- Email: [email protected]
Effects of fetal number and multifetal reduction on length of in vitro fertilization pregnancies
Effects of Fetal Number and Multifetal Reduction on Length of In Vitro Fertilization Pregnancies RAY V. HANING, Jr, MD, DAVID B. SEIFER, MD, CAROL A. WHEELER, MD, GARY N. FRISHMAN, MD, HELAYNE SILVER, MD, AND DEBORAH 1. PIERCE, BS Objective: To determine the effects of multifetal reduction and other variables on the duration of gestation of in vitro fertilization (IVF) pregnancies. Methods: All 274 IVF pregnancies from the inception of the Women and Infants’ Hospital IVF Program on May 26, 1988, until December 31, 1993, were evaluated. Results: Spontaneous reduction occurred in ten pregnancies, and multifetal reduction was elected in 28 multiple gestations. Among 260 pregnancies that remained viable beyond 20 weeks, 162 singletons (37.9 f 0.29 weeks; mean f standard error) had a longer mean gestation than did 64 twins (34.6 f 0.61 weeks), 25 pregnancies reduced to twins (33.4 f 1.0 weeks), or nine triplets (29.7 f 1.9 weeks). Triplets delivered 4.9 weeks earlier than nonreduced twins (P < .05) and 3.7 weeks before twins resulting from multifetal pregnancy reduction (P < .05). Regression analysis showed that at the S-week ultrasound, each viable fetus could be expected to reduce the duration of the gestation by about 3.6 weeks, and each fetus reduced medically or as a result of natural causes could be expected to prolong the gestation by approximately 3.0 weeks. Only 14% of triplet pregnancies underwent spontaneous multifetal reduction. Conclusion: Multifetal reduction of pregnancies with three or more fetuses was beneficial and increased the duration of gestation. (Obstet Gynecol 1996;87.964-8)
Multiple gestations are at increased risk for premature birth, and the risk resulting from in vitro fertilization (IVF) treatment appears to be greater than that of ageand fetal number-matched pregnancies without IVF treatment.’ In vitro fertilization results in increased serum relaxin concentrations in singleton pregnancies2,3 and multiple gestations4,’ and an increased risk of premature birth. c Multiple gestations after treatment with gonadotropins have a concentration of serum From the Departmed sity School of Medicine Rhode Islmd.
964
0029-78444/~6/915.00 PI1 50029.7844(96)000.59-2
of Obstetrics and Gynecology, Brozun Univerat Women rind b@zts Hospital, Providence,
relaxin above that of similarly treated singleton pregnancies,‘“,’ and it has been hypothesized that the increased serum relaxin concentration may increase the risk of premature birth by weakening the cervical tissues.3Z8 A similar phenomenon is seen in pregnancies resulting from IVF, in which the serum relaxin concentration is higher in multiple gestations than in singleton gestations’,“,‘; IVF twins have a 2.6-fold higher serum relaxin concentration than twins conceived normally.’ The development of techniques for fetal reduction has provided a means of attempting to reduce the risk of premature birth for women with multiple pregnancies.“-‘” However, because of the emotional nature of the decision, religious beliefs, and lack of accurate information about the effect of fetal reduction on the risk of premature birth in IVF patients, it is difficult to counsel patients with multiple gestations resulting from treatment with IVF. We have already reported the effects of IVF treatment on the serum relaxin concentration in a subset of these patients5 This study was conducted to provide important data on the risk factors for premature birth in IVF pregnancies, the duration of gestation in IVF pregnancies, the spontaneous reduction rate in multiple gestations resulting from IVF, and the beneficial effect of multifetal reduction on the duration of gestation in IVF pregnancies compared with appropriate controls.
Materials and Methods We evaluated all cycles resulting in pregnancy from the inception of the Women and Infants’ Hospital IVF/gamete intrafallopian transfer/zygote intrafallopian transfer program from May 26,1988 until December 31, 1993. Women and Infants’ Hospital is a nonprofit, nonsectarian, tertiary care hospital that performs deliveries for approximately 10,000 pregnant women
Obstetrics
G- Gynecology
each year. The hospital delivers the majority of all pregnant women in Rhode Island, most of our patient’s IVF pregnancies, and nearly all of our IVF multiple gestations. The population of this part of southern New England is predominantly white and of Roman Catholic religious background, although there are important contributions from all racial and religious minority groups. Because both Massachusetts and Rhode Island have mandated coverage of infertility treatment, many patients in the region, who otherwise could not afford it, have been able to access IVF treatment. The IVF patients were nearly all married and predominantly middle class or above in financial resources. The racial and religious composition of the patient population followed regional demographics, except as modified by economic factors. All decisions about whether to undergo selective reduction were made by the patients on an individual basis after considering their own feelings about the risks and benefits as well as the emotional, ethical, and religious issues involved. All serum samples were drawn for clinical management (as outlined later). The protocol for the study of pregnancy outcome was approved by the Research and Human Rights Committee of Women and Infants’ Hospital. There were 274 pregnancies with one or more fetal hearts detected by vaginal probe ultrasound 6 weeks after IVF embryo transfer, gamete intrafallopian transfer, or zygote intrafallopian transfer. Of these pregnancies, 268 (97.8%‘) resulted from IVF, five (1.8%) resulted from gamete intrafallopian transfer, and one (0.4%) resulted from zygote intrafallopian transfer. For all further analyses, the gamete intrafallopian transfer and zygote intrafallopian transfer pregnancies were combined with the IVF pregnancies because the treatment protocols were identical for the three groups, except for the transfer of oocytes or embryos. These 274 pregnancies (referred to hereafter as “IVF pregnancies”) initially consisted of 166 singleton pregnancies, 66 twin pregnancies, 36 triplet pregnancies, five quadruplet pregnancies, and one quintuplet pregnancy at 6 weeks after IVF embryo transfer. Leuprolide acetate (Lupron; Tap Pharmaceuticals Inc., North Chicago, IL) suppression was given to all patients, and it was usually started in the mid- to late luteal phase at a dose of 0.5 mg subcutaneously each day. It was continued for 2-4 weeks before treatment with gonadotropins and then continued during administration of FSH and LH until administration of hCG. All patients were treated with FSH and LH daily in two divided doses administered 12 hours apart, and all received 10,000 IU of hCG (on hCG day 0) before egg retrieval, as described previously.“,‘” Estradiol (E2) and progesterone were measured on day 6 or 7 and again on
VOL.
87, NO.
