Objective To assess the indication, technical aspects, and outcome of stenting of the right ventricular outflow tract (RVOT) in the management of symptomatic patients with severely limited pulmonary blood flow.
Methods Retrospective case note and procedure review of patients undergoing stenting of the RVOT over an 8 year period.
Patients Between 2005 and 2012, 52 selected patients underwent percutaneous stent implantation into a very narrow RVOT to improve pulmonary blood flow. Median age at stent implantation was 63 (range 4–406) days and median weight was 3.8 (1.7–12.2) kg.
Results 52 patients underwent stent implantation. Median procedure time was 57 (24–260) min and fluoroscopy time 16 (5.5–73) min. There was one procedural death (1.9%) and one emergency surgery (1.9%). Saturations increased from 71% (52–83%) to 92% (81–100%) (p<0.001). Two patients required early shunts due to inadequate palliation and two died from non-cardiac causes. Sixteen further catheter interventions were undertaken (balloon in 7, further stent in 9). Twenty-nine patients underwent delayed surgery (complete repair in 26, palliative in 3) at a median of 172 (52–758) days post-stenting. Left pulmonary artery Z score increased from a pre-interventional value of −1.75 (−4.96 to 0.67) to a pre-surgical value of −0.55 (−4.12 to 1.97), (p<0.01). Median right pulmonary artery Z score increased from −2.63 (−7.70 to 0.89) to −0.75 (−6.69 to 1.18) (p<0.01) . Seventeen patients remain well palliated after a median of 122 (40–286) days.
Conclusions Stenting of the RVOT is an effective treatment option in the initial management of selected patients with very reduced pulmonary blood flow.
Statistics from Altmetric.com
The approach to patients with normally related great arteries and a narrow right ventricular outflow tract (RVOT) with reduced pulmonary blood flow, as classically seen in tetralogy of Fallot, is mostly surgical—either by creating a systemic-to-pulmonary artery shunt, an outflow tract patch, or by undertaking complete repair. The best timing for complete repair remains a matter of debate.1 ,2 Creation of a systemic-to-pulmonary artery shunt as initial palliation remains the standard approach to symptomatic neonates and young infants in the majority of centres3 ,4 Nonetheless, this remains high risk surgery particularly in low weight neonates, or those with associated lesions such as atrioventricular septal defect (AVSD), hypoplastic pulmonary arteries or coexisting morbidities or syndromes.5 ,6
Percutaneous pulmonary balloon valvuloplasty has previously been advocated in the initial treatment of tetralogy of Fallot,7 ,8 but did not gain widespread acceptance due to rather unpredictable results. Stenting of the RVOT has first been described in 1997 by Gibbs and colleagues.9 To date only few small case series have been published.10–12
This report describes a single centre experience with stenting of the RVOT in a consecutive series of 52 patients who were judged to be at high risk for initial surgical palliation or early complete repair.
Patients and methods
Between 2005 and 2012, 57 patients with normally related great arteries, a ventricular septal defect, and a severely narrow RVOT resulting in profound desaturation or requiring prostaglandin infusion to maintain pulmonary blood flow through the ductus arteriosus, were selected by a local multidisciplinary team meeting for percutaneous catheter stent implantation as the primary intervention. Surgical intervention (creation of a systemic-to-pulmonary artery shunt, outflow tract patch or complete repair) was deemed high risk in all, due to a wide variety of factors including weight, associated lesions, syndromes, comorbidities or relevant anatomical features (table 1). Patients with pulmonary atresia and ventricular septal defect (VSD) who would have required radiofrequency perforation of the atretic pulmonary valve before subsequent stent implantation were excluded.
Fifty-two patients underwent stent implantation (91%). Five patients were excluded due to judged unfavourable anatomy (severe hypoplasia of the main pulmonary artery or very short infundibulum) or improvement with balloon valvotomy alone. Age at stent implantation was a median of 63 (range 4–406) days. Median weight was 3.8 (range 1.7–12.2) kg. Seventeen patients (32.7%) weighed <3 kg, with 11 children weighing <2.5 kg. There were 30 male patients.
Individual informed consent was obtained for each procedure from the parents, stating the decision of the multidisciplinary team meeting, the novelty of the approach, and the experience and results gained so far.
