International Journal of Gynecology and Obstetrics 125 (2014) 146–149

Contents lists available at ScienceDirect

International Journal of Gynecology and Obstetrics

j ourna l homepage: www.e lsev ie r .com/ locate / i jgo

CLINICAL ARTICLE

Invasive diagnostic procedures and risk of hypertensive disordersin pregnancy

George Daskalakis a,⁎, Angeliki Papapanagiotou b, Nikolaos Antonakopoulos a, Spyros Mesogitis a,Nikolaos Papantoniou a, Dimitrios Loutradis a, Aris Antsaklis a

a First Department of Obstetrics and Gynecology, University of Athens Medical School, Alexandra Maternity Hospital, Athens, Greeceb Department of Clinical Biochemistry, University of Athens Medical School, Athens, Greece

⁎ Corresponding author at: 8 I.Metaxa Street and 1 Vas.Athens, Greece. Tel.: +30 210 5618001; fax: +30 210 53

E-mail address: gdaskalakis@yahoo.com (G. Daskalaki

0020-7292/$ – see front matter © 2014 International Fedehttp://dx.doi.org/10.1016/j.ijgo.2013.10.015

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 3 July 2013Received in revised form 16 October 2013Accepted 15 January 2014

Keywords:AmniocentesisChorionic villus samplingPre-eclampsiaPregnancy induced hypertension

Objective: To determine whether the risk of hypertensive complications differs among low-risk womenwho un-dergo prenatal diagnosis via chorionic villus sampling (CVS) and amniocentesis. Methods: In a retrospectivestudy, data were analyzed from women who underwent prenatal diagnosis by CVS or amniocentesis atAlexandra Maternity Hospital, Athens, Greece, between 1998 and 2011. All women had either transabdominalCVS at 10–13 weeks of pregnancy with a 20-gauge needle, or amniocentesis at 17–21 weeks with a 22-gaugeneedle, both under direct ultrasound guidance. Only women who had cytogenetically normal pregnancies anddelivered at the study hospital were included. Themain outcomemeasurewas the development of hypertensivecomplications. Results: Overall, 3243 womenwho underwent CVS and 6875 woman who underwent amniocen-tesis met the inclusion criteria, and their outcomeswere analyzed. In total, 237 women (2.3%) developed hyper-

tensive disorders during their pregnancy. The incidence of pre-eclampsia (2.4% vs 0.8%) and total hypertensivedisorders (3.8% vs 1.7%) was significantly higher (P b 0.001) in the CVS group than in the amniocentesisgroup. Conclusion:Womenwho underwent CVS had a significantly higher risk of developing hypertensive disor-ders in comparison to those who underwent amniocentesis. This finding warrants further investigation via awell-designed prospective randomized trial. © 2014 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Most pregnant women have a screening test for chromosomal de-fects in the first trimester via the measurement of nuchal translucencyand biochemical markers. For most womenwho have abnormal results,the diagnostic procedure of choice is chorionic villus sampling (CVS).With the widespread use of first-trimester aneuploidy screening,the demand for CVS is increasing in everyday practice. The diagnosticaccuracy of the method is approximately 99%, which is similar to thatof amniocentesis. However, CVSprovides the opportunity for a safer ter-mination of pregnancy and, at the same time, it values the patient’s pri-vacy because the results are available in the first trimester. Althoughthere is extensive literature on the risk of pregnancy loss after an inva-sive diagnostic test [1,2], data on other complications arising from theseprocedures are scarce. Such information is extremely important for pa-tients who seek prenatal diagnosis.

A few studies have shown a relationship between CVS and increasedrisk of hypertensive disorders in pregnancy [3–5]. The underlying path-ophysiology of pre-eclampsia is not completely understood. Amongthe plausible theories, a dysfunction of the placenta is thought to be

Sophias Street, P. Penteli, 1523617224.s).

ration of Gynecology and Obstetrics. P

responsible for the occurrence of the complication. Impaired placentalformation at the beginning of pregnancy is thought to lead to hyperten-sive disorders later in the third trimester. Studies supporting the relation-ship of CVS to hypertensive complications are based on the assumptionthat focal placental disruption during the procedure increases the riskof pre-eclampsia.

