The `vanishing embryo' phenomenon in an oocyte donation programme

Manuel Rodríguez-González,1, Vicente Serra, Juan Antonio Garcia-Velasco, Antonio Pellicer and José Remohí

Instituto Valenciano de Infertilidad and Department of Pediatrics, Obstetrics and Gynaecology, Valencia University School of Medicine, Valencia, Spain


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: We studied the incidence of vanishing embryos (VE) in pregnancies achieved by oocyte donation and evaluated the obstetric and perinatal complications. METHOD: A retrospective study was carried out based on a chart review of 399 patients with multiple pregnancies from our oocyte donation programme. We defined vanishing phenomenon as the early resorption, in the first trimester, of one or more embryos in a multiple gestation, after confirming embryonic heart activity by transvaginal ultrasound. RESULTS: Vanishing embryo was observed in 75 patients (18.8%). In 60 patients (80%) this phenomenon occurred before the ninth gestational week. A higher incidence of VE was observed in patients who initially showed a higher number of gestational sacs (P < 0.03). Vaginal bleeding in the first trimester was significantly higher in patients with VE (P < 0.005). Miscarriage rate was similar in pregnancies with and without VE (P = NS). The incidence of pregnancy induced hypertension was decreased in the group with VE (P < 0.03). Preterm spontaneous rupture of membranes occurred more frequently in pregnancies with VE (P < 0.05). However, gestational age at delivery was similar in the group with VE and the controls. CONCLUSIONS: The high incidence of VE in pregnancies achieved by oocyte donation should be considered when counselling patients with high order multiple gestations.

Key words: oocyte donation/perinatal outcome/spontaneous embryo reduction/vanishing embryo


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Since assisted reproduction techniques were introduced in routine clinical practice, a rise in the incidence of multiple pregnancies has been observed (Bustillo and Zarutskie, 1998Go). This implies a higher rate of maternal as well as perinatal complications (Bernasko and Lynch, 1997Go). Among these complications is spontaneous fetal loss, whose incidence and aetiology remains unknown. In 1945, Stoeckel described a higher twin pregnancy rate than the observed twin delivery rate. Even though clinical evidence was not possible as ultrasound did not exist, the resorption of one of the fetuses (foetus papiraceous) was described (Stoeckel, 1945Go).

With the advent of transvaginal ultrasound and assisted reproduction techniques, we learned that `vanishing embryo' (VE) is not an infrequent event. However, little is known about the pathophysiology of the process, an event that has been considered as a natural adaptation mechanism (Boklage, 1995Go). Other aetiological factors that may be involved are embryo aneuploidy (Tharapel et al., 1989Go; Rudnicki et al., 1991Go; Callen et al., 1991Go; Post and Nijhuis, 1992Go; Falik-Borenstein et al., 1994Go) or congenital abnormalities (Weissman et al., 1994Go). Due to the difficulties and limitations in its definition and diagnosis, the reported frequency of vanishing phenomenon has ranged from 3.7–100% (Dickey et al., 1990Go; Legro et al., 1995Go). The great diversity of the population studied and the limitations of the diagnostic techniques employed contribute to the confusion that exists around this particular event.

We defined vanishing phenomenon as the spontaneous loss of one or more embryos after identifying heart activity. Attempting to minimize interpretative error we identified a true intrauterine gestational sac using several sonographic characteristics. These included: a double contour, identification of a yolk sac within the gestational sac, and recognition of an embryonic heart after 6 weeks of gestation (Blumenfeld et al., 1992Go).

When a pregnancy achieved by oocyte donation becomes clinically evident and fetal heart activity is evidenced by ultrasound, early pregnancy loss has been estimated at around 20% (Cano et al., 1995Go). This rate of pregnancy loss is obtained from both singleton and multiple pregnancies. The multiple pregnancy rate in oocyte donation programmes ranges from 25–30% (Remohí et al., 1997Go). Interestingly, spontaneous pregnancy loss mainly occurs between 8 and 9 weeks of gestation (Sampson and de Crespigny, 1992Go).

