Tetralogy of Fallot and Aortic Root Dilation: A Long-Term Outlook
- First Online:
- Cite this article as:
- Nagy, C.D., Alejo, D.E., Corretti, M.C. et al. Pediatr Cardiol (2013) 34: 809. doi:10.1007/s00246-012-0537-8
- 658 Views
Dilation of the sinus of Valsalva (SoV) has been increasingly observed after repaired tetralogy of Fallot (TOF). We estimate the prevalence of SoV dilation in adults with repaired TOF and analyze possible factors related to aortic disease. Adults with TOF [n = 109, median age 33.2 years (range 18.1 to 69.5)] evaluated at Johns Hopkins Hospital from 2001 to 2009 were reviewed in an observational retrospective cohort study. Median follow-up was 27.3 (range 0.1–48.8) years. SoV dilation was defined as >95 % confidence interval adjusted for age and body surface area (z-score > 2). The prevalence of SoV dilation was 51 % compared with that of a normal population with a mean z-score of 2.03. Maximal aortic diameters were ≥4 cm in 39 % (42 of 109), ≥4.5 cm in 21 % (23 of 109), ≥5 cm in 8 % (9 of 109), and ≥5.5 cm in 2 % (2 of 109). There was no aortic dissection or death due contributable to aortic disease. Aortic valve replacement was performed in 1.8 % and aortic root or ascending aorta (AA) replacement surgery in 2.8 % of patients. By multivariate logistic regression analysis, aortic regurgitation (AR) [odds ratio (OR) = 3.09, p = 0.005], residual ventricular septal defect (VSD) (OR = 4.14, p < 0.02), and TOF with pulmonary atresia (TOF/PA) (OR = 6.75, p = 0.03) were associated with increased odds of dilated aortic root. SoV dilation after TOF repair is common and persists with aging. AR, residual VSD, and TOF/PA are associated with increased odds of dilation. AA evaluation beyond the SoV is important. Indexed values are imperative to avoid bias on the basis of age and body surface area.
KeywordsTetralogy of FallotCongenital heart diseaseAortaAneurysmDissection
Since the first reported palliation of tetralogy of Fallot (TOF) by Blalock and Taussig in 1945 , and the initial intracardiac repair described by Lillehei in 1955 , adult survival has greatly improved. Although before the availability of surgical therapy it was unusual for patients with TOF to survive beyond the third decade of life  currently nearly all of those born with this disease can expect to reach adulthood [1, 19, 23]. Improved mortality has shifted our focus from the long-term care of this disease toward managing late complications to decrease morbidity. Aortic root dilation has been an increasingly observed feature of late postoperative TOF [12, 22, 30]. Aortic enlargement can lead to progressive aortic valve regurgitation (AR) [22, 30], which in turn could potentially require surgical treatment. The aortopathy associated with TOF may contribute to left-ventricular dysfunction . In rare instances, aortic dissection has been reported [16, 24]. Increased flow through the aorta due to right-to-left shunting before repair and intrinsic aortopathy are postulated to be the underlying mechanisms . The prevalence of aortic root dilation has been reported in few studies and varies between 15 and 87 % depending on used definitions [7, 15, 22]. The aim of our research was to evaluate the aorta in a cohort of surviving adult patients with TOF late after repair, establish the prevalence of aortic dilation, and analyze patient characteristics to identify possible predisposing factors.
Materials and Methods
Study Design and Patient Population
We reviewed data obtained from adult patients with TOF (age ≥ 18 years, median 33.2, range 18.1 to 69.5) who were evaluated at Johns Hopkins Hospital between 2001 and 2009 in an observational retrospective cohort study. The study was approved by the Johns Hopkins Medicine Institutional Review Board (protocol NA_00015726). Demographic, morphological, surgical, and clinical data were obtained from medical records. Patients with a diagnosis of TOF, TOF with pulmonary atresia (TOF/PA), and TOF with dysplastic pulmonary valve who had undergone at least one available echocardiographic study performed at our institution were eligible for inclusion. Exclusion criteria included incomplete data from medical records, unavailable or inadequate imaging study, and women who were pregnant at the time of echocardiography or status ≤6 months postpartum. A total of 169 patients were initially identified. Of those, 60 patients did not meet the prespecified echocardiographic reading criteria, which are outlined below [two of these underwent sinus of Valsalva (SoV) replacement]. A total of 109 patients were included in the final analysis.