In 2005, Silver et al. [3] reported that CVS performed in the late firsttrimester was associated with a higher rate of hypertensive disorders ofpregnancy compared with early amniocentesis. The same associationwas also observed among nulliparous women who underwent CVS[4]. Moreover, it was reported that women who went on to developpre-eclampsia after CVS hadhigher levels ofmaternal serumα-feto pro-tein (MSAFP) and pregnancy-associated placental protein-A (PAPP-A)after the procedure compared with those who did not develop thiscomplication [5]. However, subsequent studies failed to find any asso-ciation between CVS and increased risk of hypertensive disorders ofpregnancy [6–9].

The various methodologies used in the studies—mainly the dif-ferent control groups and the failure to adjust for several maternalcharacteristics—is likely to have led to the conflicting results. Until aprospective well-designed trial aimed at investigating the potential re-lationship between CVS and hypertensive disorders gives a definitiveanswer to this question, it remains useful to report single-center expe-rience. Therefore, the aim of the present study was to examinewhether

ublished by Elsevier Ireland Ltd. All rights reserved.

Table 1Characteristics of the study population by procedure.a

CVS(n = 3243)

Amniocentesis(n = 6875)

P value

Age, y 30.0 ± 5.9 35.5 ± 4.5 b0.001b

Gestational age at procedure, wk 11.3 (1.1) 17.6 (1.4) b0.001b

Primigravida 1461 (45.1) 2819 (41.0) b0.001c

Smoker 196 (6) 446 (6.5) 0.393c

BMI 28.3 ± 4.8 28.4 ± 5.1 0.348c

Preterm delivery (b37 wk) 417 (12.9) 928 (13.5) 0.376c

Birth weight, g 3280 (3000–3560) 3300 (3000–3600) 0.240d

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by thesquare of height in meters); CVS, chorionic villus sampling.

a Values are given as mean ± SD, number (percentage), or mean (interquartile range).b By Student t test.c By χ2 test.d By Mann–Whitney test.

Table 2Comparison of study variables between chorionic villus sampling and amniocentesisgroups.a

Variable CVS(n = 3243)

Amniocentesis(n = 6875)

P valueb

Pre-eclampsiaNo 3165 (97.6) 6822 (99.2) b0.001Yes 78 (2.4) 53 (0.8)

Gestational hypertensionNo 3199 (98.64) 6813 (99.1) 0.036Yes 44 (1.36) 62 (0.9)

Pre-eclampsia or gestational hypertensionNo 3121 (96.2) 6760 (98.3) b0.001Yes 122 (3.8) 115 (1.7)

Abbreviation: CVS, chronic villus sampling.a Values are given as number (percentage).b By χ2 test.

147G. Daskalakis et al. / International Journal of Gynecology and Obstetrics 125 (2014) 146–149

CVS entails a different risk of hypertensive complications comparedwith second-trimester genetic amniocentesis among low-risk womenattending a tertiary care center for prenatal diagnosis.

2. Materials and methods

In a retrospective study, data were reviewed from pregnant womenwho underwent CVS or amniocentesis at the Division of Feto-MaternalMedicine at Alexandra Hospital, Athens University, in Athens, Greece,between January 1, 1998, and December 31, 2011. Approval for thestudy was obtained from the ethics committee of the hospital, and allwomen gave signed consent before undergoing the procedure.

All study women had either transabdominal CVS at 10–13 weeks ofgestation with a 20-gauge needle, or an amniocentesis at 17–21 weekswith a 22-gauge needle. Each procedure was performed under ultraso-nographic visualization.

