Archives ofDisease in Childhood 1990; 65: 657-661

Fetofetal transfusion syndrome: do the neonatalcriteria apply in utero?

N M Fisk, A Borrell, C Hubinont, Y Tannirandorn, U Nicolini, C H Rodeck

AbstractThirteen fetuses (five twin, one triplet) werecompromised by fetofetal transfusion syn-drome in six pregnancies, five in the mid tri-mester, and one in the third trimester. Thisdiagnosis, which was suspected because ofultrasound findings of discordant growth, dis-cordant amniotic fluid volumes, concordantexternal genitalia, and monochorial placenta-tion, was confirmed postnatally in each. Ninefetuses underwent blood sampling to aiddiagnosis and assessment of fetal welibeing.In contrast to fetofetal transfusion syndromeinvestigated postnatalHy, a difference inhaemoglobin concentration of 50 g/l or morein utero was found in only one pregnancy,which was near term, although ali had fetalerythroblastaemia and a difference in weightof 20% or more. In vivo confirmation ofshared circulation was achieved in two pre-gnancies by transfusing adult Rh negative redcelis into the smaller fetus and then detectingthem by Kleihauer testing in blood aspiratedfrom the larger. Invasive procedures alsoyielded information on fetal blood gas mea-surements (acidaemia in four and hypoxaemiain six) and amniotic pressure (raised in two).We suggest that comparison of haemoglobinconcentrations is inaccurate in fetofetal trans-fusion syndrome in utero, the diagnosis ofwhich may necessitate detection of a sharedcirculation using a marker such as adult redcelHs.

Fetal Medicine Unit,Royal PostgraduateMedical School,Institute of Obstetricsand Gynaecology,Queen Charlotte's andChelsea Hospital,Goldhawk Road,London W6 OXGN M FiskA BorrellC HubinontY TannirandornU NicoliniC H RodeckCorrespondence to:Dr Fisk.Accepted 14 February 1990

Although vascular communications are found inthe placentas of almost all monochorial twinpregnancies, clinical signs of twin to twin trans-fusion occur in only 4-26%.' 2 The pathophy-siology of this condition, known as fetofetaltransfusion syndrome, is poorly understood. Ingeneral however, the donor twin becomesanaemic, growth is retarded, and it developsoligohydramnios, whereas the recipientdevelops hydrops, cardiomegaly, and poly-hydramnios.3Most reported cases have been diagnosed

postnatally or at necropsy. The standard neo-

natal criteria comprise a difference in cordhaemoglobin concentrations of 50 g/l or more,

and a difference in birth weights of 20% or

more,4-6 although the latter criterion is notnecessarily present in the acute form of fetofetaltransfusion syndrome that occurs in labour.7Certain antenatal ultrasound appearances are

highly suggestive of fetofetal transfusionsyndrome,> "1 principally discrepant growth,and amniotic fluid volumes between monochor-ionic diamnionic twins.8 iOu

1 Discordant

growth, however, is a common complication oftwin pregnancies, and causes other than feto-fetal transfusion syndrome exist for discordantamniotic fluid volume. Signs of hydrops fetalison ultrasound scan are occasionally found in therecipient,o1 11 rarely in the donor,4 and excep-tionally in both.'2 As all the above ultrasoundfeatures are non-specific, more accuratemethods of antenatal diagnosis are needed if itsassociated high perinatal mortality rate of40-81% is to be reduced.8'0 13

Pregnancies complicated by fetofetal trans-fusion syndrome will increasingly be evaluatedin utero now that some of its signs-forexample, intrauterine growth retardation, hyd-rops, and abnormal amniotic fluid volume, havebecome accepted indications for fetal bloodsampling. Although precise criteria exist for thepostnatal diagnosis of fetofetal transfusion syn-drome, none have as yet been defined for itsdiagnosis in utero. We describe our experienceof detailed fetal assessment in fetofetal transfu-sion syndrome using ultrasound scanning andfetal blood sampling.

