Comparison of Long-term Postoperative Sequelae in Patients with Tetralogy of Fallot versus Isolated Pulmonic Stenosis

Abstract

Patients with tetralogy of Fallot (TOF) following complete repair and pulmonic stenosis (PS) after surgical valvotomy often develop significant pulmonic regurgitation (PR), eventually requiring valve replacement (PVR). Though criteria exist for the timing of PVR in TOF, it remains less clear when to intervene in valvotomy patients and whether TOF recommendations can be applied. Our aim was to compare the structural and functional sequelae of valvotomy for pulmonic stenosis (PS) with complete repair for tetralogy of Fallot (TOF). We compared the clinical characteristics, electrocardiograms, echocardiograms, cardiac MRI and invasive hemodynamics of 109 adults (34 PS and 75 TOF) newly referred to a congenital heart disease center for evaluation of PR between 2005 and 2012. Both cohorts were similar in terms of baseline demographics and presenting NYHA function class. Valvotomy patients had a slightly greater degree of PR by echo, though it was similar by cardiac MRI. ECG QRS width was greater in TOF (114±27 vs. 150±28 ms, p<0.001). MRI right ventricular ejection fraction (49±8 vs. 41±11%, p=0.001) and left ventricular ejection fraction (59±7 vs. 52±10%, p=0.002) were lower in TOF. Pacemaker or defibrillator implantation was significantly higher in TOF (3% vs. 23%, p=0.011). In conclusion, patients post-valvotomy and complete repair present with similar degrees of PR and symptom severity. Biventricular systolic function and ECG QRS width appear less affected, suggesting morphologic changes in TOF and its repair that extend beyond the effects of PR. These findings suggest the need for developing disease-specific guidelines for patients with PR post-valvotomy.

Introduction

Tetralogy of Fallot (TOF) and isolated pulmonary valve stenosis (PS) are common congenital disorders that often require early corrective intervention.^1–4 Traditional corrective surgery for TOF (“complete repair”) and surgical valvotomy for PS have frequently resulted in significant pulmonary valve regurgitation (PR).^5–9 In both populations, the resultant PR places patients at risk for right ventricular (RV) dysfunction, arrhythmias, and sudden death.^10–13 Pulmonary valve replacement (PVR) is the gold standard therapeutic option, as medical management is often ineffective.^14–16 The timing of PVR is controversial, with early replacement limited by long-term prosthetic valve degeneration and delayed replacement risking the development of irreversible RV damage.¹⁷ For this reason, the timing of PVR following TOF repair has been well-studied, and specific guidelines exist to determine the appropriate timing of intervention.^18–23 Unfortunately, few studies exist to guide the timing of PVR in the PS population.¹⁸ Many providers therefore apply the TOF guidelines to valvotomy patients, despite the fundamentally different anatomy, pathophysiology, and surgical corrections of these conditions. To assess whether the two populations are truly comparable, we examined the structural, functional and symptomatic sequelae of patients with PR resulting from TOF repair and valvotomy referred for initial assessment at an adult congenital heart disease center.

Methods

In this retrospective, incipient cohort study, we identified 109 patients from the Cleveland Clinic Adult Congenital Heart Disease Database who were newly-referred between July 2005 and June 2012 for evaluation of moderate or greater native valve PR (as quantified by echocardiography) related to prior surgical repair of PS or TOF. All patients underwent an electrocardiogram (ECG) and an echocardiogram at the initial clinic visit, with cardiac MRI and heart catheterization subsequently performed in appropriately selected patients. Patients with left bundle branch block, prior pulmonary valve replacement or those with >mild right ventricular outflow narrowing evident on imaging were excluded. Additionally, those with other major additional cardiac malformations (i.e. double chambered right ventricle, transposition of great vessels, atrioventricular canal type defect or anomalous pulmonary venous return) were excluded. Other cardiac comorbidities, including residual atrial and ventricular septal defects and peripheral pulmonic stenosis, were individually reviewed by a committee of investigators for their contribution to cardiac structure and risk of arrhythmia before possible exclusion. The study was approved by the Institutional Review Board of the Cleveland Clinic.

Data was abstracted from the database and augmented by review of the electronic and paper medical records. This included baseline demographics, cardiac medical and surgical history, comorbidities, symptom severity as measured by New York Heart Association (NYHA) functional class, electrocardiographic data, imaging data and hemodynamic tracings.

