Ann Thorac Surg 1998;66:634-640
© 1998 The Society of Thoracic Surgeons
Supplement
Which two ventricles cannot be used for a biventricular repair? Echocardiographic assessment
Norman H. Silverman, MDa,
Doff B. McElhinney, MDa
a Division of Pediatric Cardiology, Department of Pediatrics, University of California, San Francisco, California, USA
Address reprint requests to: Dr Silverman, University of California, San Francisco, Box 0214, M342A, San Francisco, CA 94143-0214
e-mail: (norman_silverman{at}pedcardgateway.ucsf.edu)
Presented at the Workshop on "One and One-Half Ventricle Repairs," Gubbio, Italy, Dec 6–7, 1996.
Abstract
Background. A variety of factors can influence the suitability of a congenitally malformed heart for biventricular repair, including size, morphology, function, and dimensions and function of the inflow and outflow, among others. Although certain features have been identified that may indicate a lower probability of successful biventricular repair, our ability to predict whether a particular patient will be able to tolerate completely separate in-series systemic and pulmonary circulations remains imperfect.
Methods and Results. In this review, we discuss the echocardiographic evaluation of various factors that can influence a patient’s suitability for two ventricle repair. We call on our own experience, and illustrate our discussion with a number of echocardiographic images.
Conclusions. In most cases, echocardiography allows for full assessment of the anatomic and functional features that influence whether a patient is a suitable candidate for biventricular repair. Although a number of indices have been developed for determining who can and cannot be expected to undergo successful two ventricle repair, there remains substantial room for progress in this area.
A variety of conditions can prevent a successful biventricular repair. This presentation will focus on ventricles that are small, but nevertheless potentially suitable for incorporation as part of a one and a half ventricle repair. In many cases, decisions about the possibility of such a procedure will be obvious from the ventricular morphology alone, whereas in others the decision will depend on elucidating the physiology by echocardiography or other means. Ventricles may be unsuited for biventricular repair because they are too small, as with the hypoplastic left and right ventricles typically found in patients with aortic or pulmonary atresia, or because of abnormalities of the atrioventricular valves, as with mitral atresia on the left side and tricuspid atresia or Ebstein’s malformation associated with pulmonary atresia on the right. Ventricles may have abnormal connections, such as double inlet to the left (most commonly) or right ventricle. Similarly, straddling of an atrioventricular valve may be so severe that the ventricle is deemed inadequate for supporting an entire cardiac output. This is particularly true when a ventricle has tendinous cords that straddle into a right ventricle, where repair by simple patching of the ventricular septal defect or translocation of the straddling cords to the left ventricle will require the underdeveloped left ventricle to function as the systemic pumping chamber. Ventricles may also be unsuited for incorporation into the circulation on account of poor function, as is the case with adequately sized left ventricles affected by either a restrictive process (such as endocardial fibroelastosis) or a dysfunctional dilated state (as with ischemia secondary to an anomalous coronary artery), or those found in right heart conditions such as Uhl’s anomaly and some cases of Ebstein’s malformation.
Recently, the Toronto group published their experience with one and a half ventricle repair in 38 patients with the following diverse conditions: Ebstein’s malformation, transposition of the great arteries, atrioventricular and ventriculoarterial discordance, pulmonary stenosis, tetralogy of Fallot, ventricular septal defects, atrioventricular septal defects, atrial septal defects, double-inlet left ventricle, double-outlet right ventricle with atrioventricular discordance, left atrial isomerism, and atrioventricular septal defect with tetralogy of Fallot [1]. Among their patients, there were four possible indications for one and a half ventricle repair: (1) a small right ventricle, (2) chronic right ventricular dysfunction, (3) facilitation of repair without hypoplasia or functional impairment of the pulmonary ventricle, and (4) acute right ventricular dysfunction. At our institution, where we performed one and a half ventricle repair in 18 patients between 1990 and 1996, a fifth indication has been partial biventricular repair in patients who already have the bidirectional Glenn anastomosis as part of a previous palliation. In addition to the above lesions, we have performed this procedure in patients with straddling tricuspid valve and double-outlet left ventricle.
