- Email: [email protected]
Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?
IJCA-27955; No of Pages 7 International Journal of Cardiology xxx (xxxx) xxx
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation? Emily Moroney a,b,1, Diana Zannino a, Rachael Cordina c,d, Thomas Gentles e, Yves d'Udekem a,b, Dominica Zentner a,f,g,⁎,1 a
Murdoch Children's Research Institute, Melbourne, Australia Department of Paediatrics, The University of Melbourne, Melbourne, Australia Department of Cardiology, Royal Prince Alfred Hospital, Sydney, Australia d Sydney Medical School, University of Sydney, Sydney, Australia e The Paediatric and Congenital Cardiac Service, Starship Children's Hospital, Auckland, New Zealand f Department of Cardiology, Royal Melbourne Hospital, Melbourne, Australia g Department of Medicine, Royal Melbourne Hospital, University of Melbourne, Melbourne, Australia b c
a r t i c l e
i n f o
Article history: Received 15 May 2019 Received in revised form 16 July 2019 Accepted 14 August 2019 Available online xxxx Keywords: Heart Defects, Congenital Fontan procedure Pregnancy Embolism and thrombosis Heart Outcomes
a b s t r a c t Background: Pregnancy in women with a Fontan circulation is understood to have a significantly increased risk of maternal morbidity and mortality. Potential longer-term effects on health outcomes remain unknown. Aim: Ascertainment of adverse events, post-delivery, in women from the Australia and New Zealand (ANZ) Fontan Registry. Comparator data from women without pregnancy and men was utilised. Methodology/results: Living adults were identified (n = 263 women, 280 men) with Registry analysis ascertaining cardiac morphology, type of Fontan procedure, pregnancy and significant cardiac adverse events. Data are described descriptively for morphology and as median (inter quartile range) for other variables. Thirty women reported 45 pregnancies beyond 20 weeks' gestation. Despite being older, these women, predelivery, had the lowest cardiac morbidity burden, compared to both male and female controls (OR = 3.25 (95% CI 1.44–8.35, p = 0.008 and OR = 2.59 (95% CI 1.14–6.70), p = 0.03). These differences were not present post-delivery. Median follow-up time post-delivery was 3.6 (1.2–7.5) years. Thrombus documentation and thromboembolic events were more common post-delivery, although differences were not significant in a propensity analysis model (adjusting for post-Fontan follow-up time, age and type of Fontan). There were no differences in arrhythmia, cardioversion, heart transplant, Fontan conversion or mortality between women postdelivery and either male or female controls. Conclusion: The apparent increase in thromboembolic events post pregnancy requires further investigation in a larger group of women over a longer period of follow-up time. This may have potential implications for postpartum anticoagulation, and for future health outcomes. © 2019 Elsevier B.V. All rights reserved.
1. Introduction With improved survival an increasing number of women with a Fontan circulation are now of childbearing age. This poses unique questions regarding the ability of the Fontan circulation to meet the physiological challenge of pregnancy, and whether pregnancy has a long-term sequalae for maternal health outcomes. Normal pregnancy is a hypercoagulable state [1] and is associated with a measurable, though subclinical, decrease in cardiac function [2]. ⁎ Corresponding author at: Department of Cardiology, The Royal Melbourne Hospital, Vic, Australia. E-mail address: [email protected] (D. Zentner). 1 Equal first authors.
These issues may be particularly significant for women with a Fontan circulation, given the increased risk of arrhythmias [3–5], ventricular dysfunction [6,7] and heart failure [4], and thromboembolic events [3,8] in the Fontan population. Guidelines [9,10] state that pregnancy in these women is associated with significant risk of maternal morbidity and mortality. Complications include arrhythmia, heart failure and obstetric haemorrhage [11–13]. Though fetal mortality [11,12] is described, no reports of maternal mortality, during pregnancy, for women with a Fontan circulation exist in the current literature. It is therefore unknown whether the guidelines are overly cautious, or mortality events have not been published. The potential for long-term sequelae from pregnancy in women with the Fontan circulation remains unknown. Limited studies have reported late maternal deaths [13,14], and complications such as
https://doi.org/10.1016/j.ijcard.2019.08.039 0167-5273/© 2019 Elsevier B.V. All rights reserved.
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
2
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
arrhythmias, thromboembolic complications and Fontan failure [15]. However, the absence of comparator cohorts in these studies makes interpretation of long-term outcomes in this population difficult. Pre-conception counselling is a class 1 recommendation for women with CHD [16]. Ideally, this should be evidence based and include an understanding of the potential for delayed morbidity and mortality sequalae from pregnancy. As men have been observed to fare less well with a Fontan circulation [17] they may serve as a control group in studies for acceptable, even if higher, adverse event rates than non pregnant female controls. Thus, we sought to retrospectively compare morbidity and mortality, within the Australia and New Zealand (ANZ) Fontan Registry, between women who had and women who had not, experienced pregnancy, and additionally compared their outcomes to those in men. 2. Methods 2.1. Patient selection The ANZ Fontan Registry collects medical correspondence for patients who have had their Fontan procedure in either Australia or New Zealand, or who are followed up in this region. The governance of the Registry has been described in depth previously [18]. Authorization for ongoing retrospective analysis of the data held in the Registry was included in the initial design of the Fontan registry, and consequently approval for this study was as part of the ongoing ethics approval granted for the Fontan Registry, (HREC36260, HREC1888EP, NTX/10/ 07/064). Data for this study were extracted from the Registry for all participants N18 years of age as of the 1st of January 2017 (n = 689). Exclusion was then undertaken for: individuals who died before the age of 18 (n = 36), those with no medical data recorded beyond the age of 18 (n = 106), and patients whose Fontan was taken down prior to turning 18 (n = 2). Two women who had previously revealed a pregnancy but not consented to share data, and had no post partum follow up data in the Registry, were also excluded. This left a final cohort of 543 (280 males and 263 females). The female cohort was divided into the pregnancy group, being women who had recorded a pregnancy beyond 20 weeks gestation (n = 30), and the female control group, who either recorded no pregnancy or only pregnancies b20 weeks duration (n = 233). Utilisation of 20 weeks gestation was based on a low likelihood that women would have reached that gestation without seeing their cardiologist. Additionally, it forms the gestational age at which, legally, a delivery is considered a birth in Australia and New Zealand and must be registered [19,20]. Women were not excluded/ included on the basis of earlier miscarriage, as this is often poorly captured in medical records [21]. Furthermore miscarriage may not be as well recognised in a cohort with an increased frequency of irregular menstrual cycling [22]. Variables ascertained for the female and male cohorts were original cardiac morphology, age at Fontan operation, type of Fontan operation, post Fontan follow up time and documented significant cardiac and Fontan related complications. These were extracted from ANZ Registry data as previously described [3], and included NYHA class, oxygen saturation, arrhythmias, thromboembolic events, cardioversions, pacemaker insertion, Fontan conversion, heart transplantation and death. Additionally, medications and echocardiogram report data were ascertained and clinical letters revaluated for the female cohort. 2.2. Data analysis Participant data was exported from REDCap (Research Electronic Data Capture, Fontan Registry data) and analysed with SPSS Version 24.0 (IBM Corp, Armonk, NY) and R version 3.2.3 (R Foundation, Vienna, Austria) software.
Continuous variables are presented as median and interquartile range (IQR), and categorical variables as percentages. Descriptive analysis, nonparametric analysis (Median test) and Fisher exact test were used. Cardiac morbidity load was determined as the number of events per person year, including all above stipulated endpoints. In women who had had a pregnancy, this was broken down into events pre and post-delivery. Kaplan Meier analysis was used to generate data depicting freedom from heart transplant, Fontan conversion and death. The cumulative incidence of arrhythmia, cardioversion and thromboembolic events was compared between the pregnancy and control cohorts, with pregnancy accounted for as a time-dependent covariate. Cox regression was used to test for differences in hazard ratios for events between the pregnancy group and the control groups. Exploratory propensity scoring was then undertaken for variables that appeared to differ after Cox Regression. A p value of b0.05 was considered to be statistically significant. Confidence intervals are presented where possible. Last follow up recorded was the date of the last appointment or the date a Registry documented event was recorded, if later. For analysis women were considered to be pre delivery, until the date of their first (or only) delivery and post delivery, for any subsequent event. 3. Results 3.1. Demographic data The pregnancy group were older (p = 0.002) with a longer postoperative follow up time (p = 0.001). Type of Fontan procedure, and predominant ventricular morphology did not statistically differ between the groups, although there were fewer extra cardiac conduit Fontan circulation patients in the pregnancy group (Table 1). 3.2. Cardiac morbidity load Adverse cardiac events (arrhythmias, cardioversions, thromboembolic events, pacemaker insertion, heart transplant, Fontan conversion and death) were combined to create a cardiac morbidity load score, utilised to calculate events per year (Fig. 1). For the pregnancy group, this was further divided into events ‘pre’ and ‘post’ delivery. The incidence of events was lower in the pregnancy women, predelivery, in comparison to both male and female controls (OR = 3.25 (95% CI 1.44–8.35, p = 0.008 and OR = 2.59 (95% CI 1.14–6.70), p = 0.03 respectively). A greater proportion of women in the pre-delivery group reported no cardiac events (n = 23/30, 77%) compared to the male controls (n = 145/280, 52%, p = 0. 01), and to the female controls (n = 129/233, 55%, p = 0.03). This is despite the median age of the predelivery cohort being significantly older than the female and male controls (28 (25.5–30.6) years vs 25.0 (20.5–29.3), p = 0.02 and 28 (25.5–30.6) years vs 24 (20.7–29.4), p = 0.0002). When compared post partum, the incidence of events in postdelivery women compared with female controls and male controls was no longer statistically significantly different (OR = 1.40 (95% CI 0.56–3.82), p = 0.5, OR = 1.75 (95% CI 0.70–4.74), p = 0.2 respectively).The post-delivery group had a higher number of events, however, than both their own comparative pre-delivery data and female controls, ranging from 0 to 1.2 events per person year. This morbidity event rate was the same as that seen in the male cohort. 3.3. Pregnancy outcome Forty women reported a total of 82 pregnancies (Table 2), ascertained from both self-reported data (questionnaire and follow up telephone surveys, previously reported [11]) and clinical letters. Median maternal age at first delivery N20 weeks was 28.02 years (IQR 25.47–30.58). Women reported experiencing up to 10 pregnancies (a single individual, who experienced 7 miscarriages). The median number of successful pregnancies was 1 (IQR 1–2).