Maternal demographics, past obstetric history, past medical history,indication for the procedure, ultrasonographic findings, placenta site,number of attempts, needle insertion, and pregnancy outcomewere re-corded in the hospital’s computerized database. Only cytogeneticallynormal pregnancies of women who delivered in the study hospitalwere included in the analysis. Exclusion criteria were the presence ofa multiple pregnancy, known congenital abnormalities, suspected con-fined placental mosaicism, chronic hypertension, pregestational diabe-tes mellitus, chronic renal disease, autoimmune disorders, inheritedthrombophilia, and antiphospholipid antibody syndrome. Moreover,womenwho underwent amniocentesis and had a transplacental needleinsertion, and thosewhounderwent a repeated procedure (CVS, amnio-centesis, or both) were also excluded from the study.

Pre-eclampsia was defined as a blood pressure of 140/90 mm Hgor higher, as confirmed in 2 readings in 4–6 hours, and proteinuriaof 300 mg or more in a 24-hour urine specimen after 20 weeks ofgestation. Gestational hypertension was defined as a blood pressure of140/90 mmHgor higher, as confirmed in 2 readings in 4–6 hours, with-out any other systemic features of pre-eclampsia after 20 weeks ofgestation. Women who were diagnosed with hypertension before20 weeks were classified as having chronic hypertension and wereexcluded from the study.

Over the study period, there were no changes in clinical obstetricpractice at the study hospital that might confound the relationship be-tween exposure and outcome. Moreover, the same operator team, usingthe same technique, performed all of the invasive diagnostic procedures.

Statistical analyses were carried out with SPSS version 17.0 (IBM,Armonk, NY, USA). Continuous variables are described as the mean ±SD ormedian (interquartile range). Quantitative variables are expressedby the absolute frequency (percentage). Percentageswere compared viathe χ2 test. Parametric Student t test was used to compare 2means if thedistribution was approximately normal; the non-parametric Mann–Whitney test was used if the normality assumption was not satisfied.

Multiple logistic regression analysis was used to assess the associa-tion of theprocedure (CVS vs amniocentesis)with pre-eclampsia, gesta-tional hypertension, or pre-eclampsia with gestational hypertensiontogether after adjusting for maternal age, primigravidity, smoking, andbody mass index (BMI, calculated as weight in kilograms divided bythe square of height inmeters). The adjusted odds ratio (95% confidenceinterval) is presented from the results of the regression analysis. All Pvalues presented from the analyses are 2-tailed and significance wasset at a level of 0.05.

3. Results

During the study period, 3243 women who underwent CVS and6875 who underwent amniocentesis met the inclusion criteria andwere included in the analysis. The demographics of the 2 study groupsare presented in Table 1. All women were white, but women in the am-niocentesis group were significantly older.

Of thewomenwhounderwent CVS, 2694 (83.1%) had the procedurebecause both parents had the β-thalassemia trait; the remaining 549women had it for fetal karyotyping. Regarding amniocentesis, mostwomen (6378; 92.8%) had the procedure for fetal karyotyping mainlyowing to advanced maternal age (N35 years), 401 (5.8%) had it forDNA analysis, and 96 (1.4%) had it for PCR studies.

Table 2 shows the pregnancy outcome of both groups. In total, 237(2.3%) developed hypertensive disorders during their pregnancy. Therate of pre-eclampsia and that of total hypertensive complications weresignificantly higher in the CVS group than in the amniocentesis group.

Multiple logistic regression analysis showed that women in the CVSgroup had a 2.89-fold greater likelihood of having pre-eclampsia com-pared with women in the amniocentesis group (Table 3). Similarly,the same group of women had 1.50-fold greater likelihood of havinggestational hypertension and 2.21-fold greater likelihood of havingpre-eclampsia or gestational hypertension compared with women inthe amniocentesis group.

