Selective termination by intrahepatic vein alcohol injection of a monochorionic twin pregnancy discordant for fetal abnormality

Authors

  • Mark L. Denbow,

    Research Fellow , Corresponding author
    1. Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School, Hammersmith and Queen Charlotte's and Chelsea Hospitals, London
    Search for more papers by this author
  • Malcolm R. Battin,

    Research Fellow
    1. Department of Paediatrics and Neonatal Medicine, Royal Postgraduate Medical School, Hammersmith and Queen Charlotte's and Chelsea Hospitals, London
    Search for more papers by this author
  • Phillipa M. Kyle,

    Senior Registrar
    1. Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School, Hammersmith and Queen Charlotte's and Chelsea Hospitals, London
    Search for more papers by this author
  • Roberto Fogliani,

    Visiting Registrar
    1. Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School, Hammersmith and Queen Charlotte's and Chelsea Hospitals, London
    Search for more papers by this author
  • Pamela Johnson,

    Senior Lecturer
    1. Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School, Hammersmith and Queen Charlotte's and Chelsea Hospitals, London
    Search for more papers by this author
  • Nicholas M. Fisk

    Professor
    1. Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School, Hammersmith and Queen Charlotte's and Chelsea Hospitals, London
    Search for more papers by this author

Correspondence: Dr M. L. Denbow, Centre for Fetal Care, Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School, Queen Charlotte's and Chelsea Hospital, Goldhawk Road, London W6 OXG, UK.

Case report

A 33 year old woman was referred with a twin pregnancy discordant for fetal abnormality. Initial ultrasound at 17 weeks of gestation confirmed, in one of the twins, a lumbo-sacral myelomeningocoele associated with an absent cerebellum, a lemon shaped head, and bilateral talipes. The other twin appeared normal.

The ultrasound appearances of concordant external genitalia, indeterminate septal thickness (1.7 mm), and absent ‘twin peak’ sign, were suggestive of monochorionic placentation. Placental vascular anastomoses between the twins, and therefore monochorionic placentation, were confirmed by injecting donor adult red cells into the intrahepatic vein of the normal twin, and then demonstrating their substantial intertwin passage on Kleihauer testing of fetal blood sampled from the abnormal twin (16% adult cells) after a delay of 30 min.

The parents received detailed counselling but remained adamant that they would not consider continuing the pregnancy. In the absence of alternative options, they requested that both twins be terminated. After discussing this at length, they consented to experimental selective fetocide which was performed at 18 weeks of gestation. One millilitre 100% alcohol (Martindale Pharmaceuticals, Essex UK) was injected under ultrasound control into the lumen of the abnormal twin's intrahepatic vein, resulting in immediate asystole. In the healthy twin, umbilical and cerebral arterial Doppler waveforms showed no evidence of haemo-dynamic instability on continuous monitoring over the first 15 min, intermittent monitoring over the next 90 min, and then again 24 h later.

The pregnancy progressed without incident, with normal ultrasonographic indices of growth, and no evidence of fetal cortical or renal damage. A healthy male infant was born vaginally at 40 weeks of gestation following an unremarkable labour. The birthweight was 3.020 kg, and Apgar scores were 9 at 1, and 9 at 5 min of age. The terminated twin was also delivered and sent, with the placenta, for pathological examination. Monochorionic placentation was confirmed; however, the severe autolysis precluded any detailed assessment of the fetus.

There were no immediate postnatal problems and routine physical examination and early cranial ultrasound were normal. At five weeks of age the infant was investigated with detailed neurological examination and cerebral magnetic resonance imaging (MRI). No neurological deficit was identified.

Discussion

Structural anomalies are increased by 50% in mono-zygous twin fetuses and are usually discordant. Whereas selective fetocide is widely practised in dichorionic twin pregnancies discordant for fetal abnormality, its use in monochorionic twin pregnancy discordant for abnormality has been followed in almost all cases by intrauterine death of the healthy twin1,2. Accordingly, there is general agreement that selective fetocide is contraindicated in monochorionic twins. Parental choice is thus restricted to either termination of the entire pregnancy including the normal twin, or continuation in the knowledge that one child is likely to suffer long term handicap.

The substantial risks involved for the surviving monochorionic fetus following its co-twin termination are thought to be mediated through vascular anastomoses within the chorionic plate. These anastomotic vessels are known to be present in most, if not all, monochorionic placentas. The mechanism of death or neurological damage in a surviving fetus is now considered to be due to its exsanguination into the terminated twin along these placental anastomoses3. Older theories of disseminated intravascular coagulation following the passage of thromboplastins from the dead fetus to its co-twin have now been largely discredited3. Hence, any safe method of selective termination of pregnancy would require the complete and immediate occlusion of the circulation. Attempts have been made to occlude the low flow to acardiac fetuses using laser ablation or ligation of the umbilical cord. However these techniques, although sometimes successful, involve invasive approaches such as hysterotomy or fetoscopy with their attendant risks. We and others have recently reported the successful use of an ultrasound-guided approach involving absolute alcohol injection in acardiac twins4,5. We now report the use of this technique in monochorionic twins discordant for fetal abnormality.

Absolute alcohol is a known vessel sclerosant and has been widely used to interrupt blood supply to renal, cerebral, and gastrointestinal lesions in adults without causing tissue damage remote from the target organ6. The immediate effect of rapid injection into the vessel is perivascular tissue toxicity leading to necrosis. The occlusion of the vessel is brought about by the damaged erythrocytes and denatured protein which clump together, and effectively plug the lumen. Injection into the intrahepatic vein of the fetus obviates the risk of inadvertent intra-arterial administration associated with injection into the umbilical vein at the insertion of the umbilical cord into the placenta. A further reason for preferring the intrahepatic vein for injection is that the operator can be entirely confident which twin is being terminated.

This technique has several advantages over other available options. A single 20 gauge needle insertion is considerably less traumatic to the fetal membranes compared with the smallest fetoscopes (2 mm) available. The procedure is technically comparable to routine intrahepatic vein blood sampling, and therefore no additional training or equipment is required. Finally, unlike fetoscopic techniques where general anaesthesia is often required, the procedure can be performed successfully using local anaesthesia.

The theoretical risk of passage of a significant volume of alcohol to the surviving twin is very small for several reasons. Only 1 mL of alcohol is used and it is expected that this amount becomes fixed at the site of injection. Furthermore, for any alcohol to reach the surviving twin it would have to have passed through the capillary bed of the terminated twin and its placenta in significant concentration to cause damage, and this is unlikely to occur.

We conclude that absolute alcohol injection into the intrahepatic vein may be used for selective fetocide in monochorionic twins discordant for fetal abnormality. Fetal cardiac output is relatively low at the stage of pregnancy reported here and later in pregnancy incomplete vascular occlusion may endanger the normal twin. Further studies are required before this technique can be recommended in the late second or third trimesters.

Ancillary