Preeclampsia, classically characterized by the development of hypertension and proteinuria after 20 weeks gestation, affects approximately 3% of all pregnancies worldwide.1–3 Preeclampsia in its most severe form results in increased morbidity and mortality among women and their infants, particularly in low- and middle-income countries.4–6 Although the etiology of preeclampsia has yet to be fully elucidated, there is a body of research supporting an association between autoimmune cascade and preeclampsia.7–10 More recent research has demonstrated that alterations in the concentrations of circulating angiogenic factors are associated with preeclampsia.11–18
Preeclampsia research among HIV-infected women has been limited. HIV-infected, pregnant women in the era before the use of highly active antiretroviral therapy (HAART) were found to be less likely to develop preeclampsia than HIV-uninfected women.19,20 However, HIV-infected women receiving HAART seem to experience similar rates of preeclampsia as HIV-uninfected women,20–23 suggesting that HAART may restore risk of preeclampsia back to a preinfection baseline.
Access to HAART in pregnancy in low- and middle-income countries is increasing. As HAART in pregnancy is being scaled up in developing countries, it is often done in the setting of limited access to obstetric care, including limited ability to recognize and safely manage preeclampsia. Therefore, it is of public health importance to understand any association between HAART use in pregnancy and risk of preeclampsia, as well as the mechanism underlying such an association.
Using a cohort of HIV-1–infected, treatment-naive pregnant women in Botswana who initiated HAART in pregnancy, we sought to identify risk factors associated with preeclampsia. Using a smaller subset of our cohort, we explored levels of 2 angiogenic markers to determine if alterations in the concentration of the angiogenic markers mirrored the pattern observed in preeclamptic HIV-uninfected women. We also analyzed the change in levels of these angiogenic markers 1 month after HAART initiation among non-preeclamptic women.
The Mma Bana Study, which enrolled 730 HIV-1–infected, HAART-naive pregnant women, has been described in detail previously.24 Five hundred sixty participants who had CD4+ cell counts ≥200 cells per cubic millimeter and no AIDS-defining illnesses at enrollment, between 26 and 34 weeks gestation, were randomized to receive either abacavir/zidovudine/lamivudine co-formulated as Trizivir (GlaxoSmithKline, Greenford, United Kingdom) (TZV) twice daily or zidovudine/lamivudine co-formulated as Combivir (CBV) (GlaxoSmithKline) with lopinavir/ritonavir co-formulated as Kaletra (Abbott Laboratories, Abbott Park, IL) (CBV-KAL) twice daily. Another 170 women, who had a screening or previously documented CD4+ cell count of less than 200 cells per cubic millimeter, were enrolled in an observational arm of the Mma Bana study between 18 and 34 weeks gestation and monitored as they initiated CBV and nevirapine, the HAART regimen specified in Botswana's national HIV treatment guidelines. This study focuses on the 722 Mma Bana women who remained on study from enrollment through delivery.
A high rate of preeclampsia was identified while the Mma Bana study was ongoing and a sub-study was undertaken. We sought to evaluate changes in levels of placental growth factor (PlGF) and soluble FMS toll-like tyrosine kinase-1 (sFlt-1) 1 month after HAART initiation. Collection and storage of maternal plasma 1 month after HAART initiation was added in the latter half of the Mma Bana study. We obtained and stored maternal specimens until we had paired specimens (enrollment and 1 month after HAART initiation) from 60 sequentially enrolled women who had not yet delivered 1 month after HAART initiation and did not experience preeclampsia (controls). These paired specimens were analyzed to determine changes in PlGF and sFlt-1 levels after a month of HAART treatment. We also compared PlGF and sFlt-1 levels from the enrollment specimens of the 60 controls with enrollment specimens from the 11 women who experienced preeclampsia to ascertain whether HIV-infected women who later experienced preeclampsia had significantly lower levels of PlGF and higher levels of sFlt-1 before the onset of preeclampsia symptoms, a pattern observed among HIV-uninfected, preeclamptic women. Preeclamptic women were enrolled in the Mma Bana study between September 13, 2006, and April 10, 2008, whereas controls (non-preeclamptics) were enrolled between January 9, 2008, and May 12, 2008.