ECG analysis was limited to nonpaced tracings. Blinded qualitative assessments of cardiac chamber sizes by echocardiography were quantified as: 0=normal, 1=mild, 2=moderate, 3=moderate to severe and 4=severe enlargement. Similarly, right ventricular function by echo was quantified as: 0=normal, 1=mild, 2=moderate, 3=moderate to severe and 4=severe dysfunction. Cardiac catheterization data included right heart pressures, cardiac index calculated by Fick method and the presence and severity of coronary artery disease. Cardiac MRI data included chamber sizes and function, quantified valvular regurgitation and aortic dimensions. Volumetric data by MRI was indexed using body surface area.

Comparisons of dichotomous variables between groups were performed using the Pearson Chi-squared test, Fisher’s exact test or Wilcoxson sum-rank test where appropriate. Comparisons of continuous variables were performed using two-sided t-tests. For all tests, a p-value <0.05 was considered statistically significant. Data analyses were performed using JMP Pro software version 9.0.0 (SAS Institute Inc., Cary, NC).

Results

109 patients (34 PS valvotomy and 75 TOF complete repair) were included in the final analysis (Table 1). Basic demographic data, including age, gender and body mass index was similar between groups. There was no difference in mean age at corrective surgery. PS patients presented to the Adult Congenital Heart Disease Center an average 37.5±10.5 years after surgery compared to 32.3±10.6 years in TOF (p=0.023). Not surprisingly, due to anatomic limitations, patients with repaired TOF were significantly more likely to have required palliative shunting and eventual placement of a pacemaker or cardiac defibrillator. Cardiovascular and non-cardiovascular comorbidities were otherwise similar. Upon presentation, the PS group reported a higher prevalence of chest pain (29% vs. 12%, p=0.026) and shortness of breath at rest (32% vs. 15%, p=0.033). All other symptoms—including fatigue, cyanosis, orthopnea, palpitations, dyspnea on exertion, home oxygen supplementation and edema—were similar in prevalence among the two groups. Symptom severity at presentation was also similar as determined by NYHA function class. Complete operative records were available for 23% of the cohort. Of the TOF subset, 33% of the repairs included known use of a transannular patch.

Table 1.

Demographic Data at Presentation

Variable	Pulmonic Stenosis (n = 34)	Tetralogy of Fallot (n = 75)	P-value
Age (years)	47.0 ± 13.5	43.9 ± 13.3	0.260
Male	17 (50%)	43 (58%)	0.536
Body Mass Index (kg/m2)	28.6 ± 7.1	27.4 ± 6.8	0.388
Age at first repair (years)	6.8±8.5	8.5±10.3	0.411
Time from repair to presentation (years)	37.5 ± 10.5	32.3 ± 10.6	0.023
Shunt	1 (3%)	30 (40%)	<0.001
Pacer/Implantable cardiac defibrillator	1 (3%)	17 (23%)	0.011
New York Heart Association functional class	1.7 ± 0.7	1.6 ± 0.6	0.506
Hypertension	11 (32%)	19 (25%)	0.491
Pulmonary Hypertension	1 (3%)	11 (15%)	0.100
Diabetes mellitus	2 (6%)	4 (5%)	1.000
Hyperlipidemia	3 (9%)	12 (16%)	0.383
Smoker	8 (24%)	18 (24%)	1.000

Data given as mean ± SD

Non-paced electrocardiographic data was obtained from 95 patients (31 valvotomy and 64 complete repair). PR and corrected QT intervals were similar in both groups (Table 2). QRS duration, however, was significantly greater in the TOF group.

Table 2.

EKG and Echocardiogram Data

Variable	Pulmonic Stenosis	Tetralogy of Fallot	P-value
PR interval (ms)	177 ± 32	177 ± 55	0.946
QRS width (ms)	114 ± 27	150 ± 28	<0.001
QTc (ms)	421 ± 51	420 ± 53	0.888
Right atrium volume (mL)	25.1 ± 10.6	27.0 ± 10.6	0.560
Right ventricle volume†	1.5 ± 1.0	1.5 ± 1.0	0.761
Right ventricle function*†	1 (0,2)	1 (0,2)	0.512
Left ventricle diastolic width (cm)	4.4 ± 0.5	4.5 ± 0.8	0.457
Left ventricle systolic width (cm)	2.9 ± 0.5	3.1 ± 0.6	0.289
Left ventricle function*†	0 (0,0)	0 (0,0)	0.163
Aortic regurgitation*‡	0 (0,0)	0.5 (0,2)	0.021
Pulmonic regurgitation‡	3.2 ± 0.9	2.4 ± 1.6	0.004
Mitral regurgitation‡	0.9 ± 0.6	0.9 ± 0.9	0.869
Tricuspid regurgitation‡	1.5 ± 1.0	1.5 ± 0.9	0.853
Pulmonic gradient (mmHg)	8.2 ± 3.1	13.9 ± 8.2	0.007
Right ventricular systolic pressure (mmHg)	35 ± 13	46 ± 20	0.007
Left ventricle ejection fraction (%)	56 ± 5	54 ± 8	0.166

