- Email: [email protected]
Twin-to-twin transfusion syndrome: What are appropriate diagnostic criteria?
Volume 161 Number 2
the explanation for why hydrops develops in the severely anemic fetus. We gratefully acknowledge the patient care provided by the Yale Obstetrics and Gynecology housestaff and the nursing care provided by Mitzi Barger, RN, and Nancy Nickless, RN. REFERENCES 1. Phibbs RH, Johnson P, Tooley WHo Cardiorespiratory status of erythroblastotic newborn infants: II. Blood volume, hematocrit and serum albumin concentration in relation to hydrops fetalis. Pediatrics 1974;53: 13-23. 2. Grannum PA, Copel JA, Plaxe S, et a1. Intrauterine intravascular transfusion for severe erythroblastosis fetalis. N EnglJ Med 1986;314:1431-4. 3. Grannum PA, Copel JA. Moya F, et a1. The reversal of hydrops fetalis by intravascular intrauterine transfusion in severe isoimmune fetal anemia. AM J OBSTlT GYNECOL 1988;158:914-9. 4. Kleinman CS, Donnerstein RL, DeVore GR, et a1. Fetal echocardiography for evaluation of in utero congestive heart failure: a technique for the study of non-immune fetal hydrops. N EnglJ Med 1982;306:568-75. 5. Kleinman CS, Weinstein EM, Copel JA. Pulsed Doppler analysis of human fetal blood flow. Clin Diagn Ultrasound 1986; 17: 173-85. 6. Shepard MJ, Richards VA, Berkowitz RL, et a1. An evaluation of two equations for predicting fetal weight by ultrasound. AM J OBSTET GYNECOL 1982;142:47-54.
Fetal cardiac output in isoimmunized pregnancies
7. Hade L, Angelsen B. Doppler ultrasound in cardiology. 2nd ed. Philadelphia: Lea & Febiger, 1985:324. 8. Kenny JF, Plappert T, Doubilet P, et a1. Changes in intracardiac blood flow velocities and right and left ventricular stroke volumes with gestational age in the normal human fetus: a prospective Doppler echocardiographic study. Circulation 1986; 74: 1208-16. 9. Reed KL, Meijboom EJ, Sahn DJ, et a1. Cardiac Doppler flow velocitieb in human fetuses. Circulation 1986;73: 41-6. 10. Kenny JF, Plappert T, Doubilet P, et a1. Effects of heart rate on ventricular size, stroke volume, and output in the normal human fetus: a prospective Doppler echocardiographic study. Circulation 1987;76:52-8. 11. De Smedt MCH, Visser GHA, Meijboom EJ. Fetal cardiac output estimated by Doppler echocardiography during mid- and late gestation. Am J Cardiol 1987;60:338-42. 12. Nicolaides KH, Warenski JC, Rodeck CH. The relationship of fetal plasma protein concentration and hemoglobin level to the development of hydrops in rhesus isoimmunization. AMJ OBSTET GYNECOL 1985;152:341-4. 13. Barss VA, Doubilet PM, St. John-Sutton M, et a1. Cardiac output in a fetus with erythroblastosis fetalis: assessment using pulsed Doppler ultrasound. Obstet Gynecol 1987; 70:442-4. 14. Fumia FD, Edelstone DI, Holzman IR. Blood flow and oxygen delivery to fetal organs as functions of fetal hematocrit. AM J OBSTET GYNECOL 1984;150:274-82. 15. Sooth ill PW, Nicolaides KH, Rodeck CH, et al. Relationship of fetal hemoglobin and oxygen content to lactate concentration in Rh isoimmunized pregnancies. Obstet Gynecol 1987 ;69:268-71.
Twin-to-twin transfusion syndrome: What are appropriate diagnostic criteria? Fiona H. Danskin, MB, ChB,* and James P. Neilson, MD
Glasgow, Scotland One hundred seventy-eight consecutive twin pregnancies were studied to reevaluate the standard diagnostic criteria for chronic twin-to-twin transfusion syndrome of an intertwin hemoglobin difference >5 gm/dl and a birth weight difference >20%. Hemoglobin differences> 5 gm/dl were found in six pregnancies with monochorionic placentas but also in seven with dichorionic placentas. Birth weight differences >20% occurred no more commonly in monochorionic than in dichorionic pregnancies. Of the four pregnancies with a coexisting hemoglobin difference >5 gm/dl and birth weight difference >20%, only one had a monochorionic placenta and therefore likelihood of vascular anastomoses. Diagnosis of twin-to-twin transfusion syndrome cannot be definitively established by current standard diagnostic criteria. (AM J OBSTET GYNECOL 1989;161:365-9.)
Key words: Twin pregnancy, twin-to-twin transfusion syndrome, diagnosis
From the Department ofM idwzjery, University of Glasgow, the Queen Mother's Hospital. Received for publication August 7, 1988; revised December 8, 1988; accepted February 24, 1989. Reprints not available. *Dr. Danskzn holds the emden MedIcal Research Scholarship administered by the Scottish Hospital Endowments Research Trust.