Preeclampsia was defined as the presence of hypertension, as evidenced by either a systolic pressure greater than 140 mm Hg or a diastolic pressure greater than 90 mm Hg, on 2 occasions (separated by at least 6 hours) and 1+ or greater urine protein via urine dip stick as documented during a study site visit or by a nonstudy health care provider. Study participants were evaluated by a study nurse and physician at enrollment, 2 weeks postenrollment, and monthly thereafter until delivery, unless pregnancy gestational age exceeded 38 weeks and then weekly until delivery. Blood pressure was measured, and urine was checked for the presence of protein at every antenatal visit. Blood pressure was checked using either automated electronic equipment or manual sphygmomanometers. Any woman with a systolic and/or diastolic blood pressure meeting the definition of preeclampsia with 1+ proteinuria or greater was referred to the local hospital for definitive management of preeclampsia by the study physician.
Gestational age was calculated from an algorithm that relied upon maternal-reported last menstrual period (LMP) and ultrasound performed before randomization. Where the estimated gestational age of a first trimester ultrasound differed by more than 1 week from the woman's reported LMP, a second trimester ultrasound differed by more than 2 weeks, or a third trimester ultrasound differed by more than 3 weeks, ultrasound dating was employed. Otherwise, the LMP was employed. Once the gestational age was ascertained using this algorithm, it was not altered later in the pregnancy.
The Botswana Health Research Development Committee and the Harvard School of Public Health Human Subjects Committee approved the collection of blood and its utilization for research purposes in the Mma Bana study. Women who enrolled in the Mma Bana study provided written informed consent for their blood samples to be used, and human subject research guidelines of Harvard School of Public Health were followed in the conduct of this clinical research.
Sample Collection and Preparation
Samples of maternal peripheral blood were obtained from HIV-infected women at enrollment and 1 month after HAART initiation in the parent Mma Bana study between 26 and 34 weeks gestation (for women in the randomized arms) or between 18 and 34 weeks gestation (for women in the observational arm). HAART was initiated at the enrollment visit and after the blood draw. The samples were processed, and the plasma was stored at −70°C, shipped frozen, and kept below −60°C until time of the assay.
Measurement of sFlt-1 and PlGF
Assays were performed by personnel blinded to the outcome of the pregnancy. PlGF and sFlt-1 were measured using prototype ARCHITECT immunoassays (Abbott Laboratories). For each woman, the 2 biomarkers were measured from the same-banked plasma sample aliquot.
The PlGF immunoassay measures the free form of PlGF-1. The assay has a lower limit of detection of 1 pg/mL with a calibration range up to 1500 pg/mL. A 20-day imprecision study using control and panel samples ranging from 12 to 1000 pg/mL resulted in within-run covariance from 1.4% to 6.7% and total covariance from 1.8% to 6.7%. The assay has shown acceptable dilution linearity up to 6000 pg/mL using a 4-fold dilution for samples initially reading greater than 1500 pg/mL.
The sFlt-1 immunoassay measures both free and bound sFlt-1. The assay has a lower limit of detection of 0.10 ng/mL with a calibration range up to 150 ng/mL. A 10-day imprecision study using control and panel samples ranging from 5 to 100 ng/mL resulted in within-run covariance from 1.8% to 3.0% and total covariance from 2.6% to 3.8%.
