Original Article

Outcome following selective fetal reduction in monochorionic anddichorionic twin pregnancies discordant for structural, chromosomaland genetic disorders

Elena NOBILI,1 Gowrishankar PARAMASIVAM1 and Sailesh KUMAR1,2

1Centre for Fetal Care, Queen Charlotte’s and Chelsea Hospital, Imperial College London, and 2Institute of Reproductive andDevelopmental Biology, Imperial College London, London, UK

Background: To investigate the indications for offering selective fetal reduction in monochorionic (MC) and dichorionic(DC) twins and to correlate obstetric outcome with the antenatal procedure.Methods: All cases of MC and DC twins discordant for structural anomalies and for chromosomal/genetic abnormalitieswere included. Selective reductions performed for twin-to-twin transfusion syndrome or growth restriction were excluded.For DC twins, feticide was achieved using intracardiac injection of potassium chloride (KCl). For MC twins, bipolar cordocclusion (BCO), interstitial laser or radiofrequency ablation (RFA) was used.Results: There were 121 twin pregnancies discordant for structural and chromosomal abnormalities. Only 88 (56 wereMC twins and 32 were DC twins) had selective reduction. For both MC and DC twins, the leading indication forselective reduction was structural anomalies with CNS malformations the most common. For all MC fetal reductiontechniques, the overall pregnancy loss rate (<24 weeks) was 8.9% with RFA having the lowest procedure loss rate (7.7%).The preterm delivery rate was lowest with reduction in DC pregnancies. The live birth rates for MC twins were >87% and100% for DC twins.Conclusions: Selective reduction in MC pregnancies carries an increased procedure-related and preterm delivery ratecompared with DC pregnancies. The main indication for selective reduction was structural malformations, with apredominance of CNS anomalies.

Key words: bipolar cord occlusion, monochorionic, radiofrequency ablation, selective termination, twins.

Introduction

Selective fetal reduction is performed for a variety ofindications in multiple pregnancies. The indications foroffering selective fetal reduction in a multiple pregnancyshould be, and usually are, not different from those insingleton pregnancies when a structural anomaly or achromosomal abnormality is detected. However, additionalconsiderations relevant in a multiple pregnancy include theincreased procedure-related risk of pregnancy loss, theideal gestation to perform selective feticide, the possibilityof maternal coagulopathy secondary to the prolongedretention of a nonviable fetus, the appropriate techniqueand finally the psychological effects on the mother ofcarrying a dead fetus for a prolonged period of time.

In dichorionic pregnancies, the technique for fetalreduction is similar to that employed in singletonswhere intracardiac injection of potassium chloride is usedto achieve asystole.1 In contrast, fetal reduction inmonochorionic pregnancies requires a technique thatablates or interrupts blood flow in the umbilical cord ofthe affected fetus. This is because the monochorionicplacenta usually contains various anastomoses between thetwo fetal circulations that preclude the injection of acardiotoxic agent into one fetus as it may affect the othertwin.The purpose of this study was to investigate the

indications for offering selective fetal reduction inmonochorionic and dichorionic twin pregnancies, toascertain the spectrum of structural fetal anomalies seenrelative to chorionicity and to correlate obstetric outcomewith the antenatal fetal reduction procedure.

Materials and Methods

We conducted a retrospective observational study of allwomen with monochorionic and dichorionic twinpregnancies discordant for structural fetal anomalies and

Correspondence: Dr Sailesh Kumar, Centre for Fetal Care,Queen Charlotte’s and Chelsea Hospital, Imperial CollegeLondon, London, W12 0HS, UK.Email: sailesh.kumar@imperial.ac.uk

Received 20 November 2012; accepted 30 January 2013.

