Article Text

Download PDFPDF

Original article
A cohort study of women with a Fontan circulation undergoing preconception counselling
  1. Matthew Cauldwell1,
  2. Kate Von Klemperer2,
  3. Anselm Uebing2,
  4. Lorna Swan2,
  5. Philip J Steer1,
  6. Sonya V Babu-Narayan2,
  7. Michael A Gatzoulis2,
  8. Mark R Johnson1
  1. 1Academic Department of Obstetrics and Gynaecology, Imperial College London, Chelsea and Westminster Hospital, London, UK
  2. 2Adult Congenital Heart Centre, The National Heart and Lung Institute, Imperial College London, Royal Brompton Hospital, London, UK
  1. Correspondence to Professor Mark R Johnson, Academic Department of Obstetrics and Gynaecology, Imperial College London, Chelsea and Westminster Hospital, 369 Fulham Road, London SW10 9NH, UK. Mark.johnson{at}


Objective To describe outcomes for women with a Fontan circulation attending preconception counselling (PCC). Exploring (1) impact of counselling on decision to become pregnant; (2) pregnancy rates in those opting for pregnancy and (3) short-term and long-term outcome of those who succeeded in becoming pregnant.

Methods Retrospective review of women aged 16–45 years with a Fontan circulation from 1994 to 2014.

Results 58 women were offered PCC, 3 declined and 55 received PCC. Following PCC, 15 opted against conception, 16 decided to delay pregnancy, 19 attempted pregnancy and 5 were lost to follow-up. Of the 19 women, 14 succeeded, becoming pregnant a total of 43 times (median 1, range 1–9). Of these, 6 miscarried all pregnancies. 8 carried 14 pregnancies to viability. Baseline hypoxaemia and cardiac disease in pregnancy (CARPREG) score was similar in those opting for and against pregnancy, but CARPREG score was better in those delaying conception. Women exclusively miscarrying or unable to conceive were more likely to have baseline hypoxaemia and greater CARPREG score. Cardiac complications included arrhythmia requiring treatment (n=4) and one thromboembolism. Obstetric complications were greater in women with a Fontan circulation, 10 preterm births (<37 weeks) and 8 small for gestational age babies (<10th centile). There was one neonatal death. At follow-up, there was no deterioration in clinical status as determined by echo.

Conclusions Most women accept PCC and decided to pursue pregnancy; in some cases, this was despite being advised of a poor prognosis. Pregnancy outcome was related to baseline hypoxia and CARPREG scores.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Preconception counselling (PCC) is a vital part of the management of women with pre-existing heart disease.1 Their overall mortality is high, in the region of 1%, or 100 times the rate in the background population.2 Women with a Fontan circulation fall into the high-risk category of most risk-assessment scores.3 ,4 Women with a Fontan circulation have a univentricular physiology, which has a limited capacity to meet the increased haemodynamic demands of pregnancy.5 The flow of systemic venous blood into the pulmonary circulation is maintained by the central venous pressure and augmented by the negative intrathoracic pressure induced by normal breathing. Patients with a Fontan circulation have an increased risk of atrial tachycardia, thromboembolism, cardiac failure, liver failure and protein losing enteropathy. Data published in 2014 suggest a 25-year mortality of almost 25%.6 However, the information regarding management and outcomes of pregnancy for patients with a Fontan circulation is limited (table 1). Maternal mortality is uncommon, but atrial arrhythmias are frequent and pregnancy maybe associated with a temporary decline in functional status.7 Despite the absence of good data, some experts have advised against pregnancy, citing an unacceptable risk of mortality and morbidity during pregnancy and uncertainty about the long-term impact of pregnancy, others are more optimistic.7 ,8 Fertility is reported to be low in patients with a Fontan circulation, with increased rates of amenorrhoea and early miscarriage.9–11 In this study, we investigated the impact of PCC in a group of women with a Fontan circulation, documenting the number choosing not to attempt to conceive and the outcome in those who tried to become pregnant. We have related the baseline cardiac status to the decisions made after PCC, the impact on pregnancy rates and the ability to carry the pregnancy to viability. Finally, we have assessed whether or not pregnancy accelerates the rate of clinical deterioration.

