Right Ventricular Outflow Tract Stenting

• Editorials
• catheterization
• comorbidity
• cyanosis
• tetralogy of Fallot

For cyanotic young infants with tetralogy of Fallot (TOF), treatment options include, broadly, primary anatomic repair versus some form of palliation that would forestall complete anatomic repair until a more suitable time. With concerns about performing an early primary anatomic repair, particularly in a child with risk factors, the desire to identify the ideal palliative option is a pressing one. Characteristics of the ideal palliation in this situation might include the following: (1) providing a stable yet balanced source of pulmonary blood flow; (2) allowing for growth and development of the pulmonary arteries, ideally catch-up growth, because pulmonary arteries in this scenario are often hypoplastic; (3) providing sufficient durability to allow enough time to pass that the child grows, comorbidities (when present) subside, and the child reaches the time window when anatomic repair is deemed optimal; and (4) leaving behind no residue that might complicate the anatomic repair or increase the risk of subsequent reinterventions. At present, commonly used palliative options include surgical systemic–pulmonary shunt (typically a modified Blalock–Taussig shunt), ductus arteriosus stent, pulmonary balloon valvuloplasty, and right ventricular outflow tract (RVOT) stent. To date, there has been considerable debate on how many features of the ideal palliative procedure each of these options holds.

See Article by Sandoval et al

In the current issue of Circulation: Cardiovascular Interventions, Sandoval et al¹ at the Hospital for Sick Children in Toronto make a compelling case that stenting of the RVOT is a serious contender for the title of ideal palliative procedure. They performed a detailed retrospective review of their experience managing infants with TOF. Infants were treated in 1 of 4 ways: those with early cyanosis (<3 months of age) were treated with either primary repair (early-PS group in those with pulmonary stenosis and early-PA group in those with pulmonary atresia) or RVOT stenting (stent group), whereas those without early cyanosis had primary repair electively at an age deemed optimal between 3 and 11 months (surg>3mo group). The surg>3mo group was randomly selected from this center’s large experience performing elective surgical repair of TOF to maintain a relatively balanced number of patients in each group between 42 and 49. Patients were not, of course, assigned randomly to a treatment group. Instead, the decision to perform early primary repair versus RVOT stenting was based largely on the presence of risk factors. Those patients with low weight or other comorbidities and a small pulmonary valve annulus whose competency was felt unlikely to be maintained at surgical repair were more likely to be assigned to the stent group, although the authors acknowledge that this was ultimately an individualized decision. As expected, the early-PS, early-PA, and stent groups all had similar cumulative and postoperative lengths of stay, all significantly longer than that for the surg>3mo group. Most importantly, the stent group at baseline had significantly smaller pulmonary arteries than the other 3 groups at a median Nakata index of 79 mm²/m². After RVOT stenting, there was significant catch-up growth of the pulmonary arteries. Although they remained somewhat smaller at the time of ultimate anatomic repair (median 147 mm²/mm² compared with 167 mm²/m² in the surg>3mo group), this difference was not statistically significant. This is an impressive result, and the authors should be commended. On close inspection, however, there are several potential drawbacks to this strategy. First, stenting of the RVOT, particularly in a small infant with a hypoplastic or atretic pulmonary valve, can be a technically difficult procedure. The authors report in the supplementary material that inadvertent perforation of the RVOT occurred in 6 patients (13% of the 46 patients in whom an attempt to stent the RVOT was made). It is a testament to their skill that this resulted in only 2 emergent pericardiocenteses and not more numerous or more serious complications. In addition, there is likely a significant learning curve to perform this procedure. In fact, the authors have previously reported on their early experience with RVOT stenting,² a cohort that partially overlaps with the present cohort. This experience covers their learning curve, a period of time during which procedural modifications occurred, including the type of stent used and the degree to which the infundibulum was covered. Second, patients in the stent group had a high burden of reinterventions. It is unclear whether these were included in the cumulative hospital length of stay data reported, but this clearly should count as a cost. Third, stenting of the RVOT prevents any consideration for a valve-sparing anatomic repair. The authors make a reasonable argument that these pulmonary valves were too small to be considered for a valve-sparing repair anyway, but this is a commitment that needs to be recognized and acknowledged. Fourth, there is some suggestion that the presence of the stent in the RVOT may complicate the subsequent anatomic repair. It is impressive that 95% of all stents were able to be removed entirely at the time of complete anatomic repair. However, this is substantially higher than the 44% success rate reported previously by Barron et al³ in a series of RVOT stents, which included 15 patients who eventually underwent complete anatomic repair, and the source of this disparity is not clear. Furthermore, the cardiopulmonary bypass time was significantly longer in the stent group than in all other groups at a median of 139 minutes, a potential concern based on previously reported associations between longer cardiopulmonary bypass time and worse subsequent clinical outcomes at the time of elective anatomic repair.⁴ Is it possible that this additional time was related to stent removal as was suspected in the experience by Barron et al? Anecdotally, other concerns related to the incomplete removal of stent material include difficulty sewing the ventricular septal defect patch and interference with aortic valve function. It would be interesting to know whether the rate of (unintentional) residual ventricular septal defect and postoperative aortic insufficiency were different between the groups. Finally, although all 4 groups of patients seem to be doing well in the short term to midterm, ultimately the more important longer-term outcomes will be related to right ventricular dilation and function, QRS duration, and need for pulmonary valve replacement and other reinterventions. Hopefully, the authors will continue to follow this valuable cohort to address some of those questions at a later time.

