Diagnosis of HIV Infection During Early Infancy: How Early Is Early Enough?

(See the major article by Francke et al on pages 1319–28.)

There has been tremendous success in reducing the incidence of mother-to-child human immunodeficiency virus (HIV) transmission, with reductions in the annual number of new infections among children globally of 56% since 2010 and 70% since 2000 [1]. However, the number of new infections among women of reproductive age has not declined in the last 5 years, and adolescent HIV infections, accounting for 30% of new infections in 2014, occur disproportionately in girls and young women; in sub-Saharan Africa, 71% of new infections among youths age 15–19 years were in girls [2]. There remains a large unmet need for family planning in many HIV-endemic countries, with 1.5 million HIV-infected women in these areas continuing to give birth annually. Even with the dramatic scale-up in the number of HIV-infected pregnant women receiving antiretroviral treatment (ART), in 2015 >300 000 HIV-infected pregnant women received no antiretroviral drugs at all [1]; without ART, the risk of mother-to-child HIV transmission is 30%–40% in breastfeeding populations. Thus, it can be anticipated that new pediatric HIV infections will continue to occur for many years—even the 2011 goal of so-called elimination of pediatric HIV infection was to reduce new infections to <40 000 annually, not to zero [3].

Diagnosis of HIV infection during early infancy (commonly known as “early infant HIV diagnosis” [EID]) is critical because HIV disease progression is much more rapid in children than adults, with mortality as high as 68% by age 2 years in infants with perinatal infection (acquired in utero or during the intrapartum period), with much of this risk occurring in the first few months of life [4, 5]. In infants infected in utero or during the intrapartum period, studies suggest that mortality begins to increase at about 3–4 weeks of age, compared with uninfected infants, and is as high as 10% by age 2 months, reaching a peak of 30%–40% between 8 and 12 weeks of age [5–10]. While mortality appears lower for children who acquire HIV infection through breast milk, it is still considerable—an estimated 26% by age 1 year [8].

The Children with HIV Early Antiretroviral Therapy (CHER) trial showed that ART initiation before age 12 weeks in infants with a CD4⁺ T-cell percentage of ≥ 25% and minimal clinical HIV disease reduced mortality by 76% [11]. Even in the absence of mortality, there is a progressive decline in the CD4⁺ T-cell count and clinical progression of HIV disease observed in young infants not receiving ART. In the CHER study, 23% of 532 HIV-infected infants screened for the study at a median age of 7 weeks already had severe CD4⁺ T-cell depletion and/or an AIDS-defining event. In a study of 403 HIV-infected infants in South Africa initiating ART at a median age of 8.4 weeks, 62% had advanced HIV disease (defined as a CD4⁺ T-cell percentage of <25% or World Health Organization–defined clinical stage 3 or 4 disease) at the time ART was started [12]. Thus, very early diagnosis of HIV infection, with rapid initiation of ART, is critical to be able to reduce early mortality in children.

The World Health Organization recommends early infant diagnostic testing by HIV nucleic acid testing (NAT) at age 6 weeks for all HIV-exposed infants, with the rationale that it appears most cost-effective to test at a time that in utero, intrapartum, and early postpartum infection could be detected, given the expense of NAT [13]. However, with this approach, a substantial portion of perinatally infected children may die before they can benefit from diagnostic testing and ART initiation [14, 15]. Conventional EID programs typically involve dried blood spot sample collection at clinics and transport to central laboratories with NAT capacity; ineffective transportation and complicated communication systems result in prolonged turnaround and loss to follow-up before ART can be initiated [16]. For example, the EID turnaround in Zambia from sample collection to return of results to the caregiver was 92 days, and a study in rural South Africa found that ART was not started for 45% of infants with HIV-positive NAT results [17, 18]. With NAT performed at 6 weeks of age, treatment initiation under the best of circumstances may not occur until the infant is 3–4 months of age—after the initial peak in mortality has occurred.

