Summary
Background:
Small studies reported poor postpartum outcomes among young women living with perinatal HIV-infection (WLPHIV) who are now aging into adulthood and becoming pregnant. For targeted clinical intervention, we sought to identify WLPHIV at risk of poor postpartum virologic control.
Methods:
We abstracted data on pregnancy history for WLPHIV in the Pediatric HIV/AIDS Cohort Study-AMP Up protocol, a prospective study of young adults living with perinatal HIV from 14 sites in the United States. Linear models with generalized estimating equations described trends in HIV viral load (VL) through one year post-pregnancy by pregnancy outcome. Group-based trajectory modeling (GBTM) identified VL trajectory groups in the first year postpartum after live births. We then compared sociodemographic and clinical factors across identified groups. We defined viremia as ≥400 copies per mL.
Findings:
VL levels increased by 0.7 (95% confidence interval: 0.5, 1.0) log10 copies per mL through the first 12 weeks postpartum after 104 live births and subsequently stabilized from 13 weeks to one-year postpartum (slope: −0.01 (−0.3, 0.3) log10 copies per mL). In comparison, the average VL trajectory after 43 spontaneous/elective abortions remained at levels <400 copies per mL.
GBTM identified three distinct groups of VL trajectories after live births, classified as reflecting sustained suppression, rebound viremia, and persistent viremia, with 31 (30%), 55 (53%), and 18 (17%) postpartum trajectories included in each group, respectively. Women with sustained postpartum suppression were older than those with rebound or persistent postpartum viremia (median age at conception: 22.9 vs. 20.4 and 19.0 years respectively). Pre-conception viremia and immune suppression were also strong risk factors for postpartum viremia.
Interpretation:
Despite success in achieving VL suppression during pregnancy, WLPHIV have a high risk of postpartum viremia. Younger age at conception, pre-conception viremia, and pre-conception immune suppression may identify WLPHIV most likely to benefit from postpartum adherence interventions.
Funding:
National Institutes of Health
Introduction
With antiretroviral therapy (ART), young women living with perinatal HIV-infection (WLPHIV) are reaching child-bearing age and becoming pregnant. Pregnant WLPHIV are unique in their lifelong experience with HIV infection and treatment. Many have had childhood experience with mono- and dual-therapy, increasing the likelihood of drug resistant virus.1–3 There is also the challenge of maintaining adherence to ART in this young population of women who may be in the process of transitioning to adult health care, experiencing depression, isolation and stigma, and dealing with parental loss.4–6 The impact of these life stressors coupled with pregnancy and motherhood on their long-term health is unknown.
The majority of studies on pregnancies among WLPHIV have been case reports or descriptive studies, summarizing maternal age, CD4 counts, and HIV viral load (VL), as well as outcomes including elective terminations, miscarriage, live births, and perinatal transmissions.7–12 A few have commented on high rates of unintended pregnancies7,8,10 and ART resistance7,11 in this population, and one study noted challenges maintaining postpartum retention in care9. Some studies have also compared pregnancies among WLPHIV and women with non-perinatally acquired HIV.3–5,13–15 These have primarily focused on comparing pregnancy characteristics and infant outcomes, such as perinatal transmissions and growth, between these two groups of women. Only three studies have specifically evaluated the postpartum health of WLPHIV.4–6
Phillips et al found that persistent viral replication was more common for up to 6 months postpartum among 11 WLPHIV compared to 27 women with non-perinatally acquired HIV in their Bronx clinic.4 In a follow-up study, Munjal et al found that success in decreasing VL during pregnancy was not maintained postpartum in 30 WLPHIV compared to 35 women with non-perinatally acquired HIV.5 The study also observed 4 deaths among their WLPHIV over 4 years of follow-up. A recent study by Meade et al followed 22 WLPHIV at a health center in Georgia and observed very poor postpartum retention, with only 4 (20%) and 2 (12%) women retained at 12 months and 24 months postpartum, respectively.6 Among those observed at these time points, 0-15% were virally suppressed. These studies suggest a need to specifically study the postpartum health of WLPHIV and identify factors associated with poor outcomes to effectively target resources to those at most risk.
In our present study, we first estimated and compared pregnancy rates between WLPHIV and women who are perinatally HIV-exposed but uninfected (PHEU), and then evaluated postpartum health by describing VL and CD4 trajectories after pregnancy among a large cohort of WLPHIV across multiple sites in the United States (US) by pregnancy outcome (i.e., live births vs. spontaneous/elective abortions). We additionally identified correlates of VL trajectories after live births. We hypothesized that pregnancy rates would be similar between WLPHIV and women who are PHEU and that we would observe increases in VL and decreases in CD4 counts after live births relative to spontaneous/elective abortions. We further hypothesized that postpartum viremia would be associated with younger age at conception, lower education level as a measure of socioeconomic status, first pregnancy, lower pre-conception CD4 counts, and pre-conception viremia.
Methods
Study design and participants
The Adolescent Master Protocol (AMP) Up protocol of the Pediatric HIV/AIDS Cohort Study evaluates the impact of HIV infection and ART on young adults living with perinatal HIV infection through transition into adulthood (https://phacsstudy.org/About-Us/Active-Protocols). Between April 15, 2014 and May 14, 2014, 14 clinical sites in the US and Puerto Rico opened to enroll young adults ≥18 years of age with perinatally acquired HIV-infection and a comparison group of young adults who are PHEU. A complete medical chart history since birth is required, including diagnoses, ART, plasma VL concentrations, and CD4 counts. The AMP Up protocol was approved by the Institutional Review Board at each participating site. Written informed consent is obtained from each participant. We included young women enrolled as of October 1, 2017, who reported ever having heterosexual vaginal intercourse and their age at sexual debut in our study population. For analyses of post-pregnancy VL and CD4 trajectories, we also required having at least two recorded measures in the one year after the end of pregnancy.
