITICISMS OF CONCORDE
The protocol change in 1989, which allowed open label use of AZT
for those with less than 500 CD4 cells, may have skewed the
results by causing a significant number of those in the deferred
arm to receive the same therapy as those in the immediate
treatment arm. Therefore, some argue, it is not surprising that
there is no significant difference between the two arms. The
Concorde researchers respond "that a valid comparison was still
possible because of the large number of patients in the study
and the marked difference in the amount of zidovudine taken; in
the first eighteen months of the trial, participants in the
immediate group spent 85 percent of the total time before they
developed AIDS or ARC on zidovudine, compared with only 7
percent of the deferred group."
The dose of AZT used in Concorde (1000mg per day) is
significantly greater than that now used in the U.S. (500-600mg
per day). Since Concorde did not use what is believed to be the
optimum dose of AZT, some suggest the results do not provide an
accurate indication of the potential benefits of AZT.
Concorde was a monotherapy study of AZT and does not reflect the
use of combination therapy which, some claim, may provide the
most beneficial results in antiretroviral therapy.
Perspectives on Concorde
Treatment Issues surveyed a number of leading AIDS researchers
and physicians in the U.S. for their thoughts on the preliminary
results of Concorde and whether it has affected their treatment
recommendations for their HIV-positive patients. We did not use
scientific methods to select the participants. Consequently,
this survey does not necessarily represent the opinions of all
AIDS researchers and physicians in the U.S. Rather, we solicited
responses from the researchers who conducted previous U.S.
clinical trials of AZT, private physicians with large AIDS
practices, community-based researchers, and prominent critics of
previous AZT studies. We are grateful to all who participated in
1. At what CD4 level do you presently recommend your
HIV-positive patients take AZT?
Donald Abrams, M.D., San Francisco General Hospital: I have
never adopted a rigid cutoff point at which I recommend AZT. I
generally initiate therapy when a patient requests it, although
I discourage people with more than 500 CD4 cells from beginning
drug. Over the years I have followed, and continue to do so, a
cohort of people who have chosen not to initiate drug
regardless of CD4 count.
Lynn Besch, M.D., Louisiana Community AIDS Research Program:
500+1-100, or when patients feel strongly they should be on
antiretroviral drug at any CD4 cell count.
Richard E. Chaisson, M.D., Director, AIDS Service, Johns
Hopkins Hospital, Baltimore: Less than 500.
Deborah Cotton, M.D., Ph.D., Harvard Medical School: I have
always felt that a CD4 of 500 was not the appropriate cutoff at
which to initiate AZT. However, I have tended to initiate AZT
in asymptomatics with a CD4 of 350 or lower. Concorde may
change my practice: to wait until the patient has symptoms or
reaches a CD4 of 200. In other words, I was already
conservative about early intervention with AZT and am likely to
become more so. The caveat, of course, is that like everyone
else I really would like to see the complete data report.
Lawrence R. Crane, M.D., Medical Director, HIV/AIDS Programs,
Detroit Medical Center: Less than 500 CD4 cells.
Margaret Fischl, M.D., University of Miami School of Medicine;
Principal Investigator for many AZT efficacy studies: We still
initiate AZT at 500 CD4 cells. Data from David Cooper will
appear soon. His data, in patients with 400-750 CD4 cells, gave
patients AZT or placebo. Apparently, there was delayed
progression to advanced disease and/or a CD4 count below 350.
Data from 019 should also be available early next year. We are
beginning to implement early intervention starting at 750 CD4
Howard Grossman, M.D., New York City: I begin discussions when
CD4 counts decrease to less than 500 on two occasions but do
not really encourage patients strongly unless CD4 counts are
obviously dropping or the patient has symptoms.
John D. Hamilton, M.D., Professor of Medicine, Duke University;
Principal Investigator, Veterans Administration AZT Study: I
have no definite preset CD4 level at which I recommend AZT.
Below 200, however, I nearly always suggest it. Above that
level, I base my recommendation on the presence of symptoms,
their level of severity, and a frank discussion with the
patient about the other potential treatment incentives and
N. Patrick Hennessey, M.D., New York City: It depends entirely
on the individual, i.e., the chronological course of CD4
numbers and the relationship to change in CD8 and clinical
symptoms. There has never been a set number at which I suggest
Martin S. Hirsch, M.D., Harvard Medical School: Less than 500
Roberta Luskin-Hawk, M.D., Principal Investigator, Chicago
Community Program for Clinical Research on AIDS (CPCRA): 2 CD4
counts under 500.
