Secondary Logo

Journal Logo

Epidemiology and Social

Characteristics and outcomes of adolescents living with perinatally acquired HIV within Southern Africa

Tsondai, Priscilla R.a; Braithwaite, Kateb; Fatti, Geoffreyc,d; Bolton Moore, Carolyne,f; Chimbetete, Cleophasg; Rabie, Helenah; Phiri, Sami; Sawry, Shobnaj; Eley, Briank; Hobbins, Michael A.l; Boulle, Andrewa; Taghavi, Katayounm; Sohn, Annette H.n; Davies, Mary-Anna

Author Information
doi: 10.1097/QAD.0000000000002683

Abstract

Introduction

Southern Africa is home to nearly 40% of the estimated 1.6 million adolescents aged 10–19 years living with HIV globally [1–3]. This population is comprised of adolescents living with perinatally acquired HIV (ALPH) as well as those who acquired the infection nonperinatally during childhood or adolescence. ALPH have longstanding HIV infection acquired at a time when their immune systems were not fully developed and experience the rapid physical, mental, and psychosocial changes characteristic of adolescence while living with a chronic, often stigmatized disease [4–6].

Current evidence has shown that in comparison with adults and young children living with HIV, adolescents generally have poorer outcomes [7–9]. They are underserved by the current HIV care services, have lower access to antiretroviral therapy (ART), are more likely to be loss to follow-up (LTFU), and have poorer growth and adherence to treatment [10–12]. Previous research examining outcomes of ALPH within Southern Africa assumed perinatal infection among those entering care before the age of 10 years and examined outcomes of younger adolescents (10–14 years of age) [13]. Although most children with perinatally acquired HIV present for care as infants or children, up to a third have slow-progressing infection and survive into adolescence without treatment [14–20]. Data are lacking on the exact numbers of ALPH in this category, their characteristics, or how their outcomes differ from those presenting earlier.

Using data merged from 15 cohorts in the International epidemiology Databases to Evaluate AIDS in Southern Africa (IeDEA-SA) collaboration, we compared the characteristics (demographic, clinical, anthropometric) and outcomes [transfer out (TFO), LTFU, mortality, and retention] of ALPH who entered care aged less than 10 vs. aged 10–13 years; and explored the predictors of mortality after the age of 13 years.

Methods

Study population and design

We conducted an analysis of prospectively collected data of ALPH in care within sites contributing data to the IeDEA-SA collaboration between January 2004 and December 2017. IeDEA-SA is the Southern African regional collaboration of the IeDEA consortium. It combines routine, observational, prospectively collected, individual patient data from ART programs in six Southern Africa countries: Lesotho, Malawi, Mozambique, South Africa, Zambia, and Zimbabwe (http://www.iedea-sa.org). Cohorts within IeDEA-SA range in size, are predominantly government-funded and follow national HIV treatment guidelines. Each participating site transfers anonymized data to the IeDEA-SA Data Centers at the University of Cape Town in South Africa and the University of Bern in Switzerland at regular intervals (approximately annually) using a standard Data transfer and Exchange Standard for inclusion in combined analyses [21].

Ethics

All participating sites have institutional ethical approval from the relevant Institutional Review Boards (IRBs) to contribute anonymized individual patient data to the IeDEA-SA Data Centers. The Data Centers have ethical approval to receive and analyze anonymized patient data from the individual sites. Most local IRBs waived the need for informed consent for analysis of deidentified routine patient data, but where the IRB required informed consent, this was obtained from the caregiver with verbal assent from the child.

Measurements and outcomes

As mode of HIV infection was not routinely reported in the data, we used age of entry into HIV care less than 13 years as a proxy for perinatal infection, unless a nonperinatal mode of infection was recorded. We included adolescents who entered care aged 10–13 years to capture those children with slow-progressing disease who survive into adolescence without treatment. ALPH were thus defined as patients who entered HIV care aged less than 13 years, with no documentation of nonperinatal causes of acquiring HIV, and with at least one HIV care visit during adolescence. We excluded participants with missing data on sex or with less than 6 months of potential follow-up from entry into care until database closure.

We defined entry into HIV care and into our study as the first recorded HIV care visit in the respective cohort. We defined ART initiation as the earliest date recorded in the databases when a triple-drug ART regimen was dispensed. Characteristics of ALPH at entry into HIV care, at ART initiation, across various ages during adolescence, and at last visit included demographic (age, sex, calendar year, and country); clinical (WHO clinical staging, ART regimens, CD4+ absolute cell count and percentage, and HIV-RNA viral load); and anthropometric (height and BMI) variables. To describe characteristics at entry into care, measurements from the date closest to the date of entry into care up to 1 month after were used. Measurements closest to the date of ART initiation, birthday of interest, and last visit within a window of ±6 months were used to describe the characteristics at ART start, across various ages during adolescence, and at last visit, respectively. In assessing viral suppression, only adolescents in care within sites offering routine (at least annual) viral load measurements were included. HIV-RNA viral load measures from the date closest to each respective birthday, within a window of ±9 months, were used to describe viral suppression across the various ages during adolescence. We used an 18-month window to assess viral suppression, as routine viral load testing is recommended annually, and we wanted to allow enough time on either side of the date of each birthday for a viral load to be done and recorded. Viral suppression was considered as an HIV-RNA less than 400 copies/ml.

Age at entry into HIV care was categorized as less than 10 years and 10–13 years. Severe immunosuppression was defined where the CD4+ absolute cell count or percentage met classification as per WHO immunologic classifications for established HIV infection [22]. Height, and BMI measures were converted to age and sex-specific z-scores [height-for-age z-score (HAZ) and BMI-for-age z-score (BAZ)] using the WHO ‘igrowup_restricted’ Stata macro [23] for up to 5 years of age and the ‘who2007’ Stata macro [24] from 5 to 19 years of age. HAZ and BAZ were further categorized as more than −2, −2 or less to more than −3, and −3 or less.

