Fresh Autologous Pericardium to Reconstruct the Pulmonary Valve at the Annulus When Tetralogy of Fallot Requires a Transannular Patch at Midterm

Abstract

Tetralogy of Fallot often requires reconstruction of the right ventricular outflow tract with a transannular patch (TAP), but this renders the pulmonary valve incompetent and eventually leads to right ventricular dysfunction.

We retrospectively evaluated the efficacy of a reconstructed pulmonary valve and annulus in 70 patients who underwent, from December 2006 through December 2010, complete correction of tetralogy of Fallot. We divided the 70 patients into 2 groups in accordance with whether they required (n=50) or did not require (n=20) a TAP. We used autologous untreated pericardium to fashion the TAP and to create both an annulus of the correct size and a competent pulmonary valve with native leaflets. We evaluated the efficiency of this procedure both functionally and anatomically.

The median age of the patients was 11 years (range, 2–38 yr). There were 56 males, with no significant difference in sexual distribution between groups. The clinical follow-up was 88% for 57.5 months, and the echocardiographic follow-up was 80% for 36 months. There was no significant difference in outflow gradient or in the occurrence of pulmonary insufficiency between the TAP group (none, 31; mild, 12; moderate, 6; and severe, 1) and the No-TAP group (none, 16; moderate, 2; and severe, 2) (P=0.59). Nor was there any thickening or calcification in the constructed valves.

We conclude that pulmonary valves constructed of untreated autologous pericardium performed as well as native valves after total tetralogy of Fallot correction at midterm.

Pulmonary insufficiency (PI) is an unwarranted side effect of the transannular patch (TAP) used for relieving the right ventricular outflow tract (RVOT) obstruction that is characteristic of tetralogy of Fallot (TOF).¹ With time, the presence of PI causes right ventricular (RV) overload and eventually leads to RV dysfunction.² Attempts to reduce insufficiency at the pulmonary valve (PV)—such as limiting the extent of infundibular incision or using a smaller transannular patch—have met with limited success.^3,4 A TAP for RVOT reconstruction relieves its obstruction; but its expansion, especially at the level of the PV, cannot be governed, and this results in free pulmonary regurgitation. The use of a smaller TAP not only fails to relieve the RVOT obstruction efficiently but is unable to curtail the free PI. A procedure that can resolve the problem of PI with relief of RVOT obstruction must fulfill these 2 criteria: 1) limit the expanse of the TAP at the level of the native PV, while enabling it to bulge (and thereby increase the room in the RVOT to relieve obstruction) and 2) provide a competent PV. This study evaluates the efficiency of one such procedure at creating a competent PV complex (that is, a competent PV and a new pulmonary annulus) in TOF patients who need a TAP for relief of RVOT obstruction when undergoing complete intracardiac repair. This procedure was tested at midterm follow-up.

Patients and Methods

This is a retrospective analysis of 70 cases of children who underwent total cardiac repair for TOF from December 2006 through December 2010. Our institutional ethics committee granted us permission to collect data from their records. These patients were segregated into 2 groups: Group I (n=50) needed a TAP for repair of the RVOT, and Group II (n=20) could avoid the TAP. Patients with a Z value of minus 2 (determined on the basis of body weight) or greater (in accordance with the size of the pulmonary arteries at first branching) were considered for total correction. The decision for TAP across the RVOT was made on the basis of pulmonary annulus size. A Z value of minus 2 or greater made a case for the use of TAP across the RVOT.⁵

Surgical Procedure

The operation was performed through a median sternotomy. Standard ascending aortic and bicaval cannulation was performed after pericardial harvesting and systemic heparinization. Cardiopulmonary bypass was instituted and the operation was conducted at 28 °C. After the ascending aorta was cross-clamped, cold-blood St. Thomas cardioplegic solution was administered antegrade at the rate of one dose every 20 minutes. After the caval cannulas were in place, the right atrium was entered and the left atrium was then vented through the atrial septum. The ventricular septal defect was closed with a Dacron patch. The infundibular resection was started at the right atrium and completed at the pulmonary artery. In Group II cases, the PV was inspected and a commissurotomy performed if stenosis was present. In Group I cases requiring annular enlargement, the annulus was divided anteriorly and at least 2 cm into the infundibulum until adequate relief of the RVOT obstruction was achieved. Autologous pericardium, untreated, was harvested and used as a TAP and for reconstruction of the anterior PV cusp.

