Disclosing in utero HIV/ARV exposure to the HIV-exposed uninfected adolescent: is it necessary?

Introduction The tremendous success of antiretroviral therapy has resulted in a diminishing population of perinatally HIV-infected children on the one hand and a mounting number of HIV-exposed uninfected (HEU) children on the other. As the oldest of these HEU children are reaching adolescence, questions have emerged surrounding the implications of HEU status disclosure to these adolescents. This article outlines the arguments for and against disclosure of a child's HEU status. Discussion Disclosure of a child's HEU status, by definition, requires disclosure of maternal HIV status. It is necessary to weigh the benefits and harms which could occur with disclosure in each of the following domains: psychosocial impact, long-term physical health of the HEU individual and the public health impact. Does disclosure improve or worsen the psychological health of the HEU individual and extended family unit? Do present data on the long-term safety of in utero HIV/ARV exposure reveal potential health risks which merit disclosure to the HEU adolescent? What research and public health programmes or systems need to be in place to afford monitoring of HEU individuals and which, if any, of these require disclosure? Conclusions At present, it is not clear that there is sufficient evidence on whether long-term adverse effects are associated with in utero HIV/ARV exposures, making it difficult to mandate universal disclosure. However, as more countries adopt electronic medical record systems, the HEU status of an individual should be an important piece of the health record which follows the infant not only through childhood and adolescence but also adulthood. Clinicians and researchers should continue to approach the dialogue around mother–child disclosure with sensitivity and a cogent consideration of the evolving risks and benefits as new information becomes available while also working to maintain documentation of an individual's perinatal HIV/ARV exposures as a vital part of his/her medical records. As more long-term adult safety data on in utero HIV/ARV exposures become available these decisions may become clearer, but at this time, they remain complex and multi-faceted.


Introduction
With the widespread use of combination, antiretroviral therapy (ART) for the prevention of mother-to-child transmission (PMTCT) of HIV, vertical transmission rates have dwindled to B2% [1Á3].The tremendous success of PMTCT has resulted in a diminishing population of perinatally infected children on the one hand and a mounting number of HIV-exposed uninfected (HEU) children on the other. It is estimated that approximately 20% of all infants born in sub-Saharan Africa are born HEU [4].
Households comprising HIV-infected women and HEU children often face significant socio-economic stressors with limited healthcare access, high levels of perceived stigma and low levels of psycho-social support [5Á7]. Disclosure of a child's HEU status, by definition, requires disclosure of maternal HIV status, and this may be difficult given the mother's right to privacy and concern for safety, particularly with continuing stigma around HIV. Disclosure of a mother's HIV status to her children has been low with rates ranging from 20 to 60% in the United States [8,9] and 40 to 50% in sub-Saharan Africa [5,10]. While much of the disclosure literature has focused on a mother's disclosure of her HIV status to her children, exploring potential benefits to a mother's own health [8,11], little has been published on the disclosure of a child's in utero exposure to HIV and antiretroviral medications (ARVs) and whether this has direct risks or benefits to the child's health [12].
The oldest of HEU children are now reaching adolescence and early adulthood, an important transition period often marked by concerns around diminished healthcare access and utilization [13]. As HEU adolescents transition from paediatric to adult healthcare, many assume responsibility for their own healthcare decisions during an already complex phase of cognitive, psychosocial and developmental changes. This responsibility requires knowledge of their medical history, which may include information about perinatal exposures such as exposure to in utero HIV/ARV. Today, we face unknowns regarding the long-term safety of intrauterine HIV/ARV exposure into adulthood and an increasingly large and aging population of HEU children. At the intersection of these issues, the landscape of HIV disclosure is beginning to consider not only benefits/harms for the mother and her family regarding disclosure of maternal HIV status but also benefits/harms for the child regarding disclosure of a child's perinatal HIV/ARV exposure status. Researchers evaluating the long-term safety of intrauterine HIV/ARV exposures through prospective cohort studies require long-term monitoring of HEU children into adulthood necessitating consent from such individuals when they turn 18, resulting in a need to disclose perinatal HIV/ARV exposures to the HEU participant. Clinicians assuming the healthcare of HEU adolescents may struggle with how to best monitor HEU patients in the setting of a lack of conclusive data on the long-term risks of intrauterine HIV/ARV exposures. HEU adolescents and their mothers may have competing desires for privacy due to persistent stigma and the need to avert other psychosocial harms. Adolescents transitioning to adult care may not be fully emotionally and mentally prepared to assume responsibility for their own health as this can be an unstable period of experimentation and individuation which supersedes desires to participate in consistent healthcare. In this article, we summarize the arguments for and against disclosure of intrauterine HIV/ARV exposure to HEU children/ adolescents.

