Elsevier

The Lancet

Volume 356, Issue 9236, 30 September 2000, Pages 1143-1147
The Lancet

Articles
Outcome after prenatal diagnosis of hypoplastic left-heart syndrome: a case series

https://doi.org/10.1016/S0140-6736(00)02756-2Get rights and content

Summary

Background

Traditionally, after prenatal diagnosis of hypoplastic left-heart syndrome (HLHS) couples have been offered termination of pregnancy or comfort care. Success of postnatal surgical options such as the Norwood procedure have been associated with survival of up to 60%. Whether survival is affected by the congenital anomaly being identified prenatally or postnatally remains uncertain.

Methods

We reviewed all cases of prenatally diagnosed HLHS referred to the Fetal Medicine Unit at Birmingham Women's Hospital over 6 years between 1994 and 1999.

Findings

87 cases of HLHS were referred at a median gestational age (95% CI) of 23 (19–37) weeks. Of these, 53 (61%) chose prenatal karyotyping. The overall frequency of abnormal karyotype was found in seven of 59 cases (12%) and associated structural anomalies in 18 of 87 (21%). After counselling, 38 of 87 couples (44%) chose termination of pregnancy. Of the remaining 49 fetuses, 11 (23%) were not considered for postnatal surgery because of parental choice and they died after compassionate care. Of the 36 babies who had surgery postnatally, 12 survived (33%). We recorded a survival rate of 38% for the stage-1 Norwood procedure in the prenatally diagnosed HLHS in our centre. These data suggest that at the point of prenatal detection, the overall survival rate for fetuses with HLHS is 25% (if terminated pregnancies are excluded).

Interpretation

Fetal echocardiography allows early diagnosis of HLHS and gives clinicians the opportunity to triage this group dependent on prenatal findings, including karyotyping and the exclusion of other structural anomalies. These prospective data provide up-to-date information on the basis of which parents can make decisions.

Introduction

Hypoplastic left-heart syndrome (HLHS) makes up 1% of all congenital heart disease but is responsible for 25% of all cardiac deaths in the first week of life.1 Hypoplasia of the left ventricle may also be associated with aortic valve atresia or stenosis, mitral valve atresia or stenosis, and hypoplasia of the aortic arch.2 It is recognised that critical aortic stenosis could develop into HLHS.3 Before the era of prenatal diagnosis, this disorder led to unexpected, rapid deterioration in neonatal condition within the first few days of life, once the ductus arteriosus closed.2 However, HLHS is potentially detectable at about 20 weeks' gestation by routine ultrasound visualising a fourchamber view of the fetal heart.4 Although the sensitivity and specificity for detecting congenital heart disease varies in the UK, there is evidence that with adequate training in fetal echocardiography serious disease could be detected.5 The introduction of prenatal diagnosis has led to a lower frequency of live babies born with HLHS, as a result of termination of pregnancy.6, 7 The alternative to termination of pregnancy in the UK was terminal supportive comfort care, with few cases being presented to cardiac surgeons.8

Over the past 20 years, postnatal surgical treatment options have become gradually available.7 The Norwood staged palliation procedure (and modifications) for HLHS has been introduced into paediatric care, with good results, so that parents are increasingly choosing to continue with an affected pregnancy.9, 10 Such palliation has been offered in our centre, and has achieved early survival in up to 59% of babies with both prenatal and postnatally diagnosed HLHS. Most of these series consisted of babies in whom the anomaly was detected postnatally.9, 10 Since 1994, the proportion of babies with a prenatal diagnosis of HLHS referred to our centre has increased strikingly. It is difficult for health-care professionals dealing with prenatally diagnosed cases to have adequate data with which to counsel parents, when most of the data available correspond to postnatally detected cases of HLHS. Other investigators have suggested that prenatally diagnosed congenital anomalies which are surgically correctable (eg, diaphragmatic hernia) have a poorer prognosis when viewed from inutero diagnosis than by the paediatric surgeons, once the baby is born.

We aimed to review prenatal associations and outcomes of all babies referred for prenatal assessment to the Fetal Medicine Centre at Birmingham Women's Hospital between 1994 and 1999.

The ultrasonography department records and the West Midlands Congenital Anomaly register were searched to find all cases of suspected HLHS referred to the Fetal Medicine Centre between January, 1994, and December, 1999. All fetuses underwent a detailed ultrasound scan, including a fetal echocardiographical examination by a paediatric cardiologist (JGW or OS). HLHS was the only cardiac lesion evident by prenatal, postnatal echocardiography, or pathological examination. Our criteria for inclusion of a fetus with HLHS were the presence of situs solitus with normally related great vessels and isolated hypoplasia of the left ventricle. This anomaly in some cases may have been associated with aortic valve atresia, mitral valve obstruction, and hypoplasia of the aortic arch. Hypoplastic left-hearts with additional complex features (eg, transposition of the great arteries, double outlet right ventricle, atrioventricular septal defect) were excluded (29 cases over the 6 years).

