Point-of-care viral load tests at delivery improve early preventive treatment for children at risk of HIV

Point-of-care viral load testing for infants and rapid initiation of antiretroviral treatment reduced deaths among infants with HIV by 67% in the first six months after birth in a large, randomised study in Mozambique and Tanzania, LIFE study investigators reported at the 30th Conference on Retroviruses and Opportunistic Infections (CROI 2023) on Monday.

The study also showed that measuring mothers' viral load at delivery using a point-of-care test resulted in more infants being flagged as ‘high risk’ for HIV acquisition. Infants flagged as ‘high risk’ for HIV acquisition were also more likely to receive enhanced postnatal prophylaxis if their mothers had received a point-of-care viral load test.

Currently in many African countries, less than half of infants exposed to HIV are tested for the virus and many caregivers never receive test results. Even when an infant is diagnosed with HIV, less than half start antiretroviral treatment. Faster diagnosis, without the need to send samples to a laboratory and await the results, could improve infant diagnosis and treatment initiation.

Diagnosing infants quickly, retaining them in care and providing them with treatment as soon as possible is critical, as infants with HIV are at high risk of death during the first months of life without appropriate treatment. 

The LIFE study was designed to investigate whether infant point-of-care viral load tests at birth led to a reduction in deaths and improvement in viral suppression, when compared to the standard practice of testing for HIV at approximately six weeks after birth.

The study randomised 28 health facilities in Mozambique and Tanzania to provide either:

• Point-of-care early infant diagnosis and immediate antiretroviral treatment initiation for infants diagnosed with HIV (intervention arm)

Or:

• Point-of-care early infant diagnosis at week 4-6 after delivery and linkage to antiretroviral treatment (control arm).

The study enrolled 6605 infants.

During the study, 124 infants were diagnosed with HIV (1.9% of participants). More than half of infants (51%) were diagnosed at birth, 32% by six weeks of age and 16% by 12 weeks of age. Just under 9% of the infants' mothers were not on antiretroviral treatment at the time of delivery and 90% of mothers had viral load above 1000 copies/ml at delivery.

Ninety-three per cent of infants diagnosed with HIV started antiretroviral treatment within two days, including 92% of those diagnosed at birth.

Mortality was higher in infants in the control arm. Fourteen per cent in the control arm died, after a median of 14 weeks, compared to 5% in the intervention arm, after a median of 23 weeks. The risk of death was 67% lower in the intervention arm (adjusted hazard ratio 0.328, 90% confidence interval 0.09-1.09, p=0.070). However, due to the smaller than expected number of infants diagnosed with HIV, this difference was of borderline statistical significance. Longer-term follow-up found no significant difference at 12 or 18 months after birth.

There was no difference in severe clinical events (a combined endpoint of mortality, hospitalisation and severe medical events), retention or in a combination of all endpoints.

The study investigators say that it’s possible that poor viral suppression in infants due to poorly tolerated treatment with lopinavir/ritonavir granules may explain the lack of clinical benefit beyond six months. At month 6, 37% of infants in the intervention group and 23% in the control group had viral loads below 1000 copies/ml. Although these proportions improved somewhat at months 12 and 18, a substantial proportion in both arms remained virally unsuppressed and over a quarter of participants had withdrawn from the study or been lost from follow-up by month 18.

Nevertheless, they say that point-of-care viral load testing and rapid treatment for infants is feasible in rural health settings where care is provided by nurses. The study investigators also pointed to the very low HIV transmission rates observed in this study in routine care; in Tanzania, 0.6% of infants born to mothers with HIV had acquired HIV by the age of 12 weeks, and 2.6% in Mozambique. The study team had anticipated that 4% of infants would acquire HIV in this study.

Maternal point-of-care viral load testing at delivery

In 2016, the World Health Organization recommended enhanced postnatal antiretroviral prophylaxis (ePNP), usually consisting of either daily zidovudine and nevirapine, or daily zidovudine, lamivudine and nevirapine for the first six weeks of life for infants judged to be at high risk of HIV acquisition (see 2018 programmatic guidance here). The enhanced prophylaxis regimen (or nevirapine alone) should continue for a further six weeks if the mother is breastfeeding the infant.

Infants are considered to be at high risk if their mother was diagnosed with HIV at delivery, started antiretroviral treatment less than four weeks before delivery or had a viral load test result above 1000 copies/ml less than four weeks before delivery.

As this definition of high risk depends on having timely information about maternal viral load, the LIFE study assessed whether a maternal point-of-care viral load test increased the number of infants identified as higher risk for HIV acquisition (in Mozambique and Tanzania) and the subsequent provision of enhanced postnatal antiretroviral prophylaxis for infants at higher risk (in Tanzania only). Infants received zidovudine and nevirapine as prophylaxis in this study if they were classified as high risk, or nevirapine alone if they were judged to be at low risk of HIV acquisition.

Health facilities were randomised to provide either point-of-care viral load testing and clinical assessment of infant HIV exposure risk (the intervention arm) or clinical assessment of infant HIV exposure risk (the control arm). Enhanced postnatal antiretroviral prophylaxis was provided for both arms on the basis of maternal risk factors in Tanzania and for all infants in Mozambique.

Point-of-care viral load testing was carried out by nurses at the health facilities and viral load results were available on the same day.

The study enrolled 6512 mothers with HIV, of whom 72% were diagnosed before the end of the second trimester and 78% were virally suppressed at delivery. Women had a mean age of 29 years. Almost all (99%) were taking antiretroviral treatment; approximately three-quarters were taking dolutegravir, tenofovir and lamivudine or emtricitabine. Participants had been on antiretroviral treatment for a median of six months at the time of delivery but 22% were not virally suppressed at delivery.

Out of 6568 infants, 19% in the intervention arm and 4% in the control arm were classified as high risk for HIV exposure (p<0.0001). Eighty-two per cent of infants in the intervention arm were classified as high risk on the basis of the mother’s point-of-care viral load test result.

As all infants at the Mozambique study sites received enhanced antiretroviral prophylaxis, the study assessed the proportions of high-risk infants in each study arm who received enhanced antiretroviral prophylaxis at Tanzania study sites.

Higher-risk infants in the intervention arm were almost four times more likely to receive enhanced antiretroviral prophylaxis compared to the control arm (odds ratio 3.75). Fifty-nine per cent in the intervention arm and 31% in the control arm received enhanced antiretroviral prophylaxis. The remainder of high-risk infants received prophylaxis with nevirapine alone.

The study investigators say that the suboptimal rates of enhanced prophylaxis in both study arms raises the question of whether all infants born to mothers with HIV should receive enhanced antiretroviral prophylaxis in regions with a large proportion of high-risk infants. But in regions where the proportion of high-risk infants is low, they say that this strategy may lead to unnecessary exposure of many infants to zidovudine.