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
3
Table 1 Baseline characteristics. Variable
Pregnancy group (n = 30)
Female controls (n = 233)
Male controls (n = 280)
Statistical comparison
Age at last follow up, median (IQR), y Post-operative follow-up time, median (IQR), y Type of Fontan, n (%) Atriopulmonary Lateral tunnel Extra cardiac conduit Ventricle morphology, n (%) Right Left Indeterminate/biventricular Primary diagnosis, n (%) Tricuspid atresia Double inlet left ventricle Double outlet right ventricle Atrioventricular canal or AVSD Pulmonary atresia with VSD Pulmonary atresia with intact ventricular septum HLHS Ebstein's anomaly ccTGA Other
32.9 (26.6–37.3) 22.7 (20.5–26.6)
25.0 (20.5–29.3) 19.5 (15.1–23.9)
24.0 (20.7–29.4) 19.3 (15.0–23.5)
p = 0.002a p = 0.001a p = 0.14b
14 (47) 13 (43) 3 (10)
77 (33.0) 88 (37.8) 68 (29.2)
87 (31.1) 106 (37.9) 87 (31.1)
4 (13) 24 (80) 2 (7)
48 (20.6) 161 (69.1) 24 (10.3)
75 (26.8) 180 (64.3) 25 (8.9)
8 (27) 10 (33) 2 (7) 1 (3) 4 (13) 1 (3) 0 0 3 (11) 1 (3)
72 (30.9) 46 (19.7) 37 (15.9) 19 (8.2) 16 (6.9) 5 (2.1) 6 (2.6) 4 (1.7) 14 (6.0) 14 (6.0)
77 (27.5) 58 (20.7) 47 (16.8) 15 (5.4) 8 (2.9) 15 (5.4) 8 (2.9) 2 (0.7) 20 (7.1) 30 (10.7)
p = 0.3b
Abbreviations: AVSD indicates atrioventricular septal defect; ccTGA, congenitally corrected transposition of the great arteries; HLHS, hypoplastic left heart syndrome; IQR, interquartile range; and VSD, ventricular septal defect. a Non-parametric comparison, median test. b Fisher's exact test.
3.4. Functional status at last follow up I. NYHA class A recent NYHA class was recorded for 200 women in the female control cohort (86%) and 22 in the pregnancy group (73%). In the 200 female controls, 115 (57.5%) were NYHA Class 1, 78 (39%) were NYHA
Class 2, 6 (3%) were NYHA Class 3 and one woman (0.5%) was NYHA Class 4. Comparatively, in the 22 women in the pregnancy group, 15 (68%) had an NYHA recording pre-delivery, a median 1.2 years prior to pregnancy and 10 (67%) were NYHA Class 1, while 5 (33%) were NYHA Class 2. A median 3.62 years (IQR 1.19–7.50) post-delivery, 12 (54.5%) were NYHA Class 1, 9 women (41%) were NYHA Class 2 and one
Fig. 1. Number of adverse cardiac events per person year. Adverse cardiac events includes arrhythmias, cardioversions, thromboembolic events, pacemaker insertion, heart transplant, Fontan conversion and death, presented as per year of follow up data. The pregnancy group was subdivided, into a before delivery group; showing the number of events seen from birth to last recorded follow up before delivery and an after delivery group; showing the number of events seen post-delivery to last recorded follow up. Four women in the after delivery group had no data recorded post-delivery, and hence, were excluded from the after delivery group.
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
4
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
Table 2 Pregnancy outcome and maternal echocardiogram results at last follow up. Pregnancy outcome
Pregnancy outcome • Termination • Miscarriage • Delivery N20 weeks' gestation • Still born Baby outcome (live birth) (n = 43) • Neonatal deaths
Pregnancy group (30 women, 69 pregnancies)
Female controls (233 women, 13 pregnancies)
2 (3%) 22 (32%) 42 pregnancies/43 babies (1 set twins) (61%) 3 (4%)
3 (23%) 10 (77%)
3 (1 singleton, 1 set of twins)
Maternal echocardiogram results Variable
Pregnancy group (post-delivery) (n = 24)
Statistical Female controls (n = comparison 178)
Echocardiogram (y) post op, median (IQR) Echocardiogram (y) post pregnancy, median (IQR) Subjective ventricular systolic function, n (%)a Normal Reduced (mild – severe) AV regurgitation, n (%)a None/trivial Mild – moderated
24.4 (21.2–26.7)
20.6 (15.8–24.8) NA
3.5 (2.0–6.7)
p = 0.02b
143 (83.6) 28 (16.4)
11 (46) 13 (54)
70 (43.0) 92 (57.0)
IV. Echocardiogram
– p = 0.39c
18 (75) 6 (25)
Anticoagulation data (consisting of both antiplatelet and anticoagulation medications for the purpose of this analysis) was available for 193 female controls (83%), and 22 women in the pregnancy group, post delivery (73%). Two women in the pregnancy group (9%), and two women in the female control group were taking 2 or more anticoagulation medications (1%). No anticoagulation was prescribed for 3 (14%) women in the pregnancy group, and 12(6%) women in the female control cohort. At follow up, the commonest anticoagulant was warfarin, utilised by 40% (n = 9/22) of women in the pregnancy group and 55% (n = 107/193) of the female controls, p = 0.26. Alternate therapies to warfarin prescribed post-delivery were, aspirin (n = 6/22), clexane (n = 1), apixaban (n = 1), or dual therapies of warfarin and aspirin (n = 1), or aspirin and clexane (n = 1). Comparatively, in the female control cohort, aspirin was used in 35% (n = 68 /193) aspirin, with the remainder taking clexane (n = 1), clopidogrel (n = 1), rivaroxaban (n = 1), apixaban (n = 1), dual therapy with warfarin and aspirin (n = 1) or triple therapy with aspirin, warfarin and clopidogrel (n = 1).
p = 0.82c
a Owing to incomplete echocardiogram data within reports for some patients, percentage does not reflect column heads. b Calculated using non-parametric comparison, median test. c Calculated using Fisher‘s exact test. d No cases of severe AV valve regurgitation recorded.
Echocardiogram reports were available for 178 (76%) women in the female control group, and post-delivery echocardiogram recordings were available for 24 (80%) in the pregnancy group. At last echocardiogram, two (1.1%) women in the female control group had an intracardiac or Fontan circuit thrombus detected, compared to three (13%) woman in the pregnancy group post-delivery, p = 0.01. Of these, a clinical thromboembolic event was then recorded for 2 of the 3 pregnancy group women, only. In contrast, in the remaining women the thrombus was described as an incidental finding, with no clinical sequelae and not resulting in any change to medical therapy. 3.5. Late adverse cardiac events
woman (4.5%) was NYHA Class 3. Three women with NYHA data experienced a deterioration of one NYHA class post-delivery (13.6%). Of the 228 males (81%) with a recent NYHA recording, 148 (65%) were NYHA Class 1, 69 (30%) were NYHA Class 2, 10 (4%) were NYHA Class 3 and one male (0.4%) was NYHA class 4. II. Oxygen saturations
Oxygen saturation data was available for 132 (56%) women in the female control group, with a median saturation of 92% (83–95), and 159 (57%) males, with a median saturation of 93% (90–95). The lowest saturation in the female control cohort was 76% and in the male cohort 80%. In the pregnancy group, 20 women (67%) had saturation recordings available. Only 8 of the 20 pregnancy women had a pre-delivery saturation recorded, median 91.5% (90–94). The lowest pre-delivery saturation recorded was 85% in a woman 1 year before a livebirth at 31 weeks. In the 20 women with a post-delivery recording, the median saturation was 92% (89.5–92). III. Medication
Current medication data were available for 203 women in the female control group (87%) and for 23 (77%) women in the pregnancy group, a median of 3.04 (IQR 1.04–7.08) years post-delivery. The median number of cardiac medications was 1 in both groups, ranging from 0 to 6 in the pregnancy group and 0 to 8 in the control female cohort. The number of women who were on no cardiac medications was 7 (3%) in the control female cohort, and 3 (13%) in the pregnancy group.
Follow up post pregnancy, a median 3.6 years (IQR 1.2–7.5) was recorded in 26 (87%) of women. Pre-delivery, there had been a total of 22 events reported in 7 women (23%). Follow up for a minimum postpartum time of 2 years was available in 18 women, of whom 7 (38%) had experienced 26 adverse events post-delivery. Median follow up time of these 7 women with an adverse event postpartum was 7.5 years (3.9–14.4). Comparatively, among the control cohort, 104 (45%) women in the female control group, and 135 (48%) of men had experienced at least one adverse event throughout their life, (arrhythmia, thromboembolic event, cardioversion, pacemaker insertion, heart transplant, Fontan conversion or death). I. Thromboembolic events
In the pregnancy group, post-delivery, 4 thromboembolic events occurred in 3 women with an atriopulmonary Fontan. One woman experienced two strokes, one year apart, while the remaining two women experienced a right atrial thrombus with pulmonary embolism (n = 1), and right atrial thrombus causing pulmonary emboli and splenic infarct, identified post-mortem as cause of death (n = 1). Contemporaneous anticoagulation was aspirin (n = 1), warfarin (n = 2) and noncompliance with prescribed anticoagulation (n = 1). Prior to pregnancy, these women had no history of TE nor had a thrombus being reported on echocardiogram. The difference in the incidence of thromboembolic events between the pregnancy group and female and male groups is shown in Fig. 2. When assessed according to age, the pregnancy group had an increased incidence of thromboembolic events: HR (95% CI): 4.84 (1.08–21.6) p = 0.04. However, this was not statistically significant when analysed
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
5
No difference was observed between the control groups (male and female) and the pregnancy group in the time from Fontan operation to first arrhythmia, HR (95% CI): 1.56 (0.55–4.56) p = 0.4 or the age at which the first arrhythmia occurred, HR (95% CI): 1.48 (0.44–4.94) p = 0.5. Similarly, no difference was observed between the three groups in the incidence of first cardioversion, both when analysed as time since Fontan operation, HR (95% CI): 0.46 (0.06–3.43) p = 0.4 or age, HR (95% CI) 0.32 (0.04–2.49) p = 0.3. At last follow up, a pacemaker had been implanted in 40 (14%) males and in 30 (13%) women in the female control group. In the pregnancy group, 7 (23%) of women had a pacemaker. Implantation preceded pregnancy in 5 and post-dated pregnancy in 2. III. Fontan conversion, heart transplant and death - impact on longterm survival
Fontan failure endpoints were seen in 3 (10%) women in the pregnancy group, with 1 death 2.6 years post-delivery, 1 heart transplant 6 years post-delivery and 1 Fontan conversion 0.9 years post-delivery. However, survival free from Fontan conversion, heart transplant and death did not differ between the control female and male groups and the pregnancy group, either when measured in years since Fontan, HR (per year increase): 1.10 (95% CI 0.32–3.76) p = 0.9; or age, HR (per year increase):0.88 (95% CI 0.24–3.16) p = 0.8. 4. Discussion
Fig. 2. Cumulative incidence of thromboembolic events according to gender and pregnancy group, measured across A) years since Fontan and B) age. Pregnancy was accounted for as a time-dependent covariate, with women transitioning from the not pregnant (female control) to pregnant group after their first delivery.
according to years post Fontan operation: HR (95% CI): 3.12 (0.87–11.1) p = 0.08. Exploratory propensity scoring was undertaken to account for the effect of the differences in age and time since Fontan, on the difference in TE events, between the pregnancy and no-pregnancy groups. Matching age at follow up, time since Fontan and Fontan type, reduced the study cohort to 19 (63%) women in the pregnancy group and 48 (17%) women in the female control group. TE events were also reduced from 36 to 10. Propensity scoring analysis removed the difference between the pregnant and female control subgroups with respect to thromboembolic events, when assessed according to age; coefficient factor 4.77: (95% CI: 0.6692–34.07), p = 0.119. II. Arrhythmias and cardioversion
Five women (19%) in the pregnancy group experienced 12 arrhythmia events post-delivery. Pre-delivery, only one woman had experienced an arrhythmia. The median time between pregnancy and the first post-delivery arrhythmia was 7.38 (1.08–11.80) years.