The outcomes of this analysis included LTFU, TFO, mortality, and retention. We defined LTFU as having no recorded visit in the 12 months prior to the individual cohort database closure, with no record of having died or transferred care. The date of being LTFU was set as the date of last contact with the site. TFO was defined as any documented transfer of care. Mortality included all-cause mortality recorded in the dataset. We defined retention as having a recorded visit in the 12 months prior to database closure with no record of having died or been TFO.

Statistical analysis

Participants entered the analysis on the date of their first visit and exited on the earliest date of death, TFO, LTFU, or their 22nd birthday. Cross-sectional immunological, virologic, and anthropometric characteristics of ALPH as well as mortality, LTFU, TFO, and retention outcomes are presented by age of entry into care (<10 vs. 10–13 years).

In addition, separate cumulative incidence estimates for mortality, LTFU, TFO, and corresponding retention are calculated using competing risks analysis. For ALPH who entered care aged less than 10 years, person-time for cumulative incidence estimates started on the date of their 10th birthday and ended on the date of death, TFO, LTFU, or 22nd birthday, whichever occurred first. Person-time for cumulative incidence estimates for ALPH who entered care aged 10–13 years started on the date of their first visit and ended on the date of death, TFO, LTFU, or 22nd birthday, whichever occurred first.

We used Cox proportional hazards models to identify predictors of mortality following the 13th birthday. Models only included ALPH who had initiated ART by their 13th birthday and were still alive and in care at the age of 13 years, as participants started contributing person-time from the date of their 13th birthday. Models were adjusted for the following characteristics, which were known or assumed to be potential confounders associated with mortality: age at entry into care (aged <10 vs. 10–13 years), sex (male vs. female), ART duration and regimen by 13th birthday, calendar year of 13th birthday, CD4+ cell count at 13th birthday, and HAZ at 13th birthday. Furthermore, we conducted a sensitivity analysis including only those who had been on ART for at least 1 year by the age of 13 years to exclude those with a very high mortality after ART initiation.

All statistical analyses were done using Stata 15.0 (StataCorp, College Station, Texas, USA).

Results

Of the 25 496 ALPH within the IeDEA-SA database who met our eligibility criteria, we excluded 22 with missing data on sex and 348 with less than 6 months of potential follow-up. In total, 25 126 ALPH (51.5% female) from 15 IeDEA-SA sites in six Southern African countries were thus included (Table 1).

Table 1 - Characteristics of adolescents living with perinatally acquired HIV at entry into HIV care.
Characteristic First visit aged <10 years First visit aged 10–13 years All
Total number, n (%) 16 229 (64.6%) 8897 (35.4%) 25 126 (100%)
Sex: male, n (%) 8070 (49.7%) 4115 (46.2%) 12 185 (48.5%)
 Female, n (%) 8159 (50.3%) 4782 (53.8%) 12 941 (51.5%)
Year of birth, median (IQR) 2002 (2000; 2004) 1999 (1996; 2001) 2001 (1998; 2004)
Country, n (%)
 Lesotho 232 (1.4%) 134 (1.5%) 366 (1.5%)
 Malawi 2003 (12.3%) 1402 (15.8%) 3405 (13.5%)
 Mozambique 65 (0.4%) 50 (0.6%) 115 (0.5%)
 South Africa 7758 (47.8%) 3243 (36.4%) 11 001 (43.8%)
 Zambia 4980 (30.7%) 3181 (35.7%) 8161 (32.5%)
 Zimbabwe 1191 (7.3%) 887 (10.0%) 2078 (8.3%)
Age (years) at first visit, median (IQR) 6.7 (4.4; 8.4) 11.4 (10.6; 12.1) 8.6 (5.8; 10.8)
Year of first visit, median (IQR) 2008 (2006; 2010) 2010 (2008; 2013) 2009 (2006; 2011)
 2004–2006, n (%) 5108 (31.5%) 1389 (15.6%) 6497 (25.9%)
 2007–2010, n (%) 7733 (47.6%) 3433 (38.6%) 11 166 (44.4%)
 2011–2017, n (%) 3388 (20.9%) 4075 (45.8%) 7463 (29.7%)
Severely immunosuppresseda, n/N 11 195/16 229 5795/8897 16 990/25 126
n (%) 5778 (51.6%) 2954 (51.0%) 8732 (51.4%)
CD4+ percentageb, n/N 2918/4862 2918/4862
 Median (IQR) 16 (11.0; 23.0) 16 (11.0; 23.0)
CD4+ cell count (cells/μl)b, n/N 6812/11 367 5360/8897 12 172/20 264
 Median (IQR) 351 (175; 606) 283 (123; 488) 321 (148; 553)
 <350, n (%) 3380 (49.6%) 3219 (60.1%) 6599 (54.2%)
 350–500, n (%) 1166 (17.1%) 861 (16.1%) 2027 (16.6%)
 >500, n (%) 2266 (33.3%) 1280 (23.9%) 3546 (29.1%)
HAZ, n/N 8570/16 229 5245/8897 13 815/25 126
 Median (IQR) −2.11 (−3.05; −1.19) −2.20 (−3.05; −1.30) −2.15 (−3.05; −1.23)
 >−2, n (%) 3992 (46.6%) 2319 (44.2%) 6311 (45.7%)
 ≤−2 to >−3, n (%) 2316 (27.0%) 1540 (29.4%) 3856 (27.9%)
 ≤−3, n (%) 2262 (26.4%) 1386 (26.4%) 3648 (26.4%)
BAZ, n/N 8738/16 229 5296/8897 14 034/25 126
 Median (IQR) −0.41 (−1.34; 0.44) −1.07 (−2.01; −0.27) −0.67 (−1.62; 0.21)
BMI (kg/m2), n/N 8615/16 229 5189/8897 13 804/25 126
 Median (IQR) 15.1 (14.1; 16.4) 15.5 (14.3; 16.9) 15.3 (14.1; 16.6)
BAZ, BMI-for-age z-score; HAZ, height-for-age z-score; IQR, interquartile range; n/N, number of nonmissing observations/total eligible.
aAs defined in the WHO 2006 guidelines criteria.
bCD4+ percentage assessed in adolescents that entered care aged less than 5 years; CD4+ cell count assessed in adolescents that entered care aged at least 5 years.

