Incidence and Risk Factors of Recoarctation Following Surgical Repair of Aortic Coarctation in Infants: A retrospective single-centre study

TL;DR

This study found that 23% of infants who had aortic coarctation surgery developed recoarctation, with low body weight and use of a bovine patch as key risk factors.

Contribution

The study identifies novel independent predictors of recoarctation in infants after aortic coarctation repair.

Findings

Recoarctation occurred in 23% of infants following surgical repair of aortic coarctation.

Low surgical weight and bovine patch use were independently associated with recoarctation.

Infants with recoarctation had smaller aortic arch and isthmus measurements on echocardiography.

Abstract

The reported incidence and risk factors for aortic recoarctation (reCoA) after surgical repair in infants vary widely. This study aimed to determine the incidence of reCoA and to identify associated risk factors in a cohort of infants. Secondary objectives included assessing parameters associated with reintervention.

This retrospective cohort study was conducted at the National Heart Center, Muscat, Oman, and included all infants who underwent surgical repair of coarctation of the aorta (CoA) from January 2016 to December 2022. Preoperative, intraoperative and postoperative data were collected, including demographics, prostaglandin E2 use, surgical technique and echocardiographic measurements. Recoarctation was defined as requiring surgical or catheter-based reintervention within the follow-up period.

A total of 113 infants were included in this study. The incidence of reCoA was 23%. Affected patients were significantly younger (P = 0.002) and had lower body weight (P = 0.012) than those without reCoA. Dysmorphic features were more prevalent in the reCoA group (P = 0.042). End-to-end anastomosis was more frequent in the non-reCoA group, whereas bovine patch augmentation was significantly more common in the reCoA group (P = 0.006). Among patients requiring cardiopulmonary bypass, the reCoA group had significantly shorter aortic cross-clamp times (P = 0.014). Echocardiographic analysis revealed smaller transverse aortic arch (P = 0.019) and aortic isthmus (P = 0.018) measurements in the reCoA group. Multivariate analysis identified low surgical weight (odds ratio [OR] = 0.085, 95% confidence interval [CI]: 0.008–0.898; P = 0.047) and bovine patch use (OR = 5.35, 95% CI: 0.688–119; P = 0.048) as independent predictors.

Low body weight at surgery, use of bovine patch augmentation and smaller aortic arch dimensions are significant risk factors for reCoA in infants. These findings emphasise the importance of individualised surgical planning and rigorous postoperative monitoring.

Tables

Table 1.. Demographics and preoperative variables of infants who underwent aortic recoarctation (N = 113).

	n (%)
Characteristics	ReCoA (n = 26)	Non-reCoA (n = 87)	P value*
Demographics
Male	14 (53)	48 (55)	0.90
Mean age at surgery in days ± SD	12 ± 10	25 ± 31	0.002
Mean weight in kg ± SD	2.9 ± 0.4	3.1 ± 0.6	0.012
Perinatal diagnosis	3 (11)	17 (20)	0.33
Escorted	20 (77)	58 (67)	0.47
Dysmorphic	7 (27)	8 (9)	0.042
Pre-operative variables
Cardiac diagnosis			0.123
Discrete CoA	10 (40)	34 (39)
VSD + CoA	4 (16)	15 (17)
Complex CHD† + CoA	4 (16)	21 (24)
IAA	1 (4)	11 (12)
Cardiac CT	5 (19)	24 (28)	0.45
Use of prostin	22 (84)	60 (69)	0.13
Single ventricle	1 (4)	7 (8)	0.68

Table 2.. Surgical and postoperative variables of infants who underwent aortic recoarctation (N = 113).

