Click here to read this article as a CME activity
Ann Thorac Surg 2007;84:587-593
© 2007 The Society of Thoracic Surgeons
Original Articles: Cardiovascular
Repair of Neonates and Young Infants With Ebstein’s Anomaly and Related Disorders
Christopher J. Knott-Craig, MDa,*,
Steven P. Goldberg, MDb,
Edward D. Overholt, MDa,
Edward V. Colvin, MDb,
James K. Kirklin, MDb
a Department of Thoracic and Cardiovascular Surgery, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma
b Division of Cardiothoracic Surgery, University of Alabama at Birmingham, Birmingham, Alabama
Accepted for publication March 21, 2007.
* Address correspondence to Dr Knott-Craig, University of Alabama at Birmingham, 716 Ziegler Research Bldg, 703 19th St S, Birmingham, AL 35294 (Email: ckc{at}uab.edu).
Presented at the Forty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 29–31, 2007.
| Pediatric cardiac surgery:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract
|
|---|
Background: Severely symptomatic neonates and young infants with Ebstein’s anomaly usually die without surgical intervention. The relative risks and benefits of single-ventricle palliation versus a two-ventricle repair are uncertain. In a recent series, 69% early survival with single-ventricle palliation was reported in 16 neonates with Ebstein’s anomaly. Our institutional bias has been to do a two-ventricle repair in all such patients.
Methods: We reviewed our entire surgical experience with a two-ventricle repair in the severely symptomatic neonate (n = 22) and young infant (n = 5). The indications for operation were ventilator dependence, severe cardiac failure, prostaglandin-dependent circulation, and gross cardiomegaly.
Results: Between 1994 and 2006, 27 consecutive patients with Ebstein’s anomaly underwent operation. Associated comorbidities included anatomic or functional pulmonary atresia (n = 18), ventricular septal defects (n = 3), small left ventricle (n = 3), hypoplastic branch pulmonary arteries (n = 3), previous cardiac surgery (n = 4), significant intracranial hemorrhage (n = 3), hepatic necrosis and renal insufficiency (n = 3), and malignant tachyarrhythmias (n = 4). Operations consisted of tricuspid valve repair (n = 23) or valve replacement (n = 2), Blalock-Taussig shunt only (n = 1), and bilateral pulmonary arterioplasty with bidirectional Glenn (n = 1). Hospital survival was 74%, and there have been no late deaths during a median follow-up period of 5.4 years (range, 0.2 to 12 years). Three patients required tricuspid valve replacement during the follow-up period. Late arrhythmia requiring medication is present in 1 patient. All patients are currently in New York Heart Association functional class I.
Conclusions: Two-ventricle repair currently has similar early survival compared with single-ventricle palliation. The advantages of a better physiologic repair can be anticipated for a longer follow-up period.
Ebstein’s anomaly (EA) represents a downward, or apical displacement of the posterior and septal leaflets of the tricuspid valve, with a resultant redundant "atrialized" component of the right ventricle situated above the plane of the valve annulus. The anterior leaflet is abnormally broad, often described as saillike, although still based at the true annulus [1, 2]. In the symptomatic neonate with EA, the leading edge of the anterior leaflet is often attached to the free wall of the right ventricle. Associated cardiac anomalies such as pulmonary atresia are common, making a two-ventricle repair technically very difficult [3, 4]. During the past three decades, many repairs have been developed in adults [5] and older children, with variations on the theme of plication of the redundant atrialized right ventricle and utilization of the saillike anterior leaflet in construction of a competent monocuspid valve [6–9]. Symptomatic neonates and very young infants with EA usually present as critically ill, unstable patients, most of whom will die in early infancy without a surgical procedure [7]. In addition, neonates with severe tricuspid regurgitation or gross cardiomegaly, and who are otherwise asymptomatic, are reported to have an associated mortality in infancy of 45% [10, 11]. Single-ventricle palliation of the neonate with EA was first suggested by Starnes and associates in 1991 [12]. We successfully did the first two-ventricle repairs in neonates in 1994, and reported those in 2000. In 2002, Sano and coworkers [13] reported a modified technique of single-ventricle palliation using "ventricular exclusion."