Baseline characteristics at entry into HIV care

Approximately two-thirds (n = 16 229) of adolescents in our analysis entered HIV care before the age of 10 years, at a median [interquartile range (IQR)] age of 6.7 (4.4; 8.4) years. The median (IQR) age at first visit among those who entered care between the ages of 10–13 years (n = 8897) was 11.4 (10.6; 12.1) years (Table 1). Among those with measurements at their first visit, just over half presented for care severely immunosuppressed, with no difference observed between those presenting aged less than 10 vs. aged 10–13 years. Median HAZ was below the −1 z-score and median BAZ below the 0 z-score (Table 1). There were no differences observed in the HAZ and BMI between those presenting aged less than 10 vs. aged 10–13 years (Table 1).

Patient characteristics at antiretroviral therapy initiation

In total, 94.1% of all ALPH initiated ART during follow-up, with those who entered care aged less than 10 years more likely to have initiated ART [97.9%, 95% confidence interval (CI) 97.6; 98.1%] compared with those who presented aged 10–13 years (87.3%, 95% CI 86.6; 86.0%; Table 2). Median (IQR) age at ART start for adolescents who entered care aged less than 10 years was 7.2 (4.8; 8.9) vs. 11.6 (10.8; 12.3) years for those who entered aged 10–13 years. Approximately two-thirds of ALPH with data available were classified as WHO clinical stage 3 or 4, 54.6% (95% CI 53.8; 55.4%) as severely immunosuppressed, 55.1% (95% CI 54.2; 56.0%) as stunted (HAZ ≤ −2), and 16.9% (95% CI 16.3; 17.6%) as wasted (BAZ ≤ −2). Median (IQR) CD4+ cell counts, HAZ, and BAZ at ART start were generally higher among those who entered care aged less than 10 years compared with those who entered aged 10–13 years (Table 2).

Table 2 - Characteristics of adolescents living with perinatally acquired HIV at antiretroviral therapy start.
Characteristic First visit aged <10 years First visit aged 10–13 years All
Ever initiated ART, n (%) 15 883 (97.9%) 7770 (87.3%) 23 653 (94.1%)
Age (years) at ART start, median (IQR) 7.2 (4.8; 8.9) 11.6 (10.8; 12.3) 8.9 (6.1; 10.9)
 0–4.9, n (%) 4142 (26.1%) 4142 (17.5%)
 5–9.9, n (%) 10 701 (67.4%) 10 701 (45.2%)
 ≥10, n (%) 1040 (6.5%) 7770 (100) 8810 (37.3%)
Year of ART start, median (IQR) 2008 (2006; 2011) 2010 (2008; 2013) 2009 (2007; 2011)
Severely immunosuppresseda, n/N 10 823/15 883 4994/7770 15 817/23 653
n (%) 5850 (54.0%) 2791 (55.9%) 8641 (54.6%)
CD4+ percentageb, n/N 2510/4142 2510/4142
 Median (IQR) 15.0 (10.0; 21.1) 15.0 (10.0; 21.1)
CD4+ cell count (cells/μl)b, n/N 7001/11 741 4562/7770 11 563/19 511
 Median (IQR) 329 (170; 568) 251 (109; 423) 297 (142; 505)
 <350, n (%) 3710 (53.0%) 3025 (66.3%) 6735 (58.2%)
 350–500, n (%) 1204 (17.2%) 703 (15.4%) 1907 (16.5%)
 >500, n (%) 2087 (29.8%) 834 (18.3%) 2921 (25.3%)
HAZ, n/N 8702/15 883 4597/7770 13 299/23 653
 Median (IQR) −2.12 (−3.03; −1.22) −2.24 (−3.06; −1.36) −2.16 (−3.05; −1.27)
 >−2, n (%) 4021 (46.2%) 1949 (42.4%) 5970 (44.9%)
 ≤−2 to >−3, n (%) 2443 (28.1%) 1409 (30.6%) 3852 (29.0%)
 ≤−3, n (%) 2238 (25.7%) 1239 (27.0%) 3477 (26.1%)
BAZ, n/N 8860/15 883 4659/7770 13 519/23 653
 Median (IQR) −0.40 (−1.30; 0.44) −1.02 (−1.93; −0.24) −0.62 (−1.54; 0.24)
BMI (kg/m2), n/N 8765/15 883 4583/7770 13 348/23 653
 Median (IQR) 15.3 (14.1; 16.6) 15.7 (14.5; 17.1) 15.4 (14.2; 16.8)
WHO stage, n/N 11 028/15 883 5623/7770 16 651/23 653
 1 or 2, n (%) 3909 (35.5) 2319 (41.2) 6228 (37.4)
 3 or 4, n (%) 7119 (64.5) 3304 (58.8) 10 423 (62.6)
First-line regimen, n/N 14 784/15 883 7263/7770 22 047/23 653
 2 NRTI + EFV, n (%) 7074 (47.8) 3760 (51.8) 10 834 (49.1)
 2 NRTI + NVP, n (%) 6438 (43.5) 3362 (46.3) 9800 (44.4)
 2 NRTI + PI, n (%) 1272 (8.6) 141 (1.9) 1413 (6.4)
ART, antiretroviral therapy; BAZ, BMI-for-age z-score; EFV, efavirenz; HAZ, height-for-age z-score; IQR, interquartile range; NRTI, nucleoside reverse transcriptase inhibitor; NVP, nevirapine; PI, protease inhibitor.
aAs defined in the WHO 2006 guidelines criteria.
bCD4+ percentage assessed in adolescents that started ART aged less than 5 years; CD4+ cell count assessed in adolescents that started ART aged at least 5 years.

