Outcome of twin, triplet, and quadruplet in vitro fertilization pregnancies: the Norfolk experience*

FERTILITY AND STERILITY

Vol. 57, No.4, April 1992

Printed on acid-free paper in U.S.A.

Copyright c 1992 The American Fertility Society

Outcome of twin, triplet, and quadruplet in vitro fertilization pregnancies: the Norfolk experience*

Muhieddine A_-F_ Seoud, M_D_t James P. Toner, M.D., Ph.D. Catherine Kruithoff, B.S.N. Suheil J. Muasher, M.D. The Jones Institute for Reproductive Medicine, Eastern Virginia Medical School, Department of Obstetrics and Gynecology, Norfolk, Virginia

Objective: To review the maternal morbidity and neonatal morbidity and mortality associated with in vitro fertilization (IVF) multiple pregnancies. Design: Retrospective analysis of data collected from office and hospital records and from questionnaires sent to patients, their obstetricians, and pediatricians. Setting: Patients (all with private insurance carriers) enrolled in an academic IVF program (The Jones Institute for Reproductive Medicine). Patients, Participants: All IVF pregnancies resulting in one or more gestational sacs on the initial ultrasound at 6 to 7 weeks were reviewed. Main Outcome Measures: The frequency and severity of obstetrical and neonatal complications and the perinatal mortality of IVF twins, triplets, and quadruplets were compared. These were also compared with non-IVF multiple pregnancies. Results: From 1982 to 1990, 629 IVF pregnancies progressed beyond 20 weeks; 115 twins (18.3%), 15 triplets (2.4%), and 4 quadruplets (0.6%). There was a high incidence of antenatal complications such as abortions (30.3%,42%, and 20%), premature labor (41.5%, 92.3%, and 75%), pregnancyinduced hypertension (17.0%, 38.6%, and 50% ), and gestational diabetes mellitus (3.1 %, 38.5 %, and 25%) for twins, triplets, and quadruplets, respectively. The mean gestational age at delivery was 35.5 ± 3.7, 31.8 ± 2.7, and 31.0 ± 1.7 weeks, respectively. There was also a proportionate progressive increase in neonatal complications. The mean weights were 2,473 ± 745, 1,666 ± 441 and 1,414 ± 368 g, respectively. Twins (22.7%), 64.1 % of triplets, and 75% of quadruplets needed admission to the neonatal intensive care unit and remained for an average of 12.0 ± 2.3, 17.4 ± 14.0, and 57.8 ± 17.9 days, respectively. There was no difference in the mean Apgar scores or the incidence of congenital malformations in the three groups. The corrected perinatal mortality rates were 38.5, 0.0, and 0.0 per thousand live births, respectively. . Conclusion: Triplet and quadruplet IVF pregnancies have increased obstetrical and neonatal complications compared with IVF twins. The perinatal mortality and the incidence of congenital Fertil Steril 1992;57:825-34 malformations are, however, comparable in all three groups. Key Words: Twin, triplet, quadruplet, pregnancy, IVF

Received September 3, 1991; revised and accepted December 20,1991. * Presented in part at the 7th World Congress of IVF and Assisted Procreation, Paris, France, June 28 to July 3, 1991; and the 47th Annual Meeting of The American Fertility Society, Orlando, Florida, October 18 to 25, 1991. t Reprint requests: Muhieddine A-F Seoud, M.D., Eastern Virginia Medical School, Department of Obstetrics and Gynecology, 825 Fairfax Avenue, Norfolk, Virginia 23507. Vol. 57, No.4, April 1992

The incidence of naturally occurring multiple birth of higher order is very low (1, 2). The introduction of ovulation inducing agents and, more recently, of assisted reproduction has resulted in a substantial increase in the incidence of multiple pregnancies of up to 10 times (1, 2). As the number of multiple fetus increases in utero, so does the incidence of complications and the possibility of adverse outcome. Seoud et al.

IVF twins, triplets, and quadruplets

825

Most, if not all, of the recent literature concerning twin, triplet, and quadruplet pregnancies report data from naturally occurring multiple gestations alone or in combination with pregnancies resulting from ovulation induction. Few have solely reported on the outcome of in vitro fertilization (IVF) pregnancies (3). More data are available on naturally occurring twins than on higher order pregnancies (1,4-6). A recent review of 198 triplets was collected from 24 centers around the country, and only 4 centers had >15 sets (7). Another review of 71 quadruplets was drawn from 27 states and two Canadian provinces (2). Sassoon et al. (8) were among the few who retrospectively compared well-matched twin and triplet pregnancies and found worse perinatal outcome in the latter group; however, only 6.7% of their pregnancies were the result of IVF. This makes it difficult to formulate recommendations concerning the standard of care as well as the exact incidence of complications especially as it relates to the somewhat older IVF population. The recent improvement in the perinatal mortality of twins, triplets, and quadruplets, especially those resulting from ovulation induction may relate to early detection, closer follow-up, better neonatal intensive care, and stronger motivation to adhere to medical care among infertility patients. One recent review from England revealed a lower perinatal mortality in triplets as compared with twins (3). The purpose of this study is to present the experience of the Jones Institute for Reproductive Medicine with IVF multiple pregnancies in relation to the maternal morbidity and the neonatal morbidity and mortality. MATERIALS AND METHODS

The patient population included all patients enrolled in IVF from 1982 to 1990. The data were retrieved from the office charts and through detailed questionnaires sent to patients, obstetricians, and pediatricians in the various institutions in which these patients were managed. When made available, the obstetrician's office charts and the hospital records were reviewed. Standard definitions were used in evaluating outcome variables: premature labor was defined as the occurrence of uterine contractions resulting in progressive cervical change before 37 weeks. Premature rupture of the membranes (PROM) was defined as rupture of the membranes before the onset of labor without respect to the gestational age. Premature cervical dilatation was diagnosed when there was premature effacement and dilatation of the cervix 826

Seoud et al.

