Objective
Data from the international literature suggest that there may be an association between maternal human immunodeficiency virus (HIV) infection and vasculoplacental complications during pregnancy. Studies on this subject have reached discordant conclusions. The aim of this study was to assess the incidence of vasculoplacental complications during pregnancy in women with and without HIV infection.
Study Design
This single-center case-control study compared the incidence of pregnancy-related hypertension, preeclampsia, eclampsia, and vascular intrauterine growth restriction in 280 women with HIV and 560 women not infected with HIV, matched for age, parity, and geographic origin.
Results
The incidence rates of pregnancy-related hypertension, preeclampsia, eclampsia, and vascular growth restriction did not differ between the women with and without HIV infection. The overall incidence of vasculoplacental complications did not differ between the 2 groups (7.5% vs 9.8%, respectively; P = .27). The risk of these was not associated with exposure to antiretroviral treatments, viral load, or CD4 T-cell counts at the beginning of pregnancy.
Conclusion
This study shows no difference in the incidence of vasculoplacental complications between women with and without HIV infection.
The pregnancy-related diseases referred to as vasculoplacental (also called ischemic placental disease), such as preeclampsia, eclampsia, and intrauterine growth restriction (IUGR) of vasculoplacental origin, are a major cause of morbidity and mortality in both mothers and fetuses.
Some studies suggest that maternal infection by human immunodeficiency virus (HIV) may be associated with these conditions, regardless of whether or not the patient is receiving antiretroviral treatment (ARV). The data from the literature are nonetheless discordant; some report an association with an increased risk of vasculoplacental complications. Inversely, others find results suggesting a protective effect, and others show no association. These data are often difficult to interpret, especially because they frequently lack information about the stage and severity of these women’s HIV infections, their treatment, and whether confounding factors, such as geographic origin, drug use, and smoking, were taken into account.
The appearance of new ARV drugs and changes in guidelines and practices for this treatment may also explain these discordances, at least in part. The prescription of these drugs at the beginning of the second trimester is now recommended, even if the mother’s condition does not indicate a need for treatment because the combination of three antiretroval drugs (highly active antiretroviral therapy [HAART]) is associated with a reduction in the maternal viral load, infant mortality, and the rate of mother-to-child transmission. The benefits of this strategy in terms of the prevention of mother-to-child transmission are incontestable, but the safety of these drugs is nonetheless questionable, and their toxicity might promote the development of vasculoplacental diseases.
The principal objective of this retrospective case-control study was to determine whether HIV infection and ARV exposure are associated with an increased risk of vasculoplacental disease. The demonstration of an increased risk would allow obstetricians not only to provide better information to their patients but also to adapt the prenatal monitoring of infected pregnant women and their fetuses to reduce the morbidity associated with this disease.
Materials and Methods
This single-center retrospective case-control study took place in the obstetrics department of the Bichat-Claude Bernard University Hospital (Paris, France), a level IIb maternity ward. It considered for inclusion all women who gave birth from December 2006 through October 2010.
The Paris North Institutional Review Board (Comité d’Évaluation de l’Éthique des projets de Recherche Biomédicale) approved this study (approval number 11.090).
The women in the case group met the following inclusion criteria: HIV infection known at the time of delivery; delivery after 20 weeks of gestation, regardless of pregnancy outcome; management and delivery planned at Bichat-Claude Bernard, regardless of the actual place of delivery; and singleton pregnancy.
The control group, unexposed to HIV (negative serology results during pregnancy, for this screening test was routinely proposed to all pregnant women), were identified from the delivery and transfer registers for the pregnant women signed up for delivery in our institution and otherwise met the same inclusion criteria. One-to-two matching was performed according to the following 3 criteria: maternal age group (≤19, 20-22, 23-25, 26-28, 29-31, 32-34, 35-37, 38-40, ≥41 years), geographic origin (Europe, North Africa, West Indies, Asia, sub-Saharan Africa), and parity.