Twin-to-twin transfusion syndrome is an important complication of twin pregnancy, I, 2 resulting from shunting of blood from one twin to the other through placental vascular anastomoses. Injection studies of twin placentas have shown that such anastomoses are almost universally present in monochorionic placentas
365
366
Danskin and Neilson
August 1989 Am J Obstet Gynecol
Table I. Zygosity studies in 178 consecutive twin pregnancies Monochorionic placenta Blood group studies Discordant sex Intrauterine death Technical failure
51) 19 70 Monozygotic
135) 87 Dizygotic 52
4) 21 Unknown 17
but so extremely rare in dichorionic placentas l • 3 that the possibility of the syndrome occurring with this type of placenta can be disregarded. Both acute and chronic forms of the twin-to-twin transfusion syndrome have been described. 4 • 5 The acute form has been attributed to rapid transfer of blood through superficial arteryto-artery or vein-to-vein anastomoses during labor and delivery, resulting in a hypovolemic, anemic donor twin and a hypervolemic, plethoric recipient twin of similar birth weight. 4 . 5 The chronic syndrome results from gradual antepartum transfusion through deep arteryto-vein anastomoses, rendering the donor twin not only hypovolemic and anemic but also growth retarded, whereas the hypervolemic, plethoric recipient is usually larger. 6 -s These features may be accompanied by oligohydramnios in the sac of the donor and polyhydramnios in that of the recipient and may be identifiable before birth by ultrasonography.9. to Twin-to-twin transfusion syndrome is usually diagnosed or confirmed after birth by the discovery of a hemoglobin difference >5 gm/dl between twins 2 and a birth weight difference >20%, providing evidence of the chronic form of the syndrome! We have reevaluated these standard and widely accepted diagnostic criteria as part of an ongoing intensive study of twin pregnancy. Despite an exhaustive MEDLINE literature search, we have not been able to identify any other such reevaluation. Material and methods
One hundred seventy-eight consecutive twin pregnancies delivering in The Queen Mother's Hospital, Glasgow, (a tertiary referral center) between August 1984 and February 1988 were studied. Of the mothers, 160 had attended the hospital for antenatal care, mostly at a special twin pregnancy antenatal clinic, and the other 18 were transferred from other hospitals, usually in established or incipient preterm labor. Venous blood samples were usually obtained from neonates by pediatricians on the first day of life for hematologic indexes, including hemoglobin concentration. In some
cases in which both babies were well, at term, and appropriate for gestational age, the pediatric staff considered venesection unnecessary on clinical grounds. All birth weights were recorded, and the intertwin birth weight difference was calculated as a percentage of the weight of the larger twin. Zygosity was determined in 88% of pregnancies (Table I) by inspection of the placentas (a monochorionic placenta indicating a monozygotic pregnancy), sex discordance (indicating a dizygotic pregnancy), or when necessary by comparing red blood cell antigens of the ABO, rhesus, MNS, Kell, Duffy, Kidd, Lutheran, and Wright systems. II In all 21 cases in which zygosity was not established, the placenta was dichorionic and the twins were of like sex; in four there had been intrauterine death of one twin, and in the remainder blood was not sent for grouping. Total perinatally related wastage, an index that includes late abortions, stillbirths, early and late neonatal deaths, and perinatally related infant deaths, 12 is shown in Table II. In all three monochorionic monozygotic pregnancies in which one or both fetuses were stillborn, features were compatible with, although not diagnostic of, twin-to-twin transfusion syndrome. In one preterm labor occurred at 34 weeks, and the leading twin died of asphyxia after abdominal dystocia caused by ascites; the second twin weighed 530 gm less and had a hemoglobin value at birth of 11 gmt dl. In another pregnancy acute polyhydramnios occurred in the amniotic sac of the first twin, which resulted in abruptio placentae and intrauterine death. The second twin, who survived, was very pale at birth and weighed 27% less than the first. In a third pregnancy acute polyhydramnios developed at 32 weeks, which was more marked in the amniotic sac of the first twin. Labor supervened within 24 hours and stillborn female infants were born; the first twin weighed 1.71 kg and appeared very plethoric with congested organs at postmortem examination, and the second twin weighed only 1.46 kg and appeared very pale. Hemoglobin values were not obtained in the latter two pregnancies. Results
The hemoglobin concentrations of venous blood samples withdrawn from both babies on the first day of life were obtained in 118 of the 178 pregnancies. Both hemoglobin values were known in 35 of the 51 monochorionic pregnancies, and in six there was an intertwin difference >5 gm/dl. However, there were two pregnancies in the monozygotic dichorionic group, one in the dizygotic group, and four in the zygosity unknown group. all of which had dichorionic placentas, in which hemoglobin differences >5 gm/dl were also found. Oligohydramnios occurred in the amniotic sac of one twin in only two of these 13 pregnancies, both
Diagnosis of twin-to-twin transfusion syndrome 367
Volume 161 Number 2
Table II. Outcome in 178 consecutive twin pregnancies TPRWllOOOt
Zygosity I placentation
Monozygotic monochorionic Monozygotic dichorionic Dizygotic Not known
3*
0
0
0
0
1 4
2
8§1I
0 2
4
102
93
38
38
0
174 42
164 32
0 I
0
64.3 57.5 238.1
NND,Neonatal deaths; PRID, perinatally related infant deaths; TPRW, total perinatally related wastage. *Perinatally related infant deaths are all infant deaths (between 1 month and 1 year of life) from or pardy from a primary
perinatal complication or its treatment.!2 tTotal perinatally related wastage includes all late abortions, except terminations of pregnancy for fetal abnormalities, all perinatal deaths, all late neonatal deaths (during the second, third, and fourth weeks of life), and all perinatally related infant deaths.!2 tOne exomphalos. § One anencephalic. IIOne cyanotic congenital heart disease. of which had dichorionic placentas. None of the twentysix babies in this group required treatment for anemia or polycythemia in the neonatal period. Birth weight differences, which were expressed as a percentage of the weight of the larger twin, were calculated for all pregnancies. Excluding those cases in which there was a fetal abnormality or a fetus papyraceus, a birth weight difference >20% was present in 17.4% of pregnancies. The distribution of these according to zygosity is shown in Table III. Oligohydramnios was again an infrequent occurrence in this group and was found in the sac of one twin in only four pregnancies. Three of these had a dichorionic placenta and one had a monochorionic placenta. In only four pregnancies did a hemoglobin difference >5 gm/dl and a birth weight difference >20% coexist, and none of these four pregnancies showed evidence of oligohydramnios or polyhydramnios in either amniotic sac. Only one of these pregnancies was associated with a monochorionic placenta, and in it alone did the heavier twin have the higher hemoglobin value (Table IV).
