The substantial recent developments in AIDS therapy and in our ability to monitor viral dynamics have led to a major shift in the thinking about HIV therapy. Several panels have convened to reconsider the best treatment strategies for HIV infection in the face of these developments. Some of the panels making recommendations have included one convened by the Department of Health and Human Services (DHHS) and the Henry J. Kaiser Family Foundation, a USA Panel convened by the International AIDS Society, and one convened by the British HIV Association. Because of the rapid development of new therapies and in the ability to monitor the viral load, the recommendations have been based to a large degree on the principles of therapy articulated earlier and on small or short-term clinical trials with laboratory end points. It will probably be years before randomized trials with clinical end points are conducted to evaluate many of these recommendations, and in a number of cases these trials may never be undertaken.

Although there are some differences between the recommendations made by these panels, overall they were quite similar. The recommendations articulated here apply to adults and adolescents infected with HIV and are largely derived from those of the DHHS-Kaiser panel. It is anticipated that these recommendations will evolve over the next several years, and physicians treating HIV-infected patients are strongly urged to keep alert for such changes. Updated recommendations from the DHHS-Kaiser group are to be posted by the AIDS Treatment Information Services Website at Internet Uniform Resource Location http://www.hivatis.org.

Once it is established that a patient has HIV infection, the two most useful tests for guiding therapy are the plasma HIV RNA level (viral load) and the CD4+ T-cell count. Both of these tests should be performed at the time of diagnosis. The viral load should subsequently be performed every 3 to 4 months, whereas the CD4 count should be measured every 3 to 6 months. Plasma RNA levels can be acutely affected by immune stimulation, and for this reason they should generally not be measured within 4 weeks after successful treatment of intercurrent infections, resolution of symptomatic illnesses, or immunizations. Also, there are some differences in the assays used to measure viral RNA; the viral load assessments obtained with branched DNA technology are generally about one half of the magnitude of those obtained with reverse transcriptase polymerase chain reaction (RT-PCR) technology. For this reason, one assay type should ideally be used throughout in following a given patient. Guidelines given here will be keyed to the results of RT-PCR assays. Ideally, decisions of when to institute therapy should be made based on two measurements of CD4 count and viral load, except in those patients with advanced disease, in whom the dangers of a delay in treatment associated in repeating the test may outweigh the advantages. Other evaluations that should be undertaken in patients before initiating therapy include a complete history and physical examination, a complete blood cell count, and a chemistry profile. Also, if not already performed, a VDRL, a tuberculin skin test, Toxoplasma IgG serology, a gynecologic examination with a Papanicolaou smear, and other clinically appropriate tests should be obtained.

Considerations for Initiating Antiretroviral Therapy
The decision to undertake antiretroviral therapy in patients is an important one and should only be made in the setting of careful patient counseling and education. It has clearly been shown that antiretroviral therapy can benefit patients with advanced HIV infection and immunosuppression, and such patients should generally be treated. The decision to treat asymptomatic patients with established HIV infection, however, should be individualized based on a careful consideration of the patient’s disease status and the likelihood of compliance with therapy. Therapy of such patients is based on the principle that continued viral replication is always harmful. In this context, the potential benefits of preventing disease progression by initiating therapy must be weighed against the risks of drug toxicities, the inconvenience of the treatment regimens, and the risk of selecting for resistant strains of virus. Non-compliance with drug regimens can create periods of time during which there is incomplete HIV suppression and hasten the emergence of resistant strains, thus both thwarting the effectiveness of the present regimen and reducing future therapeutic options. Thus, an important factor in the decision to initiate antiretroviral therapy is the likelihood of patient adherence to the prescribed regimen after counseling and education.

There are two general approaches to initiating therapy in asymptomatic patients with established HIV infection. The more aggressive approach is heavily based on the principles of therapy discussed earlier and involves offering therapy to all patients with less than 500 CD4+ cells/mm³ or with 20,000 HIV virions/mL (by RT-PCR). The more conservative approach would observe patients with 350 to 500 CD4+ cells/mm³ and less than 20,000 virions/mL. Although there is currently a move toward the more aggressive strategy, it is worth remembering that in every case the patient should make the final decision on the acceptance of therapy after discussion of the issues concerning his or her own clinical situation.

