Unveiling ccTGA complexity: CMR and 3D modelling for surgical planning—a case series

Abstract

Aims

Congenitally corrected transposition of the great arteries (ccTGA) is a rare congenital heart defect characterized by atrioventricular and ventriculoarterial discordance and frequently associated with heterogeneous anatomical anomalies. This variability makes surgical indication and strategy particularly complex, requiring accurate preoperative anatomical and functional assessment. We report four paediatric patients with complex ccTGA who underwent anatomical repair at our institution between July 2022 and March 2024. Associated findings included dextrocardia, situs inversus, ventricular septal defects, pulmonary outflow tract obstruction, and atrioventricular valve abnormalities.

Methods and results

All patients underwent stepwise multimodal imaging. Transthoracic echocardiography was complemented by cardiovascular magnetic resonance (CMR) to assess ventricular volumes and function, septal defects, outflow tract relationships, and systemic and pulmonary venous return. CMR datasets were processed to generate patient-specific virtual and 3D-printed models through dedicated segmentation, preserving real anatomical dimensions. These models were used for surgical simulation and multidisciplinary planning. Integration of CMR and 3D modelling directly influenced operative strategy in all cases. 3D reconstructions refined ventricular septal defect tunnelling towards the aorta, optimized ventriculotomy site selection, clarified atrial switch configuration, and anticipated the need for pulmonary root translocation. Anatomical repair was successfully achieved in all four patients, with no early mortality and preserved biventricular function at short follow-up.

Conclusion

Structured integration of CMR and patient-specific 3D modelling enhances spatial understanding and supports individualized surgical planning in complex ccTGA. Incorporating advanced imaging into routine clinical workflow represents a valuable strategy to improve precision and decision-making in anatomically challenging congenital heart disease.

Introduction

Congenitally corrected transposition of the great arteries (ccTGA) is a rare congenital heart defect characterized by both atrioventricular (AV) and ventriculoarterial (VA) discordance, accounting for approximately 0.4–0.5% of all congenital heart defects and with an estimated incidence of about 1 in 33 000 live.¹ Determining the surgical indication and selecting the most appropriate operative strategy are particularly complex because of the heterogeneous nature of ccTGA and the wide variability of associated anomalies, such as situs inversus, dextrocardia, pulmonary outflow tract obstruction, or valve abnormalities.² In this context, accurate anatomical and functional characterization often requires multimodal imaging approach. While transthoracic echocardiography (TTE) remains the first-line imaging modality for congenital heart disease (CHD), in patients with ccTGA it may be limited in clearly defining complex spatial relationships between intracardiac structures and the great arteries. In such cases, cardiovascular magnetic resonance (CMR) and computed tomography (CT) provide a more comprehensive anatomical assessment and have replaced diagnostic catheterization in many scenarios.³ Moreover, the use of patient-specific 3D modelling technologies, based on segmentation of medical imaging data, has markedly increased for surgical or interventional planning in CHD, enabling both virtual 3D reconstructions and physical 3D-printed models.^4–7 In this report, we present a series of four patients with complex ccTGA, all surgically managed at our institution between July 2022 and March 2024. First-line TTE was complemented by CMR imaging combined with 3D modelling, providing a comprehensive anatomical and functional assessment that supported both surgical planning and clinical decision-making. This manuscript aims to emphasize the value of multimodal imaging in guiding surgical planning for complex CHD, reflecting evidence that such integration improves anatomical understanding and patient outcomes.⁸ To better illustrate how these imaging modalities were applied in practice, the following section provides a detailed description of the CMR acquisition protocol and the subsequent 3D modelling process.

Materials and methods

In our clinical workflow, all patients first underwent TTE as an initial assessment. When further anatomical clarification was required, CMR and CT were performed to provide high-resolution, 3D views of cardiac structures, ventricular volumes, and vessel anatomy. This stepwise integration of echocardiography and CMR allowed for a comprehensive evaluation, combining functional insights from TTE with detailed anatomical information from advanced imaging modalities.

