Pregnancy Outcomes in Women With Transposition of the Great Arteries After an Arterial Switch Operation

Key Points

Question

What are the outcomes of pregnancy in women with transposition of the great arteries after arterial switch surgery?

Findings

In this cohort study of 15 women, there were 24 completed pregnancies with live, healthy infants and no adverse maternal cardiac events. No neoaortic root progression was seen, and only 2 delivery modes were dictated by maternal cardiac considerations.

Meaning

In this analysis, pregnancy was well tolerated and maternal and infant outcomes were good in women with transposition of the great arteries and previous arterial switch surgery.

This cohort study seeks to determine the outcomes of pregnancy in women with transposition of the great arteries after an arterial switch operation, as well as outcomes of their offspring.

Abstract

Importance

A growing number of women are approaching childbearing age after arterial switch surgery for transposition of the great arteries. Prepregnancy counseling requires updated knowledge of the additional cardiovascular risks pregnancy poses for this cohort of women and the potential effect on their offspring; however, to our knowledge, this information is currently unknown.

Objective

To determine the pregnancy outcomes of women with transposition of the great arteries after an arterial switch operation, as well as the outcomes of their offspring.

Design, Setting, and Participants

This cohort study assessed women who had had arterial switch surgery from 1985 to the present and who were 16 years or older as of January 2018. All women with a previous arterial switch surgery for transposition of the great arteries with completed or ongoing pregnancy were included. Data were collected in a level 1 congenital cardiology center and joint obstetrics-cardiology clinic in Birmingham, United Kingdom.

Exposures

Patients were assessed before, during, and after pregnancy.

Main Outcomes and Measures

Adverse maternal cardiac events (arrhythmia, heart failure, aortic dissection, or acute coronary syndrome) and aortic root dilatation, aortic regurgitation, and left ventricular function before and after pregnancy were the main outcomes. Mode of delivery and fetal outcomes were considered secondary outcomes.

Results

A total of 25 pregnancies were identified in 15 women; 8 women had had 1 pregnancy, while 7 were multiparous. There were no adverse maternal cardiac events. Before pregnancy, 8 women (53%) had no aortic regurgitation, 1 (7%) had a trivial degree of regurgitation, 4 (26%) had mild regurgitation, and 2 (14%) had moderate regurgitation. After pregnancies, 1 woman (7%) had minor progression of aortic regurgitation. Five women (36%) had mild neoaortic root dilatation prepregnancy, but none developed progressive dilatation in the first year post-partum. A total of 24 pregnancies were completed by the end of the study, with all infants born alive and well. Nineteen modes of delivery were known; there were 7 cesarean deliveries (37%), of which 2 (11%) were recommended for aortic dilatation and 5 (26%) for obstetric indications or maternal choice.

Conclusions and Relevance

Pregnancy is well tolerated after arterial switch operation; no adverse maternal cardiac events or early progression of neoaortic root dilatation or aortic regurgitation were observed in this study. These results provide evidence to allow reassurance of women with previous arterial switch surgery who are planning pregnancies.

Introduction

The arterial switch operation (ASO) has been the surgical correction of choice for transposition of the great arteries (TGA) for more than 30 years.¹ Hence, there are a growing number of women at or approaching childbearing age. Pregnancy represents a state of sustained physiological stress,² which may exacerbate complications after ASO, including neoaortic root dilatation, aortic regurgitation, and myocardial ischemia arising from coronary abnormalities.^3,4 Much care is taken in prepregnancy counseling of these patients,⁵ but there is a paucity of contemporaneous information about pregnancy outcomes, because the largest (to our knowledge) published case study contains just 9 women with 13 successful pregnancies and 2 cardiac complications.⁶ We aimed to determine the pregnancy outcomes for a larger cohort of women after ASO.

Methods

Female patients with previous ASO who were being followed up at the Queen Elizabeth Hospital Birmingham were identified. Women who had completed pregnancies or were currently pregnant were identified from medical records and the joint obstetrics-cardiology clinic.

This study was registered with the institution’s Research and Development office. In accordance with the United Kingdom National Health Service National Research Ethics Service guidance, neither individual informed consent nor formal research ethics committee review was required, because the study was undertaken by the direct clinical team using information previously collected in the course of routine care.

