Changes in fetal cardiac axis between 8 and 15 weeks' gestation

Authors

  • A. McBrien,

    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • L. Howley,

    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • Y. Yamamoto,

    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • D. Hutchinson,

    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • A. Hirose,

    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • P. Sekar,

    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • V. Jain,

    1. Department of Obstetrics and Gynecology, University of Alberta, Edmonton, Alberta, Canada; Women's and Children's Health Research Institute, Mazankowski Alberta Heart Institute, Edmonton, Alberta, Canada
    Search for more papers by this author
  • T. Motan,

    1. Department of Obstetrics and Gynecology, University of Alberta, Edmonton, Alberta, Canada; Women's and Children's Health Research Institute, Mazankowski Alberta Heart Institute, Edmonton, Alberta, Canada
    Search for more papers by this author
  • J. Trines,

    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • W. Savard,

    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    Search for more papers by this author
  • L. K. Hornberger

    Corresponding author
    1. Fetal and Neonatal Cardiology Program, Department of Pediatrics, Division of Cardiology, University of Alberta, Edmonton, Alberta, Canada
    2. Department of Obstetrics and Gynecology, University of Alberta, Edmonton, Alberta, Canada; Women's and Children's Health Research Institute, Mazankowski Alberta Heart Institute, Edmonton, Alberta, Canada
    • Correspondence to: Dr L. K. Hornberger, Pediatric Cardiology, Stollery Children's Hospital, WMC 4C1.19 8440 - 112 Street, Edmonton, Alberta T6G 2B7, Canada (e-mail: lisa.hornberger@albertahealthservices.ca)

    Search for more papers by this author

ABSTRACT

Objectives

To document changes in the normal embryonic/fetal cardiac axis in the late first and early second trimesters of pregnancy.

Methods

Images from 188 fetal echocardiograms performed prospectively between 8 and 15 weeks' gestation in 166 healthy pregnancies and in 10 pregnancies with severe fetal heart disease were reviewed. For each echocardiogram, three measurements of the cardiac axis were taken in the axial plane at the level of the four-chamber view. Differences in mean embryonic/fetal cardiac axis at different gestational ages in the healthy pregnancies were compared.

Results

The mean ± SD embryonic/fetal cardiac axis was 25.5 ± 11.5° from 8 + 0 to 9 + 6 weeks (Group  1), 40.4 ± 9.2° from 10 + 0 to 11 + 6 weeks (Group  2), 49.2 ± 7.4° from 12 + 0 to 12 + 6 weeks (Group  3), 50.6 ± 5.7° from 13 + 0 to 13 + 6 weeks (Group 4) and 48.6 ± 7.3° from 14 + 0 to 14 + 6 weeks (Group 5). Groups 1 and 2 were significantly different from each other and all other groups (P < 0.05). The results for 22 cases with repeat measurements from 8 + 0 to 11 + 6 and 12 + 0 to 14 + 6 weeks confirmed that the embryonic/fetal cardiac axis increased significantly (P < 0.001). In the cases with severe congenital heart disease, the cardiac axis was > 90th centile in four cases and < 10th centile in two cases.

Conclusions

The embryonic cardiac axis is relatively midline at 8 weeks and levorotates in the late first trimester. By 12 weeks' gestation, the normal leftward fetal cardiac axis is established and remains stable until at least 14 + 6 weeks. Observation of an abnormal cardiac axis in some cases of severe congenital heart disease prior to 15 weeks' gestation may assist in prenatal detection. Copyright © 2013 ISUOG. Published by John Wiley & Sons Ltd.

INTRODUCTION

The cardiac axis describes the angle between the ventricular septum and the midline of the chest measured from the four-chamber view[1]. An abnormal cardiac axis is often associated with cardiac and extracardiac abnormalities and, as such, is useful in fetal anatomical screening[2-4]. Although the normal fetal cardiac axis has been well defined in the mid and third trimesters, and is relatively constant at 45 ± 10°, little is known about the normal embryonic/fetal cardiac axis at earlier gestational ages. Fetal echocardiography is increasingly used in early pregnancy to screen for congenital heart disease (CHD) in high-risk pregnancies. Knowledge of normal changes in the cardiac axis in early pregnancy is of interest developmentally and should ultimately assist in the detection of abnormalities that result in an abnormal embryonic/fetal cardiac axis at earlier gestational ages. The purpose of our investigation was to document the normal cardiac axis between 8 + 0 and 14 + 6 weeks' gestation.