6, JUNE
1996
days 11-13 after hCG; hCG assays were begun 2 weeks after embryo transfer. The number of sacs and the number of sacs with a heartbeat were determined using an ultrasound vaginal probe 4 and 6 weeks, respectively, after embryo transfer (corresponding to 6 and 8 weeks after the theoretical last menstrual period). All patients were treated with intramuscular progesterone alone for luteal phase support. An initial dose of 50 mg of progesterone in sesame seed oil was administered intramuscularly on the day of embryo transfer (hCG day 4). Subsequently, all patients received 25 mg of progesterone intramuscularly each day unless the serum progesterone/E2 molar ratio was less than 100 on hCG day 6-7 or hCG day 11-13, in which case the progesterone dose was increased to 50 mg. Once increased, the dose was not altered until therapy ceased. In pregnant patients, progesterone was discontinued at hCG day 60 as long as endogenous production maintained a serum progesterone concentration greater than 20 ng/mL. If a progesterone concentration of 20 ng/mL could not be maintained, supplemental progesterone was continued until the endogenous production could maintain it. Gestational age was defined as the day of conception plus 2 weeks. We considered conception to have occurred on the day of embryo transfer (approximately 48 hours after egg retrieval). Twenty-two selective reduction procedures were performed transvaginally at Columbia Presbyterian Medical Center, New York, New York,” and the remainder were performed transabdominally at Mt. Sinai Hospital, New York, New York” (four cases) or the Peter Bent Brigham Hospital, Boston, Massachusetts (two cases). The serum samples were stored at -20C until assay. Estradiol was measured by radioimmunoassay using the Pantex direct I-125-estradiol kit (Pantex, Santa Monica, CA). Between- and within-assay variation was 7.3 and 4.2%, respectively. Progesterone was measured with the Diagnostic Products Corporation (Los Angeles, CA) Coat-A-Count progesterone kit; betweenand within-assay variation was 7.2 and 5.8%, respectively. The data for E2, progesterone, the number of follicles, and the number of oocytes were log-transformed before statistical analysis to correct for the log-normal distribution of the data. There is only one value per patient for any given determination within a single interval of time. Differences between treatments in the analysis of variance were tested using the t test for the planned comparison (between triplets and pregnancies reduced to twins) and by Scheffe method to provide a conservative test of the remaining comparisons between the treatment means.2’ Chi-square was used to test differences for attribute data. Regression analysis and stepwise multiple linear regression (with entry and elimination of variables both set at P < .05) were used to
Haning
et al
1VF Pwpancy
Duration
965
search for variables predictive of the duration of gestation. All statistical calculations were performed using Minitab release 9.0 (Minitab Inc., State College, PA).
Results Both spontaneous and induced fetal reductions occurred. Five twins and one triplet pregnancy reduced spontaneously to singletons, and four triplet pregnancies spontaneously reduced to twins. Thus, five of 36 triplet pregnancies (14%) and five of 66 twin pregnancies (8%) underwent spontaneous reduction. Induced reduction produced 28 twin pregnancies from 22 triplet pregnancies, five quadruplet pregnancies, and one quintuplet pregnancy. Ten of 166 singleton pregnancies (6.0%), one of 66 twin pregnancies, and none of the 14 triplet pregnancies that did not undergo multifetal pregnancy reduction were lost before 20 weeks. Three (10.7%) of the 28 pregnancies with triplets or more that underwent multifetal pregnancy reduction to twins were lost before 20 weeks. Twenty-five pregnancies that underwent multifetal pregnancy reduction continued to 20 weeks or beyond. There was a significantly higher early loss of twins resulting from multifetal pregnancy reduction than of nonreduced twins (P < .Ol, 2). Correction for the alterations in the number of viable fetuses by the time of birth and removal of the 14 pregnancies not progressing to 20 weeks yielded 260 pregnancies, distributed as follows: 162 (62%) singleton pregnancies, 64 (25%,) twin pregnancies not resulting from multifetal pregnancy reduction, nine triplet pregnancies (3%), and another 25 twin pregnancies (10%) that resulted from multifetal reduction of 21 triplet pregnancies and four quadruplet pregnancies. Five of the singleton pregnancies resulted from spontaneous reduction of twins, and one resulted from spontaneous reduction of triplets. Four of the twin pregnancies resulted from spontaneous reduction of triplets. The singleton pregnancies (gestational length 37.9 ? 0.29 weeks; hJ = 162) had a longer gestation than did twins (34.6 + 0.61 weeks; N = 64), pregnancies reduced to twins (33.4 ? 1.0 weeks; N = 25), or triplets (29.7 t 1.9 weeks; N = 9), all P < .Ol by Sheffk test. Triplets delivered 4.9 weeks earlier than twins not resulting from multifetal pregnancy reduction (P < .05, Sheffe test), and twins resulting from multifetal pregnancy reduction delivered 3.7 weeks later than triplets (P < .05, t test as the planned comparison). Regression analysis of the data was performed to determine variables predictive of the duration of gestation, the direction of the slope, and their relative importance. For the following calculations, we again used only the pregnancies that remained viable to 20 weeks
966
Haning
et al
IVF
Prepuncy
Dumtion
or more. Regression analysis (with N = 260 in all cases) showed that the weeks of gestation was predicted by the number of live fetuses present in the uterus at delivery (R = -0.445, P < .Ol), the number of viable pregnancies at S-week ultrasound (R = -0.411, P < .Ol), and the serum progesterone concentration on day 6-7 after hCG (R = -0.144, P < .05). There was no significant regression of weeks of gestation on the serum E2 concentration on days 11-13 after hCG (R = -0.103), the serum progesterone concentration on days 11-13 after hCG (R = -O.lOO), the number of follicles on the day of hCG (R = -0.086), the number of oocytes retrieved (R = 0.041), the serum E2 concentration on day 6-7 after hCG (R = -0.023), or the serum E2 concentration on the day of hCG (R = -0.006). Stepwise multiple linear regression analysis using the aforementioned variables showed that after the addition of one of the measures of fetal number to the regression equation, none of the other variables listed were statistically significant predictors of gestational duration. Regression analysis gave the equation: WEEKS = 41.7 - 3.76 (LIVE), with R2 = 0.198 and P < ,001 for the test of significance for regression coefficient (3.76 versus 0), N = 260 cases; where WEEKS is the weeks of gestation completed at delivery and LIVE is the number of infants alive at birth. This equation has the disadvantage that the number of infants alive at birth can be determined only after delivery has occurred. Multiple linear regression analysis was used to obtain an expression to predict the duration of gestation from the number of viable fetuses present at the S-week ultrasound (USSWKS) and the number of fetuses that became nonviable either as a result of the reduction procedure or through natural causes (ALLREDUCED). We obtained the expression: WEEKS = 41.6 - 3.63(USSWKS) + 3,01(ALLREDUCED), with R’ = 0.201, N = 260 cases. The regression coefficient for USSWKS was significant at P < .OOl (3.63 versus 0), and that for ALLREDUCED was significant at P = .OOl (3.01 versus 0). This equation has the advantage that it is based on information available to the physician and patient at the time when decisions must be made about multifetal pregnancy reduction. The values for USSWKS ranged from 1 to 5 and the values for ALLREDUCED in the data set ranged from 0 to 3, with a total of 30 reductions by one fetus and a total of five reductions by two fetuses.
