Controversies in Antiretroviral Therapy

Managing HIV infection is like a chess game in some ways: Strategy is a key component of a successful attack on the virus. Unfortunately, the rules of the game keep changing. As much as we'd like to sometimes, we can't undo most of the moves that we've already made. In 1998, what should our first moves be? Should we start our attack now or watch and wait? Should we go for an all-out attack or should we hold back a little and keep something in reserve? If we go for an all-out attack, can we back off after a little while? Or, if we choose to keep some of our forces in reserve, can we add them on successfully if the initial attack isn't strong enough? How should we plan to move ahead if our first attack fails? Without purporting to answer these controversial questions, this article will expand on these important clinical issues.

When to Start Treatment

The lack of consensus about when to start antiretroviral therapy for someone with established HIV infection is evidenced by the differing guidelines established by various committees of experts. The need to start treatment is never an emergency, though in some circumstances it can be relatively urgent. Note that the ensuing discussion does not include persons with acute HIV infection -- that is, persons who have documented HIV in their bloodstream due to very recent infection and who have not yet seroconverted (developed antibodies to the virus). Most experts believe that persons with acute HIV infection, and perhaps those within six months of documented seroconversion should be treated. In most situations, however, the decision need not be rushed and observing the trend in a person's viral load and T-cell count over time can be of great help. The bottom line is that a number of factors have to be weighed and the decision individualized (see Table 1). Guidelines are only guidelines.

Table 1 Risks and Benefits of Early Initiation of HAART in Asymptomatic Patients*
Potential Benefits delayed progression to AIDS and prolong life controlled viral replication and prevention of the emergence of resistant mutations prevention of progressive damage to the immune system and reconstitution of a normal immune system decrease in the risk of drug toxicity	Potential Risks reduced quality of life from side-effects and inconvenience of drug regimens earlier development of resistance limitation of future treatment options due to the development of resistance unknown long-term toxicity of antiretroviral drugs unknown duration of effectiveness of current therapies
*Adapted from Guidelines for the Use of Antiretroviral Agents in HIV-Infected Adults and Adolescents, Department of Health and Human Services, June 17, 1998.

The main reason to consider early treatment of established HIV infection is to delay the onset of AIDS and prolong life. We know that antiretroviral therapy has these clinical benefits in persons with advanced HIV disease (T-cells less than 200) but they have not been demonstrated in persons with less advanced disease to date. This doesn't mean that treatment won't have these beneficial effects -- we just don't know yet. There are a number of theoretical reasons to believe that early treatment will be beneficial. These arguments, as well as the opposing viewpoint, are summarized below.

A major theoretical argument in favor of early treatment is that controlling viral replication early on may prevent the emergence of HIV which is resistant to drugs. Each time HIV replicates (divides) in the body, there is a chance that mistakes will be made in copying its genetic material (RNA). Some of these mistakes, called mutations, may give HIV the ability to divide in the presence of certain antiretroviral drugs. These resistance mutations can develop even before a person starts on antiretroviral drugs. The argument goes that the sooner a person starts on therapy, the less likely they are to develop a significant number of these resistance mutations and therefore the greater the likelihood that the drug regimen will be able to successfully suppress HIV replication.

Another important argument in favor of early treatment relates to the fact that HIV gradually damages the immune system. The earlier the infection can be controlled, the more likely that immune function can be preserved. Similarly, there may be a better chance of restoring immune function (immune reconstitution) if someone starts treatment early on. Specifically, the body's immune response to HIV may be preserved or restored by early treatment. The presence of a strong immune response to HIV is felt by many researchers to be a key distinguishing factor between so-called long-term nonprogressors -- persons who have been infected with HIV for at least 7 to 10 years and have had little decline in T-cell counts -- and persons whose HIV infection has progressed. So restoring this immune response is felt to be desirable.

On a more practical note, persons with less advanced HIV infection tend to tolerate antiretroviral therapy more easily. Fewer side effects means that persons will be more likely to adhere to and continue on their drug regimens, which should increase the likelihood of long-term success.