METHODS

Healthy women with singleton pregnancies from 8 + 0 to 14 + 6 weeks' gestation were actively recruited prospectively for early fetal echocardiography. Women referred for clinical reasons (e.g. family history of CHD or increased nuchal translucency) were also included when the pregnancy outcome was normal. All women provided informed consent prior to participation. We excluded all multiple pregnancies, all pregnancies complicated by chronic maternal illness and pregnancies with fetal cardiac or extracardiac (including chromosomal) anomalies. Pregnancies with intrauterine demise later in pregnancy were also excluded. All fetuses underwent a follow-up echocardiogram performed at approximately 20 weeks' gestation and those with any abnormalities were excluded. Confirmation of normal pregnancy outcome was available in all cases.

Echocardiograms were performed between February 2009 and April 2012. All scans were performed by physicians or sonographers with expertise in fetal echocardiography. If adequate images could not be obtained using a transabdominal approach, a transvaginal scan was also performed. Studies were performed with either of two ultrasound machines: Siemens S2000 or GE Voluson E8 using a 7-MHz curvilinear, 9-MHz linear or a 9–12-MHz transvaginal probe. Scan duration was kept as low as reasonably possible, with minimal use of Doppler. Our protocol was essentially to follow the ALARA principle: keeping scanning times as short and output as low as possible to produce useful imaging. In addition, we consciously kept our thermal and mechanical indices at ≤ 1.0 as published in recent ISUOG guidelines[5].

For measurement of the embryonic/fetal cardiac axis, an axial image of the chest was obtained at the level of the four-chamber view. The cardiac axis was measured with a single complete rib on both sides of the thorax where possible. In earlier gestations this was difficult to achieve and we scrolled through videoclip images frame by frame to familiarize ourselves with the fetal orientation and to identify the frame demonstrating a transverse, symmetrical section through the fetal thorax comparable to that obtained at later gestations. Two observers (A.M. and L.K.H.) assessed the fetal cardiac axis offline by measuring the angle between midline of the chest (from the middle of the anterior chest wall to the center of the vertebral body) and the plane of the interventricular septum (Figure 1 and Videoclip S1). This measurement was taken three times from each recorded echocardiogram from different images and the results were averaged.

Figure 1.

Examples of cardiac axis measurements in 8-week (a, b) and 12-week (c, d) fetuses, as assessed in axial view of fetal chest. (a,c) Anatomical landmarks and orientation. (b,d) Placement of lines for measurement of cardiac axis (angle measured between line bisecting chest in midline and line drawn along ventricular septum). L, left; LV, left ventricle; R, right; RV, right ventricle; S, spine.

Gestational age was calculated using the date of the patient's last menstrual period (LMP) or early ultrasound dating where available. Biometry obtained during the echocardiogram included crown–rump length, femur length and biparietal diameter where possible. Prior to 12 weeks' gestation, gestational age was adjusted if there was a > 5-day discrepancy between crown–rump length and reference values for the gestational age according to the LMP. From 12 to 13 + 6 weeks, gestational age was adjusted if there was a > 7-day discrepancy. From 14 to 14 + 6 weeks, gestational age was adjusted if the biparietal diameter/femur length differed by > 7 days from reference values for the LMP-calculated gestational age.

Ten cases of fetal CHD that had been diagnosed prior to 15 weeks' gestation were identified from the departmental database. The cardiac axis for each case was measured from the initial echocardiogram images using the abovementioned techniques.

Institutional research review board approval was obtained prior to the onset of the study.