Definition of Aortic Root Dilation
In normal adults, age, height, body weight, and sex are known to be determinants of SoV dimensions [14, 25, 28]. Aortic root dilation was defined as an aortic root diameter at the level of SoV greater than the upper 95 % confidence limit adjusted for age and body surface area (BSA) according to previously published normative data by Roman et al. . Based on the American Society of Echocardiography recommendations for chamber quantification, we defined the severity of aortic dilation (annular or SoV) as follows: none = ≤1.96 SDs above predicted value; mild dilatation = 1.97 to <3 SDs above predicted value; moderate = 3.00 to <4 SDs above predicted value; and severe = ≥4 SDs above the predicted value .
The most recent patient echocardiograms performed between 2001 and 2009 were reviewed by two blinded experienced readers (from among one adult cardiologist and four pediatric cardiologists). For patients who underwent aortic valve or SoV surgery during the study period, data from their last available preoperative echocardiogram were used for analysis. SoV size was measured from two-dimensional echocardiographic images in accordance with the American Society of Echocardiography recommendations . Image quality was assessed first. An echocardiographic study was deemed acceptable for inclusion if at least two levels of the proximal aorta were measured, including the SoV in diastole. The aortic valve annulus was measured in the parasternal long-axis view in systole and diastole at the base of the aortic valve leaflets using the inner edge–to–inner edge technique. The maximal diameter of the aorta at the SoV, the sino-tubular junction, and the proximal tubular portion of the ascending aorta (AA) were measured in the parasternal long-axis view in systole and diastole using the leading edge–to–leading edge technique. The severity of AR was assessed by Doppler characteristics and color flow mapping in accordance with the American Society of Echocardiography recommendations  and was graded as absent, mild, moderate, and severe. Left-ventricular dysfunction was assessed qualitatively and was graded as absent, mild, moderate, and severe. For quantitative measurements, the mean of the two readers was used for analysis. When only one reading was available for a specific value, this was used for analysis. For differing qualitative measurements, a third opinion was used for a definitive decision. Interobserver differences were calculated by comparing the two measured sets of values at each level of the aorta.
Categorical data were summarized with frequencies and percentages, and groups were compared using Fisher’s exact test. Most continuous data were not normally distributed and were summarized with medians and ranges. Nonparametric Wilcoxon rank sum tests were used for group comparisons. Change in measurements between the first and the last study were compared using paired Student t test. Interobserver differences in measuring aortic root size were expressed as Pearson correlation coefficients. Analysis was performed using SAS version 9.2 (SAS Institute, Cary, NC, USA). All reported p values are two-sided, and p < 0.05 was considered significant. The authors of this article have certified that the data comply with the principles of ethical publishing .
Characteristics of adult patients with repaired TOF
No. of patients
Age in years (range)
Systolic BP in mm Hg (range)
Diastolic BP in mm Hg (range)
Family history of CHD (%)
Known syndromes (%)
Age at first palliation (years)
First palliation type (%)
Second palliation (%)
Age (range) at correction (years)
Correction type (%)
Residual VSD (%)
Follow-up (range) after repair (years)
Shunt-repair (range) interval (years)
Isolated aortic valve replacement was performed in one patient with moderate to severe AR and a normal aortic root size at the age of 31 years (17 years after TOF repair). Aortic surgery was performed in two additional patients at the time of pulmonary valve replacement. One of them had an AA size of 4.9 cm and mild AR at the age of 57 years (45 years after TOF repair) and underwent an isolated AA replacement. The other patient had an AA size of 4.7 cm and mild AR at the age of 53 years (40 years after TOF repair) and underwent AA plication.