Table 1

Summary of literature published on Fontan and pregnancy


Chelsea and Westminster Hospital (C&W) and Royal Brompton Hospital (RBH) have offered a combined cardiac pregnancy service since 1994. PCC is for all women considering a pregnancy regardless of the severity of their disease (now or in the future) and is provided jointly by a cardiologist and an obstetrician. PCC focuses on what is known about the impact of pregnancy on a Fontan circulation and the known obstetric complications such a high rate of miscarriage, preterm delivery and the likelihood of babies being born small for gestational age. Women with a Fontan circulation (aged between 16 and 45 years who attended PCC were identified from the electronic patient records. After PCC, the choices made by patients were recorded: (1) avoid pregnancy, (2) to attempt to become pregnant or (3) to delay attempting to become pregnant. Patients receiving PCC but who were lost to follow-up, or who moved out of our area were excluded from the final analysis. Electronic records and case notes were reviewed from 1994 to establish health status at initial presentation, and subsequent health and pregnancy outcomes until December 2014.

Outcomes were grouped into maternal, neonatal or obstetric events. Maternal cardiovascular complications included deterioration in functional class, documented episodes of arrhythmias, hospitalisation during pregnancy for cardiac symptoms and any documented episodes of thromboembolism. Cardiac disease in pregnancy (CARPREG) scores were calculated as they have been validated in calculating maternal cardiovascular risk of complication though to date there is limited experience in their application to patients with a Fontan circulation.

Echocardiography grading of ventricular function was agreed by two expert congenital cardiologists and this was available for all patients due to the known relevance to CARPREG scores.

Obstetric complications were defined as pregnancy-induced hypertension (hypertension occurring after >20 weeks of gestation and documented blood pressure >140 mm Hg systolic or 90 mm Hg diastolic without proteinuria); pre-eclampsia (as PIH but with additional 0.3 g of proteinuria/24 h urine sample); eclampsia (pre-eclampsia with grand mal seizures); HELLP Syndrome (biochemical evidence of haemolysis, raised liver enzymes, low platelets); thromboembolic complications; admission to the High Dependency Unit or Intensive Therapy Unit; preterm birth (birth prior to 37 completed weeks of gestation; obstetric haemorrhage (blood loss at delivery ≥500 mL for vaginal delivery and ≥1000 mLs for Caesarean section). We also included women who had miscarriages (first or second trimester losses before 24 completed weeks) and/or recorded any recurrent episodes of bleeding in early pregnancy despite no obvious cause being found as threatened miscarriage.

Neonatal complications recorded included preterm birth (delivery prior to 37 completed weeks); smallness for gestational age (SGA) (birth weight <10th centile); neonatal death (death within 1 month of delivery) and the presence of congenital heart disease.

For each pregnancy in a patient with a Fontan circulation, we used two normal controls: the woman who delivered immediately before and the woman who delivered immediately after each index case. This controlled for changes in demographics and practice over time.

Data were analysed using SPSS V.22 for windows. Differences in pregnancy outcomes between the Fontan and control groups were evaluated. Continuous variables were compared with the Student’s t test if data were normally distributed, and with the Mann–Whitney U test if they were not. The χ2 test (or Fisher's exact test if cell numbers were <5) was used to test differences between frequencies of occurrence. All tests were two-tailed and p<0.05 was considered statistically significant.


Fifty-eight women of childbearing age (16–45 years) with a Fontan circulation were offered PCC (including one woman with significantly impaired ventricular function) (figure 1). Three women declined or failed to attend PCC. Five women who received PCC, but who subsequently moved out of area or failed to attend repeated follow-up, were excluded, leaving a total of 50 women (table 2). The underlying anatomy, type of Fontan procedure and systemic ventricular morphology of women who attended PCC are included in table 2. During or following PCC, 19 women decided to try to conceive, 15 women decided not to pursue pregnancy (including a patient who was diagnosed with primary ovarian failure) and 16 women confirmed that they wished to consider pregnancy in the future. Information was collected on smoking status from all women, only three women in the cohort admitted to current tobacco use (<10 cigarettes per day). All women were provided with contraceptive advice.

Table 2

Clinical characteristics of all patients offered preconception counselling (PCC)†

Figure 1

Flowchart illustrating patients referred for preconception counselling (PCC).

The 19 women who chose to conceive were of a similar age (mean 29.4 years) at PCC as the 15 who chose not to consider pregnancy (mean 29.5 years). The New York Heart Association (NYHA) and CARPREG scores, rates of hypoxia (8/19 vs 8/15) and history of arrhythmias (12/19 vs 8/15) were similar in the two groups (table 2). In contrast, women deferring a decision to try and conceive were younger (mean 23.6 years), had similar NYHA, but better CARPREG scores (p<0.01) and less of them had significant hypoxaemia 4/12 vs 16/34 (p=0.07) than those coming to a decision either to try and conceive or not; similar numbers had a history of arrhythmia and thromboembolic disease (table 2).