114 © 2013 The Authors

ANZJOG © 2013 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists

Australian and New Zealand Journal of Obstetrics and Gynaecology 2013; 53: 114–118 DOI: 10.1111/ajo.12071

Th e Australian and New Zealand Journal of Obstetrics and Gynaecology

for chromosomal abnormalities identified on antenatalultrasound, referred to a tertiary fetal medicine unitbetween December 2002 and May 2010. We specificallyexcluded selective reductions performed due to severetwin-to-twin transfusion syndrome or fetal growthrestriction. The Centre for Fetal Care at QueenCharlotte’s and Chelsea Hospital is a major referral centrefor fetal disorders. Data were collected from thedepartment’s fetal medicine database (Astraia, Hamburg,Germany), computerised medical records and individualchart review. This was a retrospective analysis performedas part of a clinical audit for which Research EthicsCommittee approval was not required. The dichorionictwin cohort served as a ‘control group’ for comparisonpurposes for the monochorionic twin cohorts.Details about karyotyping, the technique of selective

fetal reduction, procedure-related complications as well asperinatal outcome data, including details of termination ofpregnancy, gestation at delivery, mode of delivery andbirth weight, were also obtained.Selective fetal reduction was offered to patients if a

major structural anomaly or a chromosomal abnormalitywas identified prior to 24 weeks of gestation. Beyond thisthreshold, the abnormality would necessarily have to resultin significant long-term handicap for an offer oftermination to be made.After informed consent, all procedures were performed

under local anaesthesia and with antibiotic cover.Empirically, because of the longer duration andcomplexity of the procedure in monochorionic twins, thesepatients received indomethacin 100 mg rectally andprogesterone 400 mg vaginal pessaries for 3 dayspostprocedure to minimise the risk of procedure-relatedloss. All procedures were performed by accreditedspecialists in maternal–fetal medicine.For dichorionic twins, feticide was achieved using direct

intracardiac injection of potassium chloride (KCl) (1.5 gin 10 mL). For monochorionic twins, over the studyperiod, 3 different vascular ablative techniques were used.These included bipolar cord occlusion (BCO), interstitiallaser and more recently radiofrequency ablation (RFA).The description of these techniques has previously beendetailed in publications from our unit1–4 and will not berepeated. Over the study period, we ceased performinginterstitial laser because of the high procedure-related lossrate. Only RFA and BCO were subsequently used forselective reduction. BCO was reserved for use at slightlylater gestations.Fetal asystole was confirmed at the end of each procedure

and again 1 h later. The majority of patients were scannedwithin 24 h to confirm ongoing viability of the co-twin.Where repeat scan within 24 h was not possible, patientswere scanned within 7 days from the procedure.For all procedures, we considered a procedure-related

aetiology if there was a pregnancy loss or unexpecteddelivery within 2 weeks of the selective reduction. Thepreterm (<37 weeks) delivery rate was also calculated forall subgroups.

Results

From January 2002 to May 2010, 121 twin pregnanciesdiscordant for structural and chromosomal abnormalitieswere identified: 52 dichorionic diamniotic (DCDA) twins,66 monochorionic diamniotic (MCDA) twins and 3monochorionic monoamniotic (MCMA). Of these cases,only 88 women proceeded with selective fetal reduction.Of the 88 cases, 56 were monochorionic twins (53MCDA and 3 MCMA) and 32 were DCDA twins.Maternal characteristics of the monochorionic and

dichorionic cohorts are shown in Table 1. Assistedreproductive techniques (ART) were more common in thedichorionic cohort (18.8%, 6/32) compared with themonochorionic group (1.7%, 1/56). There were nodifferences in parity or ethnicity.The median gestational age at which the abnormality or

aneuploidy was detected was 17 weeks (13–22 weeks) inmonochorionic twins and 21 weeks (17–22 weeks) indichorionic twins.Table 2 shows the various indications for selective fetal

reduction in the two groups. For monochorionic twins, theleading indication for selective fetal reduction was centralnervous system abnormalities (60.7%, 34/56), the majorityof which were neural tube defects (18/34). This wasfollowed by fetal reduction for TRAP syndrome (12.5%,7/56) and cardiac abnormalities (8.9%, 5/56). Otherreasons included anterior abdominal wall defects (primarilyexomphalos) (7.1%, 4/56), discordant aneuploidy (thesewere 2 cases of trisomy 21 and 18, respectively, and onecase of heterokaryotypic sex chromosome abnormality)(5.3%, 3/56), skeletal abnormalities/dysplasias (1.8%, 1/56),hydrops of unknown origin (1.8%, 1/56) and cystichygroma (1.8%, 1/56).For dichorionic twins, central nervous system

abnormalities represented the main reason for selectivetermination (37.5%, 12/32), followed by aneuploidy(31.3%, 10/32), abdominal wall/GIT defects (12.5%, 4/32), cardiac anomalies (9.3%, 3/32), skeletal abnormalities/