Medical records were reviewed for demographic data. Imaging data from echocardiography were obtained before, during, and after pregnancy, if available. Left ventricular (LV) size and systolic function, aortic root dimensions at the sinuses of Valsalva, and severity of any aortic regurgitation, pulmonary valve stenosis or regurgitation, and tricuspid regurgitation were recorded. Aortic dilatation was defined as a diameter greater than the 95th centile for sex and age.^7,8 Neoaortic root z scores were calculated as both nonnormalized values⁹ and values normalized for body surface area, as described by Devereux et al.¹⁰ Valvular lesions were graded according to American Society of Echocardiography guidelines.¹¹ Maternal and fetal outcomes were obtained from medical records. Evidence of adverse maternal cardiac events (arrhythmias, heart failure, aortic dissection, and acute coronary syndromes) were sought. Data collection was completed from November 2017 until February 2018. The software used for data analysis was SPSS version 22 (IBM).

Results

Fifty-five women older than 16 years who had undergone ASO were identified. This group had a mean (SD) age of 23.6 (5.1) years (range, 17-41 years). Of these, 15 women (27%) had been or were pregnant with a total of 25 pregnancies during the data collection period. Twenty-four pregnancies had been completed by the end of the period, with 1 more ongoing. The mean (SD) age at first delivery was 23.0 (3.9) years (range, 17-31 years). Eight women had had 1 pregnancy, and 7 women were multiparous (of whom 4 had had 2 pregnancies each, and 3 had had 3 pregnancies each). Demographics are shown in Table 1.

Table 1. Baseline Demographics Before Pregnancy.

Variable	Patients, No. (%) (N = 15)
Age at arterial switch operation, median (IQR), mo	0.3 (0.2-5.2)
Associated cardiac lesions and interventions
Ventricular septal defect	5 (33)
Pulmonary artery band	4 (27)
Pulmonary artery interventions	6 (40)
Age at first delivery, mean (SD), y	23.0 (3.9)
Pregnancies, No.
1	8 (53)
2	4 (27)
3	3 (20)
Cardiac lesions before first pregnancy
Aortic regurgitation
None	8 (53)
Trivial	1 (7)
Mild	4 (27)
Moderate	2 (13)
Neoaortic root dilatation	5 (33)
Pulmonary stenosis
None	11 (73)
Mild	4 (27)
Pulmonary regurgitation
None	9 (60)
Trivial	2 (13)
Mild	3 (20)
Moderate	1 (7)
Tricuspid regurgitation
None	4 (27)
Trivial	6 (40)
Mild	3 (20)
Moderate	2 (13)

Abbreviation: IQR, interquartile range.

Ten women (67%) had a diagnosis of transposition with intact ventricular septum, 3 women (20%) had transposition with ventricular septal defects, and 2 (14%) had complex TGA with associated cardiovascular lesions (of whom 1 had right ventricular outflow tract obstruction and 1 had coarctation of the aorta). Most arterial switch operations (n = 10 [67%]) were performed in the neonatal period, apart from 5 patients (33%) who had undergone ASO at ages 3 months, 8 months, 2.5 years, 6.1 years, and 8.5 years. These 5 patients were born early in the era of arterial switch surgery; 4 of them, including 3 with TGA with ventricular septal defects, underwent prior pulmonary artery banding; the fifth, who had complex TGA, presented late, at 3 months of age. Six of the women (40%) had needed further surgery or interventions after ASO was completed and prior to their first pregnancy, with the most common indication (n = 4 [27%]) being branch pulmonary artery stenosis.

During and after their pregnancies, no patients experienced adverse cardiac events. Namely, there were no episodes of heart failure, arrhythmias, coronary ischemia, aortic dissection, or maternal deaths.

Twelve of 14 women (86%), including those who were primiparous or multiparous but excluding 1 with an ongoing pregnancy, had serial echocardiography with a study after the most recent pregnancy. Of these, only 1 had progression in the degree of aortic regurgitation from none to mild, between her first and third pregnancy, which occurred over a 12-year period. There was no change in neoaortic root dilatation for the 12 women with serial imaging, including the 5 women who had had mildly dilated neoaortic roots before pregnancy. The mean neoaortic root diameters and z scores before and after pregnancy are shown in Table 2. Three women underwent cardiac magnetic resonance imaging while pregnant to fully evaluate aortic dimensions; there was no progressive neoaortic root dilatation in these individuals.

Table 2. Aortic Root Dimensions and z Scores Before and After Pregnancy.

