Current guidelines from the International AIDS Society-USA (IAS-USA),(10) the U.S. Department of Health and Human Services (DHHS),(12) and the British HIV Association (BHIVA),(13) all recommend ART for individuals with symptomatic HIV infection. This includes patients with an AIDS-defining illness or symptomatic HIV infection without AIDS--defined as the presence of thrush or unexplained fever. However, if the patient is acutely ill with a serious opportunistic infection (OI), ART may, in some cases, need to be delayed until treatment of the OI has been initiated. For example, the clinician may wish to avoid drug interactions (eg, between rifampin and protease inhibitors in the treatment of Mycobacterium tuberculosis),(10,12) conditions during acute illness that may prevent adherence to the antiretroviral regimen (eg, vomiting or inability to swallow), and the risk of immune reconstitution illness.

Immune reconstitution illness occurs most often in patients with advanced immunodeficiency, when the recovery of immune function causes new immunologic and/or inflammatory responses to an existing OI (eg, disseminated Mycobacterium avium complex, cytomegalovirus, and others) causing new symptoms such as high fever, painful lymph node enlargement, or local inflammatory syndromes. Clinical presentations can be atypical compared to the same OI in patients with untreated HIV infection, and most cases improve without discontinuing ART. Some patients, however, may require anti-inflammatory drugs.(14,15)

Although the rate of disease progression among HIV-infected individuals varies greatly,(11) whenever possible, ART should be initiated before the CD4 threshold of 200 cells/µL has been reached.(1,7,8,16) Patients with CD4 counts <200 cells/µL are at greatly increased risk of serious life-threatening OIs,(17) and immune reconstitution illness is largely limited to individuals who initiate therapy with very low CD4 counts.(14) Individuals with CD4 counts >350 cells/µL, on the other hand, have a low risk of clinical progression at 3 years,(11) and concerns about the impact of antiretroviral drugs on quality of life, adherence to regimens, risks of serious adverse effects of the drugs, and emergence of viral resistance that will limit future treatment options generally outweigh the benefits of durable viral suppression.(12,18) The timing of initiation of ART must be individualized in persons with CD4 counts between 200 and 350 cells/µL. There are no sound data in terms of clinical endpoints that define a precise threshold for initiation of ART. However, as detailed in the following paragraphs, several observational studies have evaluated the correlation between CD4 level at initiation of ART and disease progression rates, virologic response, and incidence of serious infections.

First, disease progression to AIDS and death is low if ART is initiated with a CD4 count of at least 200 cells/µL (3-year survival rate 97%), but significantly worse if treatment is delayed until the CD4 count is <200 cells/µL.(7) An evaluation of 13 prospective studies found that patients who started therapy with CD4 counts >350 cells/µL had a 3.4% cumulative risk of progression to AIDS at 3 years compared to 4.7% in patients who started with 200-349 cells/µL, and 16% if CD4 count was <50 cells/µL. The authors conclude that the data might not justify starting therapy until CD4 counts are approaching 200 cells/µL.(19) An analysis of almost 5,000 HIV-infected men (Multicenter AIDS Cohort Study [MACS] trial) considered the correlation of CD4 count and viral load with rate of progression to AIDS. The analysis concluded that it was safe to defer therapy in patients with a CD4 count of 200-350 cells/µL if the viral load was <20,000 copies/mL.(20) Once ART has been initiated, continued prevention of disease progression depends on durable viral suppression.

Second, initial virologic response to ART is also better if therapy is initiated before immunodeficiency has progressed to a severe level (CD4 count <200 cells/µL). The rate of virologic failure is significantly higher if ART is started at a CD4 count <200 cells/µL, but there appear to be no differences in rates of failure if initiated at >350 cells/µL rather than at 200-350 cells/µL.(8,16) Additionally, persons with high viral loads at baseline (>100,000 copies/mL) have a slower rate of initially achieving a viral load <500 copies/mL.(8)