4. Discussion

The present found that women who underwent CVS had a signifi-cantly higher risk of developing hypertensive disorders during theirpregnancy compared with those who underwent amniocentesis. Therisk was approximately 2.2 times higher for any hypertensive disorderand approximately 2.9 times higher for pre-eclampsia. Some studieshave reported similar results in the past [3–5], although others notonly failed to find any relationship between CVS and pre-eclampsia,but also observed a protective effect [6–9]. In the absence of a prospec-tive randomized trial with a primary outcome to measure the risk of

Table 3Odds ratios for the likelihood of developing hypertensive disorders in the CVS group com-pared with the amniocentesis group.

Hypertensive disorder OR (95% CI)a P value

Pre-eclampsia 2.89 (1.74–4.81) b0.001Gestational hypertension 1.50 (1.01–2.21) 0.031Pre-eclampsia or gestational hypertension 2.21 (1.53–3.20) b0.001

Abbreviations: CI, confidence interval; CVS, chorionic villus sampling; OR, odds ratio.a Adjusted for maternal age, primigravidity smoking, and body mass index.

148 G. Daskalakis et al. / International Journal of Gynecology and Obstetrics 125 (2014) 146–149

hypertensive disorders after an invasive diagnostic procedure, no defi-nite conclusion can be drawn about this.

Silver et al. [3] were the first to report a higher rate of hypertensivecomplications among women who underwent late CVS than amongthose who had early amniocentesis [3]. They also observed a relation-ship between the degree of placental damage and the risk of these com-plications. However, there were several biases in their study such asthe possibility of misclassification of pre-eclampsia, the non-standardgestational age at which the procedures were carried out (late CVSvs early amniocentesis), and the possibility of placental disruptionamong women who underwent amniocentesis.

Grobman et al. [4] reported that there was a higher incidence of pre-eclampsia after CVS only among nulliparous women, whereas Farinaet al. [5] observed an association between the levels of MSAFP andPAPP-A after CVS and the risk of pre-eclampsia. Adusumalli et al. [6]did not find that CVS increases the risk of hypertensive disorders,although they observed a higher combined risk of the most severeforms of pre-eclampsia and gestational hypertension [6]. By contrast,2 recent large studies suggested that there is no increase in the risk ofhypertensive complications after CVS [7,8]. Notable, in one of thesestudies, women who underwent CVS had a lower incidence of hyper-tensive complications comparedwith thosewhohad no invasive testing[8]. Similarly, in a retrospective analysis from the SwedishMedical BirthRegister, no increased risk of pre-eclampsia or gestational hypertensionwas observed among women undergoing CVS [9].

The results of the present study support the theory that first-trimester placental disruption after CVS can subsequently cause abnor-mal placental function, leading to the development of pre-eclampsia orgestational hypertension. Although there is no single unifying theory toexplain the pathophysiology of pre-eclampsia, impaired placentationseems to be responsible for this complication. Supportive evidencefor this includes increasedfirst-trimester Doppler indices for the uterinearteries among women at high risk of developing pre-eclampsia, andthe resolution of pre-eclampsia after delivery of the fetus and theplacenta [10–14].

Chorionic villus sampling might theoretically cause damage to pla-centation and induce inflammation and focal hemorrhage, whichcould lead to reduced placental perfusion owing to impaired placentalvasculature and endothelial dysfunction [3]. This is consistent with thefinding that women with pre-eclampsia have abnormal placental im-plantation, leading to placental dysfunction [15,16]. An alternativetheory suggests that an abnormal immune response of the motherafter an exposure to paternally derived fetal antigens is responsiblefor the occurrence of pre-eclampsia [17,18]. Some studies haveshown increased α-fetoprotein levels, or the presence of fetal cellsin the maternal blood, indicating altered fetal–maternal transfer afterprenatal diagnostic procedures [19,20]. It has also been proposedthat a disruption of the placenta in the first trimester might lead to in-creased production of angiogenic growth factors—such as placentalgrowth factor, vascular endothelial growth factor, and soluble fms-like tyrosine kinase 1—that predispose to the development of pre-eclampsia [21,22].