	n (%)
Variables	ReCoA (n = 26)	Non-reCoA (n = 87)	P value*
Surgical variables
Surgical technique			0.01
End-to-end	14 (54)	61 (70)
End-to-side	2 (8)	12 (14)
Subclavian flap	0 (0)	3 (3)
Bovine patch augmentation	9 (34)	7 (8)
Surgical incision			0.86
Sternotomy	7 (27)	25 (29)
Thoracotomy	19 (73)	62 (71)
Other surgical correction			0.18
VSD closure	1 (4)	11 (12)
Pulmonary artery banding	4 (15)	14 (16)
Complex surgery	0 (0)	8 (9)
Mean cross-clamp time in minutes ± SD	32 ± 21	50 ± 44	0.014
Close chest	23 (88)	79 (91)	0.71
Need for urgent redo-surgery	0 (0)	0 (0)	1
Postoperative variables
Diaphragmatic palsy	2 (7.7)	4 (5)	0.62
Air way stenosis	1 (4)	3 (3)	0.92
Chylothorax	1 (4)	1 (1)	0.36
Need for ECMO	0 (0)	2 (2)	0.43
No arrythmia	24 (92)	81 (93)	0.81
Wound infection	2 (7)	5 (6)	0.66
Mean ICU stay duration in days ± SD	6.8 ± 5.6	5.9 ± 5	0.42
Mean ventilation time in days ± SD	3.1 ± 3.7	3 ± 3.2	0.89
Needed inotropes	12 (46)	43 (49)	0.82

Table 3.. Postoperative transthoracic echocardiographic variables of the cohort (N = 113).

	n (%)
Variable	ReCoA (n = 26)	Non-reCoA (n = 87)	P value*
Aortic valve anatomy			0.37
Normal	9 (35)	38 (44)
BAV present	16 (61)	40 (46)
Mean aortic valve dimension in mm ± SD	5.5 ± 0.85	5.7 ± 0.97	0.29
Aortic valve z-score ± SD	–2.76 ± 1.6	–2.47 ± 1.5	0.50
Mean ascending aorta in mm ± SD	6.6 ± 1.3	7 ± 1.4	0.33
Ascending aorta z-score ± SD	–2.7 ± 1.4	–2.3 ± 1.9	0.48
Mean transverse aortic arch in mm ± SD	3.3 ± 0.8	4 ± 1.1	0.019
Transverse aortic arch z-score ± SD	–5.3 ± 1.7	– 4.2 ±1.9	0.03
Mean aortic isthmus in mm ± SD	1.9 ± 0.58	2.3 ± 0.87	0.018
Aortic isthmus z-score ± SD	–6.7 ± 2.8	–5.3 ± 2.5	0.02
Mean aortic isthmus gradient in mmHg ± SD	42 ± 23	33 ± 21	0.16
Mean mitral valve dimension in mm ± SD	9 ± 1.8	9.6 ± 1.9	0.25
Mitral valve z-score ± SD	–1.2 ± 1.4	–1 ± 1	0.69

Table 4.. Multivariate logistic regression analysis of the factors associated with the development of recoarctation.

Variable	OR (95% CI)	P value
Demographics
Age at surgery in days	0.952 (0.826–1.1)	0.499
Weight in kg	0.085 (0.008–0.898)	0.047
Dysmorphic	0.874 (0.089–8.62)	0.909
Surgical factors
Aortic cross-clamp time	0.982 (0.952–1.01)	0.248
Bovine patch augmentation	5.35 (0.688–119)	0.048
Echocardiographic factors
Transverse aortic arch in mm	1.75 (0.233–13.1)	0.587
Transverse aortic arch z-score	0.633 (0.182–2.2)	0.472
Aortic isthmus in mm	0.006 (0.001–2.38)	0.093
Aortic isthmus z-score	4.77 (0.698–32.6)	0.111

Full text

1. Introduction

Coarctation of the aorta (CoA) is a congenital narrowing of the aortic lumen near the ductus arteriosus, accounting for 4-8% of congenital heart defects (≈4 per 10,000 live births).^12^ CoA often coexists with other cardiac anomalies such as aortic arch hypoplasia and ventricular septal defect. Surgical repair remains the mainstay of treatment and is commonly performed during the neonatal period. However, recoarctation (reCoA), defined as the recurrence of aortic narrowing necessitating reintervention, remains a notable postoperative complication. Reported rates of reCoA vary significantly across studies, ranging from 2.4% to 26%, and are especially high (up to 44%) in infants weighing less than 2 kg at the time of surgery.^34^