Discussion The increased use of IVF procedures has resulted in the accumulation of enough experience to identify potential problems with IVF pregnancies’ and to allow statistical evaluation of the results obtained with treatment interventions. Because multiple embryos are transferred to
Obstetrics
6 Gynecology
improve the chance of conception, IVF has resulted in an increase in the incidence of multiple gestation. Multiple gestation increases the risk of the ovarian hyperstimulation syndrome in IVF pregnancies.‘,“’ There is also an increase above normal in the serum relaxin concentration in IVF pregnancies,2,“,“Zx which is even greater in multiple gestations than in singleton pregnancies.” Gestations resulting from IVF are at greater risk for premature birth than gestations not resulting from IVF, after correction for fetal number. ‘*‘Jo It may be that this increased risk of premature delivery is mediated through increased serum concentrations of relaxin in IVF pregnancies, as relaxin has been shown to be 2.6-fold higher in IVF twins than in twins conceived naturally.” It is also possible that the increased concentrations of steroid hormones (secreted in supraphysiologic amounts by the gonadotropinstimulated ovaries)“,” augment the effects of the supraphysiologic concentrations of relaxin, thus producing adverse effects on the strength of the cervical tissues; steroid hormone augmentation of the effects of relaxin on connective tissue has been shown in animal models.24 The present study was designed to test the effects of various indices on the duration of gestation in IVF pregnancies. Regression analysis of our data indicated that each additional fetus presentat the time of delivery in our IVF pregnancies reduced the duration of gestation by about 3.8 weeks. Although there was little evidence that any of the steroid hormone concentrations measured affected the duration of pregnancy in this study, the study population consisted only of IVF pregnancies, and determinations of the steroid hormone concentrations were limited to the first weeks after conception. In addition, the supraphysiologic steroid hormone concentrations in IVF pregnancies are correlated with fetal number,2” making it impossible for us to exclude the possibility that elevated relaxin or steroid hormone concentrations may mediate some of the risk of premature birth in IVF pregnancies in comparison with naturally conceived controls, after correction for fetal number. l,h,22 The present study is unique in that we followed up each viable intrauterine pregnancy initiated. Our results provide evidence that multifetal reduction can reduce the risk of premature birth in IVF pregnancies, and also provide quantitative estimates of the effects of fetal reduction on the duration of gestation. Only 14% of our IVF triplet pregnancies underwent spontaneous multifetal reduction. At the time when the patient must decide about whether to undergo an attempt at multifetal reduction, only the number of viable fetuses is known. The multiple linear regression equation changed slightly with each viable fetus present at the g-week ultrasound, reducing the duration of the gestation by about 3.6 weeks, and each fetus reduced medi-
VOL.
87, NO.
6, JUKE
1946
cally or as a result of natural causes prolonged the gestation by approximately 3.0 weeks. The patient and physician attempting to deal with a multiple gestation face several difficult problems. First, the pregnancy was obtained only after a long period of difficult and expensive treatment, and the patient and physician both fear losing the pregnancy as a result of multifetal reduction. Second, the risks associated with premature birth of a multiple gestation include more than just the risk to the mother and the potential of a fatal outcome for the infants. Premature infants are at risk for long-term sequelaeand potential life-long disability.” However, if a pregnancy is lost from multifetal reduction, the risks to the mother are lower from an early loss than from a complicated delivery of a multiple gestation.‘0,‘5 In addition, although successis never guaranteed, the chance that the patient can conceive another IVF pregnancy appears fairly good, given that IVF treatment resulted in a multiple gestation. Others25,2hhave reported that multifetal reduction to twins of multiple gestations resulting from treatment with gonadotropins substantially increased the duration of gestation, reduced the incidence of prematurity, reduced neonatal mortality, and shortened the neonatal intensive care unit stay. To our knowledge, however, ours is the first study of a population exclusively made up of IVF patients. In our experience at Women and Infants’ Hospital, multiple gestations resulting from IVF are at a significantly increased risk for low birth weight and small for gestational age infants, and for morbidity associated with prematurity, in comparison with non-IVF controls.’ Although the small numbers available make it impossible to determine statistical significance accurately, the mortality of IVF infants at Women and Infants’ Hospital was 2.7-fold higher in triplets than in twins (7.41 versus 2.78%)-i However, this figure represents only the tip of the iceberg for the increased risks faced by the IVF triplets in comparison with IVF twins. Compared with IVF twin survivors, IVF triplet survivors at Women and Infants’ Hospital experienced a 4.8-fold greater risk of culture-positive neonatal sepsis(29 versus 6%; P < .05), a 3.9-fold greater risk of respiratory distresssyndrome treated with surfactant (71 versus 18%‘;P < .005), a 3.8-fold increase in the risk of patent ductus arteriosis treated with indomethacin (57 versus 15%; P < .Ol), a 77-fold increase in days of oxygen administration (54 + 47 versus 0.7 2 2; P < .02), and a 67-fold increase in days of mechanical ventilation (40 2 54 versus 0.6 2 2; P < .025).’ Multifetal reduction of pregnancies with three or more fetuses to twins added, on average, 3.7 weeks to the observed average of 29.7 weeks in unreduced triplet pregnancies. Although we recognize that there is variability in the duration of gestation for individual cases,the average of 3.7 additional weeks in utero did much to reduce the risk of death or long-term morbidity for the infants in the cases
Haning et al
IVF
Prqpancy
Duration
967
treated with multifetal reduction. Each of our patients who lost an IVF multiple gestation after multifetal reduction was able to establish a subsequent pregnancy using IVF, with a good outcome. Counseling the patient under such circumstances is challenging because of the interaction of emotions and religious beliefs, but it was made all the more difficult by the lack of sound data on the effects of either multiple gestation or multifetal reduction on the outcome of the IVF pregnancy. We hope that the data given here will help by providing statistically based information on the effects of both fetal number and multifetal reduction on the duration of gestation. The Women and Infants’ Hospital experience with multifetal reduction in IVF patients indicates that considerable benefits are achieved by multifetal reduction of pregnancies with three or more fetuses. Because of the emotional, religious, and ethical dilemmas, only the patients themselves can make the final decision. The patient must decide which side of the risk-benefit scale to discount and which side to emphasize.