The exclusion criteria were as follows: pregnancy ended before 20 weeks, fetal malformation, chromosomal abnormalities, multiple pregnancy, fetal cytomegalovirus infection, HIV serological status unknown or not reported in the file, woman already included in the study for a later pregnancy (only the last pregnancy was included).
The number of subjects necessary was calculated based on the following hypotheses: we assumed a frequency of preeclampsia and/or vascular IUGR of 8% in the population unexposed to HIV (2010 data from the Bichat-Claude Bernard obstetrics department database) and twice this rate for women with HIV and treated by antiretroviral drugs. To show such a doubled rate, the study required 270 cases and 540 controls (alpha risk, 5%, power, 80%, 2-tailed test).
All pregnant women with HIV infection were followed up and treated according to the French guidelines then applicable.
We collected data about the women’s medical and obstetric history; the characteristics of their HIV infection and its management, especially ARV treatment before and during pregnancy; and the course and outcome of the pregnancy.
Pregnancy-related hypertension (PRH) was defined by systolic blood pressure (SBP) greater than 140 mm Hg and/or diastolic blood pressure (DBP) greater than 90 mm Hg that developed after 20 weeks of gestation and disappeared before the end of the sixth week postpartum. Preeclampsia was defined by the combination of PRH and proteinuria (>0.3 g per 24 hours), and the syndrome of hemolysis, elevated liver enzymes, and low platelet count (HELLP) by the combination of hemolysis, hepatic cytolysis, and thrombocytopenia.
Preeclampsia was defined as severe when the patient also had any 1 or more of the following: severe hypertension (SBP >160 mm Hg or DBP >110 mm Hg), kidney damage (oliguria [<500 mL per 24 hours] or creatinine levels >135 μmol/L or proteinuria >5 g/d), acute pulmonary edema, persistent severe epigastric pain (Chaussier sign), HELLP syndrome, eclampsia, intractable neurological disorders (visual disorders, polykinetic deep tendon reflexes, or headaches), thrombocytopenia less than 100,000 plt/L, or abruptio placentae.
Uterine artery Doppler waveforms were defined as abnormal when the resistance index was high or when a unilateral or bilateral notch was observed, confirmed at 22 weeks.
Vascular IUGR was defined by the combination of IUGR (less than the 10th percentile) with at least 1 of the following: a history of vasculoplacental complication, abnormal uterine Doppler waveforms, elevated umbilical artery resistance index, or a vasculoplacental complication after other causes were ruled out.
A composite vasculoplacental criterion was defined by the presence of PRH or preeclampsia or IUGR less than the 10th percentile from a vascular cause.
Term at delivery was determined from the date the pregnancy began, measured by the crown-rump length from the first-trimester ultrasound, if available, and otherwise from the date of the last menstrual period.
The statistical analyses were performed with STATA software version 10 (StataCorp LP, College Station, TX).
Measurements for birthweight (in grams) and head circumference (in centimeters) at birth were standardized and expressed as percentiles after adjustment for the newborn’s gestational age (in completed weeks of gestation) and sex, according to the reference curves from the French Audipog network ( www.audipog.net/ ).
The categorical variables were compared with χ 2 tests or a Fisher exact test, as appropriate. The discrete variables in 2 classes were compared with a Student t test and continuous variables (with normal or nonnormal distributions) with a Wilcoxon’s test.
The associations between vasculoplacental complications and candidate risk factors were first explored by the calculation of the crude odds ratios. Adjusted odds ratios were then calculated with 2 different regression models. The first was a logistic regression model and considered the population of women with HIV infection and tested disease indicators (Centers for Disease Control and Prevention [CDC] stage, CD4 T-cell count, and viral load, measured at some point between the beginning of pregnancy and week 14), after adjustment for age and parity. The second was a Cox proportional hazard model suited for matched cohort. This model was used to examine the entire study population and tested the variable of HIV infection by adjusting for the known risk factors of vascular complications and for gestational age at delivery (3 classes: ≤37 weeks, 32-37 weeks, <32 weeks). Schoenfeld’s residual test was used to test the proportional hazards assumption.