Third, certain serious infections are more common in HIV-infected individuals with CD4 counts between 200 and 350 cells/µL than in those with CD4 counts >350 cells/µL; these data might support initiating ART at a threshold somewhat higher than 200 cells/µL. These infections include bacterial pneumonia, particularly with encapsulated organisms such as Streptococcus pneumoniae and Haemophilus influenzae,(21) and pulmonary tuberculosis.(17,22) Pneumococcal disease occurs 50-100 times more often in HIV-infected persons than in immunocompetent individuals, and recurrent disease is common.(23,24) The incidence rate rises progressively with decreasing CD4 count, and is abnormally high even in persons with moderate degrees of immunodeficiency.(25) Similarly, tuberculosis occurs more frequently in HIV-infected individuals, with a demonstrated progressive increase in incidence as the CD4 count falls below 350 cells/µL.(26)

Established guidelines uniformly recommend initiating ART before the CD4 count is <200 cells/µL, but differ somewhat in the recommendations for asymptomatic individuals with CD4 counts of 200-350 cells/µL.

The 2005 DHHS guidelines recommend offering ART to individuals if CD4 count falls to <350 cells/µL or the HIV-1 viral load rises to >100,000 copies/mL and state that ART should always be given before the CD4 count falls to 200 cells/µL.(12)

The 2004 IAS-USA and the 2005 BHIVA guidelines recommend that ART be initiated before CD4 counts fall to the threshold of 200 cells/µL, and that treatment decisions be individualized for persons with CD4 >200 cells/µL. Factors important in determining when to initiate treatment in individuals with CD4 counts >200 cells/µL include: rate of CD4 cell count decline, HIV viral load, and patient understanding of, and commitment to, the proposed treatment for many years. Patients with CD4 counts >350 cells/µL who have plasma HIV RNA levels >100,000 copies/mL should be followed closely, probably at 2-month intervals, because of the increased likelihood of a rapid fall in CD4 count over the ensuing 3-year period.(10,13)

A documented rapid fall in CD4 count (declines of >100 cells/µL per annum) or a very high plasma HIV RNA level (>100,000 copies/mL) would support initiation of ART at the higher end of the 200-350 cells/µL CD4 count range.

Initiation of ART is a critical decision, which should always be made jointly between the informed patient and the physician. Generally, the decision should be made only after unhurried discussions (ideally over several patient visits), which ensure that the patient has as full an understanding as possible of the potential risks and benefits of alternative therapeutic approaches and is fully committed to the ART regimen that is chosen.

Younger age has been associated with a more rapid CD4 count increase after initiating ART.(27) Advanced age (>50 years) is an independent prognostic factor for more rapid disease progression in untreated individuals (19); with adequate ART therapy, however, there is no age-associated difference in survival.

Although some studies have shown that women have a lower initial viral load than men,(28,29) there seem to be no gender-associated differences in rates of disease progression (29,30) or prognosis.(19) There are no documented gender-associated differences in CD4 counts or in the relation of the CD4 count to the risk of OIs or disease progression. Treatment recommendations are therefore the same for men and women.

The goal of therapy is to durably inhibit viral replication to maintain effective immune responses to most microbial pathogens in HIV-infected individuals,(10,31) improve quality of life, and reduce HIV-related morbidity and mortality.(12) To achieve this goal, ART needs to be individualized and monitored for adverse effects and effectiveness. (For information on currently approved and commonly used drugs for HIV infection, see Table 1 and "Antiretroviral Drug Profiles".)

Plasma HIV RNA levels should be measured 2-4 weeks after initiating therapy. Effective ART should reduce the plasma HIV RNA by >90% (1 log10) within 2 weeks of treatment; failure to achieve such a response most commonly results from poor adherence, viral resistance, or inadequate exposure to the drug. An increase in the CD4 count generally follows the decrease in HIV RNA level. The rate of increase in CD4 count varies with the CD4 count at the time of initiation of therapy and is less helpful than HIV RNA levels in therapeutic decisions following initiation of ART. After initial successful inhibition of viral replication, occasional small increases in HIV RNA levels (50-200 copies/mL) may occur despite good adherence.(32) There is no evidence that such occasional modest increases in viral load are predictive of subsequent treatment failure.