12. Berkowitz Alvarez trimester.
1. Tallo
Cl’,
Vohr
B, Oh
W,
Rubin
L, Seifer
morbidity associated 1995;127:794-800.
DB,
Haning
RV
with
in vitro
fertili-
2. Bell RJ, Eddie LW, Lester AR, Wood EC, Johnston Relaxin in human pregnancy serum measured with
transabdominal cases. Obstet
radioimmunoassay. 3. McClure N, Leya phase Hum 4. Lynch
Obstet Gynecol 1987;69:58>9. J, Radwanska E, Rawlins R, Haning
R. The effect of multifetal pregnancy Obstet Gynecol 1995;85:756-9. 5. Haning Sarmento
RV Jr, Goldsmith J, et al. Relaxin
in-vitro fertilization/gamete 1994;9:141-6.
Philipson of singleton
G, F’etrucco pregnancies
intra-fallopian
transfer.
relaxin.
Frishman G, Am J Obstet 0, Anderson following Hum
T, Dale PO, in singleton
Lunde 0, pregnancies
Obstet 8. Haning multiple
Gynecol 1995;86:188-92. RV Jr, Steinetz B, Weiss pregnancy following
Moe
N, after
Abyholm assisted
G. Elevated menotropin
JA, Goldsmith
NJ, Weiss G. The effect of ovulation of maternal serum relaxin in
twin
serum relaxin levels in treatment. Obstet Gy-
LT, Shahinian
KA,
Erinakes
induction on the concentration pregnancies. Am J Obstet
Gynecol 1996;174:227-32. 10. Timor-Tritsch IE, Peisner DB, Monteagudo A, Lerner JP, Sharma S. Multifetal pregnancy reduction by transvaginal puncture: Evaluation of the technique 1993;168:799%804.
used
in
134
cases.
Am
11. Brandes JM, Itskovits J, Timor-Tritch Reduction of the number of embryos Steril 1987;48:326-7.
IE, Drugan in multiple
968 Haning
Duration
et al
IVF
Pregnancy
J Obstet
Gynecol
A, Frydman R. pregnancy. Fertil
U,
ultrasonic
Blankier
J, Casper
1988;
First-trimester A report
of 85
Lynch L, Dumez Y, multifetal pregnancy the world’s largest A.
Puncture
proce-
Ultrasound
Ob-
E. Selective reduction in after transvaginal and Fertil Steril 1989;52:
Transvaginal
multiple
ultrasound
pregnancy.
perinatal
in
Obstet
necessity.
Gy-
N Engl
J
1988;318:1062-3.
Dahl Lyons CA, Wheeler CA, Frishman GN, Hackett RJ, Seifer DB, Haning RV Jr. Early and late presentation of the ovarian hyperstimulation syndrome (OH%): Two distinct entities with different
risk factors. Hum Reprod 21. Snedecor GW, Cochram Iowa: 22. Doyle weight 23.
M.
guidance.
RG.
Y,
selective embryo reduction for necol 1990;75:720-2. 19. Hobbins JC. Selective reduction-A
Gynecol
reduction:
M, Wapner RJ, of transabdominal experience among
transvaginal 1991;1:144-50.
MH, Struyk and quadru-
Obstet
Alvarez
416-20. 18. Gonen
resulting Haning
1994;9:792-9. WG. Statistical
The Iowa State University P, Beral V, Maconochie and small-for-gestational-age from in-vitro RV Jr, Kiggens
one, 17P-estradiol, follow treatment
and estriol with human and
methods.
6th
ed.
Ames,
Press, 1967:268-71. N. Preterm delivery, low birthin liveborn singleton babies
fertilization. AJ, Leifheit
of multiple gestation Biochem 1985;22:823-9. 24. Sherwood OD. Relaxin.
Hum Reprod TL. Maternal
are increased menopausal maternal
In: Knobil
in pregnancies gonadotropins:
endocrine E, Neil1
1992;7:425-8. serum progesterwhich Effects
status. JD,
J Steroid
Greenwald
GS,
Markert C, Pfaff DW, eds. The physiology of reproduction. Vol. 1. New York: Raven Press, 1994:861-1009. 25. Melgar outcome
Reprod
T. Obstetric reproduction.
Chitkara
issues.
M, Shale” outcome procedures.
26.
7. Tanbo outcome
ethical
17. Shalev J, Frenkel Y, Goldenberg multiple gestations: Pregnancy transabdominal needle-guided
R, Wein
on serum
LT, Seifer DB, Wheeler CA, secretion in IVF pregnancies.
Gynecol 1996;174:233-40. 6. Wang JX, Clark AM, Kirby CA, G, et al. The obstetric outcome
necol 1985;66:42-5. 9. Haning RV Jr, Canick
reduction
syndrome.
and
BW, Quigg in octuplet
Gynecol 1993;82:61-6. IE, Peisner DB, Monteagudo
dures utilizing stet Gynecol
RV Jr. Luteal
support and severe ovarian hyperstimulation Reprod 1992;7:758-64. L, Berkowitz RL, Weiss G, Goldsmith LT, Lapinski
IE, Newton termination
multifetal pregnancy Gynecol 1990;75:735-8.
centers. Obstet 16. Timor-Tritsch
Mehalek KE, in the first
1988;318:1042-7.
15. Evans MI, Dommergues Goldberg JD, et al. Efficacy reduction: Collaborative
Jr.
I’D, Nial HD. a homologous
J Med
plet pregnancies: Clinical 711289-96. 14. Lynch L, Berkowitz RL,
20.