There are also a number of arguments against early initiation of antiretroviral therapy. A key reason to question hitting the virus early is the unproven clinical benefit mentioned previously. The average time from initial infection with HIV to the development of AIDS is nearly 10 years. The longest anyone has been on combination antiretroviral therapy is closer to three years. Although the potential to keep viral replication suppressed for a few years using combination therapy has been realized, we don't know how long this suppression will last. If viral replication is completely suppressed, resistance may not develop since replication is required for new mutations to develop. However, it is not known whether complete suppression of replication happens even when viral load is less than 40 or 50 copies per milliliter, since there may be reservoirs of replicating virus which cannot be detected with the available technology.

Another concern about early initiation of antiretroviral therapy pertains to the risk of running out of treatment options in the event of failure of the initial regimen. The risk of progression to AIDS within five years for someone with a high T-cell count and low viral load is actually quite low. Such a person is likely to do well on antiretrovirals from the viral load standpoint, at least in the short term, since they are starting out with a low viral load. However, long-term adherence to a complex drug regimen may be more challenging for someone who has no symptoms to begin with and who may develop side-effects and a reduced quality of life from the medications. If such persons develop resistance to their initial regimen early in the course of their HIV disease due to poor adherence, there is a risk that they may end up switching to multiple alternate drugs and eventually running out of drugs at a time when they are clinically well, leaving no options for the future until new classes of drugs are developed.

Another possible reason for considering deferring antiretroviral therapy is the unknown potential long-term toxicities of the antiretroviral drugs. Alterations in blood sugar and lipid levels, abnormal fat redistribution and the occurrence of heart disease and stroke in young persons were not recognized as possible adverse effects of protease inhibitors until the drugs were widely used for close to two years. Perhaps other long-term toxicities will be recognized in the future.

In summary, there are a number of compelling theoretical reasons to support early initiation of antiretroviral therapy. On the other hand, there are important reasons to consider deferring therapy in healthy persons with low risks of progression to AIDS (i.e. high T-cell counts and low viral loads). Few would argue with the need to start therapy in someone with symptoms or who is at high risk of progression in the short term.

What to Start Treatment With

Although there are a large number of drug combinations to choose from, the current debate centers around whether to initiate therapy with a protease inhibitor-containing regimen or a protease-sparing regimen. The increasing recognition of metabolic complications attributed to (but not definitively proven to be caused by) protease inhibitors coupled with the promising results of studies of efavirenz (Sustiva®, formerly called DMP 266) have fueled this debate.

These metabolic complications, including diabetes, high cholesterol and triglyceride levels, and fat redistribution were discussed in detail in the Summer 1998 issue of CRIA Update and will not be elaborated on here. The efavirenz data are worth mentioning, however, since they are from the first large study comparing protease-sparing regimens to one of the "gold standard" triple combinations with a protease inhibitor.

Efavirenz is a potent non-nucleoside reverse transcriptase inhibitor (NNRTI). In the ongoing Dupont Pharmaceuticals 006 study, in which CRIA is participating, eligible patients were either antiretroviral naive or nucleoside analog experienced, with the exception of 3TC. Their viral loads at entry were at least 10,000 copies/ml and their T-cell counts greater than 50 cells/mm³. Patients were randomly assigned to receive one of three combinations: AZT + 3TC + efavirenz; efavirenz + indinavir; or AZT + 3TC + indinavir. Data from the first 450 patients who completed 24 weeks of follow-up were presented at the 12th World AIDS Conference in July 1998. The results of this interim analysis, summarized in Table 2, suggest that the protease inhibitor-sparing combination of AZT + 3TC + efavirenz is superior to the standard triple combination of AZT + 3TC +indinavir.

