Magnetic Resonance Imaging to Assess the Hemodynamic Effects of Pulmonary Valve Replacement in Adults Late After Repair of Tetralogy of Fallot
Background— Pulmonary regurgitation (PR) late after total correction for tetralogy of Fallot may lead to progressive right ventricular (RV) dilatation and an increased incidence of severe arrhythmias and sudden death. Timing of pulmonary valve replacement (PVR) is subject to discussion, because the effect of PVR on RV function in adults is unclear. In this study, MRI was used to assess the effect of PVR on RV function and PR. Clinical improvement was established by means of the NYHA classification.
Methods and Results— Twenty-six adult patients were included. Cardiac MRI was performed at a median of 5.1±3.4 months before and 7.4± 2.4 months after PVR. Mean preoperative PR was 46±10% (range, 25% to 64%). After PVR, 20 of 26 patients (77%) showed no residual PR, 5 patients showed mild residual PR, and 1 patient showed moderate PR. RV end-diastolic volume (RV-EDV) decreased from 305±87 to 210±62 mL (P<0.001), and RV end-systolic volume (RV-ESV) decreased from 181±67 to 121±58 mL (P<0.001). No significant change was found in RV-EF (42% versus 42%). However, RVEF corrected for regurgitations and shunting increased from 25.2±8.0% to 43.3±13.7% (P<0.001). Mean validity class improved from 2.0 to 1.3 (P<0.001).
Conclusions— In adult patients with PR and RV dilatation, late after total correction of tetralogy of Fallot, MRI measurements show remarkable hemodynamic improvement of RV function after PVR and improvement of validity. We therefore advocate a less restrictive management concerning PVR in these patients.
Received April 17, 2002; revision received July 8, 2002; accepted July 8, 2002.
Tetralogy of Fallot (TOF) is the most common form of cyanotic congenital heart disease,1 with a prevalence of 0.26 to 0.8 per 1000 live births.2 Total repair for TOF has been available for 50 years,3 with a favorable outcome in most patients. Now we are faced with an increasing number of patients who present with symptomatic right ventricular (RV) failure or (supra)ventricular arrhythmias. In these patients, there is a risk for sudden death.4,5⇓ Usually, these long-term adverse effects are the result of long-standing pulmonary regurgitation (PR).6 In the past it was thought that PR in Fallot patients was rather unharmful. However, precise measurements of RV volumes using MRI have visualized the important enlargement of the RV in patients with severe PR.7
Pulmonary valve replacement (PVR) can be performed electively with little risk and may improve symptoms of RV failure and provides excellent midterm survival.8–10⇓⇓ Previous echocardiographic measurements of RV dimensions in children and adolescents showed a decrease in end-diastolic volume (EDV) and end-systolic volume (ESV) after PVR.11–13⇓⇓ However, in adults, radionuclide angiography (RNA) measurements showed no effects of PVR on RV volumes and ejection fraction (EF).14
To date, MRI is the gold standard for evaluation of RV volumes and quantification of degree of pulmonary and tricuspid regurgitation (TR).15–18⇓⇓⇓ The purpose of this study was to determine the hemodynamic and clinical effects of PVR on RV function in adults late after repair of TOF.
Between 1993 and 2002, a total of 65 patients underwent PVR for PR late after correction of TOF. Since 1997, 26 consecutive adult patients late after repair of TOF, who underwent PVR and had been evaluated preoperatively with MRI, were included. All patients were reevaluated with MRI 6 to 12 months after the operation, according to our clinical protocol.
Patients characteristics are shown in Table 1. Median age at total repair was 5.0±4.2 years (mean, 5.7; range, 0.4 to 21.0 years). Eleven patients (42%) underwent a palliative procedure before repair. A trans-annular patch had been used in 10 patients. In 20 patients, RV end-diastolic volumes could be measured both before and after PVR, and PR could be measured in 25 patients. Ventricular volumes were corrected for body surface area (BSA). Left ventricular volumes and flow through the tricuspid valve were also measured. Furthermore, validity class was assessed before and after PVR.
Indication for PVR
Patients with residual lesions after repair of TOF were considered for PVR when objective evidence of important right ventricular dilatation was found, with or without deterioration of validity or the presence of tricuspid regurgitation (TR) or (supra)ventricular arrhythmias.9,19⇓ All patients had moderate or severe pulmonary regurgitation, and 13 (50%) had severe RV dilatation, defined as RV-EDV more than double the LV-EDV. A history of (supra)ventricular arrhythmias was present in 3 (12%) patients before PVR.