IVF twins, triplets, and quadruplets

in the absence of recognizable uterine activity in the second trimester. Morning sickness was considered mild when patients were able to continue taking fluid and solid foods, moderate when patients only tolerated fluids, and severe (hyperemesis gravidarum) when necessitating hospitalization or accompanied by signs of dehydration, electrolyte imbalance, or significant ketonuria. Pregnancy-induced hypertension was defined as hypertensive disorders of pregnancy occurring after 20 weeks of gestation with a blood pressure of 140/90 or greater on two occasions 6 hours apart or an increase of 30 mm Hg systolic or 15 mm Hg diastolic over first trimester blood pressure with albuminuria. It was considered severe ifthe blood pressure was> 160/110 or ifthere were visual symptoms or severe headache, abnormal liver function tests, thrombocytopenia, oliguria of <400 mL per 24 hours, or pulmonary edema (9). Anemia was defined as a hematocrit < 30%. Intrauterine growth retardation (IUGR) was diagnosed when the ultrasonographic estimation of fetal weight was less than the 10th percentile for gestational age, whereas small for gestational age was defined by neonatal weights less than the 10th percentile using the California growth tables for multiple births (10). Gestational diabetes mellitus was diagnosed when there was an abnormal fasting blood sugar or two abnormal values in the 3 hours oral glucose tolerance test (9). Congenital anomalies were considered major when they involved a major organ, caused significant cosmetic or functional problems, or required major surgical correction. When not otherwise specified in the text, pregnancy referred to those gestations with one or more fetuses with fetal heart motions. In the statistical analysis, X2 and Student's t-test were used, and P < 0.05 was considered significant.

RESULTS

From 1982 to 1990, 190 IVF patients had two or more gestational sacs seen on the initial ultrasound at 6 to 7 weeks of gestation. Of these, 7 had four sacs, 32 had three sacs, and 151 had two sacs. Of the seven patients with four sacs, two patients had fetal heart motions in only two of the four sacs, one continuing as twin and the other as a singleton beyond 20 weeks. Five patients had fetal heart motions in all four sacs, one aborted all sacs at 18 weeks, and four continued beyond 20 weeks. Ofthe 32 patients with three sacs, 26 patients had fetal heart motions in all three sacs. Of these, 15 continued as triplets, 11 as twins, and four as singletons beyond 20 weeks. Of the 151 patients with Fertility and Sterility

two sacs, 103 continued as twins and 31 as singletons after 20 weeks. Collectively, 165 patients had two fetal heart motions on the initial scan, 26 had three, and 5 had four. These resulted in 115 twin, 15 triplet, and four quadruplet (quad) gestations that progressed beyond 20 weeks. Demographic and Basic IVF Data

Table 1 lists the characteristics of the three groups of patients. The patients with twins were significantly older than the other two groups. The baseline day 3 follicle-stimulating hormone (FSH) was significantly lower in the triplets and quads, and their luteinizing hormone (LH)/FSH ratio was higher, indicating that they constituted a relatively high responding group. This, in addition to their relatively younger age, translated into a higher peak estradiol and high number of oocytes retrieved and fertilized. This was not significant except when comparing the twins to the quads. Twins also had a significantly lower number of embryos transferred. There was no difference in the primary cause of infertility among IVF patients who conceived with twins, triplets, and quadruplets. The most common cause was tubal infertility (58.3%, 42.3%, and 40.0%), followed by endometriosis (14.7%, 23.1%, and 20.0%), and male infertility (15.3%, 23.1 %, and 0.0%), respectively. The ovarian hyperstimulation protocols were also comparable in the three groups. Human menopausal gonadotropin and/or pure FSH were administered in 71.7%, 65.4%, and 60%, respectively. Lutealleuprolide acetate (LA) was added in 25.0%, 34.7%, and 40%, respectively, whereas folTable 1

licular LA was used in five patients who conceived with twins. Obstetrical Complications

Of the 115 twin pregnancies, 110 are delivered and 5 are presently continuing beyond 20 weeks. Among the twin patients who delivered, complete follow-up is available on 94 patients, whereas only partial information is available on the remaining 16 patients. Complete follow-up is available on the 4 patients with quads and 13 patients with triplets. Table 2 lists these complications. All triplet and quadruplet pregnancies were followed up and delivered in tertiary care centers, whereas only 30% of the twins were initially managed in tertiary care centers. Abortion

There was a high incidence of first trimester bleeding that increased (but not significantly so) with the increase in the order of pregnancy. Abortion was however more difficult to define. If we consider patients who lost all fetuses (with previously documented fetal heart motions), then 17 of the 165 twins (10.3%), 2 of the 26 triplets (7.7%), and 1 of the 5 quads (20%) fit that definition. On the other hand, if we include loss of one or more (but not all) fetuses with previously documented fetal heart motions, the abortion rate becomes 50 of 165 (30.3%) for the twins and 11 of 26 (30.3%) for the triplets but remains 1 of 5 (20%) for the quads, with no statistically significant difference among the three groups.

