Early antiretroviral treatment in infants

Polly Clayden, HIV i-Base

The recent report of one HIV infected infant with viral control after interrupting antiretroviral treatment (ART) – and maybe a second case – raises the possibility of functional cure for infants started on ART soon after birth.

This spotlight on early ART also raises many issues of diagnosis, treatment and implementation, regardless of whether or not infants are “cured”.

A themed discussion at CROI 2014, led by Marc Cotton from Children’s Infectious Diseases Clinical Research Unit, Stellenbosch University, Tygerberg, South Africa, focused on response to early ART initiation. [1]

Dr Cotton began the session with a recap of the results from the children with HIV early antiretroviral (CHER) randomised trial, published in The Lancet last year. [2]

In CHER 375 infants with HIV infection, diagnosed before 12 weeks with CD4 percentage > 25% were randomised to receive: deferred ART until clinical progression or CD4% drop (standard of care when the trial was conducted); or early ART until 40 weeks, then stop until HIV progression, when ART was restarted; or early ART until 96 weeks, then stop and restart with HIV progression.

The children receiving early ART did so at a median of about 7 weeks of age and those receiving deferred ART at about 20 weeks.

The study showed children receiving early ART had better clinical and immunological outcomes than those receiving deferred ART at 6 years of follow up even with treatment interruption.

Early ART was also associated with less microbial translocation and better neurological development.

Dr Cotton suggested some of the emerging issues with early ART are: virological response, co-factors and coinfection, reservoir kinetics, HIV antibodies and DNA PCR later on, and operational issues. Five posters discussed in this session showed findings from studies looking at aspects of these issues.

Routine use of confirmatory antibody tests not recommended with early ART

Data from CHER suggests that routine antibody tests should not be used to confirm HIV status in children already diagnosed and started on ART at less than six months of age. [2]

Helen Payne and colleagues explained that HIV infected infants starting early ART might become HIV-antibody seronegative after the decay of passively acquired maternal antibody and viral suppression. As immediate treatment for all children under two years is now universally recommended – and WHO guidelines have recently increased this to five years – the frequency of seronegative status following early ART gives rise to questions about clinical management in this age group.

The investigators measured HIV antibody in stored plasma from 75/125 children in the deferred ART arm and 109/126 children in the early ART until 96 week arm at median age of 91.4 weeks (IQR 90.6 – 93.4).

They used three techniques to perform these measurements: 4th generation microparticle enzyme immunoassay HIV antigen/antibody combination; HIV-1/2 qualitative immunochromographic rapid antibody test (assessed by a blinded, independent clinician; and a sensitive in-house ELISA to quantify anti-gp120 IgG and total IgG.

This showed that children in the deferred ART arm had significantly more antibody than those in the early ART until 96 weeks arm according to results from all three tests. Using automated serology, 90% vs 54% of children were seropositive, p<0.0001; with rapid test, 88% vs 47%, p<0.0001; and quantitative anti-gp120 IgG ELISA, median 6,870 ug/ul (IQR 1,706-53,645) vs 230 (IQR 133-13,129), p=0.04.

The investigators noted that levels of total IgG were similar in both groups, suggesting that this effect was specific to HIV responses. All children had detectable anti-gp120 IgG antibody by ELISA.

Eight children (10%) from the deferred ART arm and 49 (46%) from the early ART until 96 weeks arm were seronegative using automated serology. In the deferred arm there was no difference between the automated and rapid test. But in the early ART until 96 weeks arm 9 children with weakly positive automated results had negative rapid tests.

Older age at start of ART was associated with increased anti-gp120 IgG in the deferred arm, p=0.002. By automated serology, starting ART between 12 to 24 weeks of age gave 13.7-fold higher odds of being seropositive by 92 weeks old vs starting ART at 0 to 12 weeks (95% CI 3.1 to 60.2), p=0.001. All 33 children starting ART at 24 weeks of age or more were seropositive.

Higher anti-gp120 values were associated with higher cumulative viral load to week 84, p<0.0001.

In this study, almost half the children starting ART within 12 weeks of birth and 6% of those starting between 12 to 24 weeks were seronegative by commercial serology or rapid tests at approximately two years of age.

In conclusion the investigators suggested that routine antibody tests should not be used to confirm HIV status among children already diagnosed and started on ART at less than six months of age.

Further investigation is needed to determine whether persistence of anti-gp120 IgG represents slow decay or is a response to low levels of HIV replication, they added.

Varying HIV profiles in early treated children seronegative by ELISA

Data from Cameroon suggested that seroreversion is common in infants receiving early ART. [3] This presentation, by Mathurin Tejiokem and colleagues from Cameroon and France, is from an ongoing project looking at residual HIV in early treated infants using ultra-sensitive virological methods.

ANRS-PEDICAM is a cohort of HIV-infected infants followed from one week of life or diagnosed before seven months of age from 2007 to 2011 in three urban referral hospitals in Yaounde and Douala. Five-year follow up is planned.