Vanishing embryos may be observed in 21% of dichorionic twins and in up to 50% of monochorionic twins (Benson et al., 1993Go). In triplet pregnancies, VE of one of the embryos may be observed in 90% of the cases during the first 7 gestational weeks (Manzur et al., 1995Go). Spotting is the most frequent clinical sign, being observed in 15–25% of the cases (Falco et al., 1996Go). This spotting is associated with early pregnancy loss in 7.8–76.5% (Yoshida, 1995Go). When the placenta is studied, the vanished embryo is described as a placental cyst (Nerlich, 1992), degenerated chorionic villi (Rudnicki et al., 1991Go), fibrin deposition (Yoshida, 1995Go), nodules or macerated embryos (Blumenfeld et al., 1992Go).

Several variables have been investigated in order to avoid selective embryo reduction in high order multiple gestations. Serum levels of HCG (Kelly et al., 1991Go), crown–rump length (Kol et al., 1993Go), or bradycardia in early pregnancy (Falco et al., 1996Go) as well as first trimester vaginal bleeding have been evaluated, but none of them seemed to be predictive of VE.

As oocyte donation offers an excellent model to monitor multiple pregnancies from the beginning, the aim of our study was to establish the incidence of VE in a population of infertile patients undergoing this particular assisted reproduction technique, and to describe the perinatal complications that may occur.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Patients
A total of 581 pregnancies from our oocyte donation programme were retrospectively studied from January 1997 to December 1999. A total of 1189 donation cycles were carried out and the number of oocytes donated and embryos transferred were 7.4 ± 1.9 and 3.2 ± 1.5 respectively. The age of recipient patients was 37.5 ± 5.6 years. Multiple pregnancies with evidence of VE were allocated to group I (study group). Recipients with either singleton pregnancies or multiple pregnancies without VE were included in group II (control group) (Figure 1Go). The obstetric outcome of singleton and twin pregnancies after VE was compared with that of initial singleton and twin pregnancies respectively. VE was diagnosed when at least one embryo vanished after previous identification of embryonic heart activity. Cases with blighted ova, biochemical pregnancy, miscarriage, induced selective reductions and ongoing triplets were excluded from this analysis. We defined `miscarriage' as the loss of fetal heartbeat in a clinical pregnancy.



View larger version (18K):
[in this window]
[in a new window]
 
Figure 1. Patient selection and groups of study.

 
There were no differences between the established groups of recipients regarding age, cause of infertility, or distribution of donor oocytes (Table IGo). The protocol of ovarian stimulation, steroid replacement, oocyte retrieval and gamete handling in the IVF laboratory has been described elsewhere (Pellicer et al., 1989Go).


View this table:
[in this window]
[in a new window]
 
Table I. Baseline characteristics in groups I (evidence of vanishing embryo) and II (lack of vanishing embryo)
 
Methods
The protocol of steroid replacement for recipients has also been described previously (Remohi et al., 1995Go). Briefly, patients with ovarian function were desensitized with i.m. administration of depot leuprorelin acetate (Ginecrin® depot; Abbott S.A., Madrid, Spain) beginning in the secretory phase of the previous cycle. HRT started on day 1 of the cycle with the administration of 2 mg/day of estradiol valerate (Progynova®; Schering Spain, Madrid, Spain) from days 1–8; 4 mg/day from days 9–11; and 6 mg/day from day 12 onwards. After 13 days of estradiol valerate administration, recipients were ready to receive oocytes and they waited until a donation became available. At the day of recovery of donated oocytes, 800 mg/day of micronized vaginal progesterone (Progeffik®; Laboratories Effik S.A., Madrid, Spain) were administered. Embryo transfer was performed 48 h after oocyte recovery using the transcervical approach. The regimen of 6 mg/day of estradiol valerate and 800 mg/day of progesterone was maintained for 15 days, after which serum HCG analysis was performed. When pregnancy was achieved (serum ß-HCG >=5 mIU/ml, Axsym®; Abbott), gestational sac and embryo assessment were performed weekly by transvaginal ultrasound test (Siemens Sonoline SI 410). The first sonogram was performed 7 days after a positive pregnancy test (fifth gestational week). Estradiol valerate and progesterone were maintained at the same dosage until day 80 of pregnancy (Guanes et al., 1996Go).

