HIGH-FREQUENCY failure of combination antiretroviral therapy in paediatric HIV infection is associated with unmet maternal needs causing maternal NON-ADHERENCE

Abstract

Background: Early combination antiretroviral therapy (cART) reduces the size of the viral reservoir in paediatric and adult HIV infection. Very early-treated children may have higher cure/remission potential.

Methods: In an observational study of 151 in utero (IU)-infected infants in KwaZulu-Natal, South Africa, whose treatment adhered strictly to national guidelines, 76 infants diagnosed via point-of-care (PoC) testing initiated cART at a median of 26 h (IQR 18-38) and 75 infants diagnosed via standard-of-care (SoC) laboratory-based testing initiated cART at 10 days (IQR 8-13). We analysed mortality, time to suppression of viraemia, and maintenance of aviraemia over the first 2 years of life.

Findings: Baseline plasma viral loads were low (median 8000 copies per mL), with 12% of infants having undetectable viraemia pre-cART initiation. However, barely one-third (37%) of children achieved suppression of viraemia by 6 months that was maintained to >12 months. 24% had died or were lost to follow up by 6 months. Infant mortality was 9.3%. The high-frequency virological failure in IU-infected infants was associated not with transmitted or acquired drug-resistant mutations but with cART non-adherence (plasma cART undetectable/subtherapeutic, p<0.0001) and with concurrent maternal cART failure (OR 15.0, 95%CI 5.6-39.6; p<0.0001). High-frequency virological failure was observed in PoC- and SoC-tested groups of children.

Interpretation: The success of early infant testing and cART initiation strategies is severely limited by subsequent cART non-adherence in HIV-infected children. Although there are practical challenges to administering paediatric cART formulations, these are overcome by mothers who themselves are cART-adherent. These findings point to the ongoing obligation to address the unmet needs of the mothers. Eliminating the particular barriers preventing adequate treatment for these vulnerable women and infants need to be prioritised in order to achieve durable suppression of viraemia on cART, let alone HIV cure/remission, in HIV-infected children.

Funding: Wellcome Trust, National Institutes of Health.

Keywords: Hiv; Hiv cure; Mother-to-child transmission; Paediatric hiv.

Fig. 1

Study Profile. ^✟ High risk; during pregnancy the mother either seroconverted, had <4 weeks cART, poor cART adherence by history or a documented plasma viral load >1000 HIV RNA copies per mL. *Tested with point-of-care (PoC) whole blood GeneXpert (GXP) HIV-1 Qualitative total nucleic acid (TNA) PCR (Cepheid); but when not available an overnight HIV RNA test; plasma Nuclisens EasyQ v2·0 HIV-1 RNA PCR (bioMérieux). ^✦Laboratory-based dried blood spot TNA PCR (CAP/CTM HIV-1 Qualitative Test v2 Roche) **iLembe, uMgungundlovu, eThekwini and uThungulu districts. ^{✟ ✟}Confirmatory testing included; TNA PCR via CAP/CTM or GXP, HIV RNA PCR via Nuclisens or in selected cases more sensitive methods. IU: in utero, cART: combination antiretroviral therapy.

Fig. 2

Effects of very early ART initiation and universal maternal cART on baseline plasma HIV viral load and CD4+ T cell count. (A) Plasma HIV RNA viral load, (B) absolute CD4+ T cell count and (C) CD4+ T cell percentage of the IU HIV-infected infants at baseline (<21 days of life, prior to combination antiretroviral therapy (cART) initiation) (left) and at 1 month of age (right), where PoC were diagnosed using a point-of-care HIV TNA PCR or overnight RNA PCR test and treated very early (median age 26 h) and SoC were diagnosed by the standard-of-care laboratory-based HIV PCR test and treated early (median age 10 days). (D) Plasma HIV RNA viral load, (E) absolute CD4+ T cell count and (F) CD4+ T cell percentage of the IU HIV-infected infants within 21 days of birth where 2015–2019 is the current cohort (Ucwaningo Lwabantwana) and 2002–2005 are the cohort from the same population prior to universal antenatal cART (PEHSS). (G) Plasma HIV RNA viral load, (H) absolute CD4+ T cell count and (I) CD4+ T cell percentage of the mothers of the IU HIV-infected infants shown in figure d-F, within a month of delivery. Median values and interquartile range displayed. All p-values calculated using the Mann-Whitney U test.

Fig. 3

Baseline infant proviral HIV DNA levels compared to plasma RNA. Baseline infant plasma HIV RNA levels (copies per mL) (left) and proviral HIV DNA levels (copies per million peripheral blood mononuclear cells (PBMC)) as measured by droplet digital PCR (right) for a subset of infants at baseline, a comparison for PoC, diagnosed by point-of-care test and treated very early (median time 26 h) and SoC, those diagnosed by standard-of-care laboratory-based test and treated early (median time 10 days). Median values and interquartile range displayed. Open symbols are below the limit of detection (20 or 100 copies per mL for RNA and variable for each individual test for DNA (2–10 copies per million PBMC). P-values calculated using the Mann-Whitney U test.

Fig. 4

Infant outcomes: retention, mortality and plasma viral suppression. (A) Proportions of infants with the opportunity for 12 months follow-up (excluding 4 infants who were withdrawn by the investigators) who died, were lost to follow-up, those who did not achieve plasma viral suppression by 6 months of age, those who did achieve plasma viral suppression by 6 months of age but had rebounded by 12 months of age, and those who achieved plasma viral suppression by 6 months of age and maintained suppression to at least 12 months of age. (B) Kaplan-Meier analysis of time to plasma viral suppression; a comparison of PoC and SoC groups. PoC infants were diagnosed using a point-of-care HIV TNA PCR or overnight RNA PCR test and treated very early (median time 26 h) and SoC are those diagnosed by the standard-of-care laboratory based HIV PCR test and treated early (median time 10 days). (C) Kaplan-Meier analysis of time to plasma viral rebound; a comparison of PoC and SoC groups.

Fig. 5

Determinants of infant plasma viraemia. (A) Results from plasma ART quantification from Table 3 are compared to those of 11 infants of similar age (median age 21 months) with plasma viral suppression (median duration of suppression 18 months) on cART. Negative levels includes very low levels and levels below the level of detection. P-value calculated using Fisher's exact test. 3TC: lamivudine, ABC: abacavir; LPVr: ritonavir boosted lopinavir. (B) Plasma viral load levels over time for a mother-child pair following combination antiretroviral therapy (cART) initiation, as an example of infant treatment success being determined by maternal cART adherence. Dashed line shows limit of detection (20 HIV RNA copies per mL). (C) A cross-sectional analysis of 101 mother-child pairs, >6 months on cART at the most recent timepoint of contemporaneous (within 14 days) plasma viral load measurement for the mother and child. Suppression is defined as a plasma viral load lower than the detectable limit (20 or 100 RNA copies per mL). Odds ratio is shown and p-value is calculated using Fisher's Exact test. (D) Spearman's non-parametric correlation of maternal and infant plasma load shown in (C).