Interpretation
The lower than expected number of cases reflects the decrease seen in twin pregnancy perinatal mortality reported by MBRRACE-UK in 2013-20162. It is thought this decrease may be a consequence of improved care for MC twin pregnancies, particularly in recognising and treating complications unique to MC twins, and development of new treatment techniques including the Solomon Technique for FLA26. This is also linked to updated national and international guidance13, 15, 19 , which if implemented has been demonstrated, by Twins Trust, to lower adverse outcome rates27. However the latest MBRRACE-UK report published in October 2019 reported an increase in the stillbirth and neonatal death rates in twin pregnancies in 201722. It is important to consider data over a longer period alongside the use of three year rolling averages to better reflect trends in perinatal mortality2.
MC twin pregnancies complicated by sIUFD have a high risk of subsequent co-twin neurologic morbidity. From these data it appears that prenatal and indeed postnatal screening for abnormalities of the central nervous system in survivors is not routine in the UK. There is a need to strengthen professional guidance and practice amongst both obstetricians and neonatal paediatricians19.
This study strengthens the argument made in our 2019 systematic review and meta-analysis for the need for a purposely designed prospective study, ideally with antenatal and postnatal imaging and long-term follow-up 7.
There is no specific guidance regarding when and how co-twins after sIUFD should be delivered. Of note in this study there was a high rate of preterm birth (both spontaneous and iatrogenic) and a high rate of emergency caesarean sections. The majority of co-twin survivors were admitted to the neonatal unit with prematurity the commonest indication. Despite the high risk nature of the pregnancy and requirement for admission to NNU, very few babies had planned follow-up despite an association with adverse long-term outcomes10.
The uptake of post-mortem examination was low (25.4%) for the initial sIUFD but increased to 40% if the co-twin died as well. This is below the 75% recommended uptake by the RCOG28, and may reflect that parents accept that MC twin pregnancies are higher risk, and even if a cause was not apparent antenatally, the findings of the post-mortem are unlikely to affect a subsequent pregnancy as it may be linked to monochorionicity. Despite post-mortem being considered the most useful investigation for parents to find out why their baby died29, in 6/9 post-mortems in which the UKOSS reporter knew the findings, the post-mortem was inconclusive. This highlights another area of future research, as there is not currently a specific classification system for cause of death in MC twins, which is often different to the cause of death in singletons, and is the classification system which pathologists have to currently use. Since performing this study, we have proposed a new classification system of causes of death in twin pregnancies (CoDiT) which requires further validation30. These findings also raise the consideration of whether specialist perinatal pathologists are needed for MC twin pregnancy post-mortems, and whether injection studies should be performed in all MC twin pregnancies to aid determining cause of death.