Abstract
Background:
Younger age at antiretroviral therapy (ART) initiation has been associated with smaller HIV reservoirs. We investigated whether younger age of ART initiation is associated with testing negative and weaker signal on a standard HIV diagnostic test in treated children.
Methods:
At exit from a longitudinal study at two sites in Johannesburg, South Africa, 316 school-aged, HIV-infected children on continuous ART started at a median of 6.3 months of age, were tested with standard total nucleic acid (TNA) PCR used for infant diagnosis. All negative results were repeated. Simultaneous viral load (VL) and CD4+ T-cell counts/percentages, along with data collected over the prior four years, were used in multivariable regression to predict negative PCR results and higher cycle threshold (Ct) values (weaker signal).
Results:
Seven children (2.2%, 95% CI: 0.6–3.8) in the full cohort had negative PCR results; all seven were in a subset of 102 (6.9%, 95% CI: 2.0–11.8) who had initiated ART 0–4 months of age and had VL<50 copies/mL at the time of PCR testing. Only one repeat tested as negative. Younger age at ART initiation, VL<50 copies/mL at time of test, sustained VL<400 copies/mL, lower CD4+ T-cell counts and ever treated with efavirenz were significant predictors of weaker signal on the diagnostic test.
Conclusions:
In a small proportion of children who start ART in the first months of life and remain on continuous therapy, standard diagnostic HIV PCR tests may result as negative. Repeat testing may resolve uncertainty of diagnosis.
Introduction
Timely initiation of antiretroviral therapy (ART) for perinatally-infected infants is especially critical given their otherwise rapid disease progression and mortality [1]. Fortunately in South Africa, life-saving ART for infants and young children has been a possibility for more than a decade facilitated by the availability of diagnostic HIV PCR testing as part of the national roll-out of ART since the program’s inception [2].
In more recent years, early treatment has garnered renewed scientific attention given evidence that earlier age at ART initiation is associated with a smaller size of the persisting viral reservoir while remaining on ART [3–8]. Success of ART is generally measured by sustained viral load (VL) below the detection thresholds of assays that quantitate HIV RNA in plasma. However, ART does not clear infection and HIV DNA remains in latent cells and tissue compartments. This persisting viral reservoir is the primary obstacle to HIV cure and necessitates life-long adherence to ART [9, 10].
Quantitation of total HIV DNA in white blood cells is a valuable approach to describe the size of the HIV viral reservoir [11]. Although this biomarker over-estimates the size of the viral reservoir since it includes all HIV DNA regardless of its capacity to replicate, it is still a robust and clinically-informative biomarker. Notably, total HIV DNA in blood has been shown to be one of the strongest predictors of time to viral rebound in studies of supervised ART interruption [11–13]. There have been anecdotal reports of HIV-infected neonates initiating ART close to the time of birth who subsequently test PCR negative [14, 15]. Since licensed tests for infant diagnosis detect HIV total nucleic acids (RNA and DNA) in whole blood, in a sample with no detectable HIV RNA, this test is not dissimilar to the research assays developed to quantitate the size of the viral reservoir by measuring total HIV DNA.
Very early ART, i.e. within days of birth, has only recently become a realistic possibility in South Africa with the incorporation of birth testing into infant diagnosis programs [16]. For the cohorts of HIV-infected children surviving on ART in clinical programs today, ART initiation in the neonatal period was rare. Nevertheless, given the availability of PCR diagnostic testing as part of the national ART roll-out, there are sizable numbers of children who started ART in first few months of life (early infancy).
The aim of this study was to investigate the frequency and predictors of testing negative and having weak signal on a standard HIV diagnostic PCR test in a cohort of older children living with HIV and taking ART in Johannesburg, South Africa, who had been on ART for almost 10 years. We hypothesized that children who started ART at younger ages would be more likely to test negative, and have weaker signals even if positive, than children who started ART at older ages.
Methods
Study population
We selected participants from the Childhood HAART Alterations in Normal Growth, Genes, and aGing Evaluation Study (CHANGES) cohort study conducted at two sites: Empilweni Service and Research Unit (ESRU), Rahima Moosa Mother and Child Hospital and Perinatal HIV Research Unit (PHRU), Chris Hani Baragwanath Hospital in Johannesburg, South Africa. Between 7 February 2013 and 12 August 2014, CHANGES enrolled into an observational, longitudinal study 553 HIV-infected children who had been part of clinical trials and programs at these two sites. At ESRU, 237 children were enrolled into the larger CHANGES cohort from those who had completed participation in a strategies trial of switching children initially suppressed on lopinavir/ritonavir to efavirenz[17] and 46 children enrolled from those not eligible for this trial but who had been participants in an earlier trial of switching to nevirapine[18]. At PHRU, 182 children were enrolled into the larger CHANGES cohort from the CHER study[1] and 88 children were enrolled who had started ART under 2 years of age as part of the Wellness Programme at the site. As part of CHANGES, children had VL and CD4+ T-cell counts and percentages monitored every 6 months. At each 6-month visit, the child’s current ART regimen was verified or updated through an interview with the caregiver regarding the child’s medical record. At initial enrollment into CHANGES, pre-ART characteristics, including age at ART initiation, pre-ART VL and CD4+ T cell counts and percentages, and initial regimen were abstracted. Data on any further ART regimen changes and interruptions were also recorded. The study was observational and no decisions about clinical management were made as part of the study. Methods of CHANGES have been previously described [19, 20].
At the last CHANGES visit, between 22 May 2017 and 12 April 2018, we added blood sampling for an HIV diagnostic PCR along with VL and CD4+ T-cell counts and percentages. A parent or guardian was informed of the results and any child with a negative result was re-tested. In addition, for each child with a negative test, a control child from the same site with a positive result was recalled. PCR-positive controls were matched to PCR-negative cases by age at PCR test (within a 1-year window), age at start of ART as closely as possible, and date of PCR (within 2 weeks of the PCR-negative case). For cases and controls, diagnostic PCR, VL and CD4+ T-cell counts, and percentages were repeated. The overall observational study, as well as this additional testing, was approved by the Institutional Review Boards of the University of the Witwatersrand and Columbia University. Written informed consent was obtained from a parent or guardian and assent was obtained from children.