Durability of several effective ART regimens has been demonstrated over periods of 3 or more years.(33,34) Major factors contributing to the durability of an initial regimen include antiviral potency of regimen, adherence, tolerability, convenience of the regimen, and baseline virologic or immunologic status.(10)

Self-reported adherence to ART is a strong predictor of virologic and immunologic outcome,(35) and many strategies to improve adherence have been developed.(12) Drugs with long half-lives and those that impose no dietary restrictions have clear advantages in producing durable suppression of viral replication and related increase in CD4 counts.(36) Even minor adverse effects may interfere with adherence over the long term. Twice-daily regimens are associated with better adherence than regimens taken more frequently. Once-daily regimens permit directly observed therapy, which appears to be very effective in dealing with the management of HIV infection in certain hard-to-reach patients.(37)

Lack of adherence usually causes suboptimal viral suppression. This promotes the development of drug resistance, both to one or more of the drugs being taken, and often to other drugs in the same class due to cross-resistance. There is no cross-resistance among the three classes of antiretroviral drugs used in initial therapy, but there can be extensive cross-resistance within each class, most notably in the available nonnucleoside reverse transcriptase inhibitors (NNRTIs), but cross-resistance is also significant among the nucleoside and nucleotide reverse transcriptase inhibitors (NRTIs) and the protease inhibitors (PIs). Therefore, it is of tremendous importance that the initial regimen is carefully chosen to result in effective long-term suppression of viral replication.

Two types of initial combination regimens are currently recommended (12):

1. An NNRTI with 2 NRTIs; according to mid-2005 U.S. Public Health Service (PHS) guidelines, the preferred regimen consists of efavirenz + (either lamivudine or emtricitabine) + (either zidovudine or tenofovir);

2. A PI boosted with low-dose ritonavir, and 2 NRTIs; according to mid-2005 USPHS guidelines, the preferred regimen consists of lopinavir/ritonavir + (either lamivudine or emtricitabine) + zidovudine.

Other combinations, such as a PI (with low-dose ritonavir) with an NNRTI and one or 2 NRTIs, may be used in special circumstances, but are not recommended for initial therapy.(10)

One of the preferred combinations for initial therapy is an NNRTI with 2 NRTIs. Such combinations are effective and avoid the pill burden and potential adverse effects of PI-based regimens. One disadvantage of the use of NNRTI regimens is that a single mutation in the reverse transcriptase gene can produce a virus resistant to all NNRTIs. On the other hand, the NNRTIs are generally well tolerated and have favorable pharmacokinetic properties. Two NNRTIs are currently available for use in initial therapy: efavirenz and nevirapine.

Efavirenz combined with 2 NRTIs is at least as effective as a boosted PI and 2 NRTIs in reducing viral load to undetectable levels in treatment-naive patients, including patients with a high baseline viral load (>100,000 copies/mL).(38) Efavirenz is generally well tolerated, and is given once a day. Adverse effects include vivid dreams, hyperexcitability, nightmares, and hallucinations, which usually disappear after 2-5 weeks of treatment. Mood disturbances and personality changes, however, may persist in rare cases (see "Adverse Events of Antiretroviral Drugs"). Efavirenz is contraindicated in pregnancy and needs to be used with caution in women of childbearing potential. Like nevirapine, it is rarely associated with major serum lipid abnormalities and alterations in body fat distribution. A randomized trial comparing efavirenz (EFV) and nelfinavir (NFV), each in combination with zidovudine (ZDV) plus lamivudine (3TC), or with didanosine (ddI) plus stavudine (d4T), showed that EFV/ZDV/3TC was superior to NFV/ZDV/3TC in virologic outcome.(39,40) The combination of EFV/ZDV/3TC is a highly recommended choice for initial therapy. It is effective and well tolerated, with a pill burden as low as 3 pills per day. Efavirenz is therefore the preferred NNRTI except in pregnant women, or women of childbearing potential.