Data sources and definitions
We collected sociodemographic and clinical characteristics through self-report or medical chart abstraction. Participants reported their sexual history, including age at sexual debut on an annual online survey. We obtained pregnancy history through chart abstraction and participant self-report. Our analyses were based on data obtained through chart abstraction as estimated date of confinement and date of pregnancy outcome were reported for all pregnancies, therefore allowing us to define postpartum periods and corresponding VL measures. However, we provide data on pregnancy intention obtained through self-report. We also compare reporting of pregnancies between chart abstraction and self-report. For our analyses, we classified pregnancy outcomes as live births, spontaneous abortions (i.e., pregnancy loss before 20 weeks gestation), stillbirths (i.e., pregnancy loss on or after 20 weeks gestation), and elective abortions.
Statistical analysis
We calculated pregnancy rates for WLPHIV and PHEU women as the total number of pregnancies observed over person-time at risk. An individual’s person-time at risk for pregnancy began from their age at sexual debut and concluded at the latest of their last seen date, last clinic visit date, or the last date of the sexual health survey. The last seen date is the last date the participant was seen at the clinical site for clinical care. The last clinic visit date is the date of the last visit that the participant attended for the research protocol. We excluded person-time during a pregnancy (i.e., from the date of their last menstrual period to the conclusion of pregnancy) from person-time at risk and stratified pregnancy rates by age: <15 years, 15-19 years, 20-24 years, 25-29 years, and ≥30 years.
We described trends in VL and CD4 count starting one year prior to conception, during pregnancy, and through one year post-pregnancy for all pregnancies among WLPHIV by pregnancy outcome (live birth or spontaneous/elective abortion) using locally estimated scatterplot smoothing (LOESS) plots. We set VLs below the limit of detection to the limit of detection minus 0.01 copies per mL and included all VLs during the defined periods regardless of whether they coincided with a previous or subsequent pregnancy. To estimate and compare trajectories of VL and CD4 from end of pregnancy through 12 weeks post-pregnancy, and from 13 weeks through one year post-pregnancy by pregnancy outcome, we used hierarchical piecewise linear regression models with generalized estimating equations (GEE). The hierarchical component specified that pregnancies for a given participant were correlated, and that multiple VL readings within each post-pregnancy period were correlated. The models further adjusted for factors that may differ by pregnancy outcome and impact post-pregnancy trajectories: pre-conception viremia defined as percent of VLs ≥400 copies per mL in the one year prior to pregnancy, pre-conception immune suppression defined as percent of CD4 counts <200 cells per μL in the one year prior to pregnancy, age at sexual debut, race/ethnicity, highest caregiver education level attained, and prior live birth.
We used group-based trajectory modeling (GBTM) to identify whether specific patterns in postpartum VLs existed after live births (i.e., whether there were distinct groups that followed similar patterns of VL over the one-year postpartum period after live births). To select the optimal number and shape of the trajectory groups we used the Bayesian information criterion.
We considered one to four trajectory groups to determine the optimal number of groups and constant, linear, quadratic, and cubic terms to determine the optimal shape of the trajectories. GBTM assigned group membership based on the highest probability of group membership. We then compared sociodemographic and clinical factors across identified trajectory groups. We conducted analyses using SAS version 9.4 (SAS Institute, Cary NC).
Role of the Funding source
The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results
Of 323 women in AMP Up, 273 (234 WLPHIV, 39 PHEU) who reported age at sexual debut and history of heterosexual vaginal intercourse were eligible for estimation of pregnancy rates (Table 1 of the Appendix, p.2). WLPHIV were older than PHEU women, with 280/1691 (17%) of their person-years observed at ages 25 years or older compared to 0/201 (0%) among PHEU women (Figure 1a). WLPHIV also tended to be older at sexual debut and at first pregnancy conception compared to PHEU women (median: 16 vs. 15 years, and 19 vs. 18 years, respectively, Table 2 of the Appendix, p.3). Distributions of race, ethnicity, and education level, however, were similar between WLPHIV and PHEU women, with 182/273 (67%) self-reporting as Black, 80/273 (29%) as Hispanic, and 197/273 (72%) with at least a high school education. Across comparable age categories, pregnancy rates among WLPHIV were lower compared to PHEU women (Figure 1b). This difference was particularly large in the 15-19 year age-category, with rates among WLPHIV and PHEU women calculated as 9.5 (95% confidence interval (CI): 7.3, 12.2) and 20.3 (13.3, 29.8) pregnancies per 100 person-years, respectively. Unintended pregnancy was high among both groups of women (WLPHIV: 156/195 or 80% [95% CI: 74%, 85%], women who are PHEU: 23/27 or 85% [95% Cl: 66%, 96%], Table 3 of the Appendix, p.4). Chart abstraction underreported pregnancies among WLPHIV compared to self-report (Table 4 of the Appendix, p.4). Participant self-reports conversely underreported elective abortions compared to chart abstraction among women who are PHEU.
Figure 1.
Person-time at risk (a) and pregnancy rates (b) by HIV status and age category.
Eighty-one percent (190/234) of the WLPHIV were born prior to 1996, before the availability of combination ART. Median age at sexual debut was one year earlier among the 99 WLPHIV with a pregnancy compared to the 135 with no pregnancy history (Table 5 of the Appendix, p.5). Median VL closest to sexual debut was <400 copies per mL among all WLPHIV, though WLPHIV with observed pregnancies had a more substantial history of unsuppressed VLs at levels ≥400 copies per mL through date of sexual debut compared to WLPHIV with no observed pregnancies. The median percent of unsuppressed VLs up to sexual debut was 70% among WLPHIV with a pregnancy history compared to 50% among WLPHIV with no pregnancy history. In contrast to VL, WLPHIV did not have an extensive history of immune suppression (CD4<200 cells per μL) up to sexual debut and this parameter, as well as CD4 count at sexual debut, did not differ between WLPHIV with and without pregnancies.
