The increasing burden of infertility has led to an expanded rate of use of in vitro fertilization (IVF) [1]. In an increasing number of countries, elective single embryo transfer is being implemented, however in the majority of countries double embryo transfer is still the routine which has in turn increased twin conception rates and the incidence of vanishing twin (VT) pregnancies. With regard to IVF pregnancies, it is estimated to occur in 10–30% of pregnancies [2, 3, 4, 5, 6]. The etiology of the VT phenomenon is still obscure although, placental degeneration confirmed pathologically [7] and chromosomal abnormality in the vanishing embryo [8] have been documented. Other possible causes, include chronic maternal diseases, placental “crowding”, intrauterine bleeding, and inappropriate site for implantation have been proposed [9, 10, 11, 12].

As to the VT, it is clinically important to understand the effects of VT on the survival fetal. An aspect that the patients care about mostly is whether the survivor would be secondary fetal death or not affected by the vanishing one. However, seldom study did such a research and none can answer the rates of secondary fetal death after the first vanishing one. Another aspect they are concerned about is whether the aborted embryo will affect the growth and development of the surviving fetus and whether it is healthy at birth. The pathologic implications of the VT on the surviving twin have been debated. Some studies indicating a higher incidence of adverse obstetric and perinatal outcomes [2, 3, 4, 5, 13] compared with singleton pregnancies. Adverse outcomes include higher risk of preterm birth, low birth weight(LBW), small for gestational age (SGA) and birth defects. What’s more, some papers reported some serious complications, such as brain damage or disseminated intravascular coagulation in the surviving one. However, other studies demonstrated that vanishing twin syndrome(VTS) were associated with slightly higher risks of preterm birth and LBW [14] or had similar outcomes with singleton pregnancies [15, 16], or even a protective effect [17]. The main reason for these conflicting results may be that the definition of VTS varies among different authors. Gestational sacs can “vanish” at several time points, from empty gestational sac to gestational sac with evidence of cardiac activity [18, 19] or from the first trimester to all three trimesters [2].

However, only few of these studies could rule out the possibility of confounding by vanishing gestational age and the conclusion of the existing literatures were also conflicting. Furthermore, seldom study did the research about the rates of secondary fetal death after the first vanishing one. A sibling comparison design is useful to study the association between vanishing gestational age and the obstetric and neonatal outcomes for accounting for fetal loss during the whole pregnancy. The objective of the current study was therefore to estimate the rates of secondary fetal death of a vanishing twin and differential impact on perinatal outcomes according to vanishing gestational age.

This retrospective cohort study was conducted on the clinical data obtained from IVF/ICSI performed in assisted Reproductive center of Guangzhou Women and children hospital from January 2011 to January 2019.

All the patients should sigh a follow up agreement before IVF. Gestational age was calculated by the date of embryo transfer. The date of oocyte egg retrieval is day 0 (D0). If day 3 embryo is transferred, the first day of pregnancy is 17 days ahead of transfer date. If the blastocyst is transferred, the first day of pregnancy is 19 days ahead of transfer date. If pregnant, on gestational week 6, women need to return to our hospital for the first ultrasound and then they were followed by biweekly ultrasounds, gestational age of fetuses lost and the gestational sac number were calculated at that time. All the follow up data were registered in our IVF system. All the patients’ information and clinical data were obtained from our IVF system.

The inclusion criteria includes: pregnancies that meet the inclusion criteria in gestational week 6 (A) one viable fetus; (B) one fetus with cardiac activity and a gestational sac/fetus without fetal heart beat; or (C) two viable fetuses. The following were excluded: women who received pre-implantation genetic diagnosis or fetal reduction; pregnancies in which more than two fetal hearts were beating, gestations with no viable fetuses and monozygotic twins (Either of two twins developed from the same fertilized ovum). In this research, the VTS was defined as: two gestational sacs were demonstrated at gestational weeks 6 and thereafter, only one fetal pulse was demonstrated, the cases were included as the study group.

For a better understanding of the rates of secondary fetal death after the first vanishing one, we compared the pregnancy outcomes of the co-twin in VT pregnancies and determine the rates of secondary fetal death after the first vanishing one according to gestational week at the time of determination of VT.

Subsequently, we selected patients that singletons gestation without vanishing and the vanishing twin who delivered a singletons gestation into a study group. We separate the study group into four study cohorts. Singleton cohort (group I) consisted of born singletons registered with only a gestational sac/fetus and one heart beat in early pregnancy in accordance with ‘inclusion criteria A’; Vanish before the presence of fetal cardiac activity cohort (VT before CA) (group II) consisted of singletons detected as twin pregnancies on the basis of ‘criteria B’ by ultrasound during early pregnancy; Early vanish cohort (group III) consisted of singletons detected as twin pregnancies on the basis of ‘criteria C’ by ultrasound and co-twin perishing after 6 gestational weeks but prior to 12 weeks gestational; and late vanish cohort (group IV) consisted of singletons detected as twin pregnancies on the basis of ‘criteria C’ by ultrasound and co-twin disappeared after 12 weeks gestational. We compared the perinatal outcomes of the co-twin in VT pregnancies with IVF/ICSI singletons without vanishing.

