Discussion
The principal findings of the current study are as follows: 1) the risk of preterm birth was higher in EM group compared to ER group; 2) however, the risk of non-viable pregnancy loss lower in EM group than in ER group; 3) the rate of cases with at least one alive neonate was higher in EM group compared to ER group; 4) short- and long-term outcome of neonates was comparable between ER and EM groups.
The risk of preterm birth had been consistently reported to be lowered after ER in previous studies, and the result of the current study also shows increased risk of preterm birth in ER group.7,10,11 However, the risk of non-viable pregnancy loss (miscarriage) after ER is controversial. In recent meta-analysis which analyzed 6 studies, the miscarriage risk was not different between EM and ER groups.12 In ER group compared to EM group, two studies [Antsaklis et al, 2004 (n=255); Chaveeva et al, 2013 (n=494)] showed increased miscarriage rate (2.9% ->8.1%; 3.9%->7.9% respectively),10,13 two studies [Drugan et al, 2013 (n=82); Shiva et al, 2014 (n=115)] showed similar rate (from 5.6% to 6.5%; from 12.3% to 12.1% respectively),11,14 and two studies [Ata et al., 2011 (n=65); Skiadas et al., 2011 (n=156)] showed decreased miscarriage rate (from 17.9% to 7.7%; from 14.5% to 6.9% respectively).15,16 Higher miscarriage rates in ER group have been observed in studies with larger numbers of subjects.10,13 Their results are consistent with the current study. As experience of MFPR increases, it was reported that the miscarriage risk after MFPR of triplet to twin pregnancies has reached to the twin’s natural miscarriage rate.17,18 However, considering that less invasive procedures such as chorionic villi sampling or amniocentesis has procedure related fetal loss,19 it is not reasonable that MFPR is innocuous. And we are concerned that dead fetus may have acute and remote effects on the other living fetuses.
In the current study, we have shown a statistically significant difference in the rate of pregnancies with at least one alive neonate in the EM group (97.7%) compared to the ER group (92.3%) (p=0.013). This result may be important information for pregnant woman and her family who consider MFPR. In term of at least one survival, women with keeping all three fetuses (EM group) have better outcome than women with sacrificing one fetus (ER group). Ninety-seven percent of at least one survival seems to be high. However, 94.8% of at least one survival was already reported in one study which was conducted from 1986 to 201313. Considering that current study was done more recently, better outcomes of the current study is not unexpected. Even though one third of fetuses are already sacrificed in ER group, lower rate of at least one survival in ER may have an effect on the attitude to MFPR.
What is interesting in this study is that the risk of neonatal sepsis was significantly higher in the ER group compared with the EM group even after multivariable analysis. This result suggests that ER itself might be a risk factor for sepsis. We think that there is a possibility that a clinical or subclinical inflammatory response to the dead fetal and placental tissue following embryo reduction might result in release of cytokines which may affect the survived fetus more fragile to septic condition.
The results on long-term neurodevelopmental outcomes are the major strength of the current study. Although prolonged gestation and increase in birth weight are also meaningful, more critical outcomes are long-term sequelae such as cerebral palsy (CP) and developmental delay. We showed rates of developmental delay and cerebral palsy which were not different between the EM group and ER group even though the rate of early preterm birth was significantly higher in the EM group. The incidence of CP in the triplet pregnancy was 28 to 44.8 / 1000, based on data from the 1980s and, 20-22 but it decreased to 18/1000 on the data from 1990s and early 2000s.23Actually, the risk of cerebral palsy is highest in neonates delivered at less than 28 weeks of gestation. It was recently reported that the rate of CP was 5.6% (21/381) in triplets or higher-order births of extremely low birth weight infants.24 Our study revealed much lower rates of CP as only 0.44% in EM group and 2.0% in ER group. Moreover, after introduction of MgSO4 for neuroprotection in threatened early preterm delivery,25,26 CP incidence will decrease further in the near future.
Our study had several limitations. First, this was a retrospective cohort study design. Second, we could not evaluate developmental delays using an identical method for all patients, due to the wide range of age at the time of examination.