Introduction
Globally each year there are ±1.3 million pregnant people with HIV (PPHIV), the majority (±1.1 million) receiving antiretroviral therapy (ART), providing the profound dual benefit of preserving maternal health and preventing vertical HIV transmission [1]. Understanding the implications of life-long ART and the safety of antiretrovirals during pregnancy is now imperative, given the cumulatively expanding population of PPHIV on ART [2–4].
Hypertensive disorders of pregnancy (HDP) complicate 8–10% of pregnancies and are a leading cause of maternal mortality as well as long-term morbidity, doubling the risk for cardiovascular disease later in life [5–7]. HDP are key drivers of adverse birth outcomes due to associated placental compromise resulting in sub-optimal fetal growth and even fetal demise. Additionally, HDP may result in necessitated preterm delivery when the maternal or fetal well being becomes compromised due to HDP-associated organ dysfunction [8]. Resulting fetal growth restriction and preterm birth have profound long-term implications including sub-optimal learning trajectories and susceptibility to chronic cardiovascular disease during adulthood [9]. Furthermore, preterm birth is the single largest contributor to child mortality globally [10]. There has been a long-standing interest in understanding how HDP and HIV jointly impact the health of PPHIV and their children, and the potential role of ART in the risk for HDP has become increasingly relevant.
HDP are a multifactorial, partly immune-mediated, spectrum of conditions classified by the International Society for the Study of Hypertension in Pregnancy (ISSHP) into two groups: hypertension known before or present in the first 20 weeks of pregnancy (includes essential and secondary chronic hypertension, white-coat hypertension and masked hypertension) and hypertension arising de novo 20 or more weeks’ gestation (including transient gestational hypertension, gestational hypertension, preeclampsia de novo or superimposed on chronic hypertension) [11]. Both HIV and de novo HDP (dnHDP) result in chronic inflammation, increase susceptibility to long-term maternal cardiovascular disease, and adverse birth outcomes [7,12,13]. A systematic review of studies preceding ART availability found no association between HIV and dnHDP, hypothesized to be because of HIV-related immune suppression inhibiting the immune response associated with dnHDP onset [14]. As ART became available, several small studies reported that PPHIV on ART had a higher risk for dnHDP than those ART-naive, hypothesized to be due to immune reconstitution on ART or ART-related side-effects [15,16]. A subsequent systematic review found that, although all 28 included studies observed increased dnHDP in PPHIV on ART compared to those not on ART, the studies had multiple biases, poor study designs, small sample sizes and heterogeneity in definitions [17]. Maternal ART is essential for both the mother and fetus and cannot be avoided during pregnancy. A much clearer understanding of the current relationship between HIV, ART and dnHDP is urgently required.
With universal access to ART, we hypothesized that ART would protect against the dysfunctional immune response that precipitates dnHDP through reduced immune-perturbation in people starting ART earlier prior to onset of severe immune suppression, but that in those still initiating ART with severe immune suppression dnHDP would be exacerbated [18,19]. Although we could not directly test this hypothesis, we aimed to evaluate whether there was an association between HIV, ART and the prevalence of dnHDP during the universal ART era in the Western Cape province of South Africa [11]. Our primary objective was to compare the prevalence of dnHDP in all pregnant people delivering in the Western Cape during 2018 and 2019 by categories of HIV and ART initiation status (no HIV, HIV – no ART, HIV – preconception ART, HIV – pregnancy-initiated ART). To specifically evaluate associations between HIV/ART and dnHDP in the universal ART era, we subsequently restricted the comparison to include only pregnant people diagnosed with HIV in 2016 or later, when universal ART became standard of care, compared to all pregnant people without HIV. Additionally, we aimed to evaluate HIV-specific factors associated with dnHDP.
Methods
This was a retrospective population-based cohort study of pregnant people delivering at 54 public sector health facilities in the Western Cape from 1 January 2018 to 31 December 2019. The Western Cape Government provides healthcare services for ±75% of the province's population. These include free maternity (antenatal, delivery and postnatal) care to ±110 000 pregnant people annually and ART for all people living with HIV, including pregnant people in whom the antenatal HIV prevalence is ±19% [20]. There is high uptake of routine HIV testing in people without known HIV at first antenatal visit, during the third trimester, and at delivery. The Western Cape introduced universal ART in 2016 with the recommended first-line regimen being once-daily fixed-dose tenofovir/emtricitabine/efavirenz (TDF/FTC/EFV) and same-day ART initiation for newly diagnosed PPHIV [20]. During the study period the standard second-line regimen was a protease inhibitor (PI) lopinavir/ritonavir-based regimen, and no transition had yet been recommended to dolutegravir-based first-line regimen. Routine antenatal care at primary healthcare clinics includes HDP screening through measurement of blood pressure and dipstick urine protein analysis at every antenatal visit, the results of which are documented on standardized individually held maternity case records. Nurse midwives can initiate methyldopa for uncomplicated gestational hypertension at primary healthcare level, with province-wide standardized protocols for referral to higher levels of care according to HDP severity [21].