Maternal and neonatal zation (IVF). J Pediatr
N Engl
13. Evans MI, Fletcher JC, Zador CD. Selective first-trimester
Med
References
RL, Lynch L, Chitkara U, Witkins IA, E. Selective reduction of multifetal pregnancies
duced Lipitz
CA, Rosenfeld after multifetal
DL, Rawlinson K, Greenberg reduction to twins compared
multiple gestations. S, Reichman B, Uval
2nd
ed,
M. Perinatal with nonre-
Obstet Gynecol 1991;78:763-7. J, Shale” J, Achiron R, Barkai
G, et al.
A prospective comparison of the outcome of triplet pregnancies managed expectantly or by multifetal reduction to twins. Am Obstet Gynecol 1994;170:874-9.
Reprints are not available. Address correspondence Ray V. Han@, Jr, MD 5793 Crowder #B352 New Orleans,
Boulevard LA
70127
Received August 23, 1995. Received in revised form December Accepted lanuary 22, 1996. Copyright Gynecologists.
0
to:
1996 by Published
The American by Elsevier
28, 1995.
College of Obstetricians Science Inc.
Obstetrics
and
& Gynecology
J
Multiple gestations are at increased risk for premature birth, and the risk resulting from in vitro fertilization (IVF) treatment appears to be greater than that of ageand fetal number-matched pregnancies without IVF treatment.’ In vitro fertilization results in increased serum relaxin concentrations in singleton pregnancies2,3 and multiple gestations4,’ and an increased risk of premature birth. c Multiple gestations after treatment with gonadotropins have a concentration of serum From the Departmed sity School of Medicine Rhode Islmd.
964
0029-78444/~6/915.00 PI1 50029.7844(96)000.59-2
of Obstetrics and Gynecology, Brozun Univerat Women rind b@zts Hospital, Providence,
relaxin above that of similarly treated singleton pregnancies,‘“,’ and it has been hypothesized that the increased serum relaxin concentration may increase the risk of premature birth by weakening the cervical tissues.3Z8 A similar phenomenon is seen in pregnancies resulting from IVF, in which the serum relaxin concentration is higher in multiple gestations than in singleton gestations’,“,‘; IVF twins have a 2.6-fold higher serum relaxin concentration than twins conceived normally.’ The development of techniques for fetal reduction has provided a means of attempting to reduce the risk of premature birth for women with multiple pregnancies.“-‘” However, because of the emotional nature of the decision, religious beliefs, and lack of accurate information about the effect of fetal reduction on the risk of premature birth in IVF patients, it is difficult to counsel patients with multiple gestations resulting from treatment with IVF. We have already reported the effects of IVF treatment on the serum relaxin concentration in a subset of these patients5 This study was conducted to provide important data on the risk factors for premature birth in IVF pregnancies, the duration of gestation in IVF pregnancies, the spontaneous reduction rate in multiple gestations resulting from IVF, and the beneficial effect of multifetal reduction on the duration of gestation in IVF pregnancies compared with appropriate controls.
Materials and Methods We evaluated all cycles resulting in pregnancy from the inception of the Women and Infants’ Hospital IVF/gamete intrafallopian transfer/zygote intrafallopian transfer program from May 26,1988 until December 31, 1993. Women and Infants’ Hospital is a nonprofit, nonsectarian, tertiary care hospital that performs deliveries for approximately 10,000 pregnant women
Obstetrics
G- Gynecology
each year. The hospital delivers the majority of all pregnant women in Rhode Island, most of our patient’s IVF pregnancies, and nearly all of our IVF multiple gestations. The population of this part of southern New England is predominantly white and of Roman Catholic religious background, although there are important contributions from all racial and religious minority groups. Because both Massachusetts and Rhode Island have mandated coverage of infertility treatment, many patients in the region, who otherwise could not afford it, have been able to access IVF treatment. The IVF patients were nearly all married and predominantly middle class or above in financial resources. The racial and religious composition of the patient population followed regional demographics, except as modified by economic factors. All decisions about whether to undergo selective reduction were made by the patients on an individual basis after considering their own feelings about the risks and benefits as well as the emotional, ethical, and religious issues involved. All serum samples were drawn for clinical management (as outlined later). The protocol for the study of pregnancy outcome was approved by the Research and Human Rights Committee of Women and Infants’ Hospital. There were 274 pregnancies with one or more fetal hearts detected by vaginal probe ultrasound 6 weeks after IVF embryo transfer, gamete intrafallopian transfer, or zygote intrafallopian transfer. Of these pregnancies, 268 (97.8%‘) resulted from IVF, five (1.8%) resulted from gamete intrafallopian transfer, and one (0.4%) resulted from zygote intrafallopian transfer. For all further analyses, the gamete intrafallopian transfer and zygote intrafallopian transfer pregnancies were combined with the IVF pregnancies because the treatment protocols were identical for the three groups, except for the transfer of oocytes or embryos. These 274 pregnancies (referred to hereafter as “IVF pregnancies”) initially consisted of 166 singleton pregnancies, 66 twin pregnancies, 36 triplet pregnancies, five quadruplet pregnancies, and one quintuplet pregnancy at 6 weeks after IVF embryo transfer. Leuprolide acetate (Lupron; Tap Pharmaceuticals Inc., North Chicago, IL) suppression was given to all patients, and it was usually started in the mid- to late luteal phase at a dose of 0.5 mg subcutaneously each day. It was continued for 2-4 weeks before treatment with gonadotropins and then continued during administration of FSH and LH until administration of hCG. All patients were treated with FSH and LH daily in two divided doses administered 12 hours apart, and all received 10,000 IU of hCG (on hCG day 0) before egg retrieval, as described previously.“,‘” Estradiol (E2) and progesterone were measured on day 6 or 7 and again on
VOL.
87, NO.