Preterm delivery was defined as child born before 37 weeks’ gestational age. LBW was defined as birth weight under 2500 g [20]. Perinatal malformation and birth weight data were acquired by telephone interview within 1 week of birth. Perinatal outcomes were compared in terms of mean maternal age, patient Body Mass Index(BMI), type of infertility, cycle type, mode of insemination (IVF/ICSI), embryo transfer stage and sex of the offspring, gestational age, birth weight among the groups. Data are reported using the mean SD or n (%).

Statistics: Chi-squared test was used to evaluated the differences between proportions. Differences between parameters in the specific patient groups were evaluated using the t-test where appropriate. Then, multivariate analysis was performed using binary logistic regression to adjusted for all factors. A result was considered statistically significant if a two-tailed P-value was less than 0.05. All data analyses were performed using SPSS for windows 20.0 (IBM, Armonk, NY, USA).

Given the rising rates of assisted reproductive technologies (ART) pregnancies internationally, pregnancy and perinatal outcomes in ART pregnancies are very important basis for assessing ART safety. This study showed that VT pregnancy after 12 weeks gestation did not have a higher rate of secondary fetal death but was identified as a factor for determining the risk of adverse perinatal outcome to the surviving singleton. From our knowledge, it’s the first study to evaluate the secondary fetal death rates of vanishing co-twin. We elucidated the rates of secondary fetal death of a vanishing twin and differential impact on perinatal outcomes according to vanishing gestational age at the time of vanishing. In our study, the rate of secondary fetal death of vanishing twin is 15.4% with the surviving twin following VTS happening prior to 12 weeks gestation having similar secondary fetal death rates. With late vanishing twin ($>$12 weeks), the surviving singleton had a higher risk of preterm delivery (P 0.001) and LBW (P 0.001) compared to a singleton without vanishing.

In our study, 12.5% of singleton deliveries after IVF/ICSI were originated from a VT pregnancy. The result was consistent with some other studies describing approximately 10% [2, 21]. When compared the perinatal outcome data between very early vanish or early vanish and singleton group, we find no significant differences on the term of mean gestational week, mean birth weight, rates of preterm delivery and LBW. This is different from one study which reported that with vanish before 12 weeks gestation, survivors showed a higher frequency of lower birth weight and smaller gestational age when compared with a singleton cohort [3] In comparing late vanish and singleton controls, the late vanish group showed smaller gestational age, a higher rate of preterm delivery and LBW. In addition to the adverse perinatal outcomes, studies have demonstrated higher neonatal mortality and birth defects in singleton deliveries of late VT pregnancies [4, 21, 22]. In our study, those differences did not reach statistical significance.

In other studies, some investigators did not distinguish vanishing time between the first and second trimesters [2], and failed to consider vanishing gestational age during pregnancy [13, 23, 24]. For example, Pinborg [2] has suggested that a surviving co-twin from the early vanishing times group (8 weeks), and singletons from singleton pregnancies, have similar perinatal outcomes. However, pregnancy loss before and after the presence of fetal cardiac activity was not differentiated. Our data as that of others have concluded that vanishing gestational age was close correlated to poor pregnancy outcomes. The higher the vanishing gestational age, the higher the risk that the surviving infant will be subjected to a poor perinatal outcome [2, 21, 22], especially when vanishing occurred after 12 weeks gestation. As the rates of vanishing twin happened prior to 12 weeks gestation being close to 90%, this results in most of the singleton with a vanishing twin are more likely to be born a baby as healthy as the infant from singleton delivery.

The reason for the poor perinatal outcomes when the vanishing twin occurs after 12 weeks gestation might be secondary to the increase in necrotic fetoplacental tissue needing to be absorbed and the release of more prostaglandins and cytokines which can initiate an intense inflammatory response thus influencing placental function which can affect the growth and development of the surviving fetus [25]. However, the inflammatory response caused by VT before 12 weeks gestation is too small to influence the perinatal outcomes of surviving singleton.

The limitations of our current study is as follows. First, the sample size of our study is small, large sample studies are necessary for better understanding the relationship between the VT and multiple pregnancy influences. Another limitation is that it’s difficult to determine an empty gestational sac by ultrasound and it could be misinterpreted or overlooked resulting in a possibility of an underestimation of spontaneous reduction at 6 weeks gestation.

IVF/ICSI twins do carry a higher risk than IVF/ICSI singletons regarding perinatal outcome measures. Twin after 12 weeks gestation is a predictor of adverse perinatal outcome compared the singletons without VT. Thus, the adverse outcome carried in VT after 12 weeks gestation is still a major concern in ART. Hence, the elective single embryo transfer policy in IVF/ICSI patients is now highly recommended. As related to patients with VT after 12 weeks gestation, we should more carefully monitor their antenatal course. The couples with a viable fetus or a non-viable fetus and an empty gestational sac in early pregnancy (12 weeks) have a similar outcome as the baby from a singleton delivery without vanishing.

LS, JL, CLL designed the research study. LS contributed to revise the article critically for important intellectual content; JL, CLL performed the research, analyzed the data, drafted and revised the whole article; QL, ZW, FL, DC and CYL contributed to collect the data, provide help and advice on this research. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

All subjects gave their informed consent for inclusion before they participated in the study. The study was approved by the Independent Ethics Committee of Guangzhou Women and Children’s Hospital (approval number: #2020-8).

We would like to express my gratitude to all those who helped me during the writing of this manuscript. Thanks to all the peer reviewers for their opinions and suggestions.

This research received no external funding.

The authors declare no conflict of interest.