The Western Cape Provincial Health Data Centre (WCPHDC) integrates multiple sources of routinely collected electronic information according to individual unique identifiers issued to all those accessing provincial healthcare services. The WCPHDC structure, data sources and governance have been described in detail [22]. Specific to this analysis, all people identified in the WCPHDC by validated algorithms as having a pregnancy delivery date of 01/01/2018 to 31/12/2019 were included [23]. Basic parameters extracted were delivery maternal age, multiparity, multigestation pregnancy, delivery date, delivery method, pregnancy outcome (termination, miscarriage, live birth, stillbirth, early neonatal death, late neonatal death or unknown) and birth weight. For those with more than one pregnancy during the study period, only the first pregnancy was retained. Pregnancies were excluded if the outcome was ‘termination’ or ‘miscarriage’ with a birth weight of less than 500 g, or the maternal age was unknown. For those included, relevant pregnancy exposures and outcomes were extracted from additional electronic sources. These included the electronic HIV register to identify PPHIV, date of HIV diagnosis, date of ART-start and specific antiretrovirals dispensed during pregnancy; laboratory database for CD4+ cell count; ICD-10 codes indicating an HDP (ICD-10 O10-O16) from clinic and hospital administration systems; pharmacy Anatomical Therapeutic Chemical (ATC) classification codes and dispensing dates for antihypertensive drugs dispensed in pregnancy.
The primary exposure variable of combined HIV/ART status during pregnancy was defined as no HIV (no evidence of HIV during pregnancy up until the time of delivery), HIV – no ART (evidence of an HIV diagnosis but no evidence of ART received or only started at delivery or later), HIV – preconception ART (evidence of ART started ≥294 days (≥42 weeks) prior to pregnancy outcome date; corresponds to DECIPHER high-level certainty of preconception ART [24]), or HIV – pregnancy ART (evidence of ART started >0 and <294 days prior to current pregnancy outcome date). Additional exposure variables included maternal age, multiparity, multigestation pregnancy. In PPHIV these also included: HIV diagnosis period (<2013, 2013–<2016, ≥2016); ART initiation period (<2013, 2013–<2016, ≥2016); pregnancy CD4+ cell count (earliest CD4+ cell count within 294 days predelivery and 60 days postdelivery), ART regimen dichotomized as first-line TDF/FTC/EFV throughout (yes/no), any second- (or higher) line (yes/no), any protease inhibitor (PI)-based regimen (yes/no). People with evidence of preexisting hypertension were identified by ICD-10 codes for chronic hypertension (O10–O11, O16) or antihypertensives prescribed at least 140 days before delivery. The primary outcome of dnHDP was pragmatically defined as a binary variable (yes/no) according to having an appropriate ICD-10 code (O12–O15.9), including evidence of preexisting hypertension for ICD-10 codes indicating preeclampsia/eclampsia (O14–O15.9) (according to ISSHP definition of dnHDP); or evidence of first antihypertensive drug prescription less than 140 days before delivery without ICD-10 code of preexisting hypertension.
There was no a priori sample size calculation as the analytic intent was to evaluate the impact of HIV/ART on dnHDP prevalence at a population-level within the province. Baseline maternal, pregnancy and HIV characteristics were described using frequencies with percentage for categorical variables or median with interquartile range (IQR) for continuous variables and compared according to HIV/ART status. Frequency of unknown or missing variables was described but no further methods employed for missing data. For the primary objective comparing the prevalence of dnHDP in all pregnant people delivering in the Western Cape during 2018 and 2019 by HIV/ART status, unadjusted and adjusted prevalence ratios (PRs) with 95% CI for dnHDP were calculated using Poisson regression with robust variance [25,26]. Maternal age, nulliparity, multigestation pregnancy and preexisting hypertension were considered a priori as factors independently associated with dnHDP and thus potential confounders [27]. Interactions between maternal age and HIV/ART status were evaluated. Due to the possibility of an unclear temporal relationship between timing of ART started during pregnancy and onset of dnHDP, sensitivity analyses were conducted to exclude people with ART started less than 140 days prior to delivery, in whom dnHDP onset could potentially precede ART initiation. Subsequently the analysis was further restricted to people with HIV diagnosed in 2016 or later to specifically compare people with HIV diagnosed during the universal ART era to people without HIV. Finally, the associations between ART-specific characteristics and dnHDP prevalence in models including only PPHIV on ART at least 140 days prior to delivery were evaluated with the primary exposure variable of interest classified as ART initiated during pregnancy, ART initiated preconception with a total duration on ART of less than 100 weeks and ART initiated preconception with a total duration on ART of at least 100 weeks. Sensitivity analyses were conducted to understand potential misclassification of preexisting hypertension as dnHDP in the absence of gestational age information. Analyses were conducted with Stata 17.1 (StataCorp, College Station, Texas, USA).
This study was approved by the Stellenbosch University Health Research Ethics Committee and the Western Cape Government Provincial Health Research Committee. This study included routinely collected data that were anonymized and de-identified before secure transfer from the WCPHDC to the principal investigator. As such a waiver of informed consent was granted.
Results
During the study period, 210 051 people with pregnancy outcomes were recorded in the WCPHDC. After excluding ineligible pregnancies and those with missing critical data (Fig. 1), 180 553 pregnant people were included for analysis. According to maternal HIV/ART status, 81.2% (N = 146 575) had no evidence of HIV, 2.1% (N = 3827) had evidence of HIV but no ART, 11.2% (N = 20 283) had HIV with preconception ART and 5.5% (N = 9868) had HIV with pregnancy-initiated ART (Table 1). People on preconception ART were generally older [median (IQR) age 31.0 (27.0; 35.0) years] and most often multiparous (60.5%) compared to people without HIV [age: 26.2 (22.0; 31.2) years; multiparity: 41.7%], people with HIV on no ART [age: 29.2 (24.4; 33.9) years; multiparity 44.6%] and people on pregnancy-initiated ART [age 27.4 (23.7; 31.8) years; multiparity 29.9%]. Overall, 7.6% (N = 13 677) had dnHDP, highest in PPHIV on no ART (9.8%) and lowest for preconception ART (6.9%). Nineteen percent of people with dnHDP (2595/13 677) also had preexisting hypertension and preexisting hypertension alone occurred in 2.8% (N = 4960) of all people. Of the 13 677 people with dnHDP, 76.1% (N = 10 406) were classified as such according to ICD-10 codes and 23.9% (N = 3271) from pharmacy dispensing evidence only.
