Characterization of Right Ventricular Diastolic Performance After Complete Repair of Tetralogy of Fallot
Restrictive Physiology Predicts Slow Postoperative Recovery
Background Prolonged postoperative recovery caused by a low cardiac output state occurs in some patients after complete repair of tetralogy of Fallot. Biventricular systolic function is usually well preserved in these patients. The contribution of impaired diastolic function, particularly of the right ventricle, has not been studied in detail; therefore, we performed a prospective study of right ventricular diastolic function in this patient group.
Methods and Results We studied biventricular systolic and diastolic function using Doppler echocardiographic examination. Tricuspid valve, superior vena caval, pulmonary arterial, and mitral valve Doppler spectrals were obtained during the first postoperative day in 35 patients aged 6 months to 45 years who underwent complete repair of tetralogy of Fallot. Biventricular systolic function was grossly normal in all patients. Isolated restrictive right ventricular physiology characterized by pulmonary arterial antegrade flow coincident with atrial systole and associated with prominent retrograde superior vena caval flow was seen in 17 of the 35 patients (group 1). This flow was augmented during the expiratory phase of positive pressure ventilation and abolished or greatly diminished during the inspiratory phase (P<.001). An increase in the duration of pulmonary regurgitation occurred during the inspiratory phase of positive pressure ventilation in these patients (P<.01). All patients with right ventricular restriction had a clinical picture compatible with a low cardiac output state, requiring prolonged stays in intensive care and the hospital. Clinical improvement was mirrored by resolution of the Doppler markers of right ventricular restriction in most of the patients.
Conclusions Isolated right ventricular restriction is characterized by antegrade diastolic pulmonary arterial flow on Doppler echocardiography and is responsible for the slower postoperative course and clinical evidence of low cardiac output state in some patients after complete repair of tetralogy of Fallot.
The postoperative course is uncomplicated in most patients after repair of tetralogy of Fallot. However, in some patients the postoperative period is associated with episodes of low cardiac output, raised central filling pressures, prolonged ventilation and inotropic support, and the development of pleural effusions and/or ascites. Biventricular systolic function is usually well preserved in these patients, but there are few data concerning diastolic performance.
We1 and others2 have previously demonstrated abnormalities of right ventricular diastolic function in a small number of patients late after repair of severe right ventricular outflow tract obstruction. Pulsed Doppler echocardiography demonstrated forward diastolic flow in the pulmonary artery coincident with premature opening of the pulmonary valve during atrial systole.1 2 Simultaneous catheter pressure monitoring demonstrated that this flow occurred when right ventricular end-diastolic pressure equaled or exceeded pulmonary arterial diastolic pressure.2 We proposed that this antegrade diastolic pulmonary arterial flow reflected reduced right ventricular diastolic compliance, suggesting that the right ventricle is unfillable and truly “restrictive” at end diastole, so acting as a passive conduit between the right atrium and pulmonary artery during atrial systole.
This phenomenon is sometimes seen in adults with restrictive cardiomyopathy but is an inconsistent feature,3 presumably because of biventricular involvement and the presence of pulmonary arterial diastolic hypertension. The increased deceleration rate of early rapid filling has also been proposed as an index of both right and left ventricular restrictive physiology,4 but this may be difficult to measure in children with rapid heart rates and poorly defined early rapid filling and atrial systolic phases. Furthermore, those abnormalities demonstrated during spontaneous respiration may not be directly applicable to the postoperative patient receiving positive pressure ventilation in the immediate postoperative period.
In the present study we used the presence of antegrade diastolic pulmonary arterial flow as our index for the presence of severe isolated restrictive right ventricular physiology and related this to other abnormalities of right ventricular diastolic function to assess the relation of diastolic disease to clinical outcome after complete repair of tetralogy of Fallot.