Demographic and IVF Data in IVF Patients With Multiple Pregnancies: the Norfolk Experience

Twins Age (y) Gravidity Parity Previous abortion Living child Weight (kg) Day 3 FSH (mIU/mL) Day 3 LH (mIU/mL) LH/FSH Peak E2 (pg/mL) Preovulatory oocytes retrieved Preovulatory oocytes fertilized Preovulatory embryos transferred * Values are means ± SD. Vol. 57, No.4, April 1992

34.1 ± LO± 0.2 ± 0.4 ± 0.1 ± 58.4 ± 10.3 ± 14.1 ± 1.5 ± 838 ±

0.3* 0.1 0.1 0.1 0.1 0.6 0.6 0.7 0.1 39

Triplets 32.2 ± 3.5 1.0 ± 0.3 0.3 ± 0.5 0.4 ± 0.2 0.3 ± 0.7 54.8± 11.5 7.0 ± 2.8 12.3 ± 5.0 2.0 ± 1.5 907 ± 536

Quadruplets 32.8 ± 2.6 0.4 ± 0.9 0.0 0.2 ± 0.4 0.2 ± 0.4 57.7 ± 7.4 8.0± 5.0 17.5 ± 6.0 2.6 ± 1.6 1,043 ± 540

P 2 versus 3

P 2 versus 4

P 3 versus 4

<0.001 NSt <0.04 NS <0.001 <0.001 <0.001 <0.001 <0.001 NS

<0.001 <0.001 <0.001 <0.001 <0.001 NS <0.001 <0.001 <0.001 <0.001

<0.001 <0.01 NS NS NS NS NS NS NS NS

7.6 ± 0.6

7.3 ±

2.9

11.4 ±

6.3

NS

<0.001

<0.04

5.7 ± 0.6

5.1 ±

2.0

10.2 ±

5.3

<0.003

<0.001

<0.001

3.7 ± 0.1

4.2 ±

1.0

4.0±

0.0

<0.001

<0.001

NS

t NS, not significant. Seoudet al.

IVF twins, triplets, and quadruplets

827

Table 2

Obstetrical Complications in IVF Patients With Multiple Pregnancies: the Norfolk Experience

Obstetrical complications First trimester bleed Abortion Morning sickness Premature onset of labor PROM Pregnancy-induced hypertension HELLP Gestational diabetes mellitus Third trimester bleeding Weight gain (kg) IUGR Urinary tract infection Anemia (hct < 30)

P

P

P

Twins

Triplets

Quadruplets

2 versus 3

2 versus 4

3 versus 4

51/94 (54.3)* 50/165 (30.3) 51/94 (54.3) 39/94 (41.5) 13/94 (13.8) 16/94 (17.0) 2/94 (2.1) 3/94 (3.1) 5/94 (5.3) 20.6 ± 0.8* 11/94 (14.9) 8/94 (8.5) 7/48 (14.6)

14/26 (53.8) 11/26 (42.0) 6/13 (46.2) 12/13 (92.3) 2/13 (15.4) 5/13 (38.6) 1/13 (7.7) 5/13 (38.6) 1/13 (7.7) 23.0 ± 7.7 1/13 (7.7) 2/13 (15.4) 3/13 (23.1)

4/5 (80) 1/5 (20) 2/5 (40) 3/4 (75) 0

NSt

NS

NS

NS

NS

NS

NS

NS

NS

<0.001

NS

NS

NS

NS

NS

2/4 (50) 0

<0.005

NS

NS

NS

<0.01

NS

1/4 (25) 0

<0.001

<0.02

NS

NS

NS

NS

31.0 ± 4.1 0

<0.05

<0.000

NS

NS

NS NS

1/4 (25) 0

NS

NS

NS

NS

NS

NS

* Values in parentheses are percents.

* Values are means ± SD.

t NS, not significant.

Premature Onset of Labor and Premature Cervical Dilatation

The most frequent complication in the second and third trimesters was premature labor. In twins, it started at a mean gestational age of 30.2 ± 0.7 weeks. Fifteen of the 39 twins with premature labor delivered within 24 hours, whereas the rest carried for a mean of 5.4 ± 3.8 weeks. Twelve of the 13 triplets had premature labor at an average gestational age of 28.8 ± 2.9 weeks. Half of those triplets delivered within 24 hours, whereas the other half carried for a mean of 4.2 ± 3.9 weeks. Three of the quads had premature labor at an average of 28.6 ± 1.2 weeks and carried for a mean of 2.7 ± 0.6 weeks; one of these patients was delivered because of pregnancy-induced hypertension. Thus as the number of fetuses increased, the incidence of premature labor increased. In addition, premature labor tended to occur earlier, and postponing delivery became less successful. Three twin patients had premature cervical dilatation. One of the three patients had a history of cervical conization. Two patients had cervical cerclage placed and carried to 34 and 36 weeks, whereas the third was placed on bed rest and carried to 32.5 828

Seoud et al.