Patients and methodsAmong multiple pregnancies referred over athree year period for investigation of discordantfetal growth, the diagnosis of fetofetal trans-fusion syndrome was strongly suspected in six(five twin, one triplet) because of ultrasono-graphic findings of discrepant amniotic fluidvolume, concordant external genitalia, and amonochorionic placenta, in addition to discor-dant intrauterine growth retardation. Gesta-tional ages ranged from 17-36 weeks, with onlyone patient beyond 27 weeks.

Initial assessment comprised detailed ultra-sound scanning, examining in particular the siteof the placenta, the thickness of the septum,'4the depth of the amniotic fluid pools, and thepresence of hydrops. In four cases, pulsedDoppler studies of umbilical arterial flow velo-city waveforms were undertaken (Acuson,Mountain View, California, USA). Theindications for fetal blood sampling were theinvestigation (karyotype and blood gas analysis)of severe intrauterine growth retardation in six,of hydrops in three, and of polyhydramnios infour, and the exclusion of coagulopathy in onetwin after the intrauterine death of its mono-chorial sibling two to three weeks before.9 15 16Additional reasons for fetal blood samplingwere in all cases to assess fetal wellbeing and toattempt to confirm the diagnosis by detectingdifferences in haemoglobin concentrations. Allpatients gave written informed consent. Fetal

657

on April 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.65.7_S

pec_No.657 on 1 July 1990. D

ownloaded from

Fisk, Borrell, Hubinont, Tannirandorn, Nicolini, Rodeck

blood sampling was carried out with a 20 gaugeneedle inserted under ultrasound guidance. Thesite of sampling was the placental cord insertionin five, the intrahepatic vein in one, and a freeloop of cord in three. Gross polyhydramniosprecluded fetal blood sampling in the otherfetus(es) in cases 3 and 6.The purity and fetal source of samples was

checked at the time by separate red cell volumedistributions on a particle size analyser (CoulterChannelyzer C-256, Coulter Electronics) andlater confirmed by the Kleihauer test.'7 Thepacked cell volume and haemoglobin concentra-tion in fetal blood were determined by a CoulterS Plus Counter. Nucleated red cells and reticu-locyte counts were calculated from ratiosobtained from blood films, which in turn wereapplied to results from the automated cellcounter as described previously.'8 Blood gasanalysis was performed on 250 ,ul of fetal bloodcollected into heparinised capillary tubes. 9Published normal ranges were used forhaemoglobin,20 packed cell volume,2'erythroblasts,22 and blood gases.23 Karyotypeswere determined in fetal blood after short termfibroblast culture.To document the presence of placental vascu-

lar anastomoses in vivo, adult blood (group 0Rh negative and seronegative for cytomegalo-virus was transfused into two donor fetuses.Fetal blood samples were then taken from theirpresumed recipient sibling to determine, usingthe Kleihauer technique, the passage of trans-fused adult red cells from one twin to the other.The needle used for taking the sample of fetal

blood was also used to provide access to theamniotic cavity. Four patients with grosspolyhydramnios underwent therapeutic drain-age of amniotic fluid to reduce amnioticpressure.24 25 Amnioinfusion was carried out inone fetus to allow a thin septum to be seen; ithad previously not been visible, but was sus-pected because of fetal immobility.'4

Necropsy was performed by one of two peri-natal pathologists in all but one case of fetal loss,in which consent was withheld. Postpartumevaluation of the configuration of placental

membranes confirmed monochorionic diamnio-tic or triamniotic placentas in all. Meticulousinjection studies were not performed, althoughshared areas on the chorionic surface were eva-luated macroscopically. Pathological extrame-dullary haemopoiesis was defined as histologicalevidence of increased hepatic haemopoiesis outof proportion to that seen in normal fetuses ofsimilar gestational age.For reasons of clarity, sibling fetuses are

denoted by the letters A to C, with the recipientfetus in each pregnancy referred to as fetus A.