Laboratory methods
All samples were processed at the Bio Analytical Research Corporation (BARC) laboratory in Johannesburg, South Africa. The HIV diagnostic test used was the HIV-1 total nucleic acid (TNA) PCR (Roche COBAS® TaqMan® HIV-1 Qualitative Test Version 2·0 assay, Roche Molecular Systems, Inc., Branchburg, NJ). The result of the test (positive, indeterminate, negative) was recorded as well as the cycle threshold (Ct) value. At the PCR visits, VL was measured with the Abbott RealTime HIV-1 assay (Abbott Laboratories, Des Plaines, IL) with a lower detection threshold of 40 copies/mL and CD4+ T-cell counts and percentages with the TruCount Method (BD Biosciences, Germany).
During the prior four years of the CHANGES observational study, six-monthly VL and CD4+ T-cell counts were done alternately at BARC (as described above) or as part of the routine program at the National Health Laboratory Service (NHLS). When done at NHLS, both VL and CD4+ T-cell counts and percentages were completed using COBAS AmpliPrep/Taqman HIV-1 system (Roche Holding AG, Basel Switzerland). The lower detection threshold for VL on this system was also 40 copies/mL.
Statistical approach
For this analysis, we included 235/283 children enrolled in CHANGES at ESRU and still in follow-up and 81/88 children enrolled in CHANGES at PHRU who had been part of the Wellness programme for a total of 316 children (Figure 1).. We excluded 162/182 children from the CHER study at PHRU who had undergone testing because of the complexity of their clinical histories, which had included ART interruption, and which had previously identified a child in remission [21].
Figure 1:
Flow diagram describing the inclusion into this analysis of participants from the Childhood HAART Alterations in Normal Growth, Genes, and aGing Evaluation Study (CHANGES) cohort from the Perinatal HIV Research Unit (PHRU), Chris Hani Baragwanath Hospital and Empilweni Service and Research Unit (ESRU), Rahima Moosa Mother and Child Hospital, Johannesburg, South Africa.
First, we examined the frequency of negative PCR results based on the manufacturer’s defined cycle threshold (Ct) cut-off. Next we utilized the Ct value from the PCR test as an indicator of signal strength. Although not designed as a quantitative test, higher Ct values indicate the presence of less HIV nucleic acid and lower Ct values the presence of more HIV nucleic acid. PCR results that were reported out as negative were assigned a Ct value 1 point higher than the maximum observed Ct value in our sample which was 35.2. We used the result of the first PCR if more than one was obtained.
To examine predictors of the results of the PCR test, we utilized data on the children at the time of PCR test as well as that collected during the observational study prior to the test. This included sex and age, duration of ART, ART regimen and VL and CD4 T-cell count and percentage at the time of the PCR test. Based on already collected data, we also examined age at the start of ART, pre-ART VL and CD4 T-cell counts and percentages, starting regimen, and any changes in regimen. Past history of viral suppression was based on the 6-monthly VL tests undertaken during the prior 4 years of the CHANGES study.
We compared the proportions of children with negative PCR results across groups using Chi-Squared and Fisher Exact Tests. We calculated 95% confidence intervals (CI) around these proportions based on the binomial method. Associations between signal strength based on Ct values were analyzed using multivariable linear regression models. To test for group differences on continuous variables, we used Wilcoxon rank-sum tests. Correlations were measured using the Pearson correlation coefficient. We used locally-weighted scatterplot smoothing (LOWESS) to describe the relationship between age at ART start and Ct Values. Analysis was completed using SAS version 9.4 (SAS Institute Inc., Cary, NC).
Results
The 316 HIV-infected children included in this analysis had started ART at a median of 6.3 months of age and had been on uninterrupted ART for a median of 9.9 years until HIV PCR testing at a median of 10.6 years of age. Most (91.1%) had VL <50 copies/mL at the time of testing. There were some differences in the characteristics of the children between the two clinical sites (Table 1).
