SHORT COMMUNICATION

Screening for trisomy 21 in twin pregnancies in the firsttrimester: does chorionicity impact on maternal serum freeb-hCG or PAPP-A levels?

Kevin Spencer*

Endocrine Unit, Clinical Biochemistry Department, Harold Wood Hospital, Gubbins Lane, Romford RM3 0BE, UK

In a study of 180 twin pregnancies I have examined the distribution of maternal serum free b-humanchorionic gonadotrophin (b-hCG) and pregnancy-associated plasma protein-A (PAPP-A), in addition tofetal nuchal translucency thickness (NT), in twins classified as monochorionic or dichorionic, based onultrasound appearance at 10–14 weeks of gestation. In 45 monochorionic and 135 dichorionic twinpregnancies the median MoM free b-hCG was not significantly different (1.00 vs 1.01), whilst that forPAPP-A was lower (0.89 vs 1.01) but again with no statistical significance. Previous reports of an increasedfetal NT in monochorionic twins pregnancies could not be confirmed (1.03 vs 1.00). It is concluded that theexisting pseudo risk twin correction algorithm is appropriate for both monochorionic and dichorionic twinsin providing accurate first trimester risks for trisomy 21. Copyright # 2001 John Wiley & Sons, Ltd.

KEY WORDS: NT; PAPP-A; free b-hCG; prenatal screening; first trimester; twins; Down syndrome

INTRODUCTION

In the first trimester of pregnancy, screening by acombination of fetal nuchal translucency thickness(NT) and maternal serum free b-human chorionicgonadotrophin (b-hCG) and pregnancy-associatedplasma protein-A (PAPP-A) has been shown inretrospective (Spencer et al., 1999) and prospectivestudies (Spencer et al., 2000) to identify approximately90% of cases of trisomy 21 for a 5% false positive rate.Similar prospective screening performances have beenreported by De Biasio et al. (1999), De Graaf et al.(1999) and by Krantz et al. (2000). In cases of twinpregnancy both NT (Sebire et al., 1996) and mater-nal serum biochemistry can be combined together(Spencer, 2000) to provide detection rates approachingthose achieved in singleton pregnancies. Such algo-rithms have been successfully implemented in pro-spective practice in identifying twins discordant fortrisomy 21 (Spencer and Nicolaides, 2000).

In examining the distribution of NT in twins, Sebireet al. (1996) observed that in monochorionic twinsthere was a higher incidence of increased NT inchromosomally normal fetuses (8.4% vs 5.4% indichorionic twins and 5.2% in singleton pregnancies).This observation has raised the question of whethersuch differences also existed in the distribution of thebiochemical markers between monochorionic anddichorionic twins.

PATIENTS AND METHODS

Maternal serum PAPP-A and free b-hCG weremeasured in all pregnant women attending our firsttrimester OSCAR clinic (Spencer et al., 2000) using theKryptor rapid random access immunoassay analyserand the time-resolved amplified cryptate emissiontechnology (Brahms Diagnostica, Berlin, Germany).The performance of these assays has been describedpreviously (Spencer et al., 1999). All samples wereanalysed fresh and results obtained within 30 min ofblood collection. During the analysis time all womenhad their pregnancy dated by crown–rump length(CRL) and NT measured by standardised techniquesby sonographers accredited by the Fetal MedicineFoundation. Twin pregnancies were classified asmonochorionic if there was a single placental mass inthe absence of the lambda sign and dichorionic if therewas a single placental mass and the lambda signpresent or if the placentas were not adjacent to eachother. During the period June 1998 to December 2000,11 107 women were screened of which 180 had twinpregnancies.

All marker levels were converted to multiples of thenormal median (MoM) for the gestational age andcorrected for maternal weight (when appropriate) asdescribed from data in a previous study (Spencer et al.,1999) and corrected for the presence of twins (asappropriate) as previously described (Spencer, 2000).Statistical analysis of data was performed using Exceland Analyse-It, a statistical software add-in for Excel 7(Smart Software, Leeds, UK). MoM distributionsfrom monochorionic and dichorionic twin pregnancies

*Correspondence to: K. Spencer, Endocrine Unit, Clinical Bio-chemistry Department, Harold Wood Hospital, Gubbins Lane,Romford RM3 0BE, UK. E-mail: KevinSpencer1@cs.com

PRENATAL DIAGNOSIS

Prenat Diagn 2001; 21: 715–717.DOI: 10.1002 /pd.152

Copyright # 2001 John Wiley & Sons, Ltd. Received: 8 January 2001Revised: 11 April 2001

Accepted: 25 April 2001

were compared by t-tests of unequal variance afterlog10 transformation of the MoMs.