Nevirapine is generally well tolerated, causes few drug interactions, and has no unfavorable effects on lipid profiles. Additionally, it can safely be administered in pregnancy or at delivery, to prevent perinatal transmission. Its activity is similar to indinavir and nelfinavir when combined with 2 NRTIs.(10) A large, randomized, open-label study found nevirapine (administered either once or twice daily) and efavirenz to be of comparable efficacy when each was combined with stavudine and lamivudine in an initial treatment regimen.(41) Nevirapine has frequently been associated with rashes, rarely with the Stevens-Johnson syndrome, and occasionally with serious hepatotoxicity. Women with CD4 cell counts >250 cells/µL have a higher incidence of symptomatic hepatitis than women with counts <=250 cells/µL (11% vs 0.9%). Men with CD4 cell counts >400 cells/µL similarly have an increased risk of hepatic events compared to men with counts <=400 cells/µL.(42)

Delavirdine is seldom used because of its very high pill burden and inconvenient 3-times-a-day dosing.

Currently, there are 9 PIs available. Increases in the use of PIs were clearly linked to the dramatic reductions in mortality and morbidity due to HIV infection in the late 1990s.(2) The combination of a PI and 2 NRTIs has been shown to be effective in durably reducing viral levels in antiretroviral-naive patients.

One increasingly prominent concern with PI-containing regimens is the occurrence of serum lipid abnormalities and/or alterations in body fat distribution.(43) Such abnormalities have been observed in 6-80% of PI-treated individuals, the wide range reflecting lack of a clear case definition.(12) The changes in serum lipids may occur within days after initiation of PI therapy. Changes in body fat distribution may occasionally occur within weeks, but often are not recognized until 12-24 months after initiation of treatment. The relative role of PIs and NRTIs in fat accumulation or atrophy is unclear. Treatment is difficult and seldom reverses established alterations in body fat distribution.

PI-containing regimens can be difficult to adhere to due to 3-times-a-day schedules, possible food restrictions (indinavir, nelfinavir),(36) and large pill burden (nelfinavir, amprenavir). These difficulties can be partially alleviated by concomitant use of low-dose ritonavir, which boosts plasma levels of several of the PIs by inhibiting the cytochrome P450 (CYP 450) enzyme that metabolizes these drugs. Ritonavir inhibits CYP 450 in both the gastrointestinal tract and the liver. When ritonavir is coadministered with saquinavir or lopinavir, the effects are likely through inhibition at both sites. For other PIs, the enhancement is primarily due to CYP 450 inhibition in the liver. Nelfinavir is the only currently available PI that is not effectively boosted by ritonavir.(44) Boosted PI therapy leads to more convenient twice- or once-daily regimens with fewer pills, and eliminates the need for dietary constraints.(12) Therefore, boosted PI regimens are preferred to nonboosted ones for initial therapy.

Lopinavir is commercially available only in a coformulated tablet with ritonavir. The lopinavir/ritonavir combination (Kaletra) demonstrates sustained antiretroviral activity with minimal development of resistance mutations for at least 108 weeks, and virologic failures are rare in the setting of good adherence.(45)

Atazanavir offers once-daily dosing and has fewer adverse effects on lipid metabolism than other PIs. When compared with nelfinavir, atazanavir has similar efficacy and is less likely to cause diarrhea.(46) The most frequent adverse effect of atazanavir is dose-dependent hyperbilirubinemia.(47)

Fosamprenavir is a prodrug of amprenavir, and is approved for twice- or once-daily dosing when used in combination with ritonavir. For treatment-experienced patients, only the twice-daily combination of fosamprenavir + ritonavir is recommended.

Available data support the choice of lopinavir/ritonavir + lamivudine (or emtricitabine), in combination with zidovudine, stavudine, or tenofovir as an effective and durable antiretroviral regimen. For treatment-naive individuals, lopinavir/ritonavir can be administered once daily. Although some physicians recommend such regimens for initial treatment of antiretroviral-naive patients, most prefer an initial NNRTI-based regimen, largely because of the frequency of serum lipid abnormalities that may occur with any ritonavir-boosted PI regimen.

