Association of left ventricular outflow tract size with arch morphology in interrupted aortic arch

Abstract

OBJECTIVES

Left ventricular outflow tract obstruction (LVOTO) is a major cause of morbidity and mortality in infants with interrupted aortic arch (IAA). Left Ventricular Outflow Tract (LVOT) development may be flow-mediated, thus IAA morphology may influence LVOT diameter and subsequent reintervention. We investigated the association of IAA morphology [type and presence of aortic arch aberrancy (AAb)] with LVOT diameter and reintervention.

METHODS

All surgical patients with IAA (2001–2022) were reviewed at a single institution. We compared IAA-A versus IAA-B; IAA with aortic AAb versus none; IAA-B with aberrant subclavian (AAbS) artery versus others. Primary outcomes included LVOT diameter (mm), LVOTO at discharge (≥50 mmHg), and LVOT reintervention.

RESULTS

Seventy-seven infants (mean age 10 ± 19 days) were followed for 7.6 (5.5–9.7) years. Perioperative mortality was 3.9% (3/77) and long-term mortality was 5.2% (4/77). Out of 51 IAA-B (66%) and 22 IAA-A (31%) patients, 30% (n = 22) had AAb. Smaller LVOT diameter was associated with IAA-B [IAA-A: 5.40 (4.68–5.80), IAA-B: 4.60 (3.92–5.50), P = 0.007], AAb [AAb: 4.00 (3.70–5.04) versus none: 5.15 (4.30–5.68), P = 0.006], and combined IAA-B + AAbS [IAA-B + AAbS: 4.00 (3.70–5.02) versus other: 5.00 (4.30–5.68), P = 0.002]. The likelihood of LVOTO was higher among AAb [N = 6 (25%) vs N = 1 (2%), P = 0.004] and IAA-B + AAbS [N = 1 (2%) vs N = 6 (30%), P = 0.002]. Time-to-event analysis showed a signal towards increased LVOT reintervention in IAA-B + AAbS (P = 0.11).

CONCLUSIONS

IAA-B and AAb are associated with small LVOT diameter and early LVOTO, especially in combination. This may reflect lower flow in the proximal arch during development. Most reinterventions occur in IAA-B + AAbS, hence these patients should be carefully considered for LVOT intervention at the time of initial repair.

Open in new tabDownload slide

Background

Interrupted aortic arch (IAA) is a severe disease, affecting 2 in 100 000 live births, that invariably requires surgical arch reconstruction in the neonatal period to prevent a fatal outcome [1]. Surgical strategies vary in accordance with IAA morphology which follows 3 major branching patterns based on the location of the interruption: distal to the left subclavian in type A, distal to the left common carotid in type B, and distal to the innominate artery in type C. IAA may also involve an aberrant subclavian (AAbS) artery or an aberrant aortic arch [1].

Regardless of arch intervention strategy, prevention of left ventricular outflow tract obstruction (LVOTO) is a critical component of IAA management. LVOTO occurs in 20–40% of cases following IAA repair, almost always necessitates reintervention, and is associated with increased mortality [2]. Left Ventricular Outflow Tract (LVOT) formation and size are theorized to be flow-mediated [3], with previous studies alluding to the associations between aortic valve area [4, 5], aortic root diameter [6, 7], hypoplastic left heart syndrome [8] or acquired cardiac conditions [9, 10] on LVOT diameter. As such, IAA morphology may also cause abnormal or reduced flow resulting in LVOT hypoplasia. Several small series have correlated IAA morphology with incidence of LVOTO following IAA repair [5, 11]. However, this has not been proven and the clinical significance of these findings with respect to need for reintervention remain unclear. Furthermore, long-term data for patients who receive IAA repair is sparse in the contemporary era [5, 11].

We therefore sought to examine the association of IAA morphology (IAA type, right arch, aberrant left subclavian, and aberrant vertebral branch patterns) on LVOT diameter and the need for long-term reintervention and mortality.

METHODS

Ethical statement: this study was approved by the Hospital for Sick Children Research Ethics Board (REB #1000079914) on 08/04/2022. Patient consent was waived due to the retrospective nature of the study.

Patients

We reviewed all neonates undergoing surgical repair of IAA at the Hospital for Sick Children (Toronto, Canada) between 2000 and 2022 (N = 80). Patients with truncus arteriosus and a large aortopulmonary window were excluded due to their confounding effects on LVOT size. The final number of neonates studied was N = 77, of which N = 22 were IAA type A, N = 51 were IAA type B, N = 1 were IAA type C patients and N = 24 had arch aberrancy (AAb).

Comparison groups

We compared infants with IAA based on branch type (type A versus type B) as well as the presence of AAb (present versus absent). IAA-C patients were included in the total cohort but not in calculations comparing IAA-A and IAA-B subgroups. We defined AAb as the presence of an aberrant vertebral artery, subclavian artery, including aberrant left subclavian from right-sided aortic arch. We also hypothesized that individuals with both IAA-B and concomitant AAbS (IAA-B + AAbS) would have the least amount of flow through the LVOT during development. Therefore, we compared this group to all other patients in a separate comparison. Biventricular repair was generally advocated in the setting of normal LV size and a minimum LVOT dimension of Z > −3. This was also influenced by the patients’ body weight and the principle that biventricular repair is feasible in settings where the LVOT is of at least body weight +1 mm [12, 13].

Outcomes

Outcomes of interest were an anatomical end-point consisting of the patient’s native LVOT diameter before surgical intervention as well as clinical outcomes. Perioperative clinical outcomes of interest, which we defined as within the index hospital admission, included presence of LVOTO at discharge (>50 mmHg), postoperative length of hospital stay. LVOT size was measured by the first author (Malak Elbatarny) and verified by a trained paediatric echocardiographer (Alison Howell) at the study centre. Measurements were taken from the original TTE/TEE images, and any discrepancies were resolved by through joint discussion. Long-term mortality as well as reintervention on the LVOT were also investigated. To avoid confounding effects induced by the surgical intervention, individuals undergoing Norwood Stage 1, pulmonary arterial banding and Fontan procedures were excluded from the perioperative LVOTO and long-term outcomes.

Statistical methods

Clinical characteristics were summarized using descriptive statistics. Continuous variables were characterized using median and interquartile range; dichotomous variables were characterized using frequencies. Between-group differences were compared using Fisher’s exact test and T tests for normally distributed data or the Mann–Whitney U-test for nonnormally distributed data. Survival and time to LVOT reintervention were summarized using Kaplan–Meier survival analysis. The median (Q1–Q3) follow-up was 8.0 (2.1–14.3) years. During the follow-up, there were in total 7 deceased patients. We used administrative censoring of 15 years after index date. The association between LVOTO at discharge and IAA-B + AAbS was assessed using multivariable logistic regression with the covariates preoperative LVOT diameter and ventricular septal defect (VSD) enlargement or myectomy.

RESULTS

Patient baseline characteristics

Patient baseline characteristics are shown in Table 1. There were 77 patients included in total, of which 22 were infants had IAA-A morphology, 51 had IAA-B morphology, 1 patient had IAA-C morphology and IAA classification was unknown for 3 patients. Both groups were comparable in mean age and sex. Among IAA-B patients compared, 10% were premature, whereas none were premature in the IAA-A group. This difference was not statistically significant (P = 0.27). Of the total cohort, 41 (54%) infants had an ASD and 67 (88%) infants had a VSD, 2 of which were recorded in patients unknown IAA morphology. There were no differences in the prevalence of ASD in IAA-A [IAA-A: N = 10 (46%), IAA-B: N = 31 (61%), P = 0.08]. There was a significantly higher incidence of VSD in IAA-B [IAA-A: N = 14 (64%), N = 51 (100%), P < 0.001] but no difference in the need for VSD enlargement or myectomy (P = 0.19). Information on AAb was not available for 1 patient. Arch aberrancy was significantly higher in IAA-B patients [IAA-A: N = 2 (9%), IAA-B: N = 20 (40%), P = 0.015]. In descending order of prevalence, AAb types comprised mainly of right or left subclavian artery aberrancy, followed by left vertebral artery aberrancy, and right-sided aortic arch. Twenty patients (39%) had an IAA-B with concomitant AAbS.

Baseline characteristics compared by presence of AAb and comparing IAA-B + AAbS versus all others are demonstrated in Supplementary Material, Tables S1 and S2, respectively.