Several studies have proposed potential predictors of reCoA, such as low birth weight, younger age at surgery, preoperative aortic arch hypoplasia and postoperative systolic gradients exceeding 13 mmHg.^56^ The use of prostaglandin E2 prior to surgery has also been associated with increased reCoA risk, though its role as an independent predictor remains unclear.^7^

Despite these observations, the literature presents conflicting findings regarding reCoA risk factors. Furthermore, there is limited data from Middle Eastern populations. Therefore, this study aimed to determine the incidence and predictors of reCoA after CoA repair in infants treated at a tertiary cardiac centre, focusing on clinical, surgical and echocardiographic factors associated with reintervention.

2. Methods

This retrospective, single-centre cohort study included all infants who underwent surgical repair of CoA between January 2016 and December 2022 at the National Heart Center, Muscat, Oman. Recoarctation was defined as recurrent aortic obstruction with a peak gradient >20 mmHg, requiring surgical or catheter-based reintervention, as determined by the treating cardiologist.

A team of experienced congenital cardiac surgeons executed surgical repairs. The selection of surgical technique (end-to-end anastomosis versus patch augmentation) was determined by intraoperative evaluation of arch anatomy, extent of hypoplasia and the complexity of coarctation, in accordance with institutional protocols rather than individual surgeons' personal preferences.

The minimum follow-up duration required for inclusion was 6 months. Patients were excluded if they had aortic arch hypoplasia without CoA or lacked follow-up data. Demographic and preoperative variables collected included age, weight, height, associated anomalies and the presence of dysmorphic features. Dysmorphic features were identified through documented clinical assessments by paediatric cardiology and genetics teams, with genetic testing confirmation when available. Although prostaglandin E2 administration was noted, detailed information on dosing and duration was inconsistently available and therefore not included in dose–response analyses.

Surgical variables documented included the timing of surgery, use of cardiopulmonary bypass, surgical technique (e.g., end-to-end anastomosis, end-to-side anastomosis, subclavian flap or patch augmentation), surgical approach (sternotomy or thoracotomy) and whether concomitant cardiac procedures were performed.

Postoperative variables included length of stay in the intensive care unit (ICU), need for inotropic support, requirement for urgent reoperation and other surgical complications. Transthoracic echocardiographic parameters assessed at the time of discharge from the hospital included z-scores of the aortic valve, ascending aorta, proximal and distal transverse aortic arch, aortic isthmus and mean gradient across the isthmus. Z-scores were calculated using data from the Pediatric Heart Network. All data were extracted from the hospital's electronic medical records system.

Statistical analysis was conducted using Statistical Package for Social Sciences (SPSS) software, Version 25 (IBM Corp., Armonk, New York, USA). Categorical variables are presented as frequencies and percentages, while continuous variables are reported as mean ± standard deviation after using normality testing. P values were calculated using Pearson's Chi-square test or Fisher's exact test for categorical variables; the Wilcoxon rank-sum test was used for continuous or ordinal variables. Multivariate logistic regression analysis was performed to identify independent predictors of reCoA after adjustment for relevant demographic, surgical and echocardiographic variable. A P value <0.05 was considered statistically significant.

3. Results

Of 125 infants who underwent CoA repair, 113 met inclusion criteria; 12 were excluded due to incomplete data. The overall incidence of reCoA was 23% (n = 26). The mean follow-up time was 56 months (range: 6–101 months).

Patients in the reCoA group were significantly younger at the time of surgery compared to those without reCoA (mean age: 12 ± 10 days versus 25 ± 31 days; P = 0.002) and had a lower mean body weight (2.9 ± 0.4 kg versus 3.1 ± 0.6 kg; P = 0.012). Dysmorphic features were more commonly observed in the reCoA group than those without reCoA (27% versus 9%; P = 0.042). No statistically significant differences were found between groups in terms of sex, presence of complex congenital heart defects, or prostaglandin E2 administration [Table 1].

