1995 has turned out to be a turning point in the way we think about antiretroviral therapy. The era of pessimism and nihilism ushered in by the International AIDS Conference in Berlin in 1993 appears to be over, and is being replaced by an era marked both by cautious optimism and confusion. The guidelines for antiretroviral therapy released after the Berlin conference were made instantly obsolete by the results of ACTG 175 and the Delta Trial, discussed below, and it is likely that any guidelines written today would be obsolete by the time they were published. What follows is a summary of the new data on antiretroviral therapy, and my own opinion on what it all means.
The results of ACTG 175 were presented at the Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC, the country's largest infectious disease conference) in September. This was a large trial comparing four different antiretroviral regimens: AZT (zidovudine, Retrovir) alone, ddI (didanosine, Videx) alone, AZT plus ddI, and AZT plus ddC (zalcitabine, HIVID). All drugs were given at standard doses. 2467 patients with CD4 counts between 200 and 500 participated in the study, 1067 of whom had never taken antiretrovirals, and 1400 of whom had taken AZT.
The study found that compared to participants taking AZT alone, those taking ddI, AZT/ddI, or AZT/ddC were significantly less likely to progress to any of the three main endpoints: a decline of at least 50% in the CD4 cell count, an AIDS-defining condition, or death. When only the endpoints of AIDS or death were looked at, those on ddI alone or AZT/ddI did better than those on AZT. What was most surprising was that these findings applied even to those patients who had never taken AZT before.
The European-Australian Delta trial also compared AZT to combination regimens in individuals with CD4 counts less than 350. Unlike ACTG 175, however, Delta did not include a ddI monotherapy arm. Patients in the Delta trial had somewhat more advanced disease that those in 175. The conclusions, however, were consistent with those of the American trial. Both AZT/ddI and AZT/ddC were superior to AZT monotherapy. Interestingly, this was only true for those who had never taken AZT. For those with AZT experience, the differences were not significant, but this has been attributed to the very high drop-out rate among patients in the Delta 2 arm, perhaps because 3TC became available during the course of the trial.
There are several important conclusions that can be drawn from these large trials. One that is often overlooked is that for the first time antiretroviral therapy has been associated with improved survival in patients with asymptomatic, intermediate-stage disease. The second is that AZT monotherapy appears to be inferior to both combination therapy and ddI monotherapy even in patients who have never taken AZT before. This contradicts the findings of earlier studies, in which AZT was found to be superior for those who had never been on therapy. The reason for the discrepancy is still unclear, but some believe it may be due to the fact that more and more people are becoming infected with AZT-resistant virus from the outset.
These powerful studies provide us with extremely useful information, but they don't really tell us everything we need to know about managing HIV disease in 1996. Since they were carried out and completed, a number of new treatments have emerged, such as d4T, 3TC, and protease inhibitors. How these new therapies stack up against the combinations used in ACTG 175 and Delta is not known.
Protease Inhibitors
A new class of drugs known as protease inhibitors show great promise for use in combination antiretroviral regimens. Unlike the nucleoside analogs discussed above, which inhibit the reverse transcriptase enzyme at early stages of the viral life cycle, these new agents interfere with the protease enzyme, thereby inhibiting viral replication by interfering with the assembly of viral particles before release from the cell. Protease inhibitors have the potential for more dramatic reductions in viral load than the currently available nucleoside analogs. Unfortunately, treating with protease inhibitors alone appears to lead to the rapid development of resistance, so the drugs will almost certainly be used in combination with other agents.
The first of the protease inhibitors, saquinavir (Invirase) was approved in December. Ritonavir (Norvir) was approved this winter, and indinavir (Crixivan) followed a few weeks later. Because of problems obtaining adequate blood levels, saquinavir appears to be less potent than other two, resulting in viral load reductions more comparable to those caused by drugs like AZT. Some of the most exciting news from the 3rd Conference on Retroviruses and Opportunistic Infections held in Washington in January involved the ritonavir and indinavir. It was standing-room-only at the oral presentations, and you even had to wait your turn to read the posters. Once the fuss over Magic Johnson had subsided, even the press caught on, and for once their enthusiasm may not be not have been exaggerated.