Tipranavir is a new PI, approved by the U.S. Food and Drug Administration (FDA) in June 2005 for use in individuals who have HIV strains that are resistant to other PIs. To achieve adequate plasma concentrations, tipranavir requires boosting with ritonavir.

There are 8 NRTIs currently available in the U.S. The selection of drugs should be based on existing data on effectiveness and on the individual patient situation, taking into account the adverse effect profile, potential interactions, and additive toxicities in each patient. Didanosine + stavudine-containing regimens are not recommended as initial therapy, as they are associated with more toxicity than regimens containing zidovudine + lamivudine.(39,48) The use of tenofovir + didanosine in combination with an NNRTI results in high rates of early virologic failure, especially in individuals with HIV RNA levels >100,000 copies/mL and low CD4 levels, and should also be avoided in an initial regimen.

Lamivudine has been extensively studied in various combination regimens that are well tolerated and effective. Lamivudine is available in single-tablet combinations with zidovudine (Combivir), with abacavir (Epzicom), and with zidovudine and abacavir (Trizivir). Emtricitabine is comparable to lamivudine in efficacy and can be given once daily. A double-blind study comparing emtricitabine to stavudine, both in combination with didanosine and efavirenz found that emtricitabine had significantly higher rates of virologic suppression than did stavudine.(49) Emtricitabine is also available in combination with tenofovir as a single tablet (Truvada). Emtricitabine and lamivudine share nearly identical resistance profiles.

Tenofovir is a reverse transcriptase inhibitor, a nucleotide analogue of adenosine. It appears to be as efficacious as stavudine when given in combination with lamivudine and efavirenz for the treatment of antiretroviral-naive patients.(48) It is well tolerated; adverse effects are mostly gastrointestinal in nature but may include headache, and there have been reports of renal impairment.(49)

The potential advantages of triple NRTI regimens are that they are simple, generally well tolerated, and have a low incidence of drug-drug interactions (see Table 1). However, due to inferior potency compared to other combinations, triple NRTI regimens are no longer recommended for initial therapy. A randomized, double-blind, placebo-matched study was stopped in early 2003, after a planned interim analysis determined that the fixed-dose combination of abacavir + lamivudine + zidovudine was inferior to 2 efavirenz-containing regimens in terms of virologic efficacy and durability.(50) A triple NRTI regimen should only be considered as an alternative regimen for individuals unable to tolerate PIs or NNRTIs. The abacavir, zidovudine, and lamivudine combination (coformulated as Trizivir) produced a decrease in viral level at 48 weeks equivalent to that of indinavir, zidovudine, and lamivudine, but in patients with a high baseline viral load (>100,000 copies/mL), complete viral suppression was less likely.(51) A major disadvantage is the frequent (up to 5% of patients) hypersensitivity reaction to abacavir, which necessitates discontinuation of the drug.

In exceptional circumstances, 2 other types of regimens have been used effectively, but neither is recommended for initial therapy.

In patients who have experienced moderate to severe lactic acidosis as a complication of NRTI therapy, a regimen containing a ritonavir-boosted PI and an NNRTI has been used successfully,(52) but there are no available data on the long-term effectiveness of such regimens.

A regimen containing drugs from each of 3 classes (a PI, an NNRTI, and an NRTI), has been used for patients with high near-term mortality risk because of very low CD4 cell counts,(10) but there are no long-term data supporting the greater effectiveness of such a regimen. Regimens utilizing 3 classes of antiretroviral drugs are generally avoided because of the danger--especially in patients who have problems with adherence--of developing resistance to all 3 antiretroviral drug classes. The fusion inhibitor, enfuvirtide, represents a fourth antiretroviral drug class, which is not used in initial therapy.

