Introduction: HIV-1 polymerase chain reaction (PCR) testing at
birth aims to facilitate earlier initiation of antiretroviral therapy (ART) for
HIV-infected neonates. Data from two years of universal birth testing
implementation in a high-burden South African urban setting are presented to
demonstrate the prevalence and outcomes of diagnostic challenges in this
context.
Methods: HIV-exposed neonates born at Rahima Moosa Mother and Child
Hospital between 5 June 2014 and 31 August 2016 were routinely screened at birth
for HIV-1 on whole blood samples using the COBAS® AmpliPrep/COBAS®
TaqMan (CAP/CTM) HIV-1 Qualitative Test, version 2.0 (Roche Molecular Systems,
Inc., Branchburg, NJ, USA). Virological results were interpreted according to
standard operating procedures with the South African National Health Laboratory
Service. All neonates with non-negative results were actively followed-up and
categorized according to HIV infection status as positive, negative, uncertain
and lost to follow-up (LTFU).
Results: 104 (1.8%) of 5743 HIV-exposed neonates received a
non-negative birth PCR result, for which laboratory data were available for 102
(98%) cases – 78 (76%) tested positive and 24 (24%) indeterminate. HIV
infection status was confirmed positive in 83 (81%) infants, negative in 8 (8%),
uncertain in 5 (5%) and LTFU in 6 (6%) cases. The positive predictive value
(excluding cases of uncertain diagnosis and inadequate testing) following a
non-negative HIV-1 PCR screening test at birth was 0.91 (83/91; 95% confidence
interval: 0.85–0.96). Neonates testing positive at birth had
significantly higher viral load (VL) results than those testing indeterminate at
birth of 4.5 and 3.0 log copies/ml (p = 0.0007),
respectively. Similarly, mothers of neonates with positive as compared to
indeterminate birth test results had higher VLs of 4.5 and 2.7 log copies/ml
(p = 0.0013), respectively. Half of neonates
with an indeterminate birth test were shown to be HIV-infected on subsequent
confirmatory testing, with time to final diagnosis 30 days longer for
these neonates (p < 0.0001).
Conclusion: Indeterminate HIV-1 PCR results accounted for a quarter
of non-negative results at birth and were associated with a high risk of
infection in comparison to the risk of in utero transmission.
Indeterminate birth results with positive HIV PCR results on repeat testing were
associated with later final diagnosis. The HIV-1 status remains uncertain in a
minority of cases because of repeatedly indeterminate results, highlighting the
need for more sensitive and specific virological tests.