PRENATAL DIAGNOSISPrenat Diagn 2008; 28: 7–10.Published online 14 November 2007 in Wiley InterScience(www.interscience.wiley.com) DOI: 10.1002/pd.1890

Low levels of maternal serum PAPP-A in the first trimesterand the risk of pre-eclampsia

Kevin Spencer1,2*, Nicholas J. Cowans1 and Kypros H. Nicolaides2

1Prenatal Screening Unit, Department of Clinical Biochemistry, King George Hospital, Barley Lane, Goodmayes, Essex IG3 8YB,UK2Harris Birthright Research Centre for Fetal Medicine, Kings College Hospital, Denmark Hill, London, UK

Background Low levels of pregnancy associated plasma protein-A (PAPP-A) have been previously shownto be associated with pregnancies that subsequently develop pre-eclampsia. The objective of this study was toestablish the relative risk for pre-eclampsia at various PAPP-A levels as an aid to counselling and follow upof pregnancies.

Methods Maternal serum PAPP-A and free ß-human chorionic gonadotropin (ß-hCG) levels at 11 to 13 weeksof gestation from 224 singleton pregnancies that subsequently developed pre-eclampsia were compared to thosefrom 47 770 normal singleton pregnancies resulting in live births after 37 weeks with birth weight greater thanor equal to the 10th centile of normal for gestation. In all cases, the measured PAPP-A and free ß-hCG levelswere expressed as multiple of the median (MoM). The association between metabolite levels and the incidenceof pre-eclampsia was assessed by comparing the relative incidence at a number of MoM cut-offs and at variouscentiles. At various marker levels, the likelihood ratio for pre-eclampsia was also calculated.

Results In the pre-eclampsia group the median PAPP-A MoM was significantly reduced (0.772 MoM,p < 0.0001) whilst the median free β-hCG MoM was not different from controls (0.981 MoM, p = 0.26).With decreasing levels of PAPP-A, the likelihood ratio for pre-eclampsia increased. At the 5th centile of normal(PAPP-A MoM 0.415) the odds ratio was increased 4-fold and at this cut-off 15% of cases of pre-eclampsiawould be identified.

Conclusions The graphical presentation of a likelihood ratio curve for pre-eclampsia at any PAPP-A MoMlevel is likely to be useful in counselling women with low levels of PAPP-A and a normal karyotype. Useof low levels of PAPP-A for selecting women for further follow-up with uterine artery Doppler may furtherimprove the clinical discrimination. Copyright 2007 John Wiley & Sons, Ltd.

KEY WORDS: PAPP-A; pre-eclampsia; prenatal screening; uterine artery Doppler

INTRODUCTION

Pre-eclampsia occurs in about 2% of pregnancies andis a major cause of maternal death worldwide. Theunderlying cause of pre-eclampsia is thought to be acirculatory maladaptation characterised by defective tro-phoblast invasion (Lyall, 2002). The onset of trophoblastinvasion occurs in the first trimester of pregnancy andattempts are now focussed towards the early identifica-tion of women at high risk for subsequent developmentof pre-eclampsia.

Several recent studies have shown an associationbetween low first-trimester maternal serum pregnancyassociated plasma protein-A (PAPP-A) levels and thesubsequent development of pregnancy complications,including pre-eclampsia (Ong et al., 2000; Smith et al.,2002; Yaron et al., 2002; Tul et al., 2003; Dugoffet al., 2004; Krantz et al., 2004; Spencer et al., 2005a,2006a,b, 2007a,b; Cowans and Spencer, 2007).

The aim of this study is to explain the associationbetween PAPP-A and the incidence of pre-eclampsia by

*Correspondence to: Kevin Spencer, Prenatal Screening Unit,Department of Clinical Biochemistry, King George Hospi-tal, Barley Lane, Goodmayes, Essex IG3 8YB. UK. E-mail:KevinSpencer1@aol.com

comparing the relative incidence at a number of PAPP-A MoM cut-offs and at various centiles and to constructthe likelihood ratio (LR) for pre-eclampsia at variousmarker levels.