Diminished validity, assessed by the NYHA classification, was part of the indication in 10 (38%) patients (NYHA class II or higher). Two patients without reduced validity had moderate PR in combination with severe RV dilatation of >120 mL/m2, a residual ventricular septal defect (VSD), and QRS duration >180 ms.
Co-indications included the presence of (supra)ventricular arrhythmias, prolonged QRS duration (>180 ms), and a residual VSD.
PVR was performed at a median age of 29.2±9.0 years (range, 17.0 to 45.6 years). Additional procedures performed at the time of PVR are listed in Table 1. All patients were operated with normothermic or moderately hypothermic cardiopulmonary bypass. Most pulmonary valve insertions were performed on beating heart. Aortic cross-clamping was dependent on the surgeon’s preference or on concomitant procedures. Residual VSDs were closed in 3 patients. De Vega tricuspid annuloplasty was performed in 4 patients. Cryopreserved pulmonary homografts were used in all patients. Homografts were inserted in the orthotopic pulmonary position with one proximal and one distal end-to-end running suture after longitudinally opening the proximal pulmonary artery and slightly extending this incision if necessary across the former pulmonary annulus. Calcified outflow tract patch material was resected as much as possible.
MRI studies were performed with a 1.5 Tesla system (NT15 Gyroscan, Philips Medical Systems, Best). Scout images were obtained in transverse, coronal, and sagittal planes using a standard, multislice spin echo sequence.
Short-Axis Gradient Echo Images
A multiphase, ECG-triggered, multishot echoplanar gradient echo (GRE) technique was used to acquire short-axis images. We used the short-axis plane because it allows measurement of both right and left ventricular volumes in one sequence. Images were acquired during breath holds, each lasting 10 to 15 seconds. The scout images were used to acquire 10 to 12 sections that covered both ventricles in the transverse plane; slice thickness was 10 mm with a 0.8- to 1.0-mm section gap. The flip angle was 30 degrees, and the echo time was 5 to 10 ms. Eighteen to 25 frames resulted in a temporal resolution of 22 to 35 ms. A field of view of 350 to 400 mm was used depending on the size of the patient.
Quantitative Flow Measurements
Velocity mapping was performed with the use of a velocity-encoded phase contrast sequence. A section thickness of 8 mm and a field of view of 300×300 mm were used. The flip angle was 20 degrees, and the echo time was 12 ms. For velocity mapping of the pulmonary artery, sagittal and coronal spin-echo scout images were used to construct a double oblique plane perpendicular to the vessel. Pulmonary flow measurements were performed halfway between the pulmonary valve and the bifurcation or ≈2 cm proximal to the bifurcation when no pulmonary valve was present. For velocity mapping of the flow through the tricuspid valve, 2- and 4-chamber (GRE) images were used to construct a parallel plane through the valve. The sequence was encoded for through-plane velocities up to 200 cm/s. With the use of retrospective gating, 30 to 40 time frames evenly distributed over the cardiac cycle were constructed, resulting in a temporal resolution of 25 to 35 ms. A 128×128 matrix was interpolated to a display matrix of 256×256. The MR examination lasted 45 to 60 minutes. No sedation was used in any of the patients.
All images were quantitatively analyzed on an IPC workstation (SUN Microsystems Inc) using two software packages, which were developed at our institution. Velocity maps were analyzed using the FLOW analytical software package.20–22⇓⇓
Flow curves were obtained for flow in de main pulmonary artery and just distal from the tricuspid valve. Regurgitation fraction was calculated by the following formula: regurgitant flow/systolic forward flow×100. The transverse GRE sequences of the ventricles were analyzed using MASS software.20–22⇓⇓ The EDV was assessed at the phase with the largest ventricular diameters, and the ESV was assessed at the phase with the smallest ventricular diameters. Endocardial contours of both right and left ventricles were drawn for all slices within these two phases. RV stroke volume (SV) was calculated by deducting the ESV from the EDV, and RV-EF was defined as (RV-SV×100)/RV-EDV. RV-EF was then corrected for regurgitation of the tricuspid and pulmonary valves, as well as intracardiac shunts by dividing the net pulmonary flow by the RV-EDV: RV-EFcor=net pulmonary flow (pulmonary forward flow−regurgitant flow)/RV-EDV.