This study explores whether pregnancy may result in adverse outcomes for women with a Fontan circulation, and is the first to include comparison to individuals with a Fontan circulation who have not undergone pregnancy. In this binational registry cohort, maternal cardiac events appeared least likely in the women in the pregnancy group, pre-delivery. Coding in this study was such that gestational cardiac events were counted as pre-delivery cardiac events. As pregnancy is an accepted high risk time [23] for the endpoints collated, the significance of the decreased incidence of events in the pre-delivery cohort is of further clinical relevance. Furthermore, the significantly greater age of the women who underwent pregnancy, in a population where time is understood to significantly drive adverse events [3], suggests that those undergoing pregnancy are a selected group with a well functioning Fontan circulation. This is not surprising given high risk patients are likely to be counselled against pregnancy. Current guidelines do not clearly detail how to individualise the risk of pregnancy for women with a Fontan circulation [9]. While the just released ESC guidelines [24] recommend counselling against pregnancy in any woman post a Fontan operation with any associated complication, the CARPREG II guidelines offer a cumulative scoring assessment where points are accrued for predictors of poorer maternal outcomes [25]. The scoring system includes a prior history of cardiac events, a well recognised risk factor for further adverse complications during pregnancy in women with CHD [10,26–29]. A recognised determinant of live birth is maternal oxygen saturation [30,31]. In our cohort, and in contrast to both the female controls and male controls, no woman who attained a gestational age N 20 weeks had a pre-delivery saturation recording below 85%. This accords with a prior publication (24) reporting a small group of women (n = 8) with a Fontan circulation with saturations below 85% whom all experienced a miscarriage. The data in this paper may suggest that clinical assessment has historically been incorporating relevant factors when risk stratifying these women, and, further, that patients may be compliant with advice. However, as both unmanaged chronic disease [32,33] and impaired haemodynamics [34] are risk factors for miscarriage and subfertility, women in the no-pregnancy group, who experienced a greater cardiac
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
6
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
morbidity load, may have had increased difficulty in conceiving and carrying a pregnancy N20 weeks. Post-delivery, those same women with a decreased incidence of events compared to both male and female controls pre-delivery, no longer demonstrated this difference. Additionally, the events per person year data is highest for post-delivery women and men, though the numbers are small. This suggests that pregnancy may modulate future maternal health events, as despite the short follow up time of 3.6 years post pregnancy, the event rate increased in that subgroup. This highlights the importance of adequate pre pregnancy counselling and assessment. It appears realistic to counsel women that their future health quality may reduce post a pregnancy, though the data would suggest that that reduction is still within the spectrum of well being for these patients. Despite the increased time between their Fontan operation and last echocardiogram, the pregnancy group were not shown to have greater ventricular dysfunction or AV valve regurgitation. This is surprising given reports in other CHD that pregnancy may expedite ventricular dysfunction [35–37], but may reflect the short post pregnancy follow up time in this study. Additionally, it is recognised that echocardiographic assessment of the single ventricle is challenging [38,39]. It is also possible that pre conception assessment and counselling has largely seen women with better ventricular function pursue pregnancy, as supported by other CHD literature [26,28,40]. This study suggests increased thrombus detection after pregnancy. However the significant difference found with cox regression for thromboembolic events, was not seen in the propensity score model. While this may be explained by the reduced sample size and hence event rate utilised in that comparison, it may also reflect that the difference observed relates to other variables, such as time since Fontan. Nonetheless, the potential increase in TE events demonstrated raises a concern, as clinically diagnosed thrombus has been noted to be a significant predictor of later thromboembolic death, heart failure, transplantation and all cause mortality [41]. Other studies [13,15], have also reported thromboembolic events post pregnancy in women with a Fontan circulation. It is well established that pregnancy and the postpartum period are high risk times for thromboembolic events [42–45], and these have been reported in the Fontan population [12,46]. However, none of the TE events in our population occurred within the puerperium, suggestive of a lingering effect of pregnancy in this population. It is feasible that an increased risk of thromboembolic events results from pregnancy, in a population with raised venous pressures [6], due to pregnancy-related weight retention [42] and/or varicose vein development [47]. Experiencing a TE event during pregnancy is known to increase the risk of future TE events [48]. It is possible that some women with post-delivery TE events had an unrecognised TE event in pregnancy. Our study did not demonstrate an increased incidence of arrhythmias, cardioversion, Fontan conversion, heart transplant or death after pregnancy during this short post- delivery follow up time. As many of these adverse events appear likely to be time-dependent phenomenon, these women should be followed prospectively, and in a matched manner, to allow documentation of long term outcomes.
5. Limitations This study is limited by the small sample size and inherent challenges of an unmatched, retrospective study. Data was restricted to the information documented in the ANZ Fontan Registry. Echocardiogram reports were unavailable for 24% of women in the female control group and 20% of women in the pregnancy group. Furthermore, some clinical letters did not outline endpoints such as oxygen saturations, NYHA class or current medications. The Registry collects data from various centres throughout Australia and New Zealand. We recognise that the echocardiogram protocol and reporting may therefore be different between the different centres.
6. Conclusion This retrospective cohort study did not show pregnancy to be associated with an increase in ventricular dysfunction, arrhythmias, cardioversions, heart transplant or mortality, within the time frame of data available. However, a slightly higher risk of thromboembolic events was seen post-delivery. To date anticoagulation management for women with a Fontan circulation differs widely. If the concerns raised by this paper are confirmed, there may be a role, particularly for post-delivery anticoagulation. Further work is also needed to understand how clinicians undertake pre-pregnancy evaluation. It is unclear whether the better health status of women who subsequently have a pregnancy reflects accurate risk assessment and prognostication, or an increased miscarriage event rate in the women who do not have a pregnancy. We are currently setting up a Fontan pregnancy database, as an international collaborative effort, that will allow a prospective, longitudinal and matched study of post pregnancy outcomes in this population. This is essential in order to provide clinicians, women, and their families, with increasingly accurate data that addresses concern about the possible effect of pregnancy on the future health of these women. Acknowledgements The authors thank all the participants enrolled in the Fontan Registry, the research assistants for their support in maintaining the Australian and New Zealand Fontan Registry and the support provided to the Murdoch Children's Research Institute by the Victorian Government's Operational Infrastructure Support Program. Funding This work was supported by a National Health and Medical Research Council (NHMRC) Partnership grant (1076849). Prof. Yves d'Udekem is a Clinician Practitioner Fellow of the NHMRC (1082186). Declaration of competing interest Prof Yves d'Udekem is a consultant for the companies MSD and Actelion. The remaining authors report no relationships that could be construed as a conflict of interest. References [1] P. Soma-Pillay, N.-P. Catherine, H. Tolppanen, A. Mebazaa, H. Tolppanen, A. Mebazaa, Physiological changes in pregnancy, Cardiovasc. J. Afr. 27 (2) (2016 Mar-Apr) 89–94 (08/31/received, 03/04/accepted;27(2):89-94. PubMed PMID: PMC4928162). [2] D. Zentner, M. du Plessis, S. Brennecke, J. Wong, L. Grigg, S.B. Harrap, Deterioration in cardiac systolic and diastolic function late in normal human pregnancy, Clin. Sci. (Lond.) 116 (7) (2009) 599–606 Apr. (PubMed PMID: 18855763). [3] C.L. Poh, D. Zannino, R.G. Weintraub, D.S. Winlaw, L.E. Grigg, R. Cordina, et al., Three decades later: the fate of the population of patients who underwent the Atriopulmonary Fontan procedure, Int. J. Cardiol. 231 (2017 Mar 15) 99–104 PubMed PMID: (28100430). [4] Y. d'Udekem, A.J. Iyengar, J.C. Galati, V. Forsdick, R.G. Weintraub, G.R. Wheaton, et al., Redefining expectations of long-term survival after the Fontan procedure: twentyfive years of follow-up from the entire population of Australia and New Zealand, Circulation 09 (11 Suppl 1) (2014 Sep) 130 (S32-8. PubMed PMID: 25200053. Epub 2014/09/10. eng). [5] T.A. Carins, W.Y. Shi, A.J. Iyengar, A. Nisbet, V. Forsdick, D. Zannino, et al., Long-term outcomes after first-onset arrhythmia in Fontan physiology, J. Thorac. Cardiovasc. Surg. 152 (5) (2016 Nov) 1355–1363.e1 PubMed PMID: 27751239. Epub 2016/10/ 19. eng. [6] M. Gewillig, S.C. Brown, The Fontan circulation after 45 years: update in physiology, Heart 102 (14) (2016 Jul 15) 1081–1086 PubMed PMID: 27220691. PMCID: (PMC4941188). [7] M. Grattan, L. Mertens, Mechanics of the functionally univentricular heart—how little do we understand and why does it matter? Can. J. Cardiol. 32 (8) (2016) 8//. (1033.e11-.e18). [8] D.N. Rosenthal, A.H. Friedman, C.S. Kleinman, G.S. Kopf, L.E. Rosenfeld, W.E. Hellenbrand, Thromboembolic complications after Fontan operations, Circulation 92 (9) (1995) 287–293.