When presented with a critically ill neonate or young infant with EA, the first essential surgical decision that needs to be made is whether a two-ventricle repair is feasible, or whether palliation with embarkation on a single-ventricle pathway is the more reasonable approach. The relative risks and benefits of single-ventricle palliation compared with two-ventricle repair are uncertain. In 2006 Reemtsen and colleagues [4] updated their results of single-ventricle palliation (closure of the tricuspid orifice with a fenestrated patch, aortopulmonary shunt, and later Fontan completion) in 16 neonates with EA, with 69% early survival. Our institutional preference has been to perform a two-ventricle repair in all neonates and young infants with EA and similar disorders. The results of this approach are presented in this manuscript.
|
Patients and Methods
|
|---|
Between 1994 and December 2006, we operated on 27 consecutive symptomatic neonates (n = 22) and young infants (n = 5) with EA, at Children’s Hospital of Oklahoma and University of Alabama at Birmingham. Approval for this analysis was obtained from the institutional review boards. This series includes 1 neonate operated on in Indianapolis with John Brown, MD, and 1 in San Diego with Glenn Pelletier, MD. Three additional neonates responded well to nitric oxide and prolonged anesthesia, and were ultimately weaned from the ventilator and discharged without surgical intervention. One patient subsequently had a successful repair at 8 months of age. There were no neonates with Uhl’s anomaly in this series.
The indications for operation were ventilator dependence, prostaglandin-dependent circulation, severe cardiac failure, and gross cardiomegaly. Table 1
describes the associated comorbidities in our patients, most commonly pulmonary atresia (n = 18, 67%), ventricular septal defects (n = 3), small left ventricle (n = 3), hypoplastic branch pulmonary arteries (n = 3), previous cardiac operation (n = 4), significant intracranial hemorrhage (n = 3), ischemic hepatic necrosis and renal insufficiency (creatinine > 1.5; n = 3), and malignant tachyarrhythmias (n = 4). Recent clinical and echocardiography follow-up was complete. Median follow-up was 5.4 years (range, 0.2 to 12 years).
In one patient, the Starnes repair was successfully taken down, and the patient was converted to a two-ventricle repair with tricuspid valve replacement. Another critically ill infant transferred with complete heart block, device closure of an atrial septal defect, and prior tricuspid valve replacement with a tissue prosthesis. After subsequent balloon disruption of the bioprosthesis at the referring institution, the patient underwent successful repeat replacement of the tricuspid valve, removal of the atrial septal defect device, and patch closure of the atrial septum. One patient remained ventilator-dependent despite prior Blalock-Taussig shunt and balloon dilatation of the pulmonary arteries before being referred for a two-ventricle repair. Another patient had four prior sternotomies and three Blalock-Taussig shunts before being referred for repair.
The essential principles in our two-ventricle repair, as previously described [8, 14] are as follows: (1) creation of a competent monocuspid tricuspid valve, based on the anterior leaflet; (2) reduction atrioplasty; (3) partial closure of the atrial septal defect, leaving behind a small fenestration; and (4) simultaneous repair of other cardiac defects, enumerated above. The actual technique of particularly the valve repair has evolved with time, with most recent emphasis being placed on the addition of a Sebening single-stitch approach (Fig 1) to keep the anterior leaflet approximated to the interventricular septum, minimizing tricuspid regurgitation during episodes of pulmonary hypertension [15].
View larger version (66K):
[in this window]
[in a new window]
|
Fig 1. Sebening single-stitch valvuloplasty (papillary muscle of anterior leaflet through tethered septal leaflet).