Ronald Mitsuyasu, M.D., Director, UCLA Center for Clinical AIDS
Research and Education: CD4 at or below 500.
Joan Priestley, M.D., Director, Center for 21st Century
Medicine; Holistic AIDS Specialist: I do not prescribe AZT.
Robert R. Redfield, M.D., Chief, Dept. of Retroviral Research,
Walter Reed Army Research Institute; Principal Investigator,
gp160 Therapeutic Vaccine Study: I recommend that patients
consider AZT when their mean CD4 cell count is less than 300,
particularly patients who are also anergic. I recommend
combination antiretroviral therapy for patients whose CD4 count
mean value is less than 200. Although there is a risk of
additional side-effects when using both AZT and ddI and lack of
definitive proof related to the impact of this approach on
viral resistance, I recommend combination therapy (AZT/ddI)
rather than a single agent. I believe this represents a common
Joseph Sonnabend, MBBCh, MRCP. Medical Director, Community
Research Initiative on AIDS/New York: I have never been guided
by CD4 levels. As you may know, I believe that endogenous
alpha-interferon may play an important role in pathogenesis,
and as our work has shown, AZT can promptly remove interferon
from the circulation. I therefore recommend AZT for eight to
twelve weeks for patients who have indications of elevated
interferon, such as increased beta-2 microglobulin, fever and
weight loss. I believe that the recommendation for uniformly
prescribing at less than 500 CD4+ cells was inappropriate,
particularly given that no data other than a press release was
made available to prescribing physicians.
Paula Sparti M.D., Medical Director, Community Research
Initiative of South Florida: Less than 500 CD4 or higher if
persistent decline over six month period of time.
Paul Volberding, M.D., Director, AIDS Program, San Francisco
General Hospital; Principal Investigator, ACTG Study 019: I
would continue to recommend AZT initiation in HIV infected
persons with CD4 counts at or below 500 cells.
2. Will Concorde change your recommendation? Please explain.
Abrams: No, I will continue the above. The Community Consortium
is piloting a large simple trial, entitled comPACT I in order
to determine the best time to intervene with antiretroviral
therapy in asymptomatic people. The pilot, funded by AmFAR,
will test the methodology of the LST. Ultimately we hope to
enroll 20,000 people nationwide to answer the question.
Concorde, for some, will increase their uncertainty and make
randomization to immediate versus deferred treatment a more
Besch: No, have seen too many symptomatic patients get good to
excellent benefit from AZT. If asymptomatic patients want to
wait, I already agree with that and continue to monitor them.
Also, Concorde has not presented their data. Study population
more varied than ACTG study populations and design changed
during the course of the trial. Therefore, need to see more
Chaisson: No. The analysis was an intention to treat analysis.
The protocol was changed and patients with two CD4s less than
500 could go on open label AZT. Thus, the study showed that AZT
was no better than placebo when open label use is permitted for
CD4 less than 500.
Cotton: My reasoning has always been that to justify the
long-term toxicities, including resistance of AZT, early
intervention should result in clear-cut benefits: i.e. improved
survival, real increases in time to AIDS, etc.
Crane: Not at present. The letter did not provide sufficient
information to make an informed decision. I await the full
paper. Additionally, a significant change was made in study
design during the study. We need to know how this might have
Fischl: Concorde will not change my recommendation to patients
for a couple of reasons: a.) We need a lot more information
from the study before we should change clinical practice. b.)
My interpretation of the Concorde data is different from the
study investigators. I believe the data shows that immediate
AZT resulted in a significant delay in development of AIDS, but
did not prevent it. AZT and other nucleosides lose their
effectiveness over time. We need to know how long they work and
what other markers we should use to indicate drug efficacy. We
need to know what to do once AZT loses its effect--should we
switch to ddI or add a combination. Rather than throwing AZT
out, I believe the real question is how do we maximize and
prolong its effect. c.) Concorde may not be analyzed correctly;
we need a lot more analysis of the data. For instance, about 30
percent of patients under 500 CD4 cells elected to take AZT.