Comment Since the recognition of the twin-to-twin transfusion syndrome by Herlitz!3 in 1941, much interest has been focused on this rare complication of twin pregnancy. Initial publications were single case reports of the syndrome, whereas larger studies later attempted to define the incidence, pathogenesis, and associated morbidity and mortality!' 4 It was suggested that when a large volume of blood is transferred antepartum, one or both twins may die in utero! Since many early studies only considered those pregnancies resulting in the live birth of both babies, they probably underestimated the true incidence of the syndrome. In our study all three monochorionic pregnancies resulting in the stillbirth of
Table III. Birth weight difference >20% Zygosity I placentation
Monozygotic monochorionic Monozygotic dichorionic Dizygotic Not known
Btrth weIght difference >20%
%
51
8
15.7
19
4
21.1
87 21 178
14 5 31
16.0 23.8 17.4
N
one or both babies demonstrated features compatible with the twin-to-twin transfusion syndrome. However, the purpose of our study was neither to determine the significance of the twin-to-twin transfusion syndrome in contributing to perinatal mortality and morbidity nor to define its incidence (we did not have hemoglobin values from all sets of twins). Rather, we aimed to reevaluate the standard diagnostic criteria for the twin-to-twin transfusion syndrome-a matter not only of clinical importance but also vital for further perinatal research. Indeed, our study was generated by conflicting reports of fetal velocity waveforms of the umbilical artery as assessed by the new technology of Doppler ultrasonography on twin pregnancies believed complicated by the twin-to-twin transfusion syndrome. Concordant,!4 discordant,!5 and variable l6 patterns have been described, and these discrepancies may have stemmed from diagnostic confusion. Twin-to-twin transfusion syndrome has traditionally been diagnosed after birth by finding a hemoglobin difference >5 gm/dl-a criterion introduced by Rausen et al! in 1965. They had found such a hemoglobin difference in 19 of 130 monochorionic pregnancies, whereas in 43 dizygotic pregnancies no difference >3.3
368
Danskin and Neilson
August 1989 Am J Obstet Gynecol
Table IV. Birth weight difference >20% and hemoglobin difference >5 gm/dl Birth weIght Zygosity / placentation
Monozygotic monochorionic Monozygotic dichorionic Dizygotic dichorionic Dizygotic dichorionic
Twin 1 (kg)
2.85 2.39 1.47 3.16
I
Twin 2 (kg)
2.18 1.49 2.41 2.25
gm/ dl was found. Therefore they selected a hemoglobin difference> 5 gm/ dl as a suitable cutoff value, and this has been accepted by authors of subsequent publications as definitive evidence of the twin-to-twin transfusion syndrome. However, our study showed that although a hemoglobin difference> 5 gm/ dl existed in six monochorionic pregnancies, it also occurred in seven dichorionic pregnancies; in fact the largest hemoglobin differences (12.5 and 11.3 gm/ dl) occurred in the latter group. Because placental anastomoses are extremely rare in fused dichorionic placentas,3 it is unlikely that the twin-to-twin transfusion syndrome occurred in these seven cases. Thus a difference in hemoglobin concentration is not in itself sufficient to establish a diagnosis of twin-to-twin transfusion syndrome. Tan et al! studied 482 twin pairs in an attempt to differentiate the acute and chronic forms of the syndrome. They diagnosed the syndrome in 35 sets of twins on finding obvious plethora in one twin and pallor in the other, along with a hemoglobin difference >5 gm/dl. They divided the twins into two groups according to whether the birth weight difference between each pair was greater or less than 20% of the weight of the heavier twin. They found that in those twins who had a large weight difference, the twin with the lower hemoglobin was invariably lighter, whereas in those twins with a small weight difference, the twin with the lower hemoglobin value was heavier. They concluded that birth weight difference provides a simple yet reliable yardstick for differentiation of twin pregnancies with hemoglobin differences into acute and chronic forms. This additional criterion of birth weight difference >20% has been widely adopted. However, our study showed that a birth weight difference >20% occurred no more commonly in monochorionic pregnancies than in dichorionic pregnancies. In addition, of the four pregnancies in which a hemoglobin difference >5 gm/dl and a birth weight difference >20% coexisted, only one had a monochorionic placenta. In this pregnancy the heavier twin had the higher hemoglobin value, whereas in the three dichorionic pregnancies the lighter twin had the higher hemoglobin value. These presumably represent growth
Hemoglobin
I
Difference (%)
(gm I dl)
Twzn 1
23.51 37.66 39.00 28.80
24.0 14.9 22.9 13.7
I
Twzn2 (gm I dl)
16.0 27.4 17.0 25.0
I
Difference (gm I dl)
8.0 12.5 5.9 11.3
retardation in aSSOCIatIOn with polycythemia, a wellrecognized combination. Because Tan et al. used a hemoglobin difference >5 gm/dl as the sole diagnostic criterion for the twin-to-twin transfusion syndrome and did not take into account chorionicity of the placenta, it is likely that their subgroup included a similar proportion of polycythemic growth retarded babies from dichorionic pregnancies. In summary, the standard diagnostic criteria used in clinical obstetrics and perinatal research do not permit a definitive diagnosis of the twin-to-twin transfusion syndrome. Differences in hemoglobin and birth weight must be interpreted in the light of knowledge of placental morphology and whether the heavier or lighter twin had the higher hemoglobin value. Further studies are still required to elucidate this enigmatic condition. We gratefully acknowledge the contributions of Dr. A. A. M. Gibson, consultant pathologist, and the Regional Blood Transfusion Service, Law Hospital, in establishing zygosity, and the helpful criticisms of this manuscript by Professor C. R. Whitfield and Dr. B. Holland. REFERENCES I. Benirschke K. Twin placenta in perinatal mortality. NY
State] Med 1961;61:1499-508. 2. Rausen AR, Seki M, Strauss L. Twin transfusion syndrome.] Pediatr 1965;66:613-28. 3. Robertson EG, Neer KJ. Placental injection studies in twin gestation. AM ] OBSTET GYNECOL 1983; 147: 170-4. 4. Tan KL, Tan R, Tan SH, Tan AM. The twin transfusion syndrome. Clin Pediatr 1979;18: 11l-4. 5. Klebe]G, Ingomar CJ. The fetoplacental circulation during parturition illustrated by the interfetal transfusion syndrome. Pediatrics 1972;49:112-6. 6. Becker AH, Glass H. Twin to twin transfusion syndrome. Am] Dis Child 1963; 106:624-9. 7. Corney G, Aherne W. The placental transfusion syndrome in monozygotic twins. Arch Dis Child 1965;40:26470. 8. Naeye RL. Human intrauterine parabiotic syndrome and its complications. N Engl] Med 1963;268:804-8. 9. BrennanJN, Dirnan RV, Rosen MG, Bellon EM. Fetofetal transfusion syndrome: prenatal ultrasonographic diagnosis. Radiology 1982; 143 :535-6. 10. Wittman BK, Baldwin V], Nichol B. Antenatal diagnosis of twin transfusion syndrome by ultrasound. Obstet GynecoI1981;58:123-7. II. Wilson RS. Blood typing and twin zygosity. Hum Hered 1970;20:30-56.