As more medications and possible combinations become available, the focus of ART is shifting towards the use of simpler regimens, both because challenges to adherence can compromise the long-term success of treatment, and because even modest inconveniences can become significant in the context of lifelong therapy. The aim is to improve adherence and tolerability while maintaining efficacy and safety. In this regard, much focus has been on regimens that can be taken once a day (see "Options for Once-Daily Dosing of Antiretrovirals"). In addition to improving convenience and facilitating adherence, once-daily dosing also permits the administration of ART as directly observed therapy (DOT). DOT has been shown in pilot programs to be successful in improving HIV treatment, especially among hard-to reach-populations.(37)

The once-daily antiretrovirals can include single drugs with longer half-lives and ritonavir-enhanced PIs combined so that one enhances the therapeutic drug levels of the other. There are 8 medications currently approved by the FDA for once-daily dosing: didanosine, lamivudine, emtricitabine, tenofovir, efavirenz, atazanavir, ritonavir-boosted fosamprenavir, ritonavir-boosted lopinavir, and extended-release stavudine (see Table 2 in "Options for Once-Daily Dosing of Antiretrovirals," which includes drugs not yet FDA approved for once-daily dosing). Once-daily nevirapine is comparable in efficacy to twice-daily dosing (41); and, although it has not yet been approved by the FDA for once-daily dosing, nevirapine is widely used in such regimens. Other drugs are being investigated for once-daily administration. In addition, several dual PI combinations are being evaluated for once-daily dosing, using the boosting effect of ritonavir. These include indinavir/ritonavir and saquinavir/ritonavir.

Most individuals with acute HIV infection are symptomatic and could be targeted for early therapy. Information on long-term outcomes of treatment for acute HIV infection is limited. Possible benefits of early therapy include: less severe acute disease, lower initial viral "set point" with potential slower disease progression, preservation of immune function, and possible reduction in viral transmission.(53) Additionally, there are some suggestions that shortly after acute HIV infection, there is a strong, HIV-specific CD4 T-cell response, which is lost with chronic infection.(53,54) On the other hand, potential risks include drug toxicities and, if virus is not fully suppressed, drug resistance that limits further treatment options. Based on the limited available data, some clinicians now choose to treat acute HIV infection. If treatment is considered, it is recommended to start within 6 weeks following HIV infection and, if possible, to enter the patient into a controlled clinical trial.(12) The choice of antiretroviral regimens in this setting is the same as described above.

The acquisition of drug-resistant HIV is increasing in certain parts of the world, but not in others. In North America, a retrospective analysis of the prevalence of transmitted drug-resistant HIV in 377 individuals revealed that between 1995 and 2000 the frequency of high-level resistance had increased from 3.4% to 12.4%, and the prevalence of multidrug resistance increased from 1.1% to 6.2%.(55) Similarly, in parts of the developing world, high rates of drug resistance have been reported.(56) In Switzerland, on the other hand, transmission rates of resistant virus have declined.(57) Many factors may influence these regional differences, but availability of, and adherence to, effective ART seems to play a role, as resistance emerges in the setting of continued viral replication with incompletely suppressive drugs.(58) This is particularly concerning in the developing countries where only suboptimal therapeutic regimens may be available.

A retrospective survey of clinic data from patients with acute HIV infection found that resistant virus was transmitted less frequently than expected (59); nevertheless, due to the increased prevalence of resistant virus, the transmission rates are rising.

Patients infected with drug-resistant virus are less likely to respond favorably to the initial antiretroviral therapy than individuals with a sensitive virus.(55,60) Therefore, some authors favor testing for resistance to antiretroviral drugs before initiating therapy. Current guidelines favor sending a blood specimen for genotypic testing prior to initiating ART in patients identified during acute HIV infection so that the regimen chosen can be subsequently modified if resistance is detected.(9,12) In untreated individuals with established infection of longer duration (>6 months, chronic HIV infection), the resistant virus may not be detected due to the expected predominance of wild-type virus in the absence of drug selection. In this setting, genotypic testing is not recommended in the DHHS guidelines (12); the IAS-USA guidelines suggest considering genotypic testing in areas where the prevalence of primary drug resistance is known to be high.(9)

HIV InSite is a project of the UCSF Center for HIV Information. Copyright 2008, Regents of the University of California.