ResultsClinical details and outcome are shown intable 1. No immediate complications occurredas a result of the invasive procedures. Resultsof laboratory studies on fetal blood are shownin table 2. The karyotype in all in fetal bloodwas normal, and a normal karyotype wasobtained from triplet fetuses B and C by cultureof amniotic fluid cells obtained by amniocente-sis. Each pregnancy was concordant for fetalsex. Clotting studies (platelets, thrombin time,fibrinogen, and fibrin degradation products)from the surviving fetus of case 2 werenormal.

Six and 18 ml of donor blood (packed cellvolumes of 0-72 and 0-87) were transfused intothe donor twins of cases 4 and 5, respectively.Kleihauer testing of fetal blood then obtainedfrom the recipient twins 25 minutes (case 4) and16 minutes (case 5) later showed that 145 and 70cells/1000, respectively, had stained as adultcells. In both cases the needle tip was seenclearly within the umbilical vein of the recipientsuggesting that appreciable maternal contami-nation was unlikely. This procedure was carriedout in case 4 to document the sharing of circula-tions in utero, in view of the similar fetalhaemoglobin concentrations. In contrast, thiswas done in case 5 at 36 weeks to improve thecondition of the fetus before delivery bycorrecting polycythaemia and anaemia. In case5 the recipient was sampled first, polycythaemiawas confirmed, and 30 ml of blood was with-

Table I Clinical details and outcome in six pregnancies complicated by fetofetal transfusion syndrome. Gestational age at time of blood sampling shown inparentheses

Case 1 (23 weeks) Case 2 (27 weeks) Case 3 (21 weeks) Case 4 (21 weeks) Case 5 (36 weeks) Case 6 (26 weeks)

Fetus A Fetus B Fetus A Fetus B Fetus A Fetus B Fetus A Fetus B Fetus A Fetus B Fetus A Fetus B Fetus C

Volume of amnioticfluid

Ultrasonic growth(biparietal diameterand abdominalcircumference)

End diastolicfrequencies detectedby umbilicalartery Doppler

Hydrops presentOutcome

Birth weight (g)Extramedullary

haemopoiesisat necropsy

Heart: body weightratio at necropsy

Increased Nil Increased Normal Increased Nil

Normal <5th Intrauterine <5th NormalCentile death Centile

Increased Nil Increased Reduced Increased Normal Reduced

<5th NormalCentile

<5th NormalCentile

<5th Normal NormalCentile

<5thCentile

Not Not Intrauterine Not Present Present Present Absent Present Present Absent Present Absentmeasured measured death measured

No No Yes YesMiscarried one Preterm labour:

day later Stillbirth Neonataldeath

NoIntrauterine

six days

560 420 (444) 730 355No No No Yes No necrops'

No No No Yes Nodeaths - Intrauterine Live birth Livebirth

later death three Elective caesareandays later section next day

Miscarried four days later200 480 400 2510 1658

y Yes No Not applicable

Yes No NoPreterm labour five days laterStillbirth Neonatal Stillbirth

deathday 8

1680 878 430Yes No Yes

0-015 0 005 (0 013) 0-007 No necropsy 0-012 0-003 Not applicable

658

0 006 0 007 0 005

on April 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.65.7_S

pec_No.657 on 1 July 1990. D

ownloaded from

Fetofetal transfusion syndrome: do the neonatal criteria apply in utero?

Table 2 Laboratory results on fetal blood. Case 2-fetus A, Case 3-fetus B, and Case 6-fetuses B and C did not havesamples taken. Results above the normal range are shown in bold type, those below in italics

Case I Case 2 Case 3 Case 4 Case 5 Case 6

FetusA FetusB FetusB FetusA FetusA Fetus B FetusA FetusB FetusA

Haematology:Packed cell volume 0-53 0 47 0-36 0-52 0-45 0-47 0-68 0 31 0-61Haemoglobin (g/l) 186 166 128 180 110 153 230 99 185Nucleated red cell count

