Antiretroviral Therapy in Children

The Pediatric Standard 0f Care -- Especially With Regard to the Use of Protease Inhibitors -- Is Largely Based on Guesswork and Hopeful Comparisons

Pediatricians who treat children with HIV infection have long had to struggle with a nebulous standard of care, frequently based on the recommendations of consensus panels(1) and anecdotal evidence, rather than on scientifically derived data. The development of new antiretrovirals for use in children has consistently lagged behind research in adults, often by a matter of years. Although some pharmaceutical companies are reorganizing their priorities to address the needs of pediatric patients, the same pattern has occurred with regard to development of protease inhibitors.

The current guidelines for using protease inhibitors in HIV-infected children are based largely on preliminary pediatric data and extrapolation from studies in adult patients. While the standard of care for adults is far from settled, treatment paradigms in adults are far more advanced and are more strongly supported by well-executed studies.

"Extrapolation of adult treatment recommendations to pediatric care has proved to be a complicated matter. The two laboratory markers employed most frequently in therapeutic decision-making, CD4 count and viral load, are not comparable in adults and children. These differences make it hard for pediatricians to decide when to begin antiretroviral therapy and when to switch to new drugs."

The lag in pediatrics is a consequence of several factors. Part of the problem has centered on the reluctance of federal licensing agencies to test drugs in children before there is an established safety profile in adults. Indeed, the relative paucity of pediatric data is another illustration of the many ways in which the AIDS epidemic has uncovered difficulties in the development and approval process for drugs in the United States.

In addition, because infants and children frequently handle drugs in a different manner than adults, pediatric patients are usually excluded from adult clinical trials. Consequently, information regarding the safety and efficacy of drugs in children has fallen almost exclusively on the relatively small community of pediatric clinical AIDS researchers.

Finally, an ongoing challenge in generating scientific data regarding the effectiveness of AIDS drugs in pediatric patients has been the relatively low numbers of seropositive children. Efficacy trials usually require hundreds of patients to yield statistically valid results, and entry is frequently resistricted to treatment-naïve subjects. In the United States, the pediatric HIV population has always made up only a small fraction of the total caseload. Recently, the number of new pediatric cases has been moderated by the leveling-off of HIV infection rates in child-bearing women.(2) Moreover, intragestational antiretroviral therapy with ZDV, as established by ACTG 076, has reduced the transmission rate from mothers to their infants by two thirds.(3) In a very real sense, then, the ability to conduct large-scale efficacy trials in pediatric patients has been compromised by the success of past efforts.

However, extrapolation of adult treatment recommendations to pediatric care has proved to be a complicated matter. The two laboratory markers employed most frequently in therapeutic decision-making, namely CD4 count and viral load, are not comparable in adults and children. For example, an immunocompetent adult typically has an absolute CD4 count of 800 to 1000 cells/mm3; in young children, counts around 2000 cells/mm3 are not unusual. While age-specific normal values have been established for children, the application of these values to pediatric AIDS care has never been validated. Moreover, viral loads are, on average, clearly higher in children than in adults with comparable disease,(4) and the ability to achieve undetectable viral levels with combination antiretroviral therapy has proven much more difficult in children than in adults. These differences make it hard for pediatricians to decide when to begin antiretroviral therapy and when to switch to new drugs, based on the viral load values recommended in adults.

What we know about the use of antiretrovirals in children

Within the last year, small pilot studies of lamivudine (3TC) and nevirapine have made their way into the literature. Lewis et al. demonstrated that 3TC monotherapy, at loses of 4 mg/kg/day or higher, decreases viral load in antiretroviral-naïve as well as previously treated HIV-infected children.(5) However, over the 24 weeks of the study no improvement was noted in growth and neurocognitive functions, although children with encephalopathy were excluded from the study. As in adult studies that investigated the efficacy of 3TC as monotherapy, resistance to lamivudine developed in a majority of the study subject prior to the end of the trial.