Statistical analysis
Continuous data were expressed as mean ± SD. Categorical values were expressed as n (%). Student's t-test, {chi}2, and Fisher's exact test were used where appropriate. A value of P < 0.05 was considered as significant. Statistical calculations were performed using Sigmastat® for Windows, version 2.0 (Jandel Scientific Corporation, San Rafael, CA, USA).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
From 581 oocyte donation pregnancies followed up in our institution, we found 399 multiple pregnancies (68.7%). VE was observed in 75 cases (18.8%) (Figure 1Go). The incidence of VE increased with the higher number of gestational sacs initially seen (Table IIGo) (P < 0.03). Sixty patients (80%) experienced VE before completing the eighth gestational week, and only 15 (20%) between the ninth and the eleventh week. Mean gestational age at which VE was observed was 7.1 ± 1.7 gestational weeks, and there were no differences between twins, triplets and quadruplets. The incidence of VE was not influence by the age of the women, number of embryos transferred or any other baseline characteristic of the women, except for the donor serum estradiol on day 15 (Table IGo).


View this table:
[in this window]
[in a new window]
 
Table II. Vanishing embryo (VE) in relation to the number of gestational sacs initially observed
 
First trimester bleeding was more common among pregnancies with VE than in the control group (P < 0.005). However, there were no significant differences in the miscarriage rate between both groups (P = NS) (Table IIIGo).


View this table:
[in this window]
[in a new window]
 
Table III. Early pregnancy losses in oocyte donation pregnancies with and without vanishing embryo (VE)
 
In the same period of time we analysed the incidence of VE in IVF and ICSI and saw that the incidence of VE in oocyte donation was higher than in IVF (75/399; 18.8% versus 82/1058; 7.6%) or ICSI (75/399; 18.8% versus 124/1469; 8.4%) pregnancies (P < 0.001). Although the media of embryos transferred were similar in the all of them, pregnancy and multiple pregnancies rates were significantly lower in IVF and ICSI.

The perinatal outcome of pregnancies that experienced VE was also compared with that of pregnancies without the vanishing phenomenon (Table IVGo). The pregnancies with VE were singletons or twins pregnancies since the end of the first trimester. The incidence of pregnancy induced hypertension was lower in pregnancies with VE than in the controls (P < 0.03). In contrast, preterm spontaneous rupture of membranes was higher in the group with VE (P < 0.05). Term spontaneous rupture of membranes was also increased in twin pregnancies with VE (P < 0.001). Gestational age at delivery, mode of delivery and birthweight was similar in the group with VE and the controls (P = NS).


View this table:
[in this window]
[in a new window]
 
Table IV. Perinatal outcome in oocyte donation pregnancies with and without vanishing embryo (VE)
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Multiple embryo implantation is not an infrequent event in an oocyte donation programme (Remohí et al., 1997Go), but discovering more than one gestational sac before the eighth week should not be considered as definitive because VE may occur (Mansour et al., 1999Go). An expectant attitude would be adopted, trying to avoid a selective embryo reduction.