IVF twins, triplets, and quadruplets

weeks. None of the triplets or quads had this complication. Premature Rupture of the Membranes

Thirteen twin pregnancies had PROM at a mean of 33.4 ± 5.9 weeks, 8 of whom were preterm. Twelve of the 13 patients delivered within the same day, whereas the 13th patient carried 2 more weeks. Two triplets had PROM at 31 and 35.3 weeks, and both delivered within the same day. None of the quads had PROM. Bed Rest

Three of the quads were admitted to the hospital at 24 weeks of gestation, whereas the fourth was put on bed rest at home at 24 weeks. Eight of the triplets were electively admitted to the hospital at 24 weeks, and all were advised bed rest at home from 20 weeks onward. Twenty-seven of the twins were put on bed rest at home at a mean gestational age of 24.4 ± 4.3 weeks; they delivered at a mean gestational age of 36.0 ± 2.6 weeks. Seventeen patients with twins were not advised bed rest and continued their activities as usual. They delivered at a mean gestational age of 37.5 ± 2.3 weeks. The gestational age at delivery was not significantly different between the two groups (P = 0.06). Fertility and Sterility

Pregnancy-Induced Hypertension and Gestational Diabetes Mellitus

Sixteen twin pregnancies developed pregnancyinduced hypertension at an average gestational age 29.6 ± 0.2 weeks; 13 were mild, one was severe, two had HELLP syndrome hemolysis, elevated liver enzymes, and low platelet (HELLP) Syndrome. Pregnancy-induced hypertension in five triplets developed at an average of 33.0 ± 2.2 weeks, and one of them had HELLP. Two quads had pregnancy-induced hypertension at 28 and 31 weeks, respectively. There were no eclamptic patients in any group. Gestational diabetes mellitus was significantly less frequent in the twins than the triplets or quads; none required insulin therapy for adequate glycemic control. Third Trimester Bleeding

Of the five twins with third trimester bleeding, three had placenta previa, one had placental abruption, and one remained unexplained. In addition, one patient had an isolated thrombocytopenia with no clinical bleeding. Only one of the triplets had placental abruption, whereas none of the quads had antenatal bleeding. Weight Gain

Patients with twins gained an average of 20.6 ± 0.8 kg, whereas triplets added an average of 23.0 ± 7.7 kg, and the quads gained 31.0 ± 4.1 kg. The difference was only significant when comparing the twins to the triplets and quads. Intrauterine Growth Retardation

Eleven twin pregnancies were complicated by IUGR, involving 14 fetuses diagnosed at a mean gestational age of 27.8 ± 5.2 weeks (range 16 to 33 weeks). One patient who had systemic lupus erythematosus (SLE) developed IUGR in both fetuses at 22 weeks; one died in utero at 28 weeks and one survived. Another patient had IUGR of both fetuses starting 32 weeks, and both did well. Among the 9 other pregnancies with IUGR involving only one fetus, three had congenital anomalies (one had trisomy 21 and died neonatally, one had diaphragmatic hernia and died neonatally, whereas the third had Dwayne's syndrome and survived). Two other fetuses with IUGR died, one ante natally and the other neonatally, whereas the remaining four survived. In total, 3 of the 14 IUGR fetuses were congenitally abnormal, and 5 of the 14 IUGR fetuses died (two with anomalies). Vol. 57, No.4, April 1992

One triplet pregnancy was diagnosed antenatally as having IUGR involving one fetus at 28 weeks. She was delivered, and all three neonates did well. None of the quads was diagnosed antenatally as having IUGR. Intrauterine Fetal Death

Among the twin patients, five fetuses died antenatally. One fetus had nonimmune hydrops and died at 29 weeks, whereas its counterpart was delivered at 29 weeks and did well. The second patient had SLE, and one of her twins had IUGR and died at 28 weeks, whereas its counterpart had congenital heart block and was delivered at 29 weeks and did well. The third patient also had IUGR involving one fetus starting at 20 weeks leading to fetal death at 31 weeks. Its counterpart survived and was delivered at 35 weeks without complications. Two other patients had unexplained intrauterine fetal death (IUFD) at 26.5 and 39.5 weeks, and in both the other fetus survived and was delivered intact at term. The average gestational age for the IUFD was 28.9 ± 6.2 weeks. This gives us an uncorrected stillbirth rate of 5/220 (22.7%), which corrects to 4 of 219 (18.3%) when considering only those past 28 weeks. There were no IUFDs in the triplets and the quads. Other Complications

Morning sickness was a frequent complication in all three groups but was severe enough to require hospitalization in only two of twin patients. Among the twin pregnancies, one patient had neuroblastoma as a child and delivered both sacs at 22 weeks shortly after an exploratory laparotomy for small bowel obstruction. Another patient with twins was Rh isoimmunized, but both of her fetuses did well. Two twin pregnancies were complicated by nonimmune hydrops, one of whom eventually died in utero. Two twin and one triplet pregnancies were complicated by polyhydramnios, whereas two twin pregnancies had severe oligohydramnios (both in conjunction with IUGR). Three twin, one triplet, and one quadruplet pregnancies were complicated by ovarian hyperstimulation syndrome, one of which was severe. One patient with twins had severe carpel tunnel syndrome and had to undergo surgery. One patient with twins and one with triplets developed iatrogenic pulmonary edema during tocolysis for premature labor and both were easily treated. One patient with twins underwent cholecystectomy in the second trimester. One quad, two triplet, and seven twin pregnancies developed urinary tract infections. Seven of 48 Seoud et al.

IVF twins, triplets, and quadruplets

829

(14.6%) twin pregnancies had hematocrit of <30%, whereas three ofthe 13 triplets but none ofthe quads were anemic. Peripartum Events

As the number of fetuses increased, the length of gestation significantly decreased. The peripartum complications are listed in Table 3. Antenatal Hospitalization

Approximately one third of patients with twins were hospitalized antenatally for various reasons and remained for 1 to 56 days. More than half of those with triplets were admitted and remained in the hospital for 4 to 63 days, whereas three of the quads were hospitalized for 21 to 77 days. The differences among the three groups were statistically significant for both the percentage requiring hospitalization and the length of stay.