Of the 99 WLPHIV with at least one pregnancy, 59 (60%) had 1 pregnancy, 20 (20%) had 2, 12 (12%) had 3, 4 (4%) had 4, 3 (3%) had 5, and 1 (1%) had 6, with a total of 172 recorded pregnancies. Seventy-two percent (124/172) of the pregnancies resulted in a live birth, 9% (15/172) were spontaneous abortions, and 19% (33/172) were elective abortions. There were no stillbirths. Pre-conception viremia was more common and median VL near conception was higher among pregnancies resulting in a live birth compared to pregnancies resulting in a spontaneous/elective abortion, though median VL in both groups was <400 copies per mL by the end of pregnancy (Table 1). Pre-conception immune suppression and median CD4 counts near conception and at the end of pregnancy were similar by pregnancy outcome.
Table 1.
Characteristics of WLPHIV study population, by pregnancy outcome.
| Pregnancy outcome | |||
|---|---|---|---|
| Total (N=172) | Live birth (N=124) | Spontaneous/elective abortion (N=48) | |
| Age at conception, years | |||
| Median (Q1, Q3) | 20.9 (19.0, 24.0) | 20.6 (18.8, 24.3) | 21.3 (19.1, 23.6) |
| Highest education level attained* | |||
| Less than high school diploma | 56 (33%) | 39 (31%) | 17 (35%) |
| High school diploma | 51 (30%) | 41 (33%) | 10 (21%) |
| Some college but no degree | 48 (28%) | 32 (26%) | 16 (33%) |
| Associates degree or higher | 17 (10%) | 12 (10%) | 5 (10%) |
| Infant HIV status | |||
| Infected | NA | 1 (1%) | NA |
| Uninfected | NA | 103 (83%) | NA |
| Indeterminate | NA | 13 (10%) | NA |
| Unknown | NA | 7 (6%) | NA |
| Preterm birth | NA | 25 (20%) | NA |
| Number of viral loads in the one year prior to pregnancy | |||
| Median (Q1, Q3) | 4.0 (2.0, 5.5) | 4.0 (2.0, 5.0) | 4.0 (2.0, 6.0) |
| Percent of viral loads ≥ 400 copies per mL in the one year prior to pregnancy | |||
| Median (Q1, Q3) | 50 (0, 100) | 50 (0, 100) | 33 (0, 62) |
| No data for period | 4 (2%) | 4 (3%) | 0 (0%) |
| Nearest viral load to date of last menstrual period, log10 copies per mL | |||
| Median (Q1, Q3) | 2.6 (17, 3.8) | 3.1 (1.8, 3.9) | 2.1 (1.6, 3.5) |
| No data for period | 4 (2%) | 4 (3%) | 0 (0%) |
| Nearest viral load to end of pregnancy, log10 copies per mL | |||
| Median (Q1, Q3) | 1.8 (13, 2.9) | 1.8 (1.4, 3.0) | 1.7 (1.3, 2.6) |
| No data for period | 52 (30%) | 29 (23%) | 23 (48%) |
| Number of CD4s in the one year prior to pregnancy | |||
| Median (Q1, Q3) | 4.0 (2.0, 5.0) | 4 (2.0, 5.0) | 4.5 (3.0, 5.5) |
| Percent of CD4s < 200 cells per uL in the one year prior to pregnancy | |||
| Median (Q1, Q3) | 0 (0, 0) | 0 (0, 0) | 0 (0, 0) |
| No data for period | 4 (2%) | 4 (3%) | 0 (0%) |
| Nearest CD4 to date of last menstrual period, cells per μL | |||
| Median (Q1, Q3) | 458 (281, 641) | 458 (242, 645) | 458 (302, 639) |
| No data for period | 4 (2%) | 4 (3%) | 0 (0%) |
| Nearest CD4 to end of pregnancy, cells per μL | |||
| Median (Q1, Q3) | 449 (254, 645) | 448 (239, 640) | 487 (320, 645) |
| No data for period | 62 (36%) | 39 (31%) | 23 (48%) |
| First ART regimen in postpartum period | |||
| INSTI + PI + NNRTI-based | 13 (8%) | 7 (6%) | 6 (13%) |
| INSTI + PI-based | 12 (7%) | 12 (10%) | 0 (0%) |
| INSTI + NNRTI-based | 1 (1%) | 1 (1%) | 0 (0%) |
| PI + NNRTI-based | 7 (4%) | 5 (4%) | 2 (4%) |
| INSTI only-based | 9 (5%) | 9 (7%) | 0 (0%) |
| PI only-based | 79 (46%) | 60 (48%) | 19 (40%) |
| NNRTI only-based | 23 (13%) | 11 (9%) | 12 (25%) |
| NRTI only | 8 (5%) | 8 (6%) | 0 (0%) |
| No ART | 19 (11%) | 11 (9%) | 8 (17%) |
| Missing postpartum ART | 1 (1%) | 0 (0%) | 1 (2%) |
| Most complex ART regimen in postpartum period | |||
| INSTI + PI + NNRTI-based | 17 (10%) | 10 (8%) | 7 (15%) |
| INSTI + PI-based | 17 (10%) | 16 (13%) | 1 (2%) |
| INSTI + NNRTI-based | 6 (3%) | 4 (3%) | 2 (4%) |
| PI + NNRTI-based | 7 (4%) | 4 (3%) | 3 (6%) |
| INSTI only-based | 19 (11%) | 16 (13%) | 3 (6%) |
| PI only-based | 76 (44%) | 55 (44%) | 21 (44%) |
| NNRTI only-based | 19 (11%) | 11 (9%) | 8 (17%) |
| NRTI only | 4 (2%) | 4 (3%) | 0 (0%) |
| No ART | 6 (3%) | 4 (3%) | 2 (4%) |
| Missing postpartum ART | 1 (1%) | 0 (0%) | 1 (2%) |
Note: 124 live births among 86 WLPHIV females; 48 spontaneous/elective abortions among 32 WLPHIV.