Patient characteristics at various ages during adolescence

Among the 16 229 ALPH who entered care before the age of 10 years, approximately 91.9% had initiated ART by their 10th birthday (Supplementary Material 1, https://links.lww.com/QAD/B828). At age 10 years, median (IQR) CD4+ cell count was 716 (465; 1009) cells/μl, with 71.5% of ALPH having a CD4+ cell count more than 500 cells/μl; duration on ART 3.1 (1.4; 5.3) years; with 89.5% on a nonnucleoside reverse transcriptase inhibitor (NNRTI)-based regimen; and 10.5% on a protease inhibitor-based regimen. Among those on a protease inhibitor-based regimen at age 10 years, 37.2% (95% CI 34.6; 40.0%) had been previously switched from an NNRTI (Supplementary Material 1, https://links.lww.com/QAD/B828). Among ALPH in care within facilities offering routine annual HIV-RNA viral load testing and on ART by the age of 10 years (n = 7331), 72.7% (95% CI 71.6; 73.7%) had a documented viral load measurement at age 10 years, and among these, 78.0% (95% CI 76.9; 79.1%) were virally suppressed.

Comparing characteristics of ALPH across various ages during adolescence, median CD4+ cell counts for younger adolescents (10–14 years) were generally higher than for older adolescents (15–19 years). Also, across adolescence, median CD4+ cell counts for ALPH who entered care aged less than 10 years were generally higher when compared with those for ALPH who entered care aged 10–13 years. This trend was similarly observed in a separate analysis restricted to those who had been on ART for at least 1 year. Furthermore, median CD4+ cell counts for ALPH in care from 2014 onwards were generally higher than those from the previous years (Fig. 1).

Fig. 1
Fig. 1:
Median CD4+ cell counts and height-for-age z-scores among adolescents living with perinatally acquired HIV on antiretroviral therapy for at least 1 year across various ages by age of enrolment into HIV care and by calendar period of respective birthdays.

Though median HAZ was consistently below the −1 z-score across adolescence, there was some improvement with increasing age, especially after the age of 14 years (Fig. 1). As noted for CD4+ cell counts, median HAZ was slightly higher in ALPH who entered care aged less than 10 years compared with aged 10–13 years and for those in care from calendar year 2014 onwards, across all age categories (Fig. 1). Amongst a subset of ALPH receiving care within sites offering routine annual viral load assessments, the proportion of virally suppressed decreased with increasing age. However, comparing the proportions of adolescents virally suppressed across various ages after the age of 13 years showed no differences between those who entered care aged less than 10 vs. aged 10–13 years (Supplementary Material 2, https://links.lww.com/QAD/B828).

Patient characteristics at last visit

Characteristics of ALPH at last visit are shown in Table 3. Participants contributed 152 574 person-years of follow-up (median 6.2, IQR 3.1; 8.8 years), of which 91 619 person-years was during adolescence (median 3.1, IQR 1.2; 5.6 years). At last visit, the median (IQR) age was 13.6 (11.7; 16.1) years, with 3210 (12.8%) aged more than 18 years. The median (IQR) duration of follow-up for ALPH who entered care aged less than 10 years was 7.5 (5.1; 9.7) years compared with 3.1 (0.9; 6.1) years for those who entered care aged 10–13 years.

Table 3 - Characteristics of adolescents living with perinatally acquired HIV at last visit.
Characteristic First visit aged <10 years, N = 16 229 First visit aged 10–13 years, N = 8897 All, N = 25 126
Age (years) at last visit, median (IQR) 13.1 (11.4; 15.3) 14.4 (12.5; 17.6) 13.6 (11.7; 16.1)
 10–13.9, n (%) 9935 (61.2) 3978 (44.7) 13 913 (55.4)
 14–17.9, n (%) 5067 (31.2) 2936 (33.0) 8003 (31,8)
 18–22, n (%) 1227 (7.6) 1983 (22.3) 3210 (12.8)
Duration of follow-up (years), median (IQR) 7.5 (5.1; 9.7) 3.1 (0.9; 6.1) 6.2 (3.1; 8.8)
 Total person-years 119 156 33 418 152 574
Duration of follow-up (years) during adolescence (between 10 and 19 years); median (IQR) 3.1 (1.4; 5.3) 3.1 (0.9; 6.1) 3.1 (1.2; 5.6)
 Total person-years 58 201 33 418 91 619
ART duration (years): median (IQR) 7.0 (4.4–9.3) 3.4 (1.3–6.3) 6.0 (3.1–8.6)
CD4+ cell count (cells/μl), n/N 7804/16 229 4704/8897 12 508/25 126
 Median (IQR) 649.5 (418; 907) 456 (247.5; 684) 573 (341; 828)
HAZ, n/N 9055/16 229 4726/8897 13 781/25 126
 Median (IQR) −1.66 (−2.49; −0.87) −2.00 (−2.91; −1.12) −1.77 (−2.64; −0.94)
BAZ, n/N 9098/16 229 4767/8897 13 865/25 126
 Median (IQR) −0.65 (−1.43; 0.07) −0.86 (−1.79; −0.06) −0.72 (−1.54; 0.03)
BMI (kg/m2), n/N 9062/16 229 4706/8897 13 768/25 126
 Median (IQR) 17.0 (15.5; 19.0) 17.2 (15.4; 19.4) 17.1 (15.5; 19.1)
Outcomes, n (%)
 Died 229 (1.4%) 455 (5.1%) 684 (2.7%)
 LTFU 2638 (16.2%) 2970 (33.4%) 5608 (22.3%)
 TFO 2590 (16.0) 1436 (16.1%) 4026 (16.0%)
 Retained 10 772 (66.4%) 4036 (45.4%) 14 808 (58.9%)
ART, antiretroviral therapy; BAZ, BMI-for-age z-score; HAZ, height-for-age z-score; IQR, interquartile range; LTFU, lost to follow-up; TFO, transferred out.