RESULTS

Table 1 summarises the general characteristics of themonochorionic and dichorionic twin populations. Inthe monochorionic twin group the median MoM freeb-hCG was 1.00 compared with 1.01 in the dichorionictwin group. For PAPP-A the median MoM was 0.89compared with 1.01 in the dichorionic twin group.When the log10MoM distributions were comparedwith t-tests of unequal variance the difference betweenthe log distributions was not statistically significant( p>0.2). The log10SDs of each marker were alsosimilar in each twin sets (free b-hCG: monochorionic0.2637, dichorionic 0.2582; PAPP-A: monochorionic0.2032, dichorionic 0.2171).

For NT, the median MoM in cases from mono-chorionic twin pregnancies was 1.03 compared with1.00 for dichorionic twin pregnancies. After logtransformation this difference was found to be notstatistically significant, although the log SD wasconsiderably wider in the monochorionic twins(0.1645) compared to the dichorionic twins (0.1244).Of the monochorionic fetuses 5.6% (5/90) had an NTMoM above the 95th centile for normal singletonpregnancies, whilst in fetuses from dichorionic preg-nancies this was 4.4% (12/70). This indicated a relativerisk of 1.27 when comparing the incidence in mono-chorionic to dichorionic pregnancies.

DISCUSSION

The prevalence of twin pregnancies increases withmaternal age and consequently in twins the risk forchromosomal abnormalities is higher than average(Baldwin, 1994). Multiple pregnancies are also atincreased risk of structural defects and the relative riskfor such defects in twins is twice that in singletons.Chorionicity is of relevance to several aspects ofprenatal diagnosis, to risk counselling, testing proce-dures and to the management of abnormal pregnan-cies. The relative increase in risk of monochorioniccompared to dichorionic pregnancies is comparable to

that of twins compared to singletons. Zygosity is alsorelevant to prenatal diagnosis of genetic disease. Allmonochorionic twins are monozygotic and thereforethe risk for chromosomal abnormalities affecting bothfetuses is identical to that of a singleton pregnancy(Fisk and Bennett, 1995). With dichorionic pregnan-cies, the fetuses are dizygotic in 90% of cases andhence invasive diagnostic testing is necessary for bothfetuses (Fisk and Bennett, 1995). Chorionicity can bemost easily and reliably determined in the firsttrimester by observation of the ‘lambda sign’ or ‘Tsign’ on ultrasound examination. In screening forchromosomal anomalies in the first trimester it hasbeen suggested that the incidence of fetuses with NTabove the 95th centile is about 1.5 times greater inmonochorionic twins compared with dichorionic twinsor singleton pregnancies (Sebire et al., 1996).

In expanding NT screening to include first trimestermaternal serum biochemistry we have shown thatcombined screening for twins can be an effectiveprocedure (Spencer, 2000; Spencer and Nicolaides,2000). The present study has shown that in twinpregnancies there is no statistical difference betweenthe distribution of the markers free b-hCG and PAPP-A between unaffected monochorionic and dichorionictwins. The spread of the distributions in both cases arevery similar to previously reported data (Spencer,2000). These results would suggest that no allowanceneeds to be taken of chorionicity when interpretingbiochemical markers in twin pregnancies in the firsttrimester. With fetal NT, in this small series, astatistically significant increased median MoM couldnot be confirmed in monochorionic twins although thespread of results was much wider than with dicho-rionic twins. These results may not be that dissimilarto those of Sebire et al. (1996) since whilst a statisticalincreased MoM could not be proven, a 1.27 relativerisk followed the same trend as seen by Sebire andco-workers. In an extended series of 303 mono-chorionic twins (Sebire et al., 2000), in which 15%developed twin-to-twin transfusion syndrome, fetalNT was above the 95th centile in 8.2% of cases, thusreaffirming the increased incidence seen in the originalstudy (Sebire et al., 1996). In a smaller study, Monniet al. (2000) have also shown a greater relativeincidence (1.66) of raised NT in monochorionic twins.