CMR and 3D modelling protocol

All CMR examinations were performed on a 3 Tesla scanner (Ingenia, Philips Medical Systems, The Netherlands) using a standardized clinical protocol for complex CHD. In summary, cine steady-state free precession (SSFP) sequences were acquired in contiguous short-axis stacks and standard long-axis planes to assess ventricular anatomy and quantify ventricular volumes, mass, and systolic function. Additional SSFP planes were also prescribed to delineate ventricular septal defects (VSDs) and their relationship to the AV and VA valve and ventricular outflow tracts. Contrast-enhanced MR angiography was performed to visualize extracardiac anatomy, and 4D-flow sequences were acquired to quantify arterial and venous flow. Moreover, an isotropic, respiratory-gated, 3D whole-heart (WH) sequence (1.3 × 1.3 × 1.3 mm³) was obtained in diastole. 3D models were generated form 3D WH via segmentation and reproduced using 3D printing, as previously described.⁷ 3D model reconstruction was performed using 3D Slicer by initially applying a threshold algorithm to MRI datasets, followed by extensive manual slice-by-slice editing to accurately delineate complex anatomical regions. The procedure was carried out by experienced biomedical engineers specialized in cardiovascular modelling, in collaboration with clinicians skilled in cardiac imaging. The anatomical structures included in the models were those of direct clinical relevance for the specific case and necessary for surgical pre-planning. From the final segmentation masks, 3D surface models were generated and refined through mesh correction procedures. No scaling operations were applied: the printed models reproduced the patient’s real anatomical dimensions. To ensure printability while preserving internal cavity dimensions, surfaces were uniformly thickened outward by 0.8 mm. The models were printed using selective laser sintering (SLS) technology with Flexa Bright material.

Cases presentation

Case 1

A 10-year-old girl with ccTGA, situs solitus, dextrocardia, pulmonary atresia, and a large subaortic VSD extending to the inlet region, was referred to our centre for persistent desaturation and cyanosis. She had previously undergone bidirectional Glenn and pulmonary artery reconstruction at 10 months of age. TTE demonstrated two well-developed ventricles, a complex VSD, moderate aortic regurgitation, and systemic right ventricular dilation. Cardiac catheterization confirmed normal pulmonary artery pressures. While the anatomy was largely delineated on TTE, its limited 3D and spatial resolution in complex anatomy warranted further evaluation. CMR was performed to assess the extension and spatial orientation of the VSD relative to the aortic root. CMR showed a large inlet VSD extending into the outflow tract, severe dilation of the morphological right ventricle, mild dilation of the morphological left ventricle, and moderate aortic regurgitation. Therefore, CMR provides a comprehensive data for subsequent biventricular repair planning. (Figure 1A and B).

Figure 1.

(A) CMR 4-chamber view and (B) 3D whole heart MIP reconstruction showing AV discordance, large VSD and the aorta arising from the RV with regurgitation jet of the case 1; (C) Anterior view of the 3D heart model: the right ventricle outflow section reveals large VSD of the same case. LA: left atrium; LV: left ventricle; RA: right atrium; RV: right ventricle; Ao: aorta; LPA: left pulmonary artery; VSD: ventricular septal defect.

A multidisciplinary Heart Team review confirmed the indication for surgical repair. Following this, virtual and printed 3D models (see printed model on Supplementary data online, Figure S1) were created to support surgical planning. The model (Figure 1C) was particularly useful in determining the feasibility of venous drainage rerouting (hemi-Mustard procedure), as well as in planning the LV–Ao tunnel via the RV approach and the leftward placement of the RV–PA conduit. The patient underwent successful Hemi-Mustard and Rastelli repair using a Medtronic Contegra 18 mm conduit, as per the 3D-guided plan. Postoperative course was complicated by diastolic dysfunction of the systemic ventricle, resulting in pleural effusion and supraventricular tachycardia (SVT), treated by electrical cardioversion. The patient was discharged on postoperative day 20. Early postoperative TTE at follow-up showed reduced RV volume and confirmed unobstructed venous pathways.

