[afro-nets] HIV Therapy - What Do We Know, and When Do We Know It?

[Dr Rana Jawad Asghar <jawad@alumni.washington.edu>]

HIV Therapy - What Do We Know, and When Do We Know It?
------------------------------------------------------

Paul R. Skolnik, M.D.
New England Journal of Medicine
Volume 349:2351-2352 December 11, 2003 Number 24

http://content.nejm.org/cgi/content/full/349/24/2351

It has often been stated that the central questions in therapy
for human immunodeficiency virus (HIV) infection are the same
now as they were when the epidemic started. When do we start
therapy, what do we start with, when should we switch, what con-
stitutes failure of therapy, and why do our therapies fail? In-
formation from clinical trials of various anti-HIV treatments,
along with in vitro observations, has narrowed the choices. Of-
ten, the design and completion of these trials lag behind prac-
tice, but they still serve several crucial purposes. The studies
confirm what we think we know, disprove beliefs based on plausi-
ble but faulty hypotheses, and suggest novel hypotheses that can
be tested and that can alter future treatment strategies. Occa-
sionally, new ways of testing these hypotheses emerge.

The studies by Robbins et al.1 and Shafer et al.,2 reported in
this issue of the Journal, used a complicated factorial design,
but the basic conclusions are straightforward. There are cur-
rently 20 agents that have been approved by the Food and Drug
Administration for the treatment of HIV infection. We have
learned that potent combination therapies are the most effica-
cious and durable approaches to the use of these agents. The
study by Robbins et al. tells us that one of these combinations
- zidovudine, lamivudine, and efavirenz - is a particularly good
choice with which to begin therapy for patients with previously
untreated HIV infection.

Perhaps just as important, these studies provide firm data from
prospective clinical trials that confirm what many have recently
suspected: that the combination of stavudine and didanosine
should not be used for the initial treatment of HIV infection,
if other choices exist, because of unacceptably high rates of
adverse effects. The ability of clinical trials to demonstrate
what does not work is often overlooked. This point has also been
highlighted in recent reports of particularly high rates of
failure in patients treated with a once-daily combination of
abacavir, lamivudine, and tenofovir3 or of didanosine, lami-
vudine, and tenofovir.4 In vitro studies may also suggest poten-
tially dangerous drug interactions in anti-HIV therapy (e.g.,
between stavudine and zidovudine), but there is no substitute
for carefully conducted clinical trials of new drug combinations
to discover pitfalls definitively and accurately.

Beyond these lessons, the current reports raise several impor-
tant issues about study design that may alter future practice.
First, the definition of what constitutes the failure of a regi-
men is an important consideration. This issue mirrors clinical
practice, in that treatment goals differ during different stages
of this chronic viral infection, for which a "cure" will elude
us unless new pathogenic and treatment paradigms are discovered.
If sequential therapies are needed, perhaps the overall outcome,
not the effects of the initial regimen alone, should be used as
an end point. If adverse effects, or events other than strict
virologic failure, result in the failure of a regimen, then per-
haps this end point should be included in the assessment to mir-
ror clinical practice accurately. Although one may take issue
with the exact parameters used to define "premature discontinua-
tion" of therapy, the current studies attempt to use this rea-
sonable assumption in the study design.

In drawing conclusions, it is always important not to generalize
beyond the conclusions justified by the data. For example, al-
though some may be tempted to use the current data to support
the notion that an initial treatment regimen consisting of two
nucleoside reverse-transcriptase inhibitors (such as zidovudine
and lamivudine) combined with a nonnucleoside reverse-
transcriptase inhibitor (such as efavirenz) is superior to a
regimen of two nucleoside reverse-transcriptase inhibitors com-
bined with a protease inhibitor, this conclusion would be a mis-
take.5 Several lines of evidence suggest that nelfinavir, which
was tested in the current studies, may be less effective than
other protease inhibitors for the initiation of HIV treatment.
Newer strategies that make use of the pharmacologic actions of
ritonavir on the hepatic P-450 system to "boost" plasma levels
of coadministered protease inhibitors, or that use new, more po-
tent protease inhibitors, may well constitute important options
for initial treatment in combination with other agents.

Previous studies have shown that three anti-HIV drugs are often
better than two and that two drugs are better than one. Might
regimens consisting of four drugs be even better? For the par-
ticular drugs studied here, the answer is probably no. Again,
definitions of success and failure are paramount. In the study
by Shafer et al., with respect to the end points defined, there
was no advantage to the four-drug regimens; however, the data
concerning viral resistance are provocative. The four-drug regi-
mens resulted in fewer mutations that rendered the circulating
HIV quasispecies resistant to currently available drugs than all
three-drug regimens (except that consisting of zidovudine, lami-
vudine, and efavirenz - another indication of the potency of
this regimen). Since, for patients in whom initial therapy
fails, drug resistance is often the primary difficulty in devis-
ing subsequent drug regimens, this advantage might be an impor-
tant one. Future studies should include analyses of this possi-
bility to determine the efficacy of regimens based on a potent,
boosted protease inhibitor or, if tolerable, a more complex ini-
tial regimen.