The size of the anterior cusp of the soon-to-be-fabricated PV was calculated as follows:

where the desired pulmonary annulus diameter provides a Z score of −2 or greater, depending on the weight of the patient, and

This patch is stitched on the undersurface of the TAP at the level of the native PV. This maneuver creates the new pulmonary annulus and creates the anterior cusp of the PV that helps, in union with the posteriorly placed native cusps, to make the valve competent. A detailed description and early results are published in our previous study.⁶

Clinical and Echocardiographic Follow-Up

Clinical follow-up was performed largely through out-patient interview of the patients. Echocardiography was performed on a Vivid 7^® cardiac ultrasonography system (GE Medical Systems; Horten, Norway), with 3-dimensional (3D) software and a recorder. A short-axis parasternal view was used to profile the RVOT and to examine the PV. This view was used to evaluate the RVOT gradient and PI. Three-dimensional echocardiography was performed to evaluate the anatomy of the reconstructed PV. This included the thickness of the leaflet tissues, the motion of the anterior cusp of the PV, and any evidence of calcific deposits on the valve.

Evaluation of the Procedure's Efficacy

The following parameters were used to measure functional and anatomic competence of the reconstructed PV. Pulmonary insufficiency was labeled mild, moderate, or severe on the basis of American Society of Echocardiography guidelines.⁷ The RVOT mean gradient was measured and labeled good (<20 mmHg), mild (20–30 mmHg), moderate (30–50 mmHg), or severe (>50 mmHg). Tricuspid regurgitation was labeled mild, moderate, or severe; and RV systolic pressure was measured whenever a tricuspid regurgitant jet appeared on echocardiography.

The anatomic attributes of the reconstructed valves were studied by means of 3D echocardiography in patients who presented after a follow-up period of longer than 60 months.

Statistical Analysis

All the data are presented as median, so they range between minimum and maximum. The nonparametric Mann-Whitney U test was used to compare the 2 groups. The χ² test was used in a 2×2 table while comparing 2 variables. The Pearson bivariate correlation was performed among variables. Cumulative survival was performed by Kaplan-Meier analysis. The P value of <0.05 was considered statistically significant. All analysis was performed with use of SPSS version 10 for Windows (IBM Corporation; Endicott, NY).

Results

The median age at operation was 11 years (range, 2–38 yr), and 56 of the 70 children were male. Age and sex were distributed equally in the 2 groups (Table I). Right and left pulmonary artery sizes at the first branching and Nakata indices were also not significantly different between the 2 groups (Table I). Pulmonary insufficiency grades were similar, as were RVOT gradient and RV systolic pressure (Table II). Age at operation correlated positively to RV systolic pressure (r=0.433; P=0.008). Figure 1 displays a sample gradient across the RVOT and an absence of pulmonary regurgitation consequent to a competent valve.

TABLE I.

Preoperative Variables and Follow-Up Results

TABLE II.

Comparison of Postoperative Pulmonary Insufficiency, Right Ventricular Outflow Tract Gradient, and Right Ventricular Pressure

Fig. 1.

Transthoracic echocardiograms. A) A continuous-wave Doppler signal depicts a gradient across the right ventricular outflow tract. B) Color-flow Doppler mode shows a competent pulmonary valve with no pulmonary insufficiency.