Table 1:
Characteristics of 316 HIV-infected children in care on antiretroviral therapy (ART) at two sites in Johannesburg, South Africa
| ESRU (N=235) |
PHRU (N=81) |
Total (N=316) |
P | |
|---|---|---|---|---|
| Characteristics at PCR Test | ||||
| Age (years) | <0.0001 | |||
| Median | 10.1 | 11.2 | 10.4 | |
| IQR | 9.1 – 11.6 | 10.4 – 12.2 | 9.4 – 12.0 | |
| Sex, N (%) | 0.2007 | |||
| Male | 118 (50.2) | 34 (42.0) | 152 (48.1) | |
| Female | 117 (49.8) | 47 (58.0) | 164 (51.9) | |
| Time on ART (years) | <0.0001 | |||
| Median | 9.4 | 10.8 | 9.9 | |
| IQR | 8.6 – 10.8 | 10.0 – 11.5 | 8.8 – 11.0 | |
| Current ART Regimen, N (%) | <0.0001 | |||
| EFV-based | 171 (72.8) | 2 (2.5) | 173 (54.8) | |
| LPV/r-based | 62 (26.4) | 73 (90.1) | 135 (42.7) | |
| Other-based1 | 2 (0.9) | 6 (7.4) | 8 (2.5) | |
| Viral Load (copies/mL), N (%) | 0.7729 | |||
| <50 | 212 (90.2) | 76 (93.8) | 288 (91.1) | |
| 50–399 | 8 (3.4) | 2 (2.5) | 10 (3.2) | |
| ≥400 | 15 (6.4) | 3 (3.7) | 18 (5.7) | |
| CD4 Count (cells/mm3) | 0.2159 | |||
| Median | 945 | 890 | 922 | |
| IQR | 749–1147 | 678 – 1111 | 727–1147 | |
| Missing, N (%) | 1 (0.4) | 0 (0) | 1 (0.4) | |
| CD4 Percentage (%) | 0.0012 | |||
| Median | 37.3 | 33.9 | 36.7 | |
| IQR | 32.6 – 41.8 | 29.6 – 39.0 | 31.7 – 41.2 | |
| Missing, N (%) | 1 (0.4) | 0 (0) | 1 (0.3) | |
| Characteristics at Time of ART Start | ||||
| Age at ART Start (months) | 0.0170 | |||
| Median | 6.6 | 5.3 | 6.3 | |
| IQR | 3.7 – 13.1 | 3.1 – 8.6 | 3.5 – 12.3 | |
| Age at ART Start (months), N (%) | 0.2389 | |||
| ≤2 | 28 (11.9) | 16 (19.8) | 44 (13.9) | |
| 3 | 26 (11.1) | 8 (9.9) | 34 (10.8) | |
| 4 | 24 (10.2) | 8 (9.9) | 32 (10.1) | |
| 5–6 | 38 (16.2) | 18 (22.2) | 56 (17.7) | |
| 7–8 | 21 (8.9) | 10 (12.4) | 31 (9.8) | |
| 9–12 | 35 (14.9) | 8 (9.9) | 43 (13.6) | |
| 13–24 | 60 (25.5) | 13 (16.1) | 73 (23.1) | |
| 24–32 | 3 (1.3) | 0 (0) | 3 (1.0) | |
| Starting Regimen, N (%) | <0.0001 | |||
| LPV/r-based | 189 (80.4) | 77 (95.1) | 266 (84.2) | |
| NVP-based | 0 (0) | 4 (4.9) | 4 (1.3) | |
| RTV-based | 45 (19.2) | 0 (0) | 45 (14.2) | |
| Other (ABC, AZT, 3TC) | 1 (0.4) | 0 (0) | 1 (0.3) | |
| Pre-ART viral load (copies/mL), N (%) | 0.0483 | |||
| <100,000 | 22 (11.6) | 3 (4.6) | 25 (9.8) | |
| 100,000 – 749,999 | 59 (31.1) | 14 (21.5) | 73 (28.6) | |
| ≥750,000 | 109 (57.4) | 48 (73.9) | 157 (61.6) | |
| Missing | 45 (19.2) | 16 (19.8) | 61 (19.3) | |
| Pre-ART CD4 Percentage (%), N (%) | 0.6755 | |||
| <25 | 139 (59.2) | 50 (61.7) | 189 (63.6) | |
| ≥25 | 77 (32.8) | 31 (38.3) | 108 (36.4) | |
| Missing | 19 (8.1) | 0 (0) | 19 (6.0) | |
Other regimen bases include: nevirapine (N=2), bictegravir (N=3), dolutegravir (N=1), atazanavir (N=1), and lamivudine/abacavir only (N=1).
Seven (2.2%, 95% CI: 0.6–3.8) of 316 children on ART had negative PCR results. No indeterminate results were reported. All seven children were initiated on ART 0–4 months of age (range 49–128 days, median 60 days) and had VL <50 copies/ml at the time of the negative PCR test. Four of seven children with negative results were male. Only one of the seven was from PHRU. In the subset of 102 children (72 children at ESRU and 30 children at PHRU) who had initiated ART 0–4 months of age, remained on ART and had VL <50 copies/mL at the time of testing, the proportion with negative PCR results was 7/102 (6.9%, 95% CI: 2.0–11.8).
Only one child of seven with an initially negative PCR returned a negative result on re-test. One child of seven with an initially negative PCR had a VL above detection (1330 copies/mL) on re-test. Characteristics of the children testing negative and their matched PCR positive controls are shown in Table 2. The median Ct value at second test was 33.0 in PCR-negative cases and 29.6 in PCR-positive controls (p=0.0115). Median CD4+ T-cell percentage at first test was 29.5 for PCR-negative cases and 36.8 for PCR-positive controls (p=0.3714). The median CD4+ T-cell percentage at the repeat test was 29.2 and 35.8 in cases and controls (p=0.3253).