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx [9] V. Regitz-Zagrosek, C. Blomstrom Lundqvist, C. Borghi, R. Cifkova, R. Ferreira, J.-M. Foidart, et al., ESC guidelines on the management of cardiovascular diseases during pregnancy. The Task Force on the Management of Cardiovascular Diseases during Pregnancy of the European Society of Cardiology (ESC), Eur. Heart J. 32 (24) (2011) 3147–3197. [10] M.M. Canobbio, C.A. Warnes, J. Aboulhosn, H.M. Connolly, A. Khanna, B.J. Koos, et al., Management of pregnancy in patients with complex congenital heart disease: a scientific statement for healthcare professionals from the American Heart Association, Circulation 135 (8) (2017 Jan 12) e50–e87 (PubMed PMID: 28082385). [11] D. Zentner, A. Kotevski, I. King, L. Grigg, Fertility and pregnancy in the Fontan population, Int. J. Cardiol. 208 (2016 Apr 01) 97–101 PubMed PMID: (26836494). [12] M. Gouton, J. Nizard, M. Patel, F. Sassolas, M. Jimenez, J. Radojevic, et al., Maternal and fetal outcomes of pregnancy with Fontan circulation: a multicentric observational study, Int. J. Cardiol. 187 (2015) 84–89 PubMed PMID: (25828319). [13] M.M. Canobbio, F. Cetta, C. Silversides, C. Warnes, J. Aboulhosn, J. Colman, Pregnancy after Fontan operation: early and late outcomes, J. Am. Coll. Cardiol. 61 (10) (2013) E427. [14] K.N. Pundi, K. Pundi, J.N. Johnson, J.A. Dearani, C.R. Bonnichsen, S.D. Phillips, et al., Contraception practices and pregnancy outcome in patients after Fontan operation, Congenit. Heart Dis. 11 (1) (2016) 63–70. [15] S. Arif, A. Chaudhary, P.F. Clift, R. Katie Morris, T.J. Selman, S.E. Bowater, et al., Pregnancy outcomes in patients with a fontan circulation and proposal for a risk-scoring system: single centre experience, J. Congenital Cardiol. 1 (1) (2017) 10 December 08. [16] K.K. Stout, C.J. Daniels, J.A. Aboulhosn, B. Bozkurt, C.S. Broberg, J.M. Colman, et al., AHA/ACC guideline for the management of adults with congenital heart disease, Circulation 0 (0) (2018) 2018. (CIR.0000000000000603). [17] H.M. Burkhart, J.A. Dearani, D.D. Mair, C.A. Warnes, C.C. Rowland, H.V. Schaff, et al., The modified fontan procedure: early and late results in 132 adult patients, J. Thorac. Cardiovasc. Surg. 125 (6) (2003) 1252–1258 2003/06/01/. [18] A.J. Iyengar, D.S. Winlaw, J.C. Galati, T.L. Gentles, R.G. Weintraub, R.N. Justo, et al., The Australia and New Zealand Fontan Registry: description and initial results from the first population-based Fontan registry, Intern. Med. J. 44 (2) (2014 Feb) 148–155 PubMed PMID: 24393144. Epub 2014/01/08. eng. [19] Australian Institute of Health and Welfare. Mothers & babies Glossary, , [Internet]. Available from https://www.aihw.gov.au/reports-statistics/population-groups/ mothers-babies/glossary 2017. [20] Ministry of Health, Miscarriage and stillbirth [internet], Available from: https:// www.health.govt.nz/your-health/pregnancy-and-kids/services-and-support-during-pregnancy/miscarriage-and-stillbirth 2017. [21] D. Herbert, J. Lucke, A. Dobson, Pregnancy losses in young Australian women: findings from the Australian longitudinal study on women's health, Women's Health Issues, 19(1), 2009, pp. 21–29 , 2009/01/01/. [22] M.M. Canobbio, D.D. Mair, A.J. Rapkin, J.K. Perloff, B.L. George, Menstrual patterns in females after the Fontan repair, Am. J. Cardiol. 66 (2) (1990) 238–240 1990/07/15/. [23] A. Garcia Ropero, S. Baskar, J.W. Roos Hesselink, A. Girnius, D. Zentner, L. Swan, et al., Pregnancy in women with a Fontan circulation: a systematic review of the literature, Circ. Cardiovasc. Qual. Outcomes 11 (5) (2018) May. (e004575. PubMed PMID: 29752389. Epub 2018/05/13. eng). [24] V. Regitz-Zagrosek, J.W. Roos-Hesselink, J. Bauersachs, C. Blomström-Lundqvist, R. Cífková, M. De Bonis, et al., 2018 ESC guidelines for the management of cardiovascular diseases during pregnancy, Eur. Heart J. 39 (34) (2018) 3165–3241. [25] C.K. Silversides, J. Grewal, J. Mason, M. Sermer, M. Kiess, V. Rychel, et al., Pregnancy outcomes in women with heart disease. The CARPREG II study, 71 (21) (2018) 2419–2430. [26] O.H. Balint, S.C. Siu, J. Mason, J. Grewal, R. Wald, E.N. Oechslin, et al., Cardiac outcomes after pregnancy in women with congenital heart disease, Heart 96 (20) (2010 Oct) 1656–1661 PubMed PMID: 20937754. Epub 2010/10/13. eng. [27] M.A.M. Kampman, A. Balci, H. Groen, A.P.J. van Dijk, J.W. Roos-Hesselink, J.P. van Melle, et al., Cardiac function and cardiac events 1-year postpartum in women with congenital heart disease, Am. Heart J. 169 (2) (2015) 298–304 2015/02/01/. [28] S.C. Siu, M. Sermer, J.M. Colman, A.N. Alvarez, L.-A. Mercier, B.C. Morton, et al., Prospective multicenter study of pregnancy outcomes in women with heart disease, Circulation 104 (5) (2001) 515–521.
7
[29] W. Drenthen, E. Boersma, A. Balci, P. Moons, J.W. Roos-Hesselink, B.J.M. Mulder, et al., Predictors of pregnancy complications in women with congenital heart disease, Eur. Heart J. 31 (17) (2010) 2124–2132. [30] M. Cauldwell, P.J. Steer, S. Bonner, O. Asghar, L. Swan, K. Hodson, et al., Retrospective UK multicentre study of the pregnancy outcomes of women with a Fontan repair, Heart 104 (5) (2018 Mar) 401–406 (PubMed PMID: 28954835. Epub 2017/09/29. eng). [31] P. Presbitero, J. Somerville, S. Stone, E. Aruta, D. Spiegelhalter, F. Rabajoli, Pregnancy in cyanotic congenital heart disease. Outcome of mother and fetus, Circulation 89 (6) (1994) 2673–2676. [32] R.J. Bradley, M.P. Rosen, Subfertility and gastrointestinal disease: ‘unexplained’ is often undiagnosed, Obstet. Gynecol. Surv. 59 (2) (2004) 108–117 (PubMed PMID: 00006254-200402000-00023). [33] R. Pasquali, L. Patton, A. Gambineri, Obesity and infertility. Current opinion in endocrinology, Diabetes Obes. 14 (6) (2007) 482–487 (PubMed PMID: 01266029200712000-00011). [34] R.L. Bick, D. Hoppensteadt, Recurrent miscarriage syndrome and infertility due to blood coagulation protein/platelet defects: a review and update, Clin. Appl. Thromb. Hemost. 11 (1) (2005) 1–13 2005/01/01. [35] A. Guedes, L.A. Mercier, L. Leduc, L. Berube, F. Marcotte, A. Dore, Impact of pregnancy on the systemic right ventricle after a mustard operation for transposition of the great arteries, J. Am. Coll. Cardiol. 44 (2) (2004 Jul 21) 433–437 PubMed PMID: 15261944. Epub 2004/07/21. eng. [36] C.A. Kamiya, T. Iwamiya, R. Neki, S. Katsuragi, K. Kawasaki, T. Miyoshi, et al., Outcome of pregnancy and effects on the right heart in women with repaired tetralogy of fallot, Circ. J. 76 (4) (2012) 957–963 PubMed PMID: 22277318. Epub 2012/01/27. eng. [37] N. Tzemos, C.K. Silversides, J.M. Colman, J. Therrien, G.D. Webb, J. Mason, et al., Late cardiac outcomes after pregnancy in women with congenital aortic stenosis, Am. Heart J. 157 (3) (2009) 474–480 3//. [38] R. Margossian, M.L. Schwartz, A. Prakash, L. Wruck, S.D. Colan, A.M. Atz, et al., Comparison of echocardiographic and cardiac magnetic resonance imaging measurements of functional single ventricular volumes, mass, and ejection fraction (from the Pediatric Heart Network Fontan Cross-Sectional Study)††a list of participating institutions and investigators appears in the appendix, Am. J. Cardiol. 104 (3) (2009) 419–428 2009/08/01/. [39] R.V. Williams, R. Margossian, M. Lu, A.M. Atz, T.J. Bradley, M.J. Campbell, et al., Factors impacting echocardiographic imaging after the Fontan procedure: a report from the pediatric heart network Fontan cross-sectional study, Echocardiography (Mount Kisco, NY) 30 (9) (2013 04/25) 1098–1106 (PubMed PMID: PMC3732542). [40] J. Grewal, S.C. Siu, H.J. Ross, J. Mason, O.H. Balint, M. Sermer, et al., Pregnancy outcomes in women with dilated cardiomyopathy, J. Am. Coll. Cardiol. 55 (1) (2009 Dec 29) 45–52. [41] P. Khairy, S.M. Fernandes, J.E. Mayer, J.K. Triedman, E.P. Walsh, J.E. Lock, et al., Longterm survival, modes of death, and predictors of mortality in patients with Fontan surgery, Circulation 117 (1) (2008) 85–92. [42] F.A. Anderson, F.A. Spencer, Risk factors for venous thromboembolism, Circulation 107 (23 suppl 1) (2003) I–9-I-16. [43] J.A. Heit, C.E. Kobbervig, A.H. James, T.M. Petterson, K.R. Bailey, L.J. 3rd Melton, Trends in the incidence of venous thromboembolism during pregnancy or postpartum: a 30-year population-based study, Ann. Intern. Med. 143 (10) (2005 Nov 15) 697–706 (PubMed PMID: 16287790. Epub 2005/11/17. eng). [44] I.A. Greer, Thrombosis in pregnancy: maternal and fetal issues, Lancet 353 (9160) (1999 Apr 10) 1258–1265 PubMed PMID: 10217099. Epub 1999/04/27. eng. [45] A. Abdul Sultan, J. West, L.J. Tata, K.M. Fleming, C. Nelson-Piercy, M.J. Grainge, Risk of first venous thromboembolism in pregnant women in hospital: population based cohort study from England, BMJ [Br. Med. J.] (2013) 347. [46] M. Cauldwell, K. Von Klemperer, A. Uebing, L. Swan, P.J. Steer, S.V. Babu-Narayan, et al., A cohort study of women with a Fontan circulation undergoing preconception counselling, Heart 102 (7) (2016) 534–540. [47] M. Callam, Epidemiology of varicose veins, Br. J. Surg. 81 (2) (1994) 167–173. [48] A. Bergqvist, D. Bergqvist, A. Lindhagen, T. MÄTzsch, Late symptoms after pregnancy-related deep vein thrombosis, BJOG Int. J. Obstet. Gynaecol. 97 (4) (1990) 338–341.
