Many of the advances in HIV therapy during the last six years have resulted from improved ways of using existing drugs in combination therapy regimens, the availability of new, more potent drugs such as protease inhibitors, and the introduction of viral load tests to help with decisions to start or change treatment.

The majority of clinical studies have focused on adults, and while there has been progress in research into new treatments and viral load testing for children there are fewer potential participants for paediatric trials and there have been fewer clear results. A major break through in treatments for children occurred in March 1997, when the United States Food and Drug Administration licensed two protease inhibitors, ritonavir and nelfinavir, for use in children.

Subsequently, other antiretrovirals have been approved for use in children. Nevertheless, there remains a need for paediatric research because the results of studies among adults cannot always be applied directly. For example:

• Some drugs are handled differently in children's bodies, affecting the doses that are needed.

• Young children have immature immune systems which may be less effective at fighting HIV infection.

• Normal CD4 cell counts are much higher in young children than in adults, and so CD4 cell counts in HIV-infected children are dependent on age and have to be interpreted differently from adults.

• There are differences in the 'natural history' of HIV disease between children and adults.

• There has been less research among children into the tolerability and toxicities of the range of possible combination therapies.

• It can be harder to obtain anti-HIV drugs in formulations that children can take.

HIV disease in infants and children

Viral load in infected infants may rise above one million copies/ml within weeks of infection and remain at this level during the first years of life. It is thought that such high levels of HIV are a consequence of the immaturity of the infant immune system. In adults the immune system normally reduces viral load by three to four log (e.g from two million copies and above to below 20,000 copies) within six months of infection.

The disease progression rate amongst untreated HIV-infected infants is much higher than amongst adults.

Among untreated infants in African settings, the majority of babies are likely to develop HIV-related symptoms within the first year of life (Kalish). A study of over 3000 babies in Uganda found that 22% of infants born to HIV-positive mothers died before the age of one year. By the age of five years, the death rate was 313 per 1000 live births in children of mothers who were HIV-positive at the time of birth, compared to 53 and 114 per 1000 live births in children of HIV-negative mothers (Nakiyangi 2003). A study from Rwanda found that 52% of the infected infants died compared with 4% of uninfected infants after over two years of follow-up (Spira 1999). Even more daunting rates of disease progression have been reported from the Queen Elizabeth Central Hospital in Malawi. After three years, 89% of infected infants were dead and only 1% were free of symptoms of HIV (Taha 2000), despite high immunisation rates. The earlier that children develop AIDS-defining illnesses, the poorer their chances of survival (Kalish).

In the western context, availability of antiretroviral therapy has reduced illness and death amongst children born with HIV (Lindegren). Nevertheless, factors such as advanced disease in the mother (Ioannidis 2001), a poor immune response or a lack of material antibodies (Dickover) are associated with faster disease progression. A French study of 85 infants born HIV infected since 1997 and 403 HIV infected babies born pre-1997 shows the impact that highly active antiretroviral therapy (HAART) has had on health and survival of these children. Before HAART, 12% of HIV positive infants had died by the age of 18 month, 12% developed encephalopathy and 6% an opportunistic illness. After HAART, all HIV infected infants survived to the age of 18 months, none developed encephalopathy and only one had an opportunistic infection (Faye 2003).

Treatment of HIV-infected infants in Cote d'Ivoire has also improved survival and health. Babies with an AIDS-defining illness or a CD4 percentage below 15% qualified for treatment. After approximately 18 months follow-up, rates of diarrhoea and pneumonia had fallen significantly. Survival was influence by baseline CD4 percentage - 73% of infants with a percentage below 5% survived compared with 98% with a CD4 percentage between 5-15% (Msellati 2003).

In children infected through blood transfusion or Factor VIII, early viral load levels tend to resemble those seen in adults rather than the very high levels seen in vertically infected children. As a consequence, survival and illness in such children is similar to the rates seen in adults (Engels). It has been suggested that virus levels and disease progression rates are lower because children already have a more mature immune system than infants infected perinatally.

