489

Twin-twin transfusion syndrome (TTTS) occurs in ap-proximately 10% of monochorionic pregnancies. Thedisease is thought to result from an unbalanced exchangeof blood through vascular anastomoses. Although it isknown that vascular anastomoses can be present in atleast 85% of monochorionic placentas,1-3 the reasons forthe development of the syndrome in 10% of patients andnot in the other 90% of patients remain unclear. Classicsurgical pathologic studies showed that placental anasto-moses could be either superficial (arterioarterial or ven-ovenous) or deep arteriovenous.2,4 By definition,superficial anastomoses lack a terminal end in the pla-centa, whereas deep anastomoses are involved in the

shared perfusion of a cotyledon by both twins. However,although the anatomic description of the anastomoseshas been known, the specific role of each type of anasto-mosis or group of anastomoses in the development ornondevelopment of TTTS has been controversial. Deepanastomoses establish a unidirectional flow of blood be-tween the fetuses, whereas superficial anastomoses mayhave bidirectional blood flow, which is postulated to havea “protective” role against the syndrome.5-8

Since 1997, we established a referral system to receiveall placentas of patients with TTTS who are treated at ourinstitution with selective laser photocoagulation of com-municating vessels (SLPCV).9 Patients were asked toarrange for shipment of their placentas fresh, after deliv-ery, via express mail in a water-tight container. The natureof the vascular anastomosis, which was assessed endo-scopically during the operation, was reassessed by surgi-cal pathologic results. Placentas from non-TTTSmonochorionic twin pregnancies that were delivered toour institution during that time period were also in-cluded for comparison. The purpose of this study was toassess the value of a proposed classification of mono-chorionic placentas in relationship to the developmentof TTTS.

From the Florida Institute for Fetal Diagnosis and Therapy, St Joseph’sWomen’s Hospital.Received for publication August 31, 2001; revised January 3, 2002; ac-cepted February 20, 2002.Reprint requests: Rubén A. Quintero, MD, 13601 Bruce B. Downs Blvd,Suite 250, Tampa, FL 33613. E-mail: yvrq@aol.com or http://www.fe-talmd.com© 2002, Mosby, Inc. All rights reserved.0002-9378/2002 $35.00 + 0 6/1/124280doi:10.1067/mob.2002.124280

Placental types and twin-twin transfusion syndrome

Carlos Bermúdez, MD, Carlos H. Becerra, MD, Patricia W. Bornick, RN, MSN, Mary H. Allen,RN, Jorge Arroyo, MD, and Rubén A. Quintero, MD

Tampa, Fla

OBJECTIVE: The purpose of this study was to assess the value of a proposed classification of monochori-onic placenta in reference to twin-twin transfusion syndrome.STUDY DESIGN: The placentas from laser-treated patients with twin-twin transfusion syndrome and fromuncomplicated monochorionic pregnancies that were delivered between January 1997 and December 2000were included in the study. Placentas were classified as type A (no anastomoses), type B (only deep anasto-moses), type C (only superficial anastomoses), and type D (deep and superficial anastomoses). The numberand type of anastomoses were documented in each placental type. The severity of twin-twin transfusion syn-drome was assessed in stages, as previously described. The relationship between placental types and thedevelopment and severity of twin-twin transfusion syndrome was determined.RESULTS: One hundred thirty-one placentas were examined. Twin-twin transfusion syndrome developed in0% (0/4 placentas) of type A, in 100% (85/85 placentas) of type B, in 5.6% (1/18 placentas) of type C, and in79.17% (19/24 placentas) of type D placentas. An average of 4.17 (range, 1-11) vascular anastomoses wasfound. The mean number of superficial anastomoses was not different between patients with twin-twin trans-fusion syndrome and patients with no twin-twin transfusion syndrome (1.6 vs 1.71, respectively; P = .69, Stu-dent t test). The presence or absence of superficial anastomoses was not associated with differences in theseverity of twin-twin transfusion syndrome.CONCLUSION: This classification represents a practical approach to the surgical pathologic assessment ofvascular anastomoses in monochorionic placentas, with a strong clinical correlation. It also allows for theclarification of the relationship between superficial anastomoses and twin-twin transfusion syndrome.(Am J Obstet Gynecol 2002;187:489-94.)