Discussion
Monitoring of HEU children: current guidelines We begin our discussion with a related but separate question involving whether HEU children merit long-term monitoring, since the answer to this question has direct impact on whether disclosure of a child's intrauterine HIV/ARV exposure should occur. We systematically reviewed all English, French and Spanish articles identified in a PubMed/Medline database up to July 2016 on guidelines for the monitoring of HEU children and contacted several key in-country researchers leading surveillance and research initiatives in this area. While there is no consensus on the type of monitoring which should occur, several countries have developed systems and guidelines (Table 1). Wide variability in the recommended duration and intensity of longitudinal observation exists, perhaps due to the fact that this is a rapidly evolving area where emerging needs of HEU children are slowly rising to the forefront. Mechanisms of monitoring encompass registry/ surveillance programmes and national research cohorts, depending on available resources and competing national health priorities.
The U.S. Department of Health and Human Services recommends that HEU children be followed into adulthood due to the potential for carcinogenicity from nucleoside analogue ARVs [14]; Canadian guidelines mirror this and appeal for the psychosocial support of HEU children [15]. US guidelines also acknowledge a need for ''innovative methods'' to provide follow-up of these children and encourage that information regarding in utero HIV/ARV exposure be ''part of ongoing permanent medical records for children.'' In addition to several HEU research cohorts in both countries, the United States also recently reported a linking system in one state to match subjects from the Perinatal HIV Surveillance database and the state's cancer registry to monitor malignancy risk in HEU children [16]. A similar linkage system had been developed earlier in France, where the national cancer registry was linked in an anonymized fashion [17,18] to the major research cohort with longitudinal monitoring of HEU infants until 18 to 24 months [19Á21]. The UK also has a national surveillance system of HIV-infected pregnant women and their infants (National Study of HIV in Pregnancy and Childhood, or NSHPC), which follows HEU children up to 18 months. National death and cancer event data in the UK have, in turn, been linked to data in the NSHPC to monitor death and cancer rates in HEU children [22,23]. In more resource-constrained settings, such as South Africa and Thailand, national guidelines recommend routine follow-up of HEU infants until approximately 18 months [24,25]. A South African pregnancy and HEU surveillance registry is being launched, which will ultimately include three provinces Á KwaZulu-Natal, Gauteng and the Western Cape.
The differences in national guidelines on HEU longitudinal monitoring may be attributed to the differences in healthcare and research resources between countries. Regrettably, areas where high numbers of HIV/ARV-exposed pregnancies occur are also areas where healthcare, research and public health resources may be the most constrained. Despite the lack of consensus on the type of monitoring which HEU children merit, there does appear to be general agreement that some form of follow-up of HEU children is warranted [26] for the following reasons: 1) The type and timing of ARV exposures continue to evolve, at times outpacing research, making continued surveillance essential, 2) There are still many unknowns regarding long-term effects of this exposure. Given this, we now outline arguments in favour of and against disclosing perinatal HIV/ARV exposure status.

The case for disclosure
The key arguments in favour of disclosure revolve around the assumption that there are substantial benefits (psychosocial and physical) for the child, HIV-infected mother and even other family members. In addition, disclosure may facilitate the conduct of large prospective HEU research cohorts in longterm monitoring, ultimately serving a critical public health function (Table 2).

Psychosocial considerations
Despite the paucity of literature describing the impact of disclosure of a child's perinatal HIV/ARV exposure, several studies suggest positive effects on family relationships when disclosure of maternal HIV status to children occurs [11,27,28]. The Amagugu study in South Africa reported significant reduction in parental stress and child emotional/behavioural problems after an intervention to aid in disclosure of maternal HIV status [27]. In addition to higher family cohesion [29], United States studies have demonstrated lower levels of aggressiveness, poor self-esteem [11] and problem behaviours [28] in children whose mothers had disclosed compared to those who had not.

