The art and science of twins.

A joint meeting of the Section of Paediatrics and Child Health and the Forum on Maternity and the Newborn of the Royal Society of Medicine, Tuesday 20th March 2007.

Chair, morning session: Professor Lewis Spitz, President of the Section of Paediatrics & Child Health, RSM
Chair, afternoon session: Dr. Luke Zander, retired General Practitioner and teacher and founder of the Forum

Managing twin pregnancies.
Professor Kypros Nicolaides,(KN)The Harris Birthright Centre for Fetal Medicine, King's College Hospital, London

Twin pregnancies are either dichorionic or monochorionic. This distinction can and must be made accurately at 12 weeks pregnancy, because pregnancy is likely to proceed very differently according to the chorionicity.

The major cause of neonatal death is fetal abnormality; chromosomal abnormalities may be diagnosed at 12 weeks by chorionic villus sampling or at 16 weeks by amniocentesis. Improving the diagnosis rate to 98% could not justify the 1% loss of healthy babies due to the invasive procedures; reducing the number of invasive procedures still achieves 96% diagnosis while reducing miscarriages to .025%.

The risk of a 35 year-old woman having a Down's syndrome baby is 1 in 250; the risk is doubled in a DC twin pregnancy, and the same for a MC twin pregnancy at the age of 31. Accordingly 63% of twin pregnancies fulfil the age criteria for suspicion of Down's syndrome. There is of course no question of performing amniocentesis for each baby in all of these pregnancies. The detection rates for trisomy 21 using serum biochemistry in the second trimester are 65% for singleton pregnancies and 50% for twins whether DC or MC, so this is not a viable method. A high suspicion of Down's syndrome using the identification of increased nuchal translucency (NT) by ultrasound scanning is achieved in 80% of cases in both singletons and twins. Although in MC twin pregnancies both babies would have Down's syndrome, increased NT may be present in one but absent in the other; in such cases it is acceptable to use the average of the fluid collections for diagnostic purposes. The incidence of this sign in MC twins is higher because it is also a marker of twin-to-twin transfusion syndrome; here the recipient twin demonstrates the oedema of heart failure by increased NT.

It is characteristic of Down's syndrome that the nasal bone in is slow to ossify; temporary tricuspid regurgitation is another feature, as is abnormal flow in the ductus venosus, the vessel responsible for diverting oxygenated blood from the lungs to the left side of the heart and thus to the brain. In growth restricted babies the flow may be so severely impaired as to be reversed, a fatal condition. A final marker of Down's syndrome is the flat face.

Mothers who have biochemistry and ultrasound findings indicating a high risk of Down's syndrome are advised to undergo amniocentesis or chorionic villus sampling (CVS). In an indeterminate group examination of blood flow in the heart by doppler and further examination of the baby's face may lead to the invasive procedure and provide 90% accuracy in the diagnosis of this chromosome abnormality.

In a MC pregnancy CVS can prove Down's syndrome in both babies at 12 weeks, with a 1% risk of miscarriage, and a straightforward termination of pregnancy (TOP) if requested; in DC pregnancies, where one of the babies is likely to be affected, CVS is carried out at opposite ends of the placenta to ensure accurate diagnoses of both babies, and a subsequent selective termination carries a 5% risk of miscarriage. Double amniocentesis undertaken at 16 to 18 weeks and leading to embryo reduction at 18 to 20 weeks carries a 15% risk of miscarriage because of the substantial amount of dead tissue remaining in the uterus. But a more serious disadvantage of this is that it leads to a 20% risk of premature labour with a significant probability that the surviving twin is born with handicaps requiring much more attention than would a baby with Down's syndrome.

Abnormalities other than chromosomal.