Outcomes

At the end of follow-up, 684 (2.7%) ALPH had died, 4026 (16.0%) had transferred care, 5608 (22.3%) had been LTFU, and 14 808 (58.9%) were still actively in care. Compared with those that entered care aged less than 10 years, a higher proportion of ALPH entering care aged 10–13 years died (1.4 vs. 5.1%) and were LTFU (16.2 vs. 33.4%), while a lower proportion were retained in care (66.4 vs. 45.4%) at the end of follow-up (Table 3). For ALPH who entered care aged less than 10 years, cumulative 5-year incidence estimates (95% CI) from the age of 10 years were 1.5% (1.3; 1.7%) for mortality; 17.8% (17.1; 18.4%) for LTFU; and 17.6% (17.0; 18.3%) for TFO. Corresponding 5-year cumulative probability for being retained was 63.1% (61.6; 64.6%). For ALPH who entered care aged 10–13 years, cumulative 5-year incidence estimates (95% CI) from first visit were 4.7% (4.2; 5.1%) for mortality; 28.0% (27.1; 28.9%) for LTFU; and 13.9% (13.2; 14.7%) for TFO. Corresponding 5-year cumulative probability of being retained was 53.4% (51.4; 55.5%).

In the adjusted regression model, characteristics at age 13 years associated with having a decreased risk of dying after the age of 13 years were: reaching age 13 years in more recent calendar years [adjusted hazard ratio (aHR) per increasing year 0.91; 95% CI 0.84; 0.99]; having a CD4+ cell count between 350 and 500 cells/μl (aHR 0.37; 95% CI 0.21; 0.63) or more than 500 cells/μl (aHR 0.14; 95% CI 0.09; 0.23) compared with a CD4+ cell count less than 350 cells/μl; and having a higher HAZ (aHR HAZ > −2: 0.57, 95% CI 0.36; 0.91 and HAZ −2 or less to >−3: 0.61, 95% CI 0.38; 0.98 vs. HAZ≥−3) (Table 4, Multivariate Model A). Increasing age of entry into HIV care, sex, ART duration, and ART regimen were not associated with mortality. Similar results were obtained in a sensitivity analysis restricted to ALPH who had been on ART for at least 1 year by age 13 years (Table 4, Multivariate Model B).

Table 4 - Cox-proportional hazards model for predictors of mortality following the 13th birthday among adolescents living with perinatally acquired HIV-initiated antiretroviral therapy by age 13 years and still in care after the age of 13 years (N = 13 470).
Univariate models Multivariate Model A Multivariate Model B
Characteristic HR (95% CI) aHRa (95% CI) aHRa (95% CI)
Age of entry into HIV care (years)
 <10 0.45 (0.35–0.59) 0.72 (0.36–1.43) 0.72 (0.34–1.52)
 10–13 Ref Ref Ref
Sex
 Female 0.99 (0.78–1.25) 1.34 (0.91–1.98) 1.19 (0.73–1.96)
 Male Ref Ref Ref
ART duration (years) at age 13 years (per 1-year increase) 0.80 (0.75–0.85) 0.92 (0.77–1.10) 0.85 (0.67–1.09)
ART regimen at age 13 years
 Efavirenz-based Ref Ref Ref
 Nevirapine-based 1.37 (1.03–1.83) 0.88 (0.50–1.55) 1.17 (0.54–2.51)
 Protease inhibitor-based 1.56 (0.97–2.52) 1.42 (0.74–2.73) 1.52 (0.72–3.17)
 Other 3.70 (1.17–11.71) 1.50 (0.34–6.67) 9.62 (1.17–78.92)
Calendar year of 13th birthday (per 1-unit increase) 0.84 (0.80–0.88) 0.91 (0.83–0.99) 0.97 (0.86 –1.09)
CD4+ cell count (cells/μl) at age 13 years
 <350 Ref Ref Ref
 350–500 0.42 (0.27–0.65) 0.37 (0.22–0.64) 0.37 (0.19–0.72)
 >500 0.14 (0.10–0.21) 0.15 (0.09–0.25) 0.12 (0.06–0.22)
HAZ at age 13 years
 >−2 0.43 (0.29–0.64) 0.57 (0.36–0.91) 0.46 (0.24–0.86)
 ≤−2 to >−3 0.49 (0.32–0.75) 0.61 (0.38–0.98) 0.71 (0.39–1.29)
 ≥−3 Ref Ref Ref
Multivariate Model A – includes adolescents living with perinatally acquired HIV initiated ART by age 13 years and still in care after the age of 13 years.Multivariate Model B – includes only adolescents living with perinatally acquired HIV on ART for at least 1 year by age 13 years and still in care after the age of 13 years.CI, confidence interval; aHR, adjusted hazard ratio; ART, antiretroviral therapy; HAZ, height-for-age z-score; HR, hazard ratio.
aModels also adjusted for country.

Discussion

The current analysis presents the HIV treatment outcomes of over 25 000 ALPH within Southern Africa, from as early as 2004 up to 2017. We found that increasing numbers of ALPH had suboptimal outcomes during adolescence, with these outcomes substantially worse among ALPH entering HIV care during adolescence vs. before age 10 years.

In this analysis, perinatal HIV acquisition was assumed in all adolescents who had entered care before the age of 13 years without documentation of nonperinatal HIV acquisition. We used this age threshold to capture those children with slow-progressing disease who survive into adolescence without treatment and present for care during early adolescence. We found no differences in the proportion of males vs. females (48.5 vs. 51.5%) in our overall sample after increasing this age threshold, supporting the notion of a nonsexual route of transmission among this group. Had those presenting between the ages of 10–13 years been infected through sexual contact, we would have expected the proportion of females to be substantially higher due to the well described increased risk among young females in our region [25]. In addition, those who entered care aged 10–13 years showed poor HAZ during adolescence, similar to levels seen in those who entered care before the age of 10 years. Reduced height is one of the hallmarks for longstanding HIV infection in childhood due to perinatal HIV acquisition [2,19,26,27].