The group opting to try and conceive was free of other co-morbidities, whereas the group opting to avoid pregnancy included one woman with a history of right atria thrombus, one with a benign brain tumour causing epilepsy and another with liver cirrhosis due to hepatitis C.

Compared with the women who conceived and carried at least one pregnancy to viability (n=8), the women who failed to conceive or conceived and only had miscarriages (n=11) were of a similar age (26.1 vs 28.6 years; table 3). NYHA scores were similar, but CARPREG scores were better in the group with at least one successful pregnancy (p<0.05) and rates of hypoxia lower (1/8 vs 8/11; p<0.05; table 3 suggesting CARPEG scores can be a risk assessment tool prior to pregnancy).12 The rates of arrhythmia and thromboembolic events were similar in both groups (table 3). Two women were diagnosed with unexplained infertility and underwent IVF unsuccessfully. Subsequently, one spontaneously conceived trichorionic, triamniotic triplet pregnancy, but miscarried in the first trimester.

Table 3

Characteristics of women who embarked upon pregnancy during study period

Of the six women who conceived and only had miscarriages (a total of 19), one woman had nine miscarriages, two women had three miscarriages, one woman had two miscarriages and the other two women had one each. The median age at first miscarriage was 30 (range 28–34). The patient who had nine miscarriages had tricuspid atresia and an atriopulmonary Fontan with multiple collaterals; she was hypoxaemic, with O2 saturations 85% at rest, had an NYHA score of 1, had a CARPREG score of 2 and had undergone an ablation. Overall, four of the six patients who repeatedly miscarried were hypoxic, three were known to have significant collateral vessels and one had a fenestrated total cavopulmonary connection with right to left shunting. The majority of the miscarriages were in the first trimester, but two women suffered three second trimester miscarriages, all preceded by repeated vaginal bleeding for which no specific cause was identified.

Eight women had a total of 14 live births and 9 miscarriages between them, with all but one having at least one miscarriage. Their baseline anatomy, type of Fontan and time between Fontan operation and pregnancy are included in table 4. The median age at first pregnancy was 29 years (range 26–37 years). Seven were NYHA class 1 and one was class 2. All except one woman had an oxygen saturation of >91% at PCC and at their booking appointment. Six of eight women had a prior history of arrhythmia, but none had a prior history of thromboembolism. No women had a CARPREG score >1. Seven women were assessed by echo prior to pregnancy, six had good systemic ventricular function and one patient had mild impairment (ejection fraction 51%).

Table 4

Underlying cardiac anatomy and interventions for patients with successful live births

Obstetric and neonatal complications

Obstetric outcomes

All live births were singletons. Median gestation at delivery was 34 (range 29–40) weeks vs 39 in the control group (p<0.0001, table 5). Seven women were delivered by caesarean section, five were emergency (four for fetal distress and one for failed instrumental delivery) and two were elective (both in the same patient for severe fetal growth restriction). Seven women gave birth vaginally, three spontaneously and four after an assisted delivery. There were no cases of pregnancy-induced hypertension, pre-eclampsia or HELLP syndrome. Fifty per cent of deliveries were complicated by postpartum haemorrhage (PPH; p=0.059 vs the control group), and the mean blood loss was greater in the Fontan group (967 mL vs 375 mL, p<0.01, table 5). Fifty-seven per cent of babies were born small for gestational age all being below the 10th centile (p<0.0001, table 5). The mean birth weight was 1682 g (range 600–2710 g) vs 3385 (range 2715–4150 g), (p<0.0001, table 5), and the mean birth weight centile was 12 (p<0.0001, table 5).

Table 5

Obstetric and neonatal outcomes for Fontan and control group

Neonatal outcomes

One early neonatal death occurred in a preterm baby who developed necrotising enterocolitis. The pregnancy was complicated by prelabour rupture of membranes at 32 weeks gestation. The birth weight was 1460 g (18th centile).