|
|
The operation is conducted with full cardiopulmonary bypass and bicaval cannulation, using either moderate systemic hypothermia or circulatory arrest. After identifying the course of the right coronary artery (so as not to injure it during the atrial reduction), the free wall of the enlarged right atrium is widely excised. The atrial septal defect is closed with a patch, although a 3- to 4-mm fenestration is left behind. We have found that this aids in decompressing the right heart and improving cardiac output postoperatively. The crux of the valve repair, again, rests on creation of a monocuspid valve using the broad sheet of the anterior leaflet, such that when the annular diameter is reduced with an annuloplasty, the edge of the anterior leaflet will coapt directly against its facing septal wall. The annuloplasty stitch is placed with one pledgetted end in the coronary sinus, the other pledgetted end at the location of the commissure between the anterior and (if otherwise present) posterior leaflets (Fig 2). Approximation of the annuloplasty stitch effectively partitions the tricuspid valve orifice into two openings, the "main" orifice and a more "caudal" or "rightward" orifice (Fig 3A). If the valve is then competent, the caudal orifice is closed, plicating the atrialized portion of the right ventricle at the same time as reducing the annular diameter (Fig 3B). A more complex repair is required if the anterior leaflet is dysplastic, poorly fenestrated (restricting right ventricular inflow), or more firmly tethered to the underlying myocardium. In these cases, the leaflet is taken down from the annulus [16, 17], freed up from its underlying attachments (Fig 4), and then reattached to the newly reduced annulus (Fig 5). If the anterior leaflet does not reach the opposing wall, then it can be detached from the annulus and enlarged with an autologous pericardial patch to allow it to bridge the gap (Fig 6). In two of our patients, a modified Blalock-Taussig shunt was added to improve postoperative pulmonary blood flow. In all patients, the placement of a peritoneal dialysis catheter has become very helpful in draining postoperative ascites, aiding diaphragmatic excursion and postoperative pulmonary mechanics.
View larger version (64K):
[in this window]
[in a new window]
|
Fig 2. Tricuspid annuloplasty stitch placed in coronary sinus and at location of anteroposterior commissure. (ASD = atrial septal defect.)
|
|
View larger version (55K):
[in this window]
[in a new window]
|
Fig 3. (A) Approximation of annuloplasty stitch creates two openings, the "caudal" orifice containing the entrance to the atrialized right ventricle. (B) Closure of the caudal opening plicates the atrialized right ventricle and creates a competent monocuspid valve.
|
|
Postoperatively, the patients initially remain paralyzed, sedated, and ventilated. We use large tidal volumes on the ventilator (12 to 15 mL/kg) because the lungs are usually suboptimal, their full development having been difficult in a chest crowded by cardiomegaly. We institute all measures to reduce elevated pulmonary vascular resistance, including inhaled nitric oxide. Oxygen carrying capacity is optimized by maintaining a hematocrit of at least 45% to 50%, as the oxygen saturation may be low (and tolerated) for the first 3 to 5 days after surgery. We have found isoproterenol infusion (0.05 µg · kg–1
· min–1) to be the most useful inotropic agent in the perioperative period.
|
Results
|
|---|
In all but 2 patients, two-ventricle repairs were performed. In 1 patient with a small right ventricle and only mild tricuspid regurgitation, but with a prostaglandin-dependent circulation, a Blalock-Taussig shunt only was done. This was taken down 6 months later and a bidirectional Glenn added. Another infant who had three previous Blalock-Taussig shunts elsewhere, before being referred for repair, had extensive bilateral patch augmentation of the pulmonary arteries and a bidirectional Glenn.
Seventy-four percent (20 of 27 patients) survived to hospital discharge. Recent follow-up was achieved in all survivors. The median follow-up was 5.4 years (range, 0.2 to 12 years). There have been no late deaths. All patients are currently in New York Heart Association class I functional status. Three patients have required tricuspid valve replacement during the follow-up period. Bioprostheses were placed 2 to 4 years after their initial operation. Although 4 patients had symptomatic tachyarrhythmias preoperatively, and despite the well-known association of EA with accessory pathway tachydysrhythmia [2, 6], we have had only 1 patient (5%, 1 of 20 patients) manifest a transient tachyarrhythmia requiring medication during the follow-up period.
Of the 7 nonsurvivors, the factors associated with their poor outcome were variable. Of the 11 patients with anatomic pulmonary atresia, 6 died (66%) postoperatively. Three patients required support with an extracorporeal membrane oxygenator postoperatively—1 patient experienced extracorporeal membrane oxygenator pump failure on the third postoperative day; 1 required extracorporeal membrane oxygenator followed by tricuspid valve replacement, the function having been deemed "marginal" by echocardiography; the third extracorporeal membrane oxygenator patient had hypoplastic pulmonary arteries, and underwent homograft replacement of the pulmonary valve. Although this patient subsequently had excellent hemodynamics, she experienced a severe intracranial hemorrhage and support was withdrawn. No patient requiring extracorporeal membrane oxygenator support survived to hospital discharge. Two patients had small left ventricles preoperatively, and succumbed to low cardiac output after surgery (Fig 7). There was a technical error in 1 patient, involving injury to the right coronary artery during the reduction atrioplasty. After this case, we adopted a policy of mandatory marking of the right coronary artery course with a surgical pen before making any atriotomy. Finally, 1 critically ill patient came to us already having suffered a grade IV intracranial hemorrhage associated with Staphylococcus aureus septicemia. This patient underwent surgery at 10 days of life, but unfortunately had a recurrence of the intracranial hemorrhage, and despite a satisfactory repair, the situation was irretrievable and support was withdrawn.
View larger version (84K):
[in this window]
[in a new window]
|
Fig 7. Echocardiogram of a patient with an unacceptably small left ventricle. (LA = left atrium; RA = right atrium; RV = right ventricle.)
|
|
|
Comment
|
|---|
Since 1994 we have operated on 27 critically ill neonates or young infants, and have been able to achieve early survival rates that are similar to recent reports involving single-ventricle palliation [4]. In our opinion, it remains a feasible, reasonable approach, and the benefits of having a two-ventricle system will surely be realized over the course of a longer follow-up period. In contrast to the recent report by Reemtsen and associates [4] of 16 patients with single-ventricle palliation in whom the tricuspid valve was not amenable to repair, we felt comfortable repairing all but 1 neonate. This patient had prostaglandin-dependent circulation and 2/4 tricuspid regurgitation; we elected to initially place a small aortopulmonary shunt, followed 4 months later by a bidirectional Glenn anastomosis, with an excellent result.
From our experience (74% survival), it seems evident that some neonates may have had a better chance of survival with an alternative operation. This was most evident in patients with anatomic pulmonary atresia, which was associated with a 66% mortality. Many of these patients had small pulmonary arteries; others had diminutive functional right ventricles. As the primary chordae to the anterior leaflet often originate from rudimentary papillary muscles in the infundibulum, these may easily be injured during the construction of a right ventricular outflow tract patch or pulmonary valve replacement, making the tricuspid repair inadequate.
The dilemma remains how to decide which operation is most suitable for each patient: from our experience, we currently feel confident in our ability to repair a neonate who has some prograde flow across the pulmonary valve, albeit while optimizing the pulmonary vascular resistance with inhaled nitric oxide and a low-dose infusion of isoproterenol. Figure 8
represents our management algorithm when approaching a patient with EA, based on their associated anomalies, as well as their overall clinical stability. Patients with pulmonary atresia—representing greater than 60% of the patient group in our series—seem to fall into two general groups: those who are relatively stable on the ventilator, often with gross cardiomegaly, severe tricuspid regurgitation, and a dysplastic (rather than a true EA-like) valve; and those who are very unstable with ongoing progressive metabolic acidosis and functional pulmonary atresia, often with retrograde flow back through the pulmonary valve. The former usually do well with either a two-ventricle repair if they have a decent size functional right ventricle, or just an initial shunt followed by a 1
-ventricle repair at 4 to 6 months of age if they have small functional right ventricles; the latter are probably best served with initial ligation of the main pulmonary artery and a Starnes single-ventricle palliation.
View larger version (31K):
[in this window]
[in a new window]
|
Fig 8. Management algorithm of Ebstein’s patients based on associated anomalies and overall stability. (B-T = Blalock-Taussig; LV = left ventricle; MPA = main pulmonary artery; Pulm. = pulmonary; RV = right ventricle.)
|
|
The surgical management of critically ill neonates and young infants with EA remains challenging. The judgment as to whether a diminutive, poorly functional right ventricle will be adequate to maintain circulation in the setting of initially high pulmonary vascular resistance is, in our opinion, still very subjective; furthermore, the adequacy of a small, possibly anatomically abnormal, and compressed left ventricle is equally subjective. Those patients with diminutive pulmonary arteries may be best served by initially placing an aortopulmonary shunt and delaying definitive palliation or repair until the patient is 3 to 5 months old. Those with small left ventricle probably should be considered for transplantation.
It should be noted that initial Starnes palliation does not irrevocably place the patient on a one-ventricle pathway. We successfully converted a 3-month-old critically ill ventilator-dependent infant to a two-ventricle repair, and this patient is currently an asymptomatic school-going child. This is certainly possible if the right ventricle muscularizes and hypertrophies during the first months of life, particularly as a consequence of associated pulmonary stenosis, which can be corrected at reoperation.
Finally, the management of neonates and young infants with EA and functionally similar cardiac conditions remains challenging. We have indicated that a two-ventricle repair is not only feasible, but can be achieved with acceptable risk and excellent medium-term follow-up. The benefits of a two-ventricle repair over single-ventricle palliation may become more apparent during a longer follow up period, provided the repair remains durable.
|
Discussion
|
|---|
DR FRANK A. PIGULA (Boston, MA): Thank you, Dr Goldberg, for a very interesting and provocative presentation.
While discussants are lining up, I would just make one note. I would make a comment that the artwork was also done by Dr Goldberg in these presentations.
And my question to you is how do you handle the pulmonary valve in the patients that have functional pulmonary atresia? What is the strategy there?
DR GOLDBERG: Ordinarily, the pulmonary valve is normal, or just a little smaller than usual. If it is about 7 mm, we leave it alone. If it is hypoplastic, we will use a transannular patch. If the tricuspid repair is suboptimal, then we would replace the pulmonary valve with a pulmonary homograft.
DR SHUNJI SANO (Okayama, Japan): Congratulations on your excellent results.
I think the Ebstein’s anomaly is not only the tricuspid valve disease but also disease of the right ventricle and right atrium, or right heart disease.
Our approach is to exclude these diseased right ventricles and to create the 1–1/2 or one-ventricle repair because the dilated right ventricle compresses the left ventricle, so the left ventricle seems to be small. But in most of the Ebstein’s, the left ventricle is not small.
I wondered whether in your patients the dilated right ventricle will still be dilated and compress the left ventricle, so that means the intraventricular septum, the motion changed from paradoxic to normal motion later, all the hemodynamic data of your patients, because we can compare the hemodynamic result in a 1–1/2 or a one-ventricle repair versus a two-ventricle repair. Do you have any data?
DR GOLDBERG: That is an excellent question. Obviously we are all well aware of your quite elegant and novel operative repair. For a question that detailed, I might defer to Dr Knott-Craig for some of those details.
DR CHRISTOPHER J. KNOTT-CRAIG (Birmingham, AL): We have not used the ventricular exclusion technique in the neonates. I am not sure it is possible. Regarding those patients with anatomic pulmonary atresia: the functional RV (right ventricle) is often very small. And when you open that right ventricular outflow tract, the functional RV may easily be injured. In that setting, if those right ventricles are real small now, we would probably now do either a BT (Blalock-Taussig) shunt, or streamline the patient to some alternative repair other than a two-ventricle repair.
DR JOSEPH A. DEARANI (Rochester, MN): Yes. I wanted to follow up with what Dr Knott-Craig said. I think that the single-ventricle approach has its greatest application when you have severe tricuspid regurgitation and anatomic pulmonary atresia. In Ebstein’s anomaly, the right ventricle is always myopathic to a certain degree and sometimes the right ventricular dysfunction can be quite severe. The combination of progressive pulmonary valve and tricuspid valve dysfunction, usually regurgitation, is added insult to the abnormal right ventricle. While a biventricular circulation with repeated pulmonary and/or tricuspid valve replacements may be reasonable in older patients, the alternative strategy of a single ventricle approach without the need for potentially two valve replacements would be our preference for infants and young children.
|
References
|
|---|
- Epstein ML. Congenital stenosis and insufficiency of the tricuspid valveIn: Allen HD, Clark EB, Gutgesell HP, et al. editors. 2nd ed. Moss and Adams’ heart disease in infants, children, and adolescents. vol 6. Philadelphia: Lippincott Williams & Wilkins; 2001. pp. 810-819.
- Dearani JA, Danielson GK. Ebstein’s anomalyIn: Sellke FW, del Nido PJ, Swanson SJ, editors. 7th ed. Sabiston and Spencer surgery of the chest, 7th ed, vol 2. vol 2. Philadelphia: Elsevier Saunders; 2005. pp. 2219-2231.
- Dobell ARC. Ebstein’s Abnormality of the tricuspid valveIn: Arciniegas E, editor. Pediatric cardiac surgery. Chicago: Year Book Medical Publishers; 1985. pp. 315-323.
- Reemtsen BL, Fagan BT, Wells WJ, Starnes VA. Current surgical therapy for Ebstein’s anomaly in neonates J Thorac Cardiovasc Surg 2006;132:1285-1290.[Abstract/Free Full Text]
- Dearani JA, Danielson GK. Surgical management of Ebstein’s anomaly in the adult Semin Thorac Cardiovasc Surg 2005;17:148-154.[Medline]
- Danielson GK, Maloney JD, Devloo RAE. Surgical repair of Ebstein’s anomaly Mayo Clin Proc 1979;54:185-192.[Medline]
- Carpentier A, Chauvaud S, Mace L, et al. A new reconstructive operation for Ebstein’s anomaly of the tricuspid valve J Thorac Cardiovasc Surg 1988;96:92-101.[Abstract]
- Knott-Craig CJ, Overholt ED, Ward KE, et al. Neonatal repair of Ebstein’s anomaly: indications, surgical technique, and medium-term follow-up Ann Thorac Surg 2000;69:1505-1510.[Abstract/Free Full Text]
- Boston US, Dearani JA, O’Leary PW, Driscoll DJ, Danielson GK. Tricuspid valve repair for Ebstein’s anomaly in young children: a 30-year experience Ann Thorac Surg 2006;81:690-695.[Abstract/Free Full Text]
- Celermajer DS, Cullen S, Sullivan ID, et al. Outcome in neonates with Ebstein’s anomaly J Am Coll Cardiol 1992;19:1041-1046.[Abstract]
- Celermajer DS, Dodd SM, Greenwald SE, Wyse RK, Deanfield JE. Morbid anatomy in neonates with Ebstein’s anomaly of the tricuspid valve J Am Coll Cardiol 1992;19:1049-1053.[Abstract]
- Starnes VA, Pitlick PT, Bernstein D, et al. Ebstein’s anomaly appearing in the neonate J Thorac Cardiovasc Surg 1991;101:1082-1087.[Abstract]
- Sano S, Ishino K, Kawada M, et al. Total right ventricular exclusion procedure: an operation for isolated congestive right ventricular failure J Thorac Cardiovasc Surg 2002;123:640-647.[Abstract/Free Full Text]
- Knott-Craig CJ, Overholt ED, Ward KE, et al. Repair of Ebstein’s anomaly in the symptomatic neonate: an evolution of technique with 7-year follow-up Ann Thorac Surg 2002;73:1786-1793.[Abstract/Free Full Text]
- Augustin N, Schmidt-Habelmann P, Wottke M, Meisner H, Sebening F. Results after surgical repair of Ebstein’s anomaly Ann Thorac Surg 1997;63:1650-1656.[Abstract/Free Full Text]
- Quaegebeur JM, Narayanswami S, Fraser AG, et al. Surgery for Ebstein’s anomaly: the clinical and echocardiographic evaluation of a new technique J Am Coll Cardiol 1991;17:722-728.[Abstract]
- Chauvaud S. Ebstein’s malformationSurgical treatment and results. Thorac Cardiov Surg 2000;48:220-223.
Top
Abstract
Introduction