Cardiac catheterisation was carried out under general anaesthesia with endotracheal intubation and ventilation with high inspired oxygen concentrations. Twenty-nine patients were on pre-procedural propanolol medication. Emergency drugs prepared before the procedure were propanolol 0.1 mg/kg in 10 mL, and phenylepinephrine solutions 1 in 10 000.
Catheterisation was performed from the femoral vein in 46 patients and from the right internal jugular vein in six. Heparin 50–100 iU/kg was given in all. Biplane angiography of the right ventricle was used to delineate the anatomy, to make measurements, and to select reference images to guide during stent positioning. Most commonly a tube angulation of 30° right anterior oblique (RAO) with 20° cranial tilt together with 60 or 90° left anterior oblique (LAO) was used (figure 1).
Angiographic measurements, together with the preceding cardiac ultrasound images, guided in the selection of the size of the stent chosen. Generally, an attempt was made to place the stent only across the RVOT and to preserve the pulmonary valve. In patients weighing <3.5 kg and in those in whom only short to medium term palliation was required, a standard pre-mounted bare metal coronary stent was selected (Liberte, Boston Scientific, Natick, Massachusetts, USA). In larger patients, or in those with required medium to longer term palliation, a bare metal peripheral vascular stent was selected (JoStent, Abbot Vascular, Maidenhead UK; Genesis, Cordis Corp, Miami Lakes, Florida, USA). These latter stents were routinely hand mounted onto OptaPro balloon catheters of 2 cm length (Cordis Corp, Miami Lakes, Florida, USA). Of late, Formula 414 stents (Cook Europe, Bjaeverskov, Denmark) were used in patients weighing >3.5 kg. These are pre-mounted peripheral vascular stents that can be delivered through a 5 French (F) sheath or a 6 F coronary guide catheter. All material was prepared before any attempts were made to cross the RVOT (figure 2).
Early during the series a wide range of catheters, guidewires, and sheaths were used to enter the branch pulmonary arteries and to place the stent. Of late, the technique has become more uniform. In patients weighing <3 kg, a 4 F long sheath (Flexor, Cook Europe, Bjaeverskov, Denmark) is placed within the right atrium and the RVOT is crossed with a 4 F right Judkins catheter, ideally without the help of a guidewire, while monitoring pressures and performing a test angiogram in the distal pulmonary artery. A 0.014 inch Thruway coronary wire (Boston Scientific, Natick, Massachusetts, USA) is then placed within the distal pulmonary artery and the long sheath is advanced over the 4 F right Judkins catheter into the branch pulmonary artery. The Judkins catheter is then removed over the wire and the pre-mounted coronary stent is positioned, using multiple test injections through the side-arm of the long sheath. In larger patients in whom a coronary stent is used, the same technique is used but preference is for a 6 F right Judkins guide catheter instead of the 4 F long sheath, due to the greater stability it provides and the ability to perform side-arm injections during guide catheter placement. In patients in whom a vascular stent was used to achieve greater stent diameter and longer term palliation, a 7 F Mullins sheath (Cook Europe, Bjaeverskov, Denmark) was used after the pulmonary arteries were intubated with a 5 F right Judkins catheter and a 0.035 inch standard Teflon wire was placed. Of late, the production of the Jostent by Abbott Vascular has ceased and we have switched over to using Formula 414 stents (Cook Europe, Bjaeverskov, Denmark), which are pre-mounted and can be placed over a 0.014 inch wire via a 6 F guide catheter or a 5 F long sheath.
Following placement of the stent a repeat right ventricular angiogram was performed through the long sheath with the wire remaining in the pulmonary artery. A cardiac ultrasound study was performed to rule out pericardial effusion, assess ventricular function, confirm the adequate positioning of the stent, and to rule out any damage to any of the surrounding structures. The guidewire and the venous sheath were only removed when the patient was entirely stable (figure 3).
Serial measurements of branch pulmonary artery diameters at presentation, intervention, and during follow-up were obtained by cross-sectional echocardiography. These were indexed to body surface area and expressed as Z scores. The pulmonary valve z-score at the time of intervention was a median of −4 (−6.8 to −0.3).
The described study evolved from ongoing clinical practice. Patient selection was based on individual case discussion within the greater clinical team. Cases were selected for stenting of the RVOT when surgical intervention (Blalock–Taussig shunt or compete repair) was considered high risk in the individual patient due to significant comorbidities or anatomical features. Retrospective data collection and analysis was approved by the local research ethics and clinical governance bodies.