Table 2:
Individual results of each of the 6 HIV-infected children on antiretroviral therapy who tested PCR negative and their matched controls testing PCR positive
| Pair 1 (ESRU) | Pair 2 (ESRU) | Pair 3 (ESRU) | Pair 4 (ESRU) | Pair 5 (ESRU) | Pair 6 (ESRU) | Pair 7 (PHRU) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pos/Neg | PCR − | PCR + | PCR − | PCR + | PCR − | PCR + | PCR − | PCR + | PCR − | PCR + | PCR − | PCR + | PCR − | PCR + |
| Age (years) | 9.2 | 8.5 | 8.0 | 9.2 | 11.0 | 11.5 | 8.9 | 9.2 | 8.2 | 9.0 | 8.8 | 9.3 | 10.3 | 11.8 |
| Sex | Female | Male | Male | Female | Male | Female | Female | Male | Male | Male | Male | Male | Female | Female |
| Age Started ART (months | 3.6 | 2.5 | 1.6 | 2.1 | 2.3 | 3.0 | 4.2 | 4.0 | 1.6 | 1.5 | 1.8 | 1.8 | 2.0 | 2.0 |
| Current ART Regimen | EFV, | LPV/r, | EFV, | EFV, | EFV, | EFV, | EFV, | LPV/r, | LPV/r, | EFV, | EFV, | EFV, | LPV/r, | LPV/r, |
| 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | 3TC, | |
| ABC | AZT | ABC | ABC | ABC | ABC | ABC | ABC | ABC | ABC | ABC | ABC | AZT | AZT | |
| CD4 % | 47.09 | 38.96 | 37.27 | 30.59 | 26.09 | 45.37 | 29.76 | 36.77 | . | 30.61 | 29.32 | 43.37 | 24.57 | 33.96 |
| CD4 Count (cells/mm3) | 1018 | 678 | 1113 | 577 | 262 | 954 | 420 | 935 | . | 549 | 1451 | 988 | 749 | 495 |
| VL at PCR (copies/Ml) | LDL | LDL | LDL | LDL | LDL | LDL | LDL | <40 | LDL | LDL | LDL | LDL | LDL | 58 |
| Ct Value at PCR | 0 | 31.6 | 0 | 31.5 | 0 | 29.2 | 0 | 29.1 | 0 | 29.8 | 0 | 28.8 | 0 | 29.6 |
| Repeat PCR Result | positive | positive | positive | positive | positive | positive | positive | positive | positive | positive | positive | positive | negative | positive |
| Repeat PCR Ct Value | 33.7 | 31.4 | 32.9 | 30.2 | 32.5 | 29.6 | 30.6 | 29.7 | 35.1 | 27.7 | 33.0 | 29.0 | 0 | 28.4 |
| Days after first PCR that 2nd PCR was done | 36 | 36 | 16 | 28 | 28 | 19 | 15 | 13 | 15 | 15 | 19 | 26 | 13 | 14 |
| VL at Repeat PCR | LDL | 40 | LDL | LDL | LDL | LDL | LDL | LDL | 1330 | LDL | LDL | LDL | LDL | LDL |
| CD4 % at Repeat PCR | 48.56 | 35.79 | 40.48 | 31.59 | 29.23 | 41.91 | 28.64 | 36.17 | 29.21 | 31.80 | 29.75 | 44.74 | 16.20 | 36.83 |
| CD4 Count at Repeat PCR (cells/mm3) | 856 | 748 | 888 | 767 | 512 | 1286 | 811 | 1401 | 587 | 791 | 1311 | 888 | 586 | 499 |
| VL history prior to PCR | All <50 | Ever ≥50, All <400 | All <50 | Ever ≥50, All <400 | Ever ≥50, All <400 | Ever ≥50, All <400 | Ever ≥50, All <400 | Ever ≥50, All <400 | All <50 | Ever ≥50, All <400 | All <50 | All <50 | All <50 | Ever >1000 |
| Time on ART (years) | 8.9 | 8.3 | 7.9 | 9.1 | 10.8 | 11.2 | 8.6 | 8.9 | 8.0 | 8.9 | 8.6 | 9.2 | 10.1 | 11.7 |
| Pre-ART VL copies/Ml | 127,000 | >750,000 | >750,000 | Unknown | >750,000 | 140,000 | 4,500 | 46,000 | 517,000 | 440,000 | >750,000 | 616,000 | Unknown | Unknown |
Comparisons of the ART-treated children testing PCR-negative vs. PCR-positive are shown in Table 3. Children with negative PCR results were younger at ART initiation (median age at ART initiation of 2.0 months) compared to that of those returning a positive result (median age at ART Initiation of 6.1 months (p=0.0007). All children with negative results had sustained VL <400 copies/mL throughout CHANGES prior to the PCR test, while only 183 (59.2%) of the 309 children testing positive had sustained this level of virologic suppression (p=0.03). CD4+ T-cell percentage at time of PCR test was lower (median 29.3%) in those with negative results compared to those with positive results (median 36.7) (p=0.049).
Table 3:
Comparison of the HIV-infected children on antiretroviral therapy who tested PCR-positive vs. those who tested PCR-negative
| PCR-positive (N=309) |
PCR-negative (N=7) |
Overall (N=316) |
P | ||
|---|---|---|---|---|---|
| Characteristics at PCR Test | |||||
| Sex, N (%) | 0.7145 | ||||
| Male | 148 (47.9) | 4 (57.1) | 152 (48.1) | ||
| Female | 161 (52.1) | 3 (42.9) | 164 (51.9) | ||
| Age at PCR Test (Category) | 0.6564 | ||||
| 8–10 years, N (%) | 184 (59.6) | 6 (85.7) | 190 (60.1) | ||
| 11–12 years, N (%) | 103 (33.3) | 1 (14.3) | 104 (32.9) | ||
| 13–14 years, N (%) | 22 (7.1) | 0 (0) | 22 (7.0) | ||
| Cohort, N (%) | 0.6825 | ||||
| ESRU | 229 (74.1) | 6 (85.7) | 235 (74.4) | ||
| PHRU | 80 (25.9) | 1 (14.3) | 81 (25.6) | ||
| ART Regimen, N (%) | 0.5489 | ||||
| EFV-based | 168 (54.4) | 5 (71.4) | 173 (54.8) | ||
| LPV-based | 133 (43.0) | 2 (28.6) | 135 (42.7) | ||
| Other-based1 | 8 (2.6) | 0 (0) | 8 (2.5) | ||
| Ever transitioned to EFV, N (%) | 0.2499 | ||||
| Yes | 184 (59.6) | 6 (85.7) | 190 (60.1) | ||
| No | 125 (40.5) | 1 (14.3) | 126 (39.9) | ||
| Viral Load (copies/Ml), N (%) | 1.0000 | ||||
| <50 | 281 (90.9) | 7 (100.0) | 288 (91.1) | ||
| 50–399 | 10 (3.2) | 0 (0) | 10 (3.2) | ||
| ≥400 | 18 (5.8) | 0 (0) | 18 (5.7) | ||
| CD4+ T-cell Count (cells/mm3) | 0.2891 | ||||
| Median | 924 | 749 | 922 | ||
| IQR | 734–1147 | 420–1113 | 727–1147 | ||
| Missing, N (%) | 1 (0.3) | 0 (0) | 1 (0.3) | ||
| CD4 Percentage (%) | 0.0492 | ||||