(x 109/l) 52.9 42-0 7.3 8-7 1-9 4-8 2-9 17-0 27-5Reticulocytes (x 109/l) 0-71 0-51 0-45 0-48 0-37 0-67 0.46 0-31 0-37As% 16-5 13-5 19-0 13-5 11-0 19-7 7-8 12-2 8-6

Blood gases:pH 7-27 7-30 7-38 7-43 7-37 7-23 7-34 7-39 7-19PCO2(kPa) 6-66 5-98 511 4 68 5-96 8-81 5-16 4-93 7-15P02 (kPa) 186 2-39 5.10 3 63 3 02 1 52 4 47 4-08 141Baseexcess -4-5 -4*3 -1-8 0 -0-7 -2-0 -5-2 -4-3 -8-4

Fetus A, Case 4, had samples taken after transfusion of the sibling fetus.

drawn to reduce the fetal packed cell volume.The donor was then sampled and transfused, itspacked cell volume increasing from 0-31 to0-36. The volume of red cells transfused wascalculated as less than that removed from therecipient in case immediate fetofetal transfusionworsened the recipient's polycythaemia. Therecipient was then resampled, a further 6 ml ofblood withdrawn, and 15 ml normal salineinstilled intravascularly. Dilutional exchangetransfusion in utero was abandoned at this stagebecause of a transient bradycardia. Umbilicalartery flow velocity waveforms remained nor-mal throughout. At elective delivery the nextday, Kleihauer testing of the recipient's cordblood showed that 160/1000 cells were adult.The donor's neonatal course was uncompli-cated, but the recipient required dilutionalexchange transfusion after the packed cellvolume peaked at 0-72; the infant subsequentlydeveloped necrotising enterocolitis, andrecovered with conservative management.There were 11 fetal or perinatal losses, giving

a total perinatally related wastage of 85%.The diagnostic criteria by which all these

pregnancies were considered to have beencomplicated by fetofetal transfusion syndromeare shown in table 3. Discordant growth andamniotic fluid volumes in the presence of mono-chorionic placentas was seen on ultrasono-graphy in all, and hydrops was noted in threerecipient fetuses. After delivery, fetofetal trans-fusion syndrome was confirmed in each bydifferences in fetal weight of 20% or more. Inthe four that underwent necropsy, extramedul-lary haemopoiesis in the donor, with or withoutrelative macrocardia in the recipient, providedadditional confirmation of the diagnosis.26 In

the remaining two, additional confirmation wasachieved by the difference in haemoglobin con-centrations (case 5), and the clinical appearanceat delivery (case 3).3

DiscussionFetal haematology in these pregnancies withfetofetal transfusion syndrome suggests that,unlike postnatally, a haemoglobin difference of50 g/l or more is not necessarily found in utero.In this series, only one of three twin pregnanciesin which both fetuses were sampled showedmajor differences in haemoglobin values (case5), and this was in a pregnancy with a chronicwell compensated fetofetal transfusion syn-drome sampled late in the third trimester. Incontrast, the remaining two without major dif-ferences in haemoglobin concentrations were atmuch earlier gestations (21 and 23 weeks). Case1 showed a degree of polycythaemia in bothfetuses, while in case 4, polycythaemia wasfound in the donor and not in the recipient. Inthis atypical case, however, a marginalvelamentous cord insertion with abnormalumbilical Doppler flow implicated deficientplacentation in the donor's intrauterine growthretardation, itself a known cause for polycythae-mia. In the two multiple pregnancies in whichonly the recipient was sampled, raised haemo-globin concentrations and packed cell volumeswere consistent with the diagnosis of fetofetaltransfusion syndrome. Because of the abovediscrepancies, and the existence of othermechanisms for polycythaemia, however, iso-lated polycythaemia in one twin or tripletcannot be considered diagnostic.Rausen et al were the first to suggest that a

Table 3 Presence ofdiagnostic features in six pregnancies complicated byfetofetal transfusion syndrome