;)
Fig. 1: Flow diagram illustrating inclusion of pregnant people in Western Cape province-wide cohort 2018–2019.
Table 1 -
Characteristics of all pregnant people with and without
HIV delivering in 2018 and 2019 in the Western Cape Province public-sector.
|
Total |
No HIV |
HIV
No ART |
HIV
Preconception ART |
HIV
Pregnancy-initiated ART |
| Total N (row %) |
180 553 |
146 575 (81.2) |
3827 (2.1) |
20 283 (11.2) |
9868 (5.5) |
| Age at delivery |
| Median (IQR) years |
26.9 (22.6; 31.9) |
26.2 (22.0; 31.2) |
29.2 (24.4; 33.9) |
31.0 (27.0; 35.0) |
27.4 (23.7; 31.8) |
| <25 years – N (%) |
70 380 (39.0) |
63 029 (43.0) |
1069 (27.9) |
2967 (14.6) |
3315 (33.6) |
| 25–<30 years – N (%) |
49 525 (27.4) |
39 554 (27.0) |
1015 (26.5) |
5803 (28.6) |
3153 (32.0) |
| 30–<35 years – N (%) |
36 897 (20.4) |
27 240 (18.6) |
980 (25.6) |
6483 (32.0) |
2194 (22.2) |
| ≥ 35 years – N (%) |
23 751 (13.2) |
1675 (11.4) |
763 (19.9) |
5030 (24.8) |
1206 (12.2) |
| Multiparity – N (%) |
78 024 (43.2) |
61 105 (41.7) |
1707 (44.6) |
12 265 (60.5) |
2947 (29.9) |
| Multigestation – N (%) |
2158 (1.2) |
1666 (1.1) |
59 (1.5) |
288 (1.4) |
145 (1.5) |
| Caesarean section – N (%) |
36 660 (20.3) |
28 945 (19.8) |
799 (20.9) |
4561 (22.5) |
2355 (23.9) |
| Hypertensive disorders of pregnancy |
| None – N (%) |
161 916 (89.7) |
131 426 (89.7) |
3343 (87.4) |
18 167 (89.6) |
8890 (91.0) |
| dnHDP – N (%) |
13 677 (7.6) |
11 210 (7.7) |
375 (9.8) |
1400 (6.9) |
692 (7.0) |
| Preexisting hypertension only – N (%) |
4969 (2.8) |
3939 (2.7) |
109 (2.9) |
716 (3.5) |
196 (2.0) |
| Type of de novo HDP by ICD-10 codesa
|
| Gestational hypertension |
2007 (19.3) |
1707 (19.9) |
58 (19.8) |
148 (14.5) |
94 (17.8) |
| Preeclampsia |
4958 (47.6) |
4087 (47.7) |
147 (50.2) |
457 (44.8) |
267 (50.7) |
| HELLP syndrome |
196 (1.9) |
169 (2.0) |
5 (1.7) |
13 (1.3) |
9 (1.7) |
| Eclampsia |
327 (3.1) |
275 (3.2) |
14 (4.8) |
28 (2.7) |
10 (1.9) |
| HDP unspecified |
2918 (28.0) |
2327 (27.2) |
69 (23.5) |
375 (36.7) |
147 (27.9) |
ART, antiretroviral therapy; dnHDP, de novo hypertensive disorders of pregnancy; HELLP, haemolysis, elevated liver enzymes, low platelets; IQR, interquartile range.
a% in those with an ICD-10 code of dnHDP, N = 10 406.
Among 33 978 PPHIV, 42.6% (N = 14 462) had HIV diagnosed in 2016 or later, 60.0% (N = 20 283) were on ART preconception, with the majority on ART for at least 2.5 years (Table 2). Among 30 151 people on ART, median [IQR] duration was 213 [131; 321] weeks in the preconception and 21 [14; 27] weeks in the pregnancy-initiated ART groups. Similarly, 79% (N = 16 030) of people with preconception ART were on first-line TDF/FTC/EFV throughout pregnancy versus 92.1% (N = 9090) in the pregnancy-initiated ART group. In the preconception ART group 6.6% (N = 1333) received a second-(or higher) line regimen versus 1.3% (N = 127) in the pregnancy-initiated ART group. In the 1460 people on second-(or higher) line ART, 96.9% (N = 1415) were on a PI-based regimen (lopinavir-based N = 1302; atazanavir-based N = 111; darunavir-based N = 2). The non-PI-based regimens included dolutegravir-based (N = 30), raltegravir-based (N = 1), etravirine-based (N = 13) and enfuvirtide-based (N = 1) regimens. In the 62.5% (N = 21 520) of PPHIV with a recorded pregnancy CD4+ cell count, the median [IQR] was 428 (276; 601) with no substantial group differences (Table 2). Table 1, Supplemental Digital Content, https://links.lww.com/QAD/C810 summarizes characteristics of the PPHIV diagnosed in 2016 or later only (N = 14 462). These were generally younger, with slightly lower prevalence of dnHDP and shorter durations on ART compared to all PPHIV diagnosed in any period.