Analysis of risk factors associated with preeclampsia was conducted using univariate logistic regression. Age, gravida, enrollment CD4+ cell count (<200 vs ≥200 cells per cubic millimeter), pretreatment viral load (HIV-1 RNA ≥ 100,000 vs < 100,000 copies per milliliter), HAART regimen, and gestational age at HAART initiation were analyzed using data from all 722 women who had a recorded delivery during the Mma Bana study, including 11 women who developed preeclampsia. Among the subset of 71 women for whom angiogenic markers were assayed, including all 11 preeclamptic women, the association of PlGF and sFlt-1 and odds of preeclampsia were analyzed using univariate logistic regression. Variables with a P value ≤0.05 in univariate logistic regression were included in a multivariate logistic regression model using data from the same subset.
In the sub-study of angiogenic markers, characteristics of 60 non-preeclamptic women, for whom stored plasma specimens were available, were compared with all other women in the parent study who did not experience preeclampsia. Characteristics involving continuous and ordinal variables were compared using Wilcoxon rank sum test. Categorical variables were compared using Fisher exact tests. This approach was also employed to analyze differences between these 60 non-preeclamptic women and the 11 women who developed preeclampsia.
Wilcoxon rank sum test was used to compare sFlt-1 and PlGF levels at enrollment between non-preeclamptic women and preeclamptic women and to compare median changes in sFlt-1 and PlGF levels 1 month after HAART initiation among non-preeclamptic women randomized to TZV vs CBV-KAL, all of whom had enrollment CD4+ cell counts ≥200 cells per cubic millimeter. Moses confidence intervals (CIs) were calculated for the difference in median change in the 2 angiogenic markers among non-preeclamptic women randomized to TZV vs CBV-KAL. Wilcoxon signed rank test was used to analyze the change in sFlt-1 and PlGF 1 month after HAART initiation in non-preeclamptic women.
All testing used a 2-sided significance level of 0.05, with no correction for multiple testing. Statistical analyses were performed using SAS, version 9.3 (SAS Institute, Cary, NC).
Of 722 Mma Bana women who delivered, 11 women developed preeclampsia. Women who developed preeclampsia had a higher median enrollment log viral load compared with non-preeclamptic women (5.05 log10 vs 4.07 log10 copies per milliliter, P = 0.03) (Table 1).
Preeclamptic women also experienced significantly shorter duration of pregnancy, as would be expected, with a median gestational age at delivery of 35.1 weeks compared with 39.2 weeks among the 711 women who did not experience preeclampsia (P = 0.003). The proportion of women experiencing a stillbirth was significantly higher among preeclamptic women at 64% compared with 2% among non-preeclamptic women (P < 0.001).
Risk Factors Associated With Preeclampsia
Among the 722 Mma Bana study participants who experienced a delivery, only high viral load (≥100,000 copies per milliliter) was significantly associated with preeclampsia in univariate logistic regression analysis [odds ratio (OR) 5.8, 95% CI: 1.8 to 19.4, P = 0.004] (Table 2). There was a trend toward association between preeclampsia and CD4+ cell count less than 200 cells per cubic millimeter, as well as the type of HAART regimen initiated. However, neither reached statistical significance. High viral load remained significantly associated with preeclampsia in multivariate models adjusting separately for each of CD4+ cell count at enrollment, HAART regimen, and gestational age at HAART initiation.
Comparison of PlGF and sFlt-1 in Preeclamptic Women With Non-Preeclamptic Women
Pre-HAART P1GF and sFlt-1 levels were measured in all 11 women who developed preeclampsia and in a subset of 60 women who did not (see Methods). There was no significant difference between the characteristics of the 60 non-preeclamptic women in this subset and the 711 non-preeclamptic women in the overall study population with respect to median age, number of pregnancies, median enrollment CD4+ cell count and viral load, HAART treatment regimen, median gestational age at HAART initiation and at delivery, and other socioeconomic characteristics.
The median PlGF level at study enrollment, on the day of HAART initiation, was 130 pg/mL among the 11 women who later developed preeclampsia compared with 992 pg/mL for 60 women who did not experience preeclampsia (P = 0.001) (Fig. 1). The median sFlt-1 level at study enrollment for women who later developed preeclampsia was 17.5 pg/mL compared with 9.4 pg/mL for women who did not experience preeclampsia (P = 0.03) (Fig. 1).