Data given as mean ± SD;

^*Data given as median (IQ range);

^† Volume/function graded from 0 (normal) to 3 (severely dilated/decreased);

^‡Valvular regurgitation graded as 0 (none), 1 (mild) to 4 (severe)

Echocardiographically assessed right ventricular size (mildly to moderately enlarged) and function (mildly decreased) was similar in the 2 groups (Table 2). Pulmonic regurgitation, assessed qualitatively, was slightly more severe in the post-valvotomy group, while aortic regurgitation was worse in TOF. Right ventricular systolic pressure and pulmonic outflow gradient were both slightly higher in the TOF cohort.

Complete cardiac catheterization data were available in 56 patients (18 PS and 38 TOF). This demonstrated similar hemodynamics including right atrial pressure, pulmonary artery pressure and cardiac indexes in both groups (Table 3).

Table 3.

Invasive Hemodynamic Data

Variable	Pulmonic Stenosis (n = 18)	Tetralogy of Fallot (n = 38)	P-value
Right atrial mean (mmHg)	10.1 ± 5.4	11.4 ± 6.5	0.487
Systolic pulmonary artery (mmHg)	35.4 ± 11.6	42.2 ± 22.2	0.159
Diastolic pulmonary artery (mmHg)	12.6 ± 5.1	13.8 ± 10.5	0.584
Mean pulmonary artery (mmHg)	21.0 ± 5.2	23.9 ± 14.2	0.348
Systolic right ventricle (mmHg)	41.2 ± 12.9	50.3 ± 20.5	0.102
Diastolic right ventricle* (mmHg)	10 (5, 14)	10 (6, 13.5)	0.767
Mean pulmonary capillary wedge pressure (mmHg)	12.5 ± 4.9	13.0 ± 5.4	0.737
Fick Cardiac Index (L/min/m2)	2.4 ± 0.5	2.4 ± 0.9	0.963

Data given as mean ± SD;

^*Data given as median (IQ range)

63 patients (24 PS and 39 TOF) had complete and technically adequate cardiac MRI data (mean 2.2±2.7 months from the initial evaluation; Table 4). To ensure similarity of this selected cohort to the entire study population, baseline demographic data was re-compared, with no significant differences seen. As with the entire cohort, TOF patients in this subgroup were substantially more likely to have a history prior of palliative shunt placement.

Table 4.

Magnetic Resonance Imaging Subgroup Analysis

Variable	Pulmonic Stenosis (n = 24)	Tetralogy of Fallot (n = 39)	P-value
Age (years)	45.9 ± 13.3	46.2 ± 11.7	0.304
Male	11 (46%)	23 (59%)	0.435
Body mass index (kg/m2)	27.5 ± 6.3	27.1 ± 6.3	0.783
Time from Surgery to MRI (years)	36 ± 11	32 ± 9	0.152
Shunt	4%	46%	<0.001
New York Heart Association class	1.6 ± 0.6	1.5 ± 0.6	0.786
Body Surface Area (m2)	1.96 ± 0.28	1.91 ± 0.24	0.489
Right ventricular end diastolic volume (cc)	267.1 ± 77.7	256.3 ± 112.0	0.691
Right ventricular end diastolic volume indexed (cc/m2)	147.9 ± 58.8	139.2 ± 61.7	0.592
Right ventricular end systolic volume (cc)	134.5 ± 52.7	151.8 ± 97.0	0.394
Right ventricular end systolic volume indexed (cc/m2)	73.3 ± 31.2	82.5 ± 49.7	0.389
Right Ventricular Ejection Fraction (%)	49.1 ± 7.7	40.8 ± 11.0	0.001
Left ventricular end diastolic volume (cc)	150.5 ± 34.4	158.1 ± 48.1	0.584
Left ventricular end diastolic volume index (cc/m2)	79.3 ± 14.8	83.3 ± 24.6	0.466
Left ventricular end systolic volume (cc)	61.8 ± 15.7	78.4 ± 36.3	0.028
Left ventricular end systolic volume index (cc/m2)	31.7 ± 7.9	41.3 ± 19.0	0.015
Left ventricular Ejection Fraction (%)	59.3 ± 7.1	52.1 ± 9.6	0.002
Pulmonic Regurgitation (%)	32.3 ± 18.9	37.3 ± 21.6	0.394
Aortic Regurgitation (%)	2.8 ± 6.4	4.3 ± 4.5	0.341
Root Diameter (cm)	3.1 ± 0.3	4.0 ± 0.6	<0.001
Ascending Diameter (cm)	3.0 ± 0.6	3.6 ± 0.8	0.054

Data given as mean ± SD; Values indexed by body surface area

Both groups experienced similar degrees of RV dilatation, including indexed right ventricular end diastolic and end systolic volumes. RV ejection fraction was significantly higher post-valvotomy. While left ventricular (LV) volumes fell within normal range for PS patients, they were mildly increased in the TOF group and LV ejection fraction was lower.