Table 2 -
HIV-related characteristics of all pregnant people with
HIV delivering in 2018 and 2019 in the Western Cape Province public-sector.
|
Total |
HIV
No ART |
HIV
Preconception ART |
HIV
Pregnancy-initiated ART |
|
N
|
33 978 |
3827 |
20 283 |
9868 |
| HIV diagnosis period |
| Pre2013 |
11 620 (34.2) |
806 (21.1) |
9606 (47.4) |
1208 (12.2) |
| 2013–2015 |
7896 (23.2) |
458 (12.0) |
6410 (31.6) |
1028 (10.4) |
| 2016–2019 |
14 462 (42.6) |
2563 (67.0) |
4267 (21.0) |
7632 (77.3) |
| ART initiation perioda
|
| Pre2013 |
5292 (17.6) |
NA |
5292 (26.1) |
0 |
| 2013–2015 |
8323 (27.6) |
NA |
8323 (41.0) |
0 |
| 2016–2019 |
16 536 (54.8) |
NA |
6668 (32.9) |
9868 (100) |
| Duration on ARTa
b
|
| < 20 weeks |
4425 (14.7) |
NA |
0 |
4425 (44.8) |
| 20–<40 weeks |
5362 (17.8) |
NA |
0 |
5362 (54.3) |
| 40–<100 weeks |
3063 (10.2) |
NA |
2982 (14.7) |
31 (0.8) |
| ≥100 weeks |
17 301 (57.4) |
NA |
17 301 (85.3) |
0 |
| CD4+ cell count – median (IQR) |
428 (276; 601) |
439 (294; 606) |
456 (291; 625) |
377 (251; 532) |
| Pregnancy CD4+ category (N = 21 520) |
| 0–<349 |
8170 (24.0) |
529 (13.8) |
3832 (18.9) |
3809 (38.6) |
| 350–499 |
5364 (15.8) |
389 (10.2) |
2732 (13.5) |
2243 (22.7) |
| ≥ 500 |
7986 (23.5) |
618 (16.2) |
4901 (24.2) |
2467 (25.0) |
| Unknown CD4+ cell count |
12 458 (36.7) |
2291 (59.9) |
8818 (43.4) |
1349 (13.7) |
| First-line TDF-FTC-EFV throughout |
25 120 (83.3) |
NA |
16 030 (79.0) |
9090 (92.1) |
| Any second (or higher) line |
1460 (4.8) |
NA |
1333 (6.6) |
127 (1.3) |
ART, antiretroviral therapy; IQR, interquartile range; NA, not applicable.
a% in people on ART, N = 30 151.
b10 unknown ART duration in preconception ART group.
By unadjusted robust Poisson regression compared to people without HIV, PPHIV on preconception (PR 0.90; 95% CI 0.84–0.95) or pregnancy-initiated ART (PR 0.92; 95% CI 0.85–0.99) had lower prevalence of dnHDP. In contrast, PPHIV on no ART (PR 1.28; 95% CI 1.16–1.41) higher prevalence of dnHDP (Table 3). After adjusting for baseline factors of maternal age, multiparity, multigestation pregnancy and preexisting hypertension, the prevalence of dnHDP remained lower in PPHIV on preconception ART and pregnancy-initiated ART than people without HIV when including all pregnant people in the study period (Table 3, Model 1), when excluding people with pregnancy-initiated ART started less than 140 days prior to delivery (Table 3, Model 2), and when excluding people with HIV diagnosed before 2016 (Table 3, Model 3). Older maternal age, multigestation pregnancy and preexisting hypertension remained consistently associated with increased prevalence of dnHDP in all models. There was no interaction of maternal age with HIV/ART status, and preexisting hypertension remained the factor most strongly associated with dnHDP irrespective of HIV/ART status.
Table 3 -
Prevalence ratios for the outcome of de novo hypertensive disorders of
pregnancy (dnHDP) in people delivering in the Western Cape Province of
South Africa, 1 January 2018–31 December 2019.
|
Unadjusted PR (95% CI) |
Adjusted PR (95% CI)
Model 1 [N = 180 553] |
Adjusted PR (95% CI)
Model 2 [N = 176 128] |
Adjusted PR (95% CI)
Model 3 [N = 157 507] |
| HIV and ART classification |
| No HIV |
Reference |
Reference |
Reference |
Reference |
| HIV ART-none |
1.28 (1.16; 1.41) |
1.16 (1.06; 1.28) |
1.17 (1.06; 1.29) |
1.29 (1.15; 1.44) |
| HIV ART-preconception |
0.90 (0.84; 0.95) |
0.79 (0.75; 0.83) |
0.78 (0.74; 0.83) |
0.77 (0.69; 0.87) |
| HIV ART-pregnancy |
0.92 (0.85; 0.99) |
0.93 (0.87; 0.99) |
0.83 (0.75; 0.92) |
0.84 (0.75; 0.95) |
| Maternal age – in years |
1.05 (1.05; 1.05) |
– |
– |
– |
| Maternal age categories |
| <25 years |
Reference |
Reference |
Reference |
Reference |
| 25–<30 years |
1.19 (1.13; 1.24) |
1.16 (1.11; 1.22) |
1.16 (1.11; 1.21) |
1.17 (1.12; 1.23) |
| 30–<35 years |
1.58 (1.51; 1.65) |
1.46 (1.39; 1.53) |
1.46 (1.40; 1.53) |
1.47 (1.40; 1.54) |
| ≥35 years |
2.24 (2.14; 2.34) |
1.77 (1.69; 1.85) |
1.82 (1.74; 1.92) |
1.83 (1.74; 1.93) |
| Multiparous (reference: primiparous) |
1.13 (1.10; 1.70) |
0.93 (0.90; 0.96) |
0.92 (0.89; 0.95) |
0.92 (0.89; 0.95) |
| Multigestation pregnancy (reference: singleton) |
2.09 (1.90; 2.31) |
1.69 (1.55; 1.84) |
1.92 (1.74; 2.13) |
1.96 (1.76; 2.19) |
| Preexisting hypertension (reference: no preexisting hypertension) |
5.36 (5.17; 5.56) |
4.62 (4.45; 4.80) |
4.67 (4.50; 4.86) |
4.70 (4.50; 4.89) |
Model 1 includes all pregnant people 2018–2019; Model 2 excludes people with HIV if ART started less than 140 days prior to delivery; Model 3 further excludes people with HIV if diagnosed before 2016.ART, antiretroviral therapy; CI, confidence interval; PR, prevalence ratio.