Clinical follow-up was 88% complete (61 of 70 patients) at a median of 57.5 months (range, 33–84 mo), and echocardiography was performed in 80% of patients (56 of 70), at a median follow-up period of 36 months (range, 6–72 mo). Twelve patients in Group I had echocardiographic follow-up of longer than 60 months, and anatomic boundaries in those patients were evaluated with use of 3D echocardiography (Fig. 2). Follow-up results were similar in both groups.

Fig. 2.

At the 5-year follow-up evaluation, 2-dimensional (2D) and 3-dimensional (3D) transthoracic echocardiograms show a pulmonary valve that was created with fresh autologous pericardium. A) The reconstructed pulmonary valve in the open position. The left panel shows 2D transthoracic modified short-axis views. The right panel shows a 3D reconstruction. B) The reconstructed pulmonary valve in the closed position. The left panel shows 2D transthoracic modified short-axis views. The right panel shows a 3D reconstruction. The leaflet thickness, pliability, and lack of calcification are evident.

Supplemental motion image is available for Figure 2 ^{(3.2MB, mp4)} .

There were 3 deaths: 2 in Group I and one in Group II. All deaths arose from low cardiac output and multiorgan dysfunction. There were 3 cases in which major aortopulmonary collateral vessels were coil-embolized before total correction.

Kaplan-Meier analysis for the appearance of PI (Fig. 3), for the regression of RVOT gradient below 30 mmHg (Fig. 4), and for the regression of RV systolic pressure below 35 mmHg (Fig. 5) were performed. The times of appearance of PI in the 2 groups were not significantly different: Group I, 36 months (range, 0–64 mo) versus Group II, 48 months (range, 0–72 mo), P=0.063. Similarly, the times for regression of the RVOT gradient (Group I, 24 mo [range, 0–72 mo] vs Group II, 20.5 mo [range, 0–48 mo] P=0.926) and the times for regression of RV systolic pressure (Group I, 36 mo [range, 0–72 mo] vs Group II, 41 mo [range, 0–60 mo]; P=0.575) were not significantly different.

Fig. 3.

Graph of the Kaplan-Meier analysis shows no statistical difference between the groups in the appearance of pulmonary insufficiency (PI).

Fig. 4.

Kaplan-Meier analysis of regression of right ventricular systolic pressure, as evidenced by decreases of right ventricular outflow tract (RVOT) gradients below 30 mmHg. Time to regression of the residual RVOT gradient was similar for both groups.

Fig. 5.

Graph shows the Kaplan-Meier analysis for regression of peak right ventricular systolic pressure (PRVSP) below 35 mmHg. In both groups, PRVSP regressed to acceptable limits at similar durations of follow-up.

Discussion

Tetralogy of Fallot is the most prevalent cyanotic congenital heart disease.⁸ It is also the first cyanotic heart disease to have a total and successful correction and cure. In India, the median age of operation for TOF is higher than is usual in Western countries. This is a consequence of socioeconomic circumstances; similar circumstances are found in some other countries. Moreover, RV dysfunction in TOF patients who undergo correction at India's median age of 11 years is risky: immediate PI can impede a good outcome.⁹ Hence, the study of that subset of patients cannot be neglected. A large database¹⁰ has revealed that TAP is used in 20% to 57% of TOF cases, in an attempt to relieve RVOT stenosis. Unfortunately, this procedure destroys the integrity of the native PV and annulus, leading to PI that gradually worsens in the years to follow.³ Pulmonary insufficiency causes RV dysfunction and symptoms that worsen, first in exercise and then at rest.¹¹ These patients usually develop free PI in the 10- to 20-year period after total correction.¹² About 10% to 15% of these patients will need PV replacement for correction of PI.¹³

