Article Text


Congenital heart disease
Stenting of the right ventricular outflow tract in the symptomatic infant with tetralogy of Fallot
  1. G Dohlen,
  2. R R Chaturvedi,
  3. L N Benson,
  4. A Ozawa,
  5. G S Van Arsdell,
  6. D S Fruitman,
  7. K-J Lee
  1. The Cardiac Diagnostic and Interventional Unit, The Labatt Family Heart Centre, The Hospital for Sick Children, University of Toronto School of Medicine, Toronto, Canada
  1. Dr Kyong-Jin Lee, The Hospital for Sick Children, 555 University Avenue, Toronto, Canada, M5G 1X8; kyong-jin.lee{at}


Background: Repair of neonatal tetralogy of Fallot (TOF) has low mortality. Debate continues regarding the initial management of cyanotic or duct-dependent infants with TOF and adverse risk factors. While repair can and has been performed in these patients, it is associated with increased morbidity.

Objective: We review the effectiveness of right ventricular outflow tract (RVOT) stenting in the symptomatic young infant with TOF.

Methods: Clinical, echocardiographic, angiographic and haemodynamic data were reviewed for nine patients who underwent 11 RVOT stenting procedures from October 1994 to August 2007.

Results: The pulmonary valve was deemed unsalvageable in all patients (median valve diameter 3.7 mm (range 2.7–4.2), Z-score −6.7 (range −9.7 to −5.4). RVOT stenting improved arterial oxygen saturation from a median of 73% (60–85%) to 94% (90–98%) (p = 0.008). Median Z-score for the left pulmonary artery increased from −4.9 (−7.8 to −2.4) before stent implantation to −1.5 (−4.2 to −0.2) (p = 0.02) before surgical repair. Median Z-score for the right pulmonary artery increased from −3.7 (−6.8 to −1.9) to −0.8 (−2.5 to 0.1) (p = 0.008). Median Nakata index increased from 56 mm2/m2 (21–77) to 150 mm2/m2 (123–231) (p = 0.008). There were no procedural complications. Six patients have undergone successful repair. There were no deaths.

Conclusions: In the symptomatic young infant with TOF, stenting of the RVOT provides a safe and effective management strategy, improving arterial oxygen saturation and encouraging pulmonary artery growth.

Statistics from

Debate continues regarding the initial management of cyanotic or duct-dependent infants with tetralogy of Fallot (TOF) and adverse risk factors. These risk factors include low weight, prematurity, young age (<3 months), unfavourable pulmonary arterial anatomy, abnormal coronary distribution and critical preoperative condition.13 Palliative procedures include surgical right ventricular outflow tract (RVOT) enlargement or conduit, aortopulmonary shunt, stenting of the arterial duct and balloon pulmonary valvuloplasty.410 While primary repair can and has been performed in these patients, it is associated with increased morbidity.1 11

We review the effectiveness of RVOT stenting in the symptomatic young infant with TOF.


The interventional cardiac database was searched to identify infants with TOF with or without pulmonary atresia (TOF-PA) who underwent RVOT stent implantation. Medical records, echocardiography, angiography and haemodynamic data were reviewed. Measurements of pulmonary arterial branches were obtained from recorded echocardiographic studies before the first catheterisation procedure and before surgery. Proximal pulmonary artery dimensions are presented as Z-scores adjusted for body surface area based on normal values established at our institution. The Nakata index was calculated based on distal pulmonary artery diameters to represent hilar dimensions. This review was approved by the research ethics board at the Hospital for Sick Children.

Patient population

All patients had confluent true central pulmonary arteries supplying all lung segments and aortopulmonary collateral vessels when present, offered dual supply. The pulmonary valve was deemed unsalvageable in all patients. Coronary artery anatomy was normal in all patients. Our institutional approach to neonatal TOF evolved during the study period and thus indications for stent implantation varied. Prostaglandin dependency or hypercyanotic spells in infants deemed to be at higher risk for primary repair based on pulmonary artery hypoplasia, low weight or significant non-cardiac co-morbid conditions were the most common indications.