Aortic Dilation Defined by Normative Data
Echocardiographic data for adult patients with repaired TOF
No. of values
No. of values
No. of values
Most recent echocardiogram
Annulus in cm (range)
Indexed annulus in cm/m2 (range)
SoV in cm (range)
Indexed SoV in cm/m2 (range)
SoV diastole ratio to normal as z-score (range)
2.03 (−1.81 to 8.92)
0.77 (–1.81 to 1.91)
STJ in cm (range)b
Indexed STJ diastole in cm/m2 (range)b
STJ diastole ratio to normal as z-score (range)b
2.75 (–1.58 to 11.90)
1.38 (–1.58 to 4.63)
AA in cm (range)b
Indexed AA in cm/m2 (range)b
Maximum AA diameter in cm (range)
Indexed maximum AA diameter in cm/m2 (range)
Aortic regurgitation (%)
LV dysfunction (%)
Echocardiographic measurement data for all patients and by dilation status are listed in Table 2. The dilated group showed more AR than the nondilated group (p < 0.001). There was no difference in qualitative left-ventricular function between the two groups. The median aortic root dimension at the SoV measured in diastole was 3.71 cm (range 2.19–6.22 cm), and the median expected value was 1.19 cm/m2 normal. Regarding the pattern of dilation, a number of patients showed isolated SoV dilation, whereas others had either isolated enlargement of the AA or enlargement of both the SoV and AA. The median maximal AA diameter in the dilated group was 4.38 (range 2.37–6.20) cm versus 3.64 (range 2.53–4.85) cm in the nondilated group.
Interobserver differences in the measurement of SoV dimensions, as assessed by Pearson correlation coefficient r for the SoV and AA, were 0.84 and 0.88, respectively. The aortic annulus and sino-tubular junction showed lower coefficients of 0.6 and 0.76, respectively. None of the differences were significant. Intraobserver differences were not assessed.
Variables Associated With Aortic Root Dilation
Characteristics of the patients by dilation status are listed and compared in Table 1.
The group of patients with dilated aortas was more likely to have a diagnosis of TOF/PA (p = 0.002), TOF with dysplastic pulmonary valve (p = 0.05), or residual VSD (p = 0.02) (Table 1). There was no difference between the dilated and nondilated groups by sex, hypertension, history of palliation, shunt-to-repair interval, or age at correction. The two groups did not significantly differ regarding adverse outcomes, such as death, stroke, or arrhythmia.
ORs and 95 % CIs for dilation of SoV by patient characteristics from multivariable logistic regression
95 % CI
Age at correction (1-year increase)
Time to palliation
BSA (1-unit increase)
Age at echocardiogram (1-year increase)
ORs and 95 % CIs for maximal aortic root ≥4 cm by patient characteristics from multivariable logistic regression
95 % CI
Age at correction (1-year increase)
Time to palliation
Body surface area (1-unit increase)
Age at echocardiogram (1-year increase)
In our study, we found that aortic root dilation is common in adult patients with repaired TOF, consistent with previous published data, and persists with aging [7, 15, 22]. The location of dilation was not limited to the SoV. Some patients demonstrate a pattern of isolated AA enlargement, similar to the bicuspid aortic valve population [2, 29, 31]. Aortic intervention occurred in 3 % of our population, unusual in a population dominated by right heart pathology. Although we had no cases of aortic dissection, this is largely a cohort of survivors. One patient required isolated aortic valve replacement. Prophylactic aortic replacement surgery was performed in 3 % of patients, one in conjunction with aortic valve surgery and two for treatment of a dilated AA at the time of concomitant surgical pulmonary valve replacement. Of note, two additional patients who underwent SoV and AA replacement were excluded in the analysis because they lacked adequate echocardiographic images. Both of these patients underwent surgery for primary aortic root aneurysm.
Aortic Root Size and Surgical Implications
The optimal management of aortic root dilation in patients with repaired TOF is evolving because little evidence-based information is currently available. Judicious follow-up of the aortic root and AA after TOF repair is recommended. Because some patients showed isolated AA dilation, evaluation of the entire aorta is warranted. The aortic root should be imaged on a yearly basis in all adults with TOF if a progressive increase in aortic diameter is noted . Given the propensity of AA enlargement in our population, which may be independent of aortic root enlargement, assessment must include imaging of the entire AA.