Analyzing the subset of 71 women using univariate logistic regression, viral load ≥100,000 copies per milliliter was a significant predictor of preeclampsia as found in the overall study population. In univariate analyses of the association of preeclampsia with angiogenic markers, every 100 pg/mL decrease in PlGF was associated with a 34% increased odds of preeclampsia (OR: 1.34, 95% CI: 1.13 to 1.69, P = 0.004), whereas every 2 pg/mL increase in sFlt-1 was associated with a 15% increased odds of preeclampsia (OR: 1.15, 95% CI: 1.03 to 1.28, P = 0.02). In multivariate logistic regression, only PlGF (adjusted odds ratio: 1.28 per 100 pg/mL decrease, 95% CI: 1.05 to 1.66, P = 0.04) and enrollment viral load of ≥100,000 copies per milliliter (adjusted odds ratio: 7.15, 95% CI: 1.35 to 45.01, P = 0.02) remained significantly associated with preeclampsia after adjustment for sFlt-1 (which was not significant, P = 0.18).
Non-Preeclamptic Women—Changes in PlGF and sFlt-1 1 Month After HAART Initiation
Among the 53 non-preeclamptic women from whom PlGF assay results were available at enrollment and 1 month after HAART initiation, the median change in PlGF was 134 pg/mL [interquartile range (IQR): −377 to 478 pg/mL, P = 0.56 for Wilcoxon signed rank test]. The median change in sFlt-1 for the 49 non-preeclamptic women for whom sFlt-1 assay results were available at enrollment and 1 month after HAART initiation was −0.43 pg/mL (IQR: −2.60 to 1.84 pg/mL, P = 0.32 for Wilcoxon signed rank test).
To determine whether randomized antiretroviral regimens were associated with changes in angiogenic markers, we analyzed median change in PlGF and sFlt-1 1 month after HAART initiation among non-preeclamptic women with enrollment CD4 cell counts ≥200 cells per cubic millimeter at or before enrollment who were randomized to TZV or CBV-KAL. The median change in PlGF 1 month after HAART initiation was −92 pg/mL (IQR: −440 to 548 pg/mL) among women randomized to TZV compared with 80 pg/mL (IQR: −386 to 211 pg/mL) for women randomized to CBV-KAL (difference in median change: 172 pg/mL, 95% CI: −403 to 336, P = 0.82). The median change in sFlt-1 1 month after HAART initiation was −0.43 pg/mL (IQR: −2.33 to 4.09 pg/mL) for non-preeclamptic women randomized to TZV compared with −0.30 pg/mL (IQR: −2.75 to 1.85 pg/mL) for non-preeclamptic women randomized to CBV-KAL (difference in median change: 0.13 pg/mL, 95% CI: −4.4 to 2.5, P = 0.85).
Among 722 HIV-1–infected women initiating HAART in pregnancy who delivered in the Mma Bana study, viral load ≥100,000 copies per milliliter at enrollment was significantly associated with increased risk of preeclampsia. Angiogenic markers known to be associated with preeclampsia in non–HIV-infected women were similarly associated with preeclampsia in a subset of Mma Bana study participants. Angiogenic marker association was notable before the onset of maternal hypertension and before HAART initiation. For non-preeclamptic women, angiogenic markers did not change significantly during the first month after HAART initiation.
