Hostname: page-component-7b7567c8c-kwt9d Total loading time: 0 Render date: 2026-11-12T23:00:34.660Z Has data issue: false hasContentIssue false
  • English
  • Français

3D printed heart models assist in pre-procedural planning of the innominate vein to pulmonary venous atrium redirection (Hraska / Rome procedures)

Published online by Cambridge University Press:  14 April 2025

Ross Foley,
Kevin Walsh,
Jonathan J. Rome,
Jonathan McGuinness,
Liam Morris,
Eoin O’CearbhaillOpen the ORCID record for Eoin O’Cearbhaill[Opens in a new window] and
Colin J. McMahonOpen the ORCID record for Colin J. McMahon[Opens in a new window]
Ross Foley
Affiliation:
Department of Paediatric Cardiology, Children’s Health Ireland at Crumlin, Dublin, Ireland
Kevin Walsh
Affiliation:
Department of Paediatric Cardiology, Children’s Health Ireland at Crumlin, Dublin, Ireland
Jonathan J. Rome
Affiliation:
Division of Cardiology, Department of Pediatrics, Children’s Hospital of Philadelphia, Philadelphia, PA, USA
Jonathan McGuinness
Affiliation:
Department of Cardiothoracic Surgery, Children’s Health Ireland at Crumlin, Dublin, Ireland
Liam Morris
Affiliation:
Department of Mechanical and Industrial Engineering, Galway-Mayo Institute of Technology, MET Gateway, Galway, Ireland
Eoin O’Cearbhaill
Affiliation:
School of Biomedical Engineering, University College Dublin, Dublin, Ireland
Colin J. McMahon*
Affiliation:
Department of Paediatric Cardiology, Children’s Health Ireland at Crumlin, Dublin, IrelandSchool of Medicine, University College Dublin, Dublin, Ireland
*
Corresponding author: Colin McMahon; Email: cmcmahon992004@yahoo.com
Rights & Permissions [Opens in a new window]

Abstract

Innominate vein redirection to the pulmonary venous atrium has been used in single ventricle patients in order to relieve lymphatic complications resulting from systemic venous hypertension. This has been achieved both surgically, known as the Hraska procedure and via transcatheter approach, known as the Rome procedure. Determining the pathway from the innominate vein to the pulmonary venous atrium is challenging with important intra- and extra-cardiac structures close-by. We present two cases; one surgical and one transcatheter approach, where 3D-printed cardiac models were used to assist in the pre-procedural planning of this relatively novel and challenging strategy.

Keywords

3D printCHDHraska procedureRome procedurelymphatic disorder

Information

Type
Brief Report
Information
Cardiology in the Young , Volume 35 , Issue 5 , May 2025, pp. 1066 - 1069
Check for updates
Copyright
© The Author(s), 2025. Published by Cambridge University Press

Cases

Patient One was a 15-year-old boy with a background of hypoplastic left heart syndrome who underwent 3-stage surgical palliation with Fontan completion at 3 years old. Both his Glenn and Fontan were complicated by post-operative persistent hydrothorax. He also underwent balloon angioplasty, stenting, and subsequent repeat stenting of the aortic coarctation. He developed severe protein losing enteropathy and underwent transcatheter stenting of his Fontan fenestration in an attempt to reduce systemic venous pressures. Protein losing enteropathy remained and he was diagnosed with significant lymphatic malformations on MRI lymphangiogram. He underwent transhepatic lymphatic channel sclerosis on two occasions with only marginal improvement in protein losing enteropathy. Following this a decision was made to proceed to transcatheter thoracic duct decompression (Rome procedure). A 3D-printed heart model was produced as part of his pre-procedural work-up which helped determine the optimal pathway for the procedure.

CT angiography was carried out followed by an automated image segmentation algorithm using a neural network based artificial intelligence algorithm. The model was then printed using a Formlabs Form 4 stereolithography printer. Stereolithography printers exposes a tank of liquid resin to a ultraviolet light source, which hardens it to form three-dimensional objects.

The model demonstrated a very small left atrial appendage which would be the usual destination point of the stent in the Hraska procedure. This had not been appreciated on the CT and as a result he underwent transcatheter placement of a covered stent from his innominate vein extravascularly through lung parenchyma to his right atrial appendage instead of the left. He spent one night in the paediatric ICU and was discharged home well after 5 days. He is followed up regularly in cardiology outpatients where he has shown significant improvement in his symptoms. 3D model and angiography data are shown in Figure 1.

Figure 1.

