The present study in women with triplet pregnancies conceived by IVF/ICSI revealed that 38.3% of cases involved spontaneous loss of at least one fetus, and 89.3% of these cases occurred before 12 weeks of gestation. After adjustment, early MPR was related to higher risks of spontaneous fetal loss and postprocedural fetal loss than late MPR. Singletons born after early MPR were associated with a higher mean birth weight, apart from this, the outcomes after early MPR and late MPR were comparable. Our study also demonstrated that the pregnancy and neonatal outcomes of women in the early SR group were not inferior to those in the early MPR group.
In the present study, nearly 40% of triplets experienced spontaneous loss of at least one fetus, which is roughly accordant with previous reports [15, 16, 26]. There is limited research on the pregnancy outcomes after spontaneous fetal loss in triplet pregnancy. Barton et al. reported that the risk of preterm delivery before 32 weeks of gestation was significantly increased in twins born after vanished triplets, with a mean reduction in gestational age of 1.5 weeks than non-vanished twin pregnancies [17].
Overall, we found comparable pregnancy outcomes as well as maternal and perinatal complications for twin pregnancies after early transvaginal MPR before 9 weeks of gestation and late transabdominal MPR at 11–20 weeks. The median gestational age at delivery and preterm birth rate did not differ significantly. Moreover, there was no significant difference in the incidence of the pregnancy complications including GDM, HDP, and PPROM, which concurred with previous results [12, 27]. The study by Lipitz et al. elucidated that performing MPR at 13–14 weeks was related to a similarly low rate of pregnancy loss as MPR at 11–12 weeks [28]. However, Haas et al. found that fetuses born after early MPR at 6–8 weeks were associated with a lower risk for SGA [12]. Our outcomes did not show a statistical difference regarding the SGA rate between the early and late MPR groups. This inconsistency could be ascribed to that our cohort did not include quadruplet pregnancies and higher order multiple pregnancies. It has been reported that MPR at 7–8 weeks was superior to that in the second trimester; however, this advantage was only found in pregnancies with five or more fetuses [10].
Hitherto, only a few studies have described the maternal and perinatal outcomes after fetal reduction from triplets to singletons. We demonstrated that the outcomes of singleton pregnancies after early MPR appeared to be comparable to that of late MPR. Nevertheless, singletons born after early MPR were associated with a higher mean birth weight (3268 ± 516g vs. 2955 ± 626g; P = 0.010) in our study, which might be explained by more residual fetal placental tissue after late MPR of two fetuses affecting the growth of the surviving fetus. However, a previous study found no significant difference in the mean gestational age or birth weight of singleton babies born after early MPR at 6–8 weeks compared to those after late MPR at 11–20 weeks [13]. The results of Yan et al. showed that the perinatal outcome after singleton delivery following MPR at 11–14+ 6 weeks of triplet pregnancies was similar to MPR at 15–24+ 6 weeks, hence the study concluded that the timing of MPR did not affect the neonatal outcome of singletons [29]. The different findings could be attributable to the relatively small sample sizes of these studies.
Pregnancy loss is one of the primary risks associated with MPR [30]. We observed that early MPR was associated with a higher incidence of postprocedural fetal loss, which was consistent with previous studies. Kim et al. reported that the frequency of postprocedural fetal loss tended to be higher in an early transvaginal MPR group at 6–8 weeks than in a late transabdominal MPR group at 11–20 weeks [13]. Twins pregnancies after early transvaginal MPR were associated with a higher risk of losing one fetus than those after late transabdominal MPR, as reported by Haas et al. (6% vs. 0.8%; P < 0.05)[12]. In stark contrast, Lee et al. demonstrated that the immediate loss rate was lower in an early MPR group (before 8 weeks of gestation)[24]; however, in that study, all procedures were performed transvaginally. Geva et al. found that there was no statistically significant difference in the incidence of pregnancy loss between a second-trimester MPR group and a first-trimester MPR group [31]. Furthermore, several studies have compared postprocedural fetal loss rates of MPR by different methods and shown contradictory results [25, 32]. These inconsistent results may have been due to the relatively limited sample size and the different gestational weeks of MPR. Moreover, the populations in these studies included both iatrogenic and spontaneous multifetal pregnancies. The higher rate of spontaneous fetal loss in the early MPR group is to be expected, while spontaneous loss of one or more fetuses may still occur after the early operation. Otherwise, the prevalence of first-trimester vaginal bleeding and the higher incidence of subchorionic hematoma after early transvaginal MPR may correlate with the higher rate of postprocedural fetal loss [25].
To our knowledge, the present study is the very first one to compare the outcomes and complications of IVF triplet pregnancies after early transvaginal MPR with SR before 12 weeks of gestation. Our study found that compared to early transvaginal MPR, early SR of triplet pregnancies was not connected with adverse maternal and perinatal outcomes. Skiadas et al. reported that SR before 12 weeks in triplets was associated with an increased gestational age similar to that for transabdominal MPR at 11–13 weeks [21]. Ata et al. suggested that compared to dichorionic twins, the risks of preterm delivery and pregnancy loss were not significantly increased after SR or MPR at 12–14 weeks of IVF triplets [20]. In contrast, a previous study reported that the preterm delivery rate and VLBW rate of twins born after SR were significantly higher than those born after transabdominal MPR [23]. However, the sample sizes in these two studies were relatively small. Compared with that of twins after vanished triplet pregnancies, the likelihood of early pregnancy loss was lower in twin pregnancies after early MPR in the present study, but the difference was not statistically significant. The study by Bhandari et al. found that the incidence of loss of an extra fetus in the SR group was higher than that in the MPR group [22]. Our study supported that it might be feasible for triplet pregnancies to expect the occurrence of spontaneous fetal loss until 11–12 weeks of gestation, as the pregnancy and perinatal outcomes after SR were not worse than those after early MPR. However, attention should be paid to the risk of early pregnancy loss after the SR of two fetuses.
Early MPR is performed earlier and without lethal injection, which is similar to ovum retrieval. However, the occurrence of SPR eliminates the need for performing fetal reduction. When there are more than four gestational sacs in the uterine cavity, fetal abnormalities are rarely detectable by ultrasonography at 6–8 weeks of gestation. Only irregularly formed gestational sacs have an impact on the choice of reduced fetus. In our cohort, one patient underwent late MPR due to fetal hydrocephalus detected during antenatal care after early MPR. The accuracy of antenatal care in the second trimester largely determines the advantages of MPR in this period, such as detection of structural anomalies, non-invasive DNA examination, amniocentesis biopsy, and other prenatal diagnoses [33]. Therefore, the “right” fetus can likely be selected for reduction.
Considering the similar outcomes between the early MPR group and the early SR group, it appears that the mechanism causing spontaneous fetal loss has no adverse effect on the growth of the surviving fetus or pregnancy outcomes compared to early MPR. Since SR mostly occurs within the first 12 weeks of gestation, it could be advisable to expect its occurrence before considering MPR.
The strengths of this study are its relatively large sample size and sufficient follow-up time between the performance of MPR and delivery. The present study is a complete IVF cohort study and is consistent with real-world studies. Furthermore, this study divided triplet pregnancies which underwent SR or MPR into subgroups of singleton and twin pregnancies and compared the outcomes separately. One of the limitations of this study is that due to the nature of retrospective studies, it is not possible to avoid the bias associated with the study results. In our cohort, there were only 3 cases of twin-to-twin transfusion syndrome in the SR group, possibly due to the relatively small sample size, hence the specific risks of monochorionic diamniotic twins might not be fully demonstrated in our study.