Surgical techniques differed significantly between groups. End-to-end anastomosis was more frequently performed in the non-reCoA group (70% versus 54%), whereas bovine patch augmentation was significantly more common in the reCoA group (34% versus 8%; P = 0.006). There was no significant difference in the type of surgical incision (thoracotomy versus sternotomy) or the rate of simultaneous cardiac procedures. Among the subset of patients who underwent repair using cardiopulmonary bypass, the mean aortic cross-clamp time was significantly shorter in the reCoA group (32 ± 21 versus 50 ± 44 minutes; P = 0.014). Postoperative complications such as diaphragmatic palsy, chylothorax and arrhythmia did not differ significantly between the two groups. ICU stay duration, mechanical ventilation time and requirement for inotropic support were also comparable. No in-hospital mortality or urgent reoperation occurred during the early postoperative period [Table 2].

Postoperative echocardiography revealed significantly smaller dimensions in the transverse aortic arch (3.3 ± 0.8 versus 4.0 ± 1.1 mm; P = 0.019) and aortic isthmus (1.9 ± 0.6 versus 2.3 ± 0.9 mm; P = 0.018) among patients with reCoA. Corresponding z-scores for these measurements were also significantly lower. No significant differences were observed in aortic valve anatomy or dimensions, including the presence of a bicuspid aortic valve [Table 3].

On logistic regression, lower body weight at the time of surgery (OR = 0.085, 95% CI: 0.008–0.898; P = 0.047) and use of bovine patch augmentation (OR = 5.35, 95% CI: 0.688–119; P = 0.048) emerged as independent predictors of reCoA [Table 4].

4. Discussion

This study identified a 23% incidence of reCoA following surgical repair of aortic coarctation in infants, consistent with previous reports ranging from 2.4% to 26%.^34^ The findings reinforce the significance of several clinical and surgical variables in predicting reCoA, particularly lower body weight at the time of surgery, use of bovine patch augmentation and reduced aortic arch dimensions.

The elevated reCoA rate in this cohort may be attributed to the rapid somatic growth and haemodynamic changes characteristic of infancy, which can exacerbate residual narrowing or contribute to neointimal proliferation at the repair site. The current study's results align with earlier studies reporting increased reCoA risk among infants with a lower weight, emphasising the vulnerability of this patient subset.^389^ Although lower weight at the time of surgery was independently associated with reCoA, the present cohort showed overlapping weight distributions between groups. Therefore, conclusions regarding specific weight thresholds should be interpreted cautiously.

The choice of surgical technique also emerged as a relevant factor. While end-to-end anastomosis was more commonly employed in patients without reCoA, bovine patch augmentation was significantly associated with reintervention. The broad confidence interval observed for bovine patch augmentation indicates limited event numbers rather than model instability. Sensitivity analyses excluding extreme values did not significantly modify the observed association, although the precision remained constrained due to the sample size. Although patch augmentation is often necessary in anatomically complex cases, it may predispose to future stenosis due to non-growth potential and increased likelihood of tissue scarring. These findings corroborate earlier research advocating for native tissue repairs whenever feasible.^510^

Echocardiographic observations indicated that the transverse aortic arch and aortic isthmus dimensions were significantly linked to reCoA. These findings are consistent with previous research highlighting preoperative aortic arch hypoplasia and reduced aortic dimensions as significant risk factors for reCoA.^68^ Kumar et al. suggested that early postoperative blood pressure gradient may predict the occurrence of a reCoA.^6^ The current study showed that the aortic isthmus mean gradient was not significantly different between the two groups. The implementation of z-scores to standardise these aortic measurements further enhances the credibility of these results by considering variations in body size and age.^7^

This study found that end-to-end anastomosis was more commonly performed in the non-reCoA group (71%) compared to the reCoA group (57%). In contrast, bovine patch augmentation was used more frequently in the reCoA group (34% versus 8%; P = 0.006). These findings indicate that the surgical technique may affect the risk of reCoA. The preferred surgical technique is end-to-end anastomosis for its ability to enhance native tissue growth, which may reduce the risk of reCoA when compared to other techniques like patch augmentation.^511^ However, the best surgical technique may vary based on the aortic arch anatomy and other associated cardiac anomalies.