The Royal Brompton Hospital is a supraregional center for neonates, children, and adults with congenital heart defects, with approximately 500 patients undergoing surgery each year. We performed a prospective study of 35 consecutive patients undergoing repair of tetralogy of Fallot (n=28) or tetralogy of Fallot with acquired pulmonary atresia (n=7) at this hospital between March 1992 and June 1993. Table 1⇓ shows anthropometric details, including previous surgery, preoperative hemodynamics, and operative techniques. The foramen ovale was routinely closed by direct suture whenever present. Twenty-seven patients were younger than 3 years (15<1 year; 9 aged 1 to 2 years; 3 aged 2 to 3 years). Four patients were 3 to 10 years old. Four adults (aged 30 to 41 years) underwent operation.
Doppler echocardiographic studies were carried out during the first 24 hours after the operation. All patients were receiving intermittent positive pressure ventilation when the study was performed along with inotropic support with dopamine in the range of 5 to 10 μg/kg per minute and sedation by infusions of morphine (10 to 40 mcg/kg per minute) and midazolam (0.1 to 0.4 mg/kg per hour). No patient had significant right ventricular outflow tract obstruction (Doppler gradient >20 mm Hg) or a residual intracardiac shunt. Fluid administration on the first postoperative day was 0.5 L/m2 as per our standard protocol. Hemodynamic data at the time of study are presented in Table 2⇓.
All patients were studied with either transthoracic or, when there was an inadequate window, transesophageal imaging. Eight patients were studied with both modalities. Transthoracic measurements were made with a 5- or 7.5-MHz transducer, and transesophageal measurements were made with a 5-MHz single- or dual-plane probe interfaced with a Hewlett-Packard Sonus 1500 ultrasound system. All adult patients were studied by transesophageal echocardiography with a 5-MHz transducer (single plane). After routine diagnostic imaging, an M-mode of the left ventricular cavity and the following pulsed Doppler parameters were measured in each patient: (1) superior vena caval Doppler profile (1 to 2 cm proximal to the right atrium); (2) transtricuspid valve characteristics (at the level of the tips of the valve leaflets): E and A wave velocity/integral and E wave deceleration time; (3) transmitral valve characteristics (at the level of the valve leaflets): E and A wave velocity/integral and presence of base-to-apex intraventricular flow during isovolumic relaxation; and (4) pulmonary arterial systolic and diastolic Doppler characteristics. The pulsed Doppler sample volume was placed at the midpoint between the pulmonary valve leaflets and bifurcation.
Measurements were made with simultaneous electrocardiographic, phonocardiogram, and respiratory motion recordings. Spectral recordings were made with minimal filtering on paper at a speed of 100 cm/s2. Doppler recordings were analyzed according to the phase of the respiratory cycle, as previously described.1 Briefly, three consecutive inspiratory and three consecutive expiratory cardiac cycles (defined respectively as the first cardiac cycle after the onset of the inspiratory and expiratory deflection on the respirometer) were analyzed by planimetry and the results for each of the indexes averaged.
Qualitative assessments of superior vena caval Doppler spectrals were made, with particular emphasis on the presence or absence of retrograde diastolic flow. Peak velocity and velocity integral of systolic, diastolic, and retrograde superior vena caval flows coincident with atrial systole were measured. A detailed assessment of right ventricular inflow was performed using transtricuspid valve E and A wave peak velocities, duration, and E wave deceleration times. In some patients (n=17), summation flow was present so that E and A wave filling could not be separated for the purposes of analysis. Systolic and diastolic pulmonary artery antegrade Doppler spectrals (peak velocity, duration, integral) were analyzed, as was the duration of pulmonary regurgitation.
Right Ventricular Area Measurements
Right ventricular area measurements were made in the later part of this study in eight patients in whom excellent border definition was obtained, using the method of acoustic quantification.5 6 Right ventricular area was measured from the four-chamber projection, the region of interest being taken at the level of the tricuspid valve ring, and included the entire cavity outline. The percentage of area change was measured during both ventricular and atrial systole.
Group data are expressed as mean±SD. Student’s t tests were used to compare normally distributed variables; otherwise, a Mann-Whitney U test was performed. Statistical significance was inferred at a value of P<.05.