| Median | 36.7 | 29.3 | 36.7 | ||
| IQR | 32.0–41.3 | 26.1–37.3 | 31.7–41.2 | ||
| Missing, N (%) | 1 (0.3) | 0 (0) | 1 (0.3) | ||
| Virologic Control in Study Up to PCR (copies/Ml) | 0.0115 | ||||
| All <50 | 57 (18.5) | 5 (71.4) | 62 (19.6) | ||
| Ever ≥50, all <400 | 126 (40.8) | 2 (28.6) | 128 (40.5) | ||
| Ever ≥400, all <1000 | 49 (15.9) | 0 (0) | 49 (15.5) | ||
| Ever ≥1000 | 77 (24.9) | 0 (0) | 77 (24.4) | ||
| Characteristics at Time of ART Start | |||||
| Age at ART Start (months) | 0.0007 | ||||
| Median | 6.1 | 2.0 | 6.0 | ||
| IQR | 3.6–12.4 | 1.7–3.6 | 3.5–12.3 | ||
| Age at ART Start (months), N (%) | 0.0008 | ||||
| ≤2 | 39 (12.6) | 5 (71.4) | 44 (13.9) | ||
| 3 | 34 (11.0) | 0 (0) | 34 (10.8) | ||
| 4 | 30 (9.7) | 2 (28.6) | 32 (10.1) | ||
| 5–6 | 56 (18.1) | 0 (0) | 56 (17.7) | ||
| 7–8 | 31 (10.0) | 0 (0) | 31 (9.8) | ||
| 9–12 | 43 (13.9) | 0 (0) | 43 (13.6) | ||
| 13–24 | 73 (23.6) | 0 (0) | 73 (23.1) | ||
| 25–32 | 3 (1.0) | 0 (0) | 3 (1.0) | ||
| Starting Regimen, N (%) | 1.0000 | ||||
| LPV/r-based | 260 (84.1) | 6 (85.7) | 266 (84.2) | ||
| NVP-based | 4 (1.3) | 0 (0) | 4 (1.3) | ||
| RTV-based | 44 (14.2) | 1 (14.3) | 45 (14.2) | ||
| Other (ABC, AZT, 3TC) | 1 (0.3) | 0 (0) | 1 (0.3) | ||
| Pre-ART VL (copies/Ml), N (%) | 0.5474 | ||||
| <100,000 | 24 (9.6) | 12 (16.7) | 253 (9.8) | ||
| 100,000 – 749,999 | 71 (28.5) | 2 (33.3) | 73 (28.6) | ||
| ≥750,000 | 154 (61.9) | 3 (50.0) | 157 (61.6) | ||
| Missing | 60 (19.4) | 1 (14.3) | 61 (19.3) | ||
| Pre-ART CD4 percentage (%), N (%) | 0.7077 | ||||
| <25 | 185 (63.8) | 4 (57.1) | 189 (63.6) | ||
| ≥25 | 105 (36.2) | 3 (42.9) | 108 (36.4) | ||
| Missing | 19 (6.2) | 0 (0) | 19 (6.0) | ||
Other bases for regimens: nevirapine (N=2), bictegravir (N=3), dolutegravir (N=1), atazanavir (N=1), and lamivudine/abacavir only (N=1).
4500 copies/mL
Five children had pre-ART VL ≥ 1000 and <10,000 copies/mL and two children <1000 copies/mL
Next we examined associations with PCR signal strength, as determined by the cycle threshold (Ct) values returned by the test, in the entire cohort. Ct values were normally-distributed with a mean of 29.3 and a standard deviation of 2.1 in this population. Younger age at starting ART was significantly associated with returning a higher Ct value (weaker signal). Children who started ART 0–4 months of age, when re-tested at the older age in this study on ART, had a median Ct value of 30.0 (IQR: 29.2–31.4), those starting ART 5–12 months of age had a median Ct value of 29.2 (IQR: 28.0–30.1), and those starting ART 13 months or older had a median Ct value of 28.7 (IQR: 27.9–29.4). Having a VL history consistently below 400 copies/mL in CHANGES, having a VL <50 copies/mL at the time of the PCR test, and ever having been treated with efavirenz were also associated with a higher Ct value (Table 4).
Table 4:
Results of linear regression models predicting cycle threshold (Ct) value from the HIV diagnostic total nucleic acid PCR test
| Univariate Each covariate on their own predicting Ct In whole population (N=316) |
Multivariable All covariates shown entered simultaneously into a model predicting Ct In those VL <50 copies/mL at time of test (N=288) |
Multivariable All covariates shown entered simultaneously into a model predicting Ct In ESRU only and VL <50 copies/mL at time of test (N=211) |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Parameter estimate | Standard error | p-value | Parameter estimate | Standard error | p-value | Parameter estimate | Standard error | p-value | |
| Age at ART start (months) | −0.106 | 0.018 | <0.0001 | −0.115 | 0.018 | <0.0001 | −0.104 | 0.018 | <0.0001 |
| VL sustained <400 copies/mL during CHANGES (Yes vs. No) | 1.804 | 0.214 | <0.0001 | 1.572 | 0.227 | <0.0001 | 1.371 | 0.241 | <0.0001 |
| Ever treated with EFV (Yes vs. No) | 0.633 | 0.235 | 0.007 | 0.563 | 0.238 | 0.019 | 0.576 | 0.294 | 0.051 |
| CD4+ T-cell count (cells/mm3) at PCR test (divided by 100) | −0.029 | 0.037 | 0.426 | −0.104 | 0.038 | 0.006 | −0.098 | 0.038 | 0.010 |
| Site (ESRU vs. PHRU) | 0.403 | 0.265 | 0.130 | 0.504 | 0.264 | 0.057 | - | - | - |
| VL <50 copies/mL at test | 2.345 | 0.387 | <0.0001 | - | - | - | - | - | - |
In the small sub-group who had detectable VL >50 copies/mL at time of test (n=28), there was a significant correlation between VL and Ct value (r=−0.81, p<0.0001) measured at the same time point. When we excluded those with VL >50 copies/ml at time of PCR test, leaving 288 children with VL <50 copies/mL on ART in the analysis, there was a significant negative correlation between CD4+ T-cell count and Ct value (r=−0.161, p=0.006). The direction of the association between CD4+ T-cell count and Ct value was unexpected, namely a lower CD4+ T-cell count associated with higher Ct value (weaker signal, i.e. greater likelihood of test being negative). Other characteristics examined, including sex, age at PCR test, starting regimen, pre-ART VL and pre-ART CD4 percentage, were not associated with Ct values.