Ventricular morphology
Size of the ventricle is obviously an important consideration in determining whether it is adequate for incorporation as a pumping chamber. The features of a tripartite ventricle can be determined echocardiographically (Figs 1, 2). In right ventricular hypoplasia, the most important area for echocardiographic analysis is the size of the outlet component. This is particularly important when this segment is hypoplastic and small (Fig 2). Determination of ventricular volume is an important type of assessment, and is the subject of a separate report presented in this symposium.
View larger version (123K):
[in this window]
[in a new window]
|
Fig 1. Apical four-chamber view in a patient with a hypoplastic right ventricle with intact ventricular septum and diminutive tricuspid valve. The left atrium (LA), left ventricle (LV), right atrium (RA), and right ventricle (RV) are displayed. The arrows indicate the annuli of the mitral and tricuspid valves. The scale marker (left-hand panel) indicates 1 cm, with the tricuspid annulus approximating 5 mm, yielding a Z value of the tricuspid valve of -2.5.
|
|
View larger version (128K):
[in this window]
[in a new window]
|
Fig 2. (Left) Image taken in an apical four-chamber view with caudal angulation. (Right) A view with cranial angulation of the pulmonary artery and its bifurcation into the right and left pulmonary artery (arrows). The chambers of the heart, as abbreviated in Figure 1, as well as the annulus of the very diminutive tricuspid valve are identified. The right ventricle is extremely hypertrophic, with a small inlet portion and a slit-like outlet portion proximal to the pulmonary artery identified on the right. The bifurcation of the pulmonary artery can be seen to the left of the ascending aorta (AO).
|
|
Ventricular physiology
Aside from the absolute size of the ventricle, ventricular morphology and atrioventricular valve function are important considerations. For example, with Ebstein’s malformation (particularly with severe cases) the degree of tricuspid regurgitation is an important factor in assessing the suitability of the entire right heart for incorporation into the circulation (Fig 3).
View larger version (79K):
[in this window]
[in a new window]
|
Fig 3. A series of images in a patient with Ebstein’s malformation and pulmonary atresia. (Top) Subcostal sagittal view demonstrating the inferior vena cava (IVC), a very large Eustachian valve (EV) in the right atrium (RA), and the mural leaflet (ML) of the tricuspid valve. The aorta (Ao) is seen above. The area between the mural leaflet of the tricuspid valve and the Eustachian valve shows the atrioventricular groove to which the mural leaflet is normally attached in this view. The displacement between the attachment to the ventricular wall on the diaphragmatic surface and the atrioventricular groove demonstrates the marked displacement of the mural leaflet. The area confined to the right ventricle in this example is almost exclusively atrialized right ventricle (ARV). (Middle) Subcostal coronal view orthogonal to the previous frame. The area between the atrioventricular groove and the attachment of the mural leaflet again demonstrates the marked displacement of the leaflet. The anterior leaflet is adherent to the right ventricular wall, and can be seen occupying the subpulmonary area. The arrows indicate the attached anterosuperior leaflet of the tricuspid valve. (Bottom) An apical four-chamber view in the same patient. Here the left atrium (LA) and left ventricle (LV) can be easily identified. The anterosuperior leaflet (AL) can be seen from the normal position of the atrioventricular valve groove and is adherent to the right ventricular wall (arrows). The septal leaflet is entirely adherent to the ventricular wall and can be seen separated from the endocardium only toward the apex of the ventricle. The Eustachian valve can also be identified. The position normally occupied by the right ventricle is atrialized right ventricle. (PA = pulmonary artery.)
|
|
Cross-sectional echocardiography can be used to measure chamber size and ejection fraction of the ventricle, providing good measures of systolic performance (Fig 4). Just as mitral regurgitation falsely elevates the assessment of function on the left side, ejection fraction in the face of severe tricuspid regurgitation may give a false assessment of right ventricular performance. Inadequate ventricular performance in the presence of valvar insufficiency can compromise outcome even if the atrioventricular regurgitation is minimized (Fig 5). Thus the presence of regurgitation that may not be reparable at operation may be a pointer toward consideration of auxiliary support for the right ventricle that can be achieved with a bidirectional Glenn anastomosis.