Given these potential limitations of RVOT stenting, it is worth exploring the other options for management of the symptomatic young infant with TOF. First, early primary anatomic repair has been reported with reasonable results in several previous reports.^5–9 The definition of a risk factor, which precludes early primary repair, may vary from center to center, and so, consideration of early primary repair may be reasonable for some symptomatic young infants depending on a center’s comfort level and past experience (as well as their comfort and experience with alternative palliative options). One argument for early primary repair is to avoid the progressive right ventricular hypertrophy that can occur in the unrepaired form and may complicate the ultimate anatomic repair. It is possible that the longer postoperative length of stay in the stent group may have been, at least in part, because of a poorly compliant right ventricle. Additional information on the authors’ institutional practice with regard to intentionally leaving an atrial communication at the time of anatomic repair and the postoperative oxygen saturations may have been helpful in further exploring this issue. Among palliative options, Blalock–Taussig shunt placement continues to be the de facto choice at many centers.¹⁰ However, this also continues to be a relatively morbid and mortal operation,¹¹ and there are concerns about its effects on growth of the pulmonary valve and pulmonary arteries.^12,13 Palliative balloon pulmonary valvuloplasty has also been used with reasonable success,¹⁴ although issues related to durability and limited efficacy in patients who also have significant subvalvar obstruction may prevent its widespread use and applicability. Finally, ductus arteriosus stenting has more recently been introduced as a potentially appealing palliative option with numerous reports of its successful use, including encouraging effects on pulmonary artery growth.^15–17 In a few head-to-head comparisons at single centers, ductus arteriosus stenting has performed favorably compared with Blalock–Taussig shunt placement,^18–20 although it is important to note that these are small samples of patients who were not randomly assigned to treatment strategy, and so, the results are at risk of being both underpowered and subject to confounding by indication.

It is purely speculative to predict how these various palliative options would perform against each other in a randomized trial. As an example, it is not clear that the right ventricle and diminutive pulmonary arteries of a symptomatic young infant with TOF would be better served by the flow patterns provided by a RVOT stent versus (as an example) a ductus arteriosus stent. However, based on the data presented by Sandoval et al in this issue, it is safe to assume that RVOT stenting would certainly perform competitively. If such a trial could ever come to fruition, the authors’ suggestion of using key eligibility criteria of a Nakata index ≤100 mm²/m² and pulmonary valve z score <−5 is a valuable one. Regardless, in the end, a center will likely find its ideal palliative option to be the one that it performs best. For those centers comfortable with RVOT stenting, it seems to be an appealing option to consider.

• © 2016 American Heart Association, Inc.