A randomized clinical trial is optimal for determining the best
treatment practices, but the question of when to start therapy
will probably never be addressed in such a trial because of the
huge number of study subjects and extremely long study period
that would be needed. We must use our knowledge of pathogenesis,
along with clues from retrospective cohort studies,6 to answer
this question. The pendulum of opinion about treatment for HIV
infection is particularly apparent when it comes to the timing
of treatment initiation.7 The standard of care not long ago was
to initiate treatment when the CD4 cell count fell below 500 per
cubic millimeter and the plasma HIV RNA level was low; now, some
advocate delaying therapy until the CD4 cell count is below 200
per cubic millimeter. More recent analyses suggest that this
threshold is too low and that therapy should be initiated if the
CD4 cell count is below 350 per cubic millimeter or if the HIV
RNA level is higher than 55,000 copies per milliliter. Individ-
ual factors, such as a patient's "readiness," are important
here. The central point is that changes in therapeutic practice
should not be undertaken on the basis of incomplete or subopti-
mal retrospective analyses. This caution remains a challenge in
the field of therapy for HIV infection, a field in which clini-
cians are eager to change practices quickly.

Determining what constitutes the best choice for initial therapy
is a balancing act, and adherence is another factor that must be
considered. Can and will the patient take the prescribed treat-
ment reliably? Clinical trials are particularly ill-suited to
answer this question because study populations often do not re-
flect the populations of patients that doctors see in the
clinic. The high rates of failure documented in several large
urban clinics around the world bear witness to this phenomenon.8
Some general principles are clear: adherence is usually better
with twice-daily regimens than with those that include a midday
dose, and certain side effects result in unacceptably low levels
of adherence, but much remains to be learned. Perhaps, for indi-
vidual patients, the exact balance of risks and benefits -
namely, the chance for long-term viral suppression, the risk of
certain side effects, the possibility of induction of drug re-
sistance, and the ability to adhere to a particular regimen - is
best addressed by open discussions between care providers and
their patients.

Dr. Skolnik reports having received consulting fees from Scher-
ing-Plough and research grant support from Gilead Sciences and
Roche Pharmaceuticals.

Source Information
 From the Center for HIV/AIDS Care and Research, Boston Univer-
sity Medical Center, Boston University School of Medicine, Bos-
ton.

References

[1] Robbins GK, De Gruttola V, Shafer RW, et al. Comparison of se-
quential three-drug regimens as initial therapy for HIV-1 infec-
tion. N Engl J Med 2003;349:2293-2303.[Abstract/Full Text]

[2] Shafer RW, Smeaton LM, Robbins GK, et al. Comparison of four-
drug regimens and pairs of sequential three-drug regimens as
initial therapy for HIV-1 infection. N Engl J Med 2003;349:2304-
2315.[Abstract/Full Text]

[3] Gallant JE, Rodriquez AE, Weinberg W, et al. Early non-response
to tenofovir DF (TDF) + abacavir (ABC) and lamivudine (3TC) in a
randomized trial compared to efavirenz (EFV) + ABC and 3TC:
ESS30009 unplanned interim analysis. Presented at the 43rd In-
terscience Conference on Antimicrobial Agents and Chemotherapy,
Chicago, September 14-17, 2003. abstract.

[4] Toole J. High rate of virologic failure in patients with HIV in-
fection treated with a once-daily triple NRTI regimen containing
didanosine, lamivudine, and tenofovir (important drug warning
letter).

[5] Foster City, Calif.: Gilead, October 14, 2003. Panel on Clinical
Practices for Treatment of HIV Infection, Department of Health
and Human Services. Guidelines for the use of antiretroviral
agents in HIV-1-infected adults and adolescents: November 10,
2003. (Accessed November 18, 2003, at http://aidsinfo.nih.gov.)

[6] Palella FJ Jr, Deloria-Knoll M, Chmiel JS, et al. Survival bene-
fit of initiating antiretroviral therapy in HIV-infected persons
in different CD4+ cell strata. Ann Intern Med 2003;138:620-
626.[Abstract/Full Text]

[7] Lane HC, Neaton JD. When to start therapy for HIV infection: a
swinging pendulum in search of data. Ann Intern Med
2003;138:680-681.[Full Text]

[8] Ledergerber B, Egger M, Opravil M, et al. Clinical progression
and virological failure on highly active antiretroviral therapy
in HIV-1 patients: a prospective cohort study: Swiss HIV Cohort
Study. Lancet 1999;353:863-868.[CrossRef][ISI][Medline]

--
Dr Rana Jawad Asghar
Program Manager Child Survival, Mozambique
Provincial Coordinator Sofala Province, Mozambique
Health Alliance International, Seattle, WA, USA
http://depts.washington.edu/haiuw/
Coordinator South Asian Public Health Forum
http://www.saphf.org
mailto:jawad@alumni.washington.edu
http://www.DrJawad.com