In multivariable analysis, age at ART initiation was consistently associated with returning higher Ct values after adjusting for the other covariates (Table 4). Figure 2 demonstrates the independent associations between Ct value, age at ART initiation, and history of consistent viral suppression in CHANGES. In an analysis confined to ESRU and restricted to children who had VL <50 copies/mL at the time of test (n=212), younger age at ART initiation, consistent history of VL <400 copies/mL, ever on efavirenz and lower CD4+ T-cell count were associated with a higher Ct value.
Figure 2:
Scatter plots of HIV PCR cycle threshold (Ct) values (Y-axis) by age of ART initiation in months (X-axis) among 288 HIV-infected treated children who had VL <50 copies/ml when tested. Those who sustained viral load (VL) <400 copies/ml (yes) during the observational study prior to the PCR test are shown in red open circles and those who did not sustain (no) are shown in blue solid dots. Overlaid on the scatterplot are LOWESS (Locally Weighted Scatterplot Smoothing) curves with a dashed red line for those who sustained VL <400 copies/ml (yes) prior to the PCR test (N=187) and a solid blue line for those who did not sustain (no) (N=101).
Discussion
We observed that a small proportion (<3%) of school-aged HIV-infected children who had been on continuous ART for almost ten years had a negative result on a standard HIV diagnostic PCR test. The proportion was higher (7%) when the denominator was restricted to those who had started ART under 5 months of age and had VL <50 copies/ml at the time of test. All but one re-tested as positive, suggesting that these negative results on ART indicate low levels of HIV DNA at the threshold of the detection capability of the assay and not true absence of HIV infection. The clinical histories of the children with negative results were reviewed and no irregularities suggesting misdiagnosis were found. These children had been part of studies at the sites, some for more than a decade, thus the likelihood of these children being misdiagnosed initially is remote.
We, and others, have previously reported negative HIV antibody diagnostic assay results in early-treated, well-suppressed children on continuous ART [22–24]. The frequency of the negative PCR results we observed here is lower than has been reported for the occurrence of negative HIV antibody tests. For example, we previously reported that over a third of children who started ART ≤ 3 months of age tested negative on a standard HIV antibody assay in clinical usage.[22] The known relatively high proportion of “false negative” antibody tests may lead clinicians to consider repeating nucleic acid amplification tests in older children. There has been some evidence to suggest that antiretrovirals used to prevent mother-to-child HIV transmission may partially reduce the sensitivity of standard nucleic acid amplification diagnostic tests and delay diagnosis [25–29]. But in the absence of a gold standard, the frequency and magnitude of this effect has proved difficult to establish. There have also been reports of HIV-infected neonates initiating ART who subsequently test PCR negative [14, 15]. Our cases are different in that, although ART was started early i.e. under 5 months of age, it was not within the neonatal period. Although there is the possibility that standard diagnostic tests may become negative in early-treated children who remain on ART, the proportion appears to be quite low. Clinicians should be aware of the possibility of misinterpretation of both HIV antibody and HIV nucleic acid amplification diagnostic tests results in older, virally-suppressed children who have been on ART for some time. Repeat PCR testing appears to resolve any ambiguities of diagnosis in almost all cases. We advise liberal re-testing of children with negative diagnostic test results as in most cases it appears that this will clarify true HIV diagnostic status.
We observed a strong association between the Ct value of the standard HIV diagnostic PCR test and age at ART start. These analyses were done excluding those who had plasma VL >50 copies/mL and most likely reflect a correlation between younger age at ART start and lower levels of HIV DNA. Using assays designed to quantitate cell-associated HIV DNA directly, similar results have been observed in several studies [3–8]. Our results add more weight to the evidence supporting the benefits of early treatment on reducing the size of the HIV viral reservoir. Moreover, our results suggest that output from a standard diagnostic assay may be a useful proxy measure of total HIV DNA quantity when results from research assays are not available.
Since VL in the children in this analysis had been systematically monitored over the course of the observational study for several years, we could review their histories and determine whether they had maintained VL <400 copies/mL consistently over time. A history of sustained viral suppression was found to be associated with weaker signal on the PCR assay. This is consistent with studies quantitating HIV DNA as a biomarker of the size of the viral reservoir combining data from European and Thai cohorts [4]. Although unlike the aforementioned analysis, we did not observe associations with any pre-treatment characteristics including pre-treatment VL and pre-treatment CD4 count or percentage. We did not have sufficient data to determine speed of initial viral decline to confirm or refute a prior observation suggesting this factor to be associated with the size of the viral reservoir independent of age at ART start [30].
Higher Ct values (lower HIV DNA levels) were associated with use of efavirenz as part of the treatment regimen. At the one clinical site (ESRU), the children in the study reported here had been past participants in a prior trial among children initially suppressed on lopinavir/ritonavir who were randomized to either remain on lopinavir/ritonavir-based treatment or to switch to efavirenz-based ART [17]. The second clinical site (PHRU) had not participated in this trial and almost all children had been retained on lopinavir/ritonavir. We therefore examined the association between efavirenz and Ct value only at ESRU and adjusting for other covariates. The association between current or past use of efavirenz and weaker PCR signal persisted. We speculate that this may be via a non-specific effect in that introduction of an antiretroviral agent in a class not already used may have benefits on residual viremia. In addition, there may be drug-specific effects. Efavirenz has been shown to have a higher degree of tissue penetrance than some of the other classes of antiretroviral drugs, which may also contribute to lower levels of HIV DNA [31].