Patient Characteristic	Mean (SD) Values
	Prepregnancy (n = 15)			Postpregnancy (n = 12)
	Aortic Root Diameter, mm	z Score, Nonnormalized	z Score, Normalized for Body Surface Area	Aortic Root Diameter, mm	z Score, Nonnormalized	z Score, Normalized for Body Surface Area
All	33 (5.0)	1.85 (1.96)	1.35 (1.80)	34 (5.0)	2.19 (1.86)	1.73 (1.76)
No neoaortic root dilation	29 (2.8)	0.64 (1.32)	0.12 (1.06)	30 (2.3)	0.98 (1.05)	0.39 (0.91)
Neoaortic root dilatation	38 (2.0)	3.66 (1.25)	3.20 (0.89)	38 (2.0)	3.71 (1.51)	3.40 (0.96)

All 15 women had normal LV systolic function prior to their first pregnancy. Two patients were noted to have mildly dilated LV end diastolic dimensions, but these did not change after pregnancy.

One woman had a decline in LV function from normal to mildly impaired after her second pregnancy. Third-trimester echocardiography during her second pregnancy demonstrated normal LV function; mildly impaired LV systolic function was noted 8 months post-partum. However, this patient had a complex surgical history, having undergone 3 operations since ASO and prior to her 2 pregnancies, including a resection of a pulmonary artery aneurysm, a pulmonary valve replacement with right ventricular outflow tract patch enlargement, and right coronary artery bypass grafting.

Neonatal data were available for 22 of the 24 delivered pregnancies (92%). The mean (SD) gestation at delivery was 38.5 (2.0) weeks’ gestation (range, 35-42 weeks’ gestation). One delivery was preterm (<37 weeks’ gestation); another delivery involved spontaneous labor at 35 weeks’ gestation, with an emergency cesarean delivery for breech presentation. The mean (SD) birth weight was 3184 (447) g (range, 2270-4167 g). There were no neonatal deaths. No cardiac disease was noted in any children.

The mode of delivery was available from 19 deliveries (79%). Eight of the 19 women (42%) had a normal vaginal delivery, 4 (21%) had instrumental births, and 7 (37%) had cesarean deliveries. Two of the cesarean deliveries were emergency procedures for delay in the second stage and breech presentation, respectively. The remaining 5 cesarean deliveries were elective. Two were for breech presentations, 2 were advised for dilated aortic root dimensions, and 1 was for maternal choice.

Discussion

This study is to our knowledge the largest to date of pregnancy outcomes in women who have undergone arterial switch surgery for transposition of the great arteries. We found that pregnancy was well tolerated, with no adverse cardiac maternal events.

The largest (to our knowledge) previous study by Tobler et al⁶ found a 15% rate of cardiac events within their cohort, consisting of mechanical valve thrombosis and nonsustained ventricular tachycardia. They additionally found 2 women with reduction in their LV ejection fraction, which the present study found in 1 patient. In both cohorts, neoaortic root dimensions were stable during and after pregnancy. We additionally found 1 patient with minor deterioration in her degree of aortic regurgitation, from none to mild.

The patients studied by Tobler et al⁶ represent a more complex and earlier surgical cohort, which may account for the high rates (67%) of reinterventions after the ASO and maternal complications. Among the patients in this study, the rate of surgical reintervention was 40%.

In this cohort, only 2 cesarean deliveries were dictated by maternal cardiac considerations. However, there were fewer normal vaginal deliveries in our cohort (42.1%) compared with a national average of 59.4%, while there was a higher rates of cesarean deliveries (36.8%) vs the national average of 27.8%.¹²

Reimplantation of the coronary arteries remains one of the main challenges of ASO.¹³ However, despite coronary ischemia being a well-recognized late complication after ASO,³ none of these patients developed any coronary ischemia, even with the additional physiological stress and procoagulant state pregnancy confers.

The need for serial assessment of women with aortic root dilatation during pregnancy is extrapolated from observational studies in women with aortopathies that demonstrated increased rates of aortic dilatation and dissection during pregnancy.^14,15 The 5 patients in this study who had neoaortic dilatation prepregnancy all underwent postpartum imaging that did not demonstrate any further dilatation.

Limitations

This study is limited by the retrospective design and the small number of pregnancies that have occurred in these women to date. Additionally, we do not have further data on neonatal outcomes or infant growth and development. Understanding the long-term pediatric outcomes will require careful evaluation of these children as they grow. In addition, larger multicenter studies with longer postpartum follow-up will be required to fully evaluate the long-term consequences of pregnancy in women after ASO.

Conclusions

In conclusion, these data provide evidence that pregnancy is well tolerated with good maternal outcomes after ASO for TGA in the setting of multidisciplinary obstetric-cardiology care and thorough preconceptual assessment and counseling. These data should provide reassurance for women and their clinical care team planning pregnancies after ASO for TGA.