If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Most repaired adolescents with TOF have significant aortic dilatation.
Mild increase aortic size with regression of aortic growth index are common.
Aortic growth among adolescents is a part of physiologic growth.
Aortic growth does not produce adverse effect on aortic regurgitation.
Severity of initial aortic size is not a predictor for the rate of aortic growth.
The study sought to determine the rate of aortic expansion and correlation with somatic growth in patients with repaired tetralogy of Fallot (rTOF), and predictors for determining the annual growth rate of the aorta (Ao-AGR).
Ninety-four rTOF patients (mean age 14.5 ± 4.4 years) with two cardiac magnetic resonance tests (CMR) (median duration 52 months, interquartile range, IQR 24–71) were analyzed for aortic diameter (AoD) at the annulus, the sinus of Valsalva (SoV), the sinotubular junction, and the ascending aorta (AAo), and compared with the normal limit AoD (NL-AoD) values. The median age-at-repair was 60 months (IQR 36−84). Ao-AGR and its index (Ao-AGRI) were derived from changes of the AoD and AoD-index, respectively, divided by the duration between the two studies. Three potential predictors (baseline AoD, sex, and age-at-repair) for the progression of Ao-AGR were analyzed.
There was a significant larger AoD than NL-AoD (p < 0.001). Slow aortic growth was encountered in 78–85 % of patients. The Ao-AGR was slow, the median AGR ranged from 0.37 mm (IQR 0.13−0.72) at annulus to 0.56 mm (IQR 0.22−0.91) at AAo. There was a regression in Ao-AGRI, ranged from -1.41 mm (IQR -1.94, -0.87) at annulus to -2.36 mm (IQR -3.09, -1.63) at SoV. The three predictors were not correlated with severity of Ao-AGR.
Most adolescents with rTOF show significant aortic dilatation. There is a slow Ao-AGR with regression of Ao-AGRI, which may suggest that the rate of aortic growth is slower than the somatic growth. There are no significant predictors of the progression of Ao-AGR.
Patients with definitive surgical repaired tetralogy of Fallot (rTOF) require long-term follow-up to monitor right- and left-sided hemodynamic changes, including pulmonary regurgitation, right ventricular dysfunction, aortic dilatation, and left ventricular volume overload [
]. Aortic dilatation is a well-known sequela in patients with rTOF, where the dilatation process began before surgical correction as the aorta was exposed to long-term volume overload from left-to-right shunt [
]. A large study of 768 children with rTOF, using echocardiographic assessment, demonstrated a decrease in AoR Z-score over the median follow-up period of 3.7 years. Nevertheless, the sizes did not regress to within the normal range [
]. With the use of cardiovascular magnetic resonance (CMR) to follow 110 adults with rTOF, there was a high prevalence (47 %) of progressive AoR diameter during the median follow-up period of 6.3 years, at a progression rate of between of 0.2−0.4 mm/year [
]. Knowledge of the expected aortic growth may aid in determining an appropriate follow-up imaging strategy to detect patients at risks for aortic complications.
To date, there are a few studies of rTOF conducted in adolescents, which is a period of maximum growth and were the expected change in aortic size over time has not been fully established. Our study sought to evaluate the annual growth rate (AGR) of the aorta at four levels: annulus, sinus of Valsalva (SoV), sinotubular junction (STJ), and proximal ascending aorta (AAo) and compared with the somatic growth. We also sought to identify possible predictors influencing rate of aortic growth in adolescents with rTOF by using CMR.
3. Materials and methods
3.1 Study population
This was a single-center, retrospective study, which included patients with definitive surgical repair for a preoperative diagnosis of tetralogy of Fallot (TOF) or double outlet right ventricle of TOF type, who had at least 2 CMR studies between 2009 and 2019. The study excluded patients with rTOF associated with pulmonary atresia, congenital aortic valvular disease (aortic stenosis and bicuspid aortic valve), syndromic aortopathy (Marfan syndrome and Turner syndrome), and patients with percutaneous pulmonary valve implantation. A total of 113 patients were initially identified. Nineteen patients were excluded for the following reasons: incomplete CMR data (n = 17) and non-diagnostic CMR from metallic artifacts (n = 2). Thus a total of 94 patients were included in the analysis. Demographic and age at surgery were obtained from medical records. Age and aortic regurgitation (%) at initial and follow-up CMR scans was collected from CMR reports. The study was approved by the institutional research ethics committee.