29 patients in whom both fetuses were delivered vaginally, 23 were vertex/vertex, 2 were vertex/ breech, 2 were vertex/transverse, and 2 were unknown (at 22 and 23 weeks). In 3 patients, the first fetus was delivered vaginally and the second abdominally; they presented as vertexfbreech for two and vertex/vertex for one. Nine patients who delivered twins were reported to have postpartum hemorrhage; two were significant enough to have blood transfusions. Among the nine patients with twins who developed postpartum fever, four had endomyometritis, one had a ruptured appendix, one had wound dehiscence, and in three the diagnosis was not indicated. One of the triplets had postpartum endomyometritis, and one had wound dehiscence. One of the quads had postpartum mastitis. Neonatal Outcome

Delivery Mode

Neonatal Birth Weight and Discordance

Seventy-five of 107 (70.1 %) twins had a cesarean section for both twins, whereas 29 (27.1 %) delivered vaginally, and three patients had a cesarean section for the second twin. The delivery was performed emergently in most cases. All the triplets and quads were delivered abdominally without attempt at vaginal delivery. In 11 of the 13 triplets and in three of the four quads this was done emergently. Among 103 twin pregnancies in which the primary indication for delivery was documented, 16 were for premature labor, 22 for PROM, 10 for pregnancyinduced hypertension, 6 for IUGR, 3 for third trimester bleeding, and 46 for being at term or in labor at term. In the triplets, 6 were for premature labor, 3 for pregnancy-induced hypertension, 2 for PROM, 1 for third trimester bleeding, and 1 for mature lung profile (lecithin/sphingomyelin ratio of >2 and presence of phosphatidyl glycerol in the amniotic fluid). Two of the quads were delivered because of premature labor, one for pregnancy-induced hypertension, and the fourth because she reached 33 weeks and the estimated fetal weights were around 2,000 g. Premature labor was significantly more responsible for preterm delivery in the triplets than in the twins, and more twins were delivered at term than triplets or quads. Epidural anesthesia was the most frequent form of anesthesia used in all three groups. The presentation of the fetuses at delivery was not of importance in the triplets and quads because they were all delivered abdominally. However, in 32 twins an attempt at vaginal delivery was undertaken; in the

Table 4 compares the neonatal outcome in the three groups. The weight of the neonates decreased significantly with the increase in the order of the pregnancy. The incidence of small for gestational age was 11 of 220 (5%) among the twins (4 twin As and 7 twin Bs) and 3 of 39 (7.7%) in the triplets. None of the quads was considered small for gestational age. Fourteen of 95 (14.7%) twin pregnancies were complicated by discordance in the neonatal weights of> 25%, and of these, 7 (the smaller of the set) died neonatally (1 had trisomy 21, one a diaphragmatic hernia, and two were at 26 weeks; there was no information on the remaining three neonates). Fourteen of the remaining 21 neonates needed admission to the neonatal intensive care unit (NICU), but all did well. Five of 95 (5.2%) twin pregnancies had a discordance of 20% to 25%, and all did well with only one neonate needing NICU admission, whereas 11 of 95 (11.6%) pregnancies had a discordance between 15% and 20%, and among these, 2 had nonimmune hydrops, 1 had diaphragmatic hernia and died neonatally, and 9 required NICU admissions. In the 65 of95 (68.4%) twin pregnancies with <15% discordance, there were 3 stillbirths, 2 neonatal death, and 8 of the remaining neonates required NICU admission. In two of the 13 (15.4%) triplet pregnancies, at least one of the neonates was discordant from the others by >25% and did well. In three other triplets (23%), there was a discordance of 15% to 25%, and both neonates did well. In two quadruplet pregnan-

830

Seoud et al.

IVF twins, triplets, and quadruplets

Fertility and Sterility

Table 3

Peripartum Events in IVF Patients With Multiple Pregnancies: the Norfolk Experience

Gestational age (wks) Cesarian section

Twins

Triplets

Quadruplets

P 2 versus 3

P 2 versus 4

P 3 versus 4

35.5 ± 3.7*

31.8 ± 2.7

31.0 ± 0.7

<0.001

<0.01

NSt

78/107 (72.9)* 37/98 (37.8) 61/98 (62.2)

13/13 (100) 2/13 (15.4) 11/13 (84.6)

4/4 (100) 1/4 (25) 3/4 (75)

<0.02

NS

NS

NS

NS

NS

NS

NS

NS

48/68 (70.6) 11/68 (16.2)

11/13 (84.6) 2/13 (15.4)

3/4 (75) 1/4 (25)

NS

NS

NS

NS

NS

NS

16/103 (15.5) 22/103 (21.4) 10/103 (9.7)

6/13 (46.2) 2/13 (15.4) 3/13 (23.0)

2/4 (50) 0

<0.01

NS

NS

NS

NS

NS

1/4 (25)

NS

NS

<0.02

6/103 (5.8) 3/103 (2.9) 46/103 (44.7) 31/94 (32.9) 16.5 ± 18.3

0

0

NS

NS

NS

1/13 (7.7) 1/13 (7.7) 8/13 (61.5) 22.9 ± 19.4

0

NS

NS

NS

1/14 (25) 3/4 (75) 56.0 ± 30.5

<0.01

NS

NS

<0.05

<0.01

NS

NS

<0.003

03

Elective Emergent Anesthesia Epidural General Delivery indication Premature onset of labor PROM Pregnancyinduced hypertension IUGR Third trimester bleeding Term ± labor Hospital admission Hospital days

* Values are means ± SO. t NS, not significant.

cies there was a discordance of 21.3% and 22.4%, and all neonates did well. Immediate Neonatal Course