ALPH in our analysis who entered care before the age of 10 years initiated ART at a median age of 7.2 years. Also, a large number were severely immunosuppressed at ART initiation. This highlights the substantial delay in ART initiation. This could be partially explained by the fact that these adolescents were children at a time when ART treatment services were limited and the criteria for ART initiation restrictive, with guidelines requiring prior immunologic and clinical deterioration before ART start [28–30]. In addition, socioeconomic factors could have also played a role. Although the numbers of ALPH are expected to decline in the future owing to decreasing numbers of children acquiring HIV perinatally, the next wave of ALPH may have initiated ART earlier. However, the benefit of universal ART will only be realized when all children with perinatally acquired HIV initiate ART in infancy.

Regardless of age of entry into care, our study noted a steady decline in the median CD4+ cell count with increasing age, even when restricted to those who had been on ART for at least 1 year by the respective birthday. This is worrying as older adolescents are maturing and expected to be gaining increasing independence in preparation for early adulthood. The absence of adherence data mean that we were unable to establish the extent to which poor adherence and treatment failure may have influenced this finding. In contrast, as the adolescents aged, we observed improvements in the median HAZ as shown by other research [31,32]. Though these scores improved over time, they generally remained below the −1 z-score across the adolescent ages, confirming earlier research that ALPH within resource-limited settings may not reach their full height potentials [33,34]. Compared with earlier calendar years, we found improvements in the CD4+ cell counts and HAZ in more recent calendar years, reflecting the positive effects that the changing HIV treatment guidelines, such as allowing ART to be initiated earlier, could be having [35].

We noted that overall, a substantially high proportion of ALPH were LTFU, with overall retention at the end of the study period being less than 60%. The high rate of LTFU is of concern given that it may imply an under-ascertainment of mortality in our analysis. These rates of LTFU also highlight the challenges of retaining ALPH in care as they age. One study in Zimbabwe which showed better retention rates among ALPH [36] had adolescent-friendly services including peer and nonpeer counseling within their public sector cohort. In our study, LTFU was also considerably greater among ALPH who presented to care aged 10–13 years compared with those who presented before the age of 10 years. This underscores the urgent need to recognize this population of late presenters or slow progressors and identify optimal interventions to keep then retained, particularly in the early years of engaging with care.

Cross-sectionally at the end of follow-up, mortality was substantially higher among ALPH who entered care aged 10–13 years compared with those who entered aged less than 10 years. This could be attributed to the fact that those who presented earlier needed to have reached adolescence to be included in this analysis, making this a group limited to those who survived childhood with HIV. Among the ALPH who died after the age of 13 years, median age at death was 15.5 years. Immunosuppression and impaired anthropometric measures were associated with mortality, as has been shown in other studies [36,37].

The study's limitations include that mode of HIV acquisition was assumed based on age of entry into HIV care and nonreporting of other modes of infection [2,38]. We had missing data on critical measures such as CD4+ cell count, HIV-RNA, WHO clinical staging, and anthropometric measures, particularly among older adolescents. Laboratory testing and anthropometric evaluations could have been obtained more frequently among those at higher risk of poor outcomes, which may have biased our results. In addition, we used routinely collected data in our analyses, which did not include key information on adherence, disclosure, and other social measures of well being that would be needed to more fully characterize reasons for positive or negative outcomes during adolescence. Assessment of viral load outcomes was limited to the cohorts offering routine (annual) viral load measurements. LTFU was assumed for all ALPH with an unknown outcome, which may have resulted in an underestimate of mortality. Our data were obtained from facilities providing data to the IeDEA-SA collaboration; hence, results may not be generalizable to all adolescents within the region. Lastly, reasons for TFO were not documented within the database.

Conclusion

The current analysis shows outcomes of ALPH in sites offering routine HIV care services across six Southern African countries, over extended periods of time. ALPH in our cohort had progressively worsening retention with increasing age. These outcomes were substantially worse among ALPH entering HIV care during adolescence vs. those entering care before age 10 years. Interventions tailored to improving treatment outcomes in this population need urgent attention to ensure optimal health outcomes for this population.

Acknowledgements

We wish to thank all the participants and their parents and/or caregivers and all the staff within the IeDEA-SA sites. We would like to also thank the IeDEA-SA Data Centers at the Universities of Cape Town, South Africa and Bern, Switzerland.

IeDEA-SA Steering Group: Frank Tanser, Africa Centre for Health and Population Studies, University of Kwazulu-Natal, Somkhele, South Africa; Christopher Hoffmann, Aurum Institute for Health Research, Johannesburg, South Africa; Michael Vinikoor, Centre for Infectious Disease Research in Zambia, Lusaka, Zambia; Robin Wood, Desmond Tutu HIV Centre (Gugulethu and Masiphumelele clinics), Cape Town, South Africa; Andrew Boulle, Khayelitsha ART Programme and Médecins Sans Frontières, Cape Town, South Africa; Geoffrey Fatti, Kheth’Impilo Programme, South Africa; Sam Phiri, Lighthouse Trust Clinic, Lilongwe, Malawi; Janet Giddy, McCord Hospital, Durban, South Africa; Cleophas Chimbetete, Newlands Clinic, Harare, Zimbabwe; Kennedy Malisita, Queen Elizabeth Hospital, Blantyre, Malawi; Brian Eley, Red Cross War Memorial Children's Hospital and Department of Paediatrics and Child Health, University of Cape Town, Cape Town, South Africa; Olatunbosun Faturiyele, SolidarMed SMART Programme, Lesotho; Michael Hobbins, SolidarMed SMART Programme, Pemba Region, Mozambique; Kamelia Kamenova, SolidarMed SMART Programme, Masvingo, Zimbabwe; Matthew Fox, Themba Lethu Clinic, Johannesburg, South Africa; Hans Prozesky, Tygerberg Academic Hospital, Stellenbosch, South Africa; Karl Technau, Empilweni Clinic, Rahima Moosa Mother and Child Hospital, Johannesburg, South Africa; Shobna Sawry, Harriet Shezi Children's Clinic, Chris Hani Baragwanath Hospital, Soweto, South Africa