The figures in The Lancet imply this was not significant.
However, we need to see the data to be sure. For instance, what
was the self-withdrawal rate from the study? What was the
effect of the cross-over on the study results? We also need a
center by center analysis. In America it is common practice to
examine the data provided by each center to ensure there is no
bias. This was not provided in The Lancet. Since Concorde
enrolled patients in three countries with different medical
systems and differing availability of drugs, this analysis is
Grossman: No. So far, we have been given next to no data and
yet are, once again, being asked to make significant clinical
decisions. I want release of all the data. Secondly, I think it
is clear that AZT has an effect for at least several years, and
no one seriously suggests now that single agent therapy be used
for longer than that with AZT. So, I'm not sure we'll need to
change course even if the study pans out.
Hamilton: Certainly not until after I have seen the complete
analysis and commentary on the study, and perhaps not then.
Hennessey: No. It does justify in some respects my treatment of
some individuals with CD4 counts of 650 and my not treating
other individuals with CD4 counts of 350.
Hirsch: Insufficient data are available currently to indicate
whether any change is indicated. Detailed analysis of Concorde
data will be required, and the study conduct and results
compared with other related studies (ACTG 019, 016, 036 and the
Luskin-Hawk: Certainly for patients who don't tolerate AZT
well, I would stop it readily, since the only thing we may gain
is quality of life if AZT is well tolerated. I need to see the
Mitsuyasu: At this point, no. The weight of evidence from
several other large randomized studies suggest an advantage to
early antiretroviral therapy in those with CD4 below 500. The
Concorde study results do not contradict this. It speaks to the
fact, however, that with long term use, viral resistance or
host resistance may occur and limit the long term effectiveness
of prolonged AZT use.
Priestley: If anything, Concorde does not alter my
recommendation; in fact, it strengthens my opinion that AZT is
not a useful and cost-effective therapy.
Redfield: No, it confirms my previous interpretation of the
available data. AZT provides limited clinical benefit of
limited duration in a subgroup of patients with HIV infection.
In my opinion, this benefit should be reserved for patients as
they develop symptomatic or late stage disease and combination
therapy should be seriously considered when initiating
dideoxynucleoside intervention. I also believe greater
attention should be applied to determine which patients benefit
from AZT, which patients are potentially harmed by AZT
treatment, and which patients demonstrate no effects.
Sonnabend: No. Although these are preliminary results, they are
reassuring to me in confirming my decision not to routinely
prescribe AZT to asymptomatic patients.
Sparti: No, I believe that reverse transcriptase inhibitors do
delay progression of disease, and although the effect is rather
short lived, it will be prolonged by combination therapy.
Volberding: I would not change my recommendations based on the
preliminary analysis of Concorde. I would, of course, be
inclined to reconsider after the final analysis is available
for careful review, although I suspect that limitations in the
Concorde design will, even so, leave my recommendations as they
currently are. Concorde included only about half as many
patients with a baseline CD4 count of under 500 when the study
began. Also, there are suggestions in the preliminary report
that in fact AZT use did delay the onset of clinical AIDS. If
this is true in the final analysis, it could well still argue
for initiation at an asymptomatic phase of disease hoping that
second and third line therapies with non-cross resistant
nucleosides or other combinations may allow a prolongation of
the clinical benefit induced by zidovudine.
3. What do you now say to your asymptomatic HIV-positive
patients who have taken AZT for a significant period of time?
Abrams: I generally say, if it's not broken, don't fix it.
Others say the results of recent trials suggest that people
should switch, say to ddI, after a time. That trial also failed
to demonstrate a survival advantage for those who switched.
Still others say monotherapy is clearly inadequate and
combinations are the way to go. Large trials designed to answer
the question are currently maturing. I think it would be
worthwhile to see some of the data before jumping on the "more
is better" bandwagon.
Besch: What I have been saying previously will--not make any
major changes until can more critically evaluate Concorde data.
Chaisson: Stay on it if it's working.
Cotton: I always tell patients everything I know and assure
them that there are no right answers. In other words, this is a
situation where patient preference is very important. We have
no data that stopping nucleosides is bad, so I would feel
comfortable with a patient on AZT with under 200 CD4 cells
stopping it, especially if they had any side-effects. In
general, however, I am switching patients from AZT to ddI after
about one year.