Case Ultrasound measurements Polyhydramnios Hydrops in Difference in Signs ofshared circulation* Difference in Extramedullary Difference inNo of biparietal diameter" in one sac and recipient haemnoglobin birth weight of haemopoiesis heart:body weight

and abdominal not the other(s) concentration In vivo In vitro >-20% between in donor ratio of :500/ocircumference4° <5th (discordant of 50 gll fetuses betveenfetusescentile in onefetus volume) between at necropsyand not the other(s) fetuses(discordant growth)

I Yes Yes No No Not applicable Yes Yes No Yes2 Yes Yes Yes Not applicable Not applicable No Not applicablet Yes Yes3 Yes Yes No Not applicable Not applicable Not applicable Yes Not applicable Not applicable4 Yes Yes No No Yes Yes Yes No Yes5 Yes Yes Yes Yes Yes No Yes Not applicable Not applicable6 Yes Yes Yes Not applicable Not applicable No Yes Yes No

'Signs of shared circulation in vivo means that adult red cells were detected in the recipient after adult blood had been transfused into the donor, whereas in vitro sharedcirculation refers to the presence of placental superficial anastomoses.tRetrospective calculation using centile charts of the estimated fetal weight of the surviving twin suggested a difference of :200/o at that time.

659

on April 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.65.7_S

pec_No.657 on 1 July 1990. D

ownloaded from

Fisk, Borrell, Hubinont, Tannirandorn, Nicolini, Rodeck

difference of 50 g/l in neonatal haemoglobinconcentrations was diagnostic of fetofetal trans-fusion syndrome.4 In their series, 19 of 130monochorionic twins had such a difference,while no set of dichorionic twins had a discrep-ancy of greater than 33 g/l. This criterion for thediagnosis of fetofetal transfusion syndrome hassince been widely adopted.5-7 Haemoglobin dis-cordance has, however, also been reported indichorionic twins.5 A recent study reported thatseven of 13 sets of twins with neonatal haemo-globin discordance of more than 50 g/l were infact dichorionic.27 Fetofetal transfusion syn-drome occurs almost exclusively in monochorio-nic placentas, although rarely vascularcommunications have been reported in fuseddichorionic placentas.2 On the other hand,intrauterine growth retardation is common indichorionic twins, so that haemoglobin discor-dance in dichorionic twins is far more likely toreflect polycythaemia secondary to intrauterinegrowth retardation. Tan et al described 35 setsof twins with the required haemoglobin differ-ence, but made no mention of placentalchorionicity.6 In nine of their pregnancies thepolycythaemic twin was smaller than theanaemic twin, suggesting that several cases ofdichorionic twins with intrauterine growthretardation were falsely diagnosed as havingfetofetal transfusion syndrome. Although thestandard neonatal criteria have recently beenquestioned,27 it seems reasonable to apply themat birth provided that they are applied only tomonochorionic twins in which the heavier twinhas the higher haemoglobin concentration.

This discrepancy between the classic post-natal findings and our observations made inutero may be related to differences in gestationor the duration and degree of shunting, or both.Presentation in the midtrimester suggests amore acute form of fetofetal transfusion syn-drome in which there may have been insuffi-cient time for haematological compensation.Alternatively, the considerable mixing of thecirculations implicit in large shunts may notnecessarily be reflected in differences in hae-moglobin concentrations.

Evidence of erythroblastosis (increasednucleated red cells and reticulocytes) was foundin all the fetuses sampled, usually to a greaterextent in the donor. Erythroblastaemia in therecipient could result either from fetofetal trans-fusion of donor erythroblasts,28 or from theaction of donor erythropoietin on the recipientmarrow.29 This latter suggestion would explainour finding of extramedullary haemopoiesis insome recipients. In addition, erythroblastaemiain the recipient may be secondary to hypoxia,which we found in four of five of the recipientssampled.An alternative approach to diagnosis is in vivo