Key words: Twins, placental anastomoses, twin-twin transfusion syndrome, laser photocoagulation,fetal therapy

490 Bermúdez et al August 2002Am J Obstet Gynecol

Material and methods

One hundred thirty-one monochorionic diamnioticplacentas were received fresh at the Department ofPathology of St Joseph’s Women’s Hospital, in Tampa,Fla, between January 1997 and December 2000. Onehundred five placentas (80.15%) corresponded to pa-tients with TTTS that had been treated with laser; 26placentas (19.85%) were from patients with mono-chorionic pregnancies without TTTS that were deliv-ered to our institution during the same time period.Forty-five other monochorionic placentas were not suit-able for analysis because of tissue fragmentation or pre-vious fixation in formalin. Standard sonographicdiagnostic criteria for TTTS included polyhydramniosin the recipient twin (maximum vertical pocket of am-niotic fluid, ≥8 cm), oligohydramnios in the donor twin(maximum vertical pocket of amniotic fluid, ≤2 cm),single placenta, same gender, and thin-dividing mem-brane with absent λ or twin peak sign. Disease severitywas assessed according to our staging classification:stage I, bladder of the donor twin visible, normalDoppler scans; stage II, bladder of the donor twin nolonger visible; stage III, absent or reverse end-diastolicvelocity in the umbilical artery or reverse blood flow inthe ductus venosus or pulsatile umbilical venous flow;stage IV, hydrops fetalis; and stage V, demise of 1 orboth twins.10

Surgical pathologic analysis of the placenta was per-formed by inspection of the chorionic fetal surface andwith air injection.4 Superficial arterioarterial and venove-nous anastomoses were recognized as direct and con-tinuous communications between the 2 fetal circulationswithout a terminal end into the placenta. Deep arteriove-nous communications were identified as unpaired arter-ies (or veins) from one twin into an emerging vein (orartery) to the other twin. The type and number of eachanastomosis was documented in all cases. Patency ofvascular anastomoses was assessed by air injection into the vascular communications.4 Basically, the umbilicalvessels, or the terminal vascular branches themselves,were catheterized with a 23- to 27-gauge 3/4inch needle(Winged Infusion Set; Surflo; Terumo Corp, Tokyo,Japan) through which air was injected from a 20- to 60-mL syringe (Beckton Dickinson, Franklin Lakes, NJ). Thenature of the anastomoses in placentas of patients whohad laser-treated TTTS was established from the endo-scopic description of the anastomoses and correlatedwith the surgical pathologic specimen. The photocoagu-lated areas were correlated with the description of thesurgical procedure. Patency of the ablated anastomoseswas also sought with air injection.

According to the type of anastomoses found, placentaswere classified as follows: type A, placentas without vascu-lar anastomoses; type B, placentas with only arteriove-

nous anastomoses; type C, placentas with only superficialanastomoses; or type D, placentas with combined arteri-ovenous and superficial anastomoses.

The relationship between the development and theseverity of TTTS and each placental type was assessed.

Statistical analysis. Statistical analysis was conductedwith SPSS, version 9.0 (SPSS Inc, Chicago, Ill) for Win-dows 98 (Microsoft Corportion, Redmond, Wash). Cate-goric variables were analyzed with the chi-squared testand Fisher exact test. Continuous variables were ana-lyzed with the Student t test. A probability value of <.05was considered statistically significant.