Table 1 Maternal characteristics of study population

MaternalCharacteristics MC Twins DC twins P-value

Age (years) 29.3 (17–42) 32.8 (21–42)EthnicityCaucasian 82.1% (46/56) 78% (25/32) P = 0.83Asian 10.7% (6/56) 15.6% (5/32) P = 0.52Black 5.6% (4/56) 6.4% (2/32) P = 0.87

Mode of conceptionSpontaneous 98.2% (55/56) 81.3% (26/32) P = 0.42IVF 1.8% (1/56) 18.7% (6/32) P = 0.006

ParityNulliparous 66.1% (37/56) 68% (22/32) P = 0.88Multiparous 33.9% (19/56) 32% (10/32) P = 0.83

Total number ofcases

56 32

© 2013 The Authors 115ANZJOG © 2013 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists

Selective fetal reduction in twins

dysplasias (6.3%, 2/32), Aicardi–Goutieres syndrome(3.2%, 1/32).To achieve selective reduction in monochorionic twins,

interstitial laser was used in 14.3% (8/56) of cases, bipolarcord occlusion in 39.3% (22/56) of cases andradiofrequency ablation in 46.4% (26/56) of cases. Theindications for each procedure are shown in Table 2.For interstitial laser and RFA, only local anaesthetic was

used for maternal analgesia. For BCO, epiduralanaesthesia was used in 27% (6/22) of cases. Atransplacental approach was required in 33% of cases inthe interstitial laser group, in 18% of cases in the BCOcohort and in 21% of cases in the RFA group. In all cases,the selected method of fetal reduction was successful inachieving asystole of the abnormal fetus. There were nomaternal complications with any of the techniques.There were 8 cases of selective fetal reduction in twins

using interstitial laser. The median gestational age at thetime of the procedure was 14 weeks (range 12–21 weeks).The preterm (<37 weeks) delivery rate was 12.5% (1/8).There was 1 (12.5%) pregnancy loss 24 h following theprocedure at 14 weeks of gestation. The median birth

weight at delivery was 2764 g (range 1306 g–3940 g).The live birth rate was 87.5% (7/8).There were 26 cases of selective fetal reduction using

RFA. The median gestational age at the time of theprocedure was 18 weeks (range 13–27 weeks). There were2 pregnancy losses (7.7%) within 2 weeks of theprocedure (<24 weeks). The median gestational age atdelivery was 36 + 4 weeks (range 27–41 weeks) with amedian birth weight of 2670 g (range 1240–3970 g). Thepreterm delivery rate was 26.9% (7/26). The live birth ratewas 92.3% (24/26).There were 22 cases of selective reduction using bipolar

cord occlusion. The median gestational age when theprocedure was performed was 20 weeks (range 17–32 weeks). There were 2 (9.1%) procedure-related losses(<24 weeks). The median gestational age at delivery was35 weeks (range 28–41 weeks) with a median birth weightof 2700 g (range 1420–4010 g). The preterm delivery ratewas 36.4% (8/22). The live birth rate was 90.9% (20/22).The perinatal outcomes for the various procedures arepresented in Table 3.For all fetal reduction techniques for monochorionic

twins, the overall pregnancy loss rate (<24 weeks) was8.9% (5/56).There were 32 cases of selective fetal reduction in the

DC twin cohort. For all DC twin pregnancies, anintracardiac KCl injection was used to perform selectivefetal reduction. All procedures were performed under localanaesthetic and were technically successful in all cases.There were no procedure-related losses and there were nomaternal complications. The median gestation at the timeof the procedure was 19 weeks (range 12–33 weeks). Themedian gestational age at delivery was 37 + 3 weeks (24–40 weeks) with a median birth weight of 2820 g (range740–3990 g). The preterm delivery rate was 7/32 (21.9%).The live birth rate was 100%.

