Lisa Frenkel MD
Children have frequently been treated with antiretroviral combinations that are insufficiently potent to suppress the virus, leading to resistance. This leading US paediatrician treats with more complex therapy upfront as well as in a salvage setting.
While many children have experienced recoveries that are genuinely 'miraculous' in association with HAART, we know that, unfortunately, the results have not been universally successful. In particular, treatment has 'failed' those children who developed resistance to combinations that were insufficiently potent to maximally suppress the virus. Treatments demand a high level of adherence (>95%) to avoid the selection of drug resistant virus and for other children, adverse reactions from the drugs have been intolerable.
Our aim is therefore to determine regimens that suppress viral replication, minimise side effects and that are practical for a child's lifestyle. Achieving these therapeutic goals with the currently available antiretrovirals has to overcome difficulties of adherence, the presence of drug-resistant virus from previous therapies, cross-resistance to drugs within each class, and the toxicity profile of individual treatments.
Paediatric care is based on limited data. Neither an optimum combination nor time to initiate paediatric HAART has been defined. The treatment of primary infection in infants appears beneficial , but this has not been compared to delayed HAART. Furthermore, the frequency and severity of adverse reactions associated with various treatment strategies (early vs. delayed, and continuous vs. intermittent), has not been evaluated either in the short or long term.
Early 3-drug studies
Nevertheless we have some data to work with. Early comparative trials of HAART with two nucleoside analogues and a protease inhibitor demonstrated superior clinical benefits in adults to dual nucleoside therapy . Subsequently, studies were designed to evaluate the suppression of viral replication, instead of clinical outcomes. This occurred for multiple reasons, including the association of low viral loads with slower HIV-1 disease progression in multiple studies [3-5], the ease in evaluating plasma HIV-1 RNA levels, and the recognition that if viral replication was not suppressed to very low levels selection of drug resistant virus would occur within weeks to months.
Although the superior virological benefit of three drug over dual regimens was clearly shown, early studies of three-drug HAART in nucleoside experienced children resulted in suppression to HIV-1 RNA <400 copies/ml in only 25-42% of children [6-9].
There have been few comparative trials between different HAART combinations in children. However, in reviewing both the comparative and observational studies, four-drug HAART regimens, including treatment experienced children, have demonstrated higher rates of viral suppression compared to three drugs regimens.
Four-drug regimens in the ACTG 377 and ACTG 382 studies, which also included children with previous nucleoside experience, provided greater virological success (61% <400 at week 24 in ACTG 377 and 76% <400 at week 48 in ACTG 382). However both three and four-drug regimens in these studies included all three classes of agents, leaving limited salvage options for the children whose treatment failed to achieve an undetectable viral load [10, 11].
Trials comparing three- and four-drug HAART in sequential cohorts of antiretroviral naive infants are ongoing, and preliminary reports also suggest that four-drug therapy offers greater antiviral benefit over three-drug regimens.
Arguments against the use of four drugs in an initial HAART regimen are based on several reasonable suppositions, namely, that adherence to therapy could decrease as the number of drugs increase, that adverse reactions and toxicity could be increased, and that the initial use of four-drug HAART leaves the child with less of a chance for successful 'salvage' therapy. While the data examining these issues is meagre, it does not support these suppositions, and they have not been reflected in adult care.
Toxicity was not increased among children taking four-drug compared to those taking three-drug HAART in ACTG 377 study and discontinuations were higher in the 3-drug arms, primarily due to insufficient viral suppression. Adherence to the regimens was not evaluated in this study, however, the superior antiviral effect of the four-drug HAART suggests that adherence to the former was not significantly compromised.
Salvage therapy has been less well studied but adult care has produced several important general lessons, including the importance of early switching and using greater numbers of accurately targeted agents. Nevertheless, the efficacy of salvage therapy has generally been inferior to the initial therapy of untreated individuals in all studies.
Preliminary analysis of one study of salvage therapy for children indicated that it was not effective in children previously treated with three-drug HAART that included protease inhibitors (ACTG 366)  . Only 10% of 100 PI-experienced children had their plasma HIV-1 RNA replication suppressed to <400 copies/ml 12 weeks after a four-drug regimen including nucleoside analogues not previously used, nevirapine and a protease inhibitor.
Limiting the time ongoing therapy is continued after viral rebound by earlier detection and frequent monitoring of plasma HIV-1 RNA has the potential to improve the outcome of a subsequent salvage regimen. In addition, mega-HAART treatment with five or more antiretrovirals, can suppress viral replication in antiretroviral-experienced children, if they do not have high-level resistance to protease inhibitors, and assuming that they are highly (>95%) adherent to the new regimen.
Salvage therapies have not yet been systematically studied in children but in this area we should closely follow and learn from the results of adult care. While there are immunological differences with children, the virological mechanism of resistance is likely to be very similar.
Dosing and PK
Correct dosing for each drug used in antiretroviral therapy also needs to be defined. The ACTG 382 study demonstrated that levels of efavirenz and nelfinavir were lower than expected in young children , pointing to a common theme with antiretrovirals in young children. Other studies have found lower levels than expected of ritonavir, nelfinavir, and nevirapine in young infants. Didanosine levels have been very variable in children . These data argue, not only for age-specific doses, but therapeutic monitoring of drug levels might be necessary for optimal management of HAART in children.
The focus on virological suppression in most studies has sometimes lead to the immunologic benefits with only partial suppression of replication being overlooked. The determinants of immunologic recovery have not been defined in children but are likely to be age dependent. Adults with incomplete but sustained suppression of viral replication by HAART to RNA levels below their viral set point have immunologic benefits . Importantly, the duration of these benefits have not been characterized.
In young children, I have observed a prolonged (>3 years) immunologic benefit in spite of high level viral resistance and return to baseline plasma HIV-1 RNA levels. These observations suggest two phenomena that require definition. First, the spectrum of immunologic benefits provided by lowering the level of viral replication, and second, the parameters affecting immune reconstitution in children and its natural history once achieved.
When to start - and with what
The available data has led me to advocate the initial treatment of infants and children with four-drug HAART, including two NRTIs, one NNRTI and one PI. Care in dosing the protease inhibitor should be taken to ensure that levels will be sufficiently high throughout the day and night, and lopinavir/r has many advantages in this respect. This is usually combined in a twice daily regimen with lamivudine, stavudine and nevirapine, as these agents are generally well tolerated. A very high level of adherence is necessary to obtain the maximum benefit from any HAART regimen .
I do not recommend therapy until there is at least a moderate degree of immunodeficiency or HIV-1 related symptoms, and until the family is in full agreement with therapy and is confident in their ability to administer antiretroviral treatment. Due to the difficulties in administering multiple drugs, some with unpleasant tastes, I introduce and encourage the use of gastrostomy tubes, unless the child can easily (and willingly) swallow pills.
Frequent monitoring of plasma HIV-1 RNA, (at least every month) especially early after treatment is initiated provides the opportunity to modify, intensify or stop therapy if plasma RNA levels do not progressively decrease. Using this strategy the child's HIV-1 generally does not develop multi-drug resistance, and the antiretroviral regimen can usually be manipulated so that viral replication is sufficiently suppressed by this allowing adequate immune reconstitution to occur and achieve good health.
Lisa M. Frenkel M.D. is Associate Professor of Paediatrics and Laboratory Medicine, University of Washington and Children's Hospital, Seattle, Washington. She has participated in and chaired multiple trials for the PACTG and writes extensively on the subject of paediatric HIV-infection. Her particular interest is paediatric HIV resistance.