July 12, 2002
Greg Robbins presented both papers during the last late breaker session of the meeting. The presentation was a little bit convoluted because there was a lot of statistical lingo, but this is a study that we all need time to digest. ACTG 384 will be one of the most important studies also for the years to come, together with INITIO, whose preliminary results were also presented during the conference.
The main questions ACTG 384 asked were:
To answer these questions, 980 naive patients were enrolled in the U.S. and Italy using a factorial design.
This presentation deals with the last of the questions. The factorial design had six arms, d4T + ddI with efavirenz or nelfinavir or both and AZT + 3TC with efavirenz or nelfinavir or both. The two four-drug arms were compared with the four triple arms. The primary endpoint of the trial was the time until second virologic failure (in the case of the triple arm) compared to the first virologic failure in the quadruple arm (the comparison is fair because the important thing is to know when all classes of drugs have been used).
The answer was clear: there was no difference in the time to reach the primary endpoint. What it means is that it does not matter if you start with three or four drugs, so it is better to start a three-drug regimen for multiple reasons: you avoid the cost -- both economic and in toxicity -- of the quadruple regimens, and you have more options for the future in case virologic failure occurs.
The only clearly positive effect in the four-drug arms was that it delayed the time that it took the first combination regimen to fail, but, in the long run, it did not seem to matter, and it is very unlikely people will use four-drug regimens with this idea in mind. It is also clear that efavirenz and AZT + 3TC regimens were the favored ones, which I talk about in another summary.
As I mentioned, ACTG 384 will be as important in defining the future of HIV treatment as ACTG 320 was a few years ago. This study asked the question what is the best regimen to start antiretroviral therapy. It compared, in an elegant factorial design, two different nucleoside combinations (AZT + 3TC and d4T + ddI) with nelfinavir (a protease inhibitor) or efavirenz (an NNRTI) or both. This design allows the efaviranz arms to be grouped and compared to the nelfinavir arms and the same for the two different nucleoside arms. The primary endpoint was the time to the second virologic failure in the triple arms or the first virologic failure in the quadruple arms.
Here is a summary of the six arms of the study:
AZT + 3TC + EFV d4T + ddI + NFV
AZT + 3TC + NFV d4T+ ddI + EFV
d4T + ddI + EFV AZT + 3TC + NFV
d4T + ddI + NFV AZT + 3TC + EFV
AZT + 3TC + EFV + NFV
d4T + ddI + EFV + NFV
The follow up was very long (over two years).
Had you asked me three years ago, I would have said that it did not matter how one starts treatment -- that it would take approximately the same time to fail two consecutive regimens independent of the initial regimen (the primary endpoint of the trial). I was wrong: It does matter.
During these past two years, we have realized how potent efavirenz is and how important pharmacokinetics and tolerability are in the success of an antiretroviral regimen. In this study, the winning combination was AZT + 3TC + EFV. AZT + 3TC was less toxic than d4T + ddI -- something I think we all have learned the hard way over the last two years -- and EFV was more potent and durable than nelfinavir. It does not necessarily mean that AZT + 3TC + EFV is the best initial combination (we heard about other good combinations with EFV during this meeting), but combined with the results of other studies like INITIO, and the better toxicity profile of efavirenz, NNRTI-based regimens became the big winners as first-line therapy during this conference.
As I have previously commented, this trend favoring NNRTI-based regimens as first-line therapy could be reversed if primary resistance becomes a frequent problem, or if the toxicity profile of PIs improves dramatically with the new formulations that will come out in the future.