3.2 Cardiac magnetic resonance protocol
CMR images were acquired from a 3.0-tesla scanner, with the use of one of two models (Intera Achieva or Ingenia, Philips, Best, the Netherlands): between 2009 and 2013- the Intera Achieva model and from 2014 to 2019-the Ingenia model. Images of AoR in the oblique sagittal left ventricular outflow tract (LVOT) view were obtained with the use of breath-hold, retrospective electrocardiography-gated segmented k-space, cine balanced steady-state free precession (cine bSSFP) in a 6−8 mm slice thickness, with a spatial resolution of 1.5 × 1.5 mm to 1.6 × 1.8 mm, and 30 cardiac phases of the cardiac cycle. A through-plane phase contrast CMR images of the AAo at the level approximately 2.0–3.5 cm above the aortic valve annular plane was used to measure AAo dimensions and areas. The prescribed location was evaluated for the absence of turbulent flow or interference of valve leaflets prior to data acquisition. The imaging parameters of the phase contrast pulse sequence were as follows: a slice thickness of 6−8 mm, a spatial resolution of 1.5 × 1.5 mm to 2.0 × 2.0 mm, a temporal resolution of 60 milliseconds, velocity encoding at 180−200 cm/sec, and 30–40 cardiac phases throughout the cardiac cycle. Images were acquired during breath-hold at expiration. The breath-hold period was approximately 15 s.
For children under 11 years of age, CMR images were obtained under general anesthesia by an anesthesiologist.
3.3 Image analysis
CMR images were transferred to the Picture Archiving and Communications System (PACS), using a DICOM Conformance (Synapse version 3.2.0, FUJIFILM Medical Systems USA’s Synapse® PACS System, USA). Maximum aortic diameter at the levels of annulus, SoV and STJ were measured from the cine bSSFP images in the oblique sagittal LVOT plane during ventricular systole, and perpendicular to the long-axis of the AoR [
] (Fig. 1). Maximum AAo diameter, and maximum and minimum AAo areas were measured from the magnitude images of the phase contrast pulse sequence (Fig. 2). Aortic diameters (AoD) were indexed to body surface area (BSA) and expressed as aortic diameter index (mm/m2) with BSA derived from the Dubois and Dubois formula (BSA [m2] = 0.007184 x height [cm]0.725 x weight [kg]0.425) [
]. AGR and annual growth rate index of the aorta were derived from the changes of aortic diameter (mm) and diameter index (mm/m2) between the initial and follow-up CMR studies, divided by the length of the interval between the 2 CMR examinations, and expressed as mm/year and mm/m2/year, respectively.
The expected normal age range- and sex-matched AoD were calculated and expressed as upper limit of normal aortic diameter (NL-AoD). In children (≤ 17 years old), the NL-AoD data were derived from these formulas: at annulus = 19.57 x (BSA)0.47 (male) and 19.11 x (BSA)0.44 (female); at SoV = 26.95 x (BSA)0.49 (male) and 26.36 x (BSA)0.44 (female); at STJ = 22.29 x (BSA)0.47 (male) and 21.76 x (BSA)0.42 (female); and at AAo = 22.74 x (BSA)0.46 (male) and 22.20 x (BSA)0.46 (female), where BSA was derived from the Dubois and Dubois formula [
]. For patients over 17 years of age, the NL-AoD was recorded, using the mean value of the normal aortic dimeter in the 16–29 year-old population reported by Vriz O, et al., as follows: at annulus = 10.8 mm (male) and 11.1 mm (female); at SoV = 15.2 mm (male) and 15.7 mm (female); at STJ = 12.6 mm (male) and 13.3 mm (female); and at AAo = 13.3 mm (male) and 14.5 mm (female) [
]. The difference between patients’ AoD and the calculated NL-AoD were also analyzed.