To evaluate ART-specific associations with the prevalence of dnHDP, models were restricted to PPHIV on ART at least 140 days prior to delivery (Table 4). In unadjusted models, prevalence of dnHDP with pregnancy-initiated ART was similar to preconception ART of less than 100 weeks duration (PR 1.06; 95% CI 0.90–1.25) and preconception ART at least 100 weeks (PR 1.07; 95% CI 0.96; 1.21), prevalence was higher with any second (or higher) line regimen compared to first line (PR 1.58; 95% CI 1.35; 1.85) and any PI-based regimen versus non-PI regimens (PR 1.58; 95% CI 1.34; 1.85). The prevalence was lower in the more recent ART-initiation periods of 2013–2015 (PR 0.80; 95% CI 0.71–0.90) and 2016–2019 (PR 0.76; 0.68–0.85) compared to pre2013. Once adjusted for maternal baseline factors of age, multiparity, multigestation pregnancy and preexisting hypertension, preconception ART with duration at least 100 weeks was protective against dnHDP versus pregnancy-initiated ART of at least 20 weeks (aPR 0.89; 95% CI 0.79–0.99). However, more recent ART initiation periods were no longer protective. When included in the same model, preconception ART of at least 100 weeks (aPR 0.88; 95% CI 0.78; 0.98) and any second (or higher) line regimen (aPR 1.49; 95% CI 1.28; 1.75) remained associated with increased prevalence of dnHDP in the full cohort of all people on ART at least 140 days during pregnancy but did not remain associated when restricted to people diagnosed with HIV in 2016 or later (Table 4, Models 1 and 2). The effect of maternal age, multiparity, multigestation pregnancy and preexisting hypertension remained consistently robust and similar in the ART-only models (Table 4) compared to the models of all people with and without HIV (Table 3), and these specific PRs have not been shown again.
Table 4 -
Prevalence ratios for the outcome of de novo hypertensive disorders of
pregnancy (dnHDP) restricted to people with
HIV on ART at least 140 days during
pregnancy (
N = 25 726).
|
Unadjusted PR (95% CI)
N = 25 726 |
Adjusted PRa (95% CI)
N = 25 716 |
Adjusted PR (95% CI)
Model 1 N = 23 662 |
Adjusted PR (95% CI)
Model 2 N = 7763 |
| ART initiation and duration classification |
| ART-pregnancy |
Reference |
Reference |
Reference |
Reference |
| ART-preconception <100 weeks |
1.06 (0.90; 1.25) |
1.01 (0.86; 1.20) |
1.03 (0.87; 1.22) |
1.00 (0.81; 1.89) |
| ART-preconception ≥100 weeks |
1.07 (0.96; 1.21) |
0.89 (0.79; 0.99) |
0.88 (0.78; 0.98) |
0.83 (0.66; 1.03) |
| Pregnancy ART regimens |
| Any 2nd (or higher) line (reference: first line) |
1.58 (1.35; 1.85) |
1.46 (1.25; 1.71) |
1.49 (1.28; 1.75) |
1.11 (0.65; 1.89) |
| Any PI-based regimen (reference: non-PI regimens) |
1.58 (1.34; 1.85) |
1.46 (1.25; 1.70) |
– |
– |
| ART initiation period |
| Pre-2013 |
Reference |
Reference |
– |
– |
| 2013–2015 |
0.80 (0.71; 0.90) |
0.92 (0.81; 1.04) |
– |
– |
| 2016–2019 |
0.76 (0.68; 0.85) |
0.96 (0.85; 1.09) |
– |
– |
ART, antiretroviral therapy; CI, confidence interval; PI, protease inhibitor; PR, prevalence ratio.
aEach variable in row adjusted individually for maternal baseline factors (age as a continuous variable in years, multiparity, multigestation pregnancy and preexisting hypertension as binary variables) but not adjusted for other variables in the table.Model 1: includes all pregnant people with HIV on ART at least 140 days during pregnancy delivering in 2018 and 2019, adjusted for maternal baseline factors (age, multiparity, multigestation pregnancy and preexisting hypertension) and second (or higher) line regimen.Models 2: as for model 1 with additional restriction to people diagnosed with HIV in 2016 or later.