When the diagnosis of HLHS was confirmed, a fetal medicine specialist, the paediatric cardiologist, and a paediatric cardiology counselling-nurse discussed the options with the parents. The parents were informed that a palliative Norwood staged procedure might be possible but that associated chromosome and extracardiac malformations would significantly worsen the prognosis for survival.11 We recommended to all couples that prenatal karyotyping be done, including exclusion of a 22q deletion (Di George syndrome region), and that a further ultrasound examination at 28 and 34 weeks should be arranged. If the parents opted to have a termination of pregnancy (n=38), necropsy of the fetus was advised. If the parents were considering postnatal surgery as an option (n=37), they were usually advised to deliver at the Birmingham Women's Hospital for planning of neonatal intensive care and to be closer to the Birmingham Children's Hospital where surgery would be undertaken (n=32). After discussion and assessment in Birmingham, three couples chose delivery and surgery in their own region. To plan delivery, induction of labour was carried out between 38 and 39 weeks' gestation to coincide with neonatal intensive care bed availability. In 11 pregnancies, parents did not proactively opt for termination of pregnancy or postnatal surgery. In these cases, outcome would be fatal and comfort/compassionate care was given to the babies at birth.

Demographical details were obtained from referral letters, ultrasound records, and hospital notes. Follow-up data were obtained from paediatric notes and the West Midlands Congenital Anomaly Register. If the patient delivered at a different hospital, data were requested from the relevant obstetric consultant. Cytogenetic and pathology details were also obtained. By use of statistics on a Microsoft Excel (Version 4.0) database, data are presented as median (95% CI). Minitab (Version 12.0) was used to construct a Kaplan-Meier survival curve.

87 cases of HLHS referred during the 6-year period were reviewed. 85 (98%) were singleton pregnancies and in two cases (2%) there was an affected dichorionic twin. In two cases (2%) the diagnosis of HLHS was made by the fetal medicine specialist, but not confirmed by a paediatric cardiologist because the couples chose termination of pregnancy due to the presence of additional structural malformations/chromosomal abnormality. In both cases permission to do a necropsy of the aborted fetus was declined.

Overall, the median (95% CI) age of the mother at referral was 27 years (19–37). The median gestational age at referral was 23 weeks (19–37). Nine (10%) were local mothers booked at the Birmingham Women's Hospital, 52 (60%) were referred from other hospitals in the West Midlands region, and 26 (30%) were supraregional referrals, including referrals from Scotland, Ireland, and Germany.

The sex of the baby was known in 82 of 87 cases. 48 (58%) were male and 34 (42%) were female. Only 53 of 87 parents (61%) wanted prenatal karyotyping. In 28 babies, the karyotype was unknown (but six of these were alive postsurgery). The overall frequency of abnormal karyotype was 7 of 59 cases (12%). However, many of these karyotypic anomalies were confined to fetuses that had structural anomalies. Only four of 71 cases (6%) had a karyotypic anomaly if the fetus had an isolated heart lesion at presentation. Table 1, Table 2 show the exact chromosomal abnormalities and outcomes. Only chromosome deletions and translocations were associated with confirmed isolated HLHS (two cases, who had no associated structural anomaly at ultrasonography or necropsy, or both). None of the babies with abnormal karyotype survived. One neonate died after a stage-1 Norwood procedure. A cordocentesis undertaken at the referral hospital produced a karyotype of 46XX, but it is now believed that the sample was maternal blood. Amniocentesis (undertaken at the same time as the cordocentesis) and neonatal blood samples subsequently showed a microdeletion of chromosome 5.

There were extracardiac abnormalities associated with 18 (21%) of 87 cases of HLHS, including three fetuses with hydrops and one with a pericardial effusion on ultrasound scan with other abnormalities recorded at necropsy (table 2). Only one of these babies survived, with most couples opting for termination of pregnancy or terminal supportive care. Three of these cases had chromosomal abnormalities (two were trisomy 13 and one was 45XO). In two further cases despite a normal ultrasound examination at presentation, both were found to have chromosome anomalies, and in these cases structural abnormalities were noted at necropsy.