Children receive first line ART of two NRTIs plus either lopinavir/r (LPV/r) or nevirapine in accordance with local guidelines.

At the time of analysis 192/210 (91.4%) children had started ART at a median of 4.1 months. Of those, 147/192 (76.6%) children were tested with ELISA at a median age of 19.1 months (IQR 18.2 to 20.1) and 26/147 (17.7%) of them were negative for HIV antibodies. Over a third (38.5%) of the 26 children started ART before three months.

Children with negative antibody tests were younger: 3.2 vs 4.7 months, p<0.001. They were also more likely to be WHO stage 1 or 2 when they started ART: 84.6% vs 59.5%, p=0.01. Weight for age z-score, ART regimen and CD4 percentage at start of ART were not associated with a negative antibody test.

Of 12 children with a median follow up of 47.6 months (range 21.8 to 65.1), nine remained HIV-antibody negative. The investigators then explored residual HIV activity in these children using ultrasensitive virological tests.

This revealed varying HIV-DNA and HIV-RNA profiles in the children. One child had no replicating virus (<24 copies/mL) and no detectable HIV-DNA. Five had no replicating virus and varying HIV-DNA levels (range 2.22 to 2.59 log copies/106 PBMC). Three children had replicating virus (range 36 to 97 copies/mL) and HIV-DNA (1.6 to 2.7 log copies/106 PBMC).

Western blot profiles varied from only one antibody peptide (anti-p24) with no replicating virus to several of the gag and pol peptides with replicating viruses.

The investigators concluded that seroversion is common in early treated infants: 17.7% (95% CI 11.9 to 24.8). They cautioned that misleading interpretation of these results might have public health consequences due to reduced adherence.

Work is ongoing to look at the mechanism of seronegativity, especially in antibody negative HIV-infected infants with quantifiable plasma viral load, and the potential of HIV control in these infants.

Better virologic control in infants started on ART before six months of age

Greater virological control was observed in infants starting ART before 6 months of age – compared to 6 to 24 months – in three cohorts at Rahima Moosa Mother and Child Hospital, Johannesburg, South Africa. [4]

Stephanie Shiau and colleagues from the University of the Witwatersrand, Johannesburg and Columbia University, New York, evaluated viral dynamics by age at start of ART. The three cohorts were:

Cohort 1 (Neverest 2) – children started on LPV/r-based ART <24 months of age and were randomised to stay on this regimen or switch to nevirapine after viral suppression.

Cohort 2 (Neverest 3) – children, suppressed on LPV/r, and aged 3 to 5 years at enrolment were randomised to remain on LPV/r or switch to efavirenz. Only children who started ART <24 months of age were included in the analysis.

Cohort 3 (FinHDER) – children participating in a surveillance study of newly diagnosed HIV infected infants and young children. Those who started ART <24 months of age were included.

The mean age at ART start was between 8 and 10 months across the three cohorts and other characteristics were similiar before starting ART.

The investigators reported two main findings from their evaluation.

1. Age at ART start influenced initial viral suppression:

In cohort 1 (Neverest 2), of 323 children, HIV viral loads were higher before ART in 102 children <6 months vs 221 aged 6 to 24 months: 75.6 vs 58.4% had >750,00 copies/mL; 4.7 vs 10.7 had <100,000 copies/mL, p=0.02.

These children had similar time and likelihood of achieving viral suppression by age at ART start. The investigators noted that this suggested faster decline in the younger age group.

In cohort 3 (FinHDER), 145/236 children with a known date of ART start had viral load data available. Analysis of these children found those who started ART <6 months were more likely to suppress <1000 copies/mL by 6 months (58%) vs those starting ART 6 to 24 months of age (33.3%).

2. Age at ART start influenced virological control after viral suppression:

In cohort 1 (Neverest 2 once suppressed) children <6 months at ART start were less likely to fail >1000 copies/mL by 52 weeks from randomisation, p=0.02. Most failures were in the nevirapine switch arm. Children who stayed on LPV/r <6 months at ART start were less likely to blip >50 copies/mL by 52 weeks after randomisation, p=0.04.

In cohort 2 (Neverest 3), of 293 children, those <6 months of age at ART start were less likely to blip above 50 copies/mL vs 6 to 24 months.

The investigators wrote: “These results are consistent with the notion that early initiation of ART may limit formation of the long-lasting viral reservoir in infants.”

Early viral suppression improves neurocognitive outcomes

Virologic suppression during infancy or early childhood is associated with improved neurocognitive outcomes in vertically infected children according to research from IMPAACT 219C and the Paediatric HIV/AIDS Cohort Study (PHACS). [5]

Children and adolescents with perinatally-acquired HIV can have subtle to severe neurocognitive deficits.

Claudia Crowell and colleagues hypothesised that both early ART and use of regimens with better central nervous system (CNS) penetration are associated with improved neurocognitive outcomes in school-aged vertically infected children.

The investigators looked at data from two prospective cohort studies from the United States: PHACS (2007 to 2009) and IMPAACT 219C (2000 to 2006). Children enrolled in the study had completed a cognitive ability assessment (WISC) at >6 years of age, and had at least one viral load measurement before or within 6 months of starting ART.

Viral suppression was defined as two consecutive viral loads <400 copies/mL 1 to 6 months apart, and age of viral suppression was the age at first suppressed viral load.

CNS penetration effectiveness (CPE) scores were determined for first ART regimen. The investigators used multivariate models, adjusted for age, caregiver education level, ethnicity, low birth weight and child’s first language. Results were stratified by age and before and after the availability of ART. An average CPE score was calculated for each time period.

A total of 396 children were included in the analysis. At the time of the WISC assessment they were a mean age of 9.6 years and just over half were girls.

The investigators reported that virologic suppression in early childhood was associated with a 2.2 to 4.4 higher WISC score; this association reached statistical significance for ages 4 and 5, respectively p=0.03 and p=0.02. They noted that the effect size was consistent across all age groups and so non-significance in the younger age groups might be due to lower numbers.

There was no association between CPE score and improved neurological outcome in this evaluation.

Future research from this group will explore the relationships between virologic suppression, CPE scores and neurocognitive aspects including language, executive functions, memory, learning and processing speed.

CMV and clinical outcomes with early ART

Finally, another study from the CHER researchers found early cytomegalovirus (CMV) viraemia was strongly associated with breastfeeding exposure and poorer immunological status before starting ART. [6] They also found early ART appears to improve most effects of concurrent CMV infection.

Prior to use of ART, CMV was associated with faster HIV progression. In sub-Saharan Africa, 90% of infants acquire CMV postnatally in their first year of life but its effect on HIV infection in infants on ART is unknown

Nei-yuan Hsiao and colleagues conducted an evaluation to describe the first year clinical outcomes (censored at 40 weeks post enrolment) of young infants with CMV/HIV coinfection who received ART in CHER.

CMV PCR results were available for 363/451 children in the trial (children with CD4 percentage <25% at screening were included). Children were a median age of 7.7 weeks (6.7 to 9.1) at enrolment and 89 (25%) were CMV PCR positive.

Of these children, 342 (94%) started ART during follow-up: 267 100% in the early treatment group and 75 (78%) in the deferred treatment group.

At enrolment CMV viraemia was associated with any history of breastfeeding, p=0.025; lower median CD4 percentage, p<0.0001; viral load >500,000 copies/mL, p=0.0052; and lower median weight for age and height for age z-scores, respectively p=0.004 and p=0.0097.

At one-year time to disease progression and death were similar between infants receiving early ART by CMV status, respectively negative vs positive: HR 0.618 (95% CI 0.292 to 1.2079), p=2079 and HR 0.481 (95% CI 0.193 to 1.20), p=0.1167.

This suggests that early ART attenuates the effect of CMV on HIV in infants.

References:

All references are to the programme and abstracts of the 21st Conference on Retroviruses and Opportunistic Infections (CROI), 3-6 March 2014, Boston, unless otherwise stated.

Webcasts are CROI 2014 Themed discussion: Response to early ART initiation in infants 4 March 4, 2014 1.30 – 2.30 pm.

http://www.croiwebcasts.org/console/player/22102?mediaType=slideVideo&

Cotton MF. Response to early ART initiation in infants. Introduction themed discussion. 21st CROI. 3-6 March 2014, Boston.
Cotton MF et al. Early time-limited antiretroviral therapy versus deferred therapy in South African infants infected with HIV: results from the children with HIV early antiretroviral (CHER) randomised trial. The Lancet, 382(9904);1555-63, (9 November 2013).
http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(13)61409-9/abstract
Payne H et al. Routine antibody tests have no place in determining HIV status after early ART: evidence from CHER. 21st CROI. 3-6 March 2014, Boston. Poster abstract 922.
http://croiconference.org/sites/all/abstracts/922.pdf
Tejiokem MC et al. Different profiles of HIV in early treated HIV-infected children seronegative by ELISA in Cameroon. 21st CROI. 3-6 March 2014, Boston. Poster abstract 923.
http://croiconference.org/sites/all/abstracts/923.pdf
Shiau S et al. Greater virologic control in infants initiated on ART before 6 months of age. 21st CROI. 3-6 March 2014, Boston. Poster abstract 924.
http://croiconference.org/sites/all/abstracts/924.pdf
Crowell CS et al. Early viral suppression improves neurocognitive outcomes in HIV-infected children. 21st CROI. 3-6 March 2014, Boston. Poster abstract 925.
http://croiconference.org/sites/all/abstracts/925.pdf
Hsiao N et al. Early cytomegalovirus viraemia and clinical outcomes of HIV-infected children in the early ART era. 21st CROI. 3-6 March 2014, Boston. Poster abstract 926.
http://croiconference.org/sites/all/abstracts/926.pdf