Preliminary data from an ongoing study of combination therapy using indinavir were presented. 97 patients with at least six months of prior AZT therapy, CD4 counts between 50 and 400, and viral loads above 20,000 were randomized to receive indinavir, AZT + 3TC, or the combination of all three drugs. Those who received AZT + 3TC had an average reduction in viral load of 1.4 logs, which had decreased to 0.6 logs by 24 weeks. (Now that we're measuring viral loads, the term "log" is getting thrown around a lot. A log just means a 10-fold difference. For example, a reduction in viral load from 100,000 to 10,000 is a "one log" reduction. A reduction from 100,000 to 1,000 is a "two log reduction." Thus, a one log reduction in viral load means that the viral load has decreased by 90%; a two log reduction is a 99% reduction; and a three log reduction is a 99.9% reduction, etc.) In contrast, those on indinavir monotherapy had a 1.6 log reduction in viral load (dropping to 1.3 logs at 24 weeks), and those on the triple combination regimen had a greater than 2 log reduction, which was sustained at 24 weeks. Over 85% had a reduction in viral load below the limit of detection by the Roche assay (500 copies/mL), compared to 40% and 20% in the indinavir and AZT/3TC arms, respectively. Participants in the two indinavir-containing arms had CD4 cell count increases of approximately 100, compared to 30 to 40 in the AZT/3TC arm.
Results of a study of indinavir monotherapy versus AZT/ddI versus the combination of all three drugs were also presented. The study was conducted in 79 patients who had never received antiretroviral therapy and who had CD4 cell counts less than 500 and viral loads above 20,000. Those taking triple drug therapy had a maximum 2.9 log reduction in viral load, which was sustained at approximately 2.5 logs throughout the 24 week study. Only 20% of the patients randomized to the AZT/ddI arm had a 2 log decline. 59% of the patients on triple therapy had viral loads below the limit of detection (200 copies/mL in this study) at some point during the trial, compared with 10 to 20% in the other arms.
In another study patients were randomized to receive either ritonavir or placebo in addition to whatever antiretroviral therapy they were already taking, if any. Participants had viral loads over 15,000 copies and CD4 counts less than 100, and must have had over nine months of prior antiretroviral therapy with approved agents, although current therapy was not required. Over 1000 patients participated in this study, which demonstrated significant differences in clinical progression or mortality as early as 10 days. There was a significant reduction in new AIDS-indicator conditions or death with ritonavir compared to placebo (16% versus 33%), as well as a significant reduction in mortality, with 26 deaths (4.8%) among ritonavir recipients compared to 46 (8.4%) in those on placebo. Viral load measurements were available in a subset of 168 patients. There was a maximum decrease of 1.3 logs in the ritonavir group, which peaked at 2 weeks, and returned to 0.6 logs at 16 weeks. (There was no change in patients on placebo). Both CD4 and CD8 cell counts increased in patients taking ritonavir, with and increase of 40 and 200-300 cells/mL, respectively. In contrast, CD4 cell counts did not change in placebo recipients, and CD8 cell counts declined by 50 cells.
There was little discussion of saquinavir at the Washington meetings. Although the drug has been approved by the FDA, enthusiasm has been tempered by its poor bioavailability, which results in lower efficacy compared to other protease inhibitors. However, there is evidence that bioavailability and efficacy can be improved by combining it with ritonavir, and further clinical studies using this combination are planned.
The protease inhibitors should always be used in combination with nucleoside analogs, since resistance develops more rapidly with monotherapy. This also means that simply adding a protease inhibitor to a failing nucleoside regimen is not a great idea, because it's essentially the same as using the protease inhibitor alone. A better idea, in people who have been on the same antiretrovirals for a long time and have falling T-cells or a high viral load would be to add a protease inhibitor and change to a new nucleoside regimen at the same time.
Ritonavir is now available by prescription. The dose is six capsules (100 mg each) twice daily with meals. Many people experience nausea, stomach upset, or diarrhea on that dose, and Abbott recommends that you start with three twice a day and increase to six as tolerated. Don't take too long to get to the full dose, though, as it's clear that with protease inhibitors, taking too little, or taking it intermittently, is a sure way to develop resistance. There are a number of commonly used medications that cannot be taken if you are on ritonavir so it's important to go through your medication list with your doctor. Indinavir is approved, but in shorter supply, and for now is only available through Statlanders Pharmacy (800-927-8888), which provides the drug by mail. The dose is 800 mg three times daily. There are fewer drug interactions, and the drug seems to be a bit easier to take, but it's important to drink a lot of water when you're taking indinavir, to avoid getting kidney stones.