Nevirapine, an inhibitor of reverse transcriptase that works via a different mechanism than the nucleoside analogs, was found to be safe and well absorbed orally in 21 children.(6) Luzuriaga et al. reported that the only adverse reaction was rash. Nevirapine caused rapid and significant reductions in p24 antigen levels in 50% of treated subjects over eight weeks when given in a high-dose regimen. Not surprisingly, monotherapy with nevirapine resulted in viral isolates with high-level resistance to the drug, a pattern seen in treated adults; better results may be obtained when this drug is used in combination with other nucleosides or protease inhibitors.

As previously noted, combinations of antiretrovirals are rapidly becoming the standard of care in pediatric patients. While there are some studies of antiretroviral cocktails in the literature, these tend to be non-comparative safety and pharmacokinetic studies. A relatively early report showed ZDV and didanosine (ddI ) to be a potent combination over 24 weeks.(7) Using a range of doses, Husson et al. found that these two drugs produced a median increase in CD4 counts of more than 200 cells/mm3 in previously-treated children, and a median increase of 340 cells/mm3 in treatment-naïve patients. Viral loads were significantly lower at 24 weeks as well. Together, the drugs were well-tolerated, although two patients did develop pancreatitis, a known side effect of ddI therapy, and a few others developed hematological and muscle toxicities typical of ZDV therapy.

Interestingly, another study found that ddI alone was about as effective as ddI plus ZDV. AIDS Clinical Trials Group protocol 152 provided evidence that both of the arms involving ddI produced better outcomes in pediatric patients than ZDV alone. Even though ddI appeared to be equivalent to ddI plus ZDV, the combination is preferred because it offers the potential advantage of decreased development of resistant viral strains.

"A comparison study of ZDV/ddI and ZDV/3TC gave the edge to the latter combination, based on increased CD4 counts and decreased viral loads. A rare triple-combination, open-label trial recently studied ZDV, ddI, and nevirapine. The combination was associated with efficacy over six months and it was well-tolerated."

Recently, Kline et al. treated eight children with advanced HIV disease (median CD4 count: 42 cells/mm3) with a combination of stavudine (d4T) and ddI over 24 weeks.(8) All participants had received prior d4T monotherapy. The combination was well-tolerated and no significant pharmacokinetic interactions were noted. Median HIV RNA levels decreased by .88 log at 12 weeks; by 24 weeks, a 0.03-log reduction was noted. Halfway through the study, significant increases in CD4 counts (> 20%) were observed in those patients with baseline values above 50 cells/mm3.

A recent comparison study of ZDV/ddI and ZDV/3TC in small numbers of previously treated children gave the edge to the latter combination, based on a more sustained increase in CD4 counts and decrease in viral loads.(9) A rare triple-combination, open-label trial recently studied ZDV, ddI and nevirapine.(10) The combination was associated with efficacy over six months (HIV RNA load decreased by > 0.5 log) and it was well-tolerated. For more information about both of these trials, consult the NEWSLINE section of this issue (see "Combinations of reverse transcriptase inhibitors in children" and "Triple-combination therapy in infants").

Outside of ACTG 152 there are no data to date comparing the efficacy of antiretroviral combinations in children. Many pediatricians are now using ZDV/ddI and ZDV/3TC. How to fit protease inhibitors into this picture, even with two of them fully licensed and available, is a big question.

Enter the protease inhibitors

If one is, of necessity, already prescribing combination antiretroviral therapy for pediatric patient without the benefit of hard data on the efficacy of these drugs in younger patients, perhaps it is only a small leap to experiment with a new class of compounds such as the protease inhibitors. Adding a protease inhibitor to create a three-drug cocktail has emerged, on the basis of anecdotal evidence and data in adults, as a potent treatment option for pediatric patients.

