Children and HIV/AIDS: from research to policy and action in resource-limited settings

Introduction In 2006 HIV/AIDS was the leading cause of death worldwide for individuals aged 15–49 years. The pandemic is having a dramatic impact on child mortality, with 380 000 children having died of AIDS-related diseases [1]. The same year, it was estimated that 2.3 million children under the age of 15 years were living with the virus, mainly as a result of mother-to-child transmission of HIV (MTCT) [1]. More than 90% of these children were living in sub-Saharan Africa. The number of children orphaned after their parent(s) have died from AIDS is also rising dramatically, reaching 15.2 million children worldwide in 2005. More than five million children are currently living with HIV-related chronically ill family members and will become orphans unless appropriate care and treatment is provided [2]. Considerable progress has been achieved in industrialized countries towards the prevention of new paediatric HIV infections, the provision of adequate treatment to HIV-infected children, and support to vulnerable children and orphans affected by HIV/AIDS. But for many children, especially in low-incomes countries, adequate prevention, care and treatment still remains inaccessible. The aim of this article is to review the state of knowledge in the field of paediatric HIV/AIDS, to describe the research undertaken over the past decade, and to assess the level of implementation of research results, focusing mainly on the experience of African countries. The prevention of mother-to-child transmission of HIV Most paediatric HIV infections are the result of MTCT, which can occur in the peripartum, during late pregnancy and delivery, and postpartum through breastfeeding. In the absence of any intervention, the risk of MTCT is 15–30% in non-breastfeeding populations and 20–45% among populations who practise prolonged breastfeeding, which is of particular concern in Africa [3]. In developed countries, the risk of MTCT can be reduced to 2% through the combination of several preventive interventions: antiretroviral prophylaxis administered to women during pregnancy and labour and to infants during the first weeks of life, elective caesarean delivery, and the complete avoidance of breastfeeding [4]. In low-income countries, however, caesarean delivery is seldom feasible [5] and it is often neither acceptable nor safe for mothers to refrain from breastfeeding in the absence of any specific nutritional support. In these settings, efforts to prevent HIV transmission in infants were thus initially focused on peripartum MTCT. Since 1998, the efficacy of short-course peripartum antiretroviral regimens administered to HIV-infected pregnant mothers and their infants has been established within African and Thai randomized clinical trials [6–9]. These regimens, involving three antiretroviral drugs, zidovudine, lamivudine and nevirapine, resulted in MTCT risk reductions ranging from 37 to 77% compared with no intervention. Combination regimens are more efficacious than single-drug regimens [10–12], as are regimens of longer durations compared with short-course and single-dose regimens [13]. The World Health Organization (WHO) currently recommends the following antiretroviral regimen for preventing MTCT among women who do not have indications for antiretroviral therapy (ART) for their own health: zidovudine from 28 weeks of pregnancy; zidovudine and lamivudine plus single-dose nevirapine at the onset of labour; maternal zidovudine plus lamivudine for 7 days after delivery; and single-dose nevirapine plus one week of zidovudine for newborn infants [4]. Within low-income countries who are able to deliver only a minimal range of antiretroviral drugs, the single-dose nevirapine regimen remains the most feasible and least expensive strategy, with a residual peripartum transmission rate of 12% [7,14]. Although the selection of nevirapine-resistant virus after the administration of single-dose nevirapine is frequent [15,16], this biological event can be reduced by the addition to single-dose nevirapine of a short 3-day postpartum combination of maternal zidovudine plus lamivudine [17,18]. It is now clearly acknowledged that pregnant women in need of ART should initiate treatment as soon as possible [4]. ART improves the woman's health and is also expected to reduce the maternal HIV plasma viral load, the strongest determinant of MTCT [19,20]. Although ART initiated before or during pregnancy may be associated with adverse pregnancy outcomes, in particular with prematurity [21,22], to date the benefits of antiretroviral exposure greatly outweigh any observed risk for the mother or infant [23]. Research has also addressed the long-term efficacy of peripartum interventions, e.g. after the complete cessation of breastfeeding. A diminution or even a loss of efficacy of short-course antiretroviral regimens administrated in the peripartum in African breastfeeding populations was reported [24]. Postnatal interventions aimed at the prevention of HIV through breastfeeding are thus critical to achieve an overall substantial and sustainable reduction in MTCT. Breastfeeding is the most nutritionally adequate infant feeding practice, also providing protection against diarrhoea and acute respiratory infections, especially in early life [25]. In the absence of any targeted postnatal intervention, however, breastfeeding is also responsible for 8.9 HIV infections per 100 child-years of breastfeeding and accounts for 40% of