METHODS

All women booked for maternity care at the followingUK hospitals were offered screening for trisomy 21 bya combination of foetal nuchal translucency (NT) andmaternal serum free β-human chorionic gonadotropin(β-hCG) and PAPP-A at 11+0 to 13+6 weeks’ gestation:Harold Wood Hospital, Romford (between June 1998and December 2003), King George Hospital, Good-mayes (between July 2001 and December 2003), Kentand Canterbury Hospital, Canterbury (between July 2002and December 2003), William Harvey Hospital, Ashford(between July 2002 and December 2003), Queen Eliz-abeth The Queen Mother’s Hospital, Margate (betweenJuly 2002 and December 2003), King’s College Hos-pital, London (between January 1999 and February2000) and those attending The Fetal Medicine Cen-tre, London (between July 1999 and December 2003).Women received an information leaflet about the service

Copyright 2007 John Wiley & Sons, Ltd. Received: 10 July 2007Revised: 2 October 2007

Accepted: 5 October 2007Published online: 14 November 2007

8 K. SPENCER, ET AL

and gave details about their demographic characteristicsand medical history, which were entered into computerdatabases at the two centres (PIA-Fetal Database, View-Point, Webling, Germany).

Maternal serum free β-hCG and PAPP-A were mea-sured using the Kryptor analyser (Brahms AG, Berlin,Germany) as previously described (Spencer et al., 1999)and an ultrasound examination was carried out to mea-sure foetal NT, crown-rump length (CRL) and to diag-nose any major foetal abnormalities. All scans werecarried out by sonographers who had obtained TheFetal Medicine Foundation Certificate of Competence inthe 11+0 to 13+6 week scan (www.fetalmedicine.com).Patient-specific risks were calculated by a multivariateapproach using biochemical population parameters out-lined in a previous study (Spencer et al., 1999), LRsbased on delta NT outlined in a previous study (Spenceret al., 2003b) and the gestational age–related risk oftrisomy 21 at the time of screening (Snijders et al.,1999). Data on pregnancy outcome were obtained fromthe cytogenetics laboratories, the National ChromosomalAnomaly Register, the patients themselves, their generalpractitioners or the maternity units in which they deliv-ered. In the case of all women screened by the HaroldWood laboratory, pregnancy outcome was additionallyobtained from hospital midwifery or Patient Adminis-tration computer records and matched to the prenatalscreening records by a locally derived record linkagesoftware. Detailed outcome follow-up of this type is partof good screening practice and audit–and no further eth-ical approval was necessary for this process.

All biochemical markers were converted to multipleof the median (MoM) of the expected normal medianfor a pregnancy of the same gestational day usingvalues established in a previous study (Ong et al., 2000)corrected for maternal weight (Spencer et al., 2000,2003a), self-recorded smoking status (Spencer et al.,2004) and ethnicity (Spencer et al., 2000, 2005a). Totake account of gestational variation of foetal NT, weexpressed the measured foetal NT as the difference fromthe normal median NT (Snijders et al., 1998; Spenceret al., 2003a) at the measured CRL, this being the deltaNT. The delta NT approach has been shown to be moreappropriate than the conventional MoM approach forhandling NT data and calculating patient-specific risks(Spencer et al., 2003b).

The foetal database was searched for records, whichhad complete outcome information, including gestationalage at delivery (completed weeks), birth weight andpregnancy outcome.

The normal pregnancy group was defined as thosepregnancies in which a single live foetus was deliveredafter 37 complete weeks of gestation with birth weightabove the 10th centile of normal for gestational age(Yudkin et al., 1987) and for which no foetal anomalieswere observed.

