One in Five Surviving Twins Suffers Severe Brain Injury After Their Twin Dies in the Womb

When one twin dies in a shared placenta pregnancy, the surviving twin faces a hidden danger: severe brain injury. The largest study ever conducted on this complication now shows that this risk is far from rare. In an international retrospective cohort spanning thirteen fetal therapy centres in ten countries, severe brain injury was detected in 18 percent (72 of 395) of initially surviving cotwins after a single fetal demise in monochorionic twin pregnancies. The findings, published in eClinicalMedicine, quantify for the first time at this scale the neurological toll of an event that obstetricians have long feared but could only estimate from small, fragmented studies.

Monochorionic twins share a single placenta, and with it a network of blood vessels that connects their circulations. This shared circulation is the source of well-known complications such as twin–twin transfusion syndrome, twin anaemia-polycythaemia sequence, and selective fetal growth restriction. It also sets the stage for catastrophe when one twin dies: blood can drain acutely from the survivor through placental vascular anastomoses into the deceased twin, producing hypovolaemic shock, fetal anaemia, and hypoxic-ischaemic brain injury. In unselected first-trimester monochorionic cohorts, single fetal demise occurs in roughly 6 to 8 percent of pregnancies, meaning a substantial number of families encounter this scenario each year.

Previous evidence suggested that severe brain injury affected somewhere between 20 and 27 percent of surviving cotwins, but those figures came from small observational cohorts, often from single centres, with inconsistent injury definitions and non-standardised imaging protocols. The new study addressed that gap by pooling consecutive cases from centres in Argentina, Belgium, Canada, Germany, Israel, Italy, the Netherlands, Spain, Sweden, and the United Kingdom, recruited between January 2000 and January 2025. The researchers excluded pregnancies in which the demise followed selective feticide, twin anaemia-polycythaemia sequence, fetoscopic laser surgery, or the twin reversed arterial perfusion sequence, in order to isolate a population sharing a comparable underlying mechanism of acute exsanguination through bidirectional anastomoses.

Of 445 eligible pregnancies, 50 were excluded for lacking antenatal or postnatal neuroimaging, leaving 395 for analysis. The cohort comprised 323 liveborn infants, 38 cases in which the initially surviving cotwin also died, and 34 terminations of pregnancy, three quarters of which were performed because imaging suggested fetal brain injury. Antenatal neuroimaging after the demise was nearly universal, performed in 97 percent of pregnancies, including fetal ultrasound in 95 percent and fetal MRI in 58 percent. Postnatal imaging, by contrast, was performed in only half of the liveborn neonates, and its availability varied dramatically across centres and gestational-age categories.

The pattern of injury was strikingly consistent with the proposed mechanism. The presumed origin was antenatal in 82 percent of the 72 severe cases, perinatal in 5 percent, postnatal and prematurity-related in 10 percent, and indeterminate in 3 percent. The most common type was multicystic encephalomalacia, found in 32 percent of cases, followed by arterial ischaemic infarction in 20 percent, cystic periventricular leukomalacia in 14 percent, and high-grade intraventricular haemorrhage in 11 percent. Migration disorders and severe ventriculomegaly or volume loss each accounted for 8 percent. These imaging signatures, the authors note, fit hypoxic-ischaemic damage following acute intertwin exsanguination and hypovolaemic shock.

The statistical analysis identified two independent factors associated with severe brain injury. Later gestational age at the time of the fetal demise carried an adjusted odds ratio of 1.09 per week, and signs of fetal anaemia, defined as a middle cerebral artery peak systolic velocity of at least 1.5 multiples of the median within two weeks after the demise, carried an adjusted odds ratio of 2.29. Both associations held in sensitivity analyses restricted to injuries of antenatal origin and in delta-adjusted models testing missing-not-at-random assumptions for the anaemia data. The gestational-age finding has a plausible mechanical explanation: as pregnancy advances, the superficial arterio-arterial and veno-venous anastomoses grow wider, lowering vascular resistance and permitting larger volume shifts when acute exsanguination occurs.