Fortunately, there are some safety and dosing data on two of the four available protease inhibitors. Evaluation of their antiretroviral activity in pediatric patients is far from over, however. More complete clinical investigation will be needed to provide information on how the drugs affect surrogate markers such as viral load and CD4 counts, as well as how these drugs affect clinical endpoints like survival and disease progression. Large-scale trials are just getting under way.

NELFINAVIR: Agouron Pharmaceuticals' nelfinavir (Viracept®) is the first protease inhibitor approved simultaneously for use in both children and adults. The study upon which the F.D.A. based its pediatric approval was conducted at four universities nationwide. In the trial, 38 children (ages 2-13) were evaluated.

Agouron's package insert notes that the pediatric trial produced a "similar adverse event profile" to that seen in trials conducted in adult patients. In adults, the most significant side effect by far is diarrhea, which occurred in up to 20% of patients taking 750 mg of nelfinavir (the recommended dosage) in combination with ZDV and 3TC. Diarrhea occurred in 32% of patients taking 750 mg of nelfinavir in combination with d4T. Nausea was seen in 7% of patients on the nelfinavir/ZDV/3TC combination, and other side effects (abdominal pain, weakness, rash and nausea) occurred in no more than 4% of study participants.

In children, the dose of nelfinavir is based on body weight: 20Ð30 mg/kg per dose, three times daily. It is recommended that the drug be taken with a meal or light snack. For children unable to take tablets, a powder formulation of nelfinavir can be mixed with water, milk, formula, pudding, liquid soy products, or dietary supplements. Acidic juices or foods (including orange juice, apple juice and applesauce) are not recommended, because the combination may taste bitter. In late 1997, a European study using twice-a-day dosing of nelfinavir should be completed. This simplified dosing regimen, if effective, would make the drug even more attractive for pediatric use.

One potentially important aspect of nelfinavir is its apparent lack of cross-resistance with other protease inhibitors, according to data generated by the manufacturer. This means that patients who develop resistance to nelfinavir may still be able to derive some benefit from other protease inhibitors. This finding is especially good news for children, who are much more likely to begin therapy early in infection and who have a higher likelihood of needing to switch antiretrovirals down the line. Should resistance to nelfinavir emerge, young patients treated with this drug may have a number of therapeutic options.

RITONAVIR: The other protease inhibitor available by prescription for children with HIV infection is ritonavir (Norvir®), which was approved for children a year and two weeks after its approval in adults. The study upon which the F.D.A. based its approval of ritonavir for use in children was a collaboration between the National Cancer Institute (HIV and AIDS Malignancy Branch) and Abbott Laboratories, the manufacturer of the drug. Researchers evaluated 44 children (ages 2-16). For the first 12 weeks of the trial, the drug was given as monotherapy; thereafter it was given in combination with ZDV or ZDV plus ddI.

Researchers found a similar side-effect profile in children and adults. The most common adverse events include nausea, diarrhea, weakness, and changes in the way foods taste. A significant finding is that ritonavir tends to increase absorption of drugs that are metabolized by the liver. Consequently, the concomitant use of ritonavir and certain other drugs can result in life-threatening side effects. Clinicians must therefore consider the possibility of drug-drug interactions when prescribing this protease inhibitor.

The recommended dosage of ritonavir in children is 400 mg/m2 twice daily. Individual doses should not exceed 600 mg/m2 twice daily. Abbott Labs recommends starting therapy at 250 mg/m2 twice daily, and then ramping up to the 400 mg/m2 dosage. For children too small to take capsules and those unable to do so, ritonavir also comes in an orange-colored liquid solution. This formulation is the same as the adult oral solution. Its bitter taste may not be tolerated by children.

Weighing the limited data currently available, nelfinavir may be the better protease inhibitor for children. While both agents have acceptable toxicity profiles, nelfinavir offers a more favorable cross-resistance profile, and it comes in an easy-to-mix powder form that may be more palatable than the ritonavir liquid formulation.