This is the first study, of which we are aware, describing the association between higher pretreatment viral load and preeclampsia among women initiating HAART in pregnancy. In 2010, the World Health Organization revised prevention of mother-to-child HIV transmission guidelines to advocate initiation of HAART as early as the 14th week of gestation for women diagnosed with HIV in pregnancy who do not require HAART for their own health.25 Whether early initiation of HAART resulting in undetectable viral load by the second half of pregnancy reduces the risk of preeclampsia in women with viral load ≥100,000 copies per milliliter warrants further investigation. Onset of preeclampsia usually occurs in the late second trimester or third trimester of pregnancy. The earliest onset of preeclampsia in our cohort occurred at 26 weeks gestational age. If HAART is found to mitigate this risk, research will also be needed to ascertain the optimal regimen to achieve rapid viral load reduction while ensuring safety to the mother and fetus.
This is also the first study to evaluate angiogenic markers associated with preeclampsia in a cohort of HIV-infected women initiating HAART in pregnancy. The results demonstrate that women who subsequently developed preeclampsia had significantly elevated sFlt-1 and decreased PlGF before HAART initiation and antecedent to the combined end point of hypertension and proteinuria during the pregnancy. The biomarker patterns observed in our population of HIV-infected women correspond to patterns reported in cross-sectional studies of otherwise healthy, HIV-uninfected, nulliparous women who subsequently developed preeclampsia,11,26,27 suggesting a common mechanism for both HIV-infected and HIV-uninfected women.
Data from prior studies have reported conflicting findings regarding HAART in pregnancy and its association with preeclampsia. Suy et al21 found increased risk of preeclampsia in the HAART era in their cohort of women in Barcelona, composed predominantly of Caucasians who were on HAART before conception. HAART use in pregnancy was also associated with increased preeclampsia compared with pregnancies of HIV-infected mothers not receiving HAART in a study conducted by Wimalasundera et al20 in the United Kingdom. In these cohorts, the rate of preeclampsia among HIV-infected women on HAART did not differ significantly from that of HIV-uninfected controls, leading the authors to conclude that restoration of the immune system may play a pivotal role in restoring the risk of preeclampsia.20,21 However, Haeri et al28 found preeclampsia to be less common among HIV-infected women on HAART compared with HIV-uninfected women (6% vs 12%, P = 0.04) in a cohort of women in the United States.
The introduction of HAART among non-preeclamptic women did not significantly change the angiogenic marker levels after 1 month into HAART treatment. Additionally, we did not detect any significant difference in sFlt-1 or PlGF 1 month after HAART initiation among non-preeclamptic women with enrollment CD4 cell counts ≥200 cells per cubic millimeter when comparing a triple nucleoside reverse transcriptase inhibitor regimen (TZV) with a protease inhibitor–based regimen (CBV-KAL). These findings should be regarded with caution as they may not have been sufficiently powered to detect differences. Larger studies are needed to determine if specific HAART regimens, initiated during pregnancy, influence preeclampsia angiogenic markers in a significantly different manner. If HAART is associated with preeclampsia, the mechanism is not likely to cause direct adverse shifts in angiogenic markers. Rather, a restoration of a deeper underlying risk is more likely,20 with the angiogenic markers remaining useful indicators of risk among the subset who later become preeclamptic.
Whereas our study focused on risk factors of preeclampsia in HIV-infected, pregnant women, the proportion of preeclamptic women experiencing stillbirth at 64% is alarming and warrants comment, given the overall proportion of stillbirths was 2% in the study. Preeclamptic women in the Mma Bana study were appropriately referred to a local hospital in a timely manner by study personnel. The high rate of stillbirths reflects health care challenges faced in resource-limited settings. World Health Organization has set a goal for eliminating new pediatric HIV infections worldwide by 2015.29 Yet, the stillborn rate experienced by our preeclamptic women highlights the critical need for comprehensive and integrated health services in resource-limited settings for HIV-infected women and their infants, including emergency obstetric and neonatal care. Comprehensive health care will not only contribute to the elimination of pediatric HIV infections worldwide but also optimize maternal–child health.