Results

Table I shows the relationship between placental type ver-sus the development of TTTS. There were no instances ofTTTS in placentas without anastomoses (0/4 anastomoses;type A), and only in 5.56% of placentas (1/18 placentas)with only superficial anastomoses (type C). Placentas withonly arteriovenous anastomoses (type B) were more likelyto develop TTTS than those patients with only superficial(type C), (85/85 [100%] vs 1/18 [5.6%], respectively, P =.000, Fisher exact test). Placentas with deep anastomosesonly (type B) were also more likely to have TTTS than pla-centas with deep and superficial anastomoses (type D)(85/85 [100%] vs 19/24 [79.17%], chi-squared test, 18.56;degrees of freedom, 1; P = .000, Fisher exact test). However,almost 80% of patients with combined superficial and deepanastomoses (type D) had TTTS.

Overall, the mean number of anastomoses was 4.17(range, 1-11; SD, 2.58). The number of anastomoses wassignificantly higher in TTTS placentas than in placentasthat did not have TTTS (4.76 ± 2.48 vs 1.77 ± 1.27, re-spectively; P = .000). Of the placentas with deep anasto-moses (types B and D), there was a significantly highernumber of arteriovenous anastomoses in placentas thatwere associated with TTTS than in placentas without thesyndrome (4.6 ± 2.22 vs 1.4 ± 0.5, respectively; P = .002,Student t test). However, of the placentas with superficialanastomoses (groups C and D), there was no differencein the number of superficial anastomoses in placentasthat were associated with TTTS than in placentas withoutthe syndrome (1.6 ± 0.61 vs 1.71 ± 0.96, respectively; P =.69, Student t test).

Table II shows the relationship between disease severityby TTTS stage among types B (deep anastomoses only)and D (deep and superficial anastomoses). There was nodifference in disease severity relative to the absence orpresence of superficial anastomoses. Table III shows thefrequency and types of superficial anastomoses in type Dand their relationship to stage of TTTS. There was no dif-ference in the severity of the disease relative to the type ofsuperficial anastomosis present (arterioarterial or ven-ovenous or arterioarterial + venovenous; P = 1, Fisherexact test).

Volume 187, Number 2 Bermúdez et al 491Am J Obstet Gynecol

Comment

Our study shows a clear association between the devel-opment of TTTS and the type of monochorionic pla-centa. The data show that the syndrome does not developin placentas without anastomoses (type A) and is unlikelyto develop in placentas with only superficial anastomoses(type C), whereas it appears to be invariably present inplacentas with only deep anastomoses (type B). Placentaswith combined superficial and deep anastomoses (typeD), although associated with a 20% reduction in the inci-dence of TTTS relative to placentas with only deep anas-tomoses, did not result in a less severe manifestation ofthe disease.

TTTS is thought to result from an unbalanced ex-change of blood between 2 monochorionic twinsthrough placental vascular anastomoses. Although theexact pathophysiologic mechanism responsible for thesyndrome is unknown, potential explanations can be de-rived from knowledge of the types of vascular anasto-moses that exist. By definition, deep arteriovenouscommunications perfuse a common cotyledon in a unidi-rectional way. For blood exchange to remain balanced,an arteriovenous communication in the opposite direc-tion with a similar amount of blood flow would need to bepresent. In the simplest form of TTTS, a single arteriove-nous anastomosis with no other type of anastomosis isfound. In our data, this only occurred in 3.8% of cases(4/105 cases). In most cases, there are arteriovenousanastomoses in both directions (arteriovenous anastomo-sis from donor to recipient [AVDR] or arteriovenousanastomoses from recipient to donor [AVRD]).11 Indeed,

in the current set, an average of 4.17 anastomoses werefound, which is similar to that found by other authors,2,4

but not by others.5 Thus, in the presence of combinedAVDRs and AVRDs, an even exchange of blood can occuronly if the combined net output through these vessels iszero. Although this is mathematically possible, it is not ap-parent in the current study. In fact, we have previouslyshown that a net passage of blood from donor to recipi-ent appears to occur through these anastomoses becauseof a greater number and/or size of AVDRs than ofAVRDs.11

The role of superficial anastomoses is more controver-sial. Using a syllogistic analysis, most authors suggest aprotective role for these anastomoses because they areless common in patients with TTTS. Indeed, if one usesTTTS as the starting point, our data and that of othersshow that only approximately 20% of patients with TTTShave superficial (and deep) anastomoses. However, withthe use of placental anatomy as the starting point (as inthe current study), 80% of patients with combined arteri-ovenous and superficial anastomoses have TTTS (typeD). Although this does represent a 20% reduction raterelative to patients with only deep arteriovenous commu-nications, the purported protective role of superficialanastomoses is less obvious. An actual “protective” rolefor superficial anastomoses could be argued for those pa-tients in whom only superficial anastomoses were present(type C), where the incidence of TTTS was 5.6%. Propo-nents of the protective role of superficial anastomosessuggest that an uneven blood exchange through deep ar-teriovenous communications is able to flow back to theother twin through the superficial vessels.6,12,13 Yet, the

Table I. Incidence of TTTS by placental type

Placental type TTTS (No.) No TTTS (No.) Total (No.)

A (no anastomoses) 0 (0%)* [0%]† 4 (15%)* [100%]† 4 (3%)B (arteriovenous anastomoses only) 85 (81%)* [100%]† 0 (0%)* [0%]† 85 (65%)C (superficial anastomoses only) 1 (1%)* [5.6%]† 17 (65%)* [94.4%]† 18 (14%)D (arteriovenous + superficial anastomoses) 19 (18%)* [79.17%]† 5 (19%)* [20.83%]† 24 (18%)Total 105 26 131 (100%)

*Percent of placental type within group. †Percent of TTTS within groups.

Table II. Severity of TTTS in placental types B and D

B D TotalStage (No.)* (No.)† (No.)

I 14 (16.4%) 6 (31.5%) 20 (19.23%)II 32 (37.6%) 5 (26.3%) 37 (35.57%)III 30 (35.3%) 7 (36.8%) 37 (35.57%)IV 9 (10.6%) 1 (5.2%) 10 (9.61%)Total 85 (100%) 19 (100%) 104 (100%)

*Arteriovenous anastomoses only.†Arteriovenous + superficial anastomoses.

Table III. Distribution of superficial anastomoses in pla-centas type D by stage of TTTS

Arterioarterial Arterioarterial Venovenous + venovenousanastomosis anastomosis anastomosis Total

Stage only (No.) only (No.) (No.) (No.)

I 4 1 1 6II 1 1 3 5III 2 0 5 7IV 1 0 0 1Total 8 2 9 19

492 Bermúdez et al August 2002Am J Obstet Gynecol

actual mechanism as to how this can occur is not obvious,because a net flow of blood would need to take place inthe opposite direction of the normal flow of the vessel.For example, in an arterioarterial communication, bloodthat flows from opposite directions meet at a certainpoint on the length of the artery. The collision of the 2fronts can actually be seen endoscopically as a to-and-fromovement of a demarcation line in certain arterioarterialanastomoses with arterial blood of different hues (Fig 1,A). This phenomenon may correspond to the biphasicwaveform described with Doppler studies.8,14 This colli-sion front (or as we call it, the hemodynamic equator[HE]) is located anywhere along the length of the vesseland may or may not coincide with the actual vascular

equator. In a given instance and with the combined arte-rial pulsations of both twins, the HE shows a to-and-fromovement of <1 cm of displacement (Fig 1, B). This pat-tern is not to be equated with “bidirectional flow,”8,15 be-cause for this the HE would need to reach a vein on boththe recipient and donor’s territory intermittently. This isanatomically unlikely to happen.