6, JUNE
1996
days 11-13 after hCG; hCG assays were begun 2 weeks after embryo transfer. The number of sacs and the number of sacs with a heartbeat were determined using an ultrasound vaginal probe 4 and 6 weeks, respectively, after embryo transfer (corresponding to 6 and 8 weeks after the theoretical last menstrual period). All patients were treated with intramuscular progesterone alone for luteal phase support. An initial dose of 50 mg of progesterone in sesame seed oil was administered intramuscularly on the day of embryo transfer (hCG day 4). Subsequently, all patients received 25 mg of progesterone intramuscularly each day unless the serum progesterone/E2 molar ratio was less than 100 on hCG day 6-7 or hCG day 11-13, in which case the progesterone dose was increased to 50 mg. Once increased, the dose was not altered until therapy ceased. In pregnant patients, progesterone was discontinued at hCG day 60 as long as endogenous production maintained a serum progesterone concentration greater than 20 ng/mL. If a progesterone concentration of 20 ng/mL could not be maintained, supplemental progesterone was continued until the endogenous production could maintain it. Gestational age was defined as the day of conception plus 2 weeks. We considered conception to have occurred on the day of embryo transfer (approximately 48 hours after egg retrieval). Twenty-two selective reduction procedures were performed transvaginally at Columbia Presbyterian Medical Center, New York, New York,” and the remainder were performed transabdominally at Mt. Sinai Hospital, New York, New York” (four cases) or the Peter Bent Brigham Hospital, Boston, Massachusetts (two cases). The serum samples were stored at -20C until assay. Estradiol was measured by radioimmunoassay using the Pantex direct I-125-estradiol kit (Pantex, Santa Monica, CA). Between- and within-assay variation was 7.3 and 4.2%, respectively. Progesterone was measured with the Diagnostic Products Corporation (Los Angeles, CA) Coat-A-Count progesterone kit; betweenand within-assay variation was 7.2 and 5.8%, respectively. The data for E2, progesterone, the number of follicles, and the number of oocytes were log-transformed before statistical analysis to correct for the log-normal distribution of the data. There is only one value per patient for any given determination within a single interval of time. Differences between treatments in the analysis of variance were tested using the t test for the planned comparison (between triplets and pregnancies reduced to twins) and by Scheffe method to provide a conservative test of the remaining comparisons between the treatment means.2’ Chi-square was used to test differences for attribute data. Regression analysis and stepwise multiple linear regression (with entry and elimination of variables both set at P < .05) were used to
Haning
et al
1VF Pwpancy
Duration
965
search for variables predictive of the duration of gestation. All statistical calculations were performed using Minitab release 9.0 (Minitab Inc., State College, PA).
Results Both spontaneous and induced fetal reductions occurred. Five twins and one triplet pregnancy reduced spontaneously to singletons, and four triplet pregnancies spontaneously reduced to twins. Thus, five of 36 triplet pregnancies (14%) and five of 66 twin pregnancies (8%) underwent spontaneous reduction. Induced reduction produced 28 twin pregnancies from 22 triplet pregnancies, five quadruplet pregnancies, and one quintuplet pregnancy. Ten of 166 singleton pregnancies (6.0%), one of 66 twin pregnancies, and none of the 14 triplet pregnancies that did not undergo multifetal pregnancy reduction were lost before 20 weeks. Three (10.7%) of the 28 pregnancies with triplets or more that underwent multifetal pregnancy reduction to twins were lost before 20 weeks. Twenty-five pregnancies that underwent multifetal pregnancy reduction continued to 20 weeks or beyond. There was a significantly higher early loss of twins resulting from multifetal pregnancy reduction than of nonreduced twins (P < .Ol, 2). Correction for the alterations in the number of viable fetuses by the time of birth and removal of the 14 pregnancies not progressing to 20 weeks yielded 260 pregnancies, distributed as follows: 162 (62%) singleton pregnancies, 64 (25%,) twin pregnancies not resulting from multifetal pregnancy reduction, nine triplet pregnancies (3%), and another 25 twin pregnancies (10%) that resulted from multifetal reduction of 21 triplet pregnancies and four quadruplet pregnancies. Five of the singleton pregnancies resulted from spontaneous reduction of twins, and one resulted from spontaneous reduction of triplets. Four of the twin pregnancies resulted from spontaneous reduction of triplets. The singleton pregnancies (gestational length 37.9 ? 0.29 weeks; hJ = 162) had a longer gestation than did twins (34.6 + 0.61 weeks; N = 64), pregnancies reduced to twins (33.4 ? 1.0 weeks; N = 25), or triplets (29.7 t 1.9 weeks; N = 9), all P < .Ol by Sheffk test. Triplets delivered 4.9 weeks earlier than twins not resulting from multifetal pregnancy reduction (P < .05, Sheffe test), and twins resulting from multifetal pregnancy reduction delivered 3.7 weeks later than triplets (P < .05, t test as the planned comparison). Regression analysis of the data was performed to determine variables predictive of the duration of gestation, the direction of the slope, and their relative importance. For the following calculations, we again used only the pregnancies that remained viable to 20 weeks
966
Haning
et al
IVF
Prepuncy
Dumtion
or more. Regression analysis (with N = 260 in all cases) showed that the weeks of gestation was predicted by the number of live fetuses present in the uterus at delivery (R = -0.445, P < .Ol), the number of viable pregnancies at S-week ultrasound (R = -0.411, P < .Ol), and the serum progesterone concentration on day 6-7 after hCG (R = -0.144, P < .05). There was no significant regression of weeks of gestation on the serum E2 concentration on days 11-13 after hCG (R = -0.103), the serum progesterone concentration on days 11-13 after hCG (R = -O.lOO), the number of follicles on the day of hCG (R = -0.086), the number of oocytes retrieved (R = 0.041), the serum E2 concentration on day 6-7 after hCG (R = -0.023), or the serum E2 concentration on the day of hCG (R = -0.006). Stepwise multiple linear regression analysis using the aforementioned variables showed that after the addition of one of the measures of fetal number to the regression equation, none of the other variables listed were statistically significant predictors of gestational duration. Regression analysis gave the equation: WEEKS = 41.7 - 3.76 (LIVE), with R2 = 0.198 and P < ,001 for the test of significance for regression coefficient (3.76 versus 0), N = 260 cases; where WEEKS is the weeks of gestation completed at delivery and LIVE is the number of infants alive at birth. This equation has the disadvantage that the number of infants alive at birth can be determined only after delivery has occurred. Multiple linear regression analysis was used to obtain an expression to predict the duration of gestation from the number of viable fetuses present at the S-week ultrasound (USSWKS) and the number of fetuses that became nonviable either as a result of the reduction procedure or through natural causes (ALLREDUCED). We obtained the expression: WEEKS = 41.6 - 3.63(USSWKS) + 3,01(ALLREDUCED), with R’ = 0.201, N = 260 cases. The regression coefficient for USSWKS was significant at P < .OOl (3.63 versus 0), and that for ALLREDUCED was significant at P = .OOl (3.01 versus 0). This equation has the advantage that it is based on information available to the physician and patient at the time when decisions must be made about multifetal pregnancy reduction. The values for USSWKS ranged from 1 to 5 and the values for ALLREDUCED in the data set ranged from 0 to 3, with a total of 30 reductions by one fetus and a total of five reductions by two fetuses.