Among the various studies reporting the risk of hypertensive disor-ders after CVS, there are significant differences not only in their designbut also in their sample size, which might account for the conflicting

results. The main advantage of the present study is its large samplesize and its homogenous population. All women were white and deliv-ered in the study hospital; as a result, the same standardized definitionsof hypertensive complications were used. In addition, women whounderwent transplacental amniocentesis and thosewho had a repeatedinvasive procedure were excluded from analysis.

Although the incidence of hypertensive disorders was significantlyhigher in the CVS group than in the amniocentesis group (3.8% vs1.7%) in the present study, it was well within the normal reportedrange (b6%). This is mainly due to 2 factors. First, women at greatestrisk of this condition, such as those with chronic hypertension, pre-gestational diabetes mellitus, chronic renal disease, autoimmune disor-ders, inherited thrombophilia, and antiphospholipid antibody syn-drome, were excluded. Second, most CVS procedures were performedfor women with the β-thalassemia trait who are relatively anemic,and it has been observed that nulliparous women with a low hemoglo-bin concentration have a lower incidence of hypertension comparedwith non-anemic control women [23]. This has been regarded as a com-pensative mechanism for better oxygen transfer to the fetus owing tothe observation of increased trophoblastic invasion and relatively de-creased trophoblast apoptosis in anemic women [24,25]. This observa-tion led to the conclusion that anemia has a protective effect againsthypertensive disorders during pregnancy [24].

The present study is not without limitations, mainly owing to itsretrospective design. The finding of an increased risk of hypertensivecomplications after CVS cannot be considered definitive, but highlightsthe need for further investigation. A prospective randomized well-designed study, which should take into account maternal age, BMI,parity, placenta location, number of sampling passes, and the quantityof placental tissue removed after the procedure, is necessary to drawsafe conclusions about a possible relationship between CVS and hyper-tensive complications in pregnancy.

Conflict of interest

The authors have no conflicts of interest.

References

[1] Mujezinovic F, Alfirevic Z. Procedure-related complications of amniocentesis and cho-rionic villous sampling: a systematic review. Obstet Gynecol 2007;110(3):687–94.

[2] Tabor A, Vestergaard CH, Lidegaard Ø. Fetal loss rate after chorionic villus samplingand amniocentesis: an 11-year national registry study. Ultrasound Obstet Gynecol2009;34(1):19–24.

[3] Silver RK,Wilson RD, Philip J, ThomEA, Zachary JM,MohideP, et al. Latefirst-trimesterplacental disruption and subsequent gestational hypertension/preeclampsia. ObstetGynecol 2005;105(3):587–92.

[4] Grobman WA, Auger M, Shulman LP, Elias S. The association between chorionicvillus sampling and preeclampsia. Prenat Diagn 2009;29(8):800–3.

[5] Farina A, Hasegawa J, Raffaelli S, Ceccarini C, Rapacchia G, Pittalis MC, et al. Theassociation between preeclampsia and placental disruption induced by chorionicvillous sampling. Prenat Diagn 2010;30(6):571–4.

[6] Adusumalli J, Han CS, Beckham S, Bartholomew ML, Williams III J. Chorionic villussampling and risk for hypertensive disorders of pregnancy. Am J Obstet Gynecol2007;196(6):591.e1–7.

[7] Khalil A, Akolekar R, Pandya P, Syngelaki A, Nicolaides K. Chorionic villus samplingat 11 to 13 weeks of gestation and hypertensive disorders in pregnancy. ObstetGynecol 2010;116(2 Pt 1):374–80.

[8] Odibo AO, Singla A, Gray DL, Dicke JM, Oberle B, Crane J. Is chorionic villus samplingassociated with hypertensive disorders of pregnancy? Prenat Diagn 2010;30(1):9–13.

[9] Lindgren P, Cederholm M, Haglund B, Axelsson O. Invasive procedures for fetalkaryotyping: no cause of subsequent gestational hypertension or pre-eclampsia.BJOG 2010;117(11):1422–5.