Tables 2 and 3, Supplemental Digital Content, https://links.lww.com/QAD/C810 detail sensitivity analyses considering possible misclassification of preexisting hypertension as dnHDP when classified according to first antihypertensive prescription less than 140 days prior to delivery in the absence of gestational age information. There was no meaningful impact on the association between HIV/ART status (Table 2, Supplemental Digital Content, https://links.lww.com/QAD/C810) or on the ART-specific models (Table 3, Supplemental Digital Content, https://links.lww.com/QAD/C810) under scenarios of extreme or moderate misclassification.
Discussion
In this large province-wide analysis of 180 553 pregnant people in 2018 and 2019, compared to pregnant people without HIV, for PPHIV no ART was associated with a higher prevalence and ART started prior to or during pregnancy a lower prevalence of dnHDP. Preexisting hypertension, older maternal age and multigestation pregnancy were strongly associated with dnHDP regardless of HIV status. In PPHIV, being on ART for more than 100 weeks prior to delivery was reassuringly protective against dnHDP. And although the number of people on second or higher-line regimens was relatively small, being on a second or higher-line regimen compared to a first-line regimen was associated with a relative 49% increased prevalence of dnHDP.
Our findings stand in contrast to most previous studies. Where much of the work to date has observed a reduced risk for dnHDP in people with HIV not on ART and an increased risk for dnHDP in people on ART, our findings from the universal ART era found the opposite. The only previous large study leveraging a database of routinely collected data in Zambia from 2006 to 2012, prior to universal ART, observed a lower odds of what was termed ‘pregnancy associated hypertension’ in people with HIV not on ART and a higher odds in people on ART compared to people without HIV [28]. However, the prevalence of ‘pregnancy associated hypertension’ overall was only 2.1%, a likely under-ascertainment of this outcome. This analysis of the Western Cape province-wide maternity cohort is the largest study of dnHDP in PPHIV during the universal ART era and the findings from this routine data source are robust. dnHDP prevalence is in-keeping with a previous smaller detailed cohort in the province, HIV and ART prevalence are in-keeping with modeled estimates and antenatal serosurveys, and the relationships with known dnHDP risk factors such as preexisting hypertension, maternal age, multiparity and multigestation pregnancy were consistent [29,30]. This establishes the Western Cape province-wide maternity cohort, made possible by the WCPHDC, as a much-needed efficient approach to continuously evaluate the safety of ART regimens and other HIV treatment modalities during pregnancy [4].
In this cohort, longer ART duration prior to pregnancy, even considering that ART duration is associated with advancing age and greater likelihood of being on second/higher-line regimens, offered increasing protection against dnHDP. We hypothesize that this is due to the immune stabilizing effects of long-term ART as opposed to the rapid immune system changes following ART initiation soon before or during pregnancy [18,31]. The protective effect of ART against dnHDP is a strong motivator to identify and treat HIV before pregnancy through universal ‘test and treat’ approaches and adds to other benefits for infants of maternal preconception ART including reduced vertical HIV transmission and possibly reduced infant infectious morbidity [32,33].
Evidence continues to accumulate regarding abnormal placental development and adverse birth outcomes associated with ritonavir-boosted lopinavir even when compared to alternative protease-inhibitors [34]. The US, British and European guidelines already recommend against the use of ritonavir-boosted lopinavir during pregnancy [35–37]. World Health Organization guidelines recently shifted away from ritonavir-boosted lopinavir for second-line regimens in all people with HIV, but country-level changes in HIV high prevalence settings are yet to occur [38]. Although this analysis cannot conclusively implicate ritonavir-boosted lopinavir in increasing the risk for dnHDP in PPHIV, it adds weight to the need for countries to carefully consider continuing use of this agent in people of child-bearing potential as the excess maternal and infant morbidity places additional burdens on the public healthcare system.
Our findings emphasize that antenatal and HIV services should pay attention to pregnant people with preexisting hypertension, older pregnant people with and without HIV as well as PPHIV on second/higher-line regimens, to ensure they receive good basic antenatal care with regular monitoring for development of dnHDP and early initiation of its management. Fairly simple, safe and low-cost interventions such as low-dose aspirin initiated before 16 weeks’ gestation or calcium supplementation initiated at any gestation, specifically in people with low-calcium diets as is likely in HIV high prevalence settings, can potentially reduce dnHDP and their complications [39,40]. Further studies on these interventions in older PPHIV or those on second/higher-line regimens could become increasingly relevant as the HIV epidemic evolves.
By the nature of the routinely collected data source utilized our study has limitations. The finding of an increased prevalence of dnHDP in people not on ART should be interpreted with caution as PPHIV with no evidence of receiving ART likely have additional confounders that were unmeasured in this analysis. We did not have measures of maternal body mass index. South Africa has a high prevalence of maternal obesity that itself is a major risk factor for HDP and long-term cardiovascular disease [41,42]. There may be some misclassification of dnHDP, as we did not have gestational age information to accurately determine gestation at first prescription of antihypertensive drugs or direct measures of blood pressure and proteinuria for diagnosis of dnHDP. Our sensitivity analyses in this regard indicate that even with substantial misclassification the relationship between dnHDP and HIV/ART remained consistent. The dnHDP classification has been validated in a sentinel site dataset including 10% of the people in the province-wide cohort with gestational age, measured blood pressure and proteinuria data available, that will be published separately. A major strength of this study is the large number of pregnant people included over a relatively short time with a high prevalence of HIV and sufficient people on ART to interrogate differences by timing of ART initiation and to a limited extent different regimen types. Furthermore, as this cohort constitutes the entire population of pregnant people in the public sector, and not a sample of the population, there is very limited selection bias.