View larger version (132K):
[in this window]
[in a new window]
|
Fig 4. (Left) Apical four-chamber view in a patient with moderately severe Ebstein’s malformation who had previously undergone annuloplasty with a Carpentier ring elsewhere, but presented with recurrent/residual moderate-severe tricuspid regurgitation. (Right) Systolic Doppler color-flow image from the same view demonstrates marked tricuspid regurgitation, as judged by a very broad jet. (Bottom) After removal of the ring, reconstruction of the tricuspid valve, and bidirectional Glenn at our institution, the tricuspid regurgitation is reduced to two small jets, one seen at the coaptation point between the septal and anterior leaflets, and the other arising more posteriorly between the septal and mural leaflets.
|
|
View larger version (110K):
[in this window]
[in a new window]
|
Fig 5. This apical four-chamber view was taken from an infant with pulmonary atresia after right ventricular outflow reconstruction. The sizes of the right ventricular and tricuspid annulus are reasonable, but there is substantial tricuspid regurgitation. Bidirectional Glenn procedure and tricuspid annuloplasty reduced the degree of regurgitation and improved the quality of life for this patient.
|
|
In addition, the presence of sinusoids or ventriculocoronary connections may provide further determinants of whether the ventricle can be incorporated into the pulmonary circulation (Fig 6). It has been noted by Freedom that the degree of sinusoidal involvement may parallel the size of the ventricle, with greater sinusoidal development and ventricular dependent coronary circulation being present in smaller ventricles [2]. We have made a systematic study of these vessels because this observation has been made and can regularly demonstrate sinusoids by Doppler color-flow echocardiography. The flow pattern by pulsed Doppler interrogation in these vessels is typical; blood flows toward the aortic end of the coronary in systole, and toward the myocardium in diastole. Unfortunately the number of ventricular-dependent coronary arteries in our experience is small and we cannot make a statement as to the value of echocardiography in this entity.
View larger version (94K):
[in this window]
[in a new window]
|
Fig 6. (Top left) A patient with a hypoplastic right ventricle (RV) and intact ventricular septum, demonstrating the presence of ventriculocoronary connections and retrograde flow into the right coronary artery. The color-flow map also demonstrates disturbed flow as a result of the patent ductus arteriosus within the pulmonary artery (PA), the normal-sized left ventricle (LV), and the right atrium (RA). The ventriculocoronary connection can be seen coming from the diaphragmatic surface of the right ventricle and passing immediately and directly into the coronary artery (arrows). (Bottom left) This subcostal image of a different patient demonstrates a ventriculocoronary connection (S) between the cavity of the right ventricle and the surface of the heart. Note that the Nyquist limit or velocity scale on the left-hand side has been turned down to accentuate lower velocity flow. (Top right) Here the anterior ventriculocoronary connection (VCC) is seen draining from the right ventricular outflow and toward the left coronary artery. (Bottom right) A pulsed-wave Doppler spectrum in the right coronary artery in the patient seen in the top left panel. This demonstrates systolic flow away from the transducer (ie, toward the aortic end of the coronary artery) with prograde flow into the coronary artery only for a short period of diastole. This finding has been appreciated only when there is a direct connection between the coronary artery and ventricle. (AO = aorta.)
|
|
Atrioventricular valvar abnormalities
Complete valvar atresia or substantial hypoplasia is a clear indication that a ventricle cannot be incorporated into a biventricular repair. The question often arises whether a hypoplastic or stenotic valve is potentially compatible with partial ventricular flow. The report by Hanley and associates [3] showed a significant correlation between tricuspid annular size and the risk of not being a candidate for biventricular repair. In addition, they noted that the tricuspid Z score also correlated with right ventricular size and the presence of ventricular coronary connections, confirming observations by Freedom [2]. Right ventricles with tricuspid valves having a Z score less than 2.5 have approximately an 80% chance of requiring a shunt within 1 month after birth. Thus, tricuspid valve diameter and tricuspid annulus area may replace volume analysis as a measure of right ventricular size [4].