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation? Emily Moroney a,b,1, Diana Zannino a, Rachael Cordina c,d, Thomas Gentles e, Yves d'Udekem a,b, Dominica Zentner a,f,g,⁎,1 a
Murdoch Children's Research Institute, Melbourne, Australia Department of Paediatrics, The University of Melbourne, Melbourne, Australia Department of Cardiology, Royal Prince Alfred Hospital, Sydney, Australia d Sydney Medical School, University of Sydney, Sydney, Australia e The Paediatric and Congenital Cardiac Service, Starship Children's Hospital, Auckland, New Zealand f Department of Cardiology, Royal Melbourne Hospital, Melbourne, Australia g Department of Medicine, Royal Melbourne Hospital, University of Melbourne, Melbourne, Australia b c
a r t i c l e
i n f o
Article history: Received 15 May 2019 Received in revised form 16 July 2019 Accepted 14 August 2019 Available online xxxx Keywords: Heart Defects, Congenital Fontan procedure Pregnancy Embolism and thrombosis Heart Outcomes
a b s t r a c t Background: Pregnancy in women with a Fontan circulation is understood to have a significantly increased risk of maternal morbidity and mortality. Potential longer-term effects on health outcomes remain unknown. Aim: Ascertainment of adverse events, post-delivery, in women from the Australia and New Zealand (ANZ) Fontan Registry. Comparator data from women without pregnancy and men was utilised. Methodology/results: Living adults were identified (n = 263 women, 280 men) with Registry analysis ascertaining cardiac morphology, type of Fontan procedure, pregnancy and significant cardiac adverse events. Data are described descriptively for morphology and as median (inter quartile range) for other variables. Thirty women reported 45 pregnancies beyond 20 weeks' gestation. Despite being older, these women, predelivery, had the lowest cardiac morbidity burden, compared to both male and female controls (OR = 3.25 (95% CI 1.44–8.35, p = 0.008 and OR = 2.59 (95% CI 1.14–6.70), p = 0.03). These differences were not present post-delivery. Median follow-up time post-delivery was 3.6 (1.2–7.5) years. Thrombus documentation and thromboembolic events were more common post-delivery, although differences were not significant in a propensity analysis model (adjusting for post-Fontan follow-up time, age and type of Fontan). There were no differences in arrhythmia, cardioversion, heart transplant, Fontan conversion or mortality between women postdelivery and either male or female controls. Conclusion: The apparent increase in thromboembolic events post pregnancy requires further investigation in a larger group of women over a longer period of follow-up time. This may have potential implications for postpartum anticoagulation, and for future health outcomes. © 2019 Elsevier B.V. All rights reserved.
1. Introduction With improved survival an increasing number of women with a Fontan circulation are now of childbearing age. This poses unique questions regarding the ability of the Fontan circulation to meet the physiological challenge of pregnancy, and whether pregnancy has a long-term sequalae for maternal health outcomes. Normal pregnancy is a hypercoagulable state [1] and is associated with a measurable, though subclinical, decrease in cardiac function [2]. ⁎ Corresponding author at: Department of Cardiology, The Royal Melbourne Hospital, Vic, Australia. E-mail address: [email protected] (D. Zentner). 1 Equal first authors.
These issues may be particularly significant for women with a Fontan circulation, given the increased risk of arrhythmias [3–5], ventricular dysfunction [6,7] and heart failure [4], and thromboembolic events [3,8] in the Fontan population. Guidelines [9,10] state that pregnancy in these women is associated with significant risk of maternal morbidity and mortality. Complications include arrhythmia, heart failure and obstetric haemorrhage [11–13]. Though fetal mortality [11,12] is described, no reports of maternal mortality, during pregnancy, for women with a Fontan circulation exist in the current literature. It is therefore unknown whether the guidelines are overly cautious, or mortality events have not been published. The potential for long-term sequelae from pregnancy in women with the Fontan circulation remains unknown. Limited studies have reported late maternal deaths [13,14], and complications such as
https://doi.org/10.1016/j.ijcard.2019.08.039 0167-5273/© 2019 Elsevier B.V. All rights reserved.
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
2
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
arrhythmias, thromboembolic complications and Fontan failure [15]. However, the absence of comparator cohorts in these studies makes interpretation of long-term outcomes in this population difficult. Pre-conception counselling is a class 1 recommendation for women with CHD [16]. Ideally, this should be evidence based and include an understanding of the potential for delayed morbidity and mortality sequalae from pregnancy. As men have been observed to fare less well with a Fontan circulation [17] they may serve as a control group in studies for acceptable, even if higher, adverse event rates than non pregnant female controls. Thus, we sought to retrospectively compare morbidity and mortality, within the Australia and New Zealand (ANZ) Fontan Registry, between women who had and women who had not, experienced pregnancy, and additionally compared their outcomes to those in men. 2. Methods 2.1. Patient selection The ANZ Fontan Registry collects medical correspondence for patients who have had their Fontan procedure in either Australia or New Zealand, or who are followed up in this region. The governance of the Registry has been described in depth previously [18]. Authorization for ongoing retrospective analysis of the data held in the Registry was included in the initial design of the Fontan registry, and consequently approval for this study was as part of the ongoing ethics approval granted for the Fontan Registry, (HREC36260, HREC1888EP, NTX/10/ 07/064). Data for this study were extracted from the Registry for all participants N18 years of age as of the 1st of January 2017 (n = 689). Exclusion was then undertaken for: individuals who died before the age of 18 (n = 36), those with no medical data recorded beyond the age of 18 (n = 106), and patients whose Fontan was taken down prior to turning 18 (n = 2). Two women who had previously revealed a pregnancy but not consented to share data, and had no post partum follow up data in the Registry, were also excluded. This left a final cohort of 543 (280 males and 263 females). The female cohort was divided into the pregnancy group, being women who had recorded a pregnancy beyond 20 weeks gestation (n = 30), and the female control group, who either recorded no pregnancy or only pregnancies b20 weeks duration (n = 233). Utilisation of 20 weeks gestation was based on a low likelihood that women would have reached that gestation without seeing their cardiologist. Additionally, it forms the gestational age at which, legally, a delivery is considered a birth in Australia and New Zealand and must be registered [19,20]. Women were not excluded/ included on the basis of earlier miscarriage, as this is often poorly captured in medical records [21]. Furthermore miscarriage may not be as well recognised in a cohort with an increased frequency of irregular menstrual cycling [22]. Variables ascertained for the female and male cohorts were original cardiac morphology, age at Fontan operation, type of Fontan operation, post Fontan follow up time and documented significant cardiac and Fontan related complications. These were extracted from ANZ Registry data as previously described [3], and included NYHA class, oxygen saturation, arrhythmias, thromboembolic events, cardioversions, pacemaker insertion, Fontan conversion, heart transplantation and death. Additionally, medications and echocardiogram report data were ascertained and clinical letters revaluated for the female cohort. 2.2. Data analysis Participant data was exported from REDCap (Research Electronic Data Capture, Fontan Registry data) and analysed with SPSS Version 24.0 (IBM Corp, Armonk, NY) and R version 3.2.3 (R Foundation, Vienna, Austria) software.
Continuous variables are presented as median and interquartile range (IQR), and categorical variables as percentages. Descriptive analysis, nonparametric analysis (Median test) and Fisher exact test were used. Cardiac morbidity load was determined as the number of events per person year, including all above stipulated endpoints. In women who had had a pregnancy, this was broken down into events pre and post-delivery. Kaplan Meier analysis was used to generate data depicting freedom from heart transplant, Fontan conversion and death. The cumulative incidence of arrhythmia, cardioversion and thromboembolic events was compared between the pregnancy and control cohorts, with pregnancy accounted for as a time-dependent covariate. Cox regression was used to test for differences in hazard ratios for events between the pregnancy group and the control groups. Exploratory propensity scoring was then undertaken for variables that appeared to differ after Cox Regression. A p value of b0.05 was considered to be statistically significant. Confidence intervals are presented where possible. Last follow up recorded was the date of the last appointment or the date a Registry documented event was recorded, if later. For analysis women were considered to be pre delivery, until the date of their first (or only) delivery and post delivery, for any subsequent event. 3. Results 3.1. Demographic data The pregnancy group were older (p = 0.002) with a longer postoperative follow up time (p = 0.001). Type of Fontan procedure, and predominant ventricular morphology did not statistically differ between the groups, although there were fewer extra cardiac conduit Fontan circulation patients in the pregnancy group (Table 1). 3.2. Cardiac morbidity load Adverse cardiac events (arrhythmias, cardioversions, thromboembolic events, pacemaker insertion, heart transplant, Fontan conversion and death) were combined to create a cardiac morbidity load score, utilised to calculate events per year (Fig. 1). For the pregnancy group, this was further divided into events ‘pre’ and ‘post’ delivery. The incidence of events was lower in the pregnancy women, predelivery, in comparison to both male and female controls (OR = 3.25 (95% CI 1.44–8.35, p = 0.008 and OR = 2.59 (95% CI 1.14–6.70), p = 0.03 respectively). A greater proportion of women in the pre-delivery group reported no cardiac events (n = 23/30, 77%) compared to the male controls (n = 145/280, 52%, p = 0. 01), and to the female controls (n = 129/233, 55%, p = 0.03). This is despite the median age of the predelivery cohort being significantly older than the female and male controls (28 (25.5–30.6) years vs 25.0 (20.5–29.3), p = 0.02 and 28 (25.5–30.6) years vs 24 (20.7–29.4), p = 0.0002). When compared post partum, the incidence of events in postdelivery women compared with female controls and male controls was no longer statistically significantly different (OR = 1.40 (95% CI 0.56–3.82), p = 0.5, OR = 1.75 (95% CI 0.70–4.74), p = 0.2 respectively).The post-delivery group had a higher number of events, however, than both their own comparative pre-delivery data and female controls, ranging from 0 to 1.2 events per person year. This morbidity event rate was the same as that seen in the male cohort. 3.3. Pregnancy outcome Forty women reported a total of 82 pregnancies (Table 2), ascertained from both self-reported data (questionnaire and follow up telephone surveys, previously reported [11]) and clinical letters. Median maternal age at first delivery N20 weeks was 28.02 years (IQR 25.47–30.58). Women reported experiencing up to 10 pregnancies (a single individual, who experienced 7 miscarriages). The median number of successful pregnancies was 1 (IQR 1–2).