Flow measurements demonstrated similar degrees of pulmonic and aortic regurgitation in both groups. There was no significant difference in pulmonic valve regurgitant fraction, which was elevated in both groups. Though aortic regurgitant fractions were similarly small in both groups, aortas were larger in patients with TOF, including larger root and mid-ascending aorta diameters.

Discussion

In a similarly symptomatic cohort of patients with PS following valvotomy and TOF after complete repair, we found that structural and functional sequelae were significantly more evident with TOF. Though valvotomy patients appeared to have a greater degree of PR by echocardiography, MRI quantification was similar. This likely reflects the greater accuracy of cardiac MRI in PR quantification. Alternatively, the exclusion of patients with pacemaker and defibrillators (still considered a contraindication to MRI scanning) may have altered the TOF cohort. Comparison of the full cohort and those undergoing MRI scanning, however, did not suggest any major differences. Furthermore, one would suspect that patients having defibrillators previously placed would have wider QRS durations, more dilated right ventricles and thus greater PR. Excluding these patients would therefore be expected to reduce the degree of PR seen in the TOF cohort and widen the observed difference, contrary to what was seen.

Right ventricular volumes were similarly increased in both groups, although the right ventricular ejection fraction was considerably lower in TOF patients. This measure is more sensitive to changes, and it reflects the effect of slightly smaller end-diastolic volumes and larger end-systolic volumes in TOF. Left ventricular end-systolic volume was larger and ejection fraction lower in patients with TOF. Left ventricular dysfunction has recently been characterized in the TOF population.²⁴ It appears from our data that PS patients are spared this effect and suggests a pathologic effect in TOF beyond longstanding right sided volume overload from PR. It is important to note that the decrease in left ventricular function was seen in a cohort of TOF patients who did not have implantable devices, thus excluding the potentially detrimental effect of pacing, which has previously been reported in other cohorts of patients. Also important to note is the fact that invasive hemodynamics suggest no difference in pulmonary capillary wedge pressure, suggesting a smaller role of left ventricular involvement on the presence of symptoms in these patient populations.

Aortic sizes tended to be larger in patients with TOF. This is not surprising as TOF is a conotruncal abnormality with aortic override being a major component. Aortic dissection has been reported in patients with TOF; fortunately it does not appear that the PS population is prone to this problem.

Sudden death is a well-described phenomenon in patients with congenital heart disease, particularly in TOF, and risk in the latter appears to be mitigated by right ventricular size and function.²⁴ QRS complex width has been proposed as an effective marker for predicting future risk in TOF and a cut-off of 180 msec has been proposed for defining patients at particularly high risk.^10,25 Many clinicians regularly track this value to help decide when to intervene operatively for PR. Interestingly QRS width was considerably less in PS patients and only mildly abnormal in these patients despite similarly, significantly enlarged right ventricles. Complete repair in TOF is known to produce right bundle branch block, and our data suggests that physicians should ideally not apply the cut-offs derived from TOF to the PS population. Further studies should investigate sudden death risk in the post-valvotomy population and whether a lower cutpoint or change in QRS width over time is a helpful screening tool.

Current guidelines and indicators for PVR in TOF patients include criteria for RV size, QRS width and symptomatic impairment.^10,18 This study suggests that post-valvotomy patients with PR become symptomatic despite fewer structural changes and functional deficits than do TOF patients. Therefore, the guidelines developed for PVR in the TOF population do not appear to apply to the post-valvotomy population.

This study is limited by the very selective nature of the patient cohort. All patients were referred to an adult congenital heart disease center based on the presence of PR and/or symptoms of heart failure. As such, we are capturing patients who have developed complications or are otherwise unable to be managed by their general cardiologist. Therefore, one cannot be certain that these findings apply to all patients with PS and prior valvotomy or tetralogy with prior complete repair. Nonetheless, it is a comparison that has not been previously made, and its findings have the potential to change clinical practice if confirmed in other populations. Because of the retrospective nature of data collection, the details of prior surgeries were not always available. In such cases, the patients were the sole source of clinical information, a common situation in adults with congenital heart disease. We also did not attempt to examine the impact of valvular interventions in this patient population. Future studies should aim to determine the effect of pre-PVR structure and function on the functional, structural and symptomatic improvement following valve replacement in the post-valvotomy population. This would further facilitate the development of population-specific guidelines.