The Apgar scores at 1 minute were significantly lower in the quads; however, the 5-minute scores were comparable in the three groups. In addition, the number of twins, triplets, and quads with 5minute scores of <7 were 23 of 193 (11.9%), 1 of 39 (2.6%), and 1 of 16 (6.3%), respectively. There were no differences in the Apgar scores at 1 and 5 minutes between twin A (7.5 ± 2.1 and 8.5 ± 2.0) and twin B (7.2 ± 2.2 and 8.3 ± 2.1), between triplet A (7.7 ± 1.5 and 8.3 ± 1.2), triplet B (7.5 ± 1.1 and 8.5 ± 1.0), and triplet C (7.0 ± 1.8 and 8.1 ± 1.0) and between quad A (7.3 ± 1.2 and 8.3 ± 1.2), quad B (8.0 ± 0.0 and 8.7 ± 0.6), quad C (5.3 ± 3.0 and 7.7 ± 0.6), and quad D (4.7 ± 3.1 and 7.3 ± 1.5). However, there was a tendency for the Apgar scores to be lower as the number of fetuses increased and as the order of delivery of the fetuses increased. This was significant for the 1-minute scores in the quads but not for the rest. Vol. 57, No.4, April 1992

* Values in parentheses are percents.

The ratio of male over female was 1.04, 3.09, and 1.00 for the twins, triplets, and quads, respectively, with a significant difference only between the first two groups. The ratio for the entire IVF population was 1.20. Respiratory distress syndrome secondary to hyaline membrane disease occurred in 35 of 181 (19.3%), 17 of39 (43.6%), and 12 of16 (75%) of twins, triplets, and quads. The percentage of neonates needing admission to the NICU, the days spent in the NICU, the percentage of neonates requiring assisted ventilation' the total number of days on the ventilator, and the total number of days spent in the hospital increased significantly from twins to triplets to quads as listed in Table 4. Hyperbilirubinemia occurred in 23 of 181 twins (12.7%),28 of39 triplets (71.8%), and 12 of 16 quads (75%). Three triplets had intraventricular hemorrhage (IVH). Two had grade I and did well, and one had grade III and died at 7 weeks of age. One quad had a grade I IVH and did well. Because not all the twins were screened for IVH, the exact frequency cannot be computed. Seoud et al.

IVF twins, triplets, and quadruplets

831

Table 4

Neonatal Complications in IVF Patients With Multiple Pregnancies: the Norfolk Experience

Twins

Triplets

Quadruplets

Weight (g) NICU

2473 ± 745* 4/181

NICU (d) Ventilatory assistance

12.0 ± 2.3 23/181 (12.7) 2.4 ± 2.3 8.9 ± 10.2

1,666 ± 441 25/39 (64.1) 17.4 ± 4.0 14/39 (36.9) 6.3 ± 7.0 28.1 ± 16.2

1,414 ± 368 12/16 (75) 57.8 ± 17.9 11/16 (68.8) 9.8 ± 11.5 52.7 ± 25.1

7.4 ± 2.5 8.5 ± 1.9 112/108

7.3 ± 1.1 8.3 ± 1.1 34/11

(22.m

Ventilatory (d) Hospital (d) Apgar 1 min 5 min Male/female Anomalies Minor

0/220 (0.0) 7/220 (3.9) 8/208 (38.5)

Major Perinatal mortality rate >28 wks

1/45 (2.2) 0/45 (0.0) 0.0

* Values are means ± SD. t Values in parentheses are percents.

Congenital Abnormalities

Among the twins, there were seven newborn babies with major congenital abnormalities. Three died neonatally (1 had trisomy 21 and 2 had diaphragmatic hernias). The four surviving congenitally abnormal neonates required major reconstructive surgeries: one had congenital renal uretero-pelvic stricture, one had Dwayne facial deformity syndrome, one had congenital hip dislocation, and one had a cleft palate. The rate of major anomalies was 7 of 220 (3.2%) including the undelivered patient. Among the triplets, one patient had an isolated macroglossia resulting in a minor abnormality rate of 1 of 39 (2.56%). Among the quads, one patient had an extra renal pelvis and a possible congenital lung cyst (with a normal karyotype compliment) and needed major corrective surgery. Another quad had minor clubfeet that did not necessitate surgical correction. The major abnormality rate here was 1 of 16 (6.25%). Neonatal Death Rate

A total of 11 twin newborn babies died in the neonatal period. Two sets of twins delivered after premature labor at 22 and 24.5 weeks; another set had PROM and delivered at 26 weeks. Five neonatal death occurred at >28 weeks. One had trisomy 21 and died at 7 hours. Two had IUGR starting 19 and 27 weeks and were delivered at 29 and 29.5 weeks and expired at 7 and 6 days of age (their karyotype 832

Seoud et al.