Right ventricular restriction as defined by antegrade diastolic pulmonary blood flow coincident with atrial systole (pulmonary arterial diastolic flow; Table 3⇓) was present in 17 of the 35 patients (group 1, Fig 1⇓). This accounted for 10% to 30% of the total forward pulmonary artery Doppler integral in these patients. The duration of pulmonary regurgitation was significantly shortened in those patients with antegrade pulmonary arterial diastolic flow (214.6±60 versus 153±52 milliseconds, P<.01). There was a marked variation in the Doppler signal and other indexes of diastolic performance with respiratory cycle phase (see below). The remaining 18 patients (group 2) had no evidence of diastolic dysfunction.
There was no significant difference in heart rate at the time of study between the two groups (137±18 versus 128±30 beats per minute, P=.4). All patients were in sinus rhythm at the time of study. Complete right bundle branch block was present on the resting electrocardiograms in all patients.
Right Ventricular Function
Right ventricular systolic function was qualitatively normal in the patient group, with almost complete obliteration of the right ventricular cavity at end systole in most of the patients. Right ventricular systolic area change was measured in eight patients and was greater than 60% in all. None of the patients had more than trivial tricuspid regurgitation detectable by color flow Doppler.
Transtricuspid E and A wave velocities, integrals, and their ratios could be reliably measured in 18 of 35 of the study group (Table 4⇓). There was transtricuspid summation flow in the remainder, making similar analysis impossible. The peak E wave velocity was markedly lower in the restrictive group, during both inspiration and expiration (P<.001). Despite apparent right ventricular “filling” during atrial systole with demonstrable transtricuspid flow in all, right ventricular area change with atrial systole was 5±4% in the restrictive group (group 1) compared with 30±5% (P<.0006) in group 2 when assessed by acoustic quantification. There was no significant difference in systolic right ventricular area change between the two groups (P=.25). In those 8 patients in whom both transthoracic and transesophageal recordings were made, there was no significant difference in the results obtained.
Effect of Positive Pressure Ventilation
In patients with right ventricular restriction (group 1), the inspiratory cycle of positive pressure ventilation led to a decrease in the transtricuspid E wave integral (Fig 2⇓, Table 4⇑) (P<.05) and diminution or total abolition of the antegrade pulmonary artery diastolic spectral signal (Fig 3⇓, Table 3⇑). A concomitant increase in the duration of pulmonary regurgitation occurred during positive pressure inspiration (P<.01). Antegrade diastolic pulmonary arterial integral was increased during the expiratory phase (P<.001). There was little change in pulmonary arterial systolic antegrade integral during the respiratory cycle (11.4 versus 12.2 cm, P<.79) and so the total forward flow integral was also significantly increased (P<.05) in the expiratory beats. In patients with restrictive right ventricular physiology, the antegrade diastolic pulmonary arterial integral accounted for 7±8% and 22±10% of the total forward pulmonary arterial integral during inspiration and expiration (P<.004), respectively.
Superior Vena Caval Flow
In the nonrestrictive patients (group 2), there was antegrade flow during ventricular systole in all but one patient in whom systolic and diastolic flow velocities were equal. Retrograde flow coincident with atrial systole was present in only one patient in this group (6%).
All patients with restrictive right ventricular physiology (group 1) had retrograde superior vena caval flow coincident with atrial systole (mean peak velocity, 23 cm/s). In addition, predominant antegrade flow in the superior vena cava occurred during early diastole in 15 of the 17 patients (Fig 4⇓). Central venous filling pressures were consistently higher in this group than in the nonrestrictive group (14±2 versus 7.1±1 mm Hg, P<.01).
In all but 4 of the 17 patients with restrictive right ventricles, Doppler markers of abnormal diastolic function were completely resolved by the ninth postoperative day, ie, loss of antegrade diastolic pulmonary arterial flow and return of predominant systolic antegrade flow in the superior vena cava.