Ascending aortic distensibility was defined by the formula: ([AAo maximum area- AAo minimum area]/ AAo minimum area x pulse pressure in 10−3 mmHg-1), where pulse pressure was derived from systolic blood pressure minus diastolic blood pressure [
Predictors (baseline AoD, sex and age at repair) which may influence the rate of progression of aortic size at SoV and AAo levels were analyzed. Baseline AoD was divided into 4 grades based on aortic root Z-scores proposed by Gautier et al. [
] as follows: normal (Z-score <2), mild dilatation (Z-score ≥2 to <4), moderate dilatation (Z-score ≥4 to <6), and severe dilatation (Z-score ≥6). Eighteen patients who were older than 18 years were excluded because of the ages were outside the upper limit. A total of 76 out of 94 patients with the mean age of 12.9 ± 3.1 year were included in the analysis.
Age at repair was divided into 3 groups as followed: repair at age ≤3 years, 4−6 years, and >6 years.
3.4 Statistical analysis
Statistical analyses were performed using STATA software version 16.0 (Stata Corp, College Drive, Texas, USA). Continuous variables were presented as mean ± standard deviation (SD) or median (interquartile range [IQR], 25th, 75th percentile) and categorical variables were summarized as percentages. Wilcoxon signed-rank test was used to assess change in aortic regurgitation from the initial exam to follow-up. Statistically significant changes in AGR of aorta and aortic distensibility based on baseline aortic size at SoV and AAo were analyzed using median regression analysis and linear regression analysis, respectively. A univariate logistic regression model was use to analyze whether sex and age at repair were significantly associated with AGR of the aorta. A p-value of <0.05 was considered as statistically significant. Intra-observer agreement was displayed as mean ± SD and 95 % confidence interval (CI), using a Bland-Altman analysis.
4.1 Patients data
Patient clinical characteristics and hemodynamic data for the 94 patients with definitive surgical repair by closure of the ventricular septal defect and transannular patch are summarized in Table 1. The median operative age was 60 months (interquartile range [IQR] 36, 84). Median duration between the initial and the follow-up CMR examinations was 52 months (IQR 24, 71). Most patients at the initial and follow-up CMR studies belonged to the pediatric (74 %) and adult (64 %) age groups, respectively. For the initial CMR study, all patients had trace to mild aortic regurgitation (regurgitation fraction <30 %). In the follow-up CMR, ninety-one patients had trace to mild aortic regurgitation, and three patients had moderate aortic regurgitation (regurgitation fraction 30–49 %). There was no statistically significant difference in aortic distensibility (p = 0.56) and aortic regurgitation (p = 0.78) between the two CMR studies.
Table 1Patients clinical characteristics and hemodynamic data (N = 94).
4.2 Aortic size and annual growth rate of the aorta
CMR measurement data of the aorta presented as AoD index, AoD, and calculated NL-AoD are shown in Table 2. For the initial CMR study, there was a statistically significant difference in AoD at all levels when compared with the NL-AoD (p < 0.001). The mean difference in aortic size was predominately at SoV and AAo. Table 3 demonstrates the changes of aortic size during the median follow-up of 52 months (IQR 24, 71). An increase in aortic size was noted in 78 %–85 % of patients depended on different aortic level. Moreover, an increase in aortic size at all four levels was encountered in 64 patients (68.1 %). There was a concomitant significant decline in AoD index during the follow-up period (p < 0.001). A growth rate per year of the aorta was analyzed and presented in Table 4. AAo was the region with the greatest AGR, followed by STJ, SoV and annulus, respectively. Nevertheless, there was a decline in AGR index at all levels of the aorta.
Table 2CMR data of aortic diameters at the four different levels (N = 94).
Except for at STJ, the aortic growth per year in three patients who had moderate aortic regurgitation was within the range of the study population as follows: at annulus, SoV, STJ, and AAo were 0.21 (-0.03, 1.12), 0.32 (0.22, 2.90), 0.60 (0.05, 4.80), and 0.45 (0.21, 1.17), respectively.