Our study had some limitations. Because women with baseline CD4+ cell counts less than 200 cells per cubic millimeter were eligible for nevirapine-based HAART initiation as early as the 18th week of gestation, whereas women with CD4+ cell counts ≥200 cells per cubic millimeter initiated CBV-KAL or TZV after the 26th week of gestation, the Mma Bana trial design introduced potential confounding among maternal baseline CD4+ cell count, baseline viral load, gestational age at HAART initiation, and HAART treatment regimen. We separately controlled for CD4 cell count, HAART regimen, and gestational age at HAART initiation with viral load and did not detect confounding or collinearity. However, future preeclampsia studies controlling for these possible confounders are needed.
We lacked serial measurements of sFlt-1, PlGF, or CD4+ cell counts. This precludes analysis of the contribution of HAART to the outcome of preeclampsia among women who had angiogenic marker profiles consistent with the evolution of preeclampsia. Further studies characterizing baseline sFlt-1 and PlGF before HAART initiation and serial changes in both angiogenic biomarkers and CD4+ cell counts thereafter would allow for quantitative description of HAART's contribution to these markers.
In conclusion, women enrolling in the Mma Bana cohort with a baseline viral load ≥100,000 copies per milliliter faced 7-fold higher risk of preeclampsia, after adjusting for 2 angiogenic markers predictive of preeclampsia. Data from existing large observational studies should be used to validate the association between high viral load and increased risk of preeclampsia. If the association is noted in other studies, it would be of public health importance to determine whether viral load reduction earlier in pregnancy mitigates the risk of preeclampsia. Our results demonstrate that sFlt-1 levels are elevated and PlGF levels are low antecedent to the onset of preeclampsia among HIV-infected women, in a manner comparable with prior studies involving HIV-uninfected women. Among women who did not develop preeclampsia, the analysis did not detect a statistically significant change in the levels of PlGF and sFlt-1 from baseline to 1 month after HAART initiation. Considering the conflicting findings regarding HAART's association with preeclampsia, larger studies are needed to determine the potential mechanism for preeclampsia risk restoration among women receiving HAART. Finally, the striking rate of stillbirths in our cohort of preeclamptic women speaks to the importance of proactively managing preeclampsia to ensure optimal maternal and infant outcomes.
The authors are indebted to the women and infants who participated in the Mma Bana study; the Mma Bana study team staff; the administration and staff at Scottish Livingston, Deborah Retief Memorial, Athlone Hospital, and Princess Marina Hospital; and the staff at the referring health clinics. The authors are grateful to Abbott Pharmaceuticals, GlaxoSmithKline, and the government of Botswana for the provision of study drugs. The authors wish to specifically thank Botswana Harvard AIDS Institute Partnership and Harvard School of Public Health Staff: Lillian Makori, Gloria Mayondi, Agnes Modise, Venice Modikwa, Ria Madison, Tlhongbotho Masoloko, Daisy Ramalekane, Molly Pretorius Holme, Heather Carey, Sara Mazzola, Carrie Kachoria, Raabya Rossenkahn, Vlad Novitsky, Chris Rowley, Jennifer Boyle, Michael Roy, Lendsey Melton, Chikezie Nwankwo, Scott Dryden Peterson, Onalenna Nthase, Norah Mawoko, Elias Woldegabriel, Kasonde