Authors’ contributions: P.T., M.A.D., and A.H.S. conceptualized the concept and designed the study. K.B., G.F., C.B., C.C., H.R., S.P., S.S., B.E., M.A.H., and A.B. were the lead site investigators. K.T. reviewed and provided extensive input into the article. P.T. analyzed the data and wrote up the first draft of the article. All authors reviewed the script and approved the final article.

The current work was funded by the U.S. National Institutes of Health's National Institute of Allergy and Infectious Diseases, the Eunice Kennedy Shriver National Institute of Child Health and Human Development, the National Cancer Institute, the National Institute of Mental Health, the National Institute on Drug Abuse, the National Heart, Lung, and Blood Institute, the National Institute on Alcohol Abuse and Alcoholism, the National Institute of Diabetes and Digestive and Kidney Diseases, the Fogarty International Center, and the National Library of Medicine for IeDEA-SA (U01AI069924) and for GRADUATE (R21HD089859).

Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the article.

Conflicts of interest

Authors have no conflicts of interest to declare.

Meetings at which parts of the data were presented: The 11th International Workshop on HIV Pediatrics, Mexico City, Mexico, 19–20 July 2019. The 10th IAS Conference on HIV Science, Mexico City, Mexico, 21–24 July 2019.

References

1. Johnson LF, Davies MA, Moultrie H, Sherman G, Bland R M, Rehle T M, et al. The effect of early initiation of antiretroviral treatment in infants on pediatric AIDS mortality in South Africa: a model-based analysis. Pediatr Infect Dis J 2012; 31:474–480.
2. Ferrand RA, Munaiwa L, Matsekete J, Bandason T, Nathoo K, Ndhlovu C E, et al. Undiagnosed HIV infection among adolescents seeking primary healthcare in Zimbabwe. Clin Infect Dis 2010; 51:844–851.
3. The Joint United Nations Programme on HIV/AIDS. UNAIDS 2018 estimates. 2018; Geneva, Switzerland: UNAIDS, Available at: http://aidsinfo.unaids.org/. [Accessed on 14 August 2019].
4. Sohn AH, Hazra R. The changing epidemiology of the global paediatric HIV epidemic: keeping track of perinatally HIV-infected adolescents. J Int AIDS Soc 2013; 16:18555.
5. Hazra R, Siberry GK, Mofenson LM. Growing up with HIV: children, adolescents, and young adults with perinatally acquired HIV infection. Annu Rev Med 2010; 61:169–185.
6. Lowenthal ED, Bakeera-Kitaka S, Marukutira T, Chapman J, Goldrath K, Ferrand RA. Perinatally acquired HIV infection in adolescents from sub-Saharan Africa: a review of emerging challenges. Lancet Infect Dis 2014; 14:627–639.
7. Maskew MBJ, MacLeod W, Carmona S, Sherman G, Fox MP. The youth treatment bulge in South Africa: increasing numbers, inferior outcomes among adolescents on ART.International AIDS Conference; 2016; Durban, South Africa.
8. Enane LA, Davies MA, Leroy V, Edmonds A, Apondi E, Adedimeji A, et al. Traversing the cascade: urgent research priorities for implementing the ‘treat all’ strategy for children and adolescents living with HIV in sub-Saharan Africa. J Virus Erad 2018; 4: (Suppl 2): 40–46.
9. Evans D, Menezes C, Mahomed K, Macdonald P, Untiedt S, Levin L, et al. Treatment outcomes of HIV-infected adolescents attending public-sector HIV clinics across Gauteng and Mpumalanga, South Africa. AIDS Res Hum Retroviruses 2013; 29:892–900.
10. Auld A F, Agolory SG, Shiraishi RW, Wabwire-Mangen F, Kwesigabo G, Mulenga M, et al. Antiretroviral therapy enrollment characteristics and outcomes among HIV-infected adolescents and young adults compared with older adults – seven African countries, 2004–2013. MMWR Morb Mortal Wkly Rep 2014; 63:1097–1103.
11. Nachega JB, Hislop M, Nguyen H, Dowdy DW, Chaisson RE, Regensberg L, et al. Antiretroviral therapy adherence, virologic and immunologic outcomes in adolescents compared with adults in Southern Africa. J Acquir Immune Defic Syndr 2009; 51:65.
12. Stagi S, Galli L, Cecchi C, Chiappini E, Losi S, Gattinara C, et al. Final height in patients perinatally infected with the human immunodeficiency virus. Horm Res Paediatr 2010; 74:165–171.
13. Slogrove AL, Schomaker M, Davies MA, Williams P, Balkan S, Ben-Farhat J, et al. Collaborative Initiative for Paediatric HIV Education and Research (CIPHER) Global Cohort Collaboration. The epidemiology of adolescents living with perinatally acquired HIV: a cross-region global cohort analysis. PLoS Med 2018; 15:e1002514.
14. Ferrand R, Lowe S, Whande B, Munaiwa L, Langhaug L, Cowan F, et al. Survey of children accessing HIV services in a high prevalence setting: time for adolescents to count?. Bull World Health Organ 2010; 88:428–434.
15. Walker AS, Mulenga V, Sinyinza F, Lishimpi K, Nunn A, Chintu C, Gibb DM. CHAP Trial Team. Determinants of survival without antiretroviral therapy after infancy in HIV-1-infected Zambian children in the CHAP trial. J Acquir Immune Defic Syndr 2006; 42:637–645.
16. Ferrand RA, Corbett EL, Wood R, Hargrove J, Ndhlovu CE, Cowan FM, et al. AIDS among older children and adolescents in Southern Africa: projecting the time course and magnitude of the epidemic. AIDS 2009; 23:2039–2046.