Statistical analysis

Continuous variables are reported as mean ± SD. Embryos/fetuses were grouped by gestational age for comparison: 8 + 0 to 9 + 6 weeks (Group 1), 10 + 0 to 11 + 6 weeks (Group 2), 12 + 0 to 12 + 6 weeks (Group 3), 13 + 0 to 13 + 6 weeks (Group 4) and 14 + 0 to 14 + 6 weeks (Group 5). Cardiac axis measurements for these gestational age groups were compared using one-way ANOVA followed by Tukey's multiple comparison test. In the 26 cases with more than one scan during the gestational age period assessed, the earliest scan was included and the later scan excluded for all analyses except the Wilcoxon signed ranks test. Twenty-two cases had paired measurements, the first from 8 to 11 + 6 weeks and the second from 12 to 14 + 6 weeks. For these paired cases, separate analysis was performed using the Wilcoxon signed ranks test. A P-value of < 0.05 was considered significant. Four cases underwent repeat fetal echocardiograms in which the axis was measured twice between 8 and 11 + 6 or 12 and 14 + 6 weeks; in these cases the second study was excluded. Interobserver variability was assessed on a random sample of 20% of the study group (38 echocardiograms) using the percent difference: (absolute difference between measurement by Observer 1 and measurement by Observer 2)/((measurement by Observer 1 + measurement by Observer 2)/2). Statistics were analyzed using the Statistical Package for the Social Sciences (SPSS 19.0; SPSS Inc., Chicago, IL, USA).

RESULTS

In total, 263 scans were attempted and it was possible to measure the cardiac axis in 192 (73%) of these echocardiograms (from 166 pregnancies). Successful cardiac axis measurement rates increased with increasing gestational age (Figure 2a). Among successful examinations, a transvaginal approach was predominantly used at < 10 weeks' gestation and a transabdominal approach at >10 weeks (Figure 2b).

Figure 2.

(a) Successful (image) and unsuccessful (image) echocardiographic measurement of the fetal cardiac axis, showing success increasing with gestational age. (b) Proportion of pregnancies in which transvaginal (TV) imaging was required in addition to transabdominal (TA) imaging, according to gestational age (image, TA only; image, TA + TV). Data include pregnancies with duplicate measurements.

Cardiac axis measurements and gestational age for each embryo/fetus are shown in Figure 3. For the 166 echocardiograms included, the cardiac axis was 46.0 ± 10.5°. One-way ANOVA revealed a significant difference between gestational age group mean values for the embryonic/fetal cardiac axis (P < 0.001). Tukey's multiple comparison test demonstrated that the cardiac axis was significantly more midline at 8 + 0 to 9 + 6 weeks' gestation (25.5 ± 11.5°) than at 10 + 0 to 11 + 6 weeks' gestation (40.4 ± 9.2°; P < 0.05). The cardiac axis between 10 + 0 and 11 + 6 weeks' gestation was also significantly more midline than at later gestations (P < 0.05). The cardiac axis did not change significantly between 12 + 0 and 14 + 6 weeks' gestation (49.2 ± 7.4° at 12 weeks, 50.6 ± 5.7° at 13 weeks and 48.6 ± 7.3° at 14 weeks). Percentile values for the normal embryonic/fetal cardiac axis are shown in Table 1 and Figure 4.

Figure 3.

Scatterplot demonstrating fetal cardiac axis measurements according to gestational age. Quadratic line of best fit (R2 quadratic = 0.482) and 95% range are indicated.

Table 1. Fetal cardiac axis percentiles from 8 to 14 + 6 weeks of gestation
GroupGestational age (weeks)nCardiac axis (°) percentile
10th25th50th75th90th
18 + 0 to 9 + 6149.914.625.134.543.7
210 + 0 to 11 + 62928.334.338.946.752.6
312 + 0 to 12 + 65237.644.349.852.758.7
413 + 0 to 13 + 64942.846.850.954.758.8
514 + 0 to 14 + 62238.542.647.854.359.3
Figure 4.

Box-and-whisker plot of fetal cardiac axis measurements against gestational age. Median, interquartile range, range and outliers are shown. Group 1, 8 + 0 to 9 + 6 weeks; Group 2, 10 + 0 to 11 + 6 weeks; Group 3, 12 + 0 to 12 + 6 weeks; Group 4, 13 + 0 to 13 + 6 weeks; Group 5, 14 + 0 to 14 + 6 weeks.