The risk of aortic dissection appears to be lower than in other arthropathies (e.g., Ehlers–Danlos, Loeys–Dietz, Turner, Marfan, or bicuspid aortic valve syndromes) . In patients with arthropathies associated with connective-tissue disorders, the risk of aortic dissection and rupture increases once the aortic diameter is >5 cm . Although there is no defined threshold for aortic surgery in patients with TOF, in the absence of a family history of aortic dissection or aneurysm, a rapidly expanding aneurysm (>1 cm/year) or progressive aortic insufficiency (more than mild), some experts suggest that SoV surgery in adults with repaired TOF appears unnecessary until the aortic size exceeds 5.5 cm .
Associated anomalies, patient comorbidities, anticipated life expectancy, and the risk of perioperative and long-term complications associated with surgical therapy are also factored into the decision-making process. One particular circumstance in this patient population is the need for surgical pulmonary valve replacement. Currently, there are no evidence-based data available to guide us to consider concomitant prophylactic aortic surgery at the time of surgical pulmonary valve replacement to avoid one additional future open-heart intervention. Additional risk factors, such as family history of aortopathy and especially associated aortic dissection, should be factored into management decisions. Consideration of patient body size might offer an additional tool to aid in risk stratification .
We stress the importance of evaluation of aortic diameters based on indexed values to avoid selection bias, especially in view of important management decisions, such as consideration of aortic surgery. Our data suggest that arbitrary cut-offs bias toward older, larger, and male population. Younger patients with significant aortic enlargement may be missed.
Variables Associated With SoV Dilation
Factors more commonly found in the group of patients with a SoV >95 % CI when adjusted for age and BSA were (1) diagnosis of TOF/PA, (2) residual VSD, and (3) aortic regurgitation. The diagnosis of TOF/PA associated with larger SoV has been previously reported [6, 20]. In patients with TOF/PA, left-to-right shunting attributable to palliative arterial shunts before repair had a significant volume effect on the aortic size, conceivably contributing to early aortic dilation . Although a diagnosis of TOF/PA appears to predispose to SoV dilation, its clinical significance is unclear because none of the patients who underwent aortic surgery in our study cohort had a diagnosis of TOF/PA.
Tan et al. , Niwa et al.  and Chowdhury et al.  have shown cellular abnormalities of the aortic wall in patients with TOF. Therefore, the presence of AR may serve more as a marker of significant dilation as opposed to a precursor to dilation secondary to volume load. Niwa et al.  showed that a longer shunt-to-repair interval was found to be associated with SoV dilation. Others found that pre-existing aortic root dilation normalized by 7 years of age in patients with TOF who were repaired in infancy, whereas it persisted into adulthood in a group repaired after infancy . In our cohort, aortic dilation is unrelated to age, type of palliation, age at correction, or shunt-to-repair interval. Furthermore, our measurements do not suggest that the aortic dilation normalizes over time. With >50 % of our cohort >95 % percentile adjusted for age and BSA, chronic volume load is an insufficient explanation for aortic enlargement.
The study is retrospective and likely has not captured all patients with a diagnosis of TOF evaluated at Johns Hopkins Hospital during the study time due to lack of a centralized database, which could have introduced selection bias. The analysis is limited to an 8-year period due to a finite time limit for archiving echocardiographic studies. Therefore, not every patient had at least two echocardiograms available for evaluation, which curtailed meaningful longitudinal analysis. Echocardiographic measurements were not available for all levels of the aorta in each study, partially due to inadequate image quality in adult patients and partially because older protocols did not include routine assessment of the AA. Although there were some other imaging modalities available for some patients (magnetic resonance imaging or computed tomography), we elected not to base our analysis on mixed-imaging data.
Aortic dilation after TOF repair is common and persists with aging. Significant root dilation is associated with AR, residual VSD, and a diagnosis of TOF/PA. The need for aortic valve, aortic root, or AA surgery is low but still higher than in the general population. Evaluation of aortic dimensions using indexed values is imperative to avoid bias on the basis of age and BSA. Evaluation of the AA beyond the SoV is important. Ongoing lifelong follow-up with individual patient-based management is recommended.
Many thanks to the Knepper Foundation, whose generous gift help fund this effort. Some of Kathryn A. Carson’s work on this project was supported by Grant Number UL1 RR 025005 from the National Center for Research Resources, a component of the National Institutes of Health (NIH) and NIH Roadmap for Medical Research.