Physical health considerations
Clear physical harms from intrauterine HIV/ARV exposure would necessitate disclosure to the HEU individual. Several scientific arguments may be made to demonstrate current concerns for physical harms which may exist as a result of the    exposure. First, developing theories on the origins of disease have suggested that foetal programming and the in utero milieu have a durable effect on the long-term health of an individual [30]. The in utero period represents a critical window during which changes may alter the biological setting of a foetus, thus placing the foetus at risk for future disease well into adulthood. For example, direct intrauterine toxins have the capacity to cause harmful effects even decades after the initial exposure, as in the case of antenatal diethylstilboestrol exposure and the increased risk of cervical, vaginal and breast cancer as well as infertility in adulthood [31]. Furthermore, in utero effects may present much later in life [32], such as with increased schizophrenia risk from maternal influenza and toxoplasmosis during pregnancy [33,34] or adult insulin resistance and cardiovascular disease from intrauterine growth restriction [35]. These long-term effects on neurobiological and metabolic pathways may not present with clear disease early in life but as the individual progresses through life and other adult exposures increase, there may be an accumulation of risk along the life spectrum, which places pressure on the programming a foetus may have undergone in utero, thereby increasing the risk of chronic diseases in adulthood [32]. Therefore, to avert the potential for major physical harm such as in the case of diethylstilboestrol exposure, disclosure is necessary in order to properly monitor HEU individuals into adulthood. Second, one could argue that in addition to childhood malignancies [16,17,23,36,37]

Malignancy
Though some studies with less follow-up time have reported low cancer incidence rates, which have not exceeded population norms [16,23,36,37], the French EPF recently reported 10 cases of cancer in 53,052 person-years of follow-up as well as an increased risk (hazard ratio (HR)013.6, 95% CI: 2.5Á73.9) associated with didanosine (ddI)'3TC containing regimens versus zidovudine (AZT) monotherapy in HEU children [18]. In a subsequent study with an extended 153,939 person-years of follow-up of HEU children born between 1984 and 2014, the same group reported no differences in the incidence of cancer amongst HEU children compared to the general population but an increased risk with exposure to first trimester ddI (HR 05.5, 95% CI: 2.1Á14.4) [17].

Mitochondrial toxicity
In France, combination ARVs compared to AZTmonotherapy have been found to be associated with mitochondrial dysfunction (relative risk (RR)02.5, 95% CI: 1.0Á6.5, p00.046), and several infants have shown clinical symptomatology [38,39]. Other studies have shown increased mitochondrial DNA in both AZTexposed versus -unexposed [40,41] as well as HIV/ART-exposed versus -unexposed infants [42,43]. Aberrant mitochondrial morphology has also been demonstrated in infants exposed to in utero HIV/ART [44]. What remains unanswered is if and when these early mitochondrial effects translate into poor long-term health outcomes.

Mental health
A US study of HEU and perinatally HIV-infected children observed a higher prevalence of mental health problems in HEU children (38% vs. 25%, p 00.01) in unadjusted analyses [45]. In the U.S. Child and Adolescent Self-Awareness and Health study of perinatally HIV-infected and HEU youth, both groups exhibited high rates of any psychiatric disorder (49% in HEU youth) [46], and during the one to two years of follow-up, this rate did not decrease (57% at baseline to 54% later) in HEU youth [47].

Cardiovascular and metabolic health
Recent studies have shown decreased left ventricular mass index and early diastolic annular velocity in HIV/ARV-exposed versus -unexposed infants [52]. In addition, increased risk of elevated cardiac troponin T in abacavir-exposed infants (OR 02.33, 95% CI: 1.03Á5.26) and decreased risk of elevated N-terminal pro-brain natriuretic peptide in stavudine-exposed infants (OR 00.13, 95% CI: 0.02Á0.99) have been reported [58], the long-term significance of either of which remains unclear. Lastly, studies have shown acylcarnitine and amino acid analytes, products of intermediary metabolism, were increased in ARV-exposed infants (43% vs. 0%, p 00.02) [57] as well as lower insulin levels and abnormal fuel substrate utilization in HEU infants at six weeks of life [56], which may affect the long-term metabolic health of HEU children.

