Trials & Errors

For many, the scare of fat redistribution that seems linked to PI's has people looking for good drug alternatives that don't include a PI. The most significant naïve therapy trial presented shows that a regimen that includes three NNRTI's work as well or better than the classic cocktails with a PI. The update on the trial conducted by DuPont, compared three treatment groups:

• Efavirenz (Sustiva), zidovudine, and 3TC (one NNRTI and two nRTIs)
• Efavirenz and indinavir (Crixivan) (one NNRTI and a PI)
• Indinavir, zidovudine, and 3TC (two nRTIs and a PI)

After 48 weeks of follow-up, efavirenz plus the two nRTIs is at least as effective as the indinavir plus two nRTIs regimen. Based on an intent to treat analysis (including all patients who began on the trial) after 48 weeks, 71 percent of people receiving efavirenz/zidovudine/3TC had viral loads of less than 400 copies/ml, compared to only 48 percent for people assigned to indinavir/zidovudine/3TC.

When evaluated based on people actually on treatment, after 48 weeks, 98 percent of people on the efavirenz regimen had viral load levels below 400, compared to 86 percent for the indinavir-containing regimen. (There was a much higher dropout rate in the indinavir arm, compared to the efavirenz arm, explaining the very different results in the two analyses.)

After 48 weeks, viral load levels were below 50 in 71 percent of people assigned to the three-drug efavirenz arm, compared to only 47 percent for the three-drug indinavir arm. When viral loads were examined at 48 weeks, below 90 percent in the efavirenz arm were below 50 copies, compared to 80 percent with a below 50 viral load in the indinavir arm. There has been concern that an NNRTI regimen may not be as good as a PI-containing regimen, but the DuPont 006 trial shows that even in people with viral loads above 100,000, the NNRTI regimen performed as well, or better, than the PI-containing regimen.

The most common toxicity resulting from efavirenz was central nervous system complaints, noted in 55 percent of people. Those who were instructed about these problems before beginning therapy seemed to have less trouble. The most common complaints were vivid dreams and feeling confused or even depressed. These symptoms were usually mild and resolved after 3 weeks in almost everyone.

The Atlantic Trial.

Another large treatment strategy trial presented was the Atlantic Trial, which compared three different treatments:

• d4T, ddI, and 3TC (three nRTIs)
• d4T, ddI, and neirapine (two nRTIs and an NNRTI)
• d4T, 3TC, and indinavir (two nRTIs and a PI)

There were about 80 people in each group, all had CD4 counts above 200 and no prior therapy, and the follow-up was 24 weeks. At that time, the number of people with viral loads below 400 was 71 percent with three NRTIs, 69 percent with two NRTIs and one NNRTI, and 78 percent with two NRTIs and one PI. Based on all patients who began the trial, the numbers with viral loads below 50 were 56 percent, 67 percent, and 71 percent respectively. After 24 weeks, the numbers of people who stayed on therapy who had viral loads of less than 400 were 80 percent, 89 percent, and 91 percent, respectively.

CNA 3003.

This study can be summed up by saying that the three drug combination in the study is better than the two, but still not good enough. The follow-up data presented by Margaret Fischl on this trial compared two drug combinations:

• AZT and 3TC (two NRTIs)
• AZT, 3TC, and abacavir (three NRTIs)

Abacavir is also called Ziagen, and is the most recently FDA-approved nRTI antiretroviral agent. The study was modified last year when it was observed that the triple-drug regimen was much better than the two-drug regimen. After 48 weeks of therapy with the triple nRTI combination, the number of people with viral loads less than 400 was 60 percent. However, in people who had viral load levels above 100,000 at the beginning of treatment, only 33 percent of people had a viral load less than 400 after 48 weeks. So, while this triple nRTI regimen may be effective in controlling HIV replication in people with low viral load levels, it is not potent enough to control HIV replication over the long term in people with higher viral loads. This combination had very low toxicity and has the advantage of saving two major classes of drugs -- the NNRTIs and the PIs -- for later use.

CNA 3005.

Preliminary data was also present by Schlomo Staszewski on a trial of two drug combinations in people with no prior therapy:

• Abacavir, zidovudine, and 3TC
• Indinavir, zidovudine, and 3TC

The follow-up is only 24 weeks. In both groups, 65 percent of people had achieved a viral load of less than 400, based on all people who began the trial. For people who were still on the therapy, the numbers increased to 85 percent. Longer follow-up from this trial will be available later this year.

Other Trials.

There are several similar treatment strategy trials ongoing at the AIDS Clinical Trials Group (ACTG), and in the next year or so it is hoped there will be more long-term data comparing similar types of regimens as first-line selections.

However, it will still be difficult to know whether it is better to begin a regimen based on three NRTIs, or one that includes an NNRTI, a PI, or even two PIs in addition to two NRTIs. It may well be that one size does not fit all.

Also, the efficacy of the second-line regimen must be known, as really you need to know which sequence of regimens provides the best control of HIV over the course of many years. It could be that A is better than B, as a first-line regimen, but the sequence B before A is better, overall, than A before B.