Volume 161 Number 2
12. Whitfield CR, Smith NC, Cockburn F, Gibson AAM. Perinatally related wastage-a proposed classification of primary obstetric factors. Br J Obstet Gynaecol 1986;93:694703. 13. Herlitz G. Zur Kennitis der anamischen und polyzytamischen Zutstande bei Neugeborenen, sowie des Icterus gravis neonatorum. Acta Paediatr (Uppsala) 1941 ;29:211. 14. Giles WB, Trudinger BJ, Cook CM. Fetal umbilical artery flow velocity-time waveforms in twin pregnancies. Br J Obstet Gynaecol 1985;92:490-7.
Diagnosis of twin-to-twin transfusion syndrome
15. Farmakides G, Schulman H, Saldana LR, Bracero LA, Fletcher A, Rochelson B. Surveillance of twin pregnancy with umbilical arterial velocimetry. AMJ OBSTET GVNECOL 1985; 153:789-92. 16. Erskine RLA, Ritchie JWK, Murnaghan GA. Antenatal diagnosis of placental anastomoses in a twin pregnancy using Doppler ultrasound. Br J Obstet Gynaecol 1986; 93:955-9.
The prevalence and biologic significance of lupus anticoagulant and anticardiolipin antibodies in a general obstetric population C. J. Lockwood, MD, R. Romero, MD, R. F. Feinberg, MD, PhD, L. P. Clyne, FIMLS, MS, B. Coster, BS, andJ. C. Hobbins, MD New Haven, Connecticut Circulating antibodies to negatively-charged phospholipids have been implicated in the genesis of adverse pregnancy outcomes. However, it has yet to be established that these antibodies are causative or that they are invariably associated with untoward perinatal outcomes. To address this issue, the prevalence of lupus anticoagulant and anticardiolipin antibodies was recorded in a low-risk obstetric population, and the outcomes of untreated pregnancies were evaluated. Two of 737 patients (0.27%) had lupus anticoagulant documented by a prolonged activated partial thromboplastin time that did not correct with mixing studies. In comparison, greatly elevated concentrations of immunoglobulin M-anticardiolipin antibodies or immunoglobulin G-anticardiolipin antibodies were identified in 16/737 (2.2%) patients by means of an enzyme-linked immunosorbent assay. Spontaneous abortions occurred in both lupus anticoagulant-positive patients, suggesting that the activated partial thromboplastin time used was a relatively insensitive but specific marker for anti phospholipid antibody-associated adverse pregnancy outcomes. In contrast, although 12 of 16 anticardiolipin antibodies-positive pregnancies were complicated by perinatal loss, preterm delivery, or fetal growth retardation, four patients had uncomplicated pregnancies. Moreover, the distribution of anti cardiolipin antibodies concentrations in these four patients was not clustered among the lowest anti cardiolipin antibodies values, and anticardiolipin antibodies concentrations correlated weakly with adverse outcomes. These findings suggest that antiphospholipid antibodies are related to adverse pregnancy outcomes in a complex fashion and that therapy is not always required for acceptable outcomes in patients without other risk factors. (AM J OSSTET GVNECOL 1989;161:369-73.)
Key words: Lupus anticoagulant, anticardiolipin antibodies, activated partial thromboplastin time Circulating lupus anticoagulants are either identical to or closely associated with immunoglobulin M (IgM) and immunoglobulin G (IgG) antibodies to negativelycharged phospholipids.' A laboratory characteristic of lupus anticoagulant is the inhibition of the activated partial thromboplastin time and its derivative assays.' From the Departments of Obstetrics and Gynecology and MediCine, Yale University School of MediCine. Received for publication july 9,1988; revised September 18,1988; accepted February 24,1989. Reprint requests: Charles J. Lockwood, MD, Director of Pennatal Research, St. Margaret's Hospital, Tufts Unzversl/'V, School of Medicine, 90 Cushing Ave., Boston, MA 02125.