Figure 1 shows a Kaplan-Meier survival curve for all fetuses in our cohort with a prenatal dignosis of HLHS, and a modified curve excluding those cases in which the parents chose terminations of pregnancy. Survival time was based on gestational age calculated from early ultrasound findings. Figure 2 summarises outcomes of babies with an in-utero diagnosis of HLHS. Just under half the parents chose termination of the pregnancy in the second trimester. Of the pregnancies that continued only one case of in-utero death occurred (Turner's syndrome complicated by hydrops fetalis). In about a fifth of the pregnancies that continued, parents chose compassionate, comfort care instead of surgery. Parents usually made this decision in the prenatal period, by personal choice, after identification of a co-existent chromosomal anomaly, adverse features on ultrasound (such as hydrops, or substantial tricuspid regurgitation), or extracardiac structural abnormalities. One fetus had undergone a failed in-utero aortic valvotomy (at another tertiary, fetal cardiology centre) for congenital aortic atresia and a secondary hypoplastic left ventricle (noted on progressive ultrasound examinations). The procedure was undertaken because of development of fetal hydrops at 29 weeks' gestation. The mother developed chorioamnionitis shortly after the procedure and required an emergency caesarean section. The preterm baby died within a few minutes of birth. Ten of 11 fetuses died in the neonatal period: seven in the first week of life, and one baby with an associated congenital anomaly (a Dandy-Walker variant) died aged 10 months.

In the remaining pregnancies there was intention to treat the baby by a modified Norwood operation in the postnatal period by both parents and health-care professionals. One baby died in the first hour of life after being delivered after a massive placental abruption. The other babies were operated on after delivery. Most of these neonates underwent a stage-1 modified Norwood procedure at Birmingham Children's Hospital done by two consultant cardiac surgeons (WJB and BS). One baby had an emergency aortic valvotomy because of critical aortic stenosis and a hypoplastic left ventricle. The parents chose for their baby not to have a modified Norwood procedure and this baby died postprocedure.

The remaining three couples were referred to Birmingham for fetal assessment and counselling but they chose to deliver and have surgery locally. A modified stage-1 Norwood procedure was carried out in all three babies at another regional centre, because of parental wishes. One fetus was recognised to have severe hypoplasia of the ascending aorta, confirmed on postnatal echocardiography. The parents still requested surgery, acknowledging the possible higher risks of the procedure. One child is alive and two died postsurgery. Of the 32 cases that underwent modified stage-1 Norwood procedures at Birmingham Children's Hospital, 12 (38%) survived this procedure.

Overall, from the time of diagnosis of HLHS, only 48 of 87 (55%) survived to birth. Of these 48 babies, 11 (23%) died after compassionate care, 12 (25%) were alive after surgery and 24 (50%) died after surgery. One baby died after an acute placental abruption (2%). If the complete cohort of 87 HLHS babies is studied, only 14% are still alive.

Of the 48 continuing pregnancies, the median gestation at delivery was 38 (35·4–40) weeks. Preterm delivery was reported in eight of 48 cases (17%). In five cases (10%) there was spontaneous preterm labour (including one twin pregnancy at 36 weeks), and in three cases the preterm delivery was iatrogenic because of maternal or fetal reasons, including suspected chorioamnionitis, fetal hydrops, and fetal bradycardia. The median (95% CI) birthweight for babies born at term was 3·09 kg (2·29–3·51). Three babies born at term had a low birthweight (less than 2·5 kg), including one twin, one baby with an abnormal karyotype, and one baby with additional structural anomalies (hemivertebrae and tracheo-oesophageal fistula). In terms of mode of delivery, there were 36 normal vaginal deliveries, two ventouse extractions, five elective caesarean sections, and four emergency caesarean sections. In one case, mode of delivery was unknown.

At the point of prenatal diagnosis potential survival for a fetus with HLHS was only 14%. Further examination of the fetus by detailed ultrasonography and karyotyping to exclude other abnormalities, which if present will usually lead the parents to chose termination of pregnancy, produced a cohort of intention-to-treat cases with a survival rate of 25%. However, once the baby is presented to the cardiothoracic surgeon for postnatal surgery, survival in our cohort was 33% (38% if operated on at Birmingham Children's Hospital).

Section snippets

Discussion

Although there is no universal agreement on whether routine ultrasound scanning can alter outcome, most pregnant women in the UK have at least one ultrasound examination of the fetus, mainly between 18 and 22 weeks.12 With adequate training and the examination of the four chamber view of the heart during this scan, detection rates for congenital heart disease rise above the national average.5 Data also indicate that HLHS is increasingly likely to be diagnosed in utero.13 Our findings might

References (25)

Cited by (82)

  • Prenatal diagnosis of hypoplastic left heart syndrome in current era

    2011, American Journal of Cardiology
    Citation Excerpt :

    Preoperative morbidity, including right ventricular dysfunction, the need for, and duration of, mechanical ventilation before surgery, multiorgan dysfunction, and the need for inotropic support, was lower in our preDx group. These findings are consistent with both our previous experience4 and that reported in other series.2,4,5,11,13 However, our preDx group had a lower median gestational age at delivery and lower average BW than the postDx group.

View all citing articles on Scopus
View full text