The 1 per 1000 risk of spina bifida doubles to 1 per 500 in the case of DC twins, but selective termination of one of the pregnancies is quite possible. The risk of this abnormality increases to 4 per 500 in MC pregnancies, and here the options for selective termination are limited. The injection of potassium chloride into the heart of one baby would kill both unless the two placentas are separated; terminating the abnormal pregnancy by clamping the cord would leave a substantial amount of dead tissue in the uterus, with a 20% probability of premature delivery and a 25% risk that the healthy baby would die or suffer cerebral hypoxia due to retrograde haemorrhage into the placenta of the dead baby.

There is no such thing as twins; there are DC twins, and there are MC twins. Consider these increased risks in MC twin pregnancies: miscarriage from 11 to 23 weeks 10% (DC 2%); perinatal death after 23 weeks 3% (DC 1.5%) ; fetal growth restriction 30% (DC 20% ); preterm delivery before 32 weeks 10% (DC 5%); major defects 4% (DC 1%).

Discordant NT in an MC pregnancy, significant fluid in one twin but not in the other, may indicate a lethal situation, one of them having a major cardiac defect. Severe discordancy of this type may well precede twin-to-twin transfusion syndrome, so the finding requires close monitoring to identify signs of this serious problem. The only treatment with a chance of success is complete surgical division of the two placentas; if any connecting vessels remain and one of the babies dies, the stage is set for retrograde haemorrhage and brain damage of the surviving twin.

Please forget the definition of premature delivery as birth before 37 weeks; substitute 33 weeks. And the earlier the baby is born the more likely it is to be dead or damaged. So far none of the social or obstetric measures taken in the past 50 years to reduce the incidence of severe prematurity have been effective. The incidence for singletons is 1 to 2%, for twins 5 to 10%, and for triplets 25 to 30%; although measurement of the length of the cervix at 22 weeks provides a means of predicting premature birth, cerclage is not only useless but it increases the risk. However, it is now known that taking progesterone daily from the time that a short cervix is identified reduces the incidence by half in singletons; it has yet to be established whether it is also useful for multiple pregnancies. It is also possible to predict pre-eclampsia (PET) and fetal growth restriction severe enough to require urgent delivery before 33 weeks; doppler studies provide the information necessary, and by demonstrating reversal of flow in the ductus whether a baby, singleton or twin, is on the brink of death. We are now in a position to identify these complications by serial ultrasound scans in both MC and DC twins.


KN: Since a monozygotic twin pregnancy will be present in one DC pregnancy in seven, it is wise, indeed essential, to undertake CVS for both twins unless this presents great practical difficulty; this is so even when there is marked discordancy in NT, to avoid both the tragedy of missing chromosomal abnormality in both twins and the necessity of repeating an invasive diagnostic procedure at a later stage in a pregnancy.

In cases of twin-to-twin transfusion and the imminent death of both babies we divide the placentas using an intra-uterine endoscope and laser, without consideration of details of the vascular connections. The outcome of this procedure is about 20% for the death of both twins and 40% for loss of one twin or the survival of both. Septicaemia is a rare complication of this procedure for the mother in the case of the death of one twin.

Neuro-developmental outcomes in twins.
Professor Richard Cooke (RC), Professor of neonatal medicine, Liverpool Women's Hospital

"The problems twins face in intrauterine life and infancy are well known. But what of their long-term mental and physical development? Is this affected by these earlier problems, or indeed by the twin situation?" (Elizabeth Bryan 1983). Intrauterine growth restriction (IUGR), prematurity, chorionicity, and the mere presence of the other twin in early life may all be harmful to development.

Cognitive development may be impaired. Mean IQ is lower in twins, although within the normal range. At 11-15 years it is one third of a standard deviation below that of singletons (Husen 1963). Deficits in early life are associated with but not entirely due to IUGR (Drillien 1964), although there is no significant difference by six years (Wilson 1974). Similar small deficits have been shown in closely spaced singleton siblings, perhaps for lack of parental attention (Zazzo 1979). Lower verbal reasoning scores in the 11 plus examination appeared in twins, four points below singletons and triplets, eight points below singletons, and these could only be explained by differences in rearing (Record, McKeown & Edwards 1969). A more recent analysis of older data concluded that the deficits are due to differences in birth weight and gestation (Ronalds, de Stavola & Leon 2005), while another review has shown no differences between twins and their related siblings, even though the twins may have weighed almost a kilogram less at birth (Christensen et al. 2006). Surviving twins of very low birth weight have shown deficits unrelated to multiple gestation, but related to social class and chronic lung disease (Leonard et al. 1994). My own research in children aged 7 in school showed identically reduced IQ for twins and singletons (Cooke 2005). Having a clinically significant patent ductus at birth and the measurement of head circumference at age seven were the only independent predictors of IQ.