Standard statistical methods were used. Variables are expressed as median with ranges, or mean values together with SDs. Differences between parameters were compared by Student t test, with p<0.05 being considered significant.
Fifty-two patients underwent stent implantation, with one procedural death (1.9%) due to guidewire perforation of the pulmonary artery, haemoptysis, and cardiac tamponade. A stent was placed as a bailout during resuscitation but the patient did not survive. One patient required emergency surgery for perforation of the RVOT and subsequent haemopericardium (1.9%). One patient required cardiac massage post-stent placement and recovered without sequelae. Two patients underwent successful stenting, but remained prostaglandin dependent for adequate pulmonary blood flow (3.8%) due to persistent significant RVOT obstruction. Both underwent systemic pulmonary artery shunting within 2 weeks post-procedure. Thus, the early mortality rate was 1.9% and the early surgical reintervention rate was 5.7% (3/52). Two patients died from non-cardiac causes: anal atresia with postoperative necrotising enterocolitis in one and extensive tracheal stenosis in the other within 30 days of the procedure.
Procedure time ranged from 24–260 (median 57) min with a median fluoroscopy time of 16 (range 5.5–73) min and median radiation dose exposure of 1.3 (range 0.1–16.6) Gycm2.
Sixty stents were implanted during the procedure in 52 patients, with eight patients requiring two stents due to inadequate covering of the entire length of obstruction. Stent length ranged from 12–24 mm with a median of 16 mm. Five of the eight patients requiring two stents were within the group of the first 10 patients treated with this novel technique until the end of 2008. Median stent diameter chosen was 5.0 (4.0–7.0) mm.
Systemic arterial oxygen saturations increased from a median of 71% (52–83%) to 92% (81–100%) (p<0.001). Rise in arterial saturations ranged from 9–36% (median 20%). Prostaglandin infusions were stopped at the end of the procedure in 11/12 patients. Patients with a rise of saturations in excess of 20% were commenced on oral diuretics. Three patients developed transient significant pulmonary oedema requiring intravenous diuretics. Two patients suffered from pericardial effusion, which did not require drainage. Four patients were admitted to intensive care unit for overnight ventilation after the procedure and two required blood transfusion.
Sixteen patients underwent further catheter interventions during the follow-up period. These included further balloon angioplasty of the originally placed stent in seven patients and implantation of a further, larger, stent in nine (figure 4). Two of these patients had concomitant embolisation of major aortopulmonary collateral arteries to prepare for complete surgical correction. One patient developed endocarditis on the previously placed stent some 3 weeks after hepatobiliary surgery, requiring 4 weeks of intravenous antibiotic treatment and, subsequently, surgical removal of the stent and complete repair of her underlying tetralogy of Fallot. There were no cases of ventricular arrhythmias, atrioventricular block, aortic regurgitation or myocardial ischaemia either early or late after stenting of the RVOT. There were no cases of stent fracture over the observation period.
In total, 17 patients remain well palliated a median of 122 (range 40–286) days after stenting of the RVOT.
Twenty-nine patients underwent delayed surgery (complete repair in 26, palliative in three) after a median of 172 (52–758) days post-stenting. Left pulmonary artery Z score at the time of surgery had increased from a pre-stent value of −1.75 (−4.96 to 0.67) to a pre-surgical value of −0.55 (−4.12 to 1.97) (p<0.01). Median right pulmonary artery Z score had increased from −2.63 (−7.70 to 0.89) to −0.75 (−6.69 to 1.18) (p<0.01).
At the time of surgery the stents were seen firmly embedded in the right ventricular myocardium with significant fibromuscular tissue ingrowth. Part of the posterior aspect of the stent was commonly left behind so as to limit myocardial resection and potential damage. There was one postoperative death after creation of a left Blalock–Taussig shunt and pulmonary artery augmentation some 80 days post-RVOT stenting. There was one death 62 days after complete repair of tetralogy of Fallot due to pneumonia in the context of chronic lung disease. There was one death after repair of a complete AVSD with tetralogy of Fallot, requiring extended extracorporeal membrane oxygenation support postoperatively.