Discussion

Discordancy for structural, chromosomal or geneticabnormalities raises both ethical and management issues.

Table 2 Indications for selective reduction

Indications for selective reductionMCtwins

DCtwins

CNS anomalies 34 12TRAP sequence 7 –CVS anomalies 5 3Abdominal wall/Gastrointestinal tractanomalies

4 4

Chromosomal anomalies. 3 10Skeletal abnormalities 1 2Hydrops 1 –Aicardi-Goutieres syndrome – 1Cystic hygroma 1 –

MC, monochorionic; DC, dichorionic; CNS, central nervoussystem; CVS, cardiovascular system; TRAP, twin reversed arterialperfusion.

Table 3 Procedure-related outcomes

MC twins DC twinsIL BCO RFA Intracardiac KCl

8 cases (14.3%) 22 cases (39.3%) 26 cases (46.4%) 32 cases

Median GA at procedure(wks)

14 (12–21) 20 (17–32) 18 (13–27) 19 (12–33)

Loss <2 weeks 12.5% 9.1% 7.7% 0%Preterm delivery 12.5% 36.4% 26.9% 21.9%Median BW (g) 2764 (1306 – 3940) 2700 (1420 – 4010) 2670 (1240 – 3970) 2820 (740 – 3990)Live birth rate 87.5% 90.9% 92.3% 100%Median GA at delivery (wks) 36 (26 – 40) 35 (28 – 41) 36 + 4 (27–41) 37 + 3 (24–40)

MC, monochorionic; DC, dichorionic; IL, interstitial laser; BCO, bipolar cord occlusion; RFA, radiofrequency ablation; KCl, potassiumchloride; BW, birth weight; GA, gestational age. wks, weeks.

116 © 2013 The Authors

ANZJOG © 2013 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists

E. Nobili et al.

Although considerations with regard to the ethical ormoral justification for the termination of a fetus are notdissimilar to a singleton pregnancy, there are additionalfactors that need to be considered in the context of amultiple pregnancy. These include the procedure-relatedrisk to the co-twin as well as the small maternal risks fromretention of a volume of nonviable fetal tissue followingthe selective reduction.Monozygotic twins account for a third of all twin

pregnancies but are associated with a higher incidence ofcomplications. There is a higher risk of perinatal mortality,early preterm delivery and fetal growth restriction inmonochorionic pregnancies. Structural anomalies are alsothree times higher compared with dizygotic twins andabout five times higher than in singleton pregnancies.Monochorionic twins are particularly at risk ofcardiovascular abnormalities5 compared with the generalpopulation (relative risk [RR], 9.18; 95% CI, 5.51–15.29).The incidence of fetal abnormalities has been reported

to be higher in multiple pregnancy than in singletons5–7

although the rate of anomalies in dizygotic twins is notincreased per twin. In a monozygotic twin gestation,although the rate of genetic and chromosomalabnormalities is similar to that of a singleton pregnancy,there is however an increased risk of structuralmalformations. In our study, monochorionic twins referredfor a structural malformation were more than 1.5 X morelikely to have a CNS abnormality compared withdichorionic twins. Although this may reflect referral bias,in our view CNS abnormalities appear to predominate inmonochorionic pregnancies.Sun et al.8 using matched population-based multiple

births data set from the United States showed that in 2.5%of cases, one fetus had a structural anomaly. In 1% ofcases, both fetuses had abnormalities. Compared with thecontrols, these investigators demonstrated that thepresence of an abnormal fetus significantly increased therisk of preterm delivery before 32 weeks, very low birthweight (<1500 g), small for gestational age as well as fetal,neonatal and infant death.The technique of selective fetal reduction in dichorionic