paediatric HIV infections [3,26]. Furthermore, the longer the duration of breastfeeding, the higher the resulting risk of the postnatal transmission of HIV [24,27,28]. Breastfeeding beyond 6 months is a strong determinant of HIV transmission [28]. The assessment and implementation of interventions to prevent postnatal HIV transmission that will balance this risk of MTCT with the possible adverse outcomes for mother and child health is therefore a challenge. The first alternative to prolonged breastfeeding is the complete avoidance of breastfeeding, substituted by infant formula. This strategy is currently enforced in Europe and the United States, but the situation is more complex in low-income countries. A second alternative is to shorten the duration of breastfeeding, with an early cessation implemented at approximately 6 months of age. A third potential intervention consists of the promotion of exclusive breastfeeding, e.g. without introducing any other fluids or solids than breastmilk during the first months of life [29–31]. The combination of these last two interventions reduces the cumulative risk of HIV postnatal transmission while retaining the benefits of exclusive breastfeeding during the first months of life. Given the necessary support, the acceptability of formula feeding and early breastfeeding cessation was high in Abidjan, Côte d'Ivoire [32,33]. In this research setting, no excess in mortality at 18 months was observed in children exposed to alternatives to prolonged breastfeeding when taking into account HIV status [34]. Similarly, in Botswana, the time-to-mortality distributions through 18 months of age were not significantly different between infants receiving 6 months of breastfeeding plus prophylactic infant zidovudine or formula feeding plus one month of zidovudine [35]. Replacement feeding was, however, associated with higher mortality, morbidity, and stigma in less supported field settings [36,37]. Preliminary results from studies conducted in Zambia, Malawi and Kenya suggest increased rates of diarrhoea among children breastfed for 6 months [38–40]. HIV-infected pregnant women must be counselled in choosing a feeding practice adapted to their individual situation, and must be supported in their feeding choice after delivery [41,42]. Other postnatal interventions for the prevention of MTCT include the inactivation of the virus in breastmilk via heat treatment, but their use in a domestic situation requires further practical developments [43,44]. Finally, studies have recently explored the benefits of antiretroviral regimens designed to provide maternal treatment, reducing the maternal plasma HIV viral load, or postexposure prophylaxis to infants during breastfeeding, thus reducing the risk of MTCT in settings in which breastfeeding is common [35,45]. Research is currently underway to assess the usefulness of these strategies [46]. Overall, research has demonstrated that the combination of peripartum and postnatal interventions considerably reduces MTCT rates with a long-term benefit sustained until the age of 18 months [47]. HIV counselling and testing, antiretroviral prophylactic regimens and infant feeding interventions constitute the basic package of prevention of MTCT services. The proportion of pregnant women being offered prevention of MTCT services has slightly increased from 7.6% in 2003 to 9% in 2005 in low-income countries [48], but global prevention of MTCT coverage remains unacceptably low worldwide. In 30 African countries with the highest HIV prevalence, only 5% of HIV-infected women currently access prevention of MTCT interventions [48]. Operational research initiatives have explored some of the barriers to the prevention of MTCT [49–51]. The implementation of prevention of MTCT services is strongly hindered by the quality of operating health systems [52,53], especially in rural areas: lack of decentralized services, poor monitoring, frequent stock ruptures of test kits and antiretroviral drugs. The promotion of an ‘opt-out’ approach to prenatal HIV testing [54,55] and, more generally, the adoption of the recently recommended WHO strategy for provider-initiated HIV testing [56] are likely to improve prevention of MTCT coverage. The lack of involvement of male partners is also underlined; pregnant women who are unable to share their HIV status with their partner may be reluctant to accept interventions that would identify them as being HIV infected [57,58]. Finally, innovative family approaches linking the access to HIV care with HIV prevention efforts are a critical step towards improving the transition from prevention of MTCT research to wide-scale practice, but need to be rolled out [19]. Paediatric HIV/AIDS care In low-income countries, it is estimated that 50% of HIV-infected infants will die before the age of 2 years [59]. Early paediatric HIV diagnosis is thus critical to allow the timely start of appropriate treatment, reduce morbidity and mortality, guide decisions related to child nutrition and improve the quality of life of HIV-infected children [60]. In many low-income countries, however, access to early paediatric HIV diagnosis depends on the local laboratory capacity and the availability of tests, and is often considered too costly and complex. To date, real time polymerase chain reaction is the most valuable and least expensive (less than €20) technology to detect HIV RNA among infants under 18 months of age [61]. Although this technology requires good laboratory infrastructure and human skills, its