For those women delivering at Harold Wood Hospi-tal, data on pregnancy outcome were obtained from theexamination of individual patient notes and labour wardrecords. Women developing pre-eclampsia were definedaccording to the guidelines of the International Soci-ety for the Study of Hypertension in Pregnancy. This

requires two recordings of systolic/diastolic blood pres-sure (at least one of which is >140/90 mmHg) measuredat least 4 h apart in previously normotensive women,and proteinuria of 300 mg or more in 24 h, or two read-ings of at least 2+ on dipstick analysis of midstream orcatheter urine specimens if no 24 h collection is avail-able (Brown et al., 2001). Outcome records for womendeveloping pre-eclampsia were matched to their respec-tive first-trimester biochemical screening records.

Some of the pre-eclamptic cases included in thisstudy were part of our previous studies (Spencer et al.,2005b, 2007c). The association between PAPP-A and theincidence of pre-eclampsia was assessed by comparingthe relative incidence at a number of MoM cut-offs andat various centile cut-offs. At various marker levels theLR for pre-eclampsia was also calculated.

Statistical analysis was carried out with the proceduresas described using Analyse-It (Analyse-It Software Ltd,Leeds) a statistical add in for Excel or SPSS Version 13(SPSS Ltd, Woking).

RESULTS

The normal pregnancy group comprised of 47 770 preg-nancies and the characteristics of this group have beenpreviously described (Spencer et al., 2006a). We identi-fied 222 cases with pre-eclampsia and the characteristicsof the cases are described in Table 1.

In the pre-eclampsia group, the median PAPP-A MoMwas 0.772 (95% confidence interval 0.6905–0.8605)compared with 1.0431 (95% confidence interval 1.0373–1.0497) in the controls. The mean log10 PAPP-A MoMwas significantly lower than in the controls (unpairedt-test with unequal variance p < 0.001). In the pre-eclampsia group, the median free β-hCG MoM was0.981 (95% confidence interval 0.896–1.128) comparedwith 1.052 (95% confidence interval 1.046–1.060) inthe controls. The mean log10 free β-hCG MoM was notsignificantly different between the two groups (unpairedt-test with unequal variance, p = 0.26).

Table 1—Summary of characteristics of the normal outcomeand pre-eclamptic population

Variable

Normaloutcome

(n = 47 770)Pre-eclampsia

(n = 222)

Age, mean (SD) (years) 31.5 (5.5) 29.7 (5.6)Nulliparity % 43.2% 65%Ethnicity %Caucasian 86% 85%Asian 6% 3%African 5% 10%Other 5% 2%GA at enrolment, mean(SD) (days)

84 (4.2) 86 (4.4)

GA delivery (SD)(weeks)

39.6 (1.2) 36.9 (2.5)

Birth weight (SD) (g) 3510 (410) 2285 (590)

Copyright 2007 John Wiley & Sons, Ltd. Prenat Diagn 2008; 28: 7–10.DOI: 10.1002/pd

LOW PAPP-A AND THE RISK OF PRE-ECLAMPSIA 9

Table 2—Summary of various studies of PAPP-A and free β-hCG in pre-eclampsia

Pre-eclampsia PAPP-A Free ß-hCG

5th centile 5th centile

Study N MoMOR or

RR DROR or RR<0.5 MoM MoM

OR orRR DR

OR or RR<0.2 MoM

Ong et al., 2000 80 0.903 11.1 0.879 7.4Smith et al., 2002 331 0.963 2.1 7.6 1.1 4.1Yaron et al., 2002 27 1.7Dugoff et al., 2004 764 1.54 7.85Spencer et al., 2005b 64 0.844 0.923 7.8This study 222 0.772 3.7 14.6 0.981 1.8 7.7

OR, odds ratio; RR, relative risk.