pregnancies is less complicated (intracardiac KCl injectionwith a thin gauge needle) compared with selectivetermination in monochorionic pregnancies, which requiresablation of blood flow in the umbilical cord. Interruptionof blood flow is required to prevent an acute intrafetaltransfusion as one fetus dies. Various techniques9 havebeen tried, including bipolar cord occlusion, injection ofvascular sclerosants and fetoscopic ligation of the umbilicalcord. All of these techniques carry considerableprocedure-related risks. Other techniques includeinterstitial laser10 and more recently, radiofrequencyablation.3,11,12 Previous publication from our unit showedthat interstitial laser was associated with higher procedure-related loss (27% per pregnancy and 24% per fetus),rupture of membranes and co-twin demise as well as an8% association with aplasia cutis congenita.4 The highprocedure-related loss rate in earlier series may have been

due to the inclusion of high-order multiples in contrast tothis study which only including twins. However, we havestopped using interstitial laser as a technique for selectivefetal reduction and use radiofrequency ablation as ourpreferred method for both early and mid-trimesterintervention.This is a large single-centre series of selective fetal

reduction in monochorionic and dichorionic twinsdiscordant for structural, chromosomal and geneticabnormalities. We have demonstrated that selective fetalreduction for structural, chromosomal and geneticindications is associated with excellent perinatal outcomesin dichorionic twin pregnancies. For monochorionic twins,our results are good with a live birth rate of 90.3%. Ouroverall pregnancy loss rate for monochorionic twins priorto 24 weeks of gestation was similarly good at 8.9%. Thehighest live birth rate was seen in the RFA cohort(92.3%), and the lowest in the interstitial laser cohort(87.5%). However, the number of cases in the interstitiallaser group is small.Evans et al13 published the largest series on selective

termination (using intracardiac KCl) in multiplepregnancies for structural or chromosomal anomalies.However, this study included high-order multiples andtherefore may not be directly comparable to our study.Nevertheless, the results for our dichorionic cohortcompare favourably with the results reported in that paper(mean gestation at delivery was 35 + weeks with a livebirth rate >90%). Furthermore, in a recent systematicreview by Rossi and Addario14 of 345 cases of selectivefeticide in monochorionic pregnancies from 12publications, the perinatal outcomes following theprocedure appeared better if it was performed >18 weeksof gestation irrespective of the chosen technique orindication for the procedure. These authors found thatcompared to BCO, RFA was associated with highersurvival (86 vs 82%) and lower co-twin demise rates. Ourfindings in this study support their conclusions andcompare very favourably to the figures reported in theirsystematic review. However, caution is warranted as otherauthors have found the converse. A more recent study byBebbington et al.15 comparing outcomes between bipolarcord occlusion and radiofrequency ablation demonstratedthat overall survival was 85.2% in the BCO group vs70.7% in the RFA group (P = 0.014). This was mainlydue to a significantly reduced survival rate (10.5%) in theRFA group compared with the BCO group (31.6%) forcases where delivery occurred before 28 weeks’ gestation.Premature rupture of the membranes and preterm labourwere more common in the BCO group (27.3 vs 13.7%and 22.4 vs 7%, respectively). These authors cautionedthat despite the smaller calibre of the instrument, RFA wasnot associated with a decrease in the overall complicationrate for selective termination procedures.Our study does have some limitations, in that the cases

seen in our institution probably reflect some degree ofreferral bias as we are acknowledged as a tertiary centreperforming complex interventional procedures for

© 2013 The Authors 117ANZJOG © 2013 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists

Selective fetal reduction in twins

monochorionic pregnancies. Over the last 3 years, we havemoved away from performing BCO other than in selectedcases at more advanced gestations in favour of RFAbecause of the simplicity of its technique and apparent goodoutcomes. The disadvantages of RFA relate mainly to thecost (approximately $1200) and length of the RFA needle.The radiofrequency needle conventionally used is only12 cm long, which may preclude its use in very obesepatients and in advanced gestation. Although longer needlesare available, when deployed their tines can result in amuch larger area of coagulation, which may not beappropriate at earlier gestations. Additional considerationsare the procedural skills required for early fetal intervention,which in our opinion are greater for this technique.In summary, this study demonstrates that selective

termination for fetal anomalies in both monochorionic anddichorionic pregnancies is safe and effective with relativelylow loss and prematurity rates. Whilst there is little debate,if any, about the best technique to use in dichorionictwins, the choice of any one particular technique formonochorionic twins still depends very much on theability of the technique to reliably occlude blood flowwithin the umbilical cord. On the basis of our experienceto date (we have now performed more than 100 cases ofRFA with similar outcomes), our preference is for RFA.However, larger series are required to further validatethese early results.