4.3 Predictors for increasing rate of aortic growth
Annual growth rates of the aorta at the level of SoV and AAo based on grading severity of aortic dilatation expressed as aortic root Z-scores are summarized in Table 5. Compared with patients with normal baseline aortic size, there was a tendency toward an increased growth rate at both SoV and AAo as the aorta changed from normal to mild dilatation. However, in patients with a baseline moderate aortic dilatation, there was relatively lower AGR than in the mild aortic dilatation group. There was a tendency for progression in aortic stiffness when the SoV was progressively dilated. Nevertheless, the degree of baseline aortic dilatation was not a significant predictor for annual aortic progression, using median regression analysis. When comparing a group of normal aortic size with the mild dilated and moderate dilated aorta, the p-values were as follows; at sinus: p = 0.77 and 0.36, respectively and at AAo: p = 0.27 and 0.69, respectively.
Table 5Annual growth rate at sinus of Valsalva and ascending aorta: Categorized into four grades of aortic dilatation based on aortic Z-score (N = 76).
Four grading severity of aortic dilatation
Normal (Z-score <2)
Mild dilatation (Z-score ≥2 to <4)
Moderate dilatation (Z-score ≥4 to <6)
Severe dilatation (Z-score ≥6)
Sinus of Valsalva
23 (30.3 %)
44 (57.9 %)
9 (11.8 %)
12.5 ± 2.8
13.1 ± 3.4
12.8 ± 2.7
27.77 ± 4.17
32.42 ± 3.90
37.62 ± 3.91
Annual growth rate (mm/year)
0.55 (0.22, 0.97)
0.56 (0.23, 1.37)
0.22 (0.15, 0.97)
Aortic regurgitation (%)
2.3 (1.1, 4.3)
3.5 (2.0, 5.6)
5.2 (3.2, 7.4)
Aortic distensibility (10−3 mmHg-1)
5.43 ± 1.67
5.01 ± 1.74
4.92 ± 1.96
9 (11.8 %)
53 (69.7 %)
13 (17.1 %)
1 (1.3 %)
13.3 ± 3.4
12.7 ± 3.3
13.6 ± 1.8
23.02 ± 3.98
27.79 ± 3.98
33.58 ± 3.99
Annual growth rate (mm/year)
0.42 (0.22, 0.68)
0.69 (0.28, 1.04)
0.54 (0.21, 0.72)
Aortic regurgitation (%)
2.0 (1.2, 4.5)
3.1 (1.9, 4.9)
4.9 (3.1, 11.5)
Aortic distensibility (10−3 mmHg-1)
5.40 ± 1.37
5.02 ± 1.84
5.32 ± 1.65
Continuous variables presented as mean ± SD or median and interquartile range (25th, 75th percentile) as appropriate, and categorical variables presented as N (%).
Annual growth rate presented as median difference and interquartile range (25th, 75th percentile).
Although male had slightly faster aortic growth per year than female, sex was not a significant factor influencing aortic growth, except at the STJ region (p = 0.04) (Table 6). There was no patient underwent definitive surgical repair at age less than 1 year in our study and only 3 out of 82 patients had definitive repair at 1 year of age. There were no statistically significant difference in AGR of SoV and AAo between the three groups of age at repair (Table 7).
Table 6Sex differences in annual growth rate of aorta.
Annual growth rate (mm/year)
Male (N = 63)
Female (N = 31)
0.38 [0.11, 0.74]
0.30 [0.13, 0.67]
0.47 [0.15, 1.19]
0.35 [0.18, 0.62]
0.60 [0.23, 1.05]
0.29 [0.06, 0.52]
0.58 [0.29, 1.07]
0.54 [0.09, 0.70]
Annual growth rate presented as median difference and interquartile range (25th, 75th percentile).
Data (p value) derived from median regression analysis.