HIV disease in children may manifest as AIDS-defining illnesses such as PCP (a form of pneumonia and the most common opportunistic infection in infants), candidiasis, cytomegalovirus or tuberculosis, or it may take a more non-specific form. Failure to thrive, unexplained persistent fever, swollen glands and spleen, and diarrhoea are frequent features of this syndrome. Growth impairment in HIV-infected children is associated with reduced food intake and advanced diseased. Cancers among HIV-infected children are associated with a high Epstein-Barr virus viral load in children with CD4 counts above 200 (Pollock 2003).

HIV-infected children undergo puberty on average at a later stage than uninfected children. In part, this seems related to the HIV-infected child's immune function: those with severe immune damage (defined as a CD4 percentage below 15%) undergo puberty later than others with normal or mildly impaired immune function (Buchacz 2003).

Children exposed to antiretrovirals may exhibit blood disorders such as anaemia, and liver toxicity (Taha 2002).

Prophylaxis and CD4 cell counts

Due to age-related variations in CD4 cell counts it has been difficult to estimate the need for PCP prophylaxis by monitoring the CD4 cell count. As a consequence United States guidelines now propose PCP prophylaxis for all children born of HIV-infected mothers, with the recommendation that it should start not later than four to six weeks after birth.

After the age of one it is recommended that all HIV-infected children with a CD4 cell count below 500 cells/mm³ should receive PCP prophylaxis up until the age of six, and all children above the age of five should receive PCP prophylaxis if their CD4 cell counts fall below 200 cells/mm³.

Viral load as a marker in children

Although, as described above, viral load is much higher in young children than in adults, a reduction in viral load produced by antiretroviral therapy is a reliable marker of reduced risk of disease progression.

A review of five major US paediatric trials found that changes in viral load over 24 weeks of treatment can be used just as well in children as in adults to measure the reduction in risk of death. Viral load reductions on treatment also predicted the extent of growth failure and the extent of cognitive impairment in children who started treatment above the age of one, with those children who experienced the poorest viral load reductions showing the greatest failure to thrive (Lindsey). The ACTG 152 trial of nucleoside analogues in children also found that suppressing viral load in children resulted in a similar reduction in the risk of disease progression as seen in adults (Palumbo).

Immune reconstitution in children

The pattern of immune reconstitution after commencing HAART differs in adults and children. While adults tend to gain an absolute number of CD4 cells regardless of baseline CD4 cell count, children with lower baseline CD4 cell counts tend to have the largest increase in CD4 cell numbers once they start treatment. Baseline viral load does not appear to affect immune reconstitution or response to HAART in children (Walker 2004).

There is also little difference in immune reconstitution between infants and adolescents, suggesting that the production of naﶥ CD4 cells in the thymus is not impaired as children mature. Production of new immune cells by the thymus appears to be the main source of CD4 cell re-population in children taking antiretroviral therapy (De Rossi 2002).

Treatment interruptions in children have not yet been reported, and there is disagreement among doctors over when this approach should be attempted.

Effectiveness of treatment

Children treated with triple therapy have a lower likelihood than adults of maintaining viral load below the limits of detection. This is partly because viral load in children may be higher than that of adults when they start therapy, making undetectable viral load more difficult to achieve. However, adherence to treatment amongst children is also acknowledged to be a big problem, and poor adherence does undermine anti-viral effect.

Few studies have looked at the effectiveness of treatment in very young children and infants in whom viral load may be highest. Research has focussed on children above the age of five, which represents a selection bias. Children who have survived with HIV to this age are likely to have lower viral load and may have other virus or host factors which make them less predisposed to disease progression.