Acknowledgement

EN was supported by a grant from the A Griffini andJ Miglierina Provincia di Varese Foundation. SK wasfunded by the Imperial College Healthcare NHS Trustcomprehensive Biomedical Research Centre (BRC)scheme.

Conflict of Interests

None.

References1 Pasquini L, Pontello V, Kumar S. Intracardiac injection of

potassium chloride as method for feticide: experience from asingle UK tertiary centre. BJOG 2008; 115 (4): 528–531.

2 Taylor MJ, Shalev E, Tanawattanacharoen S et al.Ultrasound-guided umbilical cord occlusion using bipolar

diathermy for Stage III/IV twin-twin transfusion syndrome.Prenat Diagn 2002; 22 (1): 70–76.

3 Paramasivam G, Wimalasundera R, Wiechec M et al.Radiofrequency ablation for selective reduction in complexmonochorionic pregnancies. BJOG 2010; 117 (10): 1294–1298.

4 O’Donoghue K, Barigye O, Pasquini L et al. Interstitial lasertherapy for fetal reduction in monochorionic multiplepregnancy: loss rate and association with aplasia cutiscongenita. Prenat Diagn 2008; 28 (6): 535–543.

5 Bahtiyar MO, Dulay AT, Weeks BP et al. Prevalence ofcongenital heart defects in monochorionic/diamniotic twingestations: a systematic literature review. J Ultrasound Med2007; 26 (11): 1491–1498.

6 Luke B, Keith LG. Monozygotic twinning as a congenitaldefect and congenital defects in monozygotic twins. FetalDiagn Ther 1990; 5 (2): 61–69.

7 Hardin J, Carmichael SL, Selvin S et al. Increased prevalenceof cardiovascular defects among 56,709 California twin pairs.Am J Med Genet A 2009; 149A (5): 877–886.

8 Sun LM, Chen XK, Wen SW et al. Perinatal outcomes ofnormal cotwins in twin pregnancies with one structurallyanomalous fetus: a population-based retrospective study. Am JPerinatol 2009; 26 (1): 51–56.

9 Challis D, Gratacos E, Deprest JA. Cord occlusion techniquesfor selective termination in monochorionic twins. J PerinatMed 1999; 27 (5): 327–338.

10 Jolly M, Taylor M, Rose G et al. Interstitial laser: a new surgicaltechnique for twin reversed arterial perfusion sequence in earlypregnancy. BJOG 2001; 108 (10): 1098–1102.

11 Lee H, Wagner AJ, Sy E. Efficacy of radiofrequency ablationfor twin-reversed arterial perfusion sequence. Am J ObstetGynecol 2007. 196 (5): p. 459 e1-4.

12 Moise KJ Jr, Johnson A Moise KY, Nickeleit V. Radiofrequencyablation for selective reduction in the complicatedmonochorionic gestation. Am J Obstet Gynecol 2008. 198 (2): p.198 e1-5.

13 Evans MI, Goldberg JD, Horenstein J et al. Selectivetermination for structural, chromosomal, and mendeliananomalies: international experience. Am J Obstet Gynecol 1999;181 (4): 893–897.

14 Rossi AC, D’Addario V. Umbilical cord occlusion for selectivefeticide in complicated monochorionic twins: a systematicreview of literature. Am J Obstet Gynecol 2009; 200 (2): 123–129.

15 Bebbington MW, Danzer E, Moldenhauer J et al.Radiofrequency ablation vs bipolar umbilical cord coagulationin the management of complicated monochorionic pregnancies.Ultrasound Obstet Gynecol 2012; 40 (3): 319–324.

118 © 2013 The Authors

ANZJOG © 2013 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists

E. Nobili et al.