Research/Public health considerations
Though they may not be feasible in all settings, prospective cohort studies can provide detailed, closely monitored, and well-described long-term outcomes data on HEU children. In order to continue these studies, it is ethically necessary to consent HEU individuals when they turn 18 since the HEU individual may have been an infant/child at enrolment when original consent was provided by a parent. This re-consenting in adolescence would require disclosure of the child's HEU status.
The case against disclosure The central argument against disclosure is that the harms of disclosure (psychological stress to the mother and child, the need to maintain privacy of the mother's HIV diagnosis, etc.) are greater than any benefit that might occur, or more simply, that there is no benefit due to the fact that no substantial health risks from intrauterine HIV/ARV exposure have been identified. Cumulative evidence strongly supports the continued use of ARVs in pregnancy, and data surrounding harmful HEU child outcomes are reassuring.

Psychosocial considerations
Though several studies discussed above have indicated psychosocial benefits to the mother and child from disclosure of maternal HIV status, there are almost an equal number citing worsening psychosocial functioning in children of mothers who disclose compared to those whose mothers do not [5,59Á63]. This increased stressor on an already fragile household environment may produce enough psychosocial harm to argue against disclosure. Lower emotional and social functioning [59] as well as increased externalizing behavioural problems [5] have been reported in cross-sectional studies of children whose mothers disclosed. Adolescents whose mothers disclosed may appear to be at risk for early parentification out of a felt need to support their HIV-infected mother [60,64]. Other reports have shown that these adolescents reported higher rates of emotional distress [28,63], highrisk behaviours [63,65] and negative school performance [62] (Table 3). Even pre-term birth, which has been shown in several studies to be associated with ART [74,89Á92,95] is still an early infant outcome and would occur and be managed well before adolescence, the time of disclosure.

Birth weight
Despite one large study in Botswana which reported an increased risk for small-for-gestational age (SGA)outcomes in HEU infants [88] (odds ratio (OR) for SGA01.5, 95% CI: 1.2Á 1.9), the vast majority of reports have not found a consistent association between in utero HIV/ARVs and low birth weight (LBW) or SGA outcomes. Exposure to antenatal AZT has not been found to be associated with SGA in the United States [66] or LBW in Europe [67]. In addition, studies in the United States [72,74] and another multi-country study [75] have reported no associations between antenatal ART and LBW/ SGA. A large study in Latin America also did not find risks for LBW when comparing classes of ARVs [73]. Lastly, two US studies [69,70] did not find increased risks for SGA, and one Ugandan study did not find increased risks for LBW [71] with intrauterine TDF exposure.

Congenital defects
In general, there has not been evidence for an increased rate of birth defects (overall rates 1.4Á6.2%) associated with HIV/ARV exposure [76Á81, 98,99]. The two largest surveillance registries for congenital anomalies in the UK [77] and the United States [100] have found low rates of birth defects consistent with other cohorts in Europe [76] and the United States [78]. Few reports have emerged from low-income countries, but one pilot ART registry from South Africa and Zambia identified a 6.2% prevalence rate for all and 2.2% for major congenital anomalies [79]. Despite earlier reports in humans revealing neural tube defects in infants exposed to efavirenz (EFV) early in gestation [99,109], a more recent meta-analysis of 2026 infants countered these results and found no risk (RR 00.78, 95% CI: 0.56Á1.08) [93]. In addition, the French EPF [80] and a recent US study [101] found overall low rates of congenital anomalies associated with EFV.

Neurodevelopment
The Pediatric AIDS Clinical Trial Group (PACTG) 219 noted no differences in mental or psychomotor developments in 1840 HIV/ARV-exposed versus -unexposed children [82]. Similar findings were found in the PHACS cohort when assessing the effects of ART exposure as well as differing ART class regimens [83]. A more recent study within PHACS also evaluated cognitive outcomes in older HEU children and did not find associations between any perinatal ART class regimens and cognitive and academic scores [84]. One study from Thailand reported small reductions in Wechsler Intelligence Scale testing comparing HEU to HUU children but acknowledged the uncertainty around the long-term clinical significance of these findings [108].