The optimal treatment strategy for children with severe RVOT obstruction remains an issue of debate. Complete neonatal repair is advocated by some centres, whereas others, including our own institution, prefer delayed complete repair after 3–4 months of life.1–5 If intervention was needed before this age, the creation of a modified Blalock–Taussig shunt remained our standard approach. This was despite the numerous small pioneering reports by other centres to propose balloon valvuloplasty or stent implantation into the RVOT as an effective interim palliative procedure.7–12
In 2005, we started to consider percutaneous stent implantation into the RVOT in patients who were judged to be at high risk for initial surgical palliation due to severe associated comorbidities. During the further course of this study patients with very low weight, in whom shunting was deemed to be a very high risk procedure, were considered for stenting of the outflow tract. Based on our initial results, we then considered children with a history of severe cyanotic spells for treatment and those in whom complete repair would constitute a very complex surgical undertaking—such as patients with anomalous systemic or pulmonary venous drainage or associated complete AVSD.
Despite a very steep learning curve and some setbacks—three of the four major complications (death in one and emergency or urgent surgical intervention) occurred in the first 12 patients treated with this technique—we continued to employ and refine this technique. In symptomatic neonates and young infants with standard tetralogy of Fallot we now advocate stenting of the RVOT as the initial palliation, as an alternative to a surgical shunting procedure. This is particularly so in children weighing <3 kg, in whom shunting is associated with high mortality rates.6 In children with tetralogy of Fallot and associated AVSD, stenting of the RVOT establishes the most efficient supply of pulmonary blood flow by ensuring systemic venous blood enters the pulmonary arteries, rather than systemic arterial blood through a systemic-to-pulmonary artery shunt. This in turn limits the volume loading of the ventricle and improves the resultant rise in arterial oxygen saturation per unit of volume load.
Selection of the stent implanted must be guided by the underlying anatomy and the angiographic and echocardiographic measurements. Based on our initial results, stent length should be ample and should exceed the measured length, due to the likelihood of foreshortening in standard angiographic projections. The use of a guidewire with distance markers is helpful to verify angiographic length measurements. A decision has to be made as to whether medium term rather than short term palliation is required. Coronary stents are limited with regards to maximal stent dilation due to their design. Peripheral vascular stents have greater potential for future progressive dilatation.
Even though interventional cardiac catheter equipment has improved over time, there is a very urgent need to design bespoke kits for the admittedly small market of interventional paediatric cardiac interventions. This is even more important now, as catheter interventions in neonates and premature babies have become a reality.
It is important to perform pulmonary balloon valvuloplasty at the same time of the procedure, as the majority of patients with tetralogy of Fallot have multilevel obstructions. This can be performed with the same balloon catheter as was used during stent placement.
Initially we deselected patients for stent treatment if this would have involved placing the stent across the pulmonary valve. During the latter part of this series this has changed. Placing a stent across the pulmonary valve leads to a situation akin to placing a non-valved right ventricle to pulmonary artery conduit or outflow tract patch. These surgical techniques have very good haemodynamic effects with systemic venous blood entering the pulmonary arterial tree, and in the Norwood palliation have been shown to promote improved growth of the pulmonary artery tree compared to a modified Blalock–Taussig shunt.13 ,14
Three patients had an anomalous left coronary artery arising from the right coronary artery. The stent within the RVOT addresses the anterior deviation of the outlet septum, pushing it backwards. There were no instances of coronary ischaemia. We would suggest that anomalous coronary artery anatomy is a good indication for an initial stent rather than early repair using a small conduit.
This study describes an evolving clinical practice and procedural details and outcomes. The study population comprised high risk patients for surgical intervention who were selected for catheter intervention on an individual basis. The range of associated lesions and comorbidities makes this a highly diverse group, which limits longitudinal analysis of pulmonary arterial growth for the entire cohort or compared against a surgical group.
Based on our experience and results, we now consider stenting of the RVOT as the first line treatment for selected patients with severe RVOT obstruction as an alternative to systemic-to-pulmonary artery shunting procedures and in those in whom early complete repair carries significant additional risk or complexity.
We would like to thank Dr A Chikermane, Dr T Desai, Dr S Viswanathan and Dr A Sands for referring patients for treatment, and express our gratitude for the support by the cardiac catheter, cardiac anaesthetic, and cardiac surgical teams at Birmingham Children's Hospital.
Contributors All authors have contributed significantly by being involved in the conception and design, acquisition of data or analysis and interpretation of data, drafting the article or revising it critically for important intellectual content, and final approval of the version published.
Competing interests None.
Ethics approval Institutional approval was obtained for the procedure.
Parental consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.