demonstration of appreciable sharing of circula-tions. This was achieved in two pregnancies bygiving a readily detectable safe marker withsubsequent demonstration of its passage insubstantial quantities into the recipient. Thefinding of a considerable proportion of adult redcells in the recipients (14-5 and 7% of red cellsin cases 4 and 5, respectively) after only shortintervals suggests that a large amount of transfer

occurred rapidly. We found no such transfer inan Rh alloimmunised monochorionic twinpregnancy uncomplicated by fetofetal trans-fusion syndrome in whom intrauterine trans-fusion was indicated for fetal anaemia, but thisis anecdotal evidence and the techniquewarrants further evaluation in the diagnosis offetofetal transfusion syndrome in utero.

Fetal blood sampling in this series suppliedadditional information about fetal wellbeing.Four of nine fetuses were acidaemic, while twoothers in addition to these four were alsohypoxaemic. As with other high risk preg-nancies, fetal acid base balance may contributeto the assessment of fetal compromise and to thetiming of intervention or delivery, or both.

This series confirms the extremely high mort-ality (85%) in fetofetal transfusion syn-drome,'3 30 and reflects both the severity ofdisease and the early gestation of the patientsreferred. All fetal losses in this series occurredbefore 28 weeks, and survival after diagnosis offetofetal transfusion syndrome in the secondtrimester is most unusual. A 100% perinatalmortality rate has been reported in a series ofeight monochorial pregnancies complicated bypolyhydramnios in the midtrimester.30 Thusour finding of fetal hypoxaemia or acidaemia,or both, in four of five pregnancies that endedin fetal loss is hardly surprising. It seemsunlikely that the losses in our series were relatedto the invasive procedures; most complicationsof fetal blood sampling occur immediately, andnone were observed here.

In view of the uniformly poor outcome offetofetal transfusion syndrome in the secondtrimester, aggressive therapeutic measures maybe warranted. Laser ablation of superficialplacental anastomoses has been investigated inanimal studies31; it is, however, the deeper andnot the more accessible superficial anastomosesthat seem to be responsible for the haemo-dynamic imbalance in fetofetal transfusion syn-drome.9 28 Selective fetocide of one twin hasbeen reported in three pregnancies with feto-fetal transfusion syndrome, two of whichresulted in live singletons born at full term.32-34In no case was there evidence of hydrops orother fetal compromise, this drastic procedurebeing carried out primarily to allow prolonga-tion of pregnancies threatened by gross poly-hydramnios. In such cases simpler treatmentsdeserve consideration. Serial amniocentesis hasbeen recommended to allow symptomatic reliefand to delay preterm labour,35 although somehave found it unsuccessful.3 3 We haverecently shown that removal of small quantitiesof amniotic fluid in gross polyhydramnios re-stores raised amniotic fluid pressure tonormal.25 Although this procedure seems toalleviate maternal symptoms, it is not knownhow long thereafter it takes for amniotic press-ure to rise again. Treating the mother withindomethacin has also been suggested to allevi-ate polyhydramnios,36 as it reduces fetal urineoutput.37 This approach is contraindicated inthe presence of oligohydramnios in one twin,36however, a feature in five of the six pregnanciesreported here.An alternative to selective fetocide is sug-

6V0

on April 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.65.7_S

pec_No.657 on 1 July 1990. D

ownloaded from

Fetofetal transfusion sndrome: do the neonatal criteria apply in utero? 661

gested by the report of Vetter and Schneider3"who noted resolution of fetofetal transfusionsyndrome in two pregnancies after iatrogenicpuncture of a chorionic plate vessel at amnio-centesis with consequent intra-amniotichaemorrhage. Repeated phlebotomy of the reci-pient and transfusion of the donor has beenconsidered as a treatment for fetofetal trans-fusion syndrome.33 The degree of shunting thatwe found in vivo in cases 4 and 5 wassubstantial, however, indicating that serial 'giveand take' transfusion would be unlikely to be oflong term benefit. Given that the hydropicrecipient twin seems relatively more at risk thanits donor, venepuncture of the recipient alonewarrants consideration as a treatment for feto-fetal transfusion syndrome.