Conclusion
These findings are reassuring, suggesting in the context of universal ART, that ART is not associated with an increased prevalence of dnHDP for most people with HIV. However, certain PPHIV sub-groups, particularly older pregnant people, those with preexisting hypertension and those on second-(or higher) line regimens, require more attention. The optimal ART regimen that balances maternal HIV as well as other obstetric co-morbidity management along with the interests of the fetus needs careful consideration that will be aided by ongoing high quality evidence generation. This Western Cape province-wide cohort approach will be able to continue to efficiently monitor the relationship between HIV and newer ART regimens or HIV treatment modalities in PPHIV as they are implemented.
Acknowledgements
ALS designed the study. F.P. curated and managed the data. A.B. provided essential data engineering oversight. A.L.S. and S.dB. conducted the data analyses. G.T., M.F.C., E.J.A., P.L.W., M.A.D., A.B. provided essential subject matter expertise to the design, analyses and interpretation of the research. A.L.S. wrote the original draft of the paper. All authors have read and approved of the final manuscript. Research reported in this publication was supported by the Fogarty International Center of the National Institutes of Health under award number 1K43TW010683 to A.L.S. and the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health under award number R01HD080465 to A.B. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This research was also supported by the CIPHER Grant Programme of the International AIDS Society under award number 2017/518-SLO to A.L.S.
Conflicts of interest
M.-A.D. receives funding from Viiv Healthcare for an unrelated project. Research reported in this publication was supported by the Fogarty International Center of the National Institutes of Health under award number 1K43TW010683 to A.L.S. and the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health under award number R01HD080465 to A.B.
References
1. Joint United Nations Programme on
HIV/AIDS (UNAIDS). 2022 UNAIDS estimates. 2022. Available at:
https://aidsinfo.unaids.org [Accessed 19 October 2022]
2. United Nations. Political declaration on
HIV and AIDS: ending inequalities and getting on track to end AIDS by 2030 United Nations General Assembly. 2021: 28. Available at:
https://www.unaids.org/sites/default/files/media_asset/2021_political-declaration-on-hiv-and-aids_en.pdf.
3. Slogrove AL.
It is a question of equity: time to talk about children who are HIV-exposed and “HIV-free”.
J Int AIDS Soc 2021; 24:2–4.
4. World Health Organization. Approaches to enhance and accelerate study of new drugs for
HIV and associated infections in pregnant women: meeting report. 2021. Available at:
https://www.who.int/publications/i/item/9789240040182 [Accessed 13 March 2022]
5. Abalos E, Cuesta C, Grosso AL, Chou D, Say L.
Global and regional estimates of preeclampsia and eclampsia: a systematic review.
Eur J Obstet Gynecol 2013; 170:1–7.
6. Say L, Chou D, Gemmill A, Tunçalp Ö, Moller AB, Daniels J, et al.
Global causes of maternal death: a WHO systematic analysis.
Lancet Glob Heal 2014; 2:323–333.
7. McDonald SD, Malinowski A, Zhou Q, Yusuf S, Devereaux PJ.
Cardiovascular sequelae of preeclampsia/eclampsia: a systematic review and meta-analyses.
Am Heart J 2008; 156:918–930.
8. Mol BWJ, Roberts CT, Thangaratinam S, Magee LA, De Groot CJM, Hofmeyr GJ.
Preeclampsia.
Lancet 2016; 387:999–1011.
9. Victora CG, Adair L, Fall C, et al.
Maternal and child undernutrition: consequences for adult health and human capital.
Lancet 2008; 371:340–357.
10. Perin J, Mulick A, Yeung D, et al.
Global, regional, and national causes of under-5 mortality in 2000–19: an updated systematic analysis with implications for the Sustainable Development Goals.
Lancet Child Adolesc Heal 2022; 6:106–115.
11. Brown MA, Magee LA, Kenny LC, et al.
The hypertensive disorders of pregnancy: ISSHP classification, diagnosis & management recommendations for international practice.
Pregnancy Hypertens 2018; 13:291–310.
12. So-Armah K, Benjamin LA, Bloomfield GS, et al.
HIV and cardiovascular disease.
Lancet HIV 2020; 7:e279–e293.
13. Zash R, Jacobson DL, Diseko M, et al.
Comparative safety of antiretroviral treatment regimens in pregnancy.
JAMA Pediatr 2017; 171:e172222.
14. Browne JL, Schrier VJMM, Grobbee DE, Peters SAE, Klipstein-Grobusch K.
HIV, antiretroviral therapy, and hypertensive disorders in pregnancy: a systematic review and meta-analysis.
J Acquir Immune Defic Syndr 2015; 70:91–98.
15. Wimalasundera RC, Larbalestier N, Smith JH, et al.
Preeclampsia, antiretroviral therapy, and immune reconstitution.
Lancet 2002; 360:1152–1154.
16. Hall DR.
Is preeclampsia less common in patients with HIV/AIDS?.
J Reprod Immunol 2007; 76:75–77.
17. Premkumar A, Dude AM, Haddad LB, Yee LM.
Combined antiretroviral therapy for HIV and the risk of hypertensive disorders of pregnancy: a systematic review.
Pregnancy Hypertens 2019; 17:178–190.
18. Gras L, May M, Ryder LP, et al.
Determinants of restoration of CD4 and CD8 cell counts and their ratio in HIV-1 – positive individuals with sustained virological suppression on antiretroviral therapy.
J Acquir Immune Defic Syndr 2019; 80:292–300.
19. Naicker T, Govender N, Abel T, et al.
HIV associated preeclampsia: a multifactorial appraisal.