Procedural details

All cardiac catheterisations were performed under general anaesthesia from a femoral venous access approach. In infants with TOF-PA, the pulmonary valve was opened using radiofrequency energy and balloon dilated using techniques previously published.12 A 0.014-inch Wizdom (Cordis Corporation, Miami, FL, USA) or the stiffer ASAHI Grand Slam (Abbott Laboratories, Redwood City, CA, USA) wire was used to cross the pulmonary valve and its tip was placed in either a distal pulmonary artery branch or in the descending aorta through the arterial duct. In earlier procedures, 10-mm Palmaz stents (P1007, P104, Johnson & Johnson Interventional Systems, Warren, NJ, USA) mounted on 7-mm balloons were deployed using a long sheath technique. Since 2002, pre-mounted coronary stents (Multi-Link Ultra, Guidant Corporation, Santa Clara, CA, USA) were implanted without a long sheath guide. Utilisation of the low profile, flexible pre-mounted coronary stent that can be deployed without a long sheath, has greatly simplified this procedure. Stent diameter was chosen to be 1–2-mm larger than the diameter of the infundibulum during diastole to maintain stability of stent position. Although coronary artery anatomy was normal in all our patients, a left anterior descending coronary artery crossing the RVOT would not have precluded stent implantation. Significant over-stretching of the coronary artery was unlikely to occur in an outflow tract dilated 1–2 mm larger than the diastolic dimension.


Descriptive statistics are reported as medians and ranges. Comparison of pre-stent and post-stent parameters was performed using the Wilcoxon signed rank test. Statistical significance was defined as p<0.05. All data analysis was performed in R



Nine infants (seven male) underwent 11 RVOT stent implantation procedures between October 1994 and August 2007. Indications to undergo palliative therapy were variable and are listed in table 1.

Table 1 Patient data at initial catheterisation procedure

Two infants had TOF-PA and two had significant multiple aortopulmonary collateral arteries. Figure 1 illustrates the treatment path to stent implantation of all patients. The median time from previous balloon dilation to RVOT stent implantation was 39 days (range 6–42). At first stent implantation, three infants had prostaglandin-dependent pulmonary blood flow (patients 3, 6, 7), two had severe cyanosis (patients 1 and 4), and five were having severe hypercyanotic episodes of whom two required mechanical ventilation (patients 6 and 8). One patient was on a continuous phenylephrine infusion (patient 9) and the other who, despite morphine, esmolol and phenylephrine boluses, had a cardiac arrest (patient 6). One infant who was transferred from another centre with severe hypercyanotic episodes had evidence of significant neurological injury (grade 3 intraventricular haemorrhage and hyoxaemic ischaemic encephalopathic changes) and seizures before the procedure (patient 9). The median oxygen saturation with supplemental oxygen was 73% (range 60–85%) (table 2). Median age at initial stent intervention was 23 days (3–119), and median weight 3.1 kg (2.1–4.1). Two infants (patients 5 and 7) required a second stent implant 109 days and 46 days after the initial implant.

Figure 1 Management path to right ventricular outflow stent procedure. RF, radiofrequency energy; RVOT, right ventricular outflow tract; TOF, tetralogy of Fallot; TOF-PA, tetralogy of Fallot with pulmonary atresia.
Table 2 RVOT stent procedure

Anatomical features before stenting

All infants had severe RVOT obstruction by echocardiography and right ventriculography. Proximal left pulmonary artery (LPA) stenosis related to ductal tissue was present in two infants and at the insertion of an aortopulmonary artery in one infant. The median pulmonary valve diameter and Z-score were 3.7 mm (2.7–4.2) and −6.7 (−9.7 to −5.4) respectively. The median LPA and right pulmonary artery (RPA) diameters were 2.5 mm (1.7–3.1) and 2.9 mm (2.1–3.9) respectively. The median Z-score for the LPA and RPA were −4.9 (−7.8 to −2.4) and −3.7 (−6.8 to −1.9) respectively. The median Nakata index was 56 mm2/m2 (21–77).

Procedural details of RVOT stenting

Two patients (patients 3 and 7) had stent implantation during the same catheterisation as the RVOT balloon dilation because of unsustained improvement in arterial oxygen saturation. Implanted stents are listed in table 2. In the initial experience, if there was a significant increase in arterial saturation, the procedure was concluded even if the entire infundibulum was not covered. Patients 5 and 7 developed significant cyanosis secondary to progressive muscular narrowing below the stent. A second stent was placed in both infants with dramatic improvement in arterial oxygen saturation (fig 2). Thereafter, our approach has been to pre-emptively cover the RVOT and pulmonary valve with a single long or tandem coronary stents. In one procedure (patient 8), three stents were placed. Concurrent bilateral pulmonary artery angioplasties were performed with 3-mm, 4-mm and 5-mm coronary balloons (Voyager RX, Guidant Corporation, Santa Clara, CA, USA) in two patients. In patient 6, a duct-related proximal LPA stenosis was not significantly improved with balloon dilation. Median fluoroscopy time was 35 minutes (12–53). The procedures resulted in an immediate increase in arterial oxygen saturations from a median of 73% (60–85%) with supplemental oxygen, to a median of 94% (90–98%) (p = 0.008) in room air. The continuous phenylephrine infusion in patient 9 was discontinued in the catheterisation laboratory after stent deployment. There were no significant procedural complications. Pulmonary over-perfusion was not problematic in any cases with a few patients requiring only short-term diuretic therapy. Median time to discharge was 4 days.1 12