The DHHS-Kaiser panel recommended that for patients with over 500 CD4+ cells/mm³ and less than 20,000 virions/mL, observation alone is reasonable to consider. They noted, however, that some experts would treat such patients. It is worth noting that these recommendations are based largely on the principles of therapy presented earlier. Over the years, the recommendations of when to begin therapy for HIV have fluctuated back and forth, and a prior trend to treat most patients with fewer than 500 CD4+ cells/mm³ with zidovudine was subsequently modified by the results of a large randomized study (the Concorde trial) showing that such early zidovudine monotherapy did not yield an improvement in survival. The present recommendations are backed by a substantially greater understanding of disease pathogenesis and involve more potent regimens, but they may again be modified as the field evolves.

A number of patients experience an acute retroviral syndrome at the time of their HIV infection. The symptoms of this syndrome often include fever, sweats, lymphadenopathy, pharyngitis, and myalgias. Physicians should be alert for the possibility of HIV infection in such patients and obtain appropriate laboratory tests, including plasma HIV RNA. Treating such patients has the potential of decreasing the spread of HIV through the body, reducing the viral load in the asymptomatic period after the acute infection, and potentially reducing the rate of viral mutation. With this in mind, it is generally recommended that patients with acute HIV infection be offered antiretroviral therapy. Some experts also recommend treating patients in whom seroconversion has been documented to have occurred within the past 6 months. In either case, it is recommended that antiretroviral therapy be utilized for at least 1 year, and some experts would continue the therapy indefinitely.

Regimens for Initial Antiretroviral Therapy
The goal of therapy should be to suppress HIV replication to undetectable levels. This will provide the greatest protection against further immune destruction and permit some degree of immune reconstitution to occur. To accomplish this, potent regimens should be used with each drug given at full dose if possible. This is particularly important when instituting therapy for the first time. The initial therapy provides the single best opportunity to achieve a prolonged suppression of HIV replication, at least below the limits of detection, and thus a delay in (or even prevention of) the emergence of resistant strains. As noted earlier, there is substantial cross-resistance among drugs within each of the classes, and for this reason it is much harder to achieve effective HIV suppression once resistance has emerged to this first regimen. It is thus essential that the initial regimen used be one that is very likely to yield complete suppression and that patient compliance is maximized by the patient wholeheartedly undertaking the therapy after careful education and counseling.

At present, the preferred regimens for this initial therapy generally involve the use of two NRTIs and either a potent single PI or efavirenz (

Table 418-4) . In general, one of the NRTIs should be thymidine-based, while the other should not be thymidine-based. However, the combination of stavudine and zalcitabine, both of which can cause peripheral neuropathy, is not recommended (

Table 418-5) . The role of abacavir for initial therapy is expected to be considered in the near future. With regard to the PIs, all are considered preferred except for the current hard-capsule formulation of saquinavir (because of its relatively poor absorption and low plasma levels). However, the combination of saquinavir and ritonavir is acceptable in the initial regimen. Alternative regimens that can be considered acceptable but are generally considered less desirable include two NRTIs along with either nevirapine or delavirdine.

The International AIDS Society USA panel suggested that use of two NRTIs was also acceptable as a first-line regimen, especially in asymptomatic patients with a relatively lower risk of progression. However, although such a regimen can provide clinical benefit, a randomized trial with clinical end points has since shown better short-term results with a three-drug regimen and there is now a movement away from this approach. Another concern with the use of only two NRTIs as initial therapy is that sustained viral suppression is generally not achieved and thus resistance is not optimally suppressed. The choice of therapy in patients with more advanced HIV disease can be complicated by the potential for drug interactions with their other medications and their greater susceptibility to certain toxicities (such as bone marrow suppression). A particular problem in this regard is the effects of PIs and NNRTIs on the cytochrome P-450 pathway. For this reason, antiretroviral therapy should be initiated in any patient only after a careful drug history is obtained (including over-the-counter and alternative medicines), and patients on antiretroviral therapy should be carefully instructed not to change any medications without discussing the change with their physician or other health care provider.

Patients should be carefully instructed to take their antiretroviral therapy as instructed once it is initiated. Suboptimal doses, omitting one of the drugs in a regimen, or missed doses all can lead to suboptimal viral suppression and increase the likelihood of resistance emerging. There is a misconception that once antiretroviral therapy is started, it can never be stopped or resistance will emerge. Actually, if all the drugs are stopped simultaneously, there is no longer evolutionary pressure for resistance to develop, and in fact, this is the preferred strategy if one or two of the drugs need to be stopped for any extended period of time.