As of earliest report in the AMP Up protocol
WLPHIV: women living with perinatally acquired HIV-infection, ART: antiretroviral therapy, INSTI: integrase strand transfer inhibitor, PI: protease inhibitor, NNRTI: non-nucleoside reverse transcriptase inhibitor, NRTI: nucleoside/nucleotide reverse transcriptase inhibitor.
Of the 172 pregnancies recorded among WLPHIV, 147 (85%, 104 live births, 43 spontaneous/elective abortions) were eligible for post-pregnancy VL trajectory analyses (i.e., had ≥ two VLs in the one year after end of pregnancy). There were no differences in eligibility by pregnancy outcome. LOESS trajectories showed mean VLs in the one year pre-conception period increasing overtime through the beginning of pregnancy (Figure1 of the Appendix, p.7). VLs then suppressed on average by the end of pregnancy before diverging by pregnancy outcome in the one year after the end of pregnancy. Figure 2 shows LOESS and model-based VL load trajectories in the one year post-pregnancy period by pregnancy outcome. From a mean VL of 2.4 log10 copies per mL at the end of pregnancy, VL levels increased by an average of 0.7 (95% Cl): 0.5, 1.0) log10 copies per mL through the first 12 weeks after live births and subsequently stabilized from 13 weeks to one-year postpartum (slope (95% Cl): −0.01 (−0.3, 0.3) log10 copies per mL). In comparison, the average post-pregnancy VL trajectory after spontaneous/elective abortions remained at levels <400 copies per mL. Multivariable adjustment did not substantially change slope parameters.
Figure 2.
LOESS and model-based postpartum HIV viral load trajectories by pregnancy outcome.
In contrast to VL, no differences were observed in post-pregnancy CD4 trajectories by pregnancy outcome. LOESS trajectories essentially showed stable CD4 counts from one year pre-conception through one year post-pregnancy, with a small decrease in CD4 counts during pregnancy prior to live births (Figure 2 of the Appendix, p.8). Unadjusted and adjusted model-based post-pregnancy CD4 trajectories reflected the pattern of stable CD4 counts, around 500 cells per μL, regardless of pregnancy outcome (Figure 3 of the Appendix, p.9).
Given observed increases in postpartum VLs after live births, we further explored patterns in VL over the one year postpartum period in this subset. GBTM identified three distinct groups of postpartum VL trajectories, with a quadratic model as the optimal fit of the data (Figure 3).
Figure 3.
Group-based trajectories of postpartum HIV viral load after live births.
These trajectory groups were classified as reflecting sustained suppression, rebound viremia, and persistent viremia, with 31 (30%), 55 (53%), and 18 (17%) postpartum trajectories included in each group, respectively. Women with sustained postpartum suppression were more likely to be older at conception than those with rebound or persistent postpartum viremia (median: 22.9 vs. 20.4 and 19.0 years respectively, Table 2). While highest level of education attained did not differ by postpartum VL trajectory group, a larger proportion of postpartum VL trajectories reflected sustained suppression after repeat pregnancies (14/38 or 37%) compared to first pregnancies (17/66 or 26%). Pre-conception viremia was a strong risk factor for postpartum viremia: 72% (13/18) of pregnancies with persistent postpartum viremia had all VLs ≥400 copies per mL in the one year prior to pregnancy, compared to 36% (20/55) and 3% (1/31) of pregnancies with postpartum rebound viremia and sustained suppression, respectively. Correspondingly, postpartum viremia after an earlier live birth and pre-conception immune suppression were also strongly associated with postpartum viremia.
Table 2.
Sociodemographic and clinical characteristics by group-based trajectories of postpartum HIV viral load after live births.
| Trajectory group | ||||
|---|---|---|---|---|
| Total (N=104) | Sustained suppression (N=31) | Rebound viremia (N=55) | Persistent viremia (N=18) | |
| Race | ||||
| Black | 64 (62%) | 20 (65%) | 32 (58%) | 12 (67%) |
| White/Other | 33 (32%) | 8 (26%) | 19 (35%) | 6 (33%) |
| Missing | 7 (7%) | 3 (10%) | 4 (7%) | 0 (0%) |
| Ethnicity | ||||
| Hispanic | 38 (37%) | 10 (32%) | 21 (38%) | 7 (39%) |
| Not Hispanic | 66 (63%) | 21 (68%) | 34 (62%) | 11 (61%) |
| Highest education level attained* | ||||
| Less than high school diploma | 32 (31%) | 10 (32%) | 16 (29%) | 6 (33%) |
| High school diploma | 36 (35%) | 10 (32%) | 20 (36%) | 6 (33%) |
| Some college but no degree | 26 (25%) | 7 (23%) | 15 (27%) | 4 (22%) |
| Associates degree or higher | 10 (10%) | 4 (13%) | 4 (7%) | 2 (11%) |
| Age at sexual debut, years | ||||
| Median (Q1, Q3) | 16 (15, 17) | 16 (15, 17) | 16 (14, 17) | 16 (15, 17) |
| Had a prior live birth | 29 (28%) | 9 (29%) | 14 (25%) | 6 (33%) |
| Postpartum viral load trajectory group after previous live birth | ||||
| Rebound or persistent viremia | 22 (21%) | 3 (10%) | 13 (24%) | 6 (33%) |
| Sustained suppression | 6 (6%) | 6 (19%) | 0 (0%) | 0 (0%) |