3D model and angiography data for patient 1 underwent the Rome procedure. ( a ) anterior view of 3D model demonstrating potential pathway (dotted line) from innominate vein (IV) to right atrial appendage (RAA). ( b ) the same model viewed from the left to demonstrate the small left atrial appendage (LAA). ( c ) pre-procedural IV angiography demonstrating flow to the pulmonary arteries (PA) via the superior vena cava (SVC) and Glenn. ( d ) post-procedural angiography demonstrating flow from the IV via the covered stent (asterisks) to right atrium (RA). ( e ) post-procedural angiography demonstrating flow from the SVC to the PAs via the Glenn.

Patient Two was a 3-year-old boy with a background of tricuspid and pulmonary atresia who underwent hybrid first stage single ventricle palliation with a PDA stent. He subsequently required surgical placement of a left-sided emergency Blalock–Tausig–Thomas shunt due to acute desaturation and decompensation. He underwent bidirectional Glenn at 6 months of age. Postoperatively he had prolonged chest drain output and chylothorax requiring re-insertion of his chest drain. He was investigated for lymphatic abnormalities given his previous complications. An MRI lymphangiogram revealed a grade 3 lymphatic malformation. On review of the lyphangiogram data, it was decided that he should undergo an innominate vein turn-down (Hraska procedure) at the time of his Fontan completion.

A 3D-printed cardiac model was produced from his CT as part of his pre-operative work up.

CT angiography and image segmentation were carried out as above. This model was printed using a Stratasys J5 MediJet Polyjet printer. Polyjet printers jet small droplets of different resins onto a build plate which are then exposed to a UV light source, hardening them into three-dimensional objects. This allows for printing of multiple colours and textures in the same model.

This patient’s model demonstrated that the innominate vein could be turned down and directly anastomosed to the left atrial appendage without a conduit. It also demonstrated the distance from the jugular vein to the innominate vein which helped with access. Based on these findings, he underwent Fontan completion using a non-fenestrated extracardiac conduit along with innominate vein turn down to the left atrial appendage and banding of the left internal jugular vein. The procedure was uncomplicated. His chest drains were removed after ten days. A postoperative cardiac catheterisation was carried out which showed a patent Hraska and favourable systemic venous pressures. He was discharged home fifteen days postoperatively. He remained well at six-week follow-up. 3D model and angiography data are shown in Figure 2.

Figure 2.

3D model and angiography data for patient 2 underwent the Hraska procedure. ( a ) anterior view of 3D model demonstrating the proximity of the innominate vein (IV) to left atrial appendage (LAA). ( b ) the same model viewed posteriorly to demonstrate the Glenn anastomoses between the SVC and the PA. ( c ) pre-procedural IV angiography demonstrating flow to the PAs via the SVC and Glenn. ( d ) post-procedural IV angiography demonstrating flow from the IV to the LA.

Discussion

Systemic venous hypertension is an inevitable consequence of single ventricle palliation caused by the lack of a subpulmonic ventricle to shield the systemic venous system from pulmonary arterial pressures. Systemic venous hypertension causes increased afterload in the lymphatic system via the thoracic duct which drains into the angle of the left subclavian and internal jugular vein. This induces pathophysiologic changes resulting in complications such as pleural effusions, ascites, plastic bronchitis and protein losing enteropathy. Reference Safi, Hjortdal and Gewillig1 Medical therapies and direct interventional approaches on the lymphatic system have shown limited success.

Hraska developed a surgical strategy to reduce afterload to the lymphatic system by detaching the innominate vein from the superior vena cava and anastomosing it to the left atrial appendage, either directly or using an interposition graft, thereby allowing thoracic duct drainage to the relatively lower pressure pulmonary venous atrium. Reference Hraska, Hjortdal, Dori and Kreutzer2,Reference Hraska, Mitchell, Woods, Hoffman, Kindel and Ginde3 Rome subsequently developed a transcatheter approach that achieved the same objective using a covered stent to redirect the innominate vein flow to the left atrium either directly or via extravascular approach. Reference Smith, Hoffman, Dori and Rome4,Reference Smith, Dori, O’Byrne, Glatz, Gillespie and Rome5

Both approaches are complex and require considerable pre-procedural planning including cross-sectional imaging in order to determine the optimum pathway to the left atrium that avoids important adjacent structures, particularly the pulmonary arteries and the aortic arch.

3D-printed heart models have been used in the pre-procedural planning of CHD over the past ten years and their use has been proven to add benefit in specific lesions with complex 3-dimensional relationships between important structures. Reference Valverde, Gomez-Ciriza and Hussain6 The two cases described above also involve complex 3-dimensional relationships between the important structures. In both cases, we needed to determine if and how a pathway could be identified which would connect the innominate vein to the atria, while avoiding the pulmonary arteries and the aortic arch. Given this challenge, we believed that 3D-printed heart models would better delineate the anatomy and relationships between the important structures. Both the image segmentation and model generation were carried out by axial 3D printing and industry partner who specialise in medical 3D printing. They provide models with customisable sections connected with magnets which allow easy assessment of the intracardiac anatomy.