We randomly selected 31 patients (every 3rd patient) from the initial CMR study for determination of intra-observer agreement using Bland-Altman analysis. There was no significant differences between the two measurements [mean difference (95 % CI)] with details as followed: at annulus = -0.18 mm (-5.12, 4.77), at SoV = 0.04 mm (-2.15, 2.22], at STJ = 0.06 mm (-2.37, 2.43), and at AAo = 0.14 mm (-2.87, 3.15), AAo-maximum area= -0.96 mm2 (-64.70, 58.77), AAo-minimum area = -7.08 mm2 (-64.27, 80.42) (Fig. 3A–F).
The present retrospective study evaluates the changes of aortic size in 94 patients with rTOF with the median follow-up period of 52 months (IQR 24, 71) and reveals three findings:
Most adolescents with rTOF show significant aortic dilatation compared with the calculated normal age range- and sex-matched aortic size. The greatest difference is at the AAo.
There is a slow progression of aortic diameter per year with a concomitant regression of the AGR index of the aorta, which could imply that the somatic growth progresses faster than the aortic growth in this study population.
Baseline aortic size, sex, and age at repair are not significant predictors of increased rate of aortic progression.
5.1 Aortic diameter and expansion rate
Our study found that the baseline aortic size in adolescents with rTOF was significant larger than the calculated NL-AoD. The greatest size difference was found at the AAo region. The significant greater in aortic size in our study could be slightly underestimated owing to the difference measurement methods comparing with the studies that we used as references. Both Gautier M, et al. and Vriz O, et al. measured aorta using leading-edge to leading-edge at end-diastole [
], while our study used inner-edge to inner-edge at systole at the moment of maximum aortic dimension. The study in adult population (42 ± 15 years, 18–80 years) showed that aorta measurements in leading-edge to leading-edge method resulted in about 2 mm larger than the measurements performed in inner-edge to inner-edge method which owing to the aortic wall thickness. However, average differences between aortic diameters at end-diastole were around 1 mm smaller than at end-systole [
]. The information in this issue and aortic wall thickness in children or adolescent was limited. Therefore, with the difference measurement method, our results should had a total smaller aortic dimension comparing with normal references data, at approximately ≤ 1 mm as children and adolescent had a lesser aortic wall thickness comparing with adults.
There was a higher incidence of aortic expansion and growth rate per year compared with a previous study using CMR as a follow-up tool [
]. With a median follow-up duration period of 52 months (IQR 24, 71), AoD increased in 78–85 % of the population and most commonly occurred at the AAo and STJ levels. The previous study reported that the incidence of aortic growth in adults with rTOF was 25–35 % of the population, and the most common encounter at the AAo which was similar to our study. A slow progressive rate of aortic diameter in adults with rTOF has been reported [
]. Our current study reveals a slightly faster AGR of the aorta, ranging from 0.41 mm/year at SoV and 0.45 mm/year at STJ to 0.56 mm/year at AAo.
The slight differences in results could be explained by the different study populations and the normal physiologic growth of the aorta for each age. Growth of the heart and great arteries after birth is at a predictable rate, reaching 50 % of the adult aorta size at birth, 75 % at 5 years, and 90 % at 12 years [
]. Therefore, increasing aortic dimensions are normal physiologic changes from newborn to late adulthood. In addition, the major growth spurt (expressed as BSA) in life occurs maximally at adolescent period [
]. Our study population belonged to early to middle adolescents (mean age 14.5 ± 4.4 years) while the population from the previous study was in early adulthood (median age 30.9 years), hence, the faster aortic growth per year in this study may reflect age-related increase aortic size.
Our data suggest that aortic dilatation and aortic growth are common in patients with rTOF, however, the growth rate may be dynamic over different stages of development. Comparisons of aortic growth should take this into consideration. This study increases our knowledge of aortic growth in the adolescent population.