| Missing | 1 (1%) | 0 (0%) | 1 (2%) | 0 (0%) |
| No prior live birth | 76 (73%) | 22 (71%) | 41 (75%) | 12 (67%) |
| Pregnancy sequence | ||||
| First pregnancy | 66 (63%) | 17 (55%) | 37 (67%) | 12 (67%) |
| Repeat pregnancy | 38 (37%) | 14 (45%) | 18 (33%) | 6 (33%) |
| Age at conception, years | ||||
| Median (Q1, Q3) | 20.4 (18.6, 23.5) | 22.9 (19.4, 25.9) | 20.4 (18.8, 22.2) | 19.0 (17.7, 20.5) |
| Number of viral loads in the one year prior to pregnancy | ||||
| Median (Q1, Q3) | 4.0 (2.0, 6.0) | 4.0 (3.0, 6.0) | 4.0 (3.0, 6.0) | 2 (1.0, 4.0) |
| Percent of viral loads ≥ 400 copies per mL in the one year prior to pregnancy | ||||
| 0% | 24 (23%) | 18 (58%) | 6 (11%) | 0 (0%) |
| 1 to 49% | 19 (18%) | 10 (32%) | 9 (16%) | 0 (0%) |
| 50 to 99% | 23 (22%) | 2 (6%) | 18 (33%) | 3 (17%) |
| 100% | 34 (33%) | 1 (3%) | 20 (36%) | 13 (72%) |
| Missing | 4 (4%) | 0 (0%) | 2 (4%) | 2 (11%) |
| Number of CD4s in the one year prior to pregnancy | ||||
| Median (Q1, Q3) | 4.0 (2.0, 6.0) | 4.0 (3.0, 6.0) | 4.0 (2.0, 5.0) | 2.0 (1.0, 4.0) |
| Percent of CD4s < 200 cells per μL in the one year prior to pregnancy | ||||
| 0% | 79 (76%) | 26 (84%) | 44 (80%) | 9 (50%) |
| 1 to 49% | 2 (2%) | 1 (3%) | 1 (2%) | 0 (0%) |
| 50 to 99% | 10 (10%) | 2 (6%) | 5 (9%) | 3 (17%) |
| 100% | 9 (9%) | 2 (6%) | 3 (5%) | 4 (22%) |
| Missing | 4 (4%) | 0 (0%) | 2 (4%) | 2 (11%) |
| Received prenatal care | ||||
| Yes | 99 (95%) | 29 (94%) | 53 (96%) | 17 (94%) |
| No | 1 (1%) | 0 (0%) | 0 (0%) | 1 (6%) |
| Missing | 4 (4%) | 2 (6%) | 2 (4%) | 0 (0%) |
| Mode of delivery | ||||
| Vaginal | 37 (36%) | 13 (42%) | 23 (42%) | 1 (6%) |
| C-section | 66 (63%) | 18 (58%) | 31 (56%) | 17 (94%) |
| Missing | 1 (1%) | 0 (0%) | 1 (2%) | 0 (0%) |
| Preterm birth | 20 (19%) | 4 (13%) | 13 (24%) | 3 (17%) |
| Infant born with birth defect | 6 (6%) | 1 (3%) | 4 (7%) | 1 (6%) |
| Infant HIV status | ||||
| Infected | 1 (1%) | 0 (0%) | 1 (2%) | 0 (0%) |
| Uninfected | 89 (86%) | 28 (90%) | 47 (85%) | 14 (78%) |
| Indeterminate | 10 (10%) | 3 (10%) | 4 (7%) | 3 (17%) |
| Missing | 4 (4%) | 0 (0%) | 3 (5%) | 1 (6%) |
As of earliest report in the AMP Up protocol
Discussion
Small, single site studies of WLPHIV raised concern of increased postpartum viremia and HIV disease progression.4–6 Our study uniquely sought to compare post-pregnancy VL and CD4 trajectories by pregnancy outcome (i.e., live births vs. spontaneous/elective abortions) and identify distinct postpartum VL trajectories after live births as well as correlated factors among a large cohort of WLPHIV from multiple sites across the US. We confirmed that on average, VLs increased after live births from suppressive levels at the end of pregnancy among WLPHIV. Our estimated VL and CD4 trajectories after live births were similar to the studies evaluating these trajectories among WLPHIV in the Bronx.4,5 However, while one study ended follow-up at six months postpartum and the other averaged postpartum laboratory data up to 9-12 months into a single time point, we estimated longitudinal trends in VL and CD4 through one year postpartum, showing not only the significant increase in VL immediately postpartum but also the persistence of elevated average VLs through one year of follow-up. We additionally observed that this trend was specific to live births. Whether our findings reflect contextual factors associated with having a newborn that present challenges to ART adherence or the health burden of pregnancies carried to term requires further studies. With regards to CD4 counts, we observed no differences in post-pregnancy CD4 trajectories by pregnancy outcome, with average CD4 counts remaining around 500 cells per μL.
Studies among women living with HIV have shown postpartum viral rebound in this general population as well.16–19 However, the prevalence of postpartum viremia may be higher among WLPHIV.4,5 We found viremia to be present in 70% of our observed postpartum periods after live births. In comparison, postpartum viremia among women living with HIV in the US is reported to range from 20-57% with differing definitions of viremia between studies.16–19 Women with non-perinatally acquired HIV, however, may have higher rates of postpartum loss to follow-up (LTFU) compared to WLPHIV. In our study, 15% were missing two or more VLs in the one year postpartum period. Similar reported rates of postpartum LTFU among women living with HIV in the US range from 24% to 63%.16–21 Qualitative studies among women living with HIV have identified access to and cost of transportation, competing work and childcare responsibilities, lack of social support, experiences of institutionalized stigma, and depression as barriers to postpartum retention.22,23 While similar qualitative work has not been conducted specifically among postpartum WLPHIV, a history of contact with the health care system for HIV care since birth and therefore stronger relationships with healthcare providers may be associated with relatively better postpartum retention among this population.