Neurological disability.

At 7% the rate of cerebral palsy (CP) for twins is about three times that for singletons (Benda 1952). Twins of like sex are more likely to have CP, particularly if a co-twin has died, where monochorionicity is the significant factor (Pharoah et al. 2002); and the smaller of twins is more likely to have CP, explained by the association with growth retardation (Russell 1961, Griffiths 1967). Whereas under 32 weeks gestation the rates for CP in twins and singletons are similar, the rate increases greatly towards term in twins (Alberman 1964). These observations were confirmed by Australian data (Peterson et al. 1993) and by a huge study over a decade in USA and Australia (Scher et al. 2002), showing a five times risk of fetal death and a four times risk of neonatal death in twins.

A literature review has shown that the risk of death for a co-twin after one twin has died is for monozygotic (MZ) twins 12%, dizygotic (DZ) 4%. The risk of neurological abnormality in the survivor is MZ 18%, DZ 1% (Ong et al. 2006). The "vanishing twin" hypothesis proposes that the undiagnosed loss of a twin could be responsible for much unexplained CP. It is known that MZ twins have a high mortality for one or both twins and that CP is common in a surviving twin. About 1 in 11 singleton births were originally twins, with the maximum loss of a twin occurring in the first trimester. The lost twin main go unnoticed, buried in the placenta of the survivor.

When compared with singletons the behavioural disorders of twins may be more common in those with neurological problems or poor growth, but are otherwise similar. Psychiatric disorders are not significantly more frequent in adult life.

References: To follow.


RC: If twins are to be followed up, this needs to be long-term. CP will appear at age 2 to 3, attention deficit disorder at 6-7, and psychological disorders in early adolescence. It would be ambitious to hope for effective prevention of these problems. MRI scans frequently show structural abnormalities in the brains of affected twins. But they do need good management - a good teacher in a small classroom.

The death of a twin.
Ms Barbara Read (BR), Counsellor, Neonatal Unit, St George's Hospital, London

Many more parents of twins (and especially higher multiple births) are likely to experience the death of one baby than those with singletons, chiefly, but not exclusively, because they are more likely to be born prematurely. The death of a twin in pregnancy, as a stillbirth or while on the neonatal unit, creates particular and complex issues for parents. Such loss is often underestimated, and the death of a twin is less likely to be acknowledged by professionals, as well as family and friends, than the death of a singleton. Grieving while simultaneously remaining hopeful for the surviving baby puts parents in a nigh impossible position.

After fourteen years working in the Multiple Births Foundation (MBF) and four years in my present post I am at last beginning to help staff of this neonatal unit to appreciate the special needs of these parents. As I move around within the unit I can get to know the babies and their parents, and provide support during what can be a rocky journey.

The sizeable unit where I work is running 32 cots at present, but this provides very limited knowledge and experience for nursing and medical staff faced with the deaths of twins – 8 in 2005, 3 in 2006. When I was at the MBF I learned how much parents appreciated good support when their babies had died, and how sad it was if the opportunity had been missed. Even lately I have met a mother who was upset when the obstetrician who had been involved with her IVF congratulated her on the birth of her baby and ignored the twin's stillbirth. To our shame Dr. Elizabeth Bryan, founder of the MBF, had described a similar case in her book 'Twins and the Family' more than twenty years ago. Twins are individual and separate.

