ddI: Good News, Questions Remain

New data released April 13 at a meeting of the AIDS Clinical
Trials Group suggested that patients who had been using AZT and
then switched to low-dose ddI did better than those who continued
with AZT. But several major questions remain, and important new
information is expected over the next several months. We will be
watching to see how the consensus of leading AIDS physicians
develops -- especially after the International Conference on AIDS
in July, where important additional data is expected.

The new information concerns patients who had been on AZT
more than 16 weeks before being randomized either to continue AZT
or to switch to one of two doses of ddI. To qualify for the
study, they either had to have a T-helper count of 300 or less
and have AIDS or ARC, or have a count of 200 or less, in which
case they could be asymptomatic. Over 900 volunteers
participated in this study, with only five percent lost to
followup.

The volunteers were randomly assigned to three groups:
continue AZT (600 mg per day), switch to 500 mg ddI (low dose),
or switch to 750 mg ddI (high dose). [Note on doses: The newer
tablet form of ddI is better absorbed than the powdered form used
in the study. A 400 mg dose of the tablets is equivalent to a
500 mg dose of the powder.]

The basic results were:

* Patients assigned to the low-dose ddI did best. Those who
had ARC or were asymptomatic had significantly fewer new AIDS-
defining events than those assigned to continue AZT; the high-
dose ddI group was intermediate. However, this difference was
not seen in patients with AIDS.

* Those assigned to either ddI group had higher T-helper
counts over time than those in the AZT group.

* However, there was no difference in survival between any
of the groups.

* Also, it did not matter how long the study volunteers had
been taking AZT before switching to ddI; the benefit was the same
in any case.

* The AZT group had more side effects than the ddI groups,
as measured by the average time until the first dose modification
(if any) was required. But the high-dose ddI group had the most
serious side effects, with 31 cases of pancreatitis, including
two deaths.

Major questions remain:

* Why was there no survival benefit if there was a benefit
(of low-dose ddI compared to continued AZT) in reducing serious
opportunistic infections? It is possible that this surprising
result occurred because of the way the study was conducted. For
ethical reasons, patients' treatments were changed if they were
not doing well on the treatment to which they were randomly
assigned; 80 percent of the deaths occurred 30 days or more after
the assigned treatment had been discontinued. At this time no
one knows if there really was no survival benefit, or if this
study did not show it because those who died had gone off the
assigned treatment long before.

* Why was benefit seen only for those less seriously ill,
but not for those with AIDS? No one knows the answer at this
time. [We hope that in addressing this question researchers will
consider some of the less orthodox theories, including the
possibility that some of the benefit of AZT may be due to immune
suppression -- of autoimmune or otherwise inappropriate immune
responses caused by the disease.]

* A major mystery to researchers is why the length of prior
AZT use (before the study) had no effect on the benefit of
switching to ddI. The expectation had been that those who had
taken AZT longer might have viral strains which had developed
more resistance to AZT, and therefore they would benefit most
from ddI. This did not happen.

There was a great range of length of prior AZT use. All
volunteers had to have tolerated AZT for at least 16 weeks to
qualify, but many had taken that drug for over 48 weeks. The
reason is that this study (which is named ACTG 116-B/117) had
originally begun as two separate studies; part of ACTG 116 (those
patients who had taken AZT for more than 16 weeks) was combined
with ACTG 117 (for patients who had taken AZT for over 48 weeks);
those recruited later in the combined study had to have taken AZT
for over four months.

Could ddI be the better drug for some patients even for
first-line therapy (for those who had no prior use of AZT)? This
study cannot answer that question. Researchers are waiting for
the results of ACTG 116A, comparing AZT and ddI in patients who
had taken AZT for less than 16 weeks. That study is not finished
yet. (ACTG 116-B/117 was stopped early, because it had already
shown that one treatment was working better than another.)

Comment

All that is known today from ACTG 116-B/117 is based on a
rapid, very early analysis of the data. More will be learned as
analysis proceeds, and as other trials finish.

A number of people are guessing -- based partly on rumors
from ongoing European studies -- that ddI may prove to be a
better drug in general than AZT. It is also possible that the
best dose of ddI will be even less than the "low" dose used in
ACTG 116-B/117. But at this time, much of the information we
would want is not available.

We do expect that the information already released will
influence medical practice. An April 13 backgrounder, "ACTG
116-B/117: Questions and Answers," released by the U. S.
National Institute of Allergy and Infectious Diseases, concludes:

"Further analysis of this study will need to be completed
before any final recommendations can be made. Completion of
other, on-going comparative studies of ddI (ACTG 116-A and ACTG
118) will help in determining the full impact of these results on
patient management. While it is not possible at this time to
make recommendations for all HIV-infected patients, patients
should discuss this information with their primary care providers
in order to devise treatment strategies to meet individual
needs."