Crane: Continue drug as along as clinically stable, no major
changes in laboratory parameters such as CD4, beta-2, HIV
Fischl: For this group, the question is how long does the
effect of AZT last. John Phair will be presenting an abstract
in Berlin from the MACS study which says that AZT is effective
for about 2.5 years. Drs. Volberding and Lagakos examined
patients in ACTG 019 who continued on long-term follow-up. They
found that AZT is effective for 36 months. I believe you should
switch to a new nucleoside after the presence of syncytia
inducing variants of HIV and CD4 count drops. You should
probably switch to ddI. I am reluctant to switch to combination
therapy in advanced disease, because the efficacy may be
limited. For example, we see a very limited impact on viral
burden with combination therapy. Perhaps early intervention
with combination nucleoside therapy will keep viral burden low
enough to prevent the development of resistance.
Grossman: What's a significant period? I recommend combination
therapy to all patients who have been on AZT for more than one
to two years or who show progression, and switch therapies if
Hamilton: Depending on how long the patient has been on AZT,
his/her stage of disease, and the nature of the discussion and
expectations at the time of the initial discussion, I would
tend to persevere with the present regimen, certainly until the
impact of Concorde has been assessed.
Hennessey: If they were started on AZT in our offices, I review
the study as well as the earlier U.S. studies of asymptomatics
and I reiterate the rationale we used in starting that
individual on meds at the time we did.
Hirsch: This depends on CD4 count. At the higher end of the
scale (e.g. 400-500), I would recommend no change. At the lower
end of the scale (e.g. under 300), I would recommend switch to
ddI or ddC, particularly if the slope of the CD4 counts is
precipitously downward. Ongoing studies, e.g. ACTG 175, should
help answer some of these questions.
Luskin-Hawk: Continue AZT if tolerated.
Mitsuyasu: The study suggests that the effect of AZT long-term
does dissipate with time and that alternate or combination
therapy will be needed as time progresses, even in those who
have remained relatively stable on AZT alone.
Priestley: I recommend that current users of AZT slowly taper
off taking the drug--drop 200mg from your daily dose each week
until you are down to nothing.
Redfield: Since I have not recommended early AZT therapy, I
have few patients in this situation. However, if asked, I would
review their clinical course on treatment, evaluate the
tolerance of drug in terms of side effects, consider testing
patient's virus for sensitivity to AZT, ddI, and ddC and adjust
therapy based on this information. If a patient lacks side
effects and clinical course is stable (i.e no evidence of CD4
decline or increase in immunological dysfunction) it is
reasonable for this subgroup to consider continuation of single
agent AZT treatment. However, if CD4 decline has continued
despite AZT and/or defects in functional immunity development
(i.e. anergy), I would consider viral sensitivity testing and
combination antiretroviral therapy (based on results) or
encourage the individual to consider other experimental
Sonnabend: I don't have many patients in this category. For the
few I have, I would reiterate my recommendation that they
consider other approaches.
Sparti: If they are tolerating AZT well, I advise continuing
medication, with a low threshold for making the decision to add
ddI or ddC.
Volberding: With asymptomatic patients who have taken AZT for a
significant period of time, I would recommend that they
continue the use of zidovudine until there has been some
laboratory or clinical evidence of disease progression. If, for
example, the CD4 cell count is less than 50 percent of what it
was when they began zidovudine, or if they developed clear
symptoms or signs of HIV disease, I would urge that they either
add ddI or ddC or switch to one of these drugs based on the
results from ACTG trials 116A and 116B/117.
4. What are your thoughts on the CD4 data from Concord?
Abrams: I am becoming more concerned about what CD4 changes in
response to drugs really means. Intuitively, one would love to
believe increased counts translate into clinical
benefit--decreased progression and prolonged survival. More and
more studies bring this into question. Concorde, the VA study,
116B/117, and our recent CPCRA ddI versus ddC study. In the
CPCRA trial, a CD4 response was appreciated in the ddI
recipients, but there was a trend towards increased survival in
those initially treated with ddC. We really need to see some
mega meta-analysis of all CD4/survival data presented at some
forum somewhere to help clarify the issue. We may be discarding
agents that could have some potential benefit for lack of a CD4
response and approving others with little effect other than a
small transient CD4 cell increase.