Discussion The increased use of IVF procedures has resulted in the accumulation of enough experience to identify potential problems with IVF pregnancies’ and to allow statistical evaluation of the results obtained with treatment interventions. Because multiple embryos are transferred to
Obstetrics
6 Gynecology
improve the chance of conception, IVF has resulted in an increase in the incidence of multiple gestation. Multiple gestation increases the risk of the ovarian hyperstimulation syndrome in IVF pregnancies.‘,“’ There is also an increase above normal in the serum relaxin concentration in IVF pregnancies,2,“,“Zx which is even greater in multiple gestations than in singleton pregnancies.” Gestations resulting from IVF are at greater risk for premature birth than gestations not resulting from IVF, after correction for fetal number. ‘*‘Jo It may be that this increased risk of premature delivery is mediated through increased serum concentrations of relaxin in IVF pregnancies, as relaxin has been shown to be 2.6-fold higher in IVF twins than in twins conceived naturally.” It is also possible that the increased concentrations of steroid hormones (secreted in supraphysiologic amounts by the gonadotropinstimulated ovaries)“,” augment the effects of the supraphysiologic concentrations of relaxin, thus producing adverse effects on the strength of the cervical tissues; steroid hormone augmentation of the effects of relaxin on connective tissue has been shown in animal models.24 The present study was designed to test the effects of various indices on the duration of gestation in IVF pregnancies. Regression analysis of our data indicated that each additional fetus presentat the time of delivery in our IVF pregnancies reduced the duration of gestation by about 3.8 weeks. Although there was little evidence that any of the steroid hormone concentrations measured affected the duration of pregnancy in this study, the study population consisted only of IVF pregnancies, and determinations of the steroid hormone concentrations were limited to the first weeks after conception. In addition, the supraphysiologic steroid hormone concentrations in IVF pregnancies are correlated with fetal number,2” making it impossible for us to exclude the possibility that elevated relaxin or steroid hormone concentrations may mediate some of the risk of premature birth in IVF pregnancies in comparison with naturally conceived controls, after correction for fetal number. l,h,22 The present study is unique in that we followed up each viable intrauterine pregnancy initiated. Our results provide evidence that multifetal reduction can reduce the risk of premature birth in IVF pregnancies, and also provide quantitative estimates of the effects of fetal reduction on the duration of gestation. Only 14% of our IVF triplet pregnancies underwent spontaneous multifetal reduction. At the time when the patient must decide about whether to undergo an attempt at multifetal reduction, only the number of viable fetuses is known. The multiple linear regression equation changed slightly with each viable fetus present at the g-week ultrasound, reducing the duration of the gestation by about 3.6 weeks, and each fetus reduced medi-
VOL.
87, NO.
6, JUKE
1946
cally or as a result of natural causes prolonged the gestation by approximately 3.0 weeks. The patient and physician attempting to deal with a multiple gestation face several difficult problems. First, the pregnancy was obtained only after a long period of difficult and expensive treatment, and the patient and physician both fear losing the pregnancy as a result of multifetal reduction. Second, the risks associated with premature birth of a multiple gestation include more than just the risk to the mother and the potential of a fatal outcome for the infants. Premature infants are at risk for long-term sequelaeand potential life-long disability.” However, if a pregnancy is lost from multifetal reduction, the risks to the mother are lower from an early loss than from a complicated delivery of a multiple gestation.‘0,‘5 In addition, although successis never guaranteed, the chance that the patient can conceive another IVF pregnancy appears fairly good, given that IVF treatment resulted in a multiple gestation. Others25,2hhave reported that multifetal reduction to twins of multiple gestations resulting from treatment with gonadotropins substantially increased the duration of gestation, reduced the incidence of prematurity, reduced neonatal mortality, and shortened the neonatal intensive care unit stay. To our knowledge, however, ours is the first study of a population exclusively made up of IVF patients. In our experience at Women and Infants’ Hospital, multiple gestations resulting from IVF are at a significantly increased risk for low birth weight and small for gestational age infants, and for morbidity associated with prematurity, in comparison with non-IVF controls.’ Although the small numbers available make it impossible to determine statistical significance accurately, the mortality of IVF infants at Women and Infants’ Hospital was 2.7-fold higher in triplets than in twins (7.41 versus 2.78%)-i However, this figure represents only the tip of the iceberg for the increased risks faced by the IVF triplets in comparison with IVF twins. Compared with IVF twin survivors, IVF triplet survivors at Women and Infants’ Hospital experienced a 4.8-fold greater risk of culture-positive neonatal sepsis(29 versus 6%; P < .05), a 3.9-fold greater risk of respiratory distresssyndrome treated with surfactant (71 versus 18%‘;P < .005), a 3.8-fold increase in the risk of patent ductus arteriosis treated with indomethacin (57 versus 15%; P < .Ol), a 77-fold increase in days of oxygen administration (54 + 47 versus 0.7 2 2; P < .02), and a 67-fold increase in days of mechanical ventilation (40 2 54 versus 0.6 2 2; P < .025).’ Multifetal reduction of pregnancies with three or more fetuses to twins added, on average, 3.7 weeks to the observed average of 29.7 weeks in unreduced triplet pregnancies. Although we recognize that there is variability in the duration of gestation for individual cases,the average of 3.7 additional weeks in utero did much to reduce the risk of death or long-term morbidity for the infants in the cases
Haning et al
IVF
Prqpancy
Duration
967
treated with multifetal reduction. Each of our patients who lost an IVF multiple gestation after multifetal reduction was able to establish a subsequent pregnancy using IVF, with a good outcome. Counseling the patient under such circumstances is challenging because of the interaction of emotions and religious beliefs, but it was made all the more difficult by the lack of sound data on the effects of either multiple gestation or multifetal reduction on the outcome of the IVF pregnancy. We hope that the data given here will help by providing statistically based information on the effects of both fetal number and multifetal reduction on the duration of gestation. The Women and Infants’ Hospital experience with multifetal reduction in IVF patients indicates that considerable benefits are achieved by multifetal reduction of pregnancies with three or more fetuses. Because of the emotional, religious, and ethical dilemmas, only the patients themselves can make the final decision. The patient must decide which side of the risk-benefit scale to discount and which side to emphasize.
12. Berkowitz Alvarez trimester.
1. Tallo
Cl’,
Vohr
B, Oh
W,
Rubin
L, Seifer
morbidity associated 1995;127:794-800.