Figure 2 Right ventricular outflow tract stenting. (A) Right ventricular (RV) angiogram in anterior projection showing severe muscular right ventricular outflow tract obstruction (RVOT) and hypoplastic pulmonary valve, main pulmonary artery and branch pulmonary arteries. (B) Lateral projection of (A). (C) Pulmonary angiogram showing growth of pulmonary arteries and stent across the RVOT pulmonary valve and main pulmonary artery. (D) Lateral projection of (C) showing two RVOT stents covering the entire infundibulum.

Pulmonary artery growth

Pre-surgical echocardiography demonstrated significant growth of the pulmonary arteries.

The median LPA and RPA diameters were 4.5 mm (3.1–7.1) and 4.9 mm (4.6–6.9) respectively. Median LPA and RPA Z- scores were –1.5 (−4.2 to −0.2) (p = 0.02) and −0.8 (−2.5 to 0.1) (p = 0.008), respectively (fig 3A, B). Median Nakata index had increased to 150 mm2/m2 (123–231) (p = 0.008) over a median time of 76 days (42–267) (fig 3C).

Figure 3 Pulmonary artery growth. First echo, echocardiogram before right ventricular outflow tract dilation or stenting. Last echo, echocardiogram before surgical repair. (A) Left pulmonary artery (LPA) Z-scores. (B) Right pulmonary artery (RPA) Z-scores. (C) Nakata index.

Subsequent corrective surgery

Six of the nine children have had surgical correction. Surgery was indicated by increasing cyanosis or hypercyanotic spells in five children. The remainder are awaiting elective repair with oxygen saturations more than 90% with no further hypercyanotic episodes. Median age at surgery was 142 days (44–413), median weight was 5.3 kg (4.0–9.3) and median time from stent implantation to surgery 125 days (40–315) (table 2). All patients underwent standard repair with stent removal and a transannular RVOT patch in five infants and conduit placement in one infant. The aortopulmonary collaterals in patient 1 were divided and ligated and in patient 5 did not require intervention. Pulmonary arterioplasty was performed only in the three patients with previously identified LPA stenosis. An atrial septal defect was left in 50%. Median bypass time was 99 minutes (66–127), cross-clamp time 51 minutes (26–78), mechanical ventilation 4 days (0–11) and intensive care unit stay 5 days (2–12). Median duration of inotropic support was 2 days (0–6) and median time to discharge was 16 days (5–29).

Follow-up after surgery

All patients are alive at a median follow-up after surgery of 2.1 years (2.5 months to 11.5 years). Two (40%) children had evidence of persistently elevated right ventricular (RV) pressures both with pre-existing LPA stenosis and a subsequent arterioplasty. Patient 1 required four postoperative catheterisations between 2.8 years and 10.9 years after surgery to address LPA and conduit stenoses and subsequently underwent conduit replacement and LPA arterioplasty 11.7 years after initial surgery. Patient 6 had persistent RV pressures that were 74% systemic and has undergone two subsequent LPA dilations at 47 days and 6 months. Four children have normal RV systolic pressures by echocardiography.


The need for early intervention is usually related to factors that also increase the risk of surgery—for example, infections or other systemic problems, hypercyanotic episodes with potential neurological injury or poor pulmonary artery anatomy.1417

Primary repair versus two-stage approach

Blalock-Taussig shunt palliation may be complicated by pulmonary artery distortion, phrenic and vocal cord nerve injury, chylothorax, shunt narrowing, overcirculation and death.14 18

Neonatal repair of tetralogy of Fallot has low mortality (<5%).19 20 Advantages of primary repair include early abolishment of cyanosis, minimisation of right ventricular hypertrophy and fibrosis, avoidance of left ventricular volume-loading from palliative shunts and potential reduction in dysrhythmias.15 16 Postoperative signs of greater physiological stress manifest as longer intensive care unit stay are reported when repair is performed under 3 months of age.1 11 16 17 19 Furthermore, the neonatal brain may be more prone to surgery-related neurological injury.2123