This in vitro laboratory phenomenom is a consequence of antibody binding to negatively-charged phospholipid surfaces on which factors Xa and Va convert prothrombin to thrombin.' Paradoxically, in vivo lupus anticoagulant and related antiphospholipid antibodies promote thrombosis l via a myriad of potential prothrombotic mechanisms. 2•6 Reportedly 94% of pregnancies occurring in patients with lupus anticoagulant who do not receive appropriate treatment result in spontaneous abortion or stillbirth. 7 A 13% to 40% prevalence of antibodies to the divalent phospholipid, cardiolipin, has been noted in patients evaluated for a history of "idiopathic" recur-
369
the explanation for why hydrops develops in the severely anemic fetus. We gratefully acknowledge the patient care provided by the Yale Obstetrics and Gynecology housestaff and the nursing care provided by Mitzi Barger, RN, and Nancy Nickless, RN. REFERENCES 1. Phibbs RH, Johnson P, Tooley WHo Cardiorespiratory status of erythroblastotic newborn infants: II. Blood volume, hematocrit and serum albumin concentration in relation to hydrops fetalis. Pediatrics 1974;53: 13-23. 2. Grannum PA, Copel JA, Plaxe S, et a1. Intrauterine intravascular transfusion for severe erythroblastosis fetalis. N EnglJ Med 1986;314:1431-4. 3. Grannum PA, Copel JA. Moya F, et a1. The reversal of hydrops fetalis by intravascular intrauterine transfusion in severe isoimmune fetal anemia. AM J OBSTlT GYNECOL 1988;158:914-9. 4. Kleinman CS, Donnerstein RL, DeVore GR, et a1. Fetal echocardiography for evaluation of in utero congestive heart failure: a technique for the study of non-immune fetal hydrops. N EnglJ Med 1982;306:568-75. 5. Kleinman CS, Weinstein EM, Copel JA. Pulsed Doppler analysis of human fetal blood flow. Clin Diagn Ultrasound 1986; 17: 173-85. 6. Shepard MJ, Richards VA, Berkowitz RL, et a1. An evaluation of two equations for predicting fetal weight by ultrasound. AM J OBSTET GYNECOL 1982;142:47-54.
Fetal cardiac output in isoimmunized pregnancies
7. Hade L, Angelsen B. Doppler ultrasound in cardiology. 2nd ed. Philadelphia: Lea & Febiger, 1985:324. 8. Kenny JF, Plappert T, Doubilet P, et a1. Changes in intracardiac blood flow velocities and right and left ventricular stroke volumes with gestational age in the normal human fetus: a prospective Doppler echocardiographic study. Circulation 1986; 74: 1208-16. 9. Reed KL, Meijboom EJ, Sahn DJ, et a1. Cardiac Doppler flow velocitieb in human fetuses. Circulation 1986;73: 41-6. 10. Kenny JF, Plappert T, Doubilet P, et a1. Effects of heart rate on ventricular size, stroke volume, and output in the normal human fetus: a prospective Doppler echocardiographic study. Circulation 1987;76:52-8. 11. De Smedt MCH, Visser GHA, Meijboom EJ. Fetal cardiac output estimated by Doppler echocardiography during mid- and late gestation. Am J Cardiol 1987;60:338-42. 12. Nicolaides KH, Warenski JC, Rodeck CH. The relationship of fetal plasma protein concentration and hemoglobin level to the development of hydrops in rhesus isoimmunization. AMJ OBSTET GYNECOL 1985;152:341-4. 13. Barss VA, Doubilet PM, St. John-Sutton M, et a1. Cardiac output in a fetus with erythroblastosis fetalis: assessment using pulsed Doppler ultrasound. Obstet Gynecol 1987; 70:442-4. 14. Fumia FD, Edelstone DI, Holzman IR. Blood flow and oxygen delivery to fetal organs as functions of fetal hematocrit. AM J OBSTET GYNECOL 1984;150:274-82. 15. Sooth ill PW, Nicolaides KH, Rodeck CH, et al. Relationship of fetal hemoglobin and oxygen content to lactate concentration in Rh isoimmunized pregnancies. Obstet Gynecol 1987 ;69:268-71.
Twin-to-twin transfusion syndrome: What are appropriate diagnostic criteria? Fiona H. Danskin, MB, ChB,* and James P. Neilson, MD
Glasgow, Scotland One hundred seventy-eight consecutive twin pregnancies were studied to reevaluate the standard diagnostic criteria for chronic twin-to-twin transfusion syndrome of an intertwin hemoglobin difference >5 gm/dl and a birth weight difference >20%. Hemoglobin differences> 5 gm/dl were found in six pregnancies with monochorionic placentas but also in seven with dichorionic placentas. Birth weight differences >20% occurred no more commonly in monochorionic than in dichorionic pregnancies. Of the four pregnancies with a coexisting hemoglobin difference >5 gm/dl and birth weight difference >20%, only one had a monochorionic placenta and therefore likelihood of vascular anastomoses. Diagnosis of twin-to-twin transfusion syndrome cannot be definitively established by current standard diagnostic criteria. (AM J OBSTET GYNECOL 1989;161:365-9.)
Key words: Twin pregnancy, twin-to-twin transfusion syndrome, diagnosis
From the Department ofM idwzjery, University of Glasgow, the Queen Mother's Hospital. Received for publication August 7, 1988; revised December 8, 1988; accepted February 24, 1989. Reprints not available. *Dr. Danskzn holds the emden MedIcal Research Scholarship administered by the Scottish Hospital Endowments Research Trust.