Table 2 DMP 266-006 (efavirenz) Study: Percent of Patients with Viral Loads Less Than 400 Copies/ml at 24 Weeks
Type of Analysis*	AZT + 3TC + Efavirenz	AZT + 3TC + Indinavir
Intent to Treat
Noncompleter = Failure	74.7%	56.2%
Last Observation    Carried Forward	86.4%	64.9%
Observed Data	94.5%	88.6%
*The data were analyzed in three different ways. The intent-to-treat analyses attempt to account for patients who did not complete 24 weeks of treatment. Since patients may have dropped out of the study because they were not responding to treatment, simply ignoring these drop-outs, as is done in the "observed data" analysis, may make the regimens appear better than they really are (because those who did not respond are not counted). The "noncompleter = failure" is a conservative analysis in which patients who did not complete 24 weeks are assumed to have had viral loads greater than 400 at 24 weeks. The "last observation carried forward" analysis takes the last viral load that a non-completing patient had before s/he dropped out of the study and uses that value for the viral load at 24 weeks. For example, if a patient had a viral load of 10,000 at week 16 of the study and dropped out at that time, his/her viral load at 24 weeks is assumed to also be 10,000.

The relative ease of taking efavirenz compared with indinavir may partly account for the apparent superiority of the triple combination with efavirenz in this study. No clinical trial is flawless, and this study has been criticized for several reasons. First, it is an open-label study rather than the traditional double-blinded type. In other words, both the investigators and participants knew which drugs the participants were taking, which can be a source of bias in a study. An additional concern is that the drop-out rate in the triple combination arm with indinavir was quite high at 37.8% compared to 20.8% in the triple arm with efavirenz. Nonetheless, it appears as if AZT + 3TC + efavirenz is at least as effective at reducing viral load and increasing CD4 counts as AZT + 3TC + indinavir at 24 weeks. The durability of these responses is not yet known, but this protease-sparing combination may be a viable option for initial therapy.

It is also not yet known whether the metabolic complications seen with protease inhibitors will also be seen with potent protease-sparing regimens. But the possiblity of reserving protease inhibitor therapy for patients whose initial regimens may fail is an attractive one to many physicians. Along the lines of reserving classes of antiretrovirals, the availability of abacavir (Ziagen®, formerly called 1592U89), a potent nucleoside analog, will provide an option allowing for the preservation of both the protease inhibitors and NNRTIs as future treatment options.

Also, the triple nucleoside analog combination of abacavir + AZT + 3TC has been shown to be potent in early clinical trials and has the advantage of being a simple regimen of two pills twice a day if the Combivir® forumulation, which combines AZT and 3TC into one pill, is used. One unresolved controversy relates to whether it is an advantage or disadvantage to start with such a regimen of three nucleoside analogs. Although this strategy may preserve other classes of drugs, there is concern that attacking the same part of the virus (the reverse transcriptase enzyme) in the same way may be less desirable than attacking different parts of the virus by using different classes of drugs. In the analogous model of cancer treatment, combinations of drugs which attack different targets of the malignant cells are superior to those attacking similar targets. An ongoing study comparing AZT + 3TC + abacavir to AZT + 3TC + indinavir should shed light on this important issue.

Aside from the decision about which drugs to start treatment with, another unresolved issue is how many drugs to use. Some authorities recommend using more than the standard three drugs for patients with high viral loads. A number of studies are addressing this question with comparisons of different combinations, like four versus three drugs.

Induction-Maintenance Strategy

A treatment strategy termed induction-maintenance has been under study recently, though the initial trials of this approach have not been successful. With this strategy, borrowed from the cancer treatment model, an aggressive regimen is used initially to induce a lowering of viral load for a set period of time or to below the level of detection, and then a less intensive maintenance regimen is used to keep the virus suppressed. The AIDS Clinical Trials Group (ACTG) conducted a study of this approach using AZT + 3TC + indinavir as induction for 24 weeks followed by AZT + 3TC, indinavir alone, or continuation of the triple combination. The results were disappointing as 17.3% of patients on AZT + 3TC and 15.8% of patients on indinavir alone had a rebound in viral load compared with only 2.9% of those who stayed on the triple combination. The French TRILEGE study was similar in design and yielded equally discouraging results.