Left Ventricular Function
Despite reversed septal motion in all patients, the left ventricular shortening fraction was normal, ranging from 33% to 50% (median, 44%). There was no evidence of reduced left ventricular diastolic abnormality as deduced from normal transmitral E and A wave velocities, integrals, and ratios in either group nor were there any significant differences between the groups (Table 5⇓). However, careful analysis revealed intraventricular base-to-apex flow in the left ventricular inflow Doppler spectrals during the isovolumic relaxation period, suggesting the presence of incoordinate left ventricular relaxation in 14 of 35 patients.7 8 There was no important mitral valve disease or regurgitation in any of the patients.
Patients with restrictive right ventricular physiology had a clinical syndrome of low cardiac output (Table 2⇑) and remained in the intensive care unit (median stay, 7 [range 6 to 45] versus 2 [1 to 41] days; P<.01) and in the hospital (median stay, 12 versus 7 days; P<.05) for a significantly longer period than those who did not demonstrate this pattern. Two patients, one in each group, remained in the intensive care unit for a prolonged period of time because of infective complications. Excluding these two patients from the analysis produced no change in statistical significance for either intensive care or hospital stay between the two groups. In addition, there was a higher incidence of pleural effusions and/or ascites (P<.05) in patients with restrictive right ventricular physiology.
There was no discernible relation between the development of restrictive right ventricular physiology and age, preoperative oxygen saturation, preoperative mean right atrial pressure or right ventricular end-diastolic pressure, length of procedure, or the use of a transannular patch or monocusp homograft (Table 1⇑).
Important abnormalities of right ventricular diastolic function occur in a significant proportion of patients after repair of tetralogy of Fallot, and our results explain the prolonged and sometimes difficult postoperative course seen in these patients. The fundamental abnormality is that of restrictive right ventricular physiology characterized by what we consider to be the hallmark of isolated severe right ventricular restriction: antegrade diastolic pulmonary blood flow coincident with atrial systole. In essence, the right ventricle becomes unfillable during atrial systole and acts as a passive conduit. Thus, some or all of the transtricuspid atrial systolic flow, demonstrable by Doppler, results in antegrade pulmonary blood flow, not right ventricular filling. The simultaneous retrograde flow in the superior vena cava and rapid deceleration of the early rapid filling velocity seen in many patients are consistent with a restrictive right ventricle. It is important to recognize that under these circumstances, traditional indexes of restrictive physiology may not be measurable (because of summation flow related to tachycardia) or valid such as atrial “filling” fraction because, to reiterate, not all transtricuspid flow equates to filling.
Similar abnormalities have been described previously in adults with restrictive myocardial processes involving the right ventricle.3 However, these findings have not been consistently documented, presumably because of biventricular restriction and hence elevated left atrial and pulmonary arterial diastolic pressures. Kisanuki et al2 studied three patients after surgery for relief of right ventricular outflow tract obstruction. They showed that antegrade pulmonary diastolic flow on Doppler echocardiography occurred when right ventricular pressure exceeded pulmonary artery pressure using simultaneous catheter measurements, reflecting reduced right ventricular compliance. In the description by Appleton et al3 of the Doppler characteristics of restrictive physiology, 8 of 14 of their patients had a pulmonary arterial diastolic pressure of 20 mm Hg or higher. Under these circumstances, antegrade diastolic flow with atrial systole is made less likely, and very high atrial systolic pressures would be required to generate it. Indeed, only half of their patients demonstrated this phenomenon, almost certainly because of biventricular restriction. Our patients had isolated right ventricular restriction, and the abnormalities of diastolic function are similar to our findings in patients studied late after repair of pulmonary atresia with intact septum.1 Those patients and the patients in our current study demonstrated a marked interdependence of right ventricular hemodynamics and respiration. Previously described indexes in patients with biventricular restrictive physiology may not be either applicable or valid in this subgroup of patients.4 Increased systemic venous return occurs during normal inspiration as well as during the expiratory phase of positive pressure ventilation.9 10 This additional right ventricular volume load shortens E wave deceleration time both in healthy individuals and in our previously reported patients after repair of pulmonary atresia, who also had increased antegrade diastolic pulmonary artery flow on inspiration.1 In the patients in the present study, positive pressure ventilation had the expected opposite effects. Not only did transtricuspid flow fall but the rise in mean airway pressure led to a reduction in antegrade diastolic flow; as a result, the duration of pulmonary regurgitation was prolonged. Thus, there is indirect evidence that cardiac output falls by several mechanisms during the inspiratory phase of positive pressure ventilation.