Tricuspid valve dysplasia associated with Ebstein’s malformation, particularly when there is outflow tract obstruction, also may interfere with the potential for a biventricular repair. Atrioventricular valves with marked regurgitation may not be reparable. We have noted that when there is a substantial degree of tricuspid regurgitation or dysplasia of the tricuspid valve (including abnormal tendinous cords), the ventricle often is not capable of supporting a full cardiac output.
Atrioventricular valve abnormalities and double-outlet ventricles
Other abnormalities of the atrioventricular valves, such as double inlet and straddle (Fig 7), can be important factors in the decision to perform one and a half ventricle repair, particularly when the systemic atrioventricular valve may not be able to support the circulation to one ventricle. In two such circumstances we have separated the circulations and supported that to the pulmonary ventricle by a cavopulmonary anastomosis.
View larger version (113K):
[in this window]
[in a new window]
|
Fig 7. (A) Apical four-chamber view of a straddling tricuspid valve. The arrows indicate the tendinous cords crossing the ventricular septum (S) through the septal defect from the right atrium (RA) into the left ventricle (LV). This patient underwent ventricular septation with transplantation of the cordal apparatus into the right ventricle (RV) and a bidirectional Glenn procedure. (B) Apical four-chamber view of another patient with straddling of the right atrioventricular valve across the ventricular septum and into the left ventricle. Doppler color flow demonstrates almost complete absence of inflow into the hypoplastic outlet chamber (v), with all right atrial and left atrial flow draining into the larger left ventricle. (Ao = descending aorta; LA = left atrium; PV = pulmonary vein.)
|
|
In the situation of double outlet of the arterial valves from a single ventricle, connecting the arterial valve to the systemic circulation may require incorporation of a substantial amount of the ventricle into the conduit, thereby diminishing the capacity and function of the pulmonary ventricle. This has occurred three times in our series, and has been performed as part of a double switch procedure. In atrioventricular septal defects where there is extreme unbalance and marked atrioventricular valve regurgitation, the septation may produce a ventricle incapable of supporting a complete cardiac output, and a cavopulmonary anastomosis may be required. Because the septation and degree of atrioventricular valve hypoplasia cannot be assessed adequately until the repair is complete, the decision may not be made until the cessation of cardiopulmonary bypass.
In some cases where function has been impaired during operation or where the function was known to be marginal before repair, we have performed the Glenn anastomosis.
In conclusion, the indications for the one and a half ventricle repair procedure are so varied that no rules can be provided at this time. It is much easier to define the ventricle that is hypoplastic with a diminutive cavity, vestigial outflow tract, and ventricular coronary connections with right ventricular-dependent coronary circulation (see Fig 6). It is also simpler to define extreme unbalance of an atrioventricular valve in an atrioventricular septal defect or determine that the atrioventricular valve and ventricle in Ebstein’s malformation will not be suitable for incorporation into a biventricular repair. Because world experience with this repair is somewhat limited, guidelines may be difficult to establish at this time.