The evidence in the literature shows a variable incidence of VE after assisted reproduction treatment. A 13% incidence of VE in 191 gestational sacs diagnosed by ultrasound was described (Kol et al., 1993Go). Other authors describe a much higher incidence, for example Dickey et al., who showed in a retrospective study of 275 multiple pregnancies that only 57% of the twin pregnancies diagnosed in the first ultrasound scan remained with two sacs after the first trimester (Dickey et al., 1990Go). In our series, the incidence of VE in oocyte donation pregnancies was higher than in IVF or ICSI pregnancies (P < 0.001). We do not know the explanation for these differences, although a possible explanation is the high multiple pregnancy rate in oocyte donation (68.7%). Landy (Landy et al., 1998), based on combined statistics from most studies, compared the incidence of pregnancy resorption in the first trimester in assisted reproductive techniques versus spontaneous conceptions. If two sacs were identified sonographically, loss of one twin could be expected in 27.1% of pregnancies achieved after assisted reproduction and in 40.5% of spontaneous pregnancies; if two embryos were seen, the loss rate was 38% in pregnancies achieved after assisted reproduction and 7.3% in spontaneous conceptions (Landy et al., 1998).

Comparing IVF, ICSI and oocyte donation pregnancies with VE, we did not find any difference in the mean gestational age at which VE was observed or in the miscarriage rate. The mean gestational age of VE was 7.5 ± 2.1 gestational weeks in IVF, 7.3 ± 1.5 gestational weeks in ICSI and 7.1 ± 1.7 gestational weeks in oocyte donation (P = NS). Miscarriage rate among IVF, ICSI or oocyte donation was 9.7% (8/82), 5.6% (7/124) and 6.6% (5/75) respectively (P = NS).

Oocyte donation offers an excellent model of controlling multiple pregnancies from the early stages of pregnancy. Oocytes are obtained from fertile women under 35 years of age, and then transferred after fertilization to a receptive endometrium previously primed with estrogen and progesterone. This HRT is maintained during the first trimester of pregnancy. However, among oocyte donation pregnancies, a higher rate of miscarriage is observed in older women (Cano et al., 1995Go). Thus, uterine ageing is also a factor influencing reproductive performance, a fact that we should consider when multiple implantation is observed in the early stages of pregnancy. However, we did not find any relationship between donor or recipient age and incidence of VE (Table IGo). Interestingly, a lower miscarriage rate was found in oocyte donation pregnancies with VE compared with those without VE (particularly among twin pregnancies), but the differences did not reach statistical significance (Table IIIGo).

Although it is difficult to think that the difference in estradiol level on day 15 is the cause of VE, we are now investigating other factors that could be involved in miscarriage cases in oocyte donation (days on waiting list, a GnRH effect, endometrial thickness).

In our study the most common complications among pregnancies with VE were first trimester bleeding and spontaneous rupture of membranes. Remnants of intrauterine fetal tissues may diminish the placental bed surface in direct contact with the uterus, reducing the transport of nutrients from the mother to the fetus and develop a subclinical inflammatory condition, starting the preterm labour syndrome (Vadillo-Ortega et al., 1990Go; Hulboy et al., 1997Go) or a preterm rupture of membranes.

As a previous pregnancy (even if it is an early miscarriage) is known to protect against pre-eclampsia, we could speculate that a VE might exert a similar effect. This hypothesis might explain the lower incidence of pregnancy induced hypertension found in the group with VE (Table IVGo).

In order to avoid medical complications of high order multiple pregnancies, and emulating the natural process of embryo selection, multifetal pregnancy reduction is a valid alternative (Yaron et al., 1998Go). However, multifetal pregnancy reduction is a psychologically traumatic intervention (Bergh et al., 1999Go), and it is not free of complications (Torok et al., 1998Go; Coffler et al., 1999Go; Mansour et al., 1999Go). In our series, the outcome of oocyte donation pregnancies with spontaneous or induced embryo reduction was similar (unpublished data).

All this information may be useful in counselling patients on the prognosis and outcome of pregnancies achieved by oocyte donation. It may be a very valuable tool in assisting with decision making about multifetal pregnancy reduction before the ninth week of gestation. Additionally, it may add interesting information to the continuous debate on the number of embryos to be transferred.


    Notes
 
1 To whom correspondence should be addressed at: Instituto Valenciano de Infertilidad, C/Guardia Civil 23, 46020 Valencia, Spain. E-mail: ivivalencia{at}ivi.es Back

Submitted on April 10, 2001, resubmitted on June 28, 2001


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Benson, C.B., Doubilet, P.M. and Laks, M.P. (1993) Outcome of twin gestations following sonographic demonstration of two heart beats in the first trimester. Ulltrasound Obstet. Gynecol., 3, 343–345.