Hypoplastic pulmonary arteries

Some centres advocating neonatal primary repair make exceptions in the setting of small pulmonary arteries, and perform palliation by balloon dilation of the pulmonary valve and arteries or ventricular septal defect patch fenestration.5 6 16 19 20 Several single centre reports have outlined the approach and results of repair of young infants with TOF with and without diminutive pulmonary arteries (Nakata index <100 mm2/m2). A two-staged approach was applied in those with diminutive pulmonary arteries while those with confluent pulmonary arteries of adequate size in the absence of haemodynamically significant aortopulmonary collaterals were deemed suitable for primary repair. In the primary repair group, early mortality was 3%20 and in the staged group, early and late mortality were 10.6% and 10%, respectively.6

Pulmonary arteries may be intrinsically diminutive with limited potential for growth or small vessel diameters may be secondary to decreased flow accompanying RVOT obstruction.4 11 The latter is demonstrated by increase in pulmonary artery dimensions after manoeuvres to augment pulmonary blood flow including surgical shunts, RVOT balloon dilation or stenting as in our series.4 5 7 11 16 24 As such, palliation may minimise or obviate the need for surgical augmentation of pulmonary arteries at time of repair. In our series, only patients with LPA stenosis related to the insertion of the ductus arteriosus or aortopulmonary collateral required surgical augmentation despite initial “diminutive” pulmonary arteries.

Muscular right ventricular outflow obstruction

The inconsistent benefit of pulmonary valve balloon dilation alone is a reflection of the frequent associated infundibular stenosis.8 11 25 Palliative stenting of the native right ventricular outflow to address muscular obstruction has been described in case reports in the settings of pulmonary atresia with intact ventricular septum, pulmonary stenosis with RV hypertrophic cardiomyopathy and tetralogy of Fallot in a 970 g infant.2527 Stent positioning is critical and even with dramatic improvement in oxygen saturation with the first implant, additional stents may be required to span the entire distance from the infundibular obstruction across the pulmonary valve. Otherwise, our experience would suggest that cyanosis will recur, requiring either an additional interventional procedure or surgical correction. This procedure secures pulmonary blood flow with dramatic improvements in systemic arterial oxygen saturataion without risk of pulmonary over-circulation as seen with aortopulmonary shunts, and provides a conduit for subsequent diagnostic and therapeutic catheterisation procedures.5 19 Stent implantation did not compromise surgical repair significantly and was easily removed in toto.

The pulmonary valve of TOF

Stents should only be implanted in patients whose RVOT will eventually require a transannular patch or conduit as the potential for pulmonary valve-sparing repair is eliminated by stent implantation. Reported transannular patch rates are high (36–100%) when surgery is performed in infancy.10 11 24 In our institution, only 7% of infants with primary repair under 2 months had preservation of the pulmonary valve (data to be published, D Fruitman, personal communication). Reports vary as to whether balloon dilation of the RVOT provides for growth of the pulmonary valve significant enough to allow for subsequent preservation.5 7 10 The pulmonary valve annulus diameters were extremely small in our series (mean pulmonary valve diameter of 3.7 mm (2.7–4.2) and and Z-score of −6.7 (−9.7 to −5.4)). Furthermore, long-term valve competency rates of surgical pulmonary valvuloplasties performed on infants requires further study.


In TOF-PA, there is a wide range of pulmonary artery anatomy, from essentially normal vessels supplied by an arterial duct to diminutive or absent central pulmonary arteries fed by aortopulmonary collaterals. Repair in the latter group has a high morbidity and mortality (35–36%).19 28 While some advocate for primary repair with unifocalisation,29 others support palliation with an aortopulmonary shunt or establishment of right ventricular to pulmonary artery continuity, generally by 3 months.6 19 30 In this regard, radiofrequency perforation of the imperforate pulmonary valve followed by stent implantation of the RV outflow tract establishes antegrade pulmonary blood flow, provides access for pulmonary artery rehabilitation while avoiding cardiopulmonary bypass.12 Concurrent coil embolisation of collateral vessels allows for continued recruitment of central pulmonary arteries and may minimise the need for unifocalisation procedures.5 19

Current institutional strategy

The heterogeneity of our patient population in whom RVOT stenting was performed reflects the evolution of our institutional strategy for the symptomatic infant with tetralogy of Fallot during this time period. At present, neonates with TOF-PA with duct-dependent pulmonary blood flow associated with good size pulmonary arteries undergo primary repair. Elective repair of standard tetralogy of Fallot is performed at 4–6 months of age. In the very young infant with a patent hypoplastic pulmonary valve and severe cyanosis in conjunction with small pulmonary arteries or medical problems, treatment by RVOT stenting is considered as a prelude to surgical repair at 6 months.


In the symptomatic young infant with TOF, stenting of the RVOT provides a safe and effective management strategy, improving arterial oxygen saturations and encouraging pulmonary artery growth.


View Abstract


  • See Editorial, p 100

  • Competing interests: None.

Request permissions

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.

Linked Articles