0

0.5

1

1.5

2

2.5

3

3.5

4

0 0.2 0.4 0.6 0.8 1 1.2 1.4

PAPP-A MoM

Like

lihoo

d ra

tio fo

r P

re-e

clam

psia

Figure 1—Likelihood ratio for pre-eclampsia based on PAPP-A MoM.Symbols represent the individual point estimates, while the smoothedline represents the best fit to the data

At the 5th centile of normal for PAPP-A (0.415 MoM)the odds ratio for pre-eclampsia was increased by 3.7-fold (95% confidence interval 2.3–4.8) and the detectionrate for pregnancies destined to develop pre-eclampsiawas 15%. The individual LRs for pre-eclampsia at eachspecific MoM level of PAPP-A is shown in Figure 1 withthe individual data points and the solid line representingthe best fit through the data points using an exponentialregression data fit. The regression coefficient was 0.935and the best fit equation was

LR pre-eclampsia = 4.815 × e−1.214 (PAPP−A MoM)

When the data on pre-eclamptic cases were exam-ined in two groups based on those delivering prior to35 weeks and those after 35 weeks, there was no signif-icant difference in the levels of PAPP-A between thosedelivering early or later.

DISCUSSION

The results of this study have confirmed data fromour previous studies (Ong et al., 2000; Spencer et al.,2005b), which have shown a reduction in PAPP-Alevels associated with pregnancies that subsequentlydevelop pre-eclampsia. There is a modest increase inLR with decreasing levels of PAPP-A and in the

presence of a normal karyotype such low levels, inaddition to being an indicator of increased risk for pre-eclampsia, also indicated increased risk for other adversepregnancy outcomes (Spencer et al., 2006a, 2007a,b).The presentation of data in terms of LR curves at anylevel of PAPP-A MoM is likely to be a more useful aidto counselling than the point estimates from previousstudies (Ong et al., 2000; Smith et al., 2002; Yaronet al., 2002; Dugoff et al., 2004; Spencer et al., 2005b),which are summarised in Table 2.

Although a low PAPP-A in itself is not a strongindicator of pre-eclampsia, we have previously showna significant improvement in detection by combin-ing first-trimester PAPP-A measurement with second-trimester uterine artery Doppler velocimetry (Spenceret al., 2005b). Thus women with a low first-trimesterPAPP-A MoM (e.g. <0.3 MoM representing the 1.5thcentile) and a normal karyotype should be followed upfurther with 22 to 24 weeks Doppler measurement toassess risk of pre-eclampsia and poor foetal growth.Whether further improvements could be made using acombination of PAPP-A and first-trimester uterine arteryDoppler remains to be seen. Certainly, early uterineartery Doppler measurement does seem to have less dis-crimination than that done at 22 to 24 weeks (Martinet al., 2001) although there is little or no correlationbetween reduced levels of PAPP-A in the first trimesterand uterine artery Doppler either in the first trimester(Prefumo et al., 2006) or second trimester (Spenceret al., 2005b). The early identification of women atincreased risk of developing pre-eclampsia by a com-bination of low PAPP-A alone or in combination withother markers such as inhibin, activin or PP13 (Onget al., 2004; Spencer et al., 2006b, 2007c; Nicolaideset al., 2006), along side uterine artery Doppler measure-ment and maternal factors (Yu et al., 2005) may allowincreased surveillance or early possible therapeutic inter-ventions with low dose aspirin in an attempt to arrestthe disease progression. Further studies are required toestablish the clinical value of these approaches.

ACKNOWLEDGEMENTS

This study was supported in part by a grant from NHSR&D (RF4: Risk Assessment in Pregnancy) to ProfessorKevin Spencer.

Copyright 2007 John Wiley & Sons, Ltd. Prenat Diagn 2008; 28: 7–10.DOI: 10.1002/pd

10 K. SPENCER, ET AL

REFERENCES

Brown MA, Lindheimer MD, de Swiet M, VanAssche A,Moutquin JM. 2001. The classification and diagnosis of the hyper-tensive disorder of pregnancy: statement from the InternationalSociety for the Study of Hypertension in Pregnancy (ISSHP).Hypertens Pregnancy 20: IX–XIV.