Bergh, C., Moller, A., Nilsson, L. et al. (1999) Obstetric outcome and psychological follow-up of pregnancies after embryo reduction. Hum. Reprod., 14, 2170–2175.[Abstract/Free Full Text]

Bernasko, J. and Lynch, L. (1997) Twin pregnancies conceived by assisted reproductive techniques: maternal and neonatal outcomes. Obstet. Gynecol., 89, 368–372.[Abstract/Free Full Text]

Blumenfeld, Z., Dirnfeld, M., Abramovici, H. et al. (1992) Spontaneous fetal reduction in multiple gestations assessed by transvaginal ultrasound. Brit. J. Obstet. Gynaecol., 99, 333–337.[ISI][Medline]

Boklage, C.E. (1995) The frequency and survival probability of natural twin conceptions. In Keith, L.G., Papiernik, E., Keith, D.M. (eds) Multiple pregnancy: epidemiology, gestation, and perinatal outcome. Parthenon, New York, pp. 41–50.

Bustillo, M. and Zarutskie, P. (1998) Assisted reproductive technology in the United States and Canada: 1995 results generated from the American Society for Reproductive Medicine/Society for Assisted Reproductive Technology registry. Fertil. Steril., 69, 389–398.[ISI][Medline]

Callen, D.F., Fernandez, H., Hull, Y.J. et al. (1991) A normal 46, XX/69, XXY placenta: a major contribution to the placenta is from a resorbed twin. Prenat. Diagn., 11, 437–442.[ISI][Medline]

Cano, F., Simón, C., Remohí, J. et al. (1995) Effect of aging on the female reproductive system: evidence for a role of uterine senescence in the decline in female fecundity. Fertil. Steril., 64, 584–589.[ISI][Medline]

Coffler, M.S., Kol, S., Drugan, A. et al. (1999) Early transvaginal embryo aspiration: a safer method for selective reduction in high order multiple gestations. Hum. Reprod., 14, 1875–1878.[Abstract/Free Full Text]

Dickey, R.P., Olar, T.T., Curole, D. N. et al. (1990) The probability of multiple births when multiple gestational sacs or viable embryos are diagnosed at first trimester ultrasound. Hum. Reprod., 5, 63–64.

Falco, P., Milano, V., Pilu, G. et al. (1996) Sonography of pregnancies with first trimester bleeding and a viable embryo: a study of prognostic indicators by logistic regression analysis. Ultrasound Obstet. Gynecol., 7, 165–169.[ISI][Medline]

Falik-Borenstein, T.C., Korenberg, J.R. and Schreck, R.P. (1994) Confined placental chimerism: prenatal and postnatal cytogenetic and molecular analysis, and pregnancy outcome. Am. J. Med. Genet., 50, 51–56.[ISI][Medline]

Guanes, P.P., Remohi, J., Gallardo, E. et al. (1996) Age does not affect uterine resistance to vascular flow in patients undergoing oocyte donation. Fertil. Steril., 66, 265–270.[ISI][Medline]

Hulboy, D., Rudolph, L. and Matrisian, L. (1997) Matrix metalloproteinases as mediators of reproductive function. Mol. Hum. Reprod., 3, 27–45.[Abstract]

Kelly, M.P., Molo, M.W., Maclin, V.M. et al. (1991) Human chorionic gonadotrophin rise in normal and vanishing twin pregnancies. Fertil. Steril., 56, 221–224.[ISI][Medline]

Kol, S., Levron, J., Lewit, N. et al. (1993) The natural history of multiple pregnancies after assisted reproduction: is spontaneous fetal demise a clinically significant phenomenon? Fertil. Steril., 60, 127–130.[ISI][Medline]

