Antiretroviral treatments have reduced morbidity and mortality from acquired immuno-deficiency syndrome (AIDS), but treatment guidelines need to be updated periodically as new drugs become available and new strategies emerge to control replication of human immunodeficiency virus (HIV). Yeni and associates of the International AIDS Society–USA Panel provide information about when to start treatment, which medications to use, and when and how to change medications.

The first antiretroviral-treatment panel issued recommendations in 1995; new data were used to update the recommendations, which reflect a change in the threshold for initiating antiretroviral therapy. In general, antiretroviral therapy should not be delayed until the CD4 cell count is 200 per mm³ (200 × 10⁶ per L) or less. However, the CD4 cell level at which to begin therapy is not as clear. Some studies have shown that response is less favorable if treatment is delayed until the count is 200 per mm³. Specifically, it seems to be more difficult to achieve a slower increase in CD4 cells and potentially more difficult to reduce the viral load. In addition, some illnesses, such as tuberculosis and pneumonia, can occur even when the CD4 cell count is above 200 per mm³. On the other hand, the risk of three-year clinical progression is low in patients whose CD4 cell counts are above 350 per mm³ (350 × 10⁶ per L), and the risks of antiretroviral treatment might be greater than the benefits.

In summary, the panel recommends that treatment be started in treatment-naive patients who are symptomatic and those who are asymptomatic but whose CD4 cell counts are no greater than 200 per mm³. Patients with CD4 cell counts above 200 per mm³ should discuss with their physician the risks and benefits of treatment. The CD4 cell count and the rate of decline (with a decline of more than 100 cells per mm³ per year considered high, even if the absolute count is greater than 350 cells per mm³), the viral load (a high load is above 50,000 to 100,000 copies per mL), and the patient's risk of poor tolerance to the medications should be considered. Treatment of serious opportunistic infection might take precedence over initiating anti-retroviral medications.

Treatment failures are common in patients with HIV and should be expected. Patients probably will need to be switched to a different regimen at some point, although it is unclear whether there is an optimal order for administering nucleoside reverse transcriptase inhibitors (NRTIs). The accompanying table lists possible drug regimens after initial failure. The preferred initial regimen will generally be one of the following: (1) a protease inhibitor (PI) with two NRTIs (with or without ritonavir); (2) a nonnucleoside reverse transcriptase inhibitor (NNRTI) plus two NRTIs; or (3) three NRTIs. The NNRTI and double-NRTI regimen can be initiated with one of the NNRTIs that are approved in the United States (nevirapine, delavirdine, and efavirenz). Direct comparisons of these drugs do not exist, but some physicians prefer efa-virenz because it can be given once daily. However, efavirenz should be avoided in women who are pregnant or who might become pregnant. Studies have shown that the triple-NRTI regimen (specifically, abacavir/zidovudine/lamivudine) is associated with a significant decrease in viral load at 48 weeks. However, if the pretreatment viral load was more than 100,000 copies per mL, the triple-NRTI regimen is less likely to cause complete viral suppression. Consequently, a patient with a high viral load or low CD4 cell count should not begin treatment with this regimen. Although the triple-NRTI regimen is typically well tolerated and has a low pill burden, about 3 percent of patients have a severe hypersensitivity reaction to abacavir. Treating patients with drugs from all three classes typically is not recommended because it heightens the risk of multiclass drug resistance and might reduce future treatment options. It might be useful in patients with advanced disease and high short-term mortality risks or in patients who have a drug-resistant strain if testing indicates that it might be effective.

Antiretroviral therapy should be monitored closely. Compliance should be checked as closely as possible, because every 10 percent deviation from total compliance is associated with a doubling of the viral load. Decreases in viral load and increases in CD4 cell counts are used to assess response to treatment. If the therapy is successful, the CD4 cell count usually increases by more than 50 cells per mm³ approximately one to two months after treatment begins. If the CD4 cell count exceeds 200 per mm³ for three to six months, some prophylactic medications can be stopped, because the risk of opportunistic infection is decreased.

Viral load is considered to be significantly improved when there is a 30 to 50 percent decrease, and therapy is considered effective when there is a 90 percent reduction in viral load after one month of treatment. Viral load and CD4 cell counts should be measured at four, eight to 12, and 16 to 24 weeks after treatment starts. Once viral loads are undetected for two sequential readings, the viral load and CD4 cell count can be monitored every two to three months. Small increases in detectable plasma virus, in the range of 50 to 400 copies per mL, are not considered predictive of treatment failure.

The standard of care in those who experience treatment failure is drug-resistance testing. It is unclear which type of testing is best, but testing should be done while the patient is still on the failing regimen (before withdrawal of the drug minimizes the chance of determining which drug is associated with a suboptimal response). Various other factors, such as compliance issues or toxic effects, could be reasons to change the medications used. It is reasonable in these cases to change individual components of the regimen (e.g., substituting stavudine for zidovudine). However, if there are acute toxic effects and it is unclear which medication is responsible, treatment should cease, at least temporarily, while symptoms of toxicity or adverse effects resolve. Changes should be made quickly to minimize development of multiple drug resistances, which will limit future treatment options. If a virologic failure leads to a decision to change treatment, at least two—and preferably three—medications should be chosen. If an NNRTI-containing regimen fails, the likelihood of NNRTI-class cross-resistance is very high. The likelihood of PI-class cross-resistance is not as predictable. A patient who has experienced multiple treatment failures will be difficult to treat, and the goal in these patients is to prevent clinical deterioration. In some cases, mega-HAART therapy (use of six or more medications) can cause increased antiretrovi-ral activity. However, toxic effects will be greater, and cost and noncompliance might also be higher.

Attempts to alter the immune response (e.g., with interleukin-2 or cyclosporin A, among others) have not been sufficiently studied to permit recommending them.