Case 2

A 6-year-old girl with ccTGA, situs solitus, dextrocardia, large atrial septal defect (ASD), and tricuspid valve (TV) dysplasia underwent TV plasty at 6 months of age. At 18 months of age, she received pulmonary artery (PA) banding and ASD calibration. She was later admitted to our centre because of persistent TV regurgitation. At admission, TTE confirmed severe TV regurgitation, significant left atrial dilation, and a high PA band gradient. CMR demonstrated a dysplastic TV with severe regurgitation and a dilated, hypertrophic systemic right ventricle (located on the left), with preserved systolic function (Figure 2A). Virtual (Figure 2B and C) and printed 3D models (see on Supplementary data online, Figure S2) confirmed the anatomical feasibility of double switch operation and supported preoperative planning by clarifying the spatial relationships between the great vessels and atrial structures, as well as venous drainage and AV valve plane. Surgical correction consisted of a modified Senning procedure combined with an arterial switch operation. Due to residual stenosis at the site of previous PA banding—a possibility anticipated during preoperative 3D assessment—pulmonary artery reconstruction was required. Postoperative recovery was complicated by transient ventricular dysfunction and episodes of SVT, both managed medically. Early postoperative TTE, performed before discharge, demonstrated unobstructed venous baffles, and satisfactory ventricular function without significant outflow tract obstruction. The patient was discharged on postoperative day 18 in stable clinical condition.

Figure 2.

(A) Cine RM 4-chamber view showing the double AV and VA discordance and dysplastic tricuspid valve; (B) 3D virtual model in a frontal view of the same case: the main external structures are visible: right atrium, ventricles, and the major vessels (aorta, main pulmonary artery and innominate vein). Green arrows indicate coronary arteries, originating from the Aorta; (C) in the posterior view, the section reveals inner atrial chambers: the left atrium is visible through the ASD. LA: left atrium; LV: left ventricle; RA: right atrium; RV: right ventricle; Ao: aorta; IV: Innominate vein; IVC: inferior vena cava; MPA: main pulmonary artery.

Case 3

A 2-year-old boy with ccTGA, situs inversus, dextrocardia, and a sub-pulmonary VSD with pulmonary stenosis was referred to our centre for surgical correction. Preoperative TTE demonstrated normal systemic venous return to the right atrium (left-sided) and pulmonary venous return to the left atrium (right-sided). The sub-pulmonary left ventricle showed outflow tract obstruction with preserved systolic function, while the systemic right ventricle (right-sided) was dilated and hypertrophic. TTE was limited in defining the posterior extension of the VSD, its relationship with the aortic root, the TV anatomy (suspected straddling), and quantifying the sub-pulmonary stenosis, so CMR was performed. Cardiac magnetic resonance identified a complex posteriorly extending VSD, AV, and VA discordance with mildly regurgitant AV valve, a sub-pulmonary LV with normal function, hypertrophic systemic RV, and severe pulmonary stenosis (Figure 3A). The great arteries were in a side-by-side orientation, with the aorta anterior and to the right. Given the complex anatomy, virtual (Figure 3B and C) and printed 3D modelling (see on Supplementary data online, Figure S3) was critical in determining the feasibility of anatomical correction using the Senning technique. They also enabled accurate planning of the LV–aorta tunnel via the RV approach and precise measurement of the distance between the right ventricle outflow tract (RVOT) and the pulmonary trunk, thereby helping to anticipate the need for root translocation. Surgery included an atrial switch (modified Shumacker technique), VSD closure with LV-to-aorta tunnelling using an autologous pericardial patch, RVOT enlargement, pulmonary root translocation with anterior pericardial patch reconstruction, and Lecompte manoeuvre. Autologous pericardium was also used for right atrial augmentation and construction of the pulmonary venous baffle, ensuring unobstructed and stable pathways. Pulmonary root translocation avoided the need for a prosthetic conduit, preserved growth potential, and reduced RV pressure overload. Postoperative complications included ascitic and pleural effusions, treated with drainage, and surgical wound dehiscence, which resolved without further intervention. The patient was discharged on postoperative day 32. Early postoperative TTE, performed 1 week after discharge, confirmed good biventricular kinetics, unobstructed venous baffles, sub-pulmonary turbulence, and free pulmonary regurgitation. Postoperative CMR demonstrated preserved biventricular function and unobstructed inflow and outflow pathways, with expected pulmonary regurgitation and mild RV dilation (Figure 3D–F).

Figure 3.