Landy, H.J. and Keith, L.G. (1998) The vanishing twin: a review. Hum. Reprod. Update, 4, 177–183.[Abstract/Free Full Text]

Legro, R.S., Wong, I.L. and Paulson, R.J. et al. (1995) Multiple implantation after oocyte donation: a frequent but inefficient event. Fertil. Steril., 63, 849–853.[ISI][Medline]

Mansour, R.T., Aboulghar, M.A., Serour, G.I. et al. (1999) Multifetal pregnancy reduction: modification of the technique and analysis of the outcome. Fertil. Steril., 71, 380–384.[ISI][Medline]

Manzur, A., Goldsman, M.P., Stone, S.C. et al. (1995) Outcome of triplet pregnancies after assisted reproductive techniques: how frequent are the vanishing embryos? Fertil. Steril., 48, 86–93.

Nerlich, A., Wisser, J. and Krone, S. (1992) Placental findings in `vanishing' twins. Gerburtshilfe Frauenheildkd, 52, 230–234.

Pellicer, A., Simón, C., Miró, F. et al. (1989) Ovarian response and outcome of in-vitro fertilization in patients treated with gonadotrophin-releasing hormone analogues in different phases of the menstrual cycle. Hum. Reprod., 6, 1387–1391.

Post, J.G. and Nijhuis, J.G. (1992) Trisomy 16 confined to the placenta. Prenat. Diagn., 12, 1001–1007.[ISI][Medline]

Remohi, J., Gutierrez, A., Cano, F. et al. (1995) Long estradiol replacement in an oocyte donation programme. Hum. Reprod., 6, 1387–1391.

Remohi, J., Gartner, B., Gallardo, E. et al. (1997) Pregnancy and birth rates after oocyte donation. Fertil. Steril., 67, 717–723.[ISI][Medline]

Rudnicki, M., Vejerslev, L.O. and Junge, J. (1991) The vanishing twin: morphologic and cytogenetic evaluation of an ultrasonographic phenomenon. Gynecol. Obstet. Invest., 31, 141–145.[ISI][Medline]

Sampson, A. and de Crespigny, L. (1992) Vanishing twins: the frequency of spontaneous fetal reduction of a twin pregnancy. Ultrasound Obstet. Gynecol., 2, 107–109.[ISI][Medline]

Stoeckel, W. (1945) Lehbuch der Geburstchilfe. Gustav Fisher, Jena (quoted in Levi, 1976).

Tharapel, A.T., Elias, S., Shulman, L.P. et al. (1989) Resorbed co-twin as an explanation for-discrepant chorionic villus results: non-mosaic 47 XX + 16 in villi (direct and cultured) with normal (46, XX) amniotic fluid and neonatal blood. Prenat. Diagn., 9, 467–472.[ISI][Medline]

Torok, O., Lapinski, R., Salafia, C.M. et al. (1998) Multifetal pregnancy reduction is not associated with an increased risk of intrauterine growth restriction, except for very-high order multiples. Am J. Obstet. Gynecol., 179, 221–225.[ISI][Medline]

Vadillo-Ortega, F., Gonzalez-Avila, G., Karchmer, S. et al. (1990) Collagen metabolism in premature rupture of amniotic membranes. Obstet. Gynecol., 75, 84–88.[Abstract]

Weissman, A., Achiron, R., Lipitz, S. et al. (1994) The first trimester growth-discordant twin: an ominous prenatal finding. Obstet. Gynecol., 84, 110–114.[Abstract]

Yaron, T., Johnson, K.D., Bryant-Greenwood, P.K. et al. (1998) Selective termination and elective reduction in twin pregnancies: 10 years experience at a single centre. Hum. Reprod., 31,2301–2304.

Yoshida, K. (1995) Documenting the vanishing twin by pathological examination. In Keith, L.G., Papiernik, E., Keith, D.M. (eds) Multiple pregnancy: epidemiology, gestation, and perinatal outcome. Parthenon, New York, pp. 51–58.

accepted on October 2, 2001.