IVF twins, triplets, and quadruplets

6.3 ± 8.0 ± 8/8

1.6 0.6

P 2 versus 3

P 2 versus 4

P 3 versus 4

<0.001

<0.001

<0.05

<0.02 <0.001

<0.001 <0.001

<0.001 <0.02

<0.002 <0.001

<0.008 <0.001

<0.04 <0.001

NS:j: NS <0.01

<0.008 NS NS

<0.01 NS NS

NS

NS

NS

NS

NS

NS

NS

NS

NS

1/16 6.25) 1/16 (6.25) 0.0

:j: NS, not significant.

was 46XY and 46XX, respectively). Two patients with diaphragmatic hernias died at 30 minutes and 33 days, respectively. This gives an uncorrected neonatal death rate of 11 of 215 (51.1 per thousand). If this rate is corrected for pregnancies beyond 28 weeks and for neonatal death before 28 days, the rate becomes 4 of 208 (19.2 per thousand). One of the triplets died neonatally at 7 weeks of age secondary to IVH; this gives an uncorrected neonatal death rate of 1 of 45 (22.2 per thousand) and a corrected one of 0.0 per thousand. None of the quadruplets died neonatally. Perinatal Mortality Rate

If we include IUFD and neonatal death beyond 20 weeks as well as neonatal deaths beyond 28 days, the uncorrected perinatal mortality rate for the twins is 16 of220 (72.7 per thousand). However, if we only include losses beyond 28 weeks and neonatal deaths occurring before 28 days, the corrected perinatal mortality rate is 8 of 208 (38.5 per thousand). The uncorrected perinatal mortality rate for the triplets was 1 of 45 (22.2 per thousand), but the corrected perinatal mortality rate for both triplets and quadruplets was 0.0 per thousand. DISCUSSION

The overall incidence of multiple pregnancies resulting from assisted reproduction is much higher than in natural reproduction. Relative to other cenFertility and Sterility

T I

ters of assisted reproduction, the rates reported here were comparable with some institutions (11, 12), but lower than others (3). We find that despite being a special group (e.g., infertile and older), the complications of multiple IVF pregnancies are comparable with those seen in naturally conceived gestations. The mean age of our patients with twins was older than most other studies of naturally occurring twins (4-6, 8), reflecting the mean age of our infertility population. The mean age of patients with triplets and quads was comparable with most other studies (1,3, 7,8) and was younger than our twins. All three groups, however, belonged to a relatively older age bracket, which is known to be associated with a higher incidence of perinatal complications. The primary cause of infertility in our population was comparable with other studies of women requiring assisted reproduction (3, 11). The incidence of first trimester bleeding was high in all three groups, but the incidence of abortions was not proportionately elevated. This is important in trying to reassure a particularly anxious infertile patient. The incidence of obstetrical complications, in general, increased with the increase in the number of fetuses. The most common was premature labor. Among the twins, our study, although not randomized, showed that there was no prolongation of gestation in patients treated with bed rest compared with those treated without this intervention. Crowther et al. (13) and Mac Lennan et al. (14), in two well-controlled studies, showed that routine hospital admission for twins was not beneficial in prolonging the gestation and should be abandoned. However, Knuppel et al. (15) showed that home uterine monitoring and the help of a perinatal nursing support group significantly prolonged the gestation. Gardner et al. (16) clearly demonstrated that intensive prenatal care of twin gestation significantly promotes fetal growth and improves perinatal outcome. In a multicenter study of triplets, Newman et al. (7), however, showed that despite intensive close outpatient management, 66.2% had premature labor and 87.9% delivered before 37 weeks. In spite of the fact that eight of our triplets were electively admitted to the hospital, the gestational age at delivery was actually shorter than most reports (32.8 to 33.8 weeks) (1, 3, 6, 8, 17, 18). Unlike premature labor, in which only 38.5% of the twins and 50% of our triplets delivered the same day, most of the twins (12/13) and triplets (2/2) with PROM delivered the same day. This is important in counseling patients in whom this complication occurs. The frequency of PROM in twins, Vol. 57, No.4, April 1992

triplets, and quads in the literature varied from 2.7% to 43.7% (1,2,4,5,8, 19). The incidence of pregnancy-induced hypertension increased with the increase in the number of fetuses and was comparable with other reports (4-6, 8). Some have attributed this to the increase in the size of the placental tissue, but an exact pathophysiology is not known. Anemia was more common in our population, despite the fact that most of our patients fall into an upper-middle class population. It should be noted that the lack of statistically significant differences when comparing the incidence of premature labor, PROM, pregnancy-induced hypertension, gestational diabetes, hospital stay, and IUGR between the quadruplets and the twins and/or triplets may be because of the limited number of subjects in the quadruplet group. Neonatal complications also increased with the increase in the number of newborns in one set. The mean gestational age for the twins was similar to other series (4-8); however, for the triplets it was shorter than most studies (1, 3, 7, 8), including the extensive review by Elster et al. (17). This is probably why the birth weight of our triplets was less than most other reports (1,720 g to 1,911 g) (4-8). The quads were delivered at a similar gestational age compared with other studies (2, 19). In general, neonatal morbidity as reflected by the percentage of admission to the NICU, the need for respiratory assistance, the length of stay in the NICU and in the hospital significantly increased from twins to triplets to quads. This is also comparable with the literature in all three groups. The perinatal mortality was surprisingly lower, however, in all the groups than what has been previously reported. It is difficult, however, to obtain a clear definition of the cutoff gestational age used in some reports, especially the older ones (1-4,6,8, 18, 19). This is the reason we expressed the perinatal mortality in an uncorrected form to include all losses after 20 weeks as well as a corrected rate after 28 weeks. Moreover, the perinatal mortality was not significantly different among the three groups. This can be partially explained by the fact that our population was of a higher economic status than other reports whether for twins (4) or for triplets (18). In addition, our patients are invariably highly motivated and tend to strictly adhere to medical instructions. The sex ratio in the twins and quadruplets was comparable with that of naturally occurring ones, which negates a selection bias for one sex over the other in assisted reproduction. We cannot, however, explain the male/female ratio of 3.09 in the Seoud et al.