Twin-to-twin transfusion syndrome is an important complication of twin pregnancy, I, 2 resulting from shunting of blood from one twin to the other through placental vascular anastomoses. Injection studies of twin placentas have shown that such anastomoses are almost universally present in monochorionic placentas
365
366
Danskin and Neilson
August 1989 Am J Obstet Gynecol
Table I. Zygosity studies in 178 consecutive twin pregnancies Monochorionic placenta Blood group studies Discordant sex Intrauterine death Technical failure
51) 19 70 Monozygotic
135) 87 Dizygotic 52
4) 21 Unknown 17
but so extremely rare in dichorionic placentas l • 3 that the possibility of the syndrome occurring with this type of placenta can be disregarded. Both acute and chronic forms of the twin-to-twin transfusion syndrome have been described. 4 • 5 The acute form has been attributed to rapid transfer of blood through superficial arteryto-artery or vein-to-vein anastomoses during labor and delivery, resulting in a hypovolemic, anemic donor twin and a hypervolemic, plethoric recipient twin of similar birth weight. 4 . 5 The chronic syndrome results from gradual antepartum transfusion through deep arteryto-vein anastomoses, rendering the donor twin not only hypovolemic and anemic but also growth retarded, whereas the hypervolemic, plethoric recipient is usually larger. 6 -s These features may be accompanied by oligohydramnios in the sac of the donor and polyhydramnios in that of the recipient and may be identifiable before birth by ultrasonography.9. to Twin-to-twin transfusion syndrome is usually diagnosed or confirmed after birth by the discovery of a hemoglobin difference >5 gm/dl between twins 2 and a birth weight difference >20%, providing evidence of the chronic form of the syndrome! We have reevaluated these standard and widely accepted diagnostic criteria as part of an ongoing intensive study of twin pregnancy. Despite an exhaustive MEDLINE literature search, we have not been able to identify any other such reevaluation. Material and methods
One hundred seventy-eight consecutive twin pregnancies delivering in The Queen Mother's Hospital, Glasgow, (a tertiary referral center) between August 1984 and February 1988 were studied. Of the mothers, 160 had attended the hospital for antenatal care, mostly at a special twin pregnancy antenatal clinic, and the other 18 were transferred from other hospitals, usually in established or incipient preterm labor. Venous blood samples were usually obtained from neonates by pediatricians on the first day of life for hematologic indexes, including hemoglobin concentration. In some
cases in which both babies were well, at term, and appropriate for gestational age, the pediatric staff considered venesection unnecessary on clinical grounds. All birth weights were recorded, and the intertwin birth weight difference was calculated as a percentage of the weight of the larger twin. Zygosity was determined in 88% of pregnancies (Table I) by inspection of the placentas (a monochorionic placenta indicating a monozygotic pregnancy), sex discordance (indicating a dizygotic pregnancy), or when necessary by comparing red blood cell antigens of the ABO, rhesus, MNS, Kell, Duffy, Kidd, Lutheran, and Wright systems. II In all 21 cases in which zygosity was not established, the placenta was dichorionic and the twins were of like sex; in four there had been intrauterine death of one twin, and in the remainder blood was not sent for grouping. Total perinatally related wastage, an index that includes late abortions, stillbirths, early and late neonatal deaths, and perinatally related infant deaths, 12 is shown in Table II. In all three monochorionic monozygotic pregnancies in which one or both fetuses were stillborn, features were compatible with, although not diagnostic of, twin-to-twin transfusion syndrome. In one preterm labor occurred at 34 weeks, and the leading twin died of asphyxia after abdominal dystocia caused by ascites; the second twin weighed 530 gm less and had a hemoglobin value at birth of 11 gmt dl. In another pregnancy acute polyhydramnios occurred in the amniotic sac of the first twin, which resulted in abruptio placentae and intrauterine death. The second twin, who survived, was very pale at birth and weighed 27% less than the first. In a third pregnancy acute polyhydramnios developed at 32 weeks, which was more marked in the amniotic sac of the first twin. Labor supervened within 24 hours and stillborn female infants were born; the first twin weighed 1.71 kg and appeared very plethoric with congested organs at postmortem examination, and the second twin weighed only 1.46 kg and appeared very pale. Hemoglobin values were not obtained in the latter two pregnancies. Results
The hemoglobin concentrations of venous blood samples withdrawn from both babies on the first day of life were obtained in 118 of the 178 pregnancies. Both hemoglobin values were known in 35 of the 51 monochorionic pregnancies, and in six there was an intertwin difference >5 gm/dl. However, there were two pregnancies in the monozygotic dichorionic group, one in the dizygotic group, and four in the zygosity unknown group. all of which had dichorionic placentas, in which hemoglobin differences >5 gm/dl were also found. Oligohydramnios occurred in the amniotic sac of one twin in only two of these 13 pregnancies, both
Diagnosis of twin-to-twin transfusion syndrome 367
Volume 161 Number 2
Table II. Outcome in 178 consecutive twin pregnancies TPRWllOOOt
Zygosity I placentation
Monozygotic monochorionic Monozygotic dichorionic Dizygotic Not known
3*
0
0
0
0
1 4
2
8§1I
0 2
4
102
93
38
38
0
174 42
164 32
0 I
0
64.3 57.5 238.1
NND,Neonatal deaths; PRID, perinatally related infant deaths; TPRW, total perinatally related wastage. *Perinatally related infant deaths are all infant deaths (between 1 month and 1 year of life) from or pardy from a primary
perinatal complication or its treatment.!2 tTotal perinatally related wastage includes all late abortions, except terminations of pregnancy for fetal abnormalities, all perinatal deaths, all late neonatal deaths (during the second, third, and fourth weeks of life), and all perinatally related infant deaths.!2 tOne exomphalos. § One anencephalic. IIOne cyanotic congenital heart disease. of which had dichorionic placentas. None of the twentysix babies in this group required treatment for anemia or polycythemia in the neonatal period. Birth weight differences, which were expressed as a percentage of the weight of the larger twin, were calculated for all pregnancies. Excluding those cases in which there was a fetal abnormality or a fetus papyraceus, a birth weight difference >20% was present in 17.4% of pregnancies. The distribution of these according to zygosity is shown in Table III. Oligohydramnios was again an infrequent occurrence in this group and was found in the sac of one twin in only four pregnancies. Three of these had a dichorionic placenta and one had a monochorionic placenta. In only four pregnancies did a hemoglobin difference >5 gm/dl and a birth weight difference >20% coexist, and none of these four pregnancies showed evidence of oligohydramnios or polyhydramnios in either amniotic sac. Only one of these pregnancies was associated with a monochorionic placenta, and in it alone did the heavier twin have the higher hemoglobin value (Table IV).