For the 22 pregnancies with repeat measurements before and after 12 weeks' gestation, the mean cardiac axis, measured between 8 + 0 and 11 + 6 weeks, was 33.4 ± 11.7° and increased to 48.1 ± 6.5° between 12 + 0 and 14 + 6 weeks (Figure 5). Wilcoxon signed ranks testing indicated that this increase in the cardiac axis was statistically significant (P < 0.001).

Figure 5.

Fetal cardiac axis measurements for paired samples showing change in axis between measurements taken before and those taken after 12 weeks' gestation.

Testing for interobserver variability in 20% of the study cohort indicated that the percentage difference was 17%.

Cardiac axis measurements for the 10 cases with CHD diagnosed prior to 15 weeks' gestation are shown in Table 2. The results demonstrate that the cardiac axis is abnormal in a high proportion of these cases. An example of the relatively levorotated axis seen in a case of tetralogy of Fallot with pulmonary atresia diagnosed at 14 weeks' gestation is shown in Figure 6.

Table 2. Fetal cardiac axis (CA) measurements and corresponding percentiles in cases of severe congenital heart disease (CHD) identified prior to 15 weeks' gestation
DiagnosisGA (weeks)Referral indicationCA ( °)CA percentile
  1. GA, gestational age at diagnosis; IVF, in-vitro fertilization; ToF, tetralogy of Fallot.

ToF, pulmonary atresia14Oligohydramnios, suspected CHD86.9>90th
ToF, atrioventricular septal defect12 + 5Increased nuchal translucency60.5>90th
Severe pulmonary and aortic regurgitation13 + 6Hydrops, twins29.9<10th
ToF13 + 1Triplet IVF pregnancy64.8>90th
Truncus arteriosus, interrupted aortic arch13 + 3IVF pregnancy46.710th–25th
Congenitally corrected transposition, pulmonary stenosis12Twins57.875th–90th
ToF12Maternal diabetes72.8>90th
Coarctation12 + 5Cystic hygroma, twin IVF pregnancy47.025th–50th
Coarctation11 + 4Cystic hygroma, suspected CHD, pleural effusion48.275th–90th
Critical aortic stenosis13Anasarca, Turner syndrome, suspected CHD27.7<10th
Figure 6.

Fetus with tetralogy of Fallot and pulmonary atresia at 14 weeks' gestation. (a) Anatomical landmarks. (b) Placement of lines for measurement of the cardiac axis. Note relative levorotation of the heart within the chest. L, left; LV, left ventricle; R, right; RV, right ventricle; S, spine.

DISCUSSION

Our study indicates that the fetal cardiac axis is relatively midline at 8 weeks' gestation and gradually rotates to the left. The normal fetal cardiac axis is established by 12 weeks' gestation and remains stable at least until 15 weeks' gestation. We also demonstrated that the cardiac axis can be assessed in the majority of cases through transabdominal imaging from 10 weeks' gestation. A high proportion of the small series with CHD in our study had an abnormal cardiac axis measurement.

Fetal echocardiography is increasingly used in the late first to early second trimesters in pregnancies at highest risk of fetal heart disease. Even in low-risk pregnancies, visualization of the four-chamber view can be accomplished in up to 87% of pregnancies examined between 11 and 14 weeks' gestation[6]. Comstock was the first to report that the normal fetal cardiac axis in the mid- and third-trimester fetus lies at a 45° angle to the left of the midline[1]. This study included a group of fetuses ranging from 13 weeks' gestation to term; however, only two fetuses were below 15 weeks' gestation.