The clinical implications of our data are clear. Patients with restrictive right ventricles have clinical features of a low cardiac output, require higher filling pressures, and although biventricular systolic performance is well preserved, they often require prolonged inotropic and volume support. This leads to a significantly greater incidence of effusions and a longer stay in intensive care.
Although reflecting adverse hemodynamics, the antegrade diastolic pulmonary arterial flow seen in these patients is important as it contributes to forward flow and shortens the duration of pulmonary regurgitation, making an important contribution to cardiac output. Conversely, anything that diminishes this effect will be detrimental. Thus, maintenance of sinus rhythm is important in patients with restrictive physiology, and the possible beneficial effects of changes in atrioventricular delay by atrioventricular pacing warrant further study. Furthermore, transient loss of these beneficial effects is seen during the inspiratory phase of positive pressure ventilation. We therefore recommend that intermittent positive pressure ventilation with a short inspiratory time and lowest possible mean airway pressure be used in these patients. An alternative strategy that we are currently investigating is the use of negative pressure ventilation, which we have shown to increase pulmonary artery flow in the atrial dependent circulation of patients after the Fontan operation.11
All of these possible therapeutic options are directed at modulating the preexisting abnormal diastolic function and do not directly address the underlying pathophysiology. The cause of the phenomenon is unclear. One might speculate several possible causes for a limited right ventricular end-diastolic volume with reduced late diastolic compliance. Endomyocardial fibrosis has been demonstrated in older patients with tetralogy of Fallot,12 and right ventricular hypertrophy is always present. The superimposition of the effects of cardiopulmonary bypass, ventriculotomy, and myocardial edema and the placement of nonfunctional patches on the ventricular septum and in the right ventricular outflow tract all might be expected to influence the diastolic performance of the right ventricle. The exact relation among these factors remains unknown, but the transient nature of our findings (resolving within 2 weeks in 13 of 17 patients) suggests that, in the short term at least, resolution or adaptation occurs in most patients. Doppler indexes of left ventricular function were normal even in those patients with restrictive right ventricular physiology. Why left ventricular diastolic function should not also be affected in our patient group is less clear. Ventriculotomy is not performed on the left ventricle, nor is noncontractile patch material incorporated into its free wall. It may be that superior left ventricular myocardial protection also plays a part as the right ventricle may be exposed to room temperature and radiant heat from overhead operating lights, despite general hypothermia and topical cooling.13 Whether the pattern we describe therefore reflects a right ventricular equivalent of the stunned myocardium syndrome described for the left ventricle remains to be resolved.14
Our patient population is unusual in including adults undergoing initial repair at 40 years of age or older but represents the referral pattern to our hospital, which has both pediatric and adult congenital heart units. Nevertheless, it was interesting to note that age at repair was not a significant factor in determining whether restrictive right ventricular physiology developed, occurring as it did across the age spectrum from infancy to adulthood. In addition, our surgeons do not perform neonatal or early total correction (<6 months), preferring when necessary to enhance pulmonary blood flow with a modified Blalock-Taussig shunt in the first instance and performing complete repair in the second year of life. Although not specifically addressed, we would suggest that the improved results of neonatal repair of the patients of DiDonato et al15 in whom a small atrial septal defect or patent foramen ovale is left (albeit at the expense of cyanosis caused by a right-to-left shunt at the atrial level) is due to decompression, and hence improved cardiac output, of the restrictive right ventricle in some of those patients.
Although it would have been useful to have information regarding sequential right ventricular size in our patients, we elected not to attempt right ventricular dimension analysis in view of the methodological problems in interpreting these data, particularly with the transesophageal approach.16 On the other hand, Doppler echocardiographic studies are easily reproduced and reflect changes in overall right ventricular hemodynamics irrespective of absolute right ventricular size. Furthermore, we did not estimate the cross-sectional area of the pulmonary artery at the site of pulsed Doppler interrogation during the cardiac cycle. Although it is possible that change in overall cross-sectional area may contribute to differences in antegrade pulmonary arterial flow velocity integrals during diastole and systole, we would suggest that any minor change would not significantly affect our interpretation of the physiology and would be more than offset by error of its measurement.