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
3
Table 1 Baseline characteristics. Variable
Pregnancy group (n = 30)
Female controls (n = 233)
Male controls (n = 280)
Statistical comparison
Age at last follow up, median (IQR), y Post-operative follow-up time, median (IQR), y Type of Fontan, n (%) Atriopulmonary Lateral tunnel Extra cardiac conduit Ventricle morphology, n (%) Right Left Indeterminate/biventricular Primary diagnosis, n (%) Tricuspid atresia Double inlet left ventricle Double outlet right ventricle Atrioventricular canal or AVSD Pulmonary atresia with VSD Pulmonary atresia with intact ventricular septum HLHS Ebstein's anomaly ccTGA Other
32.9 (26.6–37.3) 22.7 (20.5–26.6)
25.0 (20.5–29.3) 19.5 (15.1–23.9)
24.0 (20.7–29.4) 19.3 (15.0–23.5)
p = 0.002a p = 0.001a p = 0.14b
14 (47) 13 (43) 3 (10)
77 (33.0) 88 (37.8) 68 (29.2)
87 (31.1) 106 (37.9) 87 (31.1)
4 (13) 24 (80) 2 (7)
48 (20.6) 161 (69.1) 24 (10.3)
75 (26.8) 180 (64.3) 25 (8.9)
8 (27) 10 (33) 2 (7) 1 (3) 4 (13) 1 (3) 0 0 3 (11) 1 (3)
72 (30.9) 46 (19.7) 37 (15.9) 19 (8.2) 16 (6.9) 5 (2.1) 6 (2.6) 4 (1.7) 14 (6.0) 14 (6.0)
77 (27.5) 58 (20.7) 47 (16.8) 15 (5.4) 8 (2.9) 15 (5.4) 8 (2.9) 2 (0.7) 20 (7.1) 30 (10.7)
p = 0.3b
Abbreviations: AVSD indicates atrioventricular septal defect; ccTGA, congenitally corrected transposition of the great arteries; HLHS, hypoplastic left heart syndrome; IQR, interquartile range; and VSD, ventricular septal defect. a Non-parametric comparison, median test. b Fisher's exact test.
3.4. Functional status at last follow up I. NYHA class A recent NYHA class was recorded for 200 women in the female control cohort (86%) and 22 in the pregnancy group (73%). In the 200 female controls, 115 (57.5%) were NYHA Class 1, 78 (39%) were NYHA
Class 2, 6 (3%) were NYHA Class 3 and one woman (0.5%) was NYHA Class 4. Comparatively, in the 22 women in the pregnancy group, 15 (68%) had an NYHA recording pre-delivery, a median 1.2 years prior to pregnancy and 10 (67%) were NYHA Class 1, while 5 (33%) were NYHA Class 2. A median 3.62 years (IQR 1.19–7.50) post-delivery, 12 (54.5%) were NYHA Class 1, 9 women (41%) were NYHA Class 2 and one
Fig. 1. Number of adverse cardiac events per person year. Adverse cardiac events includes arrhythmias, cardioversions, thromboembolic events, pacemaker insertion, heart transplant, Fontan conversion and death, presented as per year of follow up data. The pregnancy group was subdivided, into a before delivery group; showing the number of events seen from birth to last recorded follow up before delivery and an after delivery group; showing the number of events seen post-delivery to last recorded follow up. Four women in the after delivery group had no data recorded post-delivery, and hence, were excluded from the after delivery group.
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
4
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
Table 2 Pregnancy outcome and maternal echocardiogram results at last follow up. Pregnancy outcome
Pregnancy outcome • Termination • Miscarriage • Delivery N20 weeks' gestation • Still born Baby outcome (live birth) (n = 43) • Neonatal deaths
Pregnancy group (30 women, 69 pregnancies)
Female controls (233 women, 13 pregnancies)
2 (3%) 22 (32%) 42 pregnancies/43 babies (1 set twins) (61%) 3 (4%)
3 (23%) 10 (77%)
3 (1 singleton, 1 set of twins)
Maternal echocardiogram results Variable
Pregnancy group (post-delivery) (n = 24)
Statistical Female controls (n = comparison 178)
Echocardiogram (y) post op, median (IQR) Echocardiogram (y) post pregnancy, median (IQR) Subjective ventricular systolic function, n (%)a Normal Reduced (mild – severe) AV regurgitation, n (%)a None/trivial Mild – moderated
24.4 (21.2–26.7)
20.6 (15.8–24.8) NA
3.5 (2.0–6.7)
p = 0.02b
143 (83.6) 28 (16.4)
11 (46) 13 (54)
70 (43.0) 92 (57.0)
IV. Echocardiogram
– p = 0.39c
18 (75) 6 (25)
Anticoagulation data (consisting of both antiplatelet and anticoagulation medications for the purpose of this analysis) was available for 193 female controls (83%), and 22 women in the pregnancy group, post delivery (73%). Two women in the pregnancy group (9%), and two women in the female control group were taking 2 or more anticoagulation medications (1%). No anticoagulation was prescribed for 3 (14%) women in the pregnancy group, and 12(6%) women in the female control cohort. At follow up, the commonest anticoagulant was warfarin, utilised by 40% (n = 9/22) of women in the pregnancy group and 55% (n = 107/193) of the female controls, p = 0.26. Alternate therapies to warfarin prescribed post-delivery were, aspirin (n = 6/22), clexane (n = 1), apixaban (n = 1), or dual therapies of warfarin and aspirin (n = 1), or aspirin and clexane (n = 1). Comparatively, in the female control cohort, aspirin was used in 35% (n = 68 /193) aspirin, with the remainder taking clexane (n = 1), clopidogrel (n = 1), rivaroxaban (n = 1), apixaban (n = 1), dual therapy with warfarin and aspirin (n = 1) or triple therapy with aspirin, warfarin and clopidogrel (n = 1).
p = 0.82c
a Owing to incomplete echocardiogram data within reports for some patients, percentage does not reflect column heads. b Calculated using non-parametric comparison, median test. c Calculated using Fisher‘s exact test. d No cases of severe AV valve regurgitation recorded.
Echocardiogram reports were available for 178 (76%) women in the female control group, and post-delivery echocardiogram recordings were available for 24 (80%) in the pregnancy group. At last echocardiogram, two (1.1%) women in the female control group had an intracardiac or Fontan circuit thrombus detected, compared to three (13%) woman in the pregnancy group post-delivery, p = 0.01. Of these, a clinical thromboembolic event was then recorded for 2 of the 3 pregnancy group women, only. In contrast, in the remaining women the thrombus was described as an incidental finding, with no clinical sequelae and not resulting in any change to medical therapy. 3.5. Late adverse cardiac events
woman (4.5%) was NYHA Class 3. Three women with NYHA data experienced a deterioration of one NYHA class post-delivery (13.6%). Of the 228 males (81%) with a recent NYHA recording, 148 (65%) were NYHA Class 1, 69 (30%) were NYHA Class 2, 10 (4%) were NYHA Class 3 and one male (0.4%) was NYHA class 4. II. Oxygen saturations
Oxygen saturation data was available for 132 (56%) women in the female control group, with a median saturation of 92% (83–95), and 159 (57%) males, with a median saturation of 93% (90–95). The lowest saturation in the female control cohort was 76% and in the male cohort 80%. In the pregnancy group, 20 women (67%) had saturation recordings available. Only 8 of the 20 pregnancy women had a pre-delivery saturation recorded, median 91.5% (90–94). The lowest pre-delivery saturation recorded was 85% in a woman 1 year before a livebirth at 31 weeks. In the 20 women with a post-delivery recording, the median saturation was 92% (89.5–92). III. Medication
Current medication data were available for 203 women in the female control group (87%) and for 23 (77%) women in the pregnancy group, a median of 3.04 (IQR 1.04–7.08) years post-delivery. The median number of cardiac medications was 1 in both groups, ranging from 0 to 6 in the pregnancy group and 0 to 8 in the control female cohort. The number of women who were on no cardiac medications was 7 (3%) in the control female cohort, and 3 (13%) in the pregnancy group.
Follow up post pregnancy, a median 3.6 years (IQR 1.2–7.5) was recorded in 26 (87%) of women. Pre-delivery, there had been a total of 22 events reported in 7 women (23%). Follow up for a minimum postpartum time of 2 years was available in 18 women, of whom 7 (38%) had experienced 26 adverse events post-delivery. Median follow up time of these 7 women with an adverse event postpartum was 7.5 years (3.9–14.4). Comparatively, among the control cohort, 104 (45%) women in the female control group, and 135 (48%) of men had experienced at least one adverse event throughout their life, (arrhythmia, thromboembolic event, cardioversion, pacemaker insertion, heart transplant, Fontan conversion or death). I. Thromboembolic events
In the pregnancy group, post-delivery, 4 thromboembolic events occurred in 3 women with an atriopulmonary Fontan. One woman experienced two strokes, one year apart, while the remaining two women experienced a right atrial thrombus with pulmonary embolism (n = 1), and right atrial thrombus causing pulmonary emboli and splenic infarct, identified post-mortem as cause of death (n = 1). Contemporaneous anticoagulation was aspirin (n = 1), warfarin (n = 2) and noncompliance with prescribed anticoagulation (n = 1). Prior to pregnancy, these women had no history of TE nor had a thrombus being reported on echocardiogram. The difference in the incidence of thromboembolic events between the pregnancy group and female and male groups is shown in Fig. 2. When assessed according to age, the pregnancy group had an increased incidence of thromboembolic events: HR (95% CI): 4.84 (1.08–21.6) p = 0.04. However, this was not statistically significant when analysed
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
5
No difference was observed between the control groups (male and female) and the pregnancy group in the time from Fontan operation to first arrhythmia, HR (95% CI): 1.56 (0.55–4.56) p = 0.4 or the age at which the first arrhythmia occurred, HR (95% CI): 1.48 (0.44–4.94) p = 0.5. Similarly, no difference was observed between the three groups in the incidence of first cardioversion, both when analysed as time since Fontan operation, HR (95% CI): 0.46 (0.06–3.43) p = 0.4 or age, HR (95% CI) 0.32 (0.04–2.49) p = 0.3. At last follow up, a pacemaker had been implanted in 40 (14%) males and in 30 (13%) women in the female control group. In the pregnancy group, 7 (23%) of women had a pacemaker. Implantation preceded pregnancy in 5 and post-dated pregnancy in 2. III. Fontan conversion, heart transplant and death - impact on longterm survival
Fontan failure endpoints were seen in 3 (10%) women in the pregnancy group, with 1 death 2.6 years post-delivery, 1 heart transplant 6 years post-delivery and 1 Fontan conversion 0.9 years post-delivery. However, survival free from Fontan conversion, heart transplant and death did not differ between the control female and male groups and the pregnancy group, either when measured in years since Fontan, HR (per year increase): 1.10 (95% CI 0.32–3.76) p = 0.9; or age, HR (per year increase):0.88 (95% CI 0.24–3.16) p = 0.8. 4. Discussion
Fig. 2. Cumulative incidence of thromboembolic events according to gender and pregnancy group, measured across A) years since Fontan and B) age. Pregnancy was accounted for as a time-dependent covariate, with women transitioning from the not pregnant (female control) to pregnant group after their first delivery.
according to years post Fontan operation: HR (95% CI): 3.12 (0.87–11.1) p = 0.08. Exploratory propensity scoring was undertaken to account for the effect of the differences in age and time since Fontan, on the difference in TE events, between the pregnancy and no-pregnancy groups. Matching age at follow up, time since Fontan and Fontan type, reduced the study cohort to 19 (63%) women in the pregnancy group and 48 (17%) women in the female control group. TE events were also reduced from 36 to 10. Propensity scoring analysis removed the difference between the pregnant and female control subgroups with respect to thromboembolic events, when assessed according to age; coefficient factor 4.77: (95% CI: 0.6692–34.07), p = 0.119. II. Arrhythmias and cardioversion
Five women (19%) in the pregnancy group experienced 12 arrhythmia events post-delivery. Pre-delivery, only one woman had experienced an arrhythmia. The median time between pregnancy and the first post-delivery arrhythmia was 7.38 (1.08–11.80) years.