Additionally, while in utero infection in the pre-ART era accounted for approximately 30% of all transmission, there may be a shift in the timing of perinatal HIV infection with maternal ART; although infections are fewer in number, there may be an increased proportion occurring in utero, given the high efficacy of antiretroviral drugs given around the time of birth in preventing intrapartum transmission [19–21]. Consistent with this, data from 710 HIV-infected mother-infant pairs in South Africa, the majority of whom were receiving antiretroviral prophylaxis, demonstrated that 76% of infant HIV infections detectable at age 6 weeks were detectable on day 1 of life; similar data were reported from the United States, where the proportion of in utero infection increased over time, from 27% in 1990–1992 to 80% in 1999–2000 [19, 20]. Thus, it has been suggested that the addition of NAT at birth would be clinically beneficial and cost-effective [22].

The article by Francke et al in this issue of The Journal of Infectious Diseases provides an elegant analysis (with remarkably detailed analytic information in the electronic supplemental appendices) of the expected clinical impact and cost-effectiveness of different strategies of early infant diagnosis in South Africa [23]. If NAT was limited to 1 test, then testing at 6 weeks was clinically and economically superior to testing at birth alone, with 6-week testing remaining cost-effective unless attrition rates prior to testing were >90%. Importantly, they report that performance of NAT at birth and 6 (or 10) weeks of age would further improve clinical outcomes and be cost-effective in South Africa, compared with testing at 6 weeks alone.

This result depends substantially on the time for the test result to return and ART to be initiated, as well as the extent to which earlier ART will reduce early HIV-related mortality. If testing at birth allowed initiation of treatment within 3–4 weeks of testing, it might be hypothesized that ART would have the potential to prevent very early decline in immunologic function and clinical deterioration, averting the progressive increase in mortality that begins at about age 1 month. The potential use of point-of-care technologies offers great promise and could allow initiation of ART on the same day as diagnosis is made (thus, if testing at birth, ART could be initiated as early as the first day of life) [16, 24, 25].

The benefits of very early diagnosis of HIV infection are only achieved with rapid initiation of ART. However, there are major challenges with ART in newborns, particularly preterm infants, given the limited available antiretroviral drugs with appropriate formulations for this age group [26]. Antiretroviral drugs approved for use in term infants are limited to zidovudine, lamivudine, emtricitabine, stavudine, and nevirapine; owing to potential toxicity in neonates/preterm infants, lopinavir-ritonavir is recommended only after 2 weeks of age and 42 weeks after conception. Data on appropriate dosing for preterm infants are only available for zidovudine. There remains a relative dearth of studies to evaluate newer drugs in neonates, although encouraging data have recently been presented on raltegravir pharmacokinetics in term infants [27].

The cost-effectiveness of testing at birth and 6 (or 10) weeks of age in the study by Francke et al also depended critically on avoiding attrition between the NATs at birth and 6 weeks of age; if >37% of infants receiving a negative NAT result at birth did not present for the test at 6 weeks of age, the survival benefit of adding birth testing was lost. There are limited and conflicting data on return for the 6-week test when birth testing has been performed. In 2015, South Africa initiated a policy of universal NAT at birth and 10 weeks of age; while a substantial increase in testing at birth has been observed, some studies suggest decreased NAT coverage at 6–10 weeks of age [28, 29].

In 2016, the World Health Organization modified EID guidelines to state that the addition of NAT at birth can be considered but as a conditional recommendation with low-quality evidence [13]. Potential benefits include an opportunity to avoid loss to follow-up between birth and 6 weeks of age (in 2014, NAT coverage at 6 weeks of age was only 50%) and earlier identification of infected infants with the potential to reduce early mortality and morbidity with rapid ART initiation [30]. The report by Francke et al provides convincing evidence to support this recommendation but also provides a cautionary note. For this approach to be effective, it is critical that infants with negative NAT results at birth undergo close follow-up, tracking, and repeat testing at 6–10 weeks of age and that test results are returned promptly and linked to rapid initiation of ART and HIV care. The desired outcome of EID is to save the lives of infected children, and this requires more than just a test—it requires ensuring linkage to an effective longitudinal care system with age-appropriate antiretroviral drugs and rapid initiation of therapy for HIV-infected infants (and their parents).

Note

Potential conflict of interest. Author certifies no potential conflicts of interest. The author has submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References