Changing Antiretroviral Therapy
There are three main reasons to consider changes in antiretroviral therapy: (1) for drug toxicity or incompatibilities, (2) for regimens that are believed to be suboptimal, and (3) for a failure to achieve adequate suppression or a rebound in viral load after a period of complete suppression. In the case of patients who need to change therapy because of toxicity to a particular drug, it is acceptable to substitute an appropriate alternative drug of the same class. However, for patients who have virologic drug failure, it is desirable to change at least two and preferably three new drugs whenever possible to maximize the likelihood that complete suppression will again be attained. Before changes in regimen are undertaken, however, it is important to assess compliance with the regimen and determine if poor compliance is the principal reason for failure.

The viral load should be measured immediately before initiating antiretroviral therapy, at 4 to 8 weeks after the initiation of therapy, and then every 3 to 4 months. In most patients, a potent regimen will result in a decrease of 0.5 to 0.75 log₁₀ viral load by week 4 and 1 log₁₀ by week 8. Plasma HIV RNA should be undetectable using the most sensitive assay available by months 4 to 6 after therapy is initiated. A failure to meet any of those standards suggests that the therapy is not working optimally and consideration should be given to a change in regimen.
Other factors that suggest that the regimen should be changed include the re-emergence of detectable plasma HIV RNA (especially if over 5000 copies/mL) after suppression to undetectable levels; a significant (threefold or greater) increase in the viral load not explained by intercurrent infections; or a persistent decline in the CD4 count. When the decision to change therapy is based on viral load determination, it is preferable to have a second confirmatory viral load test. Many experts believe that patients receiving therapy with only two NRTIs should have a change in therapy (rather than simply adding a PI), even if their virus is undetectable, because of their likelihood of subsequent virologic failure. Finally, clinical deterioration (such as the development of a new opportunistic infection) may prompt a reconsideration of the therapy. However, such an event may merely represent the prior immune compromise of the patient and thus may not necessitate a change in antiretroviral therapy.

In patients who have been on one or two NRTIs alone, physicians can consider switching therapy to two new NRTIs along with a potent PI (or with ritonavir and saquinavir), a new NRTI along with an NNRTI and a potent PI, or even two PIs along with an NNRTI. In patients who have been on a three-drug regimen with two NRTIs and a PI, the choice of a second-line salvage regimen depends in part on the initial PI used. It is generally desirable to select a new regimen with two new NRTIs and either a new PI, a combination of a new PI and an NNRTI, or even two new PIs (usually saquinavir plus either ritonavir or nelfinavir). It is usually best to avoid switching from ritonavir to indinavir or vice versa in the new regimen or from indinavir to nelfinavir because of the high level of cross-resistance.

Similarly, it is best to avoid changing among nevirapine, delavirdine, and etavirenz. For certain patients, one or more drugs newly licensed or available on expanded access programs may also be worth considering after the physician has studied the agents. Experience with some of these regimens, or with other four-or-more drug regimens that some physicians are using, is limited, and there remains the real possibility of unexpected interactions and results. Genotypic analysis of HIV resistance mutations may help select regimens in certain patients. However, the value of this approach in various settings remains to be fully established. There are also limited data on the value of restarting drugs to which patients have become resistant. There is some evidence that non-resistant viral strains can gradually predominate when a drug to which a patient’s HIV has become resistant is stopped. However, many copies of the resistant virus can remain in proviral form and resistant virus can very rapidly re-emerge when the drug is restarted.

Many patients have but limited options for new regimens of desired potency, and in some cases it may be rational to continue suboptimal therapy if partial viral suppression is obtained. Because of the limitations imposed by patterns of resistance, intolerance, or toxicity, some regimens that would be deemed suboptimal for initial therapy may be quite appropriate as second-or third-line regimens, especially in patients with late-stage disease. Indeed, it may be rational to withhold therapy altogether for some patients with no viable treatment options.

The experience of many physicians is that once viral strains become resistant to an initial therapy, the success of subsequently administered therapies is rather limited. Even if suppression of the viral load to undetectable levels is attained, it is often relatively short lived. This is one reason why it is so important that the initial regimen be carefully chosen and followed. Many physicians currently have a very low threshold for sequentially changing regimens in the face of persistent viral replication. A real danger of this approach is that even with 13 approved drugs, patients can rapidly use up their therapeutic options. Any regimen or change in regimen must be undertaken with attention to the effect that this decision will have on subsequent therapeutic options. In patients in whom one or more antiretroviral regimens have failed, treatment can be very challenging, and it is important for both the physician and the patient to have a realistic expectation of what can be accomplished. When possible, it is recommended that the decisions to change therapy and the design of new regimens be undertaken with the assistance of a clinician who is experienced in the care of HIV-infected patients.