The data were analyzed using SPSS for Windows (version 10.0, SPSS). The paired t test was used to evaluate preoperative and postoperative data.
PVR was performed successfully in 26 consecutive patients, with no perioperative mortality. One patient died suddenly 18 months after uncomplicated pulmonary valve replacement. No autopsy was performed.
Cardiac MRI was performed in all patients with a median of 5.1±3.4 months before PVR and repeated at a median of 7.4±2.4 months after the operation. Table 2 shows MRI and clinical data before and after PVR.
Before PVR, all patients had at least moderate PR. PR decreased from a mean of 46±10% to 4±8% (P<0.001). Twenty patients had no residual PR; mild residual PR was seen in 5 patients, whereas one patient had moderate residual PR. This patient did not show reduction of right ventricular volumes after operation.
Before PVR, mean indexed RV-EDV (RV-EDV-I) was 167± 40 mL/m2 (range, 113 to 289 mL/m2). After the operation, mean RV-EDV-I was 114±35 mL/m2 (range, 58 to 243 mL/m2). Mean RV-ESV-I decreased from 99±36 mL/m2 (range, 52 to 192 mL/m2) to 66±35 mL/m2 (range, 25 to 210 mL/m2), resulting in an average of both volumes of ≈30% (P< 0.001). Left ventricular volumes did not change significantly after operation, indicating the relative reliability of the measurements (Figure 1). After PVR, RV-EF did not change significantly, 42% versus 42%, whereas RV-EF corrected for regurgitation and shunting increased dramatically from 25±8% to 43±14% (P<0.001).
Average validity was calculated before and after PVR according to the NYHA class. Before PVR, the average validity class was 2.0±0.6, whereas after PVR, validity improved to 1.3±0.5 (P<0.001) (Figure 2).
Fourteen patients had validity class I after operation. In 12 patients, no normalization of validity was observed. In the latter group, the mean preoperative validity was 2.3±0.7 compared with a mean validity of 1.8±0.5 in the group with normalization of validity (P=0.01). Normalization of validity was not obtained in patients with preoperative validity class III.
To our knowledge, this is the first study using MRI to show improvement of RV function after PVR late after total repair for tetralogy of Fallot in patients with moderate to severe pulmonary regurgitation and right ventricular dilatation. In our study, 25 of 26 patients (96%) had no or mild residual PR, and a reduction of RV volumes of 30% was observed. Furthermore, an improvement of validity was reached; most patients with symptoms before surgery were symptom free after PVR. We also found a dramatic improvement of RV function as expressed by the right ventricular ejection fraction corrected for regurgitation and the right ventricular end-systolic volume.
The reduction of the volume of the dilated right ventricle is in accordance with the results of several previous studies that investigated the effect of PVR in children and adolescents. Bove et al13 found after PVR in a group of patients with a mean age of 14.6 years a reduction in RV volume (measured with M-mode Echo). Warner et al12 studied a group of patients with a mean age of 12 years. Their M-mode echo measurements also showed a decrease in RV volume after PVR. d’Udekem et al,11 who investigated PVR at a median age of 13 years using echocardiography, showed a significant decrease in the mean ratio between the end-diastolic diameter of RV and LV. Therrien et al14 found, in contrast to our results, no reduction in right ventricular volumes after PVR using Radionuclide angiography in a group of adults with a mean age of 33.9 years and therefore similar to our population (median age, 29.2 years).
RV-EF did not improve significantly; the reduction in RV-EDV was compensated by reduction in SV after correction of PR. Because SV is greatly determined by the volume of regurgitant flow, this method is not useful in load-dependent conditions, eg, before and after PVR. We corrected RV-EF for PR and TR as well as residual intracardiac shunts by using the net pulmonary flow, instead of the SV based on ED and ES volumes. Using this method, a dramatic increase in RV-EFcor was seen from 25% to 43%. However, RV-EDV is also load-dependent, and therefore RV-ESV might be the best parameter for systolic function of the RV. This parameter showed a decrease of 30%.
Symptomatic improvement in clinical symptoms was striking. After 6 months, validity class had improved significantly, and most patients with symptoms before surgery were now symptom free. This is in accordance with previous studies.8,14,23⇓⇓ Normalization of validity was dependent on preoperative validity class.
