ACTG 384 at Barcelona: the bottom line.

ACTG 384 is a large clinical trial by the U.S. Institute of Allergy and Infectious Diseases, which enrolled over 900 treatment-naive volunteers. Shortly before the Barcelona conference, one of the six different treatment regimens tested was found to be clearly better than the others. So volunteers will be told which regimen they were on, and advised on options for changing therapy when appropriate.

Letters were written to study volunteers before the conference; however, local IRB (institutional review board, intended to protect volunteers in clinical trials) for most of the study sites required that they review the letters before they could be distributed. Since the reviews had to wait for the IRB process, few volunteers received the news before the Barcelona meeting.

Two abstracts on this study were presented at late breakers at Barcelona (numbers LbOr20A and LbOr20B). These hard-to-read abstracts can be found by a search on the official conference site, probably reachable at: http://www.aids2002.org/Search/AbstractSearch.asp (be careful to use the Abstract Search, not the Quicklinks search confusingly placed above it); you can search for the abstract numbers.

But what matters most is:

1. Efavirenz plus AZT plus 3TC was a better starting regimen than the other five tested in this trial;

2. Regimens including d4T plus ddI had more toxicity than others; and

3. There are questions about how to apply these results to treatment today (see article by Peter Ruane, M.D., at: http://www.medscape.com/viewarticle/438430).

Comment

1. This is an important trial--despite issues of whether the questions it was designed around will ever be answered, or were the best questions to ask. The comparisons between regimens do not generalize to drug classes (whether it is better to start with a protease inhibitor or an NNRTI, for example), but apply mainly to regimens chosen from those available several years ago. Still, the trial is important because it collected good-quality data after randomized treatment assignments--data likely help in many ways to improve HIV care. We will be hearing much more from this trial over the next few years.

2. The local IRB review of the researchers' recent letter to the volunteers, which delayed its delivery to most participants, made no sense. The local IRBs will not change this letter, nor stop it from going out. So the whole exercise was an empty process. We hope the current dysfunctional relationship between local and national IRBs will be part of the badly needed review of protection of volunteers in medical research. Perhaps national studies should be reviewed by national IRBs that can be trusted--with local IRBs able to block the study initially, but not routinely involved when it is too late to make any contribution.