Our endoscopic observations suggest that both arte-rioarterial and venovenous anastomoses may be of 2 dif-ferent types: (1) In the first type (branchless type), thevessels are devoid of any branching, and the anastomosisruns uninterrupted from one fetus to the other. (2) Inthe second and more common type (branched type),branches are present along the length of the vessel, al-though the vessel itself lacks a terminal end. Thesebranches, which allow blood to perfuse cotyledons alongthe length of the vessel, are oriented toward one fetus orthe other, usually in relationship to the vascular equator.In branchless superficial anastomoses, it is difficult to un-derstand how equalization of blood flow could occur, be-cause this would imply complete reversal of blood flowthrough the vessel (Fig 2). An extreme example of thisphenomenon is reverse arterial perfusion syndrome, inwhich an artery of the pump twin flows retrogradely intothe abnormal twin.16 In branched superficial anasto-moses, no net exchange of blood can occur if the HE co-incides with the vascular equator without returning veins,resembling a tied match of tug-of-war. If the HE does notcoincide with the vascular equator, the arterioarterialmay behave as a “functional arteriovenous” anastomosis ifarterial branches of the arterioarterial anastomosis reachveins of the other twin. Thus, depending on the locationof the HE relative to the vascular equator and returningveins, an arterioarterial anastomosis may behave as afunctional arteriovenous (AVDR or AVRD) or as a neutral

Fig 1. A, In vivo endoscopic demonstration of the transverse HE in an arterioarterial anas-tomosis (arrow) within the territory of the donor twin. Darker blood from the donor twin(upper) meets bright red blood from the recipient twin (lower). In this instant, the branch(B) carries dark blood from the donor into the placenta. Because the subjacent vein re-turns to the donor, this cotyledon is normally perfused by the donor in this instant. B, Aninstant later, the HE has moved upward approximately 8 mm. The branch (B) of the arte-rioarterial anastomosis now carries bright red blood from the recipient twin into the pla-centa. Because the subjacent vein returns to the donor twin, this arterioarterialanastomosis effectively now allows blood to flow from recipient to donor (functionalAVRD). Nonetheless, the patient had stage III TTTS.

Fig 2. HE along the length of the vessel (longitudinal HE) tra-verses the vascular equator. Dark blood from the donor twin(upper) meets bright red blood from the recipient twin (lower) butcontinues in a longitudinal fashion along the length of the ves-sel. There are no adjacent veins at this level. Therefore, no netexchange of blood between the fetuses is apparent at this site.

Volume 187, Number 2 Bermúdez et al 493Am J Obstet Gynecol

vessel with normal perfusion of the placenta and zero netblood exchange between the fetuses.

We have also noted that the HE may extend itself alongthe length of the superficial anastomosis, in a longitudi-nal rather than in a transverse fashion (Fig 2). This sec-ond mechanism could allow for exchange of blood flowin both branched and unbranched superficial anasto-moses, provided that at one point the blood couldchange from artery to vein or vice versa and complete thecircle. The contribution of one or the other mechanismthat prevents or causes TTTS is unknown.

Patients with placentas with only superficial anasto-moses (type C) were unlikely to experience the syn-drome. From our data, we estimate that this is theplacental type in two thirds of non-TTTS patients. How-ever, our data show that TTTS can occur in this placentaltype. In one patient, a large arterioarterial and venove-nous anastomosis and no arteriovenous anastomoseswere present. Laser photocoagulation of the arterioarter-ial and venovenous anastomoses resulted in complete res-olution of the syndrome. We now have 2 additionalpatients with type C placentas and TTTS that are not in-cluded in this set because they fall outside the study pe-riod (unpublished data). It is likely that the HE in mosttype C placentas coincides with the vascular equator,which results in a zero net exchange of blood betweenthe fetuses. This phenomenon is possibly just a reflectionof equilibration of fetal blood pressures.