Growth
With the exception of a few studies [69,102], most large studies have not reported problems with early postnatal growth after in utero HIV/ARV exposure [70,75,85Á87]. A Thai study found no differences in weight-for-age, weightfor-length, or length-for-age z scores between infants exposed to B7.5 versus ]7.5 weeks of AZT [85]. A Spanish cohort reported similar findings when evaluating HIV/ARV-exposed versus HIV-unexposed infants and HIV-versus HIV/ARVexposed infants [86]. A multi-national study had comparable results when examining ART versus AZT monotherapy exposure [75].

Research/public health considerations
While large prospective research cohorts may be the most comprehensive method to monitor HEU children long-term, they require re-consenting an HEU adolescent and disclosure of HEU status, rendering the feasibility and sustainability of these in all settings challenging. Surveillance programmes with linkage systems for the monitoring of major events may be used instead for long-term monitoring, particularly in resource-constrained settings where the largest proportion of HEU children reside, and would not require disclosure in most circumstances.

Conclusions: to disclose or not to disclose
As we confront the many unknowns outlined above Á the continued high rates of HIV infection in women globally with an increasing and aging population of HEU children Á the tensions surrounding disclosure will need to be considered carefully. Clinicians in both high-and low-resource settings face the difficulty of balancing the need to respect a mother's rights to privacy and prevent further familial psychosocial harm versus the potential benefits to the HEU adolescent and his/her family from disclosure of exposures [12]. In these settings, careful assessment (and re-assessments) of the risk/ benefit ratio, the HEU individual's changing and maturing needs, and the mother's need for privacy should be considered during the discussion of whether to disclose or not. At present, it is not clear that we have sufficient evidence on whether long-term adverse effects are associated with in utero HIV/ARV exposures, making it difficult to mandate universal disclosure. If evidence for a particularly threatening complication from intrauterine HIV/ARV exposure unsurfaces through research, countries may grapple seriously with how best to manage and address this issue, particularly in areas where healthcare infrastructures are already fragile, or health literacy is low. Data on long-term reproductive health effects, immunologic dysfunction, risk of adult onset malignancies, cardiovascular disease, or neurodevelopmental and mental health disorders in adulthood are still inconclusive with no published reports in HEU adults. To meet this void of evidence, research and longterm monitoring likely needs to be continued, and there is           general consensus among health professionals and parents of HEU children that more data need to be collected on the longterm health of HEU individuals [26]. Research methods using anonymized surveillance systems linked to other national registries will prove indispensable as data are gathered to understand whether in utero HIV/ARV exposure may result in long-term harm, but prospective research cohorts evaluating this question will need to contend with the need for disclosure to HEU individuals in order to continue long-term follow-up into adulthood Á a conundrum where the rationale for the research clashes with the reasons for not mandating universal disclosure at present. As more countries adopt electronic medical record (EMR) systems, the HEU status of an individual will be an important piece of the health record which will follow the infant not only through childhood and adolescence but also adulthood, which may cause disclosure to be a moot point once young adults access their records. With increasing understanding of the influence of early intrauterine exposures on long-term health outcomes, this practice of early and continued documentation should become the standard as EMR systems expand, potentially rendering disclosure an easier and more natural process for parents/caregivers. Permanent documentation via EMR of perinatal exposures may also improve research and surveillance/registry efforts which are required in order to continue monitoring into adulthood and ultimately gather essential data which are still lacking. Thus, clinicians and researchers should continue to approach the dialogue around motherÁchild disclosure with sensitivity, an understanding of maternal needs in addition to a child/adolescent's development and readiness to hear information, and a cogent consideration of the evolving risks and benefits as new information becomes available but work to maintain documentation of an individual's perinatal HIV/ARV exposures as a vital part of his/her medical records. As more long-term adult safety data on in utero HIV/ARV exposures become available, these decisions may become clearer, but for the moment, they remain complex and multi-faceted.