Unexpectedly, higher Ct values (weaker signals) were associated with lower CD4+ T cell counts among children in this study who were suppressed on ART. When ART is initiated when baseline CD4+ T-cell counts are already low, it is usually associated with recovery of CD4+ T-cell counts coincident with decline in plasma VL. Discordance in viral and immune response to ART has been widely discussed and is complicated as CD4+ T-cell count is only a rough measure of immune recovery [32]. We speculate that HIV-specific immune responses stimulated by a greater amount of HIV may account for the higher CD4+ T-cell count observed in those with presumed greater amounts HIV DNA [33]. This would need to be investigated further using more detailed flow cytometric methods to disaggregate the types of CD4+ T-cells that comprise the total count. It should be noted that the CD4+ T-cell counts in our study are almost all within the normal range; thus the clinical significance of these observations is unknown.
Our study has several limitations. The PCR assay we used is validated for use in infant diagnosis and is generally not used in this age group. Thus the performance characteristics are not established for this population. Moreover, the assay is not designed and validated to quantitate HIV DNA directly. We utilized output generated by a validated assay and correlated this output with clinical parameters. Our data suggest this assay can be used as a rough proxy marker, but we do not claim that we have validated this assay to perform in this way. We undertook a cross-sectional study across two sites in Johannesburg, South Africa, consisting of predominantly adherent, healthy children who have responded well to ART. These children are not representative of all HIV-infected children initiating ART. There are undoubtedly many potential confounders that might explain our observations that we did not measure.
In conclusion, we observed that only a small proportion (7%) of HIV-infected, virally-suppressed children who started ART in the first five months of life and remained on continuous ART for almost ten years have negative results on a standard HIV diagnostic PCR test. Although there is the potential for misinterpretation of both HIV antibody and HIV nucleic acid amplification diagnostic tests results in older, virally-suppressed children who have been on ART for some time, it appears that repeat testing is sufficient to resolve ambiguities in the majority.
Acknowledgments:
This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development under Grant numbers HD 073977 and HD 073952.
References
- 1.Violari A, Cotton MF, Gibb DM, Babiker AG, Steyn J, Madhi SA, et al. Early antiretroviral therapy and mortality among HIV-infected infants. The New England journal of medicine 2008; 359(21):2233–2244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Haeri Mazanderani A, Moyo F, Sherman GG. Missed diagnostic opportunities within South Africa’s early infant diagnosis program, 2010–2015. PloS one 2017; 12(5):e0177173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kuhn L, Paximadis M, Da Costa Dias B, Loubser S, Strehlau R, Patel F, et al. Age at antiretroviral therapy initiation and cell-associated HIV-1 DNA levels in HIV-1-infected children. PloS one 2018; 13(4):e0195514. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tagarro A, Chan M, Zangari P, Ferns B, Foster C, De Rossi A, et al. Early and Highly Suppressive Antiretroviral Therapy Are Main Factors Associated With Low Viral Reservoir in European Perinatally HIV-Infected Children. Journal of acquired immune deficiency syndromes (1999) 2018; 79(2):269–276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Uprety P, Patel K, Karalius B, Ziemniak C, Chen YH, Brummel SS, et al. Human Immunodeficiency Virus Type 1 DNA Decay Dynamics With Early, Long-term Virologic Control of Perinatal Infection. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2017; 64(11):1471–1478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.van Zyl GU, Bedison MA, van Rensburg AJ, Laughton B, Cotton MF, Mellors JW. Early Antiretroviral Therapy in South African Children Reduces HIV-1-Infected Cells and Cell-Associated HIV-1 RNA in Blood Mononuclear Cells. The Journal of infectious diseases 2015; 212(1):39–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.McManus M, Mick E, Hudson R, Mofenson LM, Sullivan JL, Somasundaran M, et al. Early Combination Antiretroviral Therapy Limits Exposure to HIV-1 Replication and Cell-Associated HIV-1 DNA Levels in Infants. PloS one 2016; 11(4):e0154391. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ananworanich J, Puthanakit T, Suntarattiwong P, Chokephaibulkit K, Kerr SJ, Fromentin R, et al. Reduced markers of HIV persistence and restricted HIV-specific immune responses after early antiretroviral therapy in children. AIDS (London, England) 2014; 28(7):1015–1020. [DOI] [PubMed] [Google Scholar]
- 9.Ananworanich J, Dube K, Chomont N. How does the timing of antiretroviral therapy initiation in acute infection affect HIV reservoirs? Current opinion in HIV and AIDS 2015; 10(1):18–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kent SJ, Reece JC, Petravic J, Martyushev A, Kramski M, De Rose R, et al. The search for an HIV cure: tackling latent infection. The Lancet Infectious diseases 2013; 13(7):614–621. [DOI] [PubMed] [Google Scholar]
- 11.Avettand-Fenoel V, Hocqueloux L, Ghosn J, Cheret A, Frange P, Melard A, et al. Total HIV-1 DNA, a Marker of Viral Reservoir Dynamics with Clinical Implications. Clinical microbiology reviews 2016; 29(4):859–880. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Saez-Cirion A, Bacchus C, Hocqueloux L, Avettand-Fenoel V, Girault I, Lecuroux C, et al. Post-treatment HIV-1 controllers with a long-term virological remission after the interruption of early initiated antiretroviral therapy ANRS VISCONTI Study. PLoS pathogens 2013; 9(3):e1003211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Williams JP, Hurst J, Stohr W, Robinson N, Brown H, Fisher M, et al. HIV-1 DNA predicts disease progression and post-treatment virological control. Elife 2014; 3:e03821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Haeri Mazanderani AF, Du Plessis NM, Thomas WN, Venter E, Avenant T. Loss of detectability and indeterminate results: Challenges facing HIV infant diagnosis in South Africa’s expanding ART programme. South African medical journal = Suid-Afrikaanse tydskrif vir geneeskunde 2014; 104(8):574–577. [DOI] [PubMed] [Google Scholar]
