Treatment Recommendations

Initiation of Antiretroviral Therapy

General Considerations
Antiretroviral therapy has provided substantial clinical benefit to HIV-infected children with immunologic or clinical symptoms of HIV infection. Studies have demonstrated substantial improvements in neurodevelopment, growth, and immunologic and/or virologic status with initiation of ZDV, didanosine (ddI), lamivudine (3TC), or stavudine monotherapy (48-53). More recent pediatric trials of symptomatic children who have not previously received antiretrovirals have demonstrated that combination therapy with either ZDV and 3TC or ZDV and ddI is clinically, immunologically, and virologically superior to monotherapy with ddI or ZDV as initial therapy (36,54,55). A trial involving children who have previously received antiretrovirals has demonstrated that combination therapy that includes a protease inhibitor is virologically and immunologically superior to dual nucleoside combination therapy (56).

Data from clinical trials that address the effectiveness of antiretroviral therapy in asymptomatic infants and children with normal immune function are not available. However, initiation of therapy early in the course of HIV infection, including during the period of primary infection in the neonate, is theoretically advantageous. Control of viral replication in perinatally infected infants is inadequate, as demonstrated by the high levels of HIV RNA that are observed during the first 1-2 years of life following perinatal infection. Initiation of aggressive antiretroviral therapy during this early period of viral replication could theoretically preserve immune function, diminish viral dissemination, lower the viral set point, and result in improved clinical outcome.

In a preliminary study of early treatment of children, six HIV-infected infants aged 2-4 months were placed on a regimen of ZDV, ddI, and nevirapine; baseline HIV RNA levels were 40,000-1,500,000 copies per mL. Five of six infants had an early virologic response with a drop in RNA PCR to <10,000 copies/mL by day 14, and two of the infants maintained undetectable levels of HIV RNA through 168 days of therapy (57). These two children had persistently negative HIV cultures, undetectable RNA levels, and became HIV antibody negative, although HIV DNA PCR remained positive. Clinical trials are ongoing to assess the virologic, immunologic, and clinical response of young infants to early, aggressive antiretroviral therapy with three or four antiretroviral agents.

The theoretical problems with early therapy include the potential for short- and long-term adverse effects -- particularly for drugs being administered to infants aged <6 months, for whom information on pharmacokinetics, drug dosing, and safety is limited. These concerns are particularly relevant because life-long administration of therapy is likely to be necessary for HIV-infected infants. If viral replication is not suppressed, ongoing viral mutation is likely to result in the development of antiretroviral resistance, curtailing the duration of benefit that early therapy might confer and potentially limiting future treatment options. Therefore, intensive education of care-givers and patients about the importance of adherence to the prescribed treatment regimen should be provided before therapy is initiated so that a) potential problems and solutions can be identified and b) frequent follow-up can be provided to assess virologic response to therapy, drug tolerance, and adherence.

When to Initiate Therapy
Antiretroviral therapy is recommended for HIV-infected children with clinical symptoms of HIV infection (i.e., those in clinical categories A, B, or C) (Table 2) or evidence of immune suppression (i.e., those in immune categories 2 or 3) (Table 1) -- regardless of the age of the child or viral load (Table 7). Clinical trial data from both adults and children have demonstrated that antiretroviral therapy in symptomatic patients slows clinical and immunologic disease progression and reduces mortality (54,55,58).

Ideally, antiretroviral therapy should be initiated in all HIV-infected infants aged >12 months as soon as a confirmed diagnosis is established -- regardless of clinical or immunologic status or viral load. HIV-infected infants aged >12 months are considered at high risk for disease progression, and the predictive value of immunologic and virologic parameters to identify infants who will have rapid progression is less than that for older children. Identification of infection during the first few weeks of life permits clinicians to initiate antiretroviral therapy or intensify ongoing antiretroviral therapy used for chemoprophylaxis of perinatal transmission during the initial phases of primary infection. However, clinical trial data documenting therapeutic benefit from this approach are not available, and information on drug dosing in neonates is limited. Because resistance to antiretroviral drugs (particularly protease inhibitors) can develop rapidly when drug concentrations fall below therapeutic levels (either as a result of inadequate dosage or incomplete adherence), issues associated with adherence should be fully assessed and discussed with the HIV-infected infant's caregivers before the decision to initiate therapy is made.

Two general approaches for initiating therapy in asymptomatic children aged >1 year were outlined by the Working Group. The first approach would be to initiate therapy in all HIV-infected children, regardless of age or symptom status. Such an approach would ensure a) treatment of infected children as early as possible in the course of disease and b) intervention before immunologic deterioration. Data from prospective cohort studies indicate that most HIV-infected infants will have clinical symptoms of infection by age 1 year (59,60). Most asymptomatic infected children aged >1 year also have CD4⁺ T-lymphocyte percentages of <25% (60), which is indicative of immunosuppression (Table 1) and warrants antiretroviral therapy.