Anticoagulation management

The current practice at our institution is to advise thromboprophylaxis to all women with a Fontan circulation, and doses are based on current guidance from the Royal College of Obstetricians and Gynaecologists.13 One patient in her first pregnancy was managed solely with aspirin and delivered at 40 weeks. In her second pregnancy, she was given aspirin and prophylactic subcutaneous low molecular weight heparin (LMWH). One patient who was given aspirin and clopidogrel prior to pregnancy was advised to stop these medications and take subcutaneous LMWH at a prophylactic dose. Women who were given warfarin (5/8) were switched to taking prophylactic subcutaneous LMWH as soon as pregnancy was confirmed. In patients with additional concerns regarding thromboembolism (eg, previous thrombosis or evidence of arrhythmia during pregnancy) LMWH was given at a therapeutic dose. These patients then had antifactor Xa levels performed at least once every month to ensure appropriate dosing and changes were made as appropriate.

Cardiovascular complications

Cardiovascular complications occurred in five pregnancies in four women. There were three cases of peridelivery recurrent atrial arrhythmia without haemodynamic compromise in two women, both on β blockers and both with history of previous atrial ablation. These spontaneously terminated in two cases and required medical cardioversion with intravenous amiodarone in the other. One of these women was also on a combination of 2.5 mg amiloride and 20 mg furosemide daily. No other study patient was treated with diuretics prepregnancy or during pregnancy. One patient went into spontaneous labour at 33 weeks gestation, attended and delivered at her local unit. She became hypoxic and hypotensive and diagnostic imaging suggested a pulmonary embolism for which she was thrombolysed. She developed a PPH and was transferred to a tertiary level ITU. Another patient was admitted to Intensive Care Unit following PPH. One patient, with pulmonary atresia and MAPCAS (NYHA class 1 prior to pregnancy), became breathless and hypoxic (O2 saturation 85%) in the second trimester requiring admission for bed rest and supplemental oxygen. While her resting saturations were 91% prior to pregnancy and she desaturated to 83% on exercise. She returned to baseline status postpartum.


The Fontan procedure has improved the prognosis of women with a single ventricle physiology and many are now considering pregnancy. Our assumption, prior to the study, was that after PCC women with severe cardiac disease would opt to avoid pregnancy and conversely, those with less severe disease would opt to try to conceive. In fact, there was no difference in cardiac condition between those women who opted to try to conceive and those who did not, suggesting that the opposite was true. These data imply that other factors may be of equal or greater importance for women than the severity of their heart disease, such as deferring having family, financial independence and a stable relationship, when they are coming to a decision about pregnancy.

Retrospective studies of pregnancies in women with heart disease are limited as they typically only include pregnancies that are carried to viability and so may be biased towards women with mild heart disease. In this series, we report the reproductive experience of a complete cohort of women who considered becoming pregnant. Those who opted for pregnancy had a similar overall severity of heart disease to those who chose to avoid pregnancy. It is possible that only those who had less severe heart disease accepted the invitation to attend PCC; however, the three women who refused PCC actually had a similar spectrum of heart disease as the women who accepted PCC, suggesting that our series is representative of the overall population of women with a Fontan circulation. Furthermore, we show women who did not conceive or who only had miscarriages had a worse CARPREG score and greater rates of hypoxia. In fact, 8 of the 11 women who had at least one miscarriage demonstrated baseline hypoxaemia, which is likely to significantly contribute to high rates of miscarriage. Furthermore, we demonstrate that studies (as ours) that are limited to women who carry their pregnancy to viability are likely to be biased in favour of women who have milder heart disease.

Higher rates of infertility and miscarriage in women with a Fontan circulation have been reported before.7 ,10 However, the overall rate of miscarriage (excluding failed attempts at assisted conception) was 66%, which is much higher than described previously.11 It is three times higher than the background rate of miscarriage in women with congenital heart disease;14 and four times higher than in the healthy population.15 The association with hypoxia is well recognised and has been highlighted previously in women with congenital heart disease.16 ,17 Several women experienced recurrent vaginal bleeds, usually culminating in pregnancy loss, and which may be due to uterine congestion or anticoagulant use. This observation is also reflected in other reports.18 ,19

Our series confirms the association between Fontan and preterm delivery (71%), which has been linked to hypoxia. However, only one patient was hypoxic prior to pregnancy, suggesting that even a relatively reduced oxygen level and/or restricted cardiac output may have an adverse effect on pregnancy outcome. Similarly, Gouton et al11 reported an overall rate of preterm delivery of 69% in their cohort of 36 live births in women with a Fontan circulation, again with only one having O2 saturations of <92%.