There are a number of studies that document relatively poor results with triple combination therapy among children. A review of nearly 1,500 children in the United States found that 289 were on protease inhibitors, with nelfinavir and ritonavir the most common choices. By week 28, only 23% of the children had viral loads below 500, and only 41% were below 10,000. At 32 weeks, the average drop in viral load was -0.6 log. Two-thirds of these children stopped their drug treatment, for reasons such as side-effects, adherence problems and lack of benefit. Baseline viral load was predictive of virological response (Havens). In addition, an American study of 25 children on triple therapy found that the median viral load was not significantly lower after 12 months of treatment. Only one quarter of the group had a viral load below 400 copies at 12 months. However, the mean CD4 count increased by 657 cells and the CD4 percentage rose from 2% to 16% (Essajee 1999).

However, several studies have shown that triple combination therapy can be effective for many children, particularly those with reasonably low viral loads:

• An audit of 91 HIV-positive children attending the St George Hospital in South London found that antiretroviral regimens were changed due to virological failure in 60% of cases and due to toxicity in 10% of cases. Viral load below 400 copies/ml was achieved by 46% of children on first-line therapy and 37% of children on second-line therapy (Doerholt 2002).

• Dutch researchers reported that 28 of 35 children on AZT/3TC and indinavir achieved viral loads below 500 within three months of commencing treatment in one study (Van Rossum). Like adults, children are more likely to respond to antiretroviral therapy if they have not previously taken treatment (Steiner).

• The PACTG 382 study found that NRTI-experienced children treated with a combination of nelfinavir and efavirenz plus one or more NRTIs had substantial declines in viral load. These children had an average baseline viral load of 10,000 copies. After 48 weeks, 61% had viral load below 400 copies (Starr).

• The ACTG 338 study showed that children who have already received treatment with AZT or AZT/ddI benefitted from switching to either d4T/ritonavir or AZT/3TC/ritonavir, with approximately 60% in each arm achieving viral load below 500 copies after 12 weeks. However, the average baseline viral load in this study was only 21,000. Approximately 10% of children discontinued the liquid ritonavir treatment in the first 12 weeks of the study.

• One small, retrospective study found that a four-drug regimen including two protease inhibitors was more effective in children than a single protease combination, despite the children on the dual protease combination having more advanced HIV disease and higher viral load at baseline (Arlievsky).

Paradoxically, where treatment is successful in suppressing viral load, it may be more successful than in adults. Ten out of twelve children treated with AZT/3TC/ritonavir within the first 24 months of infection experienced viral load suppression below 400 copies and eventual disappearance of HIV antibody and cytotoxic lymphocyte responses, suggesting that viral suppression was so profound, the immune system could not respond to HIV (Luzuriaga). If treatment were stopped, it is unclear whether HIV would rebound strongly due to the absence of any HIV-specific immune responses.

Following on from the virological and immunological benefits of treatment in children, there is now data which shows antiretroviral therapy has reduced deaths and illness significantly. For example, HIV-related death in children fell by 78% in the US between 1995-1998 (Lindegren).

Tolerability and side effects

Body fat changes have been reported in some children on antiretroviral therapy, although most research is from small, clinic-based studies and thus provides little reliable information regarding incidence or risk factors in children (Verweek 2003). See Body fat changes on antiretroviral therapy (lipodystrophy) - overview in Anti-HIV therapy: Body fat and metabolic changes whilst on treatment for more details.

Amprenavir has been studied in children but is often poorly tolerated. In particular, children under four years should not take amprenavir due to the risk of serious side-effects.

Differences in diagnosis

The detection of HIV infection in children born to HIVpositive mothers is complex, and testing methods are changing rapidly. All babies born to women with HIV acquire maternal antibodies, and it is not easy to differentiate their own from their mother's. All babies have HIV antibodies from their mother's bloodstream, irrespective of whether they are HIV infected themselves. These maternal antibodies persist for approximately ten months, and may last as long as eighteen months. HIV antibody tests on these young babies will only show if the mother is infected, and are not helpful in differentiating between infected and uninfected infants. Many parents want to have an early definitive diagnosis, as repeated observation and testing can be distressing and disruptive. Many doctors would like to be able to diagnose truly HIV infected infants earlier in order to start treatment. There is an increasing availability of prophylaxes and therapies, and it is known that about 25% of babies who are HIV infected develop clinical symptoms of AIDS or die within the first year of life.