The actual number of superficial anastomoses was notdifferent between type C and type D placentas (1.6 ± 0.61vs 1.71 ± 0.96, respectively; P = 0.69, Student t test. This isin contrast with the findings of other authors.6,7 In addi-tion, the severity of the disease was unrelated to the typeor number of superficial anastomoses (Table III). Thesefindings do not support the suggestion that patients witharterioarterial anastomoses should have a better outcomeor that venovenous anastomoses are deleterious.15 Be-cause patients whom we have treated with selective laserphotocoagulation of communicating vessels have a uni-form outcome (at least one fetus survived in 86/105 preg-nancies [81.9%] of patients with TTTS in this series,regardless of stage), we could not use outcome to com-pare the possible prognostic significance of differenttypes of superficial anastomoses. Because the severity ofthe disease is related to outcome in symptomaticallytreated patients,17-19 we chose staging rather than out-come for this comparison. The lack of association be-tween disease severity and the presence or the type ofsuperficial anastomoses questions the clinical impor-tance of categorizing them prenatally in patients withTTTS.15

Our approach in the assessment of the relationship be-tween placental vascular anastomoses and the develop-ment of TTTS is somewhat similar to that of othersurgical pathology studies. Benirschke and Driscoll2 de-

scribed 7 groups in 60 placentas; Baldwin4 described 15groups in 373 placentas, and Machin et al13 described 13groups in 69 placentas according to the anastomosesfound. Our current classification into 4 placental typessuggests a rather homogenous clinical behavior, in whichTTTS is likely to develop in types B and D placentas butnot in types A or C. This classification may have impor-tant clinical implications. First, TTTS does not develop inpatients with no anastomoses (type A placentas). There-fore, the elimination of the syndrome could be achievedby surgically transforming any placenta into a type A pla-centa. Interestingly, a similar argument could potentiallybe made for type C placentas (superficial communica-tions only), in which TTTS is also unlikely to occur. Un-fortunately, surgical transformation of a type D placentainto a type C placenta by ablating only the deep arteri-ovenous anastomoses and leaving intact the superficialanastomoses may have catastrophic consequences. In-deed, as discussed earlier, the presence of only superficialanastomoses does not preclude the development of thesyndrome and, more importantly, may be responsible forit or for the development of reversal TTTS. In addition,should the death of one twin occur, fetofetal hemorrhagethrough the patent superficial anastomoses may result indeath or neurologic damage of the cotwin. Therefore,transformation of a type D placenta to a type C placenta,as suggested by some authors,20 is unwarranted. Second,in patients with established TTTS, it is probably of littlevalue to determine the presence or absence of superficialanastomoses with ultrasound, as they do not seem to beassociated with different disease severity. Finally, ultra-sound scanning of the placenta for the presence of su-perficial anastomoses8,15 in healthy monochorionic twinswould appear unwarranted since, according to our data,94% of these patients will not develop the syndrome.

Our study should help clarify the role of the differentplacental types and anastomoses in the clinical behaviorof monochorionic placentas. For the clinician, attemptsto characterize sonographically the placental type areprobably unwarranted. This simplifies the prenatal assess-ment of patients with monochorionic pregnancies. Forthe pathologist, the classification provides a practicalmethod to classify monochorionic placentas based on thetype of vascular anastomoses present, and a useful clinicalcorrelation.

We thank Mr Robert Stranahan, Ms Kristi Berlin, andMs Barbara Stroud from the Pathology Department at StJoseph’s Women’s Hospital in Tampa, Fla, for their col-laboration in the analysis of the placentas.

REFERENCES

1. Strong S, Corney G. The placenta in twin pregnancy. 1st ed. Nor-wich (UK): Pergamon Press; 1967.

2. Benirschke K, Driscoll S. The pathology of the human placenta.New York: Springer-Verlag; 1967.

494 Bermúdez et al August 2002Am J Obstet Gynecol

3. Patten RM, Mack LA, Harvey D, Cyr DR, Pretorius DH. Disparityof amniotic fluid volume and fetal size: problem of the stocktwin: US studies. Radiology 1989;172:153-7.