The left heart
With regard to the left heart, a number of factors may have implications for the suitability of borderline left ventricles for incorporation into a biventricular repair, only one of which is left ventricular cavity size (Fig 8). These have been considered carefully by Rhodes and colleagues [5] in a study in which multiple left heart features were retrospectively analyzed for association with unsuitability for biventricular repair. The variables assessed were the anteroposterior and lateral dimensions of the mitral (and tricuspid) annuli measured in apical four-chamber and parasternal long-axis views. The area of the valves was then calculated from the formula of an ellipse from the diameters D1 and D2, with area = 
{(D1 x D2)/2}. Left ventricular volume was calculated from the bullet formula, and left ventricular mass was calculated from the calculated volume x 1.04 g/mL. They measured the relative length of the left ventricle by recording a ratio of left ventricular long-axis length to the long-axis length from the crux to the apex in the four-chamber view. The aortic annulus was measured in the parasternal long-axis view in systole. All linear and area measurements were normalized to body surface area. Two similar groups of variables were used to establish a discriminant score for success after two ventricular repair. A critical value was established for these values to give a simplified score, with one point allotted for a left ventricular long-axis ratio of 0.8 or less, an indexed aortic root of 3.5 cm2/m2 or less, a mitral valve index of 4.5 cm2/m2 or less, and a left ventricular mass index of less than 35 g/m2. Mortality would be 100% in patients receiving an overall score of two or more and 8% among patients with a score of one or less. In addition, multiple regression analysis was used to determine a linear function with a discriminant score [5]: 
. A score of 0.35 or less was associated with death after biventricular repair. Indeed there has even been some concern about this in the fetus [6]. It is clear that there is more to be considered in these complex decisions than the variables included in the Rhodes discriminant score. We recently completed a retrospective study in which we found that successful biventricular repair was achieved in several patients who would have been ineligible according to the Rhodes criteria (unpublished data).
View larger version (103K):
[in this window]
[in a new window]
|
Fig 8. This patient with a small left ventricle was considered inadequate for a biventricular repair and underwent a Norwood procedure. (Top) Parasternal long-axis view demonstrating the diminutive aorta (AO), left atrium (LA), and left ventricle (LV), which has evidence of endocardial fibroelastosis and thickening of the tendinous cords supporting the mitral valve leaflets. (Bottom) This apical four-chamber view demonstrates the hypoplastic nature of the left-heart structures. The right ventricle is apex-forming. The mitral annulus is less than 6 mm. (RA = right atrium; RV = right ventricle.)
|
|
Another important point regarding the left ventricle relates to function. As has been observed in the fetus, the ventricle that is functioning against an increased afterload has compromised perfusion (Fig 9). This relates particularly to the inner aspect of the ventricle, which has the highest stress and the lowest perfusion, as coronary artery flow occurs on a gradient from outside layers of the myocardium to its inner layers [7, 8]. The fetus must perfuse both of its ventricles in diastole, as both ventricles are at "systemic" pressure in systole. In addition, because there is substantial runoff in diastole through the placental circulation, the diastolic myocardial perfusion pressure is lower. The supply of blood flow to the ventricle in stenotic lesions is therefore compromised. One advantage to the fetus is that the systolic pressure–volume area (or myocardial demand) is lower prenatally than after birth. However, the ventricle with elevated systemic pressure has greater demand. If the supply-demand ratio of either ventricle were lower, endocardial fibrosis may well result (see Figs 8, 9). This fibrosis likely causes not only a systolic abnormality of the ventricle, but also diastolic dysfunction that prevents the ventricle from filling and also from growing. The earlier in gestation this occurs, the greater the probability of ventricular hypoplasia; the later it occurs, the greater the likelihood that the ventricle will achieve an adequate size. It is for this reason that intrauterine attempts at relieving critical aortic stenosis have been undertaken [9].
View larger version (57K):
[in this window]
[in a new window]
|
Fig 9. (A) This image is from a fetus of 36 weeks’ gestation who presented initially at 26 weeks. There was disparate growth in the right and left ventricles over the course of gestation, with a progressive increase in right ventricular size, appropriate for age, but with less than adequate growth of the left ventricle. At birth, this patient had an extremely diminutive left ventricle, and a Norwood procedure was performed. (B) This frame demonstrates a shunt from the left (LA) to right (RA) atrium through the foramen ovale. (LV = left ventricle; RV = right ventricle.)
|
|
Size is not the only consideration, because if there is substantial endocardial fibrosis, then the postnatal ventricle can neither contract nor relax. This is the reason we sometimes encounter a ventricle in which all of the Rhodes criteria for successful biventricular repair are satisfied, but in which repair incorporating the left ventricle has an unfavorable result due to left ventricular dysfunction, leading to similar problems on the right side as well.
With regard to volume analysis, it has been noted for the left ventricle but could quite easily be applied to the right ventricle, that a ventricle is not adequate for supporting an entire cardiac output unless the end-diastolic volume is greater than 20 mL/m2 [10]. Volume estimation has been discussed already in another section of this symposium, but we would like to emphasize that it is important to perform volume calculations whenever possible as one means of defining the adequacy or inadequacy of the ventricle. In a small number of patients we have seen, the Ross-Konno procedure was successfully employed on left ventricles that were borderline for incorporation into the circulation. As the Ross-Konno procedure may change the prospects for reconstruction of the left ventricular outflow tract [11], the area of the aortic outflow in the Rhodes criteria may need to be reevaluated. In our limited experience with the Ross-Konno procedure in borderline hypoplastic left heart, we have found that the mitral valve often is the limiting factor.
With these caveats in mind, it is then equally difficult to define which ventricles are destined to be placed in the univentricular repair category. Ongoing endeavors will support these early results with additional data.
References
- Van Arsdell G.S., Williams W.G., Maser C.M., et al. Superior vena cava to pulmonary artery anastomosis: an adjunct to biventricular repair. J Thorac Cardiovasc Surg 1996;112:1143-1149.[Abstract/Free Full Text]
- Freedom R.M. Pulmonary atresia with intact ventricular septum. Mt. Kisco: Futura, 1989.
- Hanley F.L., Sade R.M., Freedom R.M., Blackstone E.H., Kirklin J.W., Congenital Heart Surgeons Society. Outcomes in critically ill neonates with pulmonary stenosis and intact ventricular septum: a multi-institutional study. J Am Coll Cardiol 1993;22:183-192.[Abstract]
- Wong P.C., Sanders S.P., Jonas R.C., et al. Pulmonary valve-moderator band distance and association with development of double-chambered right ventricle. Am J Cardiol 1991;68:1681-1686.[Medline]
- Rhodes L.A., Colan S.D., Perry S.B., Jonas R.A., Sanders S.P. Predictors of survival in neonates with critical aortic stenosis. Circulation 1991;84:2325-2335.[Abstract/Free Full Text]
- Hornberger L.K., Sanders S.P., Rein A.J., Spevak P.J., Parness I.A., Colan S.D. Left heart obstructive lesions and left ventricular growth in the midtrimester fetus. A longitudinal study. Circulation 1995;92:1531-1538.[Abstract/Free Full Text]
- Vincent W.R., Buckberg G.D., Hoffman J.I.E. Left ventricular subendocardial ischemia in severe valvar and supravalvar aortic stenosis. Circulation 1974;49:326-333.[Abstract/Free Full Text]
- Lewis A.B., Heymann M.A., Stanger P., Hoffman J.I.E., Rudolph A.M. Evaluation of subendocardial ischemia in valvar aortic stenosis in children. Circulation 1974;49:978-984.[Abstract/Free Full Text]
- Maxwell D., Allan L., Tynan M.J. Balloon dilatation of the aortic valve in the fetus: a report of two cases. Br Heart J 1991;65:256-258.[Abstract/Free Full Text]
- Hoffman J.I.E. Critical aortic stenosis in infancy: when is a hypoplastic left ventricle too small?. Cardiovasc J S Africa 1992;3:30-36.
- Reddy V.M., Rajasinghe H.A., Teitel D.F., Haas G.S., Hanley F.L. Aortoventriculoplasty with the pulmonary autograft: the "Ross-Konno" procedure. J Thorac Cardiovasc Surg 1996;111:158-167.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
E. T. Alboliras, C. Mavroudis, E. Pahl, S. S. Gidding, C. L. Backer, and A. P. Rocchini
Left ventricular growth in selected hypoplastic left ventricles: outcome after repair of coarctation of aorta
Ann. Thorac. Surg.,
August 1, 1999;
68(2):
549 - 555.
[Abstract]
[Full Text]
[PDF]
|
|
|