When it is known that a child is at significant risk of HIV infection (e.g. when the mother is known to be HIV-positive), a number of measures should be employed to try to ascertain if the baby is HIV infected as quickly as possible, and before maternal antibodies clear. These more sensitive tests are carried out routinely at major centres experienced in dealing with children at risk of HIV infection. Most of these centres will be happy for less HIV experienced centres to contact them for advice and assistance with care.

An accurate early diagnosis of HIV infection in infants depends on tests which establish the true presence of HIV virus, as opposed to HIV antibodies, in the baby. These tests are:

• p24 antigen or aciddissociated p24 antigen tests.

• Virus culture.

• HIV DNA Polymerase Chain Reaction tests (PCR).

PCR and virus culture tests should be performed at repeated intervals: within 48 hours of birth (testing at this point identifies about 40% of infected children), at about 48 weeks, and at 36 months. US guidelines on testing state that if two or more IgG antibody tests (see below) are negative at least six months after birth, with a one month interval between the two tests, it is reasonable to assume that the child is not infected.

In addition, tests of the infant's immune system may help to make an early diagnosis. These tests are:

• High immunoglobulin levels (IgG, IgA, IgM).

• Inverted CD4:CD8 ratio.

• Persistent HIV antibody after 18 months.

Finally, many infants have some clinical manifestations of HIV infection by 6 months of age. In the USA, researchers suggest that the majority of children can be identified with currently available technology by 34 months old.

All infants born to HIV-infected mothers should ideally receive a six week course of oral AZT, and should not be breast-fed.

Health care workers should not treat a recently delivered mother who knows she is HIV positive differently from any other mother. For some women (and families), this will be a time for planning for the future, despite the fact that the HIV status of the baby will not be known for some time. This will include decisions about any measures which might lower the chances of post natal HIV transmission, such as not breast-feeding. It will also be important to consider how to cope with the uncertainty and monitoring of the baby in the first few months. This will also be a time to plan who needs to know about the baby's risk of HIV infection. Most women will find it helpful to inform their GP and Health Visitor, but only if these professionals are sympathetic to the woman's HIV status. Knowing that the baby might be HIV infected can mean that appropriate care can more quickly be offered.

Although the majority of babies born to women with HIV are not infected, health care workers will understandably usually treat the baby as HIV infected until their HIV status is clarified. The baby may be offered prophylaxis against PCP before a definite diagnosis is made. The prophylaxis is known to be relatively safe in infants, and PCP is a significant risk to the baby in the first few months of life if she or he is HIV infected.

References

Arlievsky NZ et al. Quadruple antiretroviral therapy combining nelfinavir and saquinavir in HIV-infected children. 38th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, abstract I-4, 1998.

Buchacz K et al. Delayed onset of pubertal development in children and adolescents with perinatally acquired HIV infection. Journal of Acquired Immune Deficiency Syndromes 33: 56-65, 2003.

De Rossi A et al. Increased thymic output after initiation of antiretroviral therapy in human immunodeficiency virus type 1-infected children in the Paediatric European Network for Treatment of AIDS (PENTA) 5 Trial. Journal of Infectious Diseases 186(3): 312-320, 2002.

Dickover R et al. Early HIV gene evolution in perinatally infected infants predicts the rate of disease progression. Eighth Conference on Retroviruses and Opportunistic Infections, Chicago, abstract 510, 2001.

Doerholt K et al. Paediatric antiretroviral therapy audit in South London. HIV Medicine 3(1): 44-48, 2002.

Essajee SM et al. Immunologic and virologic responses to HAART in severely immunocompromised HIV-1-infected children. AIDS 13(18): 1523-2532, 1999.

Faye A et al. Mortality and morbidity in HIV-infected infants treated before 6 months of age. Second International AIDS Society Conference on HIV Pathogenesis and Treatment, Paris, abstract 33, 2003.