An abnormal fetal cardiac axis may be secondary to extracardiac abnormalities such as congenital diaphragmatic hernia, but it is also associated with various forms of CHD[1, 3, 7, 8]. A levorotated axis, for instance, has been found in coarctation of the aorta, conotruncal lesions and Ebstein anomaly of the tricuspid valve. In contrast, a midline or rightward axis has been observed in other more complicated cardiac defects such as corrected transposition of the great arteries and certain forms of heterotaxy. We have documented that the cardiac axis may be abnormal even in the late first trimester in the presence of more severe congenital heart disease. In our series of abnormal cases, the most abnormal axis measurements were demonstrated in the four cases of tetralogy of Fallot, all with axis measurements over the 90th percentile, similar to descriptions of findings in later gestations[3]. Our case of truncus arteriosus, however, was more midline than one would expect, and the case with congenitally corrected transposition relatively levorotated, which is out of keeping with findings from previous reports in the mid and third trimesters. Interestingly, in the latter case, levocardia, and not the usual mesocardia, was observed even later in gestation, perhaps related to right ventricular hypoplasia. More extensive study of the cardiac axis in early gestations is required to determine patterns for the various conditions. However, as is true in later gestation, it does appear likely that cardiac axis screening in early pregnancy identifies only a subset of cases affected by congenital heart disease.

The atria, ventricles, venous connections, arterial roots and intrapericardial arterial trunks are fully formed by 8 weeks' gestation[9]. Very few data exist about the normal cardiac axis in the late first and early second trimesters. In 1997, using a combination of ultrasound images and postmortem specimens, Allan et al. reported that at 9 weeks' gestation the cardiac apex appeared to point anteriorly and that it rotated to the left by 11 weeks' gestation[10]. This study reported only qualitative findings and provided no quantitative measure of the cardiac axis. This early work was also supported by additional descriptive postmortem experience[9]. In contrast, a recent publication based on fetal echocardiography suggests that the opposite occurs, with the heart rotating from the left towards the midline during the 11th week of gestation[4].

It is not entirely clear why the findings of Sinkovskaya et al. differ from our results and those of earlier studies. They studied the cardiac axis from 11 weeks' gestation, whereas our study and those by others included embryos at earlier gestations. The more midline cardiac axis is most obvious at 8 weeks' gestation and has fully levorotated to what is typically observed later in gestation by 11 weeks. Our study analyzed fetuses between 10 and 11 + 6 weeks' gestation as a group and this may in part exacerbate the difference in findings. Our observations support the earlier qualitative work, suggesting that the heart is a relatively midline structure at 8 weeks' gestation and then rotates leftwards, establishing a stable position within the chest by 12 weeks' gestation.

The rotation in the fetal cardiac axis noted during the late first trimester may result from the final phase of cardiac embryogenesis. During normal cardiac looping, a complex series of positional changes take place[11]. Early in the looping process, the ventricular relationship is craniocaudal; the ventricles then bend and twist and the right ventricle moves into a rightward, dorsal position relative to the left ventricle. In the final phase of ventricular looping, the heart ‘untwists’ anticlockwise around the basoapical axis with the right ventricle moving ventrally. During this period of embryogenesis, there is also expansion of the right ventricular inflow and apical components and delamination of the septal leaflet of the tricuspid valve. The ‘untwisting’ phase of looping and the relatively late developmental changes in right ventricular morphology may explain the change in the cardiac axis noted during early fetal echocardiography.

Limitations of this study include the difficulty in accurately measuring the cardiac axis in the earliest gestations. The embryonic/fetal bones are poorly ossified in early pregnancy and it is therefore challenging to ascertain the exact midline of the chest. Image resolution is limited by the embryonic/fetal size and the distance from the transducer; thus, it can be difficult to obtain an ideal four-chamber view. This is reflected in the relatively low proportion of cases in which the fetal cardiac axis could be ascertained prior to 11 weeks' gestation. Another potential weakness of the study is the relatively small numbers in the earliest gestation groups.

Our study demonstrates that quantitative assessment of the fetal cardiac axis is possible in the majority of cases from 10 weeks' gestation and it documents the normal embryonic/fetal cardiac axis. The reference values we provide should assist sonographers in assessing whether the cardiac axis is normal in the late first and early second trimesters. Further prospective studies are required to address whether this quantitative assessment will be useful for identifying CHD in the late first/early second trimesters.

ACKNOWLEDGMENTS

This work was supported by an operational grant from the Women's and Children's Health Research Institute and the Department of Pediatrics of the University of Alberta.

Supporting Information On The Internet

Videoclip S1 may be found in the online version of this article.

Ancillary