Comment Since the recognition of the twin-to-twin transfusion syndrome by Herlitz!3 in 1941, much interest has been focused on this rare complication of twin pregnancy. Initial publications were single case reports of the syndrome, whereas larger studies later attempted to define the incidence, pathogenesis, and associated morbidity and mortality!' 4 It was suggested that when a large volume of blood is transferred antepartum, one or both twins may die in utero! Since many early studies only considered those pregnancies resulting in the live birth of both babies, they probably underestimated the true incidence of the syndrome. In our study all three monochorionic pregnancies resulting in the stillbirth of
Table III. Birth weight difference >20% Zygosity I placentation
Monozygotic monochorionic Monozygotic dichorionic Dizygotic Not known
Btrth weIght difference >20%
%
51
8
15.7
19
4
21.1
87 21 178
14 5 31
16.0 23.8 17.4
N
one or both babies demonstrated features compatible with the twin-to-twin transfusion syndrome. However, the purpose of our study was neither to determine the significance of the twin-to-twin transfusion syndrome in contributing to perinatal mortality and morbidity nor to define its incidence (we did not have hemoglobin values from all sets of twins). Rather, we aimed to reevaluate the standard diagnostic criteria for the twin-to-twin transfusion syndrome-a matter not only of clinical importance but also vital for further perinatal research. Indeed, our study was generated by conflicting reports of fetal velocity waveforms of the umbilical artery as assessed by the new technology of Doppler ultrasonography on twin pregnancies believed complicated by the twin-to-twin transfusion syndrome. Concordant,!4 discordant,!5 and variable l6 patterns have been described, and these discrepancies may have stemmed from diagnostic confusion. Twin-to-twin transfusion syndrome has traditionally been diagnosed after birth by finding a hemoglobin difference >5 gm/dl-a criterion introduced by Rausen et al! in 1965. They had found such a hemoglobin difference in 19 of 130 monochorionic pregnancies, whereas in 43 dizygotic pregnancies no difference >3.3
368
Danskin and Neilson
August 1989 Am J Obstet Gynecol
Table IV. Birth weight difference >20% and hemoglobin difference >5 gm/dl Birth weIght Zygosity / placentation
Monozygotic monochorionic Monozygotic dichorionic Dizygotic dichorionic Dizygotic dichorionic
Twin 1 (kg)
2.85 2.39 1.47 3.16
I
Twin 2 (kg)
2.18 1.49 2.41 2.25
gm/ dl was found. Therefore they selected a hemoglobin difference> 5 gm/ dl as a suitable cutoff value, and this has been accepted by authors of subsequent publications as definitive evidence of the twin-to-twin transfusion syndrome. However, our study showed that although a hemoglobin difference> 5 gm/ dl existed in six monochorionic pregnancies, it also occurred in seven dichorionic pregnancies; in fact the largest hemoglobin differences (12.5 and 11.3 gm/ dl) occurred in the latter group. Because placental anastomoses are extremely rare in fused dichorionic placentas,3 it is unlikely that the twin-to-twin transfusion syndrome occurred in these seven cases. Thus a difference in hemoglobin concentration is not in itself sufficient to establish a diagnosis of twin-to-twin transfusion syndrome. Tan et al! studied 482 twin pairs in an attempt to differentiate the acute and chronic forms of the syndrome. They diagnosed the syndrome in 35 sets of twins on finding obvious plethora in one twin and pallor in the other, along with a hemoglobin difference >5 gm/dl. They divided the twins into two groups according to whether the birth weight difference between each pair was greater or less than 20% of the weight of the heavier twin. They found that in those twins who had a large weight difference, the twin with the lower hemoglobin was invariably lighter, whereas in those twins with a small weight difference, the twin with the lower hemoglobin value was heavier. They concluded that birth weight difference provides a simple yet reliable yardstick for differentiation of twin pregnancies with hemoglobin differences into acute and chronic forms. This additional criterion of birth weight difference >20% has been widely adopted. However, our study showed that a birth weight difference >20% occurred no more commonly in monochorionic pregnancies than in dichorionic pregnancies. In addition, of the four pregnancies in which a hemoglobin difference >5 gm/dl and a birth weight difference >20% coexisted, only one had a monochorionic placenta. In this pregnancy the heavier twin had the higher hemoglobin value, whereas in the three dichorionic pregnancies the lighter twin had the higher hemoglobin value. These presumably represent growth
Hemoglobin
I
Difference (%)
(gm I dl)
Twzn 1
23.51 37.66 39.00 28.80
24.0 14.9 22.9 13.7
I
Twzn2 (gm I dl)
16.0 27.4 17.0 25.0
I
Difference (gm I dl)
8.0 12.5 5.9 11.3
retardation in aSSOCIatIOn with polycythemia, a wellrecognized combination. Because Tan et al. used a hemoglobin difference >5 gm/dl as the sole diagnostic criterion for the twin-to-twin transfusion syndrome and did not take into account chorionicity of the placenta, it is likely that their subgroup included a similar proportion of polycythemic growth retarded babies from dichorionic pregnancies. In summary, the standard diagnostic criteria used in clinical obstetrics and perinatal research do not permit a definitive diagnosis of the twin-to-twin transfusion syndrome. Differences in hemoglobin and birth weight must be interpreted in the light of knowledge of placental morphology and whether the heavier or lighter twin had the higher hemoglobin value. Further studies are still required to elucidate this enigmatic condition. We gratefully acknowledge the contributions of Dr. A. A. M. Gibson, consultant pathologist, and the Regional Blood Transfusion Service, Law Hospital, in establishing zygosity, and the helpful criticisms of this manuscript by Professor C. R. Whitfield and Dr. B. Holland. REFERENCES I. Benirschke K. Twin placenta in perinatal mortality. NY
State] Med 1961;61:1499-508. 2. Rausen AR, Seki M, Strauss L. Twin transfusion syndrome.] Pediatr 1965;66:613-28. 3. Robertson EG, Neer KJ. Placental injection studies in twin gestation. AM ] OBSTET GYNECOL 1983; 147: 170-4. 4. Tan KL, Tan R, Tan SH, Tan AM. The twin transfusion syndrome. Clin Pediatr 1979;18: 11l-4. 5. Klebe]G, Ingomar CJ. The fetoplacental circulation during parturition illustrated by the interfetal transfusion syndrome. Pediatrics 1972;49:112-6. 6. Becker AH, Glass H. Twin to twin transfusion syndrome. Am] Dis Child 1963; 106:624-9. 7. Corney G, Aherne W. The placental transfusion syndrome in monozygotic twins. Arch Dis Child 1965;40:26470. 8. Naeye RL. Human intrauterine parabiotic syndrome and its complications. N Engl] Med 1963;268:804-8. 9. BrennanJN, Dirnan RV, Rosen MG, Bellon EM. Fetofetal transfusion syndrome: prenatal ultrasonographic diagnosis. Radiology 1982; 143 :535-6. 10. Wittman BK, Baldwin V], Nichol B. Antenatal diagnosis of twin transfusion syndrome by ultrasound. Obstet GynecoI1981;58:123-7. II. Wilson RS. Blood typing and twin zygosity. Hum Hered 1970;20:30-56.