We have emphasized the difficulties encountered if one uses traditional indexes of ventricular diastolic function in the assessment of right ventricular performance in these patients. To some extent, the same is true of the left ventricle in these postoperative patients.4 Left ventricular diastolic performance could only be assessed in 60% of our patients. In the remainder, summation rendered analysis of mitral valve inflow velocity impossible. No patient displayed restrictive left ventricular physiology based on analysis of mitral E and A wave velocities. Indeed, there was no discernible difference in these indexes between the two groups of patients (restrictive versus nonrestrictive). Therefore, there does not appear to be a uniform or predictable response in the left ventricle that would support our contention that the poor clinical progress of these patients is largely due to isolated right ventricular diastolic dysfunction. If important biventricular restriction was present, it is highly likely that no antegrade pulmonary arterial diastolic flow would occur because of raised left atrial and pulmonary arterial pressures. This is commonly the case when such conditions exist.3
Finally, we made no attempt to measure absolute cardiac output using any other method. For the same reasons as stated above, we feel that relative changes in Doppler integrals are a valid reflection of beat-by-beat changes in relative cardiac output under the circumstances of our study. Indeed, invasive techniques such as thermodilution are invalid in the presence of important pulmonary incompetence. Furthermore, the patients with restrictive right ventricular physiology had a typical clinical picture of low cardiac output (tachycardia, hypotension, increased core-peripheral temperature gap, oliguria, metabolic acidosis) and were strikingly different from those in the nonrestrictive group.
In summary, right ventricular diastolic function characteristic of restrictive physiology occurs in a significant proportion of patients after complete repair of tetralogy of Fallot and is responsible for the slower postoperative course and longer duration of ventilatory requirements in these patients.
We would like to thank Drs M. Connelly, E.A. Shinebourne, M.L. Rigby, J. Somerville, C. Lincoln, and Professor Sir M. Yacoub for allowing us to study their patients.
Presented in part at the 65th Scientific Sessions of the American Heart Association, New Orleans, La, November 1992.
- Received June 10, 1994.
- Revision received October 4, 1994.
- Accepted October 31, 1994.
- Copyright © 1995 by American Heart Association
Redington AN, Penny D, Rigby ML, Hayes A. Antegrade diastolic pulmonary arterial flow as a marker of right ventricular restriction after complete repair of pulmonary atresia with intact ventricular septum and critical pulmonary valve stenosis. Cardiol Young. 1992;2:382-386.
Rasmussen C, Dyar D, Wheeler K, Donaghey L, Dittrich HC. Comparison between automatic border detection and planimetry of right ventricular size and function following surgical treatment of thromboembolic pulmonary hypertension. J Am Coll Cardiol. 1994;12(suppl 908):239. Abstract.
Penny DJ, Rigby ML, Redington AN. Abnormal patterns of intraventricular flow and diastolic filling after the Fontan operation: evidence for incoordinate ventricular wall motion. Br Heart J. 1991;66:375-378.
Penny DJ, Hayek Z, Rawle P, Rigby ML, Redington AN. Ventilation with external high frequency oscillation around a negative baseline increases pulmonary blood flow after the Fontan operation. Cardiol Young. 1992;2:227-280.
Hegarty A, Anderson RH, Deanfield JE. Myocardial fibrosis in tetralogy of Fallot: effect of surgery or part of the natural history? Br Heart J. 1988;59:123. Abstract.
Braunwald E, Kelner RA. The stunned myocardium: prolonged postischemic ventricular dysfunction. Circulation. 1982;66:1146-1149.
Foale R, Nihoyannopoulos P, McKenna W, Kleinebenne A, Nadazim A, Rowland E, Smith G. Echocardiographic measurement of the normal adult right ventricle. Br Heart J. 1986;56:33-44.