ART resistance may be associated with postpartum viremia. While we did not assess resistance to ART received in the postpartum period, previous studies among pregnant WLPHIV have noted challenges in achieving viral suppression due to the high prevalence of ART resistance in this highly treatment-experienced population7,11 and approximately 75% of adolescents living with perinatal HIV infection in the US are estimated to have ART resistance2. We observed persistent viremia from the end of pregnancy through the postpartum period in 17% after live births. However, 53% experienced rebound viremia from suppressed levels at the end of pregnancy, suggesting that poor postpartum adherence to ART may play a larger role in postpartum viremia. Studies among women living with HIV in the US have noted significant decreases in adherence in the postpartum period relative to pregnancy.24,25 Qualitative data among these women identified having a busy schedule with a newborn, multiple children, large pill size and taste, stigma, and medication side effects as barriers to adherence in the postpartum period.22 Facilitators of postpartum adherence included wanting to stay healthy for their own well-being as well as for the care of their children, receiving family support, and medication reminders.22 Unfortunately, we did not directly assess postpartum ART adherence for women in our cohort through self-report or drug levels in biological specimens. We also did not collect information on nausea or other symptoms during the postpartum period, which may have provided insight on barriers to adherence in our population.
We identified younger age at conception, pre-conception viremia, and pre-conception immune suppression as key risk factors for postpartum viremia among our WLPHIV. These factors may be useful in identifying women who may benefit from interventions aimed at improving postpartum outcomes. They may also be useful in developing a risk score that could be utilized in clinical practice to identify women at risk for postpartum viremia, similar to one developed among women in Zambia to identify those most likely to be lost to postpartum follow-up.26 Examples of potential postpartum interventions include care coordination between maternal HIV and pediatric care, peer support, and technology interventions for appointment and medication reminders.27 Success with case management to improve postpartum retention and viral suppression at one year has been reported within a population of women living with HIV in Philadelphia.28 Such strategies may be useful for WLPHIV at high risk for postpartum viremia.
We began our analyses by estimating pregnancy rates to understand the incidence of potential risk periods for post-pregnancy viremia among WLPHIV. We observed lower pregnancy rates among WLPHIV compared to PHEU women, but comparable to sexually experienced 15-19 year olds in the US in 2013 (10.1 per 100 women).29 Overall, 42% of WLPHIV in our study had at least one recorded pregnancy, and 17% had more than one pregnancy, highlighting the aging of women born with HIV into adulthood and the need to understand their reproductive and sexual health, including pregnancy intention, potential pregnancy complications, and their postpartum health. With increasing numbers of adolescent girls with perinatal infection aging into young adulthood globally, this need will continue to grow.30
Our study enrolled WLPHIV in care. Our results therefore represent conservative estimates of postpartum VL and CD4 trends among WLPHIV. We also defined viremia as ≥400 copies per mL due to variability in detection limits during our observed calendar time horizon (range: <20 to <400, over the calendar years 2000-2017). Compared to studies utilizing lower detection limits to define viremia, the prevalence of postpartum viremia among WLPHIV may be even higher than what we report. Our reported pregnancy rates may also be underestimated as miscarriages or elective abortions may go unnoticed or unreported to health providers. We did not collect data on breastfeeding in our study. However, as our cohort includes WLPHIV in the US, breastfeeding would be rare, preventing any potential to evaluate its impact on postpartum outcomes.
In conclusion, WLPHIV, who are unique in their experience with life-long HIV, have a very high risk for postpartum viremia, despite success in achieving VL suppression during pregnancy.
Factors associated with rebound and persistent postpartum viremia in this population, including younger age at conception, pre-conception viremia, and pre-conception immune suppression may help identify women most likely to benefit from targeted postpartum adherence interventions.
Supplementary Material
Research in context.
Evidence before this study
Two small studies conducted in the Bronx, New York found an increased risk of postpartum viremia and mortality among women living with perinatal HIV infection (WLPHIV) compared to women with non-perinatally acquired HIV. To search for similar studies and those evaluating pregnancies among WLPHIV in general, we searched PubMed using the terms “perinatally” and “HIV” and “pregnancy” in December of 2016 for the study proposal and again in November of 2018 for drafting of the manuscript. We reviewed the titles and abstracts of all search results for relevant studies, defined as those that described pregnancies among WLPHIV. We then summarized these studies by noting strengths and limitations, and putting them in context. Our literature review identified the specific need to study the postpartum health of WLPHIV in a larger population.
Added value of this study
To date, only three studies have specifically evaluated the postpartum health of WLPHIV. All three were single setting studies with a maximum sample size of 30 WLPHIV. Our study uniquely sought to compare post-pregnancy viral load and CD4 trajectories by pregnancy outcome (i.e., live births vs. spontaneous/elective abortions) and identify distinct groups of postpartum VL trajectories after live births as well as correlated factors among a large cohort of WLPHIV from multiple sites across the United States (US). This study addresses the growing and increasingly urgent need to understand the reproductive and sexual health of WLPHIV. As the population of young adults born with HIV increases in countries with the highest prevalence of HIV, the US setting, which was fortunate to have early introduction of antiretroviral therapy, provides a young adult cohort of WLPHIV, whose outcomes may inform care in other settings.
Implications of all the available evidence
The cumulative evidence to date on the postpartum health of WLPHIV suggests a need for targeted postpartum adherence interventions, particularly for young women giving birth and those who had difficulty with achieving viral suppression prior to conception. WLPHIV have accumulated a lifetime of experience living with HIV and taking daily antiretroviral therapy and therefore represent a unique population with specific treatment and care management needs.