This study explores whether pregnancy may result in adverse outcomes for women with a Fontan circulation, and is the first to include comparison to individuals with a Fontan circulation who have not undergone pregnancy. In this binational registry cohort, maternal cardiac events appeared least likely in the women in the pregnancy group, pre-delivery. Coding in this study was such that gestational cardiac events were counted as pre-delivery cardiac events. As pregnancy is an accepted high risk time [23] for the endpoints collated, the significance of the decreased incidence of events in the pre-delivery cohort is of further clinical relevance. Furthermore, the significantly greater age of the women who underwent pregnancy, in a population where time is understood to significantly drive adverse events [3], suggests that those undergoing pregnancy are a selected group with a well functioning Fontan circulation. This is not surprising given high risk patients are likely to be counselled against pregnancy. Current guidelines do not clearly detail how to individualise the risk of pregnancy for women with a Fontan circulation [9]. While the just released ESC guidelines [24] recommend counselling against pregnancy in any woman post a Fontan operation with any associated complication, the CARPREG II guidelines offer a cumulative scoring assessment where points are accrued for predictors of poorer maternal outcomes [25]. The scoring system includes a prior history of cardiac events, a well recognised risk factor for further adverse complications during pregnancy in women with CHD [10,26–29]. A recognised determinant of live birth is maternal oxygen saturation [30,31]. In our cohort, and in contrast to both the female controls and male controls, no woman who attained a gestational age N 20 weeks had a pre-delivery saturation recording below 85%. This accords with a prior publication (24) reporting a small group of women (n = 8) with a Fontan circulation with saturations below 85% whom all experienced a miscarriage. The data in this paper may suggest that clinical assessment has historically been incorporating relevant factors when risk stratifying these women, and, further, that patients may be compliant with advice. However, as both unmanaged chronic disease [32,33] and impaired haemodynamics [34] are risk factors for miscarriage and subfertility, women in the no-pregnancy group, who experienced a greater cardiac
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
6
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx
morbidity load, may have had increased difficulty in conceiving and carrying a pregnancy N20 weeks. Post-delivery, those same women with a decreased incidence of events compared to both male and female controls pre-delivery, no longer demonstrated this difference. Additionally, the events per person year data is highest for post-delivery women and men, though the numbers are small. This suggests that pregnancy may modulate future maternal health events, as despite the short follow up time of 3.6 years post pregnancy, the event rate increased in that subgroup. This highlights the importance of adequate pre pregnancy counselling and assessment. It appears realistic to counsel women that their future health quality may reduce post a pregnancy, though the data would suggest that that reduction is still within the spectrum of well being for these patients. Despite the increased time between their Fontan operation and last echocardiogram, the pregnancy group were not shown to have greater ventricular dysfunction or AV valve regurgitation. This is surprising given reports in other CHD that pregnancy may expedite ventricular dysfunction [35–37], but may reflect the short post pregnancy follow up time in this study. Additionally, it is recognised that echocardiographic assessment of the single ventricle is challenging [38,39]. It is also possible that pre conception assessment and counselling has largely seen women with better ventricular function pursue pregnancy, as supported by other CHD literature [26,28,40]. This study suggests increased thrombus detection after pregnancy. However the significant difference found with cox regression for thromboembolic events, was not seen in the propensity score model. While this may be explained by the reduced sample size and hence event rate utilised in that comparison, it may also reflect that the difference observed relates to other variables, such as time since Fontan. Nonetheless, the potential increase in TE events demonstrated raises a concern, as clinically diagnosed thrombus has been noted to be a significant predictor of later thromboembolic death, heart failure, transplantation and all cause mortality [41]. Other studies [13,15], have also reported thromboembolic events post pregnancy in women with a Fontan circulation. It is well established that pregnancy and the postpartum period are high risk times for thromboembolic events [42–45], and these have been reported in the Fontan population [12,46]. However, none of the TE events in our population occurred within the puerperium, suggestive of a lingering effect of pregnancy in this population. It is feasible that an increased risk of thromboembolic events results from pregnancy, in a population with raised venous pressures [6], due to pregnancy-related weight retention [42] and/or varicose vein development [47]. Experiencing a TE event during pregnancy is known to increase the risk of future TE events [48]. It is possible that some women with post-delivery TE events had an unrecognised TE event in pregnancy. Our study did not demonstrate an increased incidence of arrhythmias, cardioversion, Fontan conversion, heart transplant or death after pregnancy during this short post- delivery follow up time. As many of these adverse events appear likely to be time-dependent phenomenon, these women should be followed prospectively, and in a matched manner, to allow documentation of long term outcomes.
5. Limitations This study is limited by the small sample size and inherent challenges of an unmatched, retrospective study. Data was restricted to the information documented in the ANZ Fontan Registry. Echocardiogram reports were unavailable for 24% of women in the female control group and 20% of women in the pregnancy group. Furthermore, some clinical letters did not outline endpoints such as oxygen saturations, NYHA class or current medications. The Registry collects data from various centres throughout Australia and New Zealand. We recognise that the echocardiogram protocol and reporting may therefore be different between the different centres.
6. Conclusion This retrospective cohort study did not show pregnancy to be associated with an increase in ventricular dysfunction, arrhythmias, cardioversions, heart transplant or mortality, within the time frame of data available. However, a slightly higher risk of thromboembolic events was seen post-delivery. To date anticoagulation management for women with a Fontan circulation differs widely. If the concerns raised by this paper are confirmed, there may be a role, particularly for post-delivery anticoagulation. Further work is also needed to understand how clinicians undertake pre-pregnancy evaluation. It is unclear whether the better health status of women who subsequently have a pregnancy reflects accurate risk assessment and prognostication, or an increased miscarriage event rate in the women who do not have a pregnancy. We are currently setting up a Fontan pregnancy database, as an international collaborative effort, that will allow a prospective, longitudinal and matched study of post pregnancy outcomes in this population. This is essential in order to provide clinicians, women, and their families, with increasingly accurate data that addresses concern about the possible effect of pregnancy on the future health of these women. Acknowledgements The authors thank all the participants enrolled in the Fontan Registry, the research assistants for their support in maintaining the Australian and New Zealand Fontan Registry and the support provided to the Murdoch Children's Research Institute by the Victorian Government's Operational Infrastructure Support Program. Funding This work was supported by a National Health and Medical Research Council (NHMRC) Partnership grant (1076849). Prof. Yves d'Udekem is a Clinician Practitioner Fellow of the NHMRC (1082186). Declaration of competing interest Prof Yves d'Udekem is a consultant for the companies MSD and Actelion. The remaining authors report no relationships that could be construed as a conflict of interest. References [1] P. Soma-Pillay, N.-P. Catherine, H. Tolppanen, A. Mebazaa, H. Tolppanen, A. Mebazaa, Physiological changes in pregnancy, Cardiovasc. J. Afr. 27 (2) (2016 Mar-Apr) 89–94 (08/31/received, 03/04/accepted;27(2):89-94. PubMed PMID: PMC4928162). [2] D. Zentner, M. du Plessis, S. Brennecke, J. Wong, L. Grigg, S.B. Harrap, Deterioration in cardiac systolic and diastolic function late in normal human pregnancy, Clin. Sci. (Lond.) 116 (7) (2009) 599–606 Apr. (PubMed PMID: 18855763). [3] C.L. Poh, D. Zannino, R.G. Weintraub, D.S. Winlaw, L.E. Grigg, R. Cordina, et al., Three decades later: the fate of the population of patients who underwent the Atriopulmonary Fontan procedure, Int. J. Cardiol. 231 (2017 Mar 15) 99–104 PubMed PMID: (28100430). [4] Y. d'Udekem, A.J. Iyengar, J.C. Galati, V. Forsdick, R.G. Weintraub, G.R. Wheaton, et al., Redefining expectations of long-term survival after the Fontan procedure: twentyfive years of follow-up from the entire population of Australia and New Zealand, Circulation 09 (11 Suppl 1) (2014 Sep) 130 (S32-8. PubMed PMID: 25200053. Epub 2014/09/10. eng). [5] T.A. Carins, W.Y. Shi, A.J. Iyengar, A. Nisbet, V. Forsdick, D. Zannino, et al., Long-term outcomes after first-onset arrhythmia in Fontan physiology, J. Thorac. Cardiovasc. Surg. 152 (5) (2016 Nov) 1355–1363.e1 PubMed PMID: 27751239. Epub 2016/10/ 19. eng. [6] M. Gewillig, S.C. Brown, The Fontan circulation after 45 years: update in physiology, Heart 102 (14) (2016 Jul 15) 1081–1086 PubMed PMID: 27220691. PMCID: (PMC4941188). [7] M. Grattan, L. Mertens, Mechanics of the functionally univentricular heart—how little do we understand and why does it matter? Can. J. Cardiol. 32 (8) (2016) 8//. (1033.e11-.e18). [8] D.N. Rosenthal, A.H. Friedman, C.S. Kleinman, G.S. Kopf, L.E. Rosenfeld, W.E. Hellenbrand, Thromboembolic complications after Fontan operations, Circulation 92 (9) (1995) 287–293.