Although some tentative steps have been taken in the right direction, much needs to change in order to achieve parity in the standard of care for HIV-infected adults and children. One important advance would be routine parallel clinical trials in both populations. Research into protease inhibitors has provided one of the first opportunities for such parallel trials. To date, only Agouron Pharmaceuticals has used this model.

Clearly there is an ongoing need to enroll as many children as possible in pediatric clinical trials. Pediatric ACTG units are present in most areas that have a high incidence of pediatric HIV disease, and referrals to these centers for recruitment into the appropriate trials is critical to shedding light on these therapeutic issues. However, even under the best circumstances the relatively small numbers of infected children render it unlikely that every question that needs to be addressed can be answered with a comparative trial.

Ultimately, large trials will need to be reserved for the most pressing questions, while many other strategies in pediatric care will have to be extrapolated from those established scientifically in adults. Such extrapolation carries considerable risk, since perinatally acquired HIV infection differs substantially from acquired disease in adults. To date, however, all similar pediatric and adult trials have yielded similar results. The need for expedient drug evaluation in the pediatric population may mandate that some agents be licensed for the use in children without the benefit of traditional comparative trials, once pediatric safety and pharmacokinetic data are available.


1. Working Group on Antiviral Therapy: National Pediatric HIV Resource Center. Antiviral therapy and medical management of the human immunodeficiency virus-infected child. Pediatr Infec Dis J 1993; 12: 513-20.

2. Davis SF, Byers RH, Lindegren ML, Caldwell B, Karon JM, Gwinn M. Prevalence and incidence of vertically acquired HIV infection in the United States. JAMA 1995; 274: 952-5.

3. Connor EM, Sperling RS, Gelber R, et al. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994; 331: 1173-80.

4. Shearer WT, Quinn TC, LaRussa P, et al. Viral load and disease progression in infants infected with human immunodeficiency virus type 1. N Engl J Med 1997; 336: 1337-42.

5. Lewis LL, Venzon D, Church J, Farley M, Wheeler S, Keller A, et al. Lamivudine in children with human immunodeficiency virus infection: a phase I/II study. J Infect Dis 1996; 174: 16-25.

6. Luzuriaga K, Bryson Y, McSherry G, Robinson J, Stechenberg B, Scott G, et al. Pharmacokinetics, safety and activity of nevirapine in human immunodeficiency virus type 1-infected children. J Infect Dis 1996; 174: 713-21.

7. Husson RN, Mueller BU, Farley M, Woods L, Kovacs A, Goldsmith C, et al. Zidovudine and didanosine combination therapy in children with human immunodeficiency virus infection. Pediatrics 1994; 93 (2): 316-22.

8. Kline MW, Fletcher CV, Harris AT, Evans KD, Rutkiewicz VL, Shearer WT, Dunkle LM. Combination therapy with stavudine and didanosine in children with advanced human immunodeficiency virus infection: pharmacokinetic properties of safety and immunologic and virologic effects. Pediatrics 1996; 97 (6): 886-90.

9. Solder B, Wintergerst, Notheis G, Eberle J, Burtler L, Belohradsky BH. Effect of antiretroviral combination therapy (zidovudine/didanosine or zidovudine/lamivudine) on quantitative plasma human immunodeficiency virus-ribonucleic acid in children and adolescents infected with human immunodeficiency virus. J Pediatrics 1997; 130 (2): 293-9.

10. Luzuriaga K, Bryson Y, Krogstad P, Robinson J, Stechenberg B, Lamson M, et al. Combination treatment with zidovudine, didanosine, and nevirapine in infants with human immunodeficiency virus type 1 infection. New Engl J Med 1997; 336: 1343-9.

Philip Toltzis, M.D., is Assistant Professor of Pediatrics at Case Western Reserve University School of Medicine, Cleveland, OH.

Tiffani Coleman, R.N., N.D., is a Nurse Case Manager at the John T. Carey Special Immunology Unit, University Hospitals of Cleveland, Cleveland, OH.