Anatomic end-points

Figure 1 compares LVOT diameter by IAA type, presence of AAb, and both in combination. Patients with IAA-B had significantly smaller LVOT diameter, compared to patients with IAA-A [IAA-A: 5.40 (4.68, 5.8), IAA-B: 4.6 (3.92, 5.50), P = 0.007, Fig. 1A]. Patients with AAb had a significantly smaller LVOT diameter [AAb: 4.00(3.70, 5.04), none: 5.15 (4.30, 5.68), P = 0.006, Fig. 1B]. IAA-B + AAbS patients had significantly smaller LVOT diameter compared to all other patients [IAA-B + AAbS: 4.00 (3.70, 5.02), other: 5.00 (4.30, 5.68), P = 0.002, Fig. 1C].

Perioperative outcomes

The total mortality rate was 7/77 patients (9.1%), of which 3 (3.8%) were perioperative deaths. We observed no significant differences in perioperative mortality or length of stay in any of the IAA subtype or AAb comparisons. However, the presence of LVOTO was significantly higher among patients with AAb [AAb: n = 6 (25%), no AAb: n = 1 (2%), P = 0.006] and those with IAA-B + AAbS compared to all others [IAA-B + AAbS: n = 6 (30%), Others: n = 1 (2%), P = 0.001] (Table 2). When performing multivariable logistic regression for LVOTO at discharge, after correction for need for VSD enlargement and preoperative LVOT diameter, IAA-B + AAbS was an independent predictor of LVOTO at discharge [odds ratio: 40.65 (3.26, 506.88), P = 0.004, Fig. 1D].

Long term outcomes

To investigate whether differences in LVOT diameter and presence of LVOTO in the short term by IAA type and AAb impacted long-term clinical outcomes, we performed time-to-event analyses for reintervention on the LVOT and long-term mortality. We excluded 7 patients who underwent either Norwood Stage 1 or pulmonary arterial banding as their index procedure from long-term survival analysis. Median follow-up for the entire cohort was 7.6 [interquartile range: 5.5–9.7] years. LVOT reintervention was performed in 16 patients (20.8%), and the long-term mortality rate was 5.2% (4/77). We observed no significant differences reintervention risk based on IAA type or AAb (Fig. 2A and B). There was a non-significant signal towards increased LVOT reintervention in the IAA-B + AAbS group compared to all others (log-rank P = 0.107, Fig. 2C).

There were no significant differences in long-term mortality based on IAA type (log rank P = 0.38, Fig. 2D). Additionally, we observed 100% survival among AAb patients.

DISCUSSION

Two key findings were demonstrated based on our study: (i) IAA morphology, specifically type B and AAb, are associated with reduced LVOT diameter, and (ii) these findings are associated with 2 main clinical findings. First, in the perioperative period, the proportion of patients with LVOTO at discharge was significantly higher in subgroups with IAA-B, AAb and especially a combination of IAA-B and AAbS. Second, most long-term reinterventions on the LVOT occurred in the IAA-B group and although time-to-event analysis did not reach significance, a signal towards increased LVOT reintervention was observed among patients with combined IAA-B + AAbS.

Previous studies have demonstrated that root intervention at the time of IAA repair can result in favourable perioperative and midterm results when performed by experienced operators [3, 14, 15]. In general, however, long-term outcomes of IAA repair are lacking in the modern era. In this study, we report close to 10-year outcomes on a substantial cohort of IAA patients and demonstrate a low reintervention rate. We also found no differences in long-term mortality based on IAA type or arch morphology, suggesting that LVOT reintervention can be performed with similar mortality when required. Nevertheless, the small sample size and low event rate of this study preclude definitive conclusions about the non-significance of our findings. A large, multicentre study will have greater power to detect possible associations between IAA morphology and LVOT reintervention or long-term mortality [5].

It generally believed that LVOT size is dependent on proximal arch flow during development. The findings of this study may offer some evidence in support of this theory in the context of IAA. Type B and aberrant subclavian anatomy theoretically may divert flow from the LVOT during development and prevent adequate LVOT growth [8]. Still, our study does not provide direct evidence of this phenomenon which warrants further study using flow MRI techniques to correlate proximal arch flow patterns with LVOT size.