Volume 161 Number 2
12. Whitfield CR, Smith NC, Cockburn F, Gibson AAM. Perinatally related wastage-a proposed classification of primary obstetric factors. Br J Obstet Gynaecol 1986;93:694703. 13. Herlitz G. Zur Kennitis der anamischen und polyzytamischen Zutstande bei Neugeborenen, sowie des Icterus gravis neonatorum. Acta Paediatr (Uppsala) 1941 ;29:211. 14. Giles WB, Trudinger BJ, Cook CM. Fetal umbilical artery flow velocity-time waveforms in twin pregnancies. Br J Obstet Gynaecol 1985;92:490-7.
Diagnosis of twin-to-twin transfusion syndrome
15. Farmakides G, Schulman H, Saldana LR, Bracero LA, Fletcher A, Rochelson B. Surveillance of twin pregnancy with umbilical arterial velocimetry. AMJ OBSTET GVNECOL 1985; 153:789-92. 16. Erskine RLA, Ritchie JWK, Murnaghan GA. Antenatal diagnosis of placental anastomoses in a twin pregnancy using Doppler ultrasound. Br J Obstet Gynaecol 1986; 93:955-9.
The prevalence and biologic significance of lupus anticoagulant and anticardiolipin antibodies in a general obstetric population C. J. Lockwood, MD, R. Romero, MD, R. F. Feinberg, MD, PhD, L. P. Clyne, FIMLS, MS, B. Coster, BS, andJ. C. Hobbins, MD New Haven, Connecticut Circulating antibodies to negatively-charged phospholipids have been implicated in the genesis of adverse pregnancy outcomes. However, it has yet to be established that these antibodies are causative or that they are invariably associated with untoward perinatal outcomes. To address this issue, the prevalence of lupus anticoagulant and anticardiolipin antibodies was recorded in a low-risk obstetric population, and the outcomes of untreated pregnancies were evaluated. Two of 737 patients (0.27%) had lupus anticoagulant documented by a prolonged activated partial thromboplastin time that did not correct with mixing studies. In comparison, greatly elevated concentrations of immunoglobulin M-anticardiolipin antibodies or immunoglobulin G-anticardiolipin antibodies were identified in 16/737 (2.2%) patients by means of an enzyme-linked immunosorbent assay. Spontaneous abortions occurred in both lupus anticoagulant-positive patients, suggesting that the activated partial thromboplastin time used was a relatively insensitive but specific marker for anti phospholipid antibody-associated adverse pregnancy outcomes. In contrast, although 12 of 16 anticardiolipin antibodies-positive pregnancies were complicated by perinatal loss, preterm delivery, or fetal growth retardation, four patients had uncomplicated pregnancies. Moreover, the distribution of anti cardiolipin antibodies concentrations in these four patients was not clustered among the lowest anti cardiolipin antibodies values, and anticardiolipin antibodies concentrations correlated weakly with adverse outcomes. These findings suggest that antiphospholipid antibodies are related to adverse pregnancy outcomes in a complex fashion and that therapy is not always required for acceptable outcomes in patients without other risk factors. (AM J OSSTET GVNECOL 1989;161:369-73.)
Key words: Lupus anticoagulant, anticardiolipin antibodies, activated partial thromboplastin time Circulating lupus anticoagulants are either identical to or closely associated with immunoglobulin M (IgM) and immunoglobulin G (IgG) antibodies to negativelycharged phospholipids.' A laboratory characteristic of lupus anticoagulant is the inhibition of the activated partial thromboplastin time and its derivative assays.' From the Departments of Obstetrics and Gynecology and MediCine, Yale University School of MediCine. Received for publication july 9,1988; revised September 18,1988; accepted February 24,1989. Reprint requests: Charles J. Lockwood, MD, Director of Pennatal Research, St. Margaret's Hospital, Tufts Unzversl/'V, School of Medicine, 90 Cushing Ave., Boston, MA 02125.
This in vitro laboratory phenomenom is a consequence of antibody binding to negatively-charged phospholipid surfaces on which factors Xa and Va convert prothrombin to thrombin.' Paradoxically, in vivo lupus anticoagulant and related antiphospholipid antibodies promote thrombosis l via a myriad of potential prothrombotic mechanisms. 2•6 Reportedly 94% of pregnancies occurring in patients with lupus anticoagulant who do not receive appropriate treatment result in spontaneous abortion or stillbirth. 7 A 13% to 40% prevalence of antibodies to the divalent phospholipid, cardiolipin, has been noted in patients evaluated for a history of "idiopathic" recur-
369