The timing of birth emerged as a separate and more delicate issue. Among liveborn infants, an exploratory analysis showed an association between earlier gestational age at birth and severe brain injury, and neonatal morbidity was markedly more frequent in infants born less than four weeks after the demise. Severe brain injury occurred in 33 percent of liveborn infants delivered within one week of the demise, compared with much lower rates among those carried longer. Yet the authors urge strong caution in interpreting this pattern: most severe brain injury was already antenatal in origin, and early delivery was frequently prompted by fetal distress, fetal anaemia, or concern about further exsanguination. The association may therefore reflect reverse causation and confounding by indication rather than a harmful effect of prolonged expectant management.

Clinical practice around the world varies enormously, and the study documents this heterogeneity in detail. The median interval between the fetal demise and live birth ranged from 2.6 to 17.4 weeks across countries. Professional guidance diverges as well: ISUOG and FIGO advise against immediate delivery in preterm monochorionic twins after a single demise, with ISUOG suggesting delivery at 34 to 36 weeks after steroid administration, while the RCOG recommends aiming for term with individualised timing, and SMFM and ACOG offer no specific guidance at all. Treatment of fetal anaemia was equally inconsistent, with the proportion of anaemic survivors receiving intrauterine transfusion ranging from 0 to 67 percent across centres, and no apparent difference in brain injury rates between transfused and expectantly managed fetuses in this retrospective data.

The study’s diagnostic findings carry practical weight. Severe brain injury was first detected by fetal neurosonography in 35 percent of cases and by fetal MRI in another 35 percent, while neonatal cranial ultrasound and neonatal MRI made the first diagnosis in 26 and 4 percent respectively. Notably, in 3 of 59 cases of antenatal injury, fetal neurosonography appeared normal and the damage was only revealed on fetal MRI, underscoring the complementary value of advanced prenatal imaging. Because one third of abnormalities were first identified on postnatal imaging, despite postnatal scans being performed in only half of liveborn infants, the authors argue that routine postnatal neuroimaging after a single fetal demise is equally essential, and that the true prevalence of postnatally acquired injury may currently be underestimated.

The investigators acknowledge the limits of their retrospective design. Differential availability of imaging may have introduced selection and ascertainment bias, clinical management and documentation were not standardised across centres or over the 25-year inclusion window, formal independent dual review of imaging was not performed, and the absence of pre-demise neuroimaging limits definitive attribution of migration disorders and ventriculomegaly to the demise event itself. Monoamniotic pregnancies, which may carry additional risks from cord entanglement, comprised only 9 percent of the cohort and could not be analysed separately. Even so, the scale and harmonised definitions of this cohort give the clearest picture yet of a rare but devastating complication. The authors conclude that routine serial antenatal and postnatal neuroimaging should follow every single fetal demise in a monochorionic pregnancy, and that in clinically stable pregnancies the demise alone should not prompt early delivery, since prematurity adds its own burden of brain injury to survivors who have already weathered an acute circulatory catastrophe.

Subject of Research: Severe brain injury in surviving cotwins after single fetal demise in monochorionic twin pregnancies

Article Title: Brain injury after single fetal demise in monochorionic twin pregnancies: an international multicentre retrospective cohort study

Article References: Rondagh, M., Lewi, L., Aertsen, M., Russo, F., Lanna, M., Casati, D., Righini, A., Shinar, S., Miller, E., Carmant, L. S., Khalil, A., Sorrenti, S., Mohammed, D., Herling, L., Bohlin, K., Yinon, Y., Komem, D. A., Sanin-Ramirez, D., Roca, P., … Torracca, F. (2026). Brain injury after single fetal demise in monochorionic twin pregnancies: an international multicentre retrospective cohort study. eClinicalMedicine, 101, Article 104290. https://doi.org/10.1016/j.eclinm.2026.104290

Image Credits: AI Generated

DOI: 10.1016/j.eclinm.2026.104290

Keywords: monochorionic twins, single fetal demise, brain injury, hypoxic-ischaemic injury, fetal anaemia, placental vascular anastomoses, retrospective cohort study, neuroimaging, intrauterine transfusion, preterm birth, twin-twin transfusion syndrome, fetal MRI