Cowans NJ, Spencer K. 2007. First trimester ADAM12 and PAPP-Aas markers of intrauterine fetal growth restriction through their rolesin the insulin-like growth factor system. Prenat Diagn 27: 264–271.

Dugoff L, Hobbins JC, Malone FD, et al. 2004. First trimestermaternal serum PAPP-A and free beta subunit human chorionicgonadotropin concentrations and nuchal translucency are associatedwith obstetric complications: A population based screening study(The FASTER Trial). Am J Obstet Gynecol 191: 1446–1451.

Krantz D, Goetzl L, Simpson JL, et al. 2004. Association of extremefirst trimester free human chorionic gonadotropin-β, pregnancyassociated plasma protein A, and nuchal translucency withintrauterine growth restriction and other adverse pregnancyoutcomes. Am J Obstet Gynecol 191: 1452–1458.

Lyall F. 2002. The human placental bed revisited. Placenta 23:555–562.

Martin AM, Bindra R, Curcio P, Cicero S, Nicolaides KH. 2001.Screening for pre-eclampsia and fetal growth restriction by uterineartery Doppler at 11–14 weeks of gestation. Ultrasound ObstetGynecol 18: 583–586.

Nicolaides KH, Bindra R, Turzan OM, et al. 2006. A novel approachto first trimester screening for early pre-eclampsia combining serumPP-13 and Doppler ultrasound. Ultrasound Obstet Gynecol 27:13–17.

Ong CYT, Liao AW, Spencer K, Munim S, Nicolaides KH. 2000.First trimester maternal serum free β human chorionicgonadotrophin and pregnancy associated plasma protein A aspredictors of pregnancy complications. Br J Obstet Gynaecol 107:1265–1270.

Ong CYT, Liao AW, Munim S, Spencer K, Nicolaides KH. 2004.First trimester maternal serum activin A in pre-eclampsia and fetalgrowth restriction. J Matern Fetal Neonatal Med 15: 176–180.

Prefumo F, Canini S, Casagrande V, Pastorino D, Venturini PL, DeBiasio P. 2006. Correlation between first trimester uterine arteryDoppler indicies and maternal serum free β-human chorionicgonadotropin and pregnancy associated plasma protein A. FertilSteril 86: 977–980.

Smith GCS, Stenhouse EJ, Crossley JA, Aitken DA, Cameron AD,Connor JM. 2002. Early pregnancy levels of pregnancy associatedplasma protein A and the risk of intrauterine growth restriction,premature birth, preeclampsia and stillbirth. J Clin EndocrinolMetab 87: 1762–1767.

Snijders RJM, Noble P, Sebire N, Souka A, Nicolaides KH. 1998.UK multicentre project on assessment of risk for trisomy 21by maternal age and fetal nuchal translucency thickness at10–14 weeks of gestation. Lancet 18: 519–521.

Snijders RJM, Sundberg K, Holzgreve W, Henry G, Nicolaides KH.1999. Maternal age and gestation specific risk for trisomy 21.Ultrasound Obstet Gynecol 13: 167–170.

Spencer K, Ong CYT, Liao AWK, Nicolaides KH. 2000. Theinfluence of ethnic origin on first trimester biochemical markersof chromosomal anomalies in the first trimester. Prenat Diagn 20:491–494.

Spencer K, Souter V, Tul N, Snijders R, Nicolaides KH. 1999. Ascreening program for trisomy 21 at 10–14 weeks using fetal

nuchal translucency, maternal serum free β-human chorionicgonadotropin and pregnancy associated plasma protein-A.Ultrasound Obstet Gynecol 13: 231–237.

Spencer K, Bindra R, Nicolaides KH. 2003a. Maternal weightcorrection of maternal serum PAPP-A and free β-hCG whenscreening for trisomy 21 in the first trimester of pregnancy. PrenatDiagn 23: 851–855.