The differences between our results and those of Therrien et al14 are remarkable. In both studies, mean age at PVR was ≈30 years. Nonetheless, our patients underwent total correction at earlier age than those of Therrien et al14 (our group versus their group, 5.7 versus 12.1 years). Additionally, in their group, 40% underwent concomitant relief of infundibular stenosis, whereas only 2 (8%) of our patients underwent such a procedure. Another, more likely, explanation is the measurement technique. In a later study by Therrien et al24 of patients with a median age of 28.2 years, transthoracic echocardiographic measurements showed a decrease in the number of patients with moderate to severe right ventricular dilatation after PVR compared with preoperatively. This is in accordance with our findings.
A regurgitant pulmonary valve is usually replaced by a pulmonary homograft. However, bioprosthetic conduits and mechanical valve prostheses have been reported as alternatives.23 We used homografts, because they are excellent pulmonary valve substitutes with very low gradients. Pulmonary homografts are preferred over aortic homografts in the pulmonary position because of their better durability.25 Pulmonary homografts may suffer from calcific degeneration and may need later replacement.25 Some surgeons use mechanical valves in the pulmonary position, but because the patients are usually young, and females may develop a wish for pregnancy, we prefer to avoid life-long anticoagulation. The lifetime of a pulmonary homograft remains unclear. Possibly, the durability of cryopreserved pulmonary homografts used for late PVR after Fallot repair is similar to that of pulmonary homografts, which are orthotopically used to reconstruct the right ventricular outflow tract in the Ross procedure. In the International Ross Registry report of 2523 documented Ross operations, freedom from reoperation on the right ventricular outflow tract was 92% at 10 years and 85% at 25 years.26
Optimal timing of pulmonary valve replacement is still a subject of debate. The amelioration of right ventricular function has to be weighed against the risk of later reoperation for homograft failure. Even in symptom-free patients with right ventricular dilatation as a result of moderate to severe pulmonary regurgitation, we advocate pulmonary valve replacement to prevent and reduce right ventricular dilatation. Additional arguments for this strategy are the predisposition of moderate to severe pulmonary regurgitation for ventricular dysrhythmias27 and the beneficial effect of pulmonary valve replacement on electrical instability and the incidence of tachyarrhythmias.24
This study has some limitations, such as the fact that we were not able to acquire full MRI data in all patients. A common problem, especially in patients with complex cardiac malformations, is the inability to get sufficient cardiac triggering. At present, this problem is minimized by using vector ECG triggering and improved imaging sequences, such as balanced FFE. These techniques were not used in the present study. In our patients, the occurrence of triggering problems was not associated with larger RV volumes or arrhythmias. Motion and metal artifacts (eg, sternal clips) are other well-known problems.
In conclusion, in adult patients late after total correction for tetralogy of Fallot who undergo PVR for moderate to severe pulmonary regurgitation and right ventricular dilatation, MRI measurements showed remarkable hemodynamic improvement. Furthermore, normalization of validity is more likely in patients with preoperative validity class II to III or better than in patients with worse validity. We therefore advocate a less restrictive strategy concerning pulmonary valve replacement in patients with moderate to severe PR and RV dilatation late after total repair for tetralogy of Fallot.
- ↵Allen HD, Gutgesell HP, Clark EB, et al. Moss and Adams’ Heart diseases in Infants, Children, and Adolescents: Including the Fetus and the Young Adult. 6th ed. Philadelphia: Lippincott Williams & Wilkins; 2001: 880–902.
- ↵Gatzoulis MA, Till JA, Somerville J, et al. Mechanoelectrical interaction in tetralogy of Fallot: QRS prolongation relates to right ventricular size and predicts malignant ventricular arrhythmias and sudden death. Circulation. 1995; 92: 231–237.
- ↵Rebergen SA, Chin JG, Ottenkamp J, et al. Pulmonary regurgitation in the late postoperative follow-up of tetralogy of Fallot: volumetric quantitation by nuclear magnetic resonance velocity mapping. Circulation. 1993; 88: 2257–2266.
- ↵Hazekamp MG, Kurvers MM, Schoof PH, et al. Pulmonary valve insertion late after repair of tetralogy of Fallot. Eur J Cardio Thorac Surg. 2001; 19: 667–670.
- ↵Therrien J, Siu SC, Harris L, et al. Impact of pulmonary valve replacement on arrhythmia propensity late after repair of tetralogy of Fallot. Circulation. 2001; 103: 2489–2494.