Further studies are required to see if the observedlower PAPP-A in these cases reaches statisticalsignificance. In the meantime, screening using theproposed pseudo risk algorithm (Spencer, 2000) willgive risks equivalent to those in singleton pregnanciesand allow families expecting twins to make aninformed choice with respect to prenatal diagnosis.

REFERENCES

Baldwin VJ. 1994. Anomalous development in twins. In Pathology ofMultiple Pregnancy, Baldwin VJ (ed.). Springer-Verlag: NewYork; 169–197.

De Biasio P, Siccardi M, Volpe G, Famularo L, Santi F, Canini S.1999. First trimester screening for Down syndrome using nuchaltranslucency measurements with free b-hCG and PAPP-A

Table 1—General characteristics of the study population

Monochorionictwins

Dichorionictwins

n 45 135Median maternal age (years) 32.81 31.50Median maternal weight (kg) 66.0 66.5Median gestation (days) 87.5 87.6Median CRL (mm) 58.1 60.6Median NT (mm) 1.3 1.4Smoker/non-smoker 8/37 27/108Median NT MoM 1.03 1.00Median free b-hCG MoM 1.00 1.01Median PAPP-A MoM 0.89 1.01

K. SPENCER716

Copyright # 2001 John Wiley & Sons, Ltd. Prenat Diagn 2001; 21: 715–717.

between 10 and 13 weeks of pregnancy – the combined test.Prenat Diagn 19: 360–363.

De Graaf IM, Pajkrt E, Bilardo CM, Leschot NJ, Cuckle HS, VanLlith JMM. 1999. Early pregnancy screening of fetal aneuploidywith serum markers and nuchal translucency. Prenat Diagn 19:458–462.

Fisk NM, Bennett PR. 1995. Prenatal determination of chorionicityand zygosity. In Multiple Pregnancy, Ward RH, Whittle M (eds).RCOG Press: London; 56–67.

Krantz DA, Hallahan TW, Orlandi F, Buchanan P, Larsen JW,Macri JN. 2000. First trimester Down syndrome screening usingdried blood biochemistry and nuchal translucency. Obstet Gynecol96: 207–213.

Monni G, Zoppi MA, Ibba RM, Putzolu M, Floris M. 2000. Nuchaltranslucency in multiple pregnancies. Croat Med J 41: 266–269.

Sebire NJ, Snijders RJM, Hughes K, Sepulveda W, Nicolaides KH.1996. Screening for trisomy 2 in twin pregnancies by maternal ageand fetal nuchal translucency thickness at 10–14 weeks ofgestation. Br J Obstet Gynaecol 103: 999–1003.

Sebire NJ, Souka A, Skentou H, Geerts L, Nicolaides KH. 2000.

Early prediction of severe twin-to-twin transfusion syndrome.Hum Reprod 15: 2008–2010.

Spencer K. 2000. Screening for trisomy 21 in twin pregnancies in thefirst trimester using free b-hCG and PAPP-A, combined with fetalnuchal translucency thickness. Prenat Diagn 20: 91–95.

Spencer K, Nicolaides KH. 2000. First trimester prenatal diagnosisof trisomy 21 in discordant twins using fetal nuchal translucencythickness and maternal serum free b-hCG and PAPP-A. PrenatDiagn 20: 683–684.

Spencer K, Souter V, Tul N, Snijders R, Nicolaides KH. 1999. Ascreening program for trisomy 21 at 10–14 weeks using fetalnuchal translucency, maternal serum free b-human chorionicgonadotropin and pregnancy associated plasma protein-A. Ultra-sound Obstet Gynecol 13: 231–237.

Spencer K, Spencer CE, Power M, Moakes A, Nicolaides KH. 2000.One stop clinic for assessment of risk for fetal anomalies: a reportof the first year of prospective screening for chromosomalanomalies in the first trimester. Br J Obstet Gynaecol 107:1271–1275.

SCREENING FOR TRISOMY 21 IN TWIN PREGNANCIES 717

Copyright # 2001 John Wiley & Sons, Ltd. Prenat Diagn 2001; 21: 715–717.