Int J Mol Sci 2021; 22:1–16.
20. Provincial Government of the Western Cape – Department of Health H (HAST) D. The Western Cape consolidated guidelines for
HIV treatment: prevention of mother-to-child transmission of
HIV (PMTCT), children, adolescents and adults, 2018 (Amended Version). 2018: 1–72. Available at:
https://www.westerncape.gov.za/assets/departments/health/wc_hiv_consolidated_guidelines_march_2018_0.pdf [Accessed 22 February 2022]
21. Moodley J, Soma-Pillay P, Buchmann E, Pattinson RC.
Hypertensive disorders in pregnancy: 2019 National guideline.
South Afr Med J 2019; 109:12723.
22. Boulle A, Heekes A, Tiffin N, et al.
Data centre profile: the provincial health data centre of the Western Cape.
Int J Popul Data Sci 2019; 4:06.
23. Heekes A, Tiffin N, Dane P, et al.
Self-enrolment antenatal health promotion data as an adjunct to maternal clinical information systems in the Western Cape Province of South Africa.
BMJ Glob Heal 2018; 3:e000565.
24. Slogrove AL, Burmen B, Davies M-A, et al.
Standardized definitions of in utero HIV and antiretroviral drug exposure among children.
Clin Infect Dis 2022; 75:347–355.
25. Espelt A, Olmo MM, Penelo E, Bosque- M.
Applied prevalence ratio estimation with different regression models: an example from a cross-national study on substance use research.
Adicciones 2016; 29:105–112.
26. Chen W, Qian L, Shi J, Franklin M.
Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification.
BMC Med Res Methodol 2018; 18:1–12.
27. Phipps EA, Thadhani R, Benzing T, Karumanchi SA.
Preeclampsia: pathogenesis, novel diagnostics and therapies.
Nat Rev Nephrol 2019; 15:275–289.
28. Stoner MCD, Vwalika B, Smid MC, et al.
A retrospective study of HIV, antiretroviral therapy, and pregnancy-associated hypertension among women in Lusaka, Zambia.
Int J Gynecol Obstet 2016; 134:299–303.
29. Hall D, Gebhardt S, Theron G, Grove D.
Preeclampsia and gestational hypertension are less common in HIV infected women.
Pregnancy Hypertens 2014; 4:91–96.
30. South African National Department of Health. 2015 National Antenatal Sentinel
HIV & Syphilis Survey Report. Pretoria,
South Africa: 2017. Available at:
www.health.gov.za [Accessed 12 July 2018]
31. Wilson EMP, Sereti I.
Immune restoration after antiretroviral therapy: the pifalls of hasty or incomplete repairs.
Immunol Rev 2013; 254:343–354.
32. Mandelbrot L, Tubiana R, Chenadec F Le, Dollfus C, Faye A.
No perinatal HIV-1 transmission from women with effective antiretroviral therapy starting before conception.
Clin Infect Dis 2015; 61:1715–1725.
33. Goetghebuer T, Smolen KK, Adler C, et al.
Initiation of antiretroviral therapy before pregnancy reduces the risk of infection-related hospitalization in human immunodeficiency virus – exposed uninfected infants born in a high-income country.
Clin Infect Dis 2019; 68:1193–1203.
34. Dunk CE, Serghides L.
Protease inhibitor-based antiretroviral therapy in pregnancy: effects on hormones, placenta, and decidua.
Lancet HIV 2022; 9:e120–e129.
35. Clinical Info
HIV Gov. Recommendations for the use of antiretroviral drugs in pregnant women with
HIV infection and interventions to reduce perinatal
HIV transmission in the United States. 2021. Available at:
https://clinicalinfo.hiv.gov/sites/default/files/guidelines/documents/Perinatal_GL.pdf [Accessed 22 February 2022]
36. British
HIV Association. BHIVA guidelines for the management of
HIV in
pregnancy and postpartum 2018 (2020 third interim update). 2020: 1–110. Available at:
https://www.bhiva.org/pregnancy-guidelines [Accessed 22 February 2022]
37. European AIDS Clinical Society. Guidelines Version 11. 2021: 631–632. Available at:
https://www.eacsociety.org/media/final2021eacsguidelinesv11.0_oct2021.pdf [Accessed 22 February 2022]
38. World Health Organization.
HIV prevention, testing, treatment, service delivery and monitoring: recommendation for public health approach. 2021: 148–182. Available at:
https://www.who.int/publications/i/item/9789240031593 [Accessed 13 March 2022]
39. Duley L, Meher S, Hunter KE, Seidler AL, Askie LM.
Antiplatelet agents for preventing preeclampsia and its complications (Review).
Cochrane Database Syst Rev 2019; 2019:CD004659.
40. Hofmeyr GJ, Manyame S, Medley N, Williams MJ.
Calcium supplementation commencing before or early in pregnancy, for preventing hypertensive disorders of pregnancy (Review).
Cochrane Database Syst Rev 2019; 9:CD011192.
41. Bengtson AM, Phillips TK, le Roux SM, et al.
Postpartum obesity and weight gain among human immunodeficiency virus-infected and human immunodeficiency virus-uninfected women in South Africa.
Matern Child Nutr 2020; 16:1–13.
42. Hoffman RM, Newhouse C, Chu B, Stringer JSA, Currier JS.
Noncommunicable diseases in pregnant and postpartum women living with HIV: implications for health throughout the life course.
Curr HIV/AIDS Rep 2021; 18:73–86.