Preliminary results of the Dutch ADAM trial were presented at the 12th World AIDS Conference. In this study, patients were induced for 26 weeks with a quadruple drug regimen consisting of d4T + 3TC + nelfinavir + saquinavir soft gel capsules. Those patients with viral loads less than 50 copies/ml were then randomly assigned to receive either d4T + nelfinavir, nelfinavir + saquinavir, or continue on quadruple therapy. Although only data from 25 patients followed out to 36 weeks were presented, there was a striking difference between the dual and quadruple therapy arms: 9 of 14 (64%) patients on one of the two dual therapies had rebounds in viral load to greater than 50 copies/ml compared with only one of 11 (9%) on the quadruple arm.

An ongoing British trial called ProCom is studying a similar approach using a four- to six-month induction with d4T + ddI + nelfinavir + saquinavir soft gel capsules. Patients with undetectable viral loads will then be randomized to d4T + ddI, nelfinavir + saquinavir, or to continue on the quadruple combination. Although the results to date have been discouraging, the induction-maintenance strategy still may be viable. Its appeal is the potential for simplifying the antiretroviral regimen after a certain period of time. Perhaps the induction phase has just not been long enough in the studies completed thus far.

Intensification Strategy

The opposite approach to induction-maintenance is called intensification. This is another cancer treatment term which refers to the addition of drugs to a regimen when the initial regimen is not yielding the desired result. The Prometheus study used this strategy in comparing the dual protease inhibitor combination of ritonavir + saquinavir to ritonavir + saquinavir + d4T. In this trial, patients in the dual therapy arm whose viral loads were not below 400 copies/ml after 12 weeks of treatment added d4T. The results presented in Geneva were interesting. The triple combination arm was superior at Week 18 but both arms were comparable at Week 48, with approximately 80 percent of patients below 400 copies/ml. In other words, intensification with d4T for patients whose viral loads did not decline sufficiently by Week 12 yielded the same net result as starting treatment with the triple combination. While these results are intriguing, further study is needed before embracing this treatment strategy.

Sequencing of Antiretrovirals

The unfortunate reality is that a significant proportion of people end up with failure of their initial antiretroviral regimens. Many factors may contribute to this, and there are different definitions of failure. It makes good sense to plan in advance for the possibility of drug failure and to choose the first attack based on the possible options for subsequent attacks if needed. This has been alluded to earlier in the discussion of initial regimens -- for example, the initial choice of a protease-sparing combination enables the use of protease inhibitors in so-called salvage regimens. Salvage regimens are the subject of an article in this issue by Tim Horn. The order in which antiretrovirals are used is sometimes referred to as the sequence. A key consideration in planning a sequence is what is the likelihood that specific drugs in a subsequent regimen will work based on the choice of the initial regimen? The likelihood of cross-resistance -- meaning that once HIV becomes resistant to one drug within a class it may also be resistant to other related drugs in the same class -- is one determinant of the potential success of a second or third regimen.

There are other potential reasons unrelated to resistance that could lessen the likelihood of success of a subsequent regimen. One controversial potential reason relates to how nucleoside analogs are processed inside of cells in order to become active drugs. The nucleoside analogs undergo a processing called phosphorylation, where phosphate groups are added onto the chemical structures of the drugs. Results of a very small study of patients who took d4T after having taken AZT extensively in the past suggested that phosphorylation of d4T took place less efficiently in their cells. This observation was felt to partly explain why patients in the European study called ALTIS who had taken AZT in the past did not respond as well to d4T +3TC compared to patients with no prior antiretroviral therapy. In contrast, a recent analysis from the Johns Hopkins Hospital's HIV Clinic Database suggested that prior treatment with AZT did not adversely affect the response rate to subsequent treatment with d4T. This controversy remains unresolved.

Conclusions

As more antiretroviral drugs become available, treatment options expand dramatically. A number of fundamental questions that impact how best to use the available drugs and those which may be approved in the future remain unanswered. Further research into long-term strategies for HIV management is ongoing and will hopefully shed light on these unanswered questions.

Dr. Marshall Glesby is CRIA's Medical Director and a Clinical Instructor at New York University School of Medicine.

Back to CRIA Update Fall 98 Contents Page