(A) 3D whole heart MIP reconstruction showing AV and VA discordance and large VSD; (B) anterior view of the 3D heart model showing right atrium, left appendage, left and right ventricles, and the major vessels (aorta, main pulmonary artery and superior vena cava); (C) internal view of the heart revealed by cutting away a portion of the model’s external surface. Longitudinal section of the right ventricle reveals VSD; (D, E, F) the post-operative CMR illustrates the unobstructed pulmonary and systemic venous return (PVR) and the surgical tunnel of the aorta to the left ventricle. Ao: aorta; La: left appendage; LV: left ventricle; RA: right atrium; RV: right ventricle; SVC: superior vena cava; IVC: inferior vena cava; MPA: main pulmonary artery; RPA: right pulmonary artery; VSD: ventricular septal defect, PVR: pulmonary venous return.

Case 4

A 3-year-old girl with ccTGA, situs inversus, levocardia, pulmonary atresia, and subaortic VSD was admitted to our centre for a surgical re-evaluation. She underwent a Blalock-Taussig shunt at 1 month of age and a Glenn procedure at 11 months. TTE confirmed AV discordance, aorta arising from the RV, a large VSD, mild aortic regurgitation, and mild tricuspid straddling. However, TTE provided insufficient spatial information regarding relationships between the VSD and systemic outflow, TV straddling, and coronary artery course, which were essential for planning definitive repair. CMR offered additional anatomical detail, particularly better visualization of the pulmonary arteries and distal RPA stenosis with hypoperfusion of the right lung, and confirmed a large ASD, subaortic VSD (Figure 4A), mild tricuspid straddling, normal biventricular volumes, and a dilated aortic root. Based on CMR data and the virtual and printed 3D models (see printed model on Supplementary data online, Figure S4), further guided surgical planning by confirming the feasibility of venous drainage rerouting, according to Hemi-Mustard technique, as well as in planning the LV–Ao tunnel via the RV approach and the rightward placement of the RV–PA conduit. (Figure 4B and C). Hemi-Mustard procedure with LV-to-aorta tunnelling via RV incision was performed. A 12 mm Hancock conduit was used for RVOT reconstruction, while the right PA was augmented with a pericardial patch. The postoperative course was uneventful, and the patient was discharged on postoperative day 23 with favourable short-term clinical outcome. Postoperative TTE at 1 week after discharge showed good biventricular kinetics, moderate tricuspid insufficiency, and unobstructed venous baffles. Postoperative CMR confirmed unobstructed systemic and pulmonary venous baffles, successful RPA enlargement, and a small residual VSD with a mild left-to-right shunt (Qp/Qs = 1.25) (Figure 4D–F).

Figure 4.

(A) 3D whole heart MIP reconstruction illustrating the double discordance AV and VA in situs inversus and a large VSD; (B) anterior view of 3D heart model. The main external structures are visible: right and left ventricles, and the major vessels (aorta, superior vena cava and left pulmonary artery). Green arrow indicates left Glenn anastomosis between SVC and left pulmonary artery; (C) The incision of the right ventricle outflow tract reveals large VSD; (D) post-operative cine-CMR showing an unobstructed pulmonary venous return; (E) in 3D whole heart MIP reconstruction is illustrated the left ventricle outflow tract (LVOT) and RVOT and a small VSD (red arrow); (F) MIP reconstruction illustrating right pulmonary artery enlargement. Ao: aorta; LA: left atrium; LV: left ventricle; RV: right ventricle; SVC: superior vena cava; PA: pulmonary artery; LPA: left pulmonary artery; RPA: right pulmonary artery; VSD: ventricular septal defect.