NMF was the recipient of a Gladys Dodds Fellowship from theRoyal College of Obstetricians and Gynaecologists and aForeman Fellowship from Royal Prince Alfred Hospital, Sydney,Australia. AB was supported by a CIRIT Fellowship, Generalitatde Catalunya, and the Prenatal Diagnosis Unit, University ofBarcelona. CH holds a Royal Society and a Council of EuropeAward. YT acknowledges the Anandamahidol Foundation andChulalongkorn University, Thailand.

1 Galea P, Scott JM, Goel KM. Feto-fetal transfusion syn-drome. Arch Dis Child 1982;57:781-3.

2 Robertson EG, Neer KJ. Placental injection studies in twingestation. Am J Obstet Gynecol 1983;147:170-4.

3 Naeye RL. Human intrauterine parabiotic syndrome and itscomplications. N Engl J Med 1%3;268:804-9.

4 Rausen AR, Seki M, Strauss L. Twin transfusion syndrome.J Pediatr 1965;66:613-28.

5 Abraham JM. Intrauterine feto-fetal transfusion syndrome:clinical observations and speculations on pathogenesis. ClinPediatr 1967;6:405-10.

6 Tan KL, Tan R, Tan SH. The twin transfusion syndrome.Clin Pediatr 1979;18:111-4.

7 Klebe JC, Ingomar CJ. The fetofetal circulation duringparturition illustrated by the interfetal transfusion syn-drome. Pediatrics 1972;49:112-6.

8 Brown DG, Benson CB, Driscoll SG, Doubilet PM. Twin-twin transfusion syndrome: sonographic findings. Radio-logy 1989;170:61-3.

9 Benirschke K, Kim CK. Multiple pregnancy. N Engl J Med1973;288:1276-84.

10 Babbington MV, Wittmann BK. Fetal transfusion syndrome:antenatal factors predicting outcome. AmJ Obstet Gynecol1989;160:913-5.

11 Brennan JN, Diwan RJ, Rosen MG, Bellon EM. Fetofetaltransfusion syndrome: prenatal ultrasonographic diagnosis.Radiology 1982;143:535-6.

12 Macafee CAJ, Fortune DW, Beischer NA. Non-immunologi-cal hydrops fetalis. Journal of Obstetrics and Gynaecology ofthe British Commonwealth 1970;77:226-37.

13 Pretorius DH, Manchester D, Barkin S, Parker S, NelsonTR. Doppler ultrasound of twin twin transfusion syn-drome. J Ultrasound Med 1988;7:117-24.

14 Mahony BS, Filly RA, Callen PW. Amnionicity and chorioni-city in twin pregnancies: prediction using ultrasound.Radiology 1985;155:205-9.

15 Moore CM, McAdams AJ, Sutherland J. Intrauterine dis-seminated intravascular coagulation: a syndrome ofmultiple pregnancy with a dead twin fetus. Pediatrics 1969;74:523-8.

16 Cox WL, Forestier F, Capella-Pavlovsky M, Daffos F. Fetalblood sampling in twin pregnancies. Fetal Therapy 1987;2:101-8.

17 Mibashan RS, Rodeck CH, Thompson JK. Prenatal diag-nosis of the haemophilias. In: Bloom AL, ed. Thehaemophilias. Edinburgh: Churchill Livingstone, 1982:176-96.

18 Fisk NM, Tannirandorn Y, Santolaya J, Nicolini U, LetskyEA, Rodeck CH. Fetal macrocytosis in association withchromosomal abnormalities. Obstet Gynecol 1989;74:611-6.

19 Nicolini U, Santolaya J, Fisk NM, et al. Changes in fetal acidbase status during intravascular transfusion. Arch Dis Child1988;63:710-4.