For all of these statistical tests, a value of P < .05 was considered statistically significant.
Results
During the study period, 8273 women gave birth to singleton infants after 20 weeks’ gestation. Among them, 307 had HIV infection and met the study’s inclusion criteria. We excluded 27 already included in the study with a more recent pregnancy. Finally, 280 women were included in the case group (273 subtype 1, 5 subtype 2, and 2 simultaneously infected by subtypes 1 and 2) and matched with 560 uninfected women.
Data for 832 of these 840 patients were available in records stored onsite. The 8 missing or incomplete records were those of women who gave birth in other maternity units; we were able to obtain all of them from the hospital of birth ( Figure ).
Table 1 summarizes the data about the status of the HIV infection of the 280 infected women. The diagnosis had been made during a previous pregnancy in 27.1% of the cases, and during the pregnancy studied here in 15.4%, no case of seroconversion during pregnancy has been identified. The CDC stage was 1 for 78.2% of the women. Of the women whose infection was known before the study pregnancy, 47.9% had received ARV before the beginning of the pregnancy. At the beginning of this pregnancy, 44.7% had an undetectable viral load (threshold 50 copies/mL) and 21% a CD4 T-cell count less than 300/mm3.
| Characteristics | HIV-positive women (n = 280) |
|---|---|
| HIV diagnosed more than 1 y ago, n (%) | 231 (82.5) |
| HIV serotype, n (%) | |
| Type 1 | 273 (97.5) |
| Type 2 | 5 (1.8) |
| Types 1 and 2 | 2 (0.7) |
| Circumstances of diagnosis, n (%) | |
| Current pregnancy | 43 (15.4) |
| Previous pregnancy | 76 (27.1) |
| Voluntary screening | 48 (17.1) |
| Marker | 28 (10.0) |
| HIV-positive partner | 9 (3.2) |
| Unknown | 76 (27.1) |
| Seroconversion during this pregnancy | 0 |
| Mode of transmission, n (%) | |
| Sexual | 93 (33.2) |
| Drug dependence | 1 (0.4) |
| Transfusion | 4 (1.4) |
| Workplace accident | 1 (0.4) |
| Unknown | 166 (59.3) |
| CDC stage, n (%) | |
| 1 | 219 (78.2) |
| 2 | 30 (10.7) |
| 3 | 13 (4.6) |
| Unknown | 18 (6.4) |
| Treatment before this pregnancy, n (%) | |
| Antiretroviral treatment | 134 (47.9) |
| Bactrim | 8 (2.9) |
| Laboratory test status before the pregnancy among the 237 patients with known HIV infection | |
| CD4 T-cell count, n (%) | |
| <300/mm 3 | 38 (16.0) |
| >300/mm 3 | 140 (59.1) |
| Unknown | 59 (24.9) |
| Viral load, n (%) | |
| <50 copies/mL | 106 (44.7) |
| >50 copies/mL | 79 (33.3) |
| Unknown | 52 (21.9) |
During the third trimester of the pregnancy, 94% of the infected women received ARV treatment. In a very large percentage of cases, this consisted of a combination of 2 nucleoside analogs and a protease inhibitor (90.5%). Much more rarely (6.8%), it was 2 nucleoside analogs with a nonnucleoside analog. Other combinations were extremely rare.