Both the interventionalist in case one and the surgeon in case two stated that the 3D models added vital information and helped to guide the procedure. In the first case, the model demonstrated a small left atrial appendage, which resulted in the procedure being altered so that the stent was guided from the innominate vein to the right atrial appendage instead. In the second case, the model demonstrated the proximity of the innominate vein to the left atrial appendage, which gave the surgeon confidence that a direct anastomosis could be achieved without a conduit. In both instances, this was the first time the specific procedure was carried out in our institution. Both the interventionalist and the surgeon stated that the model offered them reassurance that the procedure could be performed as planned and believed that having the model saved time during the procedures. They also reported the models to be beneficial in the parental consent process.

Conclusion

The innominate vein to pulmonary venous atrium redirection strategy is a relatively novel and promising management strategy for single ventricle patients with associated lymphatic abnormalities. Both the surgical and transcatheter approach involve complex pre-procedural planning in order to determine the correct pathway. We have presented one case of each approach where 3D-printed cardiac models were used to add important pre-procedural information. We suggest that the use of 3D-printed models is helpful in the planning of these innominate vein redirection cases.

References

Safi, S, Hjortdal, VE, Gewillig, M, et al. Lymphatic disorder management in pediatric patients with congenital heart disease in european pediatric cardiology centers: current status, disparities, and future considerations. J Am Heart Assoc 2024; 13 ( 22 ): e036597. doi: 10.1161/JAHA.124.036597.CrossRefGoogle ScholarPubMed
Hraska, V, Hjortdal, VE, Dori, Y, Kreutzer, C. Innominate vein turn-down procedure: killing two birds with one stone. JTCVS Tech 2021; 7: 253–260. doi: 10.1016/j.xjtc.2021.01.045.CrossRefGoogle ScholarPubMed
Hraska, V, Mitchell, ME, Woods, RK, Hoffman, GM, Kindel, SJ, Ginde, S. Innominate vein turn-down procedure for failing Fontan circulation. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu 2020; 23: 34–40. doi: 10.1053/j.pcsu.2020.01.002.CrossRefGoogle ScholarPubMed
Smith, CL, Hoffman, TM, Dori, Y, Rome, JJ. Decompression of the thoracic duct: a novel transcatheter approach. Catheter Cardiovasc Interv 2020; 95: E56–E61. doi: 10.1002/ccd.28446.Google ScholarPubMed
Smith, CL, Dori, Y, O’Byrne, ML, Glatz, AC, Gillespie, MJ, Rome, JJ. Transcatheter thoracic duct decompression for multicompartment lymphatic failure after fontan palliation. Circ Cardiovasc Interv 2022; 15 ( 7 ): e011733. doi: 10.1161/CIRCINTERVENTIONS.121.011733. Epub 2022 Jun 16. PMID: 35708032.CrossRefGoogle ScholarPubMed
Valverde, I, Gomez-Ciriza, G, Hussain, T, et al. Three-dimensional printed models for surgical planning of complex congenital heart defects: an international multicentre study. Eur J Cardio-Thoracic Surg 2017; 52: 1139–1148. doi: 10.1093/ejcts/ezx208.CrossRefGoogle ScholarPubMed
Figure 0

Figure 1.3D model and angiography data for patient 1 underwent the Rome procedure. (a) anterior view of 3D model demonstrating potential pathway (dotted line) from innominate vein (IV) to right atrial appendage (RAA). (b) the same model viewed from the left to demonstrate the small left atrial appendage (LAA). (c) pre-procedural IV angiography demonstrating flow to the pulmonary arteries (PA) via the superior vena cava (SVC) and Glenn. (d) post-procedural angiography demonstrating flow from the IV via the covered stent (asterisks) to right atrium (RA). (e) post-procedural angiography demonstrating flow from the SVC to the PAs via the Glenn.

Figure 1

Figure 2.3D model and angiography data for patient 2 underwent the Hraska procedure. (a) anterior view of 3D model demonstrating the proximity of the innominate vein (IV) to left atrial appendage (LAA). (b) the same model viewed posteriorly to demonstrate the Glenn anastomoses between the SVC and the PA. (c) pre-procedural IV angiography demonstrating flow to the PAs via the SVC and Glenn. (d) post-procedural IV angiography demonstrating flow from the IV to the LA.