Besch: No surprise--I think most researchers view CD4 counts as
available but imperfect surrogate markers--not many studies
have correlated CD4s or drug treatment with survival.
Chaisson: CD4 is a reliable surrogate marker. I do not accept
the published conclusion of the Concorde study, thus maintain
my opinion about CD4.
Cotton: The CD4 data is critical to analyze carefully when it
comes out. We have pinned a lot on CD4 as surrogate marker; we
need to constantly reassess this decision as new data becomes
available. (By we I mean the ACTG, FDA, advocates, patients.)
Crane: I think that clinical parameters more important and
agree that CD4 is not a "perfect" surrogate marker.
Fischl: I was disappointed in these comments. CD4 is not
perfect, as statistical modeling has shown. However, the
Concorde team may not have modeled CD4 correctly. CD4 is a two
dimensional marker --you have to account for increases and the
duration of response. When CD4 is modeled correctly, it becomes
Grossman: a). Showing beginning and ending levels without
looking at intervening data points is not useful. b). What
about other markers, such as CD4 percent, Beta 2 microglobulin,
anti-p24 antibody, etc? c). I would imagine that if one- third
of delayed treatment group actually took AZT for the last two
to three years of the study, it would have a large impact on
the results, especially when progression rates are so low.
Hamilton: I have no thoughts about the data until I have had
the opportunity to review the details of Concorde. Generally,
my use of CD4 in guiding decisions is somewhat limited in any
Hennessey: When the Concorde study was not stopped at the
evaluation point that stopped ACTG 019 and the other U.S.
studies, it was apparent that dramatic differences were not
being seen between the groups and results similar to those just
published in the Concorde study were expected. I have some
questions as to composition of the European/U.S. studies in
terms of just how "asymptomatic" the participants were.
Conclusion, the data is not a surprise but the clarification of
the letter will be difficult as best.
Hirsch: Initial impressions are that AZT raises CD4 counts
early, but that this effect is lost over time, confirming
earlier studies. More definitive interpretation will require
more detailed analysis of the data. Conclusions about the
prognostic value of CD4 changes, based on the available data
from Concorde, appear premature.
Luskin-Hawk: This study underscores the importance of using
clinical endpoints and survival to evaluate antiretroviral
therapy and not to rely too heavily on surrogate endpoints.
Since patients on the deferred arm could start AZT if two CD4
counts were under 500, how different were the two groups?
Severe drug toxicity was an endpoint in Concorde and a higher
dose (1000mg/day) was used.
Mitsuyasu: The finding merely confirms the fact that benefits
from small changes over the short run in CD4 counts are perhaps
overshadowed by other variables (such as development of viral
resistance, etc.) over the long run. Better prognostic markers
are clearly needed to predict overall prognosis for individual
patients and to assess benefit of therapy. CD4 change is a
good, but not perfect surrogate marker.
Priestley: I believe this study will renew interest in CD8
cells, and concentrate on therapies that raise them.
Redfield: I believe this is possible. What I would like to know
is was there a disparity between treatment arms in terms of
functional immunity (i.e. skin test reactivity) and clinical
outcome. I suspect that maintenance or lack of functional
immunity and outcome will correlate. The difference reported in
CD4 cell counts are marginal, and even of less importance to me
if these differences reflect differences in single values
rather than mean values.
Sonnabend: Again, although the results are preliminary, they
seem to justify the unease that some of us felt over the
decision to use CD4+ levels as a measure of therapeutic
Sparti: I strongly believe that CD4 count is predictive of
outcome. Survival is much better in 50-100 CD4 group than those
persons close to zero. The Concorde data will need to be
evaluated closely. Were they using multiple prophylaxis?
Volberding: The CD4 data from Concorde are obviously still
extremely preliminary. As I understand it, there was, in fact,
a significant CD4 benefit from the use of zidovudine in the
immediate therapy group. The investigators concluded that CD4
wasn't of value because they felt there wasn't an ultimate
clinical value but again I think we shouldn't be willing to
concede that point until we see the final data. If there was,
in fact, a clinical benefit from our way of thinking, then
their study may in fact reinforce the use of CD4. Obviously,
none of us believe that CD4 is a very good marker of the
disease or especially of its response to therapy but I think
that it would be dangerous to immediately abandon this marker
given that that might force us to use only clinical endpoints
which would require that drug development take much longer than
it already does. I am very hopeful that HIV quantitation, using
a variety of new methodologies, may be able to resolve this
dilemma quickly by giving us a surrogate marker that is more
central to the actual disease process.