Checking the quality of the matching, we observed that the 2 groups were comparable for age, parity, and geographic origin. No difference was observed for a history of hypertensive diseases, nephropathy, diabetes, smoking or consumption of illegal drugs, in utero death, or IUGR ( Table 2 ).
| Characteristic | HIV-positive women (n = 280) | HIV-negative women (n = 560) | P value |
|---|---|---|---|
| Age (y), mean ± SD | 31.0 ± 5 | 30.9 ± 5 | .82 |
| Height (m), mean ± SD | 1.66 ± 0.06 | 1.66 ± 0.06 | .75 |
| Prepregnancy weight (kg), mean ± SD | 69.4 ± 13.6 | 68.9 ± 13.8 | .65 |
| Prepregnancy BMI (kg/m 2 ) mean ± SD | 25.2 ± 4.9 | 25.1 ± 4.8 | .70 |
| BMI (kg/m 2 ), n (%) | |||
| >25 | 103 (36.8) | 151 (27.0) | .34 |
| >30 | 37 (13.2) | 217 (38.7) | .69 |
| Geographic origin, n (%) | |||
| Europe | 23 (8.2) | 46 (8.2) | 1.00 |
| North Africa | 12 (4.3) | 24 (4.3) | |
| West Indies | 4 (1.4) | 8 (1.4) | |
| Sub-Saharan Africa | 240 (85.7) | 480 (85.7) | |
| Asia | 1 (0.4) | 2 (0.4) | |
| History of hypertension, n (%) | |||
| Essential hypertension | 8 (2.9) | 13 (2.3) | .09 |
| Pregnancy-related hypertension | 0 | 9 (1.6) | |
| Preeclampsia | 1 (0.4) | 7 (1.2) | |
| Nephropathy, n (%) | 5 (1.8) | 14 (2.5) | .51 |
| Diabetes, n (%) | |||
| Type I | 3 (1.1) | 15 (2.7) | .32 |
| Type II | 0 | 1 (0.2) | |
| Pregnancy related | 3 (1.1) | 10 (1.8) | |
| History of tuberculosis, n (%) | 16 (5.7) | 5 (0.9) | < .001 |
| History of malaria, n (%) | 47 (16.8) | 39 (7.0) | < .001 |
| Sickle cell disease SS, n (%) | 0 | 0 | — |
| Use of toxic substances, n (%) | |||
| Smoking | 22 (8.1) | 29 (5.2) | .96 |
| Alcohol | 4 (1.5) | 15 (2.7) | .28 |
| Cocaine | 1 (0.37) | 0 | .15 |
| Crack | 1 (0.4) | 0 | .15 |
| Heroin | 1 (0.4) | 0 | .15 |
| Cannabis | 0 | 2 (0.4) | .32 |
| Number of pregnancies (mean ± SD) | 3.7 ± 2.1 | 3.4 ± 1.8 | .99 |
| Multiparous, n (%) | 203 (72.5) | 404 (72.1) | .91 |
| Obstetrical history, n (%) | |||
| Elective termination of pregnancy | 136 (48.6) | 176 (31.4) | < .001 |
| Early miscarriage | 60 (21.4) | 136 (24.3) | .36 |
| Late miscarriage | 7 (2.5) | 13 (2.3) | .87 |
| Medical termination of pregnancy | 7 (2.5) | 10 (1.8) | .49 |
| In utero fetal death | 7 (2.5) | 15 (2.7) | .88 |
| Intrauterine growth restriction | 29 (10.4) | 43 (7.7) | .19 |
| Cesarean delivery | 93 (33.2) | 104 (18.6) | < .001 |
| Loop electrosurgical excision | 19 (6.8) | 9 (1.6) | < .001 |
The case and control groups did not differ in their prevalence of the composite vasculoplacental criterion (9.8% vs 7.7%, respectively; P = .269) or PRH, preeclampsia, or severe preeclampsia. We observed a higher prevalence of abnormal uterine artery Doppler waveforms in the HIV-infected group (16.8% vs 12.5%; P = .02).