Micheal, Chandan Harikrishnan, Jane Magetse, Joyce Lubinda, Tebogo Kakhu, Thena Tumediso, Modiegi Diseko, Mosetsanagape Galekhutle, Keamogetse Rebatenne, Mavis Moeng, Kebaibphe Ntalabgwe, Ditlamelo Mareme, Victoria Kgwadi, Kaone Kgati, Keitumetse Sakana, Best Mafoko, Lazarus Moremi, Jimmy Nkgau, Ilori Adewale, Banno Janet Moorad, Dipotso Arbi, Kesego Dudu Kooreng, Selebaleng Vinoliah Simon, Maggie Mosetsanagape Nkgau, Collen Rananna, Rejoice Molefe, Nametso Dimpho Lekwape, Tebogo Ncube, Eldah Kakanyo Tshotlego, Segomotso Mapote, Radinku Tshegofatso, Emmanuel Keikotlhae, William Keboutlwe, Hanqiwe Olebeng, Seleetso Ndicky Modibedi, Tshepo Silwane, Tshepiso Patricia Morupisis, Ntsholeng Kekgethile, Sydney Kgwefane, Julius Kgangetsile, Nnahurumnanya Iwe, Tseo Khudube, Malebogo Ntshimane, Maureen Gower, Nthabiseng Kgaodi, Kate Selathwe, Lorraine Phiri, Rosemary Musonda, Phillimon Segopodi, Dorcas Moses, Terence Mohammed, Dineo Mongwato, Bonolo Khumotaka, Phibeon Munyaradzi Mangwendeza, Chishamiso Mudenyanga, Matshediso Zachariah, Gertrude Ditshotlo, Alex Ntau, and Poko Molwane; Botswana Ministry of Health: Khumo Seipone, Shenaaz El Halabi, Pilate Khulumani, Mary Kasule, Madisa Mine, Kgomotso Makhaola, Howard Moffat, Haruna Baba Jibril, and the prevention of mother-to-child HIV transmission unit; Princess Marina Hospital, Gaborone: Staff of Maternity, Postnatal, and Children's ward; Scottish Livingstone Hospital, Molepolole: Staff of Maternity, Postnatal, and Children's ward; Deborah Retief Memorial Hospital, Mochudi: Staff of Maternity, Postnatal, and Children's ward; Athlone Hospital Lobatse: Staff of Maternity, Postnatal, and Children's ward; District Health Teams (Molepolole and Mochudi). City Council Clinics team (Lobatse and Gaborone); GlaxoSmithKline: Edde Loeliger; Baylor University: Gabriel Anabwani and Elizabeth Lowenthal; Beth Israel Deaconess Medical Center: Linda Shipton; Harvard Medical School: Jennifer Chen; Oxford University: Philip Goulder and Philippa Mathews; National Institutes of Health: Lynne Mofenson; and the DSMB members and members of the Community Advisory Board.
2. Villar J, Say L, Shennan A, et al.. Methodological and technical issues related to the diagnosis, screening, prevention, and treatment of preeclampsia and eclampsia. Int J Gynecol Obstet. 2004;85(suppl 1):S28–S41.
3. Conde-Agudelo A, Villar J, Lindheimer M. Maternal infection and risk of preeclampsia: systematic review and metaanalysis. Am J Obstet Gynecol. 2008;198:7–22.
4. Firoz T, Sanghvi H, Merialdi M, et al.. Preeclampsia in low and middle income countries. Best Pract Res Clin Obstet Gynaecol. 2011;25:537–548.
5. Khan KS, Wojdyla D, Say L, et al.. WHO analysis of causes of maternal death: a systemic review. Lancet. 2006;367:1066–1074.
6. Osungbade KO, Ige OK. Public health perspectives of preeclampsia in developing countries: implications for health system strengthening. J Pregnancy. 2011;2011:481095.
7. Borchers AT, Naguwa SM, Keen CL, et al.. The implications of autoimmunity and pregnancy. J Autoimmun. 2010;34:J287–J299.
8. Siddiqui AH, Irani RA, Blackwell SC, et al.. Angiotensin receptor agonistic autoantibody is highly prevalent in preeclampsia: correlation with disease severity. Hypertension. 2010;55:386–393.
9. Xia Y, Kellems RE. Receptor-activating autoantibodies and disease: preeclampsia and beyond. Expert Rev Clin Immunol. 2011;7:659–674.
10. Wallukat G, Neichel D, Nissen E, et al.. Agonistic autoantibodies directed against the angiotensin II AT1
receptor in patients with preeclampsia. Can J Physiol Pharmacol. 2003;81:79–83.