routine use for programmatic purposes should be encouraged [4]. In rural areas, a dried blood spot (DBS) is used for the collection and storage of blood samples [62,63], although DBS samples still need to be referred to a central laboratory for HIV testing [64]. It is expected that DBS will become the standard tool for improving the coverage of early paediatric HIV diagnosis in low-income countries. After the confirmed diagnosis of HIV infection, the baseline clinical and laboratory assessment for children should include the clinical staging of HIV disease and the measurement of CD4 cells and T lymphocytes [63]. The prevention of opportunistic infections such as tuberculosis or Pneumocystis pneumonia [65] has been the standard of paediatric HIV care for many years. In South Africa, isoniazid prophylaxis reduced the incidence of tuberculosis in HIV-infected children and improved their survival [66]. A Zambian trial demonstrated the efficacy of cotrimoxazole prophylaxis in children older than 12 months [67]. All HIV-exposed children should thus receive cotrimoxazole prophylaxis from the age of 6 weeks in resource-limited settings, irrespective of locally identified resistance to this drug, as recently recommended by WHO [68]. Although cotrimoxazole costs as little as US$0.03 a day, UNAIDS estimates that 4 million children who need this drug do not access it [48]. ART helps HIV-infected children to preserve, enhance, and reconstitute their immune system and therefore reduce the risk of opportunistic infections, to suppress HIV replication, to restore their growth, to improve mental functioning, and overall their quality of life [69–72]. The decision-making process for the initiation of ART in children in low-income countries relies on clinical and immunological assessment [63]. Despite nearly 15 years of experience in the treatment of HIV-infected children in Europe and the United States [73], considerable uncertainty remains as to when to start ART. The benefits of early ART initiation [74,75] need to be balanced against the costs and drawbacks of ART (quality of life, lifelong therapy, viral resistance, adverse effects, and limited second and third-line regimens). WHO recommends that all children classified as WHO paediatric clinical stage 3 or 4 can start ART regardless of the CD4 cell count, total lymphocyte count or the availability of virological test results. Recent data from South Africa from the CHER randomized trial showed that starting antiretroviral drugs within the first 12 weeks of life, rather than starting at the current WHO recommended CD4 cell threshold of 20%, lowered the risk of death [76]. Only a handful of antiretroviral drugs in the current WHO guidelines, however, have solid formulations in doses appropriate for paediatric use and paediatric fixed-dose drug combinations are scarce. Although three recent studies showed satisfactory virological and immunological benefits in children receiving an adult fixed-dose combination ART in fractions [77–79], the lack of pharmacokinetic and pharmacodynamic data on the antiretroviral drugs available to children still contributes to HIV-infected children being underdosed in antiretroviral drugs [80]. There is an urgent need for additional drug formulations for children [81]. In 2007, the first-line antiretroviral regimen recommended by WHO for children includes nevirapine. The immunological and virological response in infants exposed to nevirapine remains of much concern [82], however, and should be further documented. Additional research is also critically needed regarding the support to families and health workers who disclose HIV infection to the child [83], and the psychological development of children. A recent South African study reported a 26% disclosure rate among children aged under 6 years [84], although the importance of paediatric disclosure was unanimously acknowledged. The absence of HIV disclosure to a child is associated with non-adherence to treatment [85–87]. Safe and adequate paediatric HIV/AIDS care deserves further research. The implementation and management of a comprehensive package of services for the care of HIV-infected children requires a minimum quality of health services, in terms of human resources and technical equipment, and is thus a major challenge for all-level health facilities in low-income countries. The provision of paediatric HIV/AIDS care at district level The expected wide-scale introduction of ART for children is indeed only one aspect of a successful care, support and treatment approach. A comprehensive continuum of care for children should also include HIV counselling and testing, cotrimoxazole prophylaxis for both HIV-exposed and HIV-infected children, regular follow-up of HIV-infected children, and community-based support and counselling of caregivers. In many low-income countries, however, such comprehensive paediatric HIV/AIDS care faces numerous challenges within the local district health system [88]. The lack of health staff and inadequate healthcare infrastructure is one of the first constraints to the scaling-up of HIV/AIDS care services. Health professionals required to care for HIV-infected children are in short supply in Africa [89]. The identification and referral of children in need of HIV/AIDS care (coming from inpatient wards, adult antiretroviral programmes, rural clinics, community-based organizations, or prevention of MTCT programmes) is also poor. In South Africa, only 10% of children requiring HIV testing at 12 months were actually tested [90]. The improvement in local patient monitoring [91], the decentralization and extension of prevention of MTCT and HIV/AIDS care and support infrastructure into poorly serviced areas, combined with the deployment of mid-level professional cadres and community-level health workers [92,93] are examples of possible solutions. Anonymous HIV screening of all infants at immunization clinics is also feasible to identify HIV-infected children early for referral into care and treatment programmes [94]. Evidence-based guidelines on paediatric HIV/AIDS care, adapted to the capacity of each health facility, are essential. The specific assessment and management of symptomatic HIV infection at the primary healthcare level was recently integrated within the WHO Integrated Management of Childhood Illness guidelines, and evaluated in South Africa [95]. Additional prospective assessments of clinical algorithms is still required to improve paediatric care and treatment [96]. Experience in adult care [97–99] and recently in African children [100,101] advocates for free access at the point of service delivery to paediatric HIV care and treatment including ART. The various operational constraints to paediatric HIV/AIDS partly explain why currently children represent only 6% of the overall population receiving ART, when this age group represents 14% (n = 600 000) of the total population in need of ART [48]. Political leadership in the fight against HIV/AIDS is, however, steadily growing and several ART programmes are now accessible to children in resource-poor settings [69,71,72,79,102–109]. The research and programmatic experience of caring for children living with HIV-infected family members is also very scarce and highlights the many gaps remaining in the economic, political and social response to children affected by HIV/AIDS. Children affected by HIV/AIDS In low-income countries, the vulnerability of children affected by HIV/AIDS is first physical, and is mainly related to the impact of HIV/AIDS on household economies. The illness or death of a breadwinner leads to difficulties in responding to the basic needs of family members, including accessing antiretroviral drugs. Uninfected children born to HIV-positive mothers thus experience their mother's illness and death at a young age, which contributes to putting them at risk of increased morbidity and mortality [59,110,111]. To date, however, there is only limited evidence of ill health among young orphans who have survived their mother's death [112,113]. The psychological impact of HIV/AIDS on children in low-income countries has been greatly underestimated, and the of a child living in a family affected by HIV/AIDS through many These children have to the and of their HIV-infected especially when are of the of the illness their parent(s) or to HIV Children are their of and their to the family after the from is a frequent and of The loss of adult and such as or is also likely to a of or Overall, much remains to be on the psychological of HIV/AIDS among children, including the specific by age, of and living Research and intervention studies on and social support for children affected by HIV/AIDS are scarce and are focused on the specific difficulties and of In rural access to mental health services was to benefit all regardless of family The number of health workers in mental however, remains in most low-income countries In programmes involving the and between orphans and have the of children affected by HIV/AIDS and also a in the development of children, especially of orphans and children affected by in countries affected by the HIV/AIDS the supply and quality of services are In orphans were less likely to be in than of the same age, as a result of the to family children and The and support to children affected by HIV/AIDS is by the of the political which is in many of the countries by the HIV/AIDS In of the sub-Saharan countries with HIV reaching or only developed a on orphans and HIV/AIDS. This in support systems may explain why families and have been as the of children affected by HIV/AIDS. In African the of child by which children are for by the HIV In the of these have been There is concern over child within the family will be able the expected in the number of orphans This concern has encouraged the development of community-based initiatives from than to or supported by of these care strategies have been to be inadequate for implementation and are are on Overall, research has that of children affected by HIV/AIDS and service to in from each with little between approaches A number of have been developed by Health or for responding to the needs of children affected by HIV/AIDS and the of nearly 30 the to to the growing of children affected by HIV/AIDS [48]. In during the past two efforts to fight against HIV/AIDS have research for child survival in low-income countries. The transition from research to however, remains the coverage of prevention of MTCT should be a There is also currently evidence to out paediatric HIV care and treatment, at least at the same and as adult This requires improving healthcare systems as as political and at and knowledge that orphans and vulnerable children a to HIV/AIDS and a Finally, the fight against HIV/AIDS requires the scaling-up and of primary HIV prevention, as of a comprehensive response that access to treatment and The would to the following for their and This was by in the of the of a for the in on 15 and was by the as a at the Africa for Health and of South was a of the and of