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039
E. Moroney et al. / International Journal of Cardiology xxx (xxxx) xxx [9] V. Regitz-Zagrosek, C. Blomstrom Lundqvist, C. Borghi, R. Cifkova, R. Ferreira, J.-M. Foidart, et al., ESC guidelines on the management of cardiovascular diseases during pregnancy. The Task Force on the Management of Cardiovascular Diseases during Pregnancy of the European Society of Cardiology (ESC), Eur. Heart J. 32 (24) (2011) 3147–3197. [10] M.M. Canobbio, C.A. Warnes, J. Aboulhosn, H.M. Connolly, A. Khanna, B.J. Koos, et al., Management of pregnancy in patients with complex congenital heart disease: a scientific statement for healthcare professionals from the American Heart Association, Circulation 135 (8) (2017 Jan 12) e50–e87 (PubMed PMID: 28082385). [11] D. Zentner, A. Kotevski, I. King, L. Grigg, Fertility and pregnancy in the Fontan population, Int. J. Cardiol. 208 (2016 Apr 01) 97–101 PubMed PMID: (26836494). [12] M. Gouton, J. Nizard, M. Patel, F. Sassolas, M. Jimenez, J. Radojevic, et al., Maternal and fetal outcomes of pregnancy with Fontan circulation: a multicentric observational study, Int. J. Cardiol. 187 (2015) 84–89 PubMed PMID: (25828319). [13] M.M. Canobbio, F. Cetta, C. Silversides, C. Warnes, J. Aboulhosn, J. Colman, Pregnancy after Fontan operation: early and late outcomes, J. Am. Coll. Cardiol. 61 (10) (2013) E427. [14] K.N. Pundi, K. Pundi, J.N. Johnson, J.A. Dearani, C.R. Bonnichsen, S.D. Phillips, et al., Contraception practices and pregnancy outcome in patients after Fontan operation, Congenit. Heart Dis. 11 (1) (2016) 63–70. [15] S. Arif, A. Chaudhary, P.F. Clift, R. Katie Morris, T.J. Selman, S.E. Bowater, et al., Pregnancy outcomes in patients with a fontan circulation and proposal for a risk-scoring system: single centre experience, J. Congenital Cardiol. 1 (1) (2017) 10 December 08. [16] K.K. Stout, C.J. Daniels, J.A. Aboulhosn, B. Bozkurt, C.S. Broberg, J.M. Colman, et al., AHA/ACC guideline for the management of adults with congenital heart disease, Circulation 0 (0) (2018) 2018. (CIR.0000000000000603). [17] H.M. Burkhart, J.A. Dearani, D.D. Mair, C.A. Warnes, C.C. Rowland, H.V. Schaff, et al., The modified fontan procedure: early and late results in 132 adult patients, J. Thorac. Cardiovasc. Surg. 125 (6) (2003) 1252–1258 2003/06/01/. [18] A.J. Iyengar, D.S. Winlaw, J.C. Galati, T.L. Gentles, R.G. Weintraub, R.N. Justo, et al., The Australia and New Zealand Fontan Registry: description and initial results from the first population-based Fontan registry, Intern. Med. J. 44 (2) (2014 Feb) 148–155 PubMed PMID: 24393144. Epub 2014/01/08. eng. [19] Australian Institute of Health and Welfare. Mothers & babies Glossary, , [Internet]. Available from https://www.aihw.gov.au/reports-statistics/population-groups/ mothers-babies/glossary 2017. [20] Ministry of Health, Miscarriage and stillbirth [internet], Available from: https:// www.health.govt.nz/your-health/pregnancy-and-kids/services-and-support-during-pregnancy/miscarriage-and-stillbirth 2017. [21] D. Herbert, J. Lucke, A. Dobson, Pregnancy losses in young Australian women: findings from the Australian longitudinal study on women's health, Women's Health Issues, 19(1), 2009, pp. 21–29 , 2009/01/01/. [22] M.M. Canobbio, D.D. Mair, A.J. Rapkin, J.K. Perloff, B.L. George, Menstrual patterns in females after the Fontan repair, Am. J. Cardiol. 66 (2) (1990) 238–240 1990/07/15/. [23] A. Garcia Ropero, S. Baskar, J.W. Roos Hesselink, A. Girnius, D. Zentner, L. Swan, et al., Pregnancy in women with a Fontan circulation: a systematic review of the literature, Circ. Cardiovasc. Qual. Outcomes 11 (5) (2018) May. (e004575. PubMed PMID: 29752389. Epub 2018/05/13. eng). [24] V. Regitz-Zagrosek, J.W. Roos-Hesselink, J. Bauersachs, C. Blomström-Lundqvist, R. Cífková, M. De Bonis, et al., 2018 ESC guidelines for the management of cardiovascular diseases during pregnancy, Eur. Heart J. 39 (34) (2018) 3165–3241. [25] C.K. Silversides, J. Grewal, J. Mason, M. Sermer, M. Kiess, V. Rychel, et al., Pregnancy outcomes in women with heart disease. The CARPREG II study, 71 (21) (2018) 2419–2430. [26] O.H. Balint, S.C. Siu, J. Mason, J. Grewal, R. Wald, E.N. Oechslin, et al., Cardiac outcomes after pregnancy in women with congenital heart disease, Heart 96 (20) (2010 Oct) 1656–1661 PubMed PMID: 20937754. Epub 2010/10/13. eng. [27] M.A.M. Kampman, A. Balci, H. Groen, A.P.J. van Dijk, J.W. Roos-Hesselink, J.P. van Melle, et al., Cardiac function and cardiac events 1-year postpartum in women with congenital heart disease, Am. Heart J. 169 (2) (2015) 298–304 2015/02/01/. [28] S.C. Siu, M. Sermer, J.M. Colman, A.N. Alvarez, L.-A. Mercier, B.C. Morton, et al., Prospective multicenter study of pregnancy outcomes in women with heart disease, Circulation 104 (5) (2001) 515–521.
7
[29] W. Drenthen, E. Boersma, A. Balci, P. Moons, J.W. Roos-Hesselink, B.J.M. Mulder, et al., Predictors of pregnancy complications in women with congenital heart disease, Eur. Heart J. 31 (17) (2010) 2124–2132. [30] M. Cauldwell, P.J. Steer, S. Bonner, O. Asghar, L. Swan, K. Hodson, et al., Retrospective UK multicentre study of the pregnancy outcomes of women with a Fontan repair, Heart 104 (5) (2018 Mar) 401–406 (PubMed PMID: 28954835. Epub 2017/09/29. eng). [31] P. Presbitero, J. Somerville, S. Stone, E. Aruta, D. Spiegelhalter, F. Rabajoli, Pregnancy in cyanotic congenital heart disease. Outcome of mother and fetus, Circulation 89 (6) (1994) 2673–2676. [32] R.J. Bradley, M.P. Rosen, Subfertility and gastrointestinal disease: ‘unexplained’ is often undiagnosed, Obstet. Gynecol. Surv. 59 (2) (2004) 108–117 (PubMed PMID: 00006254-200402000-00023). [33] R. Pasquali, L. Patton, A. Gambineri, Obesity and infertility. Current opinion in endocrinology, Diabetes Obes. 14 (6) (2007) 482–487 (PubMed PMID: 01266029200712000-00011). [34] R.L. Bick, D. Hoppensteadt, Recurrent miscarriage syndrome and infertility due to blood coagulation protein/platelet defects: a review and update, Clin. Appl. Thromb. Hemost. 11 (1) (2005) 1–13 2005/01/01. [35] A. Guedes, L.A. Mercier, L. Leduc, L. Berube, F. Marcotte, A. Dore, Impact of pregnancy on the systemic right ventricle after a mustard operation for transposition of the great arteries, J. Am. Coll. Cardiol. 44 (2) (2004 Jul 21) 433–437 PubMed PMID: 15261944. Epub 2004/07/21. eng. [36] C.A. Kamiya, T. Iwamiya, R. Neki, S. Katsuragi, K. Kawasaki, T. Miyoshi, et al., Outcome of pregnancy and effects on the right heart in women with repaired tetralogy of fallot, Circ. J. 76 (4) (2012) 957–963 PubMed PMID: 22277318. Epub 2012/01/27. eng. [37] N. Tzemos, C.K. Silversides, J.M. Colman, J. Therrien, G.D. Webb, J. Mason, et al., Late cardiac outcomes after pregnancy in women with congenital aortic stenosis, Am. Heart J. 157 (3) (2009) 474–480 3//. [38] R. Margossian, M.L. Schwartz, A. Prakash, L. Wruck, S.D. Colan, A.M. Atz, et al., Comparison of echocardiographic and cardiac magnetic resonance imaging measurements of functional single ventricular volumes, mass, and ejection fraction (from the Pediatric Heart Network Fontan Cross-Sectional Study)††a list of participating institutions and investigators appears in the appendix, Am. J. Cardiol. 104 (3) (2009) 419–428 2009/08/01/. [39] R.V. Williams, R. Margossian, M. Lu, A.M. Atz, T.J. Bradley, M.J. Campbell, et al., Factors impacting echocardiographic imaging after the Fontan procedure: a report from the pediatric heart network Fontan cross-sectional study, Echocardiography (Mount Kisco, NY) 30 (9) (2013 04/25) 1098–1106 (PubMed PMID: PMC3732542). [40] J. Grewal, S.C. Siu, H.J. Ross, J. Mason, O.H. Balint, M. Sermer, et al., Pregnancy outcomes in women with dilated cardiomyopathy, J. Am. Coll. Cardiol. 55 (1) (2009 Dec 29) 45–52. [41] P. Khairy, S.M. Fernandes, J.E. Mayer, J.K. Triedman, E.P. Walsh, J.E. Lock, et al., Longterm survival, modes of death, and predictors of mortality in patients with Fontan surgery, Circulation 117 (1) (2008) 85–92. [42] F.A. Anderson, F.A. Spencer, Risk factors for venous thromboembolism, Circulation 107 (23 suppl 1) (2003) I–9-I-16. [43] J.A. Heit, C.E. Kobbervig, A.H. James, T.M. Petterson, K.R. Bailey, L.J. 3rd Melton, Trends in the incidence of venous thromboembolism during pregnancy or postpartum: a 30-year population-based study, Ann. Intern. Med. 143 (10) (2005 Nov 15) 697–706 (PubMed PMID: 16287790. Epub 2005/11/17. eng). [44] I.A. Greer, Thrombosis in pregnancy: maternal and fetal issues, Lancet 353 (9160) (1999 Apr 10) 1258–1265 PubMed PMID: 10217099. Epub 1999/04/27. eng. [45] A. Abdul Sultan, J. West, L.J. Tata, K.M. Fleming, C. Nelson-Piercy, M.J. Grainge, Risk of first venous thromboembolism in pregnant women in hospital: population based cohort study from England, BMJ [Br. Med. J.] (2013) 347. [46] M. Cauldwell, K. Von Klemperer, A. Uebing, L. Swan, P.J. Steer, S.V. Babu-Narayan, et al., A cohort study of women with a Fontan circulation undergoing preconception counselling, Heart 102 (7) (2016) 534–540. [47] M. Callam, Epidemiology of varicose veins, Br. J. Surg. 81 (2) (1994) 167–173. [48] A. Bergqvist, D. Bergqvist, A. Lindhagen, T. MÄTzsch, Late symptoms after pregnancy-related deep vein thrombosis, BJOG Int. J. Obstet. Gynaecol. 97 (4) (1990) 338–341.
Please cite this article as: E. Moroney, D. Zannino, R. Cordina, et al., Does pregnancy impact subsequent health outcomes in the maternal Fontan circulation?, International Journal of Cardiology, https://doi.org/10.1016/j.ijcard.2019.08.039