A procedure that can avoid or modify this sequence of events will unquestionably be useful. Previously, surgeons used smaller TAPs and infundibular incisions to limit free regurgitation of the PV. These measures indeed slowed the effect of PI on the RV,¹⁴ but this mere delay in RV dysfunction led eventually to a fate similar to that of patients who developed isolated PI after the TAP.¹⁵ The more fruitful solution, as we have outlined it above, relieves the RVOT stenosis as well as possible and creates a competent PV in patients who really need a TAP. Repair of the PV (after the TAP procedure) is the most suitable of options for reconstruction of the RVOT.¹⁶

It is important to stress that the median age of repair in this group of TOF patients is 11 years, forming a subset that has not undergone evaluation in the aforementioned studies. Because of this deficiency in research, the progress of PI in this group cannot be predicted by available data. Furthermore, we ourselves did not evaluate the hemodynamics of the RV—particularly the end-diastolic pressure that can alter the appearance of PI and might be important for this group. Pulmonary insufficiency appearing after the TAP procedure is not the only variable that worsens RVOT function or predicts reoperation after successful repair.¹⁷ The right atrial area and the akinetic length of the RVOT after repair have been associated with sustained tachycardia.¹⁸ Despite these observations, maximum harm to the RV is induced by PI.^19,20

The procedure discussed in this study helps in creating a new pulmonary annulus and a competent valve for that size annulus by creating an anterior leaflet of the PV that coapts with the native leaflets. In comparing the results of this procedure with results wherein a TAP was not needed and the native valve was preserved, we observed that this correction performs in a manner similar to the correction in our No-TAP group, with regard to the development and progression of PI.

This technique has not been used in patients younger than 2 years of age. Because the growth potential of the pericardium in use is unknown (as is that of the new annulus), we do not recommend the application of this technique to infants. In children who underwent more than 5 years of follow-up, we have found the reconstructed leaflet of the new PV to be supple, effectively mobile, and free of calcification, as witnessed in 3D echocardiograms. The gradients across the RVOT and the PV in our Group I were also comparable to those in our No-TAP group. The RV systolic pressure has been similar in both groups, and the reconstructed valve has been able to maintain its performance at the peak pressure of 56 mmHg. The RVOT by age 11 is very muscular, even to the point of rotating laterally. Complete relief of obstruction is sometimes not possible, and a higher gradient is usually accepted.

We used untreated autologous pericardium for this procedure. Previous reports have not favored the use of untreated pericardium, for it is known to contract and become dysfunctional early.²¹ Because the early applications of untreated autologous pericardium failed, this method fell out of repute and was abandoned in favor of glutaraldehyde treatment. However, there is now a recent report of untreated pericardium's successful use in mitral valve repair at midterm follow-up.²² This present study also establishes the use of untreated pericardium at midterm follow-up. Glutaraldehyde-treated pericardium, although easy to handle and effective in early follow-up, does calcify and become stiff, because the fixation process kills the cells and breaks the cross-linkages.²³

Several surgical groups that have spared the RV infundibulum have experienced higher rates of repeat intervention for residual infundibular stenosis.²⁴ Attempts at sparing the annulus have given us very early results: mid- and long-term effects are not yet known.²⁵ Attempts to create pulmonary leaflets for bicuspid PVs with biologic and polytetrafluoroethylene (PTFE) materials have shown good early results: some studies have proved their efficacy even at midterm or longer.^26,27 Conversely, the construction of a monocusp from a homograft has been associated with destruction consequent to ABO blood group incompatibility, and with poor valvular function at midterm.²⁸ Although thin and impervious to the growth of fibrous tissue, PTFE is known to calcify in the long term.²⁹ It definitely has no growth potential. Evidence supports the idea that untreated pericardium, if it is still thin and freely mobile, will be able to function as a valve at midterm. Long-term follow-up of this attempt to avoid PI is yet to be carried out.

Conclusion

We conclude that this method of creating a competent PV in patients who need TAP while undergoing total correction for TOF is effective at midterm. Fresh untreated autologous pericardium holds promise for use at the pulmonary position. However, long-term evaluation of this procedure is essential.