It is this lack of acknowledgement of the dead twin (triplet, quadruplet) which grieves parents; it can even be apparent in the labour ward. Midwives now have the necessary skills to provide good support when a singleton dies, but such skills are not the rule when a twin has died. It is particularly important not to overburden parents with details of the care of the surviving twin when they are grieving.

Numbers of women hold on to their status as mothers of twins even when one has died, and it is distressing for them to meet other mothers whose twins are there in front of them. They may find themselves pretending that the other baby is at home, and may even punish the 'successful' mother by bluntly stating “he's dead”.

Now they fear for the fate of the surviving twin; many parents believe that the neonatal unit is a place for their babies to die, whereas of course the survival rate is good. Some seek advice as to how they can help the survivor him/herself not to suffer feelings of loss. It is important to provide parents with information about organizations, national or regional, which can help them along the way when they are ready.

The dilemma of ongoing anxiety combined with recent bereavement can result in a reluctance of mothers to return to the unit, where for example they may be needed to provide breast milk. We must understand and forgive this, and even recommend time being spent at home. Not only professionals but parents too are grateful to see the surviving baby acknowledged on the cot label and wristband as “Twin One”; this is not a singleton, but one of twins. Whenever possible the cot of a survivor should not be placed between twins, and parents must be made aware if this is unavoidable. In such ways parents can be helped to feel that their baby may be lost but is not forgotten.

Difficult situations cannot always be avoided. A bereaved mother encounters a member of her NCT group who has delivered healthy twins at term; how must she feel? Not like visiting the unit, to be sure. A twin dies in hospital; let us leave his name on the bottle of expressed breast milk and (until it is updated) on the unit's list of babies under its care, and talk to the parents. Understanding and consideration like this are greatly appreciated.

The collection of mementoes of a dead twin, the lock of hair, wrist bands, photographs, routinely made for singletons, is easily forgotten as work continues around the survivor. Parents should also be asked whether they would like a photograph of their twins together, as this is likely to be the only opportunity. Would they like to involve a chaplain? By being comfortable discussing a dead baby we can help parents to come to their decisions as to what they wish.

If your units do not use practices I have described here, please consider making changes; we must not still be letting parents down twenty years from now.

Suggested Reading
Bryan, E.M. Twins Triplets and More. St Martins Press. N.Y. (1992) Bryan, E.M & Hallett, F. Guideline for Professionals: Bereavement Sandbank, A. Twin and Triplet Psychology. Routledge (1999) Woodward, J. The Lone Twin – understanding twin bereavement and loss. Free Association Press N.Y. (1998) When a Twin Dies – Meetings held at the Multiple Births Foundation.

Useful Organisations
Twins And Multiple Births Association – bereavement support group.
The Lone Twin Network PO Box 5653, Birmingham B29 7JY
The Child Bereavement Charity Ashton House, West Wycombe, High Wycombe, Bucks. HP14 3AG
The Child Death Helpline 0800282986
Multiple Birth Foundation
SANDS: Stillbirth and Neonatal Death Society National Helpline 020 7436 5881


BR: Perhaps because they fear that they are tempting fate parents often prefer not to agree to the postmortem of a dead twin , in case the other twin dies; for the same reason a funeral is often deferred.

I am fortunate in being able to teach nursing and junior medical staff , although time for this is inadequate. Clearly there is a need for it.

The Lone Twin network exists to support surviving twins of any age, whether their twin died before or soon after birth or late in life. The Network List is the means by which lone twins can make contact with one another as soon as they join. Contact is by post only, at P.O. Box 5653, Birmingham B29 7JY. When I introduced mothers who had lost twins to older survivors it was the mothers who benefited.

Elizabeth Bryan: A difficulty arises when a baby who died in utero has not been named, which gives the surviving twin a problem in relating to him. Even more problematic is the distressing situation where the survivor has not been told soon enough that he is a twin; the MBF policy is to advise giving the survivor this information as early as feasible. BR: Also naming both babies helps the parents to distinguish them in their minds.