4. Baldwin V. Pathology of multiple pregnancy. New York:Springer-Verlag; 1994.

5. Bajoria R. Abundant vascular anastomoses in monoamniotic ver-sus diamniotic monochorionic placentas. Am J Obstet Gynecol1998;179:788-93.

6. Bajoria R, Wigglesworth J, Fisk NM. Angioarchitecture of mono-chorionic placentas in relation to the twin-twin transfusion syn-drome. Am J Obstet Gynecol 1995;172:856-63.

7. Denbow ML, Cox P, Taylor M, Hammal DM, Fisk NM. Placentalangioarchitecture in monochorionic twin pregnancies: relation-ship to fetal growth, fetofetal transfusion syndrome, and preg-nancy outcome. Am J Obstet Gynecol 2000;182:417-26.

8. Denbow ML, Cox P, Talbert D, Fisk NM. Colour Doppler energyinsonation of placental vasculature in monochorionic twins: ab-sent arterio-arterial anastomoses in association with twin-to-twintransfusion syndrome. Br J Obstet Gynaecol 1998;105:760-5.

9. Quintero R, Morales W, Mendoza G, et al. Selective photocoagula-tion of placental vessels in twin-twin transfusion syndrome: evolu-tion of a surgical technique. Obstet Gynecol Surv 1998; 53:S97-103.

10. Quintero R, Morales W, Allen M, Bornick P, Johnson P, KruegerM. Staging of twin-twin transfusion syndrome. J Perinatol 1999;19:550-5.

11. Quintero R, Quintero L, Bornick P, Allen M, Johnson P. Thedonor-recipient (D-R) score: in vivo endoscopic evidence to sup-port the hypothesis of a net transfer of blood from donor to re-cipient in twin-twin transfusion syndrome. Prenat Neonat Med2000;5:84-91.

12. Benirschke K. Multiple pregnancy. In: Polin R, Fox W, editors.Fetal and neonatal physiology. Vol 1. Philadelphia: WB Saun-ders; 1992.

13. Machin G, Still K, Lalani T. Correlations of placental vascularanatomy and clinical outcomes in 69 monochorionic twin preg-nancies. Am J Med Genet 1996;61:229-36.

14. Hecher K, Jauniaux E, Campbell S, Deane C, Nicolaides K.Artery-to-artery anastomosis in monochorionic twins. Am J Ob-stet Gynecol 1994;171:570-2.

15. Taylor MJ, Talbert DG, Fisk NM. Mapping the monochorionicequator: the new frontier. Ultrasound Obstet Gynecol 1999;14:372-4.

16. Van Allen M, Smith D, Shepard T. Twin reversed arterial perfu-sion (TRAP) sequence: a study of 14 twin pregnancies with ac-ardius. Semin Perinatol 1983;7:285-93.

17. Mari G, Detti L, Oz U, Abuhamad AZ. Long-term outcome intwin-twin transfusion syndrome treated with serial aggressive am-nioreduction. Am J Obstet Gynecol 2000;183:211-7.

18. Zikulnig L, Hecher K, Bregenzer T, Bäz E, Hackelöer B. Dopplersonography and prognostic factors in severe twin-twin transfu-sion syndrome with endoscopic laser therapy [abstract]. Ultra-sound Obstet Gynecol 1998;12:9.

19. Hecher K, Ville Y, Snijders R, Nicolaides K. Doppler studies ofthe fetal circulation in twin-twin transfusion syndrome. Ultra-sound Obstet Gynecol 1995;5:318-24.

20. van Gemert MJ, Major AL, Scherjon SA. Placental anatomy, fetaldemise and therapeutic intervention in monochorionic twinsand the transfusion syndrome: new hypotheses. Eur J Obstet Gy-necol Reprod Biol 1998;78:53-62.