20 Nicolaides KH, Clewell WH, Mibashan RS, Soothill PW,Rodeck CH, Campbell S. Fetal haemoglobin measurementin the assessment of red cell isoimmunisation. Lancet1988 g: 1073-5.

21 Cox WL, Daffos F, Forestier F, et al. Physiology andmanagement of intrauterine growth retardation: a biologicapproach with fetal blood sampling. Am J Obstet Gynecol1988;159:36-41.

22 Millar DS, Davis LR, Rodeck CH, Nicolaides KH,Mibashan RS. Normal blood cell values in the early mid-trimester fetus. Prenat Diagn 1985;5:367-73.

23 Soothill PW, Nicolaides KH, Rodeck CH, Campbell S.Effect of gestational age on fetal and intervillous bloodgases and acid-base values in human pregnancy. FetalTherapy 1986;1:168-75.

24 Nicolini U, Fisk NM, Talbert DG, et al. Intrauterinemanometry: technique and application to fetal pathology.Prenat Diagn 1989;9:243-54.

25 Fisk NM, Tannirandorn Y, Nicolini U, Talbert DG, RodeckCH. Amniotic pressure in disorders of amniotic fluidvolume. Obstet Gynecol (in press).

26 Naeye RL. The fetal and neonatal development of twins.Pediatrics 1964;33:546-53.

27 Danskin FH, Neilson JP. Twin-to-twin transfusion syn-drome: what are the appropriate diagnostic criteria? Am JObstet Gynecol 1989;161:365-9.

28 Aherne W, Strong JS, Corney G. The structure of theplacenta in the twin transfusion syndrome. Biol Neonat1968;12:121-35.

29 Dudley DKL, D'Alton ME. Single fetal death in twingestation. Semin Perinatol 1986;10:65-72.

30 Weir PE, Ratten GJ, Beischer NA. Acute polyhydramnios-a complication of monozygous twin pregnancy. BrJ' ObstetGynaecol 1979;86:849-53.

31 DeLia JE, Gukierski MA, Lundergan DK, Kochenour NK.Neodymium: yttrium-aluminum-garnet laser occulsion ofrhesus placental vasculature via fetoscopy. Am J ObstetGynecol 1989;160:485-9.

32 Wittmann BK, Farquharson DF, Thomas WDS, BaldwinVJ, Wadsworth L. The role of feticide in the managementof severe twin transfusion syndrome. Am J Obstet Gynecol1986;155:1023-6.

33 Weiner CP. Diagnosis and treatment of twin to twintransfusion in the mid-second trimester of pregnancy. FetalTherapy 1987;2:71-4.

34 Chescher NC, Seeds JW. Polyhydramnios and oligo-hydramnios in twin gestations. Obstet Gynecol 1988;71:882-4.

35 Feingold M, Cetrulo CL, Newton ER, Weiss J, Shakr C,Shmoys S. Serial amniocenteses in the treatment of twin totwin transfusion complicated with acute polyhydramnios.Acta Genet Med Gemellol 1986;35:107-13.

36 Lange IR, Harman CR, Ash KM, Manning FA,Menticoglou S. Twins with hydramnios: treating pre-mature labor at source. Am J Obstet Gyncol 1989;160:552-7.

37 Kirshon B, Moise KJ, Wasserstrum N, Ou C-N, Huhta JC.Influence ofshort-term indomethacin therapy on fetal urineoutput. Obstet Gynecol 1988;72:51-3.

38 Vetter K, Schneider KTM. Iatrogenous remission of twintransfusion syndrome. Am J Obstet Gynecol 1988;158:221.

39 Campbell S, Newman GB. Growth of fetal biparietal diameterduring normal pregnancy. Br J Obstet Gynaecol 1971;78:513-9.

40 Campbell S, Wilkin D. Ultrasonic assessment of fetalabdominal circumference in the estimation of fetal weight.BrJ Obstet Gynaecol 1975;82:689-91.

on April 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.65.7_S

pec_No.657 on 1 July 1990. D

ownloaded from