In general, the identification of type B and AAbS should focus attention on the risk of potentially small LVOT, particularly when present in combination. Clinically, this may translate to a lower threshold to perform VSD enlargement or myectomy in these patients, provided the additional intervention does not add excessive operative risk. Previous studies have demonstrated root intervention at the time of IAA repair can result in favourable anatomical results when attempted by experienced operators [16–18]. In our centre, pre-emptive surgical myectomy was performed where feasible prior to VSD closure in patients where a deviated septum protruded beneath the aortic valve annulus and thus obstructed the outflow tract. Nevertheless, we observed no increased risk of death in our data among IAA patients requiring reintervention for LVOTO in the long term. For these patients, careful imaging assessment via transoesophageal echocardiography and sometimes additional CT with 3D reconstruction was used to determine the nature of subaortic stenosis. Subaortic resection involved excision of all fibro-muscular stenosis for localized outflow obstruction. More extensive narrowing may have required removal of the original patch, VSD enlargement and re-patching (usually performed through a right ventriculotomy). Complex multilevel stenosis with a degree of hypoplasia may require formal Konno or Ross–Konno procedure if involving the aortic annulus (Supplementary Material, Table S3). Therefore, immediate perioperative risk should always be counterbalanced with long-term benefit. This is particularly important given that long-term reintervention can be performed safely if needed. Nevertheless, mitigating LVOTO in the perioperative period, at least, remains a primary concern. Morphological features of IAA branching may be informative factors in deciding whether to perform concomitant LVOT intervention and for long term surveillance of LVOTO.

Limitations

This study must be interpreted in the context of certain limitations. Firstly, despite being numerous for IAA, the overall sample size was still relatively small and data were accumulated from a single centre only. Furthermore, we had only 1 patient with IAA type C, and as a result were unable to perform additional analyses with this cohort. Owing to retrospective review from charts over 20 years ago, we did not have reliable access to long-term morbidity outcomes such as detailed echocardiographic parameters, quality of life metrics or NYHA class. We observed no mortality events in the aberrant subclavian arm and overall event rate was low.

We were unable to perform LVOT diameter indexing by weight due to the absence of weight measurements recorded from the exact time of the echo. Additionally, we observed more prematurity and small body size in the IAA-B group. Therefore, body size may be a confounding factor in LVOT diameter beyond the flow-mediated mechanism.

The definition of AAb was debated amongst coauthors, in particular the role of a right arch and aberrant vertebral versus aberrant subclavian in mediating flow patterns. Since the exact contribution of each morphological alteration to flow patterns has not been quantified, we included any abnormal morphology in AAb and performed sensitivity analyses where right arch and aberrant vertebral were removed and results were not altered substantially.

CONCLUSION

Type B IAA and Arch Aberrancy are associated with smaller LVOT diameter in IAA, especially when in combination. These findings may reflect reduced flow in the proximal arch during development. Smaller LVOT diameter was also clinically significant with an increased incidence of LVOTO perioperatively. Additionally, most late reinterventions occurred in the type B group, although the signal was not statistically significant. Late mortality was not significantly different based on IAA morphology suggesting that reintervention on the LVOT over the long term can be performed safely when needed. The identification of type B IAA and AAbS should focus attention on the risk of potentially small LVOT and possible need for concomitant LVOT intervention.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

Malak Elbatarny is funded by the Canadian Institute of Health Research Vanier Award and University of Toronto Surgeon-Scientist Training Program Funding.

Conflict of interest: none declared.

DATA AVAILABILITY

The data underlying this article will be shared on reasonable request to the corresponding author.

Author contributions

Malak Elbatarny: Formal analysis; Methodology; Visualization; Writing—original draft; Writing—review & editing. Grace Lee: Data curation; Formal analysis; Methodology; Writing—original draft; Writing—review & editing. Alison Howell: Data curation; Methodology; Writing—review & editing. Marisa Signorile: Data curation; Formal analysis; Writing—review & editing. Osami Honjo: Formal analysis; Methodology; Writing—review & editing. David J. Barron: Conceptualization; Methodology; Project administration; Supervision; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Pankaj Garg, Christian Pierre Brizard and the other anonymous reviewers for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• AAb

  Arch aberrancy

 
• AAbS

  Aberrant subclavian

 
• IAA

  Interrupted aortic arch

 
• LVOTO

  Left ventricular outflow tract obstruction

 
• VSD

  Ventricular septal defect