Acknowledgments
We thank the children and families for their participation in PHACS, and the individuals and institutions involved in the conduct of PHACS. The study was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development with co-funding from the National Institute on Drug Abuse, the National Institute of Allergy and Infectious Diseases, the Office of AIDS Research, the National Institute of Mental Health, the National Institute of Neurological Disorders and Stroke, the National Institute on Deafness and Other Communication Disorders, the National Institute of Dental and Craniofacial Research, and the National Institute on Alcohol Abuse and Alcoholism, through cooperative agreements with the Harvard T.H. Chan School of Public Health (HD052102) (Principal Investigator: George Seage; Program Director: Liz Salomon) and the Tulane University School of Medicine (HD052104) (Principal Investigator: Russell Van Dyke; Co-Principal Investigator: Ellen Chadwick; Project Director: Patrick Davis). Data management services were provided by Frontier Science and Technology Research Foundation (PI: Suzanne Siminski), and regulatory services and logistical support were provided by Westat, Inc (PI: Julie Davidson).
Funding
The Pediatric HIV/AIDS Cohort Study (PHACS) was supported by the Eunice Kennedy Shriver National Institute Of Child Health & Human Development (NICHD) with co-funding from the National Institute Of Dental & Craniofacial Research (NIDCR), the National Institute Of Allergy And Infectious Diseases (NIAID), the National Institute Of Neurological Disorders And Stroke (NINDS), the National Institute On Deafness And Other Communication Disorders (NIDCD), Office of AIDS Research (OAR), the National Institute Of Mental Health (NIMH), the National Institute On Drug Abuse (NIDA), and the National Institute On Alcohol Abuse And Alcoholism (NIAAA), through cooperative agreements with the Harvard T.H. Chan School of Public Health (HD052102) and the Tulane University School of Medicine (HD052104).
Note: The conclusions and opinions expressed in this article are those of the authors and do not necessarily reflect those of the National Institutes of Health or U.S. Department of Health and Human Services.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Declaration of interests
AM has received honorariums as an advisory board member from Antiva, Merck and GlaxoSmithKline. All other authors declare no competing interests.
Clinical Site Investigators and Staff
The following institutions, clinical site investigators and staff participated in conducting PHACS AMP and AMP Up in 2017, in alphabetical order: Ann & Robert H. Lurie Children’s Hospital of Chicago: Ram Yogev, Margaret Ann Sanders, Kathleen Malee, Yoonsun Pyun; Baylor College of Medicine: William Shearer, Mary Paul, Norma Cooper, Lynnette Harris; Bronx Lebanon Hospital Center: Murli Purswani, Mahboobullah Mirza Baig, Alma Villegas; Children’s Diagnostic & Treatment Center: Ana Puga, Sandra Navarro, Patricia A. Garvie, James Blood; Boston Children’s Hospital: Sandra K. Burchett, Nancy Karthas, Betsy Kammerer; Jacobi Medical Center: Andrew Wiznia, Marlene Burey, Ray Shaw, Raphaelle Auguste; Rutgers - New Jersey Medical School: Arry Dieudonne, Linda Bettica, Juliette Johnson; St. Christopher’s Hospital for Children: Janet S. Chen, Maria Garcia Bulkley, Taesha White, Mitzie Grant; St. Jude Children’s Research Hospital: Katherine Knapp, Kim Allison, Megan Wilkins, Jamie Russell-Bell; San Juan Hospital/Department of Pediatrics: Midnela Acevedo-Flores, Heida Rios, Vivian Olivera; Tulane University School of Medicine: Margarita Silio, Medea Gabriel, Patricia Sirois; University of California, San Diego: Stephen A. Spector, Kim Norris, Sharon Nichols; University of Colorado Denver Health Sciences Center: Elizabeth McFarland, Emily Barr, Carrie Glenny, Jennifer Dunn; University of Miami: Gwendolyn Scott, Grace Alvarez, Gabriel Fernandez, Anai Cuadra.
References
- 1.Van Dyke RB, Patel K, Siberry GK, et al. ; Pediatric HIV/AIDS Cohort Study. Antiretroviral treatment of US children with perinatally acquired HIV infection: temporal changes in therapy between 1991 and 2009 and predictors of immunologic and virologic outcomes. J Acquir Immune Defic Syndr 2011; 57: 165–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Van Dyke RB, Patel K, Kagan RM, et al. ; Pediatric HIV/AIDS Cohort Study (PHACS). Antiretroviral Drug Resistance Among Children and Youth in the United States With Perinatal HIV. Clin Infect Dis 2016; 63: 133–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Badell ML, Kachikis A, Haddad LB, Nguyen ML, Lindsay M. Comparison of pregnancies between perinatally and sexually HIV-infected women: an observational study at an urban hospital. Infect Dis Obstet Gynecol 2013; 2013: 301763. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Phillips UK, Rosenberg MG, Dobroszycki J, et al. Pregnancy in women with perinatally acquired HIV-infection: outcomes and challenges. AIDS Care 2011; 23: 1076–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Munjal I, Dobroszycki J, Fakioglu E, et al. Impact of HIV-1 infection and pregnancy on maternal health: comparison between perinatally and behaviorally infected young women. Adolesc Health Med ther 2013; 4: 51–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Meade CM, Hussen SA, Momplaisir F, Badell M, Hackett S, Sheth AN. Long term engagement in HIV care among postpartum women with perinatal HIV infection in the United States. AIDS Care 2018; 30: 488–492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Centers for Disease Control and Prevention (CDC). Pregnancy in perinatally HIV-infected adolescents and young adults--Puerto Rico, 2002. MMWR Morb Mortal Wkly Rep 2003; 52:149–51. Erratum in: MMWR Morb Mortal Wkly Rep 2003; 52: 191. [PubMed] [Google Scholar]