Discussion

In our series, anatomical repair was successfully achieved in all four paediatric patients with complex ccTGA, despite marked heterogeneity in ventricular morphology and function, septal anatomy, and great vessel relationships. Our experience demonstrates how structured preoperative integration of TTE, CMR, and 3D modelling directly informed surgical strategy in each individual patient. CMR proved essential for defining the relationship between the VSD and the surrounding anatomy, evaluating arterial and venous anatomy and flow, and assessing ventricular volumes and function. These features are difficult to capture with echocardiography alone, yet they are critical to determine the feasibility and safety of anatomic repair. Despite its potential, systematic use of CMR for surgical planning in ccTGA remains poorly represented in current literature.⁹ The most innovative component of our workflow was the incorporation of CMR-derived 3D digital and printed models. In recent years, the application of 3D printing in CHD has expanded substantially.¹⁰ A systematic review and meta-analysis evaluating the impact of 3D printing in congenital cardiac surgery reported that patient-specific models led to modification of the planned surgical strategy in a significant proportion of cases and were associated with improved anatomical understanding and potential reduction in operative times.¹¹ Similarly, comprehensive reviews focusing on 3D printing in CHD have emphasized its utility in transposition physiology, complex VSDs, and outflow tract reconstructions.¹² These reports highlight improved spatial comprehension, facilitation of surgical simulation, and enhanced communication within multidisciplinary teams. Applications have been described not only in classic transposition of the great arteries but also in complex intraventricular rerouting procedures, where understanding the spatial relationship between the VSD and great arteries is critical for avoiding postoperative obstruction. The conceptual foundations of this approach can be traced back to early pioneering experiences with rapid prototyping in structural heart disease, which first demonstrated the feasibility and practical value of patient-specific physical models in guiding interventional and surgical decision-making.¹³ More recently, case-specific experiences focusing directly on congenitally corrected transposition of the great arteries have reported successful integration of 3D-printed models into surgical planning, demonstrating improved visualization of VSD alignment, outflow tract geometry, and coronary relationships, with favourable early postoperative outcomes.¹⁴ Although these reports remain limited in number and are largely descriptive, they provide proof-of-concept evidence supporting model-guided planning in this rare and heterogeneous condition.

In our cohort, 3D modelling provided meaningful adjunctive support in preoperative assessment and planning:

• In two patients, virtual simulation refined the VSD tunnelling trajectory towards the aorta, minimizing the theoretical risk of left ventricular outflow tract obstruction.

• In another, 3D reconstruction of systemic and pulmonary venous return supported optimization of atrial switch configuration.

• In two cases, 3D modelling guided RVOT reconstruction, anticipating the need for pulmonary root translocation in one and leftward placement of the RV–PA conduit in another, allowing the most favourable surgical approach.

These findings are consistent with published data indicating that 3D modelling may alter or confirm surgical planning in complex CHD.^11,12 Despite growing evidence, limitations remain. Segmentation and printing require dedicated expertise, time, and financial resources. Static models cannot reproduce myocardial compliance or dynamic intracardiac flow, and long-term outcome data demonstrating definitive superiority over conventional planning are still lacking.¹¹ Nevertheless, in rare and anatomically heterogeneous conditions such as ccTGA, where surgical margins for error are narrow, incremental improvements in preoperative spatial understanding may translate into clinically meaningful benefits. At short-term follow-up, all four patients demonstrated preserved biventricular function and unobstructed outflow tracts, with no early mortality. While larger studies are required to validate reproducibility and cost-effectiveness, our experience supports the structured integration of CMR and patient-specific 3D modelling in selected complex congenital cases.

Conclusion

In our cohort of four anatomically heterogeneous ccTGA patients, we observed that a CMR-based, patient-specific planning approach was instrumental in tailoring surgical strategies to individual anatomical complexity. In all cases, CMR combined with 3D modelling provided detailed spatial understanding of ventricular relationships, outflow tract anatomy, and associated defects, supporting preoperative planning and surgical management. While our experience suggests that advanced imaging may support postoperative assessment in selected cases, further data are needed to define its systematic role in follow-up. Long-term follow-up will be required to monitor for potential complications and to confirm the durability and clinical impact of this patient-specific, CMR-guided approach in complex ccTGA.

Supplementary data

Supplementary data are available at European Heart Journal - Imaging Methods and Practice online.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

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

Lead author biography

Irene Cao was born on 22 April 1992. She obtained her Scientific High School Diploma in 2011 and subsequently enrolled in the Faculty of Medicine at the University of Cagliari, where she graduated in 2019. She is a trainee (last year) in Cardiac Surgery at the University of Padova. During her training period she attended the department of paediatric and congenital cardiac surgery at the Heart Hospital of Massa and is currently working at the Department of paediatric and congenital cardiac surgery at the ‘Monaldi Hospital’ of Naples.

Vera Cetera was born on 12 October 1989. She obtained her Classical High School Diploma in 2008 and subsequently enrolled in the Faculty of Medicine at the University of Parma, where she graduated in 2014. She completed her specialty training in Cardiac Surgery at the University of Verona. She is currently a Paediatric Cardiac Surgeon at the Heart Hospital of Massa, where her clinical activity focuses on the surgical management of congenital heart disease.