Identifying predictors of reCoA such as lower weight, bovine patch augmentation and smaller aortic dimensions has important surgical and clinical consequences. Using these clinical risk factors in preoperative risk stratification can contribute to surgical planning and guiding the postoperative course. For example, patients with lower weight or smaller aortic arch dimensions may benefit from earlier follow-up. Additionally, surgeons may consider substitute techniques to patch augmentation in high-risk cases to reduce the risk of reCoA.

Patients with reCoA had smaller transverse arch, isthmus diameters and lower z-scores. Performing early echocardiograms provides an objective measurement which can be an indicator of postoperative aortic arch measurements. Compared to other studies, the presence of bicuspid aortic valve did not significantly affect the outcomes in the current study.^8^

Multivariate analysis showed that low weight and the use of bovine patch were the independent predictors of reCoA. These findings are clinically important, as they aid surgical planning and postoperative follow-up.

The selection of the surgical technique was inherently influenced by the underlying aortic anatomy, including arch hypoplasia and the complexity of coarctation. Due to the incomplete availability of detailed pre-operative morphometric data, adjustments for all anatomical determinants of the surgical strategy were not feasible; this should be taken into account when interpreting the association between surgical technique and reCoA.

This study was subject to several limitations. First, the retrospective design increases the risk of selection bias and the chance of missing data. Second, the study was conducted at a single centre, which may limit the generalisability of the findings to other populations or healthcare systems. Third, the definition of reCoA depends on the assessment of reintervention by the cardiology team, which could vary among institutions. Fourth, pre-operative aortic arch measurements were not consistently available for all patients due to the retrospective design and variability in echocardiographic reporting, limiting their inclusion in comparative and multivariate analyses. Fifth, subgroup analysis based on weight thresholds (<2.5 versus >2.5 kg) was not performed due to limited sample size, preventing reliable statistical conclusions. Finally, this study did not account for other confounding factors such as the diagnosis of genetic syndromes or long-term haemodynamic changes.

Future research should focus on prospective multicentre studies that can help in validating these findings and analysing other risk factors associated with reCoA. An exploration into the long-term outcomes of different surgical techniques, including the use of alternative materials for patch augmentation, could provide valuable insights into optimising surgical repair. Additionally, studies that examine the role of genetic and molecular factors in the development of reCoA may help identify new therapeutic targets.

5. Conclusion

ReCoA remains a common complication after surgical repair of CoA of the aorta in infants. Lower body weight at the time of surgery and the use of bovine patch augmentation were identified as independent predictors of reCoA, while smaller postoperative aortic arch dimensions were associated with increased risk of reintervention. These findings highlight the importance of individualised surgical planning, cautious selection of repair technique based on anatomic considerations and structured postoperative echocardiographic surveillance. Prospective multicentre studies with standardised pre- and postoperative arch measurements are required to validate these findings and optimise strategies to reduce reCoA.

Authors' contribution

Hamid Atiyaq Al Badi: Conceptualization, Methodology, Formal analysis, Investigation, Data Curation, Writing - Original Draft, Writing - Review & Editing. Hamood Al Kindi: Conceptualization, Methodology, Validation, Investigation, Resources, Data Curation, Writing - Original Draft, Writing - Review & Editing. Khalid Al Alawi: Conceptualization, Supervision, Validation; Data Curation, Writing - Original Draft, Writing - Review & Editing. Abdullah Al Furqani: Conceptualization, Supervision, Validation; Data Curation, Writing - Original Draft, Writing - Review & Editing. Laith Al Kindi: Conceptualization, Supervision, Validation; Data Curation, Writing - Original Draft, Writing - Review & Editing. Madan Mohan Maddali: Methodology, Formal analysis, Investigation, Data Curation, Writing - Original Draft, Writing - Review & Editing

Ethics statement

Institutional ethical approval was obtained (MoH/CSR/24/27963) from The Scientific Research Committee, The Royal Hospital, Muscat, Oman. Due to the retrospective nature of the study, individual parental consent was waived.

Conflict of interest

The authors declare no conflicts of interest.

Funding

No funding was received for this study.

Data availability

Data are available upon reasonable request from the corresponding author.