Spencer K, Bindra R, Nix ABJ, Heath V, Nicolaides KH. 2003b.Delta-NT or NT MoM: which is the most appropriate method forcalculating accurate patient specific risks for trisomy 21 in the firsttrimester? Ultrasound Obstet Gynecol 22: 142–148.

Spencer K, Bindra R, Cacho AM, Nicolaides KH. 2004. The impactof smoking status when screening for chromosomal anomaliesusing maternal serum biochemistry and fetal nuchal translucencythickness in the first trimester of pregnancy. Prenat Diagn 24:169–173.

Spencer K, Heath V, El-Sheikhah A, Ong CYT, Nicolaides KH.2005a. Ethnicity and the need for correction of biochemical andultrasound markers of chromosomal anomalies in the first trimester:a study of Oriental, Asian and Afro-Caribbean populations. PrenatDiagn 25: 365–369.

Spencer K, Yu CKH, Cowans NJ, Otigbah C, Nicolaides KH. 2005b.Prediction of pregnancy complications by first trimester maternalserum PAPP-A and free β-hCG and with second trimester uterineartery Doppler. Prenat Diagn 25: 949–953.

Spencer K, Cowans NJ, Avgidou K, Nicolaides KH. 2006a. Firsttrimester ultrasound and biochemical markers of aneuploidy andthe prediction of impending fetal death. Ultrasound Obstet Gynecol28: 637–643.

Spencer K, Yu CKH, Savvidou M, Papageorghiou AT, Nicolaides KH.2006b. Prediction of pre-eclampsia by uterine artery Dopplerultrasonography and maternal serum pregnancy associated plasmaprotein A, free β-human chorionic gonadotropin, activin A andinhibin A at 22 + 0 to 24 + 6 weeks’ gestation. Ultrasound ObstetGynecol 27: 658–663.

Spencer K, Cowans NJ, Avgidou K, Molina F, Nicolaides KH.2007a. First trimester ultrasound and biochemical markers ofaneuploidy and the prediction of fetal growth restriction. UltrasoundObstet Gynecol. In Press.

Spencer K, Cowans NJ, Molina F, Kagan KO, Nicolaides KH. 2007b.First trimester ultrasound and biochemical markers of aneuploidyand the prediction of pre-term or very early pre-term delivery.Ultrasound Obstet Gynecol. In Press.

Spencer K, Cowans NJ, Chefetz I, Tal J, Meiri H. 2007c. Firsttrimester maternal serum PP-13, PAPP-A and second trimesteruterine artery Doppler pulsatility index as markers of pre-eclampsia.Ultrasound Obstet Gynecol 29: 128–134.

Tul N, Punsenjak S, Osredkar J, Spencer K, Novak-Antolic Z. 2003.Predicting complications of pregnancy with first trimester maternalserum free beta hCG, PAPP-A and inhibin-A. Prenat Diagn 23:990–996.

Yaron Y, Heifetz S, Ochshorn Y, Lehavi O, Orr-Urtreger A. 2002.Decreased first trimester PAPP-A is a predictor of adversepregnancy outcome. Prenat Diagn 22: 778–782.

Yu CKH, Smith GCS, Papageorghiou AT, Cacho AM, Nicolaides KH.2005. An integrated model for the prediction of pre-eclampsiausing maternal factors and uterine artery Doppler velocimetry inunselected low risk women. Am J Obstet Gynecol 193: 429–436.

Yudkin PL, Aboualfa M, Eyre JA, Redman CW, Wilkinson AR.1987. New birth weight and head circumference centiles forgestational ages 24 to 42 weeks. Early Hum Dev 15: 45–52.

Copyright 2007 John Wiley & Sons, Ltd. Prenat Diagn 2008; 28: 7–10.DOI: 10.1002/pd