An alternative approach would be to defer treatment in asymptomatic children aged >1 year with normal immune status in situations in which the risk for clinical disease progression is low (e.g., low viral load) and when other factors (e.g., concern for adherence, safety, and persistence of antiretroviral response) favor postponing treatment. In such cases, the health-care provider should regularly monitor virologic, immunologic, and clinical status. Factors to be considered in deciding when to initiate therapy include a) high or increasing HIV RNA levels, b) rapidly declining CD4⁺ T-lymphocyte number or percentage to values approaching those indicative of moderate immune suppression (i.e., immune category 2 [Table 1]), or c) development of clinical symptoms. The level of HIV RNA considered indicative of increased risk for disease progression is not well defined for young children. Regardless of age, any child with HIV RNA levels of >100,000 copies/mL is at high risk for mortality (Table 4), and antiretroviral therapy should be initiated -- regardless of clinical or immune status. HIV RNA levels in asymptomatic children aged >30 months that are the same as levels for which there are treatment recommendations for HIV-infected adults (e.g., >10,000-20,000 copies/mL) also may indicate the need to initiate treatment (Table 6). In addition, any child with HIV RNA levels that demonstrate a substantial increase (more than a 0.7 log₁₀ [fivefold] increase for children aged <2 years and more than a 0.5 log₁₀ [threefold] increase for those aged >2 years) on repeated testing should be offered therapy -- regardless of clinical or immunologic status or absolute level of viral load. These recommendations are based on limited data and may need revision as more information becomes available.

Issues associated with adherence to treatment are especially important in considering whether and when to initiate therapy. Antiretroviral therapy is most effective in patients who have never received therapy and who therefore are less likely to have antiretroviral-resistant viral strains. Lack of adherence to prescribed regimens and subtherapeutic levels of antiretroviral medications, particularly protease inhibitors, may enhance the development of drug resistance. Participation by the caregivers and child in the decision-making process is crucial, especially in situations for which definitive data concerning efficacy are not available.

Choice of Initial Antiretroviral Therapy
Combination therapy is recommended for all infants, children, and adolescents who are treated with antiretroviral agents (Table 8). When compared with monotherapy, combination therapy a) slows disease progression and improves survival, b) results in a greater and more sustained virologic response, and c) delays development of virus mutations resistant to the drugs being used. Monotherapy with the currently available antiretroviral drugs is no longer recommended to treat HIV infection. ZDV monotherapy is appropriate, however, when used in infants of indeterminate HIV status during the first 6 weeks of life to prevent perinatal HIV transmission. Infants who are identified as being HIV-infected while receiving ZDV chemoprophylaxis should be changed to a combination antiretroviral drug regimen.

Aggressive antiretroviral therapy for primary perinatal infection with three drugs is recommended because it provides the best opportunity to preserve immune function and delay disease progression. The goal of antiretroviral therapy is to maximally suppress viral replication, preferably to undetectable levels. Based on clinical trials involving infected adults, the preferred regimen is combination therapy with two NRTIs and one protease inhibitor. Although these combinations have had limited evaluation in clinical trials involving children, they can reduce HIV RNA to undetectable levels in some children (61,62). An interim analysis from a clinical trial of children (i.e., PACTG protocol 338) has demonstrated that therapy with drug combinations that include a protease inhibitor is more effective than therapy with two NRTI antiretroviral drugs in reducing viral load to undetectable levels and increasing CD4⁺ T-lymphocyte number (56). Recent data, primarily from adults, with the use of efavirenz (Sustiva) in place of the protease inhibitor may support the substitution of efavirenz for the protease inhibitor. The only available data regarding safety, dosing and virologic and immunologic efficacy of efavirenz in children are from an open-label study of efavirenz combined with nelfinavir and NRTIs in 57 pediatric patients (PACTG 382), some as young as age 3 years. In a preliminary intent-to-treat analysis, after 20 weeks of therapy, 65% of children had plasma HIV RNA levels <400, and 52% had HIV RNA levels <50 copies/mL (71,72). However, there are currently no pharmacokinetic data available on appropriate dosage of efavirenz in children under age 3 years, and although a liquid preparation is currently under study, only a capsular formulation is currently available. New antiretroviral drugs and combinations are being studied in infected adults and children. Other drug combinations that demonstrate sustainable viral load suppression and acceptable toxicity and dosing profiles most likely will become available, which will increase treatment options for children in the future. Since antiretroviral therapy will need to be administered for many years, considerations related to the choice of initial antiretroviral regimen should include potential limitations in subsequent treatment options should resistance develop.