A number of questions remain unanswered. What are the best nucleoside or nucleosides to use with these agents? At what CD4 count or viral load should therapy be initiated or changed? How will cross- resistance between protease inhibitors affect later therapy? In other words, does the development of resistance to one protease inhibitor lead to resistance to others? That appears to be the case with ritonavir and indinavir. Whether it's true for the other protease inhibitors isn't clear yet. Finally, how will we pay for double and triple-drug combinations containing these expensive new agents? The enthusiasm and renewed optimism generated by the availability of the protease inhibitors is certainly justified, but we'll need answers to those questions as we start to use them.
Summary
The wealth of new data and the growing number of antiretroviral agents are cause for both optimism and confusion. Things were certainly easier a few years ago when the rule was "CD4 less than 500, start AZT," but no one is suggesting that we go back. And while there is a call for more therapeutic guidelines to help physicians care for people with HIV infection, it's likely than any guidelines written today would be obsolete tomorrow.
Rather than try to come up with guidelines myself, I'll just give you my opinion, as of February 1996. In terms of the nucleosides, I think that ACTG 175 and the Delta trial were large enough and powerful enough to allow us to conclude that AZT by itself is not enough. However, because new drugs have been approved since those trials were conducted, we can't really say what's best as first-line therapy. There are certainly strong data to support using AZT/ddI or AZT/ddC. For someone wanting a combination with a lot of data behind it and who doesn't mind taking pills, I might pick one of those. But both cause neuropathy, and ddI is notoriously inconvenient and unpalatable. For those who want something a little easier, there's AZT/3TC. If you asked me which is better, I wouldn't be able to say, since they've never been compared. AZT/3TC is being used as first-line therapy, which is a very reasonable option. On the other hand, some would argue that it would make more sense to start with something like AZT/ddI or AZT/ddC, since you could always take AZT/3TC later after resistance had developed. Finally, for people who really hate taking medications and want something easy and convenient, there's always d4T by itself.
Of course, there are other combinations of those five nucleosides that you can come up with, if you don't mind flying by the seat of your pants. For example, we don't know much about using 3TC with nucleosides other than AZT. Preliminary data suggested that 3TC might actually make you resistant to ddI, but it now appears that combining 3TC with ddI or d4T might work, though perhaps not as well as with AZT. Even d4T and ddI are now being combined with good results and without added toxicity, even though we used to advise against combining two "d-drugs" because of the concern over neuropathy.
What about protease inhibitors? Saquinavir was first, but since it's less potent than ritonavir and indinavir, which are now available, I see little reason to use it, at least until Roche develops a new formulation that is better absorbed or until we learn more about combining it with other protease inhibitors. There's no question that right now you can't do any better than combining either ritonavir or indinavir with at least one nucleoside (see my discussion of nucleoside choices above). But that doesn't mean that everyone with HIV infection should rush out and start triple combination. The final question involves timing.
When do we start antiretroviral therapy, and when do we change agents? Is starting after the CD4 falls below 500 as obsolete as AZT monotherapy? It probably is, now that we have the ability to measure viral load, which allow us to tailor treatment to the individual, rather than relying on averages based on large clinical trials. Using viral load measurements (bDNA PCR or RNA PCR) to make treatment decisions is rapidly becoming standard practice, but we still need more information on how to use these as yet unlicensed tests. For example, someone with a very viral load (over 100,000 copies/mL) or with a dramatic rise in viral load over time is probably a candidate for therapy regardless of CD4 count. Similarly, if virus is undetectable by one of the standard viral load assays, therapy may not be necessary, even if the CD4 count is below 500. The decision becomes harder with counts in the middle range, since such counts may simply represent laboratory variation. Fortunately, clinical trials all incorporate viral load measurements now, so with time we should get the information we need.
Obviously, these are just opinions, not official guidelines. And they are opinions that may change rapidly with the emergence of new data and new drugs. I'll try to keep this update up-to-date, as the new information comes in.