17. Marston M, Becquet R, Zaba B, Moulton LH, Gray G, Coovadia H, et al. Net survival of perinatally and postnatally HIV-infected children: a pooled analysis of individual data from sub-Saharan Africa. Int J Epidemiol 2011; 40:385–396.
18. Stover J, Walker N, Grassly NC, Marston M. Projecting the demographic impact of AIDS and the number of people in need of treatment: updates to the spectrum projection package. Sex Transm Infect 2006; 82: (Suppl 3): iii45–iii50.
19. Ferrand RA, Bandason T, Musvaire P, Larke N, Nathoo K, Mujuru H, et al. Causes of acute hospitalization in adolescence: burden and spectrum of HIV-related morbidity in a country with an early-onset and severe HIV epidemic: a prospective survey. PLoS Med 2010; 7:e1000178.
20. Marston M, Zaba B, Salomon JA, Brahmbhatt H, Bagenda D. Estimating the net effect of HIV on child mortality in African populations affected by generalized HIV epidemics. J Acquir Immune Defic Syndr 2005; 38:219–227.
21. Egger M, Ekouevi DK, Williams C, Lyamuya RE, Mukumbi H, Braitstein P, et al. Cohort profile: the international epidemiological databases to evaluate AIDS (IeDEA) in sub-Saharan Africa. Int J Epidemiol 2012; 41:1256–1264.
22. World Health Organization. WHO case definitions of HIV for surveillance and revised clinical staging and immunological classification of HIV-related disease in adults and children. 2019; Geneva, Switzerland: World Health Organization, Available at: http://www.who.int/hiv/pub/guidelines/HIVstaging150307.pdf?ua=1. [Accessed on 14 April 2019].
23. World Health Organization. WHO child growth standards. STATA igrowup package. 2019; Geneva, Switzerland: World Health Organization, Available at: http://www.who.int/childgrowth/software/en/. [Accessed on 14 April 2019].
24. World Health Organization. WHO child growth standards. STATA WHO 2007 package. 2007; Geneva, Switzerland: World Health Organization, Available at: http://www.who.int/entity/growthref/tools/who2007_stata.zip [Accessed 14 August 2019].
25. Gregson S, Nyamukapa CA, Garnett GP, Mason PR, Zhuwau T, Caraël M, et al. Sexual mixing patterns and sex-differentials in teenage exposure to HIV infection in rural Zimbabwe. Lancet 2002; 359:1896–1903.
26. Buchacz K, Rogol A D, Lindsey F C, Wilson CM, Hughes MD, Seage GR, et al. Delayed onset of pubertal development in children and adolescents with perinatally acquired HIV infection. J Acquir Immune Defic Syndr 2003; 33:56–65.
27. Majaliwa ES, Mohn A, Chiarelli F. Growth and puberty in children with HIV infection. J Endocrinol Invest 2009; 32:85–90.
28. World Health Organization. HIV/AIDS Programme. Strengthening health services to fight HIV/AIDS. Antiretroviral therapy for HIV infection in adults and adolescents: recommendations for a public health approach – 2006 revision. Geneva, Switzerland: WHO; 2006.
29. World Health Organization. Antiretroviral therapy of HIV infection in infants and children: recommendations for a public health approach. Geneva, Switzerland: WHO; 2006.
30. World Health Organization. Antiretroviral therapy for HIV infection in adults and adolescents: recommendations for a public health approach – 2010 revision. Geneva, Switzerland: WHO; 2010.
31. Sutcliffe C G, van Dijk F H, Munsanje B, Hamangaba F, Sinywimaanzi P, Thuma P E, et al. Weight and height z-scores improve after initiating ART among HIV-infected children in rural Zambia: a cohort study. BMC Infect Dis 2011; 11:54.
32. Weigel R, Phiri S, Chiputula F, Gumulira J, Brinkhof M, Gsponer T, et al. Growth response to antiretroviral treatment in HIV-infected children: a cohort study from Lilongwe, Malawi. Trop Med Int Health 2010; 15:934–944.
33. Gsponer T, Weigel R, Davies M A, Bolton C, Moultrie H, Vaz P, et al. Variability of growth in children starting antiretroviral treatment in Southern Africa. Pediatrics 2012; 130:e966–e977.
34. Bakeera-Kitaka S, McKellar M, Snider C, Kekitiinwa A, Piloya T, Musoke P, et al. Antiretroviral therapy for HIV-1 infected adolescents in Uganda: assessing the impact on growth and sexual maturation. J Pediatr Infect Dis 2007; 3:97–104.
35. World Health Organization. Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection. Recommendations for a public health approach. 2nd ed.Geneva, Switzerland: WHO; 2016.
36. Shroufi A, Gunguwo H, Dixon M, Nyathi M, Ndebele W, Saint-Sauveur F F, et al. HIV-infected adolescents in Southern Africa can achieve good treatment outcomes: results from a retrospective cohort study. AIDS 2013; 27:1971–1978.
37. Chokephaibulkit K, Kariminia A, Oberdorfer P, Nallusamy R, Bunupuradah T, Hansudewechakul R, et al. Characterizing HIV manifestations and treatment outcomes of perinatally infected adolescents in Asia. Pediatr Infect Dis J 2014; 33:291.
38. Pegurri E, Konings E, Crandall B, Haile-Selassie H, Matinhure N, Naamara W, Assefa Y. The missed HIV-positive children of Ethiopia. PLoS One 2015; 10:e0124041.
Keywords:

adolescents; HIV; outcomes; perinatal HIV; Southern Africa

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.