Protease inhibitors with formulations appropriate for infants and children who cannot swallow pills include nelfinavir (Viracept®, manufactured by Agouron Pharmaceuticals, Inc., La Jolla, California), available in a powder formulation that can be mixed with water or food, and ritonavir (Norvir®, manufactured by Abbott Laboratories, North Chicago, Illinois), and amprenavir (Agenerase™, manufactured by Glaxo Wellcome, Inc., Research Triangle Park, NC), both available in a liquid formulation. Optimal dosing of these drugs in children aged <2 years is not known but is being evaluated in clinical trials (See Appendix). Indinavir (Crixivan®, manufactured by Merck and Company, Inc., West Point, Pennsylvania) and saquinavir (hard gel capsule, Invirase™, and soft gel capsule, Fortovase™, manufactured by Hoffman-LaRoche, Inc., Nutley, New Jersey) are not available in liquid formulations. Indinavir is recommended for consideration for children who can tolerate swallowing capsules. Optimal dosing of these drugs in infants and children is not known but is being evaluated in clinical trials (See Appendix). The hard-gel capsule formulation of saquinavir (Invirase™) has limited bioavailability and thus is not recommended for use with two NRTIs. Some studies have indicated substantial increases in saquinavir drug levels when coadministered with other protease inhibitors (e.g., ritonavir) or other drugs that inhibit the cytochrome P450 enzyme system. However, data regarding such combinations in children are not available. The soft-gel formulation of saquinavir (Fortovase™) with enhanced bioavailability has been approved by the Food and Drug Administration (FDA) for treatment of HIV infection in adults; however, data regarding appropriate dosing of this formulation in pediatric patients are not available.

Amprenavir in combination with 2 NRTIs is a regimen which may be offered in special circumstances in selected pediatric patients as initial antiretroviral therapy and also may have utility in antiretroviral-experienced patients. The combination of zidovudine, lamivudine and amprenavir resulted in an HIV RNA of <400 copies/ml in 53% of treatment naïve adults after 24 weeks of therapy (80). In an ongoing open label, randomized phase III trial of 504 NRTI and NNTRI experienced but PI naïve adults comparing amprenavir and indinavir in combination with 2 NRTIs (PROAB3006), 43% and 53% of patients respectively had HIV RNA <400 copies/ml after 24 weeks of therapy using an intent-to-treat analysis (83) . In a study of 41 treatment naïve adults, the combination of amprenavir and abacavir resulted in a fall in HIV RNA levels to <5 copies/ml in 58% of patients after 60 weeks of therapy (79). In HIV-infected pediatric patients, the combination of amprenavir with 2 NRTIs resulted in a decrease in HIV RNA to <400 copies/ml in 41% of antiretroviral- experienced patients treated for 8 weeks (82).

The FDA approved formulation of amprenavir contains 46 IU of vitamin E/ml of oral solution and 109 IU vitamin E per 150 mg capsule. The recommended dose of amprenavir results in a dose of 138 IU/kg/day of vitamin E using the oral solution with a maximum dose of 8,587 IU vitamin E per day. Patients receiving the recommended adult dose of amprenavir in capsule form receive 1,744 IU/day of vitamin E. In comparison, the daily recommended dose for vitamin E in children is 10 IU per day and for adults is 30 IU per day. The liquid formulation also contains propylene glycol in a concentration that exceeds WHO standards for use in infants. Young infants have immature levels of alcohol dehydrogenase enzymes which are involved in the metabolism of propylene glycol. The serum half-life of propylene glycol in neonates is prolonged at 16.9 hours compared with a half-life of 5 hours in adults. There is concern that the propylene glycol contained in the liquid formulation may not be metabolized adequately and could cause toxicity. High levels of propylene glycol have been associated with hyperosmolality, lactic acidosis, seizures and respiratory depression (87).

Because of the lack of long term data on the use of amprenavir and the lack of data on its use in treatment naïve children, it should be recommended as initial therapy only in special circumstances. This agent should not be used in children <3 years of age because of the lack of data in children <3 years of age, the paucity of data in children in general, the uncertain impact of extremely high doses of vitamin E, and the propylene glycol content of the oral liquid preparation.

Amprenavir may be included as a component of a treatment regimen for children who have failed prior protease inhibitor therapy. Therapy with amprenavir induces mutations in HIV-1 protease gene at codons 46, 47, 50, 54 and 84. At least 2-3 mutations are required at codons 46, 47 and 50 to produce >10 fold decrease in sensitivity. None of the other PIs induced a mutation at codon 50 in vitro.