[10] RobertsonWB, Brosens I, LandellsWN. Abnormal placentation. Obstet Gynecol Annu1985;14:411–26.

[11] Khong TY, De Wolf F, Robertson WB, Brosens I. Inadequate maternal vascularresponse to placentation in pregnancies complicated by pre-eclampsia and bysmall-for-gestational age infants. Br J Obstet Gynaecol 1986;93(10):1049–59.

[12] Meekins JW, Pijnenborg R, Hanssens M, McFadyen IR, van Asshe A. A study ofplacental bed spiral arteries and trophoblast invasion in normal and severe pre-eclamptic pregnancies. Br J Obstet Gynaecol 1994;101(8):669–74.

149G. Daskalakis et al. / International Journal of Gynecology and Obstetrics 125 (2014) 146–149

[13] Plasencia W, Maiz N, Bonino S, Kaihura C, Nicolaides KH. Uterine artery Doppler at11+0 to 13+6 weeks in the prediction of pre-eclampsia. Ultrasound Obstet Gynecol2007;30(5):742–9.

[14] Poon LC, Maiz N, Valencia C, Plasencia W, Nicolaides KH. First-trimester maternalserum pregnancy-associated plasma protein-A and pre-eclampsia. UltrasoundObstet Gynecol 2009;33(1):23–33.

[15] Brosens IA. Morphological changes in the utero-placental bed in pregnancy hyper-tension. Clin Obstet Gynaecol 1977;4(3):573–93.

[16] Zhou Y, Damsky CH, Fisher SJ. Preeclampsia is associated with failure ofhuman cytotrophoblasts to mimic a vascular adhesion phenotype. One causeof defective endovascular invasion in this syndrome? J Clin Invest 1997;99(9):2152–64.

[17] WilsonML, Goodwin TM, Pan VL, Ingles SA.Molecular epidemiology of preeclampsia.Obstet Gynecol Surv 2003;58(1):39–66.

[18] Moffett A, Hiby SE. How Does the maternal immune system contribute to thedevelopment of pre-eclampsia? Placenta 2007;28(Suppl. A):S51–6.

[19] Thomsen SG, Isager-Sally L, Lange AP, Saurbrey N, Grønvall S, Schiøler V. Elevatedmaternal serum α-fetoprotein caused by midtrimester amniocentesis: a prognosticfactor. Obstet Gynecol 1983;62(3):297–300.

[20] Shulman LP, Meyers CM, Simpson JL, Andersen RN, Tolley EA, Elias S. Fetomaternaltransfusion depends on amount of chorionic villi aspirated but not on method ofchorionic villus sampling. Am J Obstet Gynecol 1990;162(5):1185–8.

[21] Rana S, Karumanchi SA, Levine RJ, Venkatesha S, Rauh-Hain JA, Tamez H, et al.Sequential changes in antiangiogenic factors in early pregnancy and risk of develop-ing preeclampsia. Hypertension 2007;50(1):137–42.

[22] Widmer M, Villar J, Benigni A, Conde-Agudelo A, Karumanchi SA, Lindheimer M.Mapping the theories of preeclampsia and the role of angiogenic factors: a system-atic review. Obstet Gynecol 2007;109(1):168–80.

[23] Murphy JF, O’Riordan J, Newcombe RG, Coles EC, Pearson JF. Relation of haemoglobinlevels in first and second trimesters to outcome of pregnancy. Lancet 1986;1(8488):992–5.

[24] Kadyrov M, Schmitz C, Black S, Kaufmann P, Huppertz B. Pre-eclampsia and mater-nal anaemia display reduced apoptosis and opposite invasive phenotypes ofextravillous trophoblast. Placenta 2003;24(5):540–8.

[25] Kadyrov M, Kingdom JC, Huppertz B. Divergent trophoblast invasion and apoptosisin placental bed spiral arteries from pregnancies complicated by maternal anemiaand early-onset preeclampsia/intrauterine growth restriction. Am J Obstet Gynecol2006;194(2):557–63.