Excerpt from a letter to Treatment Issues by Ian Weller M.D.,
University College London Medical School; Principal
Investigator, Concorde Study:
At the moment, I do feel strongly that physicians and patients
should not change what they are on until the final report is
published...If a patient with CD4 levels at 350, and who had
been on AZT for two years, wanted me to decide for them then I
would recommend that they stop (taking AZT). However, I am one
of the few that has seen many of the analyses that will make up
the final report of Concorde, and outside of the trial I have
been very conservative in the use of zidovudine monotherapy in
asymptomatic patients and my patients have been comfortable
with this but of course they may self-select in terms of coming
to see me.
Excerpt from the statement of the National Institute of Allergy
and Infectious Diseases:
Daniel F. Hoth, M.D., Director of the Division of AIDS at the
National Institute of Allergy and Infectious Diseases (NIAID),
says, "The Lancet letter is based on a preliminary analysis,
with many important details to be clarified. We look forward to
meeting with our European colleagues and reviewing this data in
detail. However, at this time we see no basis for changing the
recommendation to initiate antiretroviral therapy for
HIV-infected persons whose CD4+ T-cell count falls below 500
per cubic millimeter of blood."
Excerpt from the statement of the Food and Drug Administration:
FDA believes this report clearly shows the need for continued
studies of AZT and other AIDS products to ensure that the
usefulness of the drugs is fully understood. It will be
important to explore what factors, such as viral resistance,
may be responsible for decreases in benefits of the drugs over
The Concorde study documents the need for trials that include
clinical outcome data-the effect of the drug on the course of
disease--and the need for postmarketing surveillance for drugs
approved on the basis of surrogate markers such as CD4 cell
FDA is looking forward to reviewing the data referred to in The
Lancet letter, particularly those pertaining to the use of
surrogate markers to predict clinical outcome. Careful review
and discussion with expert consultants will allow the agency to
place the study in perspective and will indicate whether any
changes in labeling for AZT are warranted.
Excerpt from the statement of Project Inform:
The bottom line: Far more data and analysis are required before
any final conclusions should be drawn about the meaning of this
study. The overall rate of disease progression seen in this
study, both on and off drug, should be extremely encouraging to
The relevance of these findings may or may not be relevant to
the way this drug is used in the US. If one assumes they are
basically accurate, then the prudent course of action is not to
abandon therapy, but to seek to improve upon it, especially if
one is still using AZT as a sole therapy after a year or more
If the findings are relevant, they would seem to back up
conclusions drawn earlier in the US that, at most, people
should remain on a single drug therapy for no longer than one
year, or that small doses of a combination of drugs is likely
to be both more effective and longer lasting because of the
inhibition of drug-resistant strains of virus.
The study provides clear evidence that long-term
antiretroviral, even using a dose two times higher than
standard, for four years, does not augment disease progression.
Different and more appropriate statistical models must be
employed or enforced in future studies to guarantee that
assessment will reflect the actual behavior of study
participants, rather than an idealized course of treatment
abandoned in the course of study.
Excerpt from the statement of Burroughs Wellcome, the
manufacturer of AZT:
CD4 cell counts were significantly higher for people who
received immediate treatment with AZT compared to CD4 counts
for people who received deferred treatment with AZT after a
variable time on placebo.
It shows an encouraging low rate of adverse reactions
associated with 1g AZT for up to four years. This dose is
higher than the currently recommended dose.
The study reports no difference in benefits in terms of disease
progression or mortality between immediate and deferred AZT
treatment. However, during the trial approximately 40 percent
of people who received placebo initially changed to AZT.
There was a surprising discrepancy between the benefit observed
for CD4 cell counts related to immediate use of AZT and the
lack of difference in benefit measured by disease progression,
as specified in the protocol. The Concorde team concludes that
there are doubts about the value of CD4 cell counts as a
measurement of the benefit of therapy.
Copyright (c) 1993 - Gay Men's Health Crisis, New York City, NY
Noncommercial reproduction encouraged. Distributed by AEGIS,
your online gateway to a world of people, knowledge, and