Alternative regimens, although not ideal, may be considered for initial therapy in circumstances in which the caregiver has concerns regarding the feasibility of adherence to a complex drug regimen or when the patient and caregivers prefer an alternative regimen. Alternative regimens have been clinically beneficial in adult and pediatric patients, but these regimens may not suppress viral load to below detectable levels as consistently as does combination therapy with two NRTIs and a protease inhibitor. An example of such alternative regimens include combination regimens of two NRTIs with nevirapine substituted for the protease inhibitor or two NRTIs alone. However, drug regimens that do not result in sustained viral suppression may result in the development of viral resistance to the drugs being used.

The combination of abacavir, ZDV, and 3TC resulted in a viral load of <400 copies/mL in 74% of treatment-naive adults at 48 weeks of therapy, results similar to those of a protease inhibitor-containing regimen. (73,74) However, in a study of infected infants who had received only prior ZDV preventive therapy, this combination produced suppression of viral replication to <400 copies/mL in only 4 of 11 subjects by 12 weeks of treatment. (Catherine Wilfert, M.D., personal communication) This triple NRTI regimen spares the initial use of protease inhibitors and non-nucleoside reverse transcriptase inhibitors and can be administered twice a day in children, which may facilitate adherence. (75, 76, 77) However, because the uncertain long-term durability of viral load suppression with a regimen comprised of three drugs of a single class (NRTIs), because results in the infected infant are disappointing, and because abacavir is associated with a potentially life-threatening hypersensitivity syndrome, a protease inhibitor-containing or efavirenz-containing regimen is preferred. Insufficient clinical trial data are available to guide the optimal use of abacavir in combination with protease inhibitors and/or non-nucleoside reverse transcriptase inhibitors.

The initial antiretroviral regimen chosen for infected infants theoretically could be influenced by the antiretroviral regimen their mother may have received during pregnancy. However, data from PACTG protocol 076 indicate that ZDV resistance did not account for most infants who became infected despite maternal ZDV treatment (63), and data from PACTG protocol 185 indicate that duration of prior ZDV therapy in women with advanced HIV disease, many of whom received prolonged ZDV before pregnancy, was not associated with diminished ZDV efficacy for reduction of transmission (64). Data do not suggest that the antiretroviral regimen for infected infants should be chosen on the basis of maternal antiretroviral use. However, continuing to monitor the frequency of perinatal transmission of antiretroviral-resistant HIV isolates is crucial, because maternal therapy with multiple antiretroviral agents is becoming more common and the prevalence of resistant viral strains in the HIV-infected population may increase over time.

Issues Regarding Antiretroviral Dosing in Neonates
Data regarding the appropriate dosing of antiretroviral drugs in neonates are limited; ZDV is the best studied antiretroviral drug in this age group. The recommended ZDV dosage for infants was derived from pharmacokinetics studies performed in full-term infants (65). Because ZDV is primarily cleared through hepatic metabolism (i.e., glucuronidation), which is immature in neonates, the half-life and clearance of ZDV are prolonged in neonates compared with older infants, thus requiring adjustments in dosing (See Appendix).

Premature infants have even greater immaturity in hepatic metabolic function than do full-term infants, and further prolongation in clearance has been documented in very premature infants (e.g., those born before 34 weeks' gestation) (66). Appropriate ZDV dosing for premature infants has not been defined but is being evaluated in a phase I clinical trial of premature infants born before 34 weeks' gestation (i.e., PACTG protocol 331) (See Appendix).

The safety and pharmacokinetics of 3TC administered alone or in combination with ZDV in pregnant women and administered for 1 week to their newborns have been evaluated (67,68). Clearance was prolonged in these infants. On the basis of data from this study, the dose recommended for use in newborns is half the dose recommended in older children (See Appendix). No data are available regarding 3TC pharmacokinetics among infants aged 2-6 weeks, and the exact age at which 3TC clearance begins to approximate that in older children is not known.

Nevirapine administration to HIV-infected pregnant women during labor and as a single dose to their newborns at age 2-3 days has been studied in a phase I trial (69). The half-life of nevirapine was prolonged in neonates compared with that in older children, indicating that some modification of nevirapine dosage is required for administration to neonates (See Appendix).

Although phase I studies of several protease inhibitors (i.e., indinavir, ritonavir, nelfinavir, or saquinavir in combination with ZDV and 3TC) in pregnant infected women and their infants are ongoing in the United States, no data are available regarding drug dosage, safety, and tolerance of any of the protease inhibitors in neonates.

previous page | next page

return to table of contents