11. Levine RJ, Maynard SE, Qian C, et al.. Circulating angiogenic factors and the risk of preeclampsia. N Engl J Med. 2004;350:672–683.
12. Vaisbuch E, Whitty JE, Hassan SS, et al.. Circulating angiogenic and antiangiogenic factors in women with eclampsia. Am J Obstet Gynecol. 2011;204:152e1–152e9.
13. Masuyama H, Suwaki N, Nakatsukasa H, et al.. Circulating angiogenic factors in preeclampsia, gestational proteinuria, and preeclampsia superimposed on chronic glomerulonephritis. Am J Obstet Gynecol. 2006;194:551–556.
14. Robinson CJ, Johnson DD, Chang EY, et al.. Evaluation of placental growth factor and soluble Fms-like tyrosine kinase 1 receptor levels in mild and severe preeclampsia. Am J Obstet Gynecol. 2006;195:255–259.
15. Vatten LJ, Eskild A, Nilsen TIL, et al.. Changes in circulating level of angiogenic factors from the first to second trimester as predictors of preeclampsia. Am J Obstet Gynecol. 2007;196:239.e1–239.e6.
16. Yelumalai S, Muniandy S, Omar SZ, et al.. Pregnancy-induced hypertension and preeclampsia: levels of angiogenic factors in Malaysian women. J Clin Biochem Nutr. 2010;47:191–197.
17. Powers RW, Jeyabalan A, Clifton RG, et al.. Soluble fms-like tyrosine kinase 1 (sFlt-1), endoglin and placental growth factor (PlGF) in preeclampsia among high risk pregnancies. PLoS One. 2010;5:e13263.
18. Moore Simas TA, Crawford SL, Solitro MJ, et al.. Angiogenic factors for the prediction of preeclampsia in high-risk women. Am J Obstet Gynecol. 2007;197:244.e1–244.e8.
19. Stratton P, Tuomala RE, Abboud R, et al.. Obstetric and newborn outcomes in a cohort of HIV-infected pregnant women: a report of the women and infants transmission study. J Acquir Immune Defic Syndr Hum Retrovirol. 1999;20:179–186.
20. Wimalasundera RC, Larbalestier N, Smith JH, et al.. Preeclampsia, antiretroviral therapy, and immune reconstitution. Lancet. 2002;360:1152–1154.
21. Suy A, Martinez E, Coll O, et al.. Increased risk of preeclampsia and fetal death in HIV-infected pregnant women receiving highly active antiretroviral therapy. AIDS. 2006;20:59–66.
22. Martin F, Taylor GP. The safety of highly active antiretroviral therapy for HIV-positive pregnant mother and her baby: is “the more the merrier”? J Antimicrob Chemother. 2009;64:895–900.
23. Parekh N, Ribaudo H, Souda S, et al.. Risk factors for very preterm delivery and delivery of very-small-for-gestational-age infants among HIV-exposed and HIV-unexposed infants in Botswana. Int J Gynecol Obstet. 2011;115:20–25.
24. Shapiro R, Hughes M, Ogwu A, et al.. Antiretroviral regimens in pregnancy and breast-feeding in Botswana. N Engl J Med. 2010;362:2282–2294.
26. Levine RJ, Lam C, Qian C, et al.. Soluble endoglin and other circulating antiangiogenic factors in preeclampsia. N Engl J Med. 2006;355:992–1005.
27. McKeeman GC, Ardill JE, Caldwell CM, et al.. Soluble vascular endothelial growth factor receptor-1 (sflt-1) is increased throughout gestation in patients who have preeclampsia develop. Am J Obstet Gynecol. 2004;191:1240–1246.
28. Haeri S, Shauer M, Dale M, et al.. Obstetric and newborn infant outcomes in human immunodeficiency virus-infected women who receive highly active antiretroviral therapy. Am J Obstet Gynecol. 2009;201:315.e1–315.e5.