5.2 Aortic expansion and somatic growth
Although, few studies have been conducted on the aortic growth indexed to BSA, Grotenhuis HB, et al. reported that aortic root dimension remained stable in the majority of children population. The mean marginal increase of AAo was 0.04 Z-score unit per year during the median follow-up period of 3.7 years (IQR 0.5, 6.9). Our study found that despite an increment in aortic diameter (ranging from 0.3 to 0.5 mm/year), there was a small negative growth rate index of the aorta (ranging from -0.3 to -0.4 mm/m2/year) in the adolescents with rTOF. This is probably explained by the rate of body growth in the adolescents being faster than in children. In addition, the growth spurt in the adolescent age group progresses faster than the rate of aortic growth. Our study also found that there was no statistically significant increase in aortic regurgitation during the median follow-up of 52 months (IQR 24, 71). Aortic regurgitation in the initial and follow-up CMR was 3.2 % (IQR 2.0–6.2 %) and 3.4 % (IQR 2.1–5.5 %). Therefore, aortic growth in adolescent population may be a part of physiologic growth and does not produce a significant change in aortic regurgitation. In addition, a few cases who had moderate aortic regurgitation in our study showed no difference in aortic growth compared with patients with mild aortic regurgitation.
5.3 Predictors for increasing rate of aortic growth
Baseline aortic size was not a significant predictor for an increased rate of aortic expansion, even though there was a propensity for increasing rate of aortic growth per year and aortic stiffness with the baseline aortic size up to mild dilatation. Conversely, a decreasing rate of aortic growth was observed when the baseline aorta had moderate dilatation. However, the results in the moderate dilatation could be affected by the relatively small numbers of patients. Such findings could be explained by the histologic changes of aorta in rTOF. The decrease aortic distensibility is an important factor in progressive aortic dilatation rather than the enlarged aorta itself, therefore, baseline AoD does not always lead to subsequent progressive dilatation [
]. In addition, a progressive rate of aortic dilatation could be a multifactorial cause.
Sex was not correlated with the rate of aortic growth, except at STJ region where sex was a marginally significant factor. While age at repair was not a significant predictor for aortic expansion rate in patients who had definite repair after 1 year of age.
5.4 Study limitations
There are several limitations in this study. First, this was a single-center, retrospective study and the data may have been influenced by selection bias, with only survivors referred to our tertiary imaging center. Second, although we used the same imaging plane to measurement aortic size in both initial and follow-up CMR studies, the measurement of aortic root from a single sagittal oblique LVOT view may cause an inadequate view of all aortic root regions in some cases. Third, since the data of ascending aortic dimension was derived from phase contrast studies where the prescribed image may not be acquired at the exact location in all patients and as in the follow-up CMR, this may degrade the accuracy in estimation of ascending aortic growth rate and aortic regurgitation grade. Fourth, the reference age and sex-matched aortic size was estimated instead of compared directly in a population-based case-control study. Another limitation is that our data was derived from CMR while the normal reference aortic data is derived from echocardiography. Difference imaging modalities may influence the aortic dimension. In addition, different methods of measurement in our study (inner-edge to inner-edge at systole) and the reference normal limit aortic size (leading-edge to leading-edge in end-diastole) underestimates aortic size up to 1 mm. Finally, the follow-up duration between the 2 CMR examinations was not of uniform duration and limited to a median duration of 52 months. This may have influenced the estimate of the aortic growth rate per year.
Our results demonstrates that most adolescents with repaired TOF have significant aortic dilatation, and that a slow increment in aortic size is commonly encountered. Nevertheless, we observe that a regression of aortic growth indexed to body growth suggests that aortic growth among adolescents may be a part of normal physiologic growth, and does not necessarily produce an adverse effect on aortic regurgitation. The severity of initial aortic size, sex, and age-at-repair are not significant predictors of the rate of aortic growth. Data from this study bridge the gap between previously conducted studies in children and adults. Despite the new observation of regression of aortic growth index in the adolescent age group, long term follow-up aortic dilatation among patients with rTOF is still required due to aortic dilatation and aortic growth are of multifactorial effects.
All authors did not receive any funding for this work.
The study was approved by the institutional Human Research Ethics Committee, Faculty of Medicine, Ramathibodi Hospital, Mahidol University.
All procedures performed in this studies were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Declaration of Competing Interest
The authors report no declarations of interest.
Factors associated with impaired clinical status in long-term survivors of tetralogy of Fallot repair evaluated by magnetic resonance imaging.