The rates of preterm birth before 37 weeks and before 32 weeks were significantly higher in the HIV group (18.6% vs 8.0%, P < .001 and 5.0% vs 2.3%, P = .04, respectively) ( Table 3 ). This difference affected only spontaneous preterm births; induced preterm births did not differ at all (3.6% vs 3.6%; P = 1.00).
| Outcomes | HIV-positive women (n = 280) | HIV-negative women (n = 560) | P value |
|---|---|---|---|
| Long-term aspirin during pregnancy, n (%) | 22 (7.9) | 30 (5.4) | .160 |
| Vasculoplacental complication, n (%) | |||
| Pregnancy-related hypertension | 4 (1.43) | 11 (2.0) | .690 |
| Preeclampsia | 10 (3.6) | 28 (5.0) | |
| Severe preeclampsia a | 4 (1.4) | 10 (1.8) | .700 |
| Composite vasculoplacental variable b | 21 (7.5) | 55 (9.8) | .269 |
| Abnormal uterine Doppler, n (%) c | 47 (16.8) | 70 (12.5) | .021 |
| Gestational age at diagnosis of vasculoplacental complications, wks (mean ± SD) | 31.2 ± 9.2 | 34.2 ± 5.0 | .080 |
| Pregnancy-related diabetes, n (%) | |||
| Dietary treatment only | 19 (6.8) | 43 (7.7) | .430 |
| Insulin | 5 (1.8) | 18 (3.2) | |
| Acute pregnancy-related cholestasis, n (%) | 4 (1.4) | 4 (0.7) | .310 |
| Threatened preterm delivery, n (%) | 34 (12.1) | 31 (5.5) | .001 |
| Fetal lung maturation by corticosteroid therapy, n (%) | 30 (10.7) | 38 (6.8) | .049 |
| Gestational age at delivery (wks), mean ± SD | 37.7 ± 3.2 | 39.0 ± 2.3 | < .001 |
| Preterm delivery, n (%) | |||
| <37 wks | 52 (18.6) | 45 (8.0) | < .001 |
| <32 wks | 14 (5.0) | 13 (2.3) | .040 |
| Induced premature birth | 10 (3.6) | 20 (3.6) | 1.00 |
| Mode of delivery, n (%) | |||
| Spontaneous vaginal delivery | 119 (42.5) | 377 (67.3) | < .001 |
| Instrumental vaginal delivery | 8 (2.9) | 35 (6.2) | |
| Cesarean delivery before labor | 121 (43.2) | 79 (14.1) | |
| Cesarean delivery during labor | 32 (11.4) | 69 (12.3) | |
| Perineal lesions among women with vaginal deliveries, n (%) | |||
| Episiotomy | 19/127 (15.0) | 135/412 (32.8) | < .001 |
| None (except episiotomy) | 87/127 (68.5) | 267/412 (64.8) | .65 0 |
| Simple tear | 37/127 (29.1) | 134/412 (32.5) | |
| Complete perineum, uncomplicated | 2/127 (1.6) | 10/412 (2.4) | |
| Complete perineum, complicated | 1/127 (0.8) | 1/412 (0.2) | |
| Child’s sex, n (%) | |||
| Female | 139 (49.6) | 297 (53.0) | .350 |
| Birthweight (mean ± SD) | |||
| Birthweight, g | 2902 ± 705 | 3201 ± 591 | < .001 |
| Birthweight, percentile | 41.6 ± 28.5 | 46.6 ± 28.6 | .020 |
| Less than third percentile, n (%) | 17 (6.1) | 21 (3.7) | .120 |
| Less than 10th percentile, n (%) | 45 (16.1) | 67 (12.0) | .090 |
| Head circumference (mean ± SD) | |||
| Head circumference, cm | 34.1 ± 2.4 | 34.4 ± 1.9 | .030 |
| Head circumference, percentile | 48.3 ± 30.8 | 52.2 ± 28.1 | .070 |
| 5 minute Apgar score <7, n (%) | 20 (7.1) | 14 (2.5) | .001 |
| Arterial pH at birth | |||
| <7.10, n (%) | 5 (1.8) | 22 (3.9) | .110 |
| Hospitalization of the child, n (%) d | 46 (16.4) | 66 (11.8) | .030 |
| Neonatal jaundice (phototherapy) , n (%) | 22 (7.9) | 71 (12.7) | .090 |
| Neonatal deaths, n (%) | 10 (3.6) | 6 (1.1) | .010 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