DB,
Haning
RV
with
in vitro
fertili-
2. Bell RJ, Eddie LW, Lester AR, Wood EC, Johnston Relaxin in human pregnancy serum measured with
transabdominal cases. Obstet
radioimmunoassay. 3. McClure N, Leya phase Hum 4. Lynch
Obstet Gynecol 1987;69:58>9. J, Radwanska E, Rawlins R, Haning
R. The effect of multifetal pregnancy Obstet Gynecol 1995;85:756-9. 5. Haning Sarmento
RV Jr, Goldsmith J, et al. Relaxin
in-vitro fertilization/gamete 1994;9:141-6.
Philipson of singleton
G, F’etrucco pregnancies
intra-fallopian
transfer.
relaxin.
Frishman G, Am J Obstet 0, Anderson following Hum
T, Dale PO, in singleton
Lunde 0, pregnancies
Obstet 8. Haning multiple
Gynecol 1995;86:188-92. RV Jr, Steinetz B, Weiss pregnancy following
Moe
N, after
Abyholm assisted
G. Elevated menotropin
JA, Goldsmith
NJ, Weiss G. The effect of ovulation of maternal serum relaxin in
twin
serum relaxin levels in treatment. Obstet Gy-
LT, Shahinian
KA,
Erinakes
induction on the concentration pregnancies. Am J Obstet
Gynecol 1996;174:227-32. 10. Timor-Tritsch IE, Peisner DB, Monteagudo A, Lerner JP, Sharma S. Multifetal pregnancy reduction by transvaginal puncture: Evaluation of the technique 1993;168:799%804.
used
in
134
cases.
Am
11. Brandes JM, Itskovits J, Timor-Tritch Reduction of the number of embryos Steril 1987;48:326-7.
IE, Drugan in multiple
968 Haning
Duration
et al
IVF
Pregnancy
J Obstet
Gynecol
A, Frydman R. pregnancy. Fertil
U,
ultrasonic
Blankier
J, Casper
1988;
First-trimester A report
of 85
Lynch L, Dumez Y, multifetal pregnancy the world’s largest A.
Puncture
proce-
Ultrasound
Ob-
E. Selective reduction in after transvaginal and Fertil Steril 1989;52:
Transvaginal
multiple
ultrasound
pregnancy.
perinatal
in
Obstet
necessity.
Gy-
N Engl
J
1988;318:1062-3.
Dahl Lyons CA, Wheeler CA, Frishman GN, Hackett RJ, Seifer DB, Haning RV Jr. Early and late presentation of the ovarian hyperstimulation syndrome (OH%): Two distinct entities with different
risk factors. Hum Reprod 21. Snedecor GW, Cochram Iowa: 22. Doyle weight 23.
M.
guidance.
RG.
Y,
selective embryo reduction for necol 1990;75:720-2. 19. Hobbins JC. Selective reduction-A
Gynecol
reduction:
M, Wapner RJ, of transabdominal experience among
transvaginal 1991;1:144-50.
MH, Struyk and quadru-
Obstet
Alvarez
416-20. 18. Gonen
resulting Haning
1994;9:792-9. WG. Statistical
The Iowa State University P, Beral V, Maconochie and small-for-gestational-age from in-vitro RV Jr, Kiggens
one, 17P-estradiol, follow treatment
and estriol with human and
methods.
6th
ed.
Ames,
Press, 1967:268-71. N. Preterm delivery, low birthin liveborn singleton babies
fertilization. AJ, Leifheit
of multiple gestation Biochem 1985;22:823-9. 24. Sherwood OD. Relaxin.
Hum Reprod TL. Maternal
are increased menopausal maternal
In: Knobil
in pregnancies gonadotropins:
endocrine E, Neil1
1992;7:425-8. serum progesterwhich Effects
status. JD,
J Steroid
Greenwald
GS,
Markert C, Pfaff DW, eds. The physiology of reproduction. Vol. 1. New York: Raven Press, 1994:861-1009. 25. Melgar outcome
Reprod
T. Obstetric reproduction.
Chitkara
issues.
M, Shale” outcome procedures.
26.
7. Tanbo outcome
ethical
17. Shalev J, Frenkel Y, Goldenberg multiple gestations: Pregnancy transabdominal needle-guided
R, Wein
on serum
LT, Seifer DB, Wheeler CA, secretion in IVF pregnancies.
Gynecol 1996;174:233-40. 6. Wang JX, Clark AM, Kirby CA, G, et al. The obstetric outcome
necol 1985;66:42-5. 9. Haning RV Jr, Canick
reduction
syndrome.
and
BW, Quigg in octuplet
Gynecol 1993;82:61-6. IE, Peisner DB, Monteagudo
dures utilizing stet Gynecol
RV Jr. Luteal
support and severe ovarian hyperstimulation Reprod 1992;7:758-64. L, Berkowitz RL, Weiss G, Goldsmith LT, Lapinski
IE, Newton termination
multifetal pregnancy Gynecol 1990;75:735-8.
centers. Obstet 16. Timor-Tritsch
Mehalek KE, in the first
1988;318:1042-7.
15. Evans MI, Dommergues Goldberg JD, et al. Efficacy reduction: Collaborative
Jr.
I’D, Nial HD. a homologous
J Med
plet pregnancies: Clinical 711289-96. 14. Lynch L, Berkowitz RL,
20.
Maternal and neonatal zation (IVF). J Pediatr
N Engl
13. Evans MI, Fletcher JC, Zador CD. Selective first-trimester
Med
References
RL, Lynch L, Chitkara U, Witkins IA, E. Selective reduction of multifetal pregnancies
duced Lipitz
CA, Rosenfeld after multifetal
DL, Rawlinson K, Greenberg reduction to twins compared
multiple gestations. S, Reichman B, Uval
2nd
ed,
M. Perinatal with nonre-
Obstet Gynecol 1991;78:763-7. J, Shale” J, Achiron R, Barkai
G, et al.
A prospective comparison of the outcome of triplet pregnancies managed expectantly or by multifetal reduction to twins. Am Obstet Gynecol 1994;170:874-9.
Reprints are not available. Address correspondence Ray V. Han@, Jr, MD 5793 Crowder #B352 New Orleans,
Boulevard LA
70127
Received August 23, 1995. Received in revised form December Accepted lanuary 22, 1996. Copyright Gynecologists.
0
to:
1996 by Published
The American by Elsevier
28, 1995.
College of Obstetricians Science Inc.
Obstetrics
and
& Gynecology
J