Havens PL et al. Efficacy of protease inhibitor therapy in children and adolescents with HIV infection. 38th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, abstract I-3, 1998.

Hoffmann F et al. Effect of Antiretroviral Triple Combinations Including the Protease Inhibitor Nelfinavir in Heavily Pretreated Children with HIV-1 Infection. European Journal of Medical Research 7(7): 330-334, 2002.

Ioannidis JP et al. Perinatal transmission of human immunodeficiency virus type 1 by pregnant women with RNA virus loads <1000 copies/ml. Journal of Infectious Diseases 183(4): 539-545, 2001.

Kalish LA et al. Evaluation of human immunodeficiency virus (HIV) type 1 load, CD4 T cell level, and clinical class as time-fixed and time-varying markers of disease progression in HIV-1-infected children. Journal of Infectious Diseases 180(5): 1514-1520, 1999.

Lindgren ML et al. Trends in selected causes of death among children (less than 13 years) who died with HIV infection, 1987-1998. Eighth Conference on Retroviruses and Opportunistic Infections, Chicago, abstract 676, 2001.

Luzuriaga K et al. Early therapy of vertical human immunodeficiency virus type 1 (HIV-1) infection: control of viral replication and absence of persistent HIV-1-specific immune responses. Journal of Virology 74(15): 6984-6991, 2000.

Melvin AJ et al. Efficacy and toxicity of antiretroviral therapy using 4 or more agents: application of a strategy for antiretroviral management in human immunodeficiency virus-infected children. Archives of Pediatric Adolescent Medicine 156(6): 568-573, 2002.

Msellati P et al. Highly active antiretroviral therapies among HIV-1-infected children in Abidjan, Cote d'Ivoire (NARS 1244). Second International AIDS Society Conference on HIV Pathogenesis and Treatment, Paris, 2003.

Nakiyangi JS et al. Child survival in relation to mothers HIV infection and survival: evidence from a Ugandan cohort study. AIDS 17: 1827-1834, 2003.

Pollock BH et al. Risk factors for pediatric human immunodeficiency virus-related malignancy Journal of the American Medical Association 289: 2393-2399, 2003.

Spira R et al. Natural history of human immunodeficiency virus type 1 infection in children: a five year prospective study in Rwanda. Pediatrics 104(5): e56, 1999.

Starr ST et al. Combination therapy with efavirenz, nelfinavir, and nucleoside reverse-

transcriptase inhibitors in children infected with human immunodeficiency virus type 1. New England Journal of Medicine 341(25): 1874-1881, 1999.

Steiner E et al. Highly active antiretroviral treatment is more effective in treatment naive than in nonnaive HIV-1 infected children. Fourth International Congress on Drug Therapy in HIV Infection, Glasgow, abstract 302, 1998.

Taha TE et al. Association of HIV-1 load and CD4 lymphocyte count with mortality among untreated African children over one year of age. AIDS 14(4): 453-459, 2000.

Taha TE et al. Effect of HIV-1 antiretroviral prophylaxis on hepatic and hematological parameters of African infants. AIDS 16(6): 851-858, 2002.

Van Rossum A et al. Clinical and virologic response to combination treatment with indinavir, zidovudine, and lamivudine in children with human immunodeficiency virus-1 infection: a multicenter study in the Netherlands. On behalf of the Dutch Study Group for Children with HIV-1 infections. Journal of Pediatrics 136(6): 780-788, 2000.

Van Rossum A et al. Therapeutic drug monitoring of indinavir and nelfinavir to assess adherence to therapy in human immunodeficiency virus-infected children. Pediatric Infectious Diseases Journal 21(8): 743-747, 2002.

Verweel G et al. Clinical and laboratory assessment of lipodystrophy syndrome in HIV-1-infected children treated with HAART. Second International AIDS Society Conference, Paris, abstract 1099, 2003.

Walker SA et al. Response to highly active antiretroviral therapy varies with age: the UK and Ireland Collaborative HIV Paediatric Study. AIDS 18: 1915-1924, 2004.