IVF twins, triplets, and quadruplets

833

triplets, believing that it is probably because of the relatively small size of this group. Little and Bryan (20) found the incidence of congenital anomalies in twins in over 30 studies to be approximately 0.3% to 4.0%, which is similar to that in our patients. The incidence of congenital anomalies in triplets and quads was also similar to other reports (1, 3, 19). In recent years, there has been an increasing trend toward selective reduction in multiple pregnancies of higher order including triplets and quads (21-23), the argument being that the perinatal mortality and morbidity are lower in the twins. Although some authors have achieved an acceptably low rate ofloss after such procedures, it should be noted that, as demonstrated by Kingsland et al. (3), Sassoon et al. (8), and our study the perinatal mortality oftriplets and perhaps quads is comparable with that of twins (3, 8). Some have attributed this to the fact that most physicians treat twins as any other low-risk pregnancies without referral to a high-risk center, whereas, most physicians refer higher order pregnancies to a tertiary care center early in the pregnancy. It is important, however, to explain to any prospective parents that pregnancies of higher order carry a high incidence of antenatal complications that require prolonged and expensive hospitalizations and that there is also a very high incidence of neonatal morbidity, which translates into prolonged and expensive hospitalization for the neonates. Follow-up in our study was not long enough to ascertain the long-term effects on the child development, but preliminary data are very encouraging. We hope to report on this issue later. REFERENCES 1. Olofsson P. Triplet and quadruplet pregnancies-a forthcoming challenge for the "general obstetrician". Eur J Obstet Gynecol Reprod BioI 1990;35:159-71. 2. Collins MS, Bleyl JA. Seventy-one quadruplet pregnancies. Management and outcome. Am J Obstet GynecoI1990;162: 1384-92. 3. Kingsland CR, Steer CV, Pampiglione JS, Mason BA, Edwards RG, Campbell S. Outcome of triplet pregnancies resulting from IVF at Bourn Hall 1984-1987. Eur J Obstet Gynecol Reprod BioI 1990;34:197-203. 4. Thompson SA, Lyons TL, Makowski EL. Outcome of twin gestation at the University of Colorado Health Sciences Center, 1973-1983. J Reprod Med 1987;32:328-39.

834

Seoud et al.

IVF twins, triplets, and quadruplets

5. Kovacs BW, Kirschbaum TH, Richard P. Twin gestations: 1. Antenatal care and complications. Obstet GynecoI1989;74: 313-7. 6. Spellacy WN, Handler H, Ferre CD. Case-control study of 1253 twin pregnancies from 1982-1987 perinatal data base. Obstet GynecoI1990;75:168--71. 7. Newman RB, Hamer C, Miller C. Outpatient triplet management: A contemporary review. Am J Obstet Gynecol 1989;161:547-55. 8. Sassoon DA, Castro LC, Davis JL, Hobel CJ. Perinatal outcome in triplet vs twin gestations. Obstet Gynecol 1990;75: 817-20. 9. Cunningham FG, Mac Donald P, Gant NF. Williams obstetrics. 18th ed. Norwalk (CT): Appleton and Lange 1989: 653-6. 10. Williams RL, Creasy RK, Cunningham GC, Hawes WE, Norris FD, Tashiro M. Fetal growth and perinatal viability in California. Obstet Gynecol 1982;59:624-9. 11. Bollen N, Camus M, Staessen C, Tournaye H, Devroey P, Van Steirteghem AC. The incidence of multiple pregnancy after in vitro fertilization and embryo transfer, gamete, or zygote intrafallopian transfer. Fertil Steril 1991;55:314-8. 12. IVF and GIFT pregnancies, Australia and New Zealand, 1987 National perinatal statistics unit. Sydney: Fertility Society of Australia, 1988:1030-4711. 13. Crowther CA, Nelson JP, Verkuyl DAA, Bannerman C, Asthust HM. Preterm labor in twins; can it be prevented by hospital admission? Br J Obstet GynaecoI1989;96:850-3. 14. Mac Lennan AH, Green RC, O'Shea R, Brookes C, Morris D. Routine hospital admission in twin pregnancy between 26-30 weeks gestation. Lancet 1990;335:267-9. 15. Knuppel RA, Lake MF, Watson DC, Welch RA, Hill WC, Fleming AD, et al. Preventing preterm birth in twin gestation: home uterine activity monitoring and perinatal nursing support. Obstet Gynecol 1990;76:24S-7S. 16. Gardner MO, Amoya MA, Sakakini J. Effects of prenatal care on twin gestations. J Reprod Med 1990;35:519-21. 17. Elster AD, Bleyl J, Crown TE. Birth weight standards for triplets under modern obstetrics care in the United States 1984-1989. Obstet Gynecol 1991;77:387-93. 18. Deale CJC, Cronje HS. A review of 367 triplet pregnancies. S Afr Med J 1984;66:92-4. 19. Lipitz S, Frenkel Y, Watts C, Ben-Rafael Z, Barkai G, Reichman B. High order multifetal gestation: management and outcome. Obstet Gynecol 1990;76:215-8. 20. Little J, Bryan E. Congenital anomalies in twins. Semin Perinatol 1986;10:50-64. 21. Hobbins JC. Selective reduction-a perinatal necessity? N Engl J Med 1988;318:1062-3. 22. Lynch L, Berkowitz RC, Chitkara U, Alvarez M. First trimester transabdominal multifetal reduction: a report of 85 cases. Obstet Gynecol 1990;75:735-8. 23. Tabsh KMA. Transabdominal multifetal pregnancy reduction: a report of 40 cases. Obstet Gynecol 1990;75:739-41.

Fertility and Sterility