- 8.Levine AB, Aaron E, Foster J. Pregnancy in perinatally HIV-infected adolescents. J Adolesc Health 2006; 38: 765–8. [DOI] [PubMed] [Google Scholar]
- 9.Meloni A, Tuveri M, Floridia M, et al. Pregnancy care in two adolescents perinatally infected with HIV. AIDS Care 2009; 21: 796–8. [DOI] [PubMed] [Google Scholar]
- 10.Kenny J, Williams B, Prime K, Tookey P, Foster C. Pregnancy outcomes in adolescents in the UK and Ireland growing up with HIV. HIV Med 2012; 13: 304–8. [DOI] [PubMed] [Google Scholar]
- 11.Cruz ML, Santos E, Benamor Teixeira Mde L, et al. Viral Suppression and Resistance in a Cohort of Perinatally-HIV Infected (PHIV+) Pregnant Women. Int J Environ Res Public Health 2016; 13: 568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Prieto LM, Fernandez McPhee C, Rojas P, et al. ; Madrid Cohort of HIV-Infected Mother-Infant Pairs. Pregnancy outcomes in perinatally HIV-infected young women in Madrid, Spain: 2000-2015. PLoS One 2017; 12: e0183558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lundberg P, Andersson R, Machado ES, Costa TPD, Hofer CB. Pregnancy outcomes in young mothers with perinatally and behaviorally acquired HIV infections in Rio de Janeiro. Braz J Infect Dis 2018; 22: 412–417 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jao J, Kacanek D, Williams PL, et al. ; Pediatric HIV/AIDS Cohort Study and the International Maternal Pediatric Adolescent AIDS Clinical Trials P1025 Protocol; Pediatric HIV/AIDS Cohort Study and the International Maternal Pediatric Adolescent AIDS Clinical Trials P1025 Protocol. Birth Weight and Preterm Delivery Outcomes of Perinatally vs Nonperinatally Human Immunodeficiency Virus-Infected Pregnant Women in the United States: Results From the PHACS SMARTT Study and IMPAACT P1025 Protocol. Clin Infect Dis 2017; 65: 982–989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Goodenough CJ, Patel K, Van Dyke RB; Pediatric HIV/AIDS Cohort Study (PHACS). Is There a Higher Risk of Mother-to-child Transmission of HIV Among Pregnant Women With Perinatal HIV Infection? Pediatr Infect Dis J 2018; 37: 1267–1270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sha BE, Tierney C, Cohn SE, et al. ; Aids Clinical Trials Group ACTG A5150 Team. Postpartum viral load rebound in HIV-1-infected women treated with highly active antiretroviral therapy: AIDS Clinical Trials Group Protocol A5150. HIV Clin Trials 2011; 12: 9–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Adams JW, Brady KA, Michael YL, Yehia BR, Momplaisir FM. Postpartum Engagement in HIV Care: An Important Predictor of Long-term Retention in Care and Viral Suppression. Clin Infect Dis 2015; 61: 1880–1887. [DOI] [PubMed] [Google Scholar]
- 18.Swain CA, Smith LC, Nash D, et al. Postpartum Loss to HIV Care and HIV Viral Suppression among Previously Diagnosed HIV-Infected Women with a Live Birth in New York State. PLoS One 2016; 11: e0160775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Patel M, Tedaldi E, Armon C, et al. ; HOPS Investigators. HIV RNA Suppression during and after Pregnancy among Women in the HIV Outpatient Study, 1996 to 2015. J Int Assoc Provid AIDS Care 2018; 17: 2325957417752259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Al Rana, Gillani FS, Flanigan TP, Nash BT, Beckwith CG. Follow-up care among HIV-infected pregnant women in Mississippi. Journal of Women’s Health 2010; 19: 1863–1867. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Siddiqui R, Bell T, Sangi-Haghpeykar H, Minard C, Levison J. Predictive factors for loss to postpartum follow-up among low income HIV-infected women in Texas. AIDS Patient Care and STDs 2014; 28: 248–253. [DOI] [PubMed] [Google Scholar]
- 22.Boehme AK, Davies SL, Moneyham L, Shrestha S, Schumacher J, Kempf MC. A qualitative study on factors impacting HIV care adherence among postpartum HIV-infected women in the rural southeastern USA. AIDS Care 2014; 26: 574–581. [DOI] [PubMed] [Google Scholar]
- 23.Buchberg MK, Fletcher FE, Vidrine DJ, et al. A mixed-methods approach to understanding barriers to postpartum retention in care among low-income, HIV-infected women. AIDS Patient Care and STDs 2015: 29: 126–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bardeguez AD, Lindsey JC, Shannon M, et al. ; PACTG 1025 Protocol Team. Adherence to antiretrovirals among US women during and after pregnancy. J Acquir Immune Defic Syndr 2008; 48: 408–417. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Mellins CA, Chu C, Malee K, et al. Adherence to antiretroviral treatment among pregnant and postpartum HIV-infected women. AIDS Care 2008; 20: 958–968. [DOI] [PubMed] [Google Scholar]
- 26.Bengtson AM, Chibwesha CJ, Westreich D, et al. A risk score to identify HIV-infected women most likely to become lost to follow-up in the postpartum period. AIDS Care 2016; 28:1035–1045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Momplaisir FM, Storm DS, Nkwihoreze H, Jayeola O, Jemmott JB. Improving postpartum retention in care for women living with HIV in the United States. AIDS 2018; 32:133–142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Anderson EA, Momplaisir FM, Corson C, Brady KA. Assessing the Impact of Perinatal HIV Case Management on Outcomes Along the HIV Care Continuum for Pregnant and Postpartum Women Living With HIV, Philadelphia 2005-2013. AIDS Behav 2017; 21:2670–2681. [DOI] [PubMed] [Google Scholar]
- 29.Kost K, Maddow-Zimet I, Arpaia A. Pregnancies, Births and Abortions Among Adolescents and Young Women in the United States, 2013: National and State Trends by Age, Race and Ethnicity, New York: Guttmacher Institute, 2017. Accessed at https://www.guttmacher.org/report/us-adolescent-pregnancy-trends-2013 on 2/1/2019 [Google Scholar]
- 30.Slogrove AL, Mahy M, Armstrong A, Davies MA. Living and dying to be counted: What we know about the epidemiology of the global adolescent HIV epidemic. J Int AIDS Soc 2017; 20: 21520. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.