- 15.Veldsman KA, Maritz J, Isaacs S, Katusiime MG, Janse van Rensburg A, Laughton B, et al. Rapid decline of HIV-1 DNA and RNA in infants starting very early antiretroviral therapy may pose a diagnostic challenge. AIDS (London, England) 2018; 32(5):629–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Moyo F, Haeri Mazanderani A, Barron P, Bhardwaj S, Goga AE, Pillay Y, et al. Introduction of Routine HIV Birth Testing in the South African National Consolidated Guidelines. The Pediatric infectious disease journal 2018; 37(6):559–563. [DOI] [PubMed] [Google Scholar]
- 17.Coovadia A, Abrams EJ, Strehlau R, Shiau S, Pinillos F, Martens L, et al. Efavirenz-Based Antiretroviral Therapy Among Nevirapine-Exposed HIV-Infected Children in South Africa: A Randomized Clinical Trial. JAMA 2015; 314(17):1808–1817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kuhn L, Coovadia A, Strehlau R, Martens L, Hu CC, Meyers T, et al. Switching children previously exposed to nevirapine to nevirapine-based treatment after initial suppression with a protease-inhibitor-based regimen: long-term follow-up of a randomised, open-label trial. The Lancet Infectious diseases 2012; 12(7):521–530. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ramteke SM, Shiau S, Foca M, Strehlau R, Pinillos F, Patel F, et al. Patterns of Growth, Body Composition, and Lipid Profiles in a South African Cohort of Human Immunodeficiency Virus-Infected and Uninfected Children: A Cross-Sectional Study. J Pediatric Infect Dis Soc 2018; 7(2):143–150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Murnane PM, Sigamoney SL, Pinillos F, Shiau S, Strehlau R, Patel F, et al. Extent of disclosure: what perinatally HIV-infected children have been told about their own HIV status. AIDS care 2017; 29(3):378–386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Violari A, Cotton MF, Kuhn L, Schramm DB, Paximadis M, Loubser S, et al. A child with perinatal HIV infection and long-term sustained virological control following antiretroviral treatment cessation. Nature communications 2019; 10(1):412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kuhn L, Schramm DB, Shiau S, Strehlau R, Pinillos F, Technau K, et al. Young age at start of antiretroviral therapy and negative HIV antibody results in HIV-infected children when suppressed. AIDS (London, England) 2015; 29(9):1053–1060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.de Souza MS, Pinyakorn S, Akapirat S, Pattanachaiwit S, Fletcher JL, Chomchey N, et al. Initiation of Antiretroviral Therapy During Acute HIV-1 Infection Leads to a High Rate of Nonreactive HIV Serology. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2016; 63(4):555–561. [DOI] [PubMed] [Google Scholar]
- 24.Stefic K, Novelli S, Mahjoub N, Seng R, Molina JM, Cheneau C, et al. Nonreactive Human Immunodeficiency Virus Type 1 Rapid Tests After Sustained Viral Suppression Following Antiretroviral Therapy Initiation During Primary Infection. The Journal of infectious diseases 2018; 217(11):1793–1797. [DOI] [PubMed] [Google Scholar]
- 25.King CC, Kourtis AP, Persaud D, Nelson JA, Ziemniak C, Hudgens MG, et al. Delayed HIV detection among infants exposed to postnatal antiretroviral prophylaxis during breastfeeding. AIDS (London, England) 2015; 29(15):1953–1961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Balasubramanian R, Fowler MG, Dominguez K, Lockman S, Tookey PA, Huong NNG, et al. Time to first positive HIV-1 DNA PCR may differ with antiretroviral regimen in infants infected with non-B subtype HIV-1. AIDS (London, England) 2017; 31(18):2465–2474. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Mitchell C, Dross S, Beck IA, Micek MA, Frenkel LM. Low concentrations of HIV-1 DNA at birth delays diagnosis, complicating identification of infants for antiretroviral therapy to potentially prevent the establishment of viral reservoirs. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2014; 58(8):1190–1193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Burgard M, Blanche S, Jasseron C, Descamps P, Allemon MC, Ciraru-Vigneron N, et al. Performance of HIV-1 DNA or HIV-1 RNA tests for early diagnosis of perinatal HIV-1 infection during anti-retroviral prophylaxis. The Journal of pediatrics 2012; 160(1):60–66.e61. [DOI] [PubMed] [Google Scholar]
- 29.Mazanderani AH, Moyo F, Kufa T, Sherman GG. Brief Report: Declining Baseline Viremia and Escalating Discordant HIV-1 Confirmatory Results Within South Africa’s Early Infant Diagnosis Program, 2010–2016. Journal of acquired immune deficiency syndromes (1999) 2018; 77(2):212–216. [DOI] [PubMed] [Google Scholar]
- 30.Persaud D, Palumbo PE, Ziemniak C, Hughes MD, Alvero CG, Luzuriaga K, et al. Dynamics of the resting CD4(+) T-cell latent HIV reservoir in infants initiating HAART less than 6 months of age. AIDS (London, England) 2012; 26(12):1483–1490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Curley P, Rajoli RK, Moss DM, Liptrott NJ, Letendre S, Owen A, et al. Efavirenz Is Predicted To Accumulate in Brain Tissue: an In Silico, In Vitro, and In Vivo Investigation. Antimicrobial agents and chemotherapy 2017; 61(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Kelly C, Gaskell KM, Richardson M, Klein N, Garner P, MacPherson P. Discordant Immune Response with Antiretroviral Therapy in HIV-1: A Systematic Review of Clinical Outcomes. PloS one 2016; 11(6):e0156099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Lederman MM, Calabrese L, Funderburg NT, Clagett B, Medvik K, Bonilla H, et al. Immunologic failure despite suppressive antiretroviral therapy is related to activation and turnover of memory CD4 cells. The Journal of infectious diseases 2011; 204(8):1217–1226. [DOI] [PMC free article] [PubMed] [Google Scholar]