There are few reports of SGA rates in patients with a Fontan circulation. Gouton reported an overall rate of SGA of 11%.11 Canobbio et al and Drenthen et al did not give any figures.7 ,8 However, we have reported previously that a larger group of women with a variety of congenital heart diseases had an overall SGA rate of 25% and almost 40% if there was concurrent maternal cyanosis.20 The risk of SGA in our series (57%) is double that of the general CHD population and women considering a pregnancy should be warned about this outcome.

Outside of pregnancy, arrhythmia is a frequent complication of a Fontan circulation often as a consequence of an enlarged right atrium or surgery resulting in atrial scarring.21–24 We observed an arrhythmia in 21% of women in this study, but this was not associated with a worse fetal outcome.

Only one case experienced a functional deterioration during pregnancy. However, after pregnancy, we found no evidence of an accelerated deterioration, in agreement with other reports.10 ,11 However, mean follow-up was only 3.9 years and only five women had complete imaging prepregnancy and postpregnancy. Consequently, we cannot be certain what the impact of pregnancy is on the natural history of a Fontan circulation, although Cannobio et al24 questioned the link between pregnancy and late death.

In summary, we suggest that women with a Fontan circulation undergoing PCC require comprehensive counselling that should include the increased risk of miscarriage, preterm delivery and poor fetal growth. These risks are higher and the chance of a successful pregnancy is lower in women who are hypoxic. Additionally, women should be counselled that they may require thromboprophylaxis and maybe more prone to arrhythmias in pregnancy. Women should be informed that their clinical status may deteriorate during pregnancy and that at present it is unclear how pregnancy affects the natural history of a Fontan circulation.


This study was retrospective in nature and while most information was complete there are some missing data, particularly information regarding patients at follow-up. Patients’ views of pregnancy before PCC were not known; it was not possible to explore how well-informed patients felt following PCC and whether it directly influenced their decisions. Five women were lost to follow-up either because they moved overseas or declined to attend. In addition, treatment was not standardised, but the same core team of obstetricians and cardiologists cared for all patients over the study period. Our study population was limited to women who were NYHA class 1 or 2 and only a single patient referred to PCC had severely impaired ventricular function. We cannot exclude that the outcome of tertiary centre adult congenital heart disease cardiologist care and advice was such that some women with a poorly functioning Fontan chose not to attend or were not referred to joint PCC clinic. We did not have any patients who conceived or who had a miscarriage that were NYHA class 3 or 4, or who were known to have a protein losing enteropathy. Finally, our sample size was small and so any conclusions are consequently limited.


Most women with a Fontan circulation are willing to attend PCC, but their choices regarding pregnancy appear to be multifactorial. For those who conceive, rates of miscarriage are high and correlate with maternal hypoxaemia. Women who go on to have a live birth are likely to have a higher proportion of pregnancy complications compared with controls.

Key messages

What is already known on this subject?

  • A parturient with a Fontan circulation is considered high risk among all current risk assessments. Obstetric risks include that of high rates of miscarriage, preterm delivery and babies that are born small for gestational age. Current case series are small and counselling women with regard to pregnancy remain challenging.

What might this study add?

  • This study examines a whole cohort of women with a Fontan circulation who have undergone prepregnancy counselling and looked at their pregnancy choices. The majority of women still wished to opt for pregnancy, but a significant proportion chose to delay pregnancy. Those opting to delay had a better CAREPREG score. Rates of miscarriage are higher than previously reported and are related to baseline hypoxaemia.

How might this impact on clinical practice?

  • Our study may help to provide greater information to patients at the point of preconception care, especially with regard to the risk of miscarriage and knowing that those with a higher cardiac disease in pregnancy (CARPREG) score and hypoxaemia are likely to have poorer outcomes.



  • Contributors MC, MRJ, KVK and LS contributed to the conception and design of the manuscript. MC and KVK were responsible for collating and analysing the data. MC, KVK and MRJ were responsible for the drafting of the manuscript, critical revision and final approval of the manuscript submitted. PJS, SVB-N, MG, LS and AU contributed to the critical revision and final approval of the manuscript submitted.

  • Funding SVB-N was supported by an Intermediate Clinical Research Fellowship from the British Heart Foundation (FS/11/38/28864). This project was supported by the NIHR Cardiovascular Biomedical Research Unit of Royal Brompton and Harefield NHS Foundation Trust and Imperial College London. This report is independent research by the National Institute for Health Research Biomedical Research Unit Funding Scheme.

  • Competing interests None declared.

  • Ethics approval Royal Brompton and Harefield Local Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles