| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;80:1431-1439
© 2005 The Society of Thoracic Surgeons
a Divisions of Cardiovascular-Thoracic Surgery and Cardiology, Children's Memorial Hospital, Chicago, Illinois
b Departments of Surgery and Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, Illinois
Accepted for publication April 6, 2005.
* Address reprint requests to Dr Mavroudis, Division of Cardiovascular-Thoracic Surgery, Children's Memorial Hospital, 2300 Children's Plaza, M/C 22, Chicago, IL 60614 (Email: cmavroudis{at}childrensmemorial.org).
Presented at the Forty-first Annual Meeting of The Society of Thoracic Surgeons, Tampa, FL, Jan 24–26, 2005.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
METHODS: From 1997 through 2004, 102 consecutive patients underwent repair of tetralogy of Fallot at a median age of 5.9 months. Twenty-five patients had a prior shunt. Eighty-two patients (80%) had pulmonary valve-sparing procedures, predominantly through a transatrial and transpulmonary approach (n = 52). Twenty patients had a transannular patch (20%). Intraoperative measurements included the pulmonary valve annulus size and the postoperative pressure ratio between the right and left ventricles.
RESULTS: Eighty of 85 (94%) patients with z-score greater than –4 had a pulmonary valve-sparing procedure compared with 2 of 17 patients (12%) with pulmonary valve annulus z-scores less than –4 (p < 0.0001). All patients with a tricuspid pulmonary valve (n = 26) had a pulmonary valve-sparing procedure compared with 56 of 76 (74%) patients with a bicuspid pulmonary valve (p = 0.0016). Five patients with initial pulmonary valve-sparing operations required reoperation for residual stenoses; 4 pulmonary valve-sparing right ventricular outflow tract resections and 1 transannular patch. The only death occurred after reoperation elsewhere. Three of 9 patients (33%) who had a postoperative pressure ratio between the right and left ventricles greater than 0.7 after their initial pulmonary valve-sparing procedure required reoperation compared with 2 of 73 with postoperative pressure ratio between the right and left ventricles less than 0.7 (3%; p = 0.008). Fifteen of 25 patients (60%) with prior shunts had pulmonary valve-sparing procedures.
CONCLUSIONS: A pulmonary valve-sparing approach to the repair of tetralogy of Fallot was applied successfully in 80% of patients. Significant markers for success were a measured pulmonary annulus z-score of –4 or larger, a tricuspid pulmonary valve, and a postoperative pressure ratio between the right and left ventricles less than 0.7.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
A simultaneous trend in surgery for TOF is primary repair in infancy. The reported experience with primary repair of TOF in neonates and very young infants has shown excellent early results, but with a high utilization of TAP [6–10]. The resulting pulmonary valve insufficiency (PI) has been thought by some to be very well tolerated, and several long-term follow-up studies support this [11–13]. Conversely, some individuals do not tolerate moderate or severe PI, and the long-term sequelae of PI include right ventricular dysfunction [14–16], poor functional status [17, 18], arrhythmias, and sudden death [19, 20].
At our institution we have made a conscious effort to avoid the use of transannular patching to preserve a functional pulmonary valve. To this end we have used a strategy of transatrial, transpulmonary repair of TOF with aggressive pulmonary valve commissurotomy, extensive subvalvar resection, pantaloon pericardial supravalvar pulmonary arterial patching, and, when required, subvalvar infundibular augmentation with a polytetrafluoroethylene (Gore-Tex, WL Gore & Associates, Flagstaff, AZ) patch. As part of our strategy to minimize the use of TAP, we continue to favor the selective use of palliative systemic-to-pulmonary artery shunts for symptomatic infants in the first weeks of life when the pulmonary valve is quite small.
The purpose of this review is to assess our strategy of valve preservation on pulmonary valve function, and subsequent requirement for reoperation. Furthermore, we sought to identify patient and pulmonary valve characteristics that mark the limits of this strategy.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
All patients undergoing repair of TOF between January 1997 and July 2004 at our institution were identified from our computerized patient database. Patients with prior complete repair, TOF with pulmonary atresia, TOF with absent pulmonary valve, and TOF with atrioventricular canal defects were excluded.
Reparative techniques were similar for all patients and included aorto–bicaval cardiopulmonary bypass with moderate hypothermia (28°C). Antegrade cold blood cardioplegia was used in conjunction with topical ice saline slush and was repeated at 15- to 20-minute intervals. In the majority of cases, the RVOT was inspected through a long medial right atrial incision with gentle retraction of the tricuspid valve. The resection began through this approach, and in several cases, could be completed without a pulmonary incision. If there was evidence of residual subvalvar, valvar, or supravalvar stenosis after the transatrial resection, then a longitudinal pulmonary arteriotomy was made and the incision carried into both the right and left sinuses on a tricuspid and vertically oriented bicuspid valve (Fig 1A) or into the anterior sinus on a horizontally oriented bicuspid valve. These incisions were extended to the level of the valve. After noting the morphology of the pulmonary valve, any commissural fusion of the pulmonary valve leaflets was sharply incised. After the pulmonary valvotomy, the right ventricular (RV)–pulmonary artery junction, which we will refer to as the "annulus," was measured with dilators and recorded as the size of the largest dilator that was completely accepted. The z-score for the annular size was calculated from a published nomogram [21].
|
In cases in which the annulus was prohibitively small, a transannular incision was made with resection of any dysplastic valve leaflets, and a transannular Gore-Tex patch inserted. Before the analysis of these data, this decision had been empiric. In the first 2 years of this series, a monocusp valve fashioned from 0.1 mm Gore-Tex was inserted into the RVOT in 7 patients. However, not seeing a significant advantage with this approach, we have closed the transannular incisions without a monocusp since 1998 (n = 13).
The ventricular septal defect was closed in the majority of cases through the right atriotomy. Interrupted pledget-supported 5-0 Ethibond sutures (Ethicon, Somerville, NJ) and polyethylene terephthalate fiber (Dacron) patch material were used to close the ventricular septal defect in all but the smallest infants in whom glutaraldehyde-tanned pericardium and running 6-0 or 7-0 polypropylene suture technique was used.
After separation from cardiopulmonary bypass, the RV and left ventricular (LV) pressures were directly measured through an RV and transseptal puncture using a 21-gauge spinal needle connected to a pressure transducer. If RV pressures were greater than 80% of LV pressure, then the RVOT was revised. If transesophageal echocardiography demonstrated the residual obstruction at the level of the annulus, then a transannular incision was made and a Gore-Tex patch was inserted. If the residual gradient was in the subvalvar RVOT, then an infundibular incision and subvalvar Gore-Tex patch were placed, allowing for valve preservation.
Preoperative and postoperative data on all patients were extracted from the medical records. Preoperative echocardiography data were taken from the most recent preceding study, if available; otherwise the intraoperative transesophageal echocardiography findings were reported. The postoperative echocardiography data reported were acquired in the last study performed before discharge home. Follow-up echocardiography data represent the most recent available study and must have been acquired at least 4 months postoperatively.
Comparisons were made between the patients who underwent transannular patching and those who had valve preservation. Continuous variables were compared with two-tailed, unpaired Student's t test, and categorical data were compared with two-tailed Fisher's exact test.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
|
|
|
The outcomes of the TOF repairs are compared in Table 3. There was a similar mean RV/LV pressure ratio, although the postoperative RVOT gradient was greater in the valve-sparing group. The incidence of moderate or severe PI, however, was substantially greater in the TAP group. Follow-up echocardiographic data were available for 55 of the patients at a mean of 34 months, range 4 to 84 months. These data demonstrate a decrease in the peak RVOT gradient for the valve-sparing group but a mild increase in gradient in the TAP group. The incidence of at least moderate PI increased in both groups, but remained substantially higher in the TAP group. There was a similar incidence of arrhythmias and heart block for both surgical strategies. One patient had a residual ventricular septal defect, which was closed at the time of reoperation for RVOT obstruction.
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
The rationale for our avoidance of the use of a TAP comes from the myriad reports showing the problems of long-term PI, although this remains a controversial issue. There have been several studies demonstrating good long-term outcomes after repair of TOF with a TAP. Bacha and colleagues [11] reported the long-term functional outcomes of 45 patients operated on for TOF as infants at Boston Children's Hospital and showed that the functional status as measured by New York Heart Association class was similar for TAP and valve-sparing procedures. Kirklin's group at the University of Alabama, Birmingham examined 814 patients and found that the risk of reoperation for PI at 20 years was 7%; they ascribed this low rate of reoperation, despite a significant incidence of PI, to the adaptive properties of the right ventricle [12]. They also found no difference in New York Heart Association class between TAP and valve-sparing approaches. These findings are contradicted by other studies showing the deleterious effects of pulmonary insufficiency on the right ventricle when more-sensitive measurements are made. Exercise testing has shown that functional status is impaired in TOF patients with long-term PI [17, 18]. Cardiac magnetic resonance cine examination of TOF patients late after repair have demonstrated the PI is closely associated with TAP and that PI results in significant RV dysfunction, even in asymptomatic patients [15, 16]. Gatzoulis and associates [20] showed that PI and the use of TAP are associated with development of ventricular tachycardia and sudden death late after TOF repair. The group at Great Ormond Street has used tissue Doppler assessments of isovolumetric myocardial acceleration to more accurately assess RV function in the face of highly variable loading conditions that exist with PI. They found in a group of 124 patients in which the use of TAP highly correlated with PI that PI was associated with both RV and LV dysfunction [14]. The late effect of TAP and PI on both RV and patient functional status are further documented by the growing number of reports on the ventricular improvement and symptomatic relief achieved with pulmonary valve insertion late after TOF repair [22–24].
Although we use the term valve-sparing repair, it is important to point out that our procedure focuses on preserving the annulus as much or more than preserving the valve cusps. To preserve the annulus with minimal gradient at that level, we perform an aggressive valvotomy and sometimes detach the cusps at the commissures. It is quite possible that the valvotomy accounts for the relatively high incidence (36%) of PI at midterm follow-up. In those cases in which an infundibular patch is required, we believe that two patches above and below the annulus are preferable to a single patch crossing the annulus.
In our series the reoperation rate for the patients with pulmonary valve-sparing procedures (6%) was not statistically different from the reoperation rate for patients with TAPs (10%); it is noteworthy that a TAP was required in only one of the operative revisions. Had a more aggressive resection been preformed on these patients, perhaps several of these reoperations could have been avoided. The use of the RV/LV pressure ratio of less than 0.8 may be inadequate. Our cutoff value had been based on the work by Naito and coworkers [25] demonstrating decreased survival with an RV/LV pressure ratio greater than 0.8 and that from the University of Alabama, Birmingham demonstrating that an RV/LV pressure ratio greater than 0.85 was associated with a 2.5 times increased risk of death and a 7.3 times increased risk of reoperation [26]. The University of Alabama, Birmingham group also showed a decrease in the RV/LV pressure ratio with time, with the greatest reduction seen by those patients with the highest immediate postoperative ratios. Our approach of favoring a higher pressure load to avoid a TAP is also supported by the work by Nollert and colleagues [27], who found no association between an RV/LV pressure ratio of 0.7 and early or late survival, but found a significant decrease in early survival for TAP and decrease in late survival for any outflow tract patch. Hirsch and associates [8] at Michigan found that an RV/LV pressure ratio greater than 0.7 was a risk factor for reoperation among 61 neonates after primary repair of TOF. Based on the 33% incidence of reoperation among our patients with an RV/LV pressure ratio greater than 0.7, we have altered our strategy to accept a maximum RV/LV pressure ratio of 0.7, not 0.8. Given that 2 of these 3 individuals had subsequent valve-sparing revisions, and such revisions could have been done at the initial operation under the more conservative RV/LV pressure ratio criteria, the overall valve salvage rate should not necessarily decrease significantly although the reoperation rate should.
In this series the presence of a tricuspid pulmonary valve was an indicator of good outcome. All 26 tricuspid pulmonary valves were salvaged. The fact that they were associated with significantly larger pulmonary annuli may be that they indicate a lesser form of TOF. Alternatively, a tricuspid pulmonary valve allows for a commissurotomy at three points instead of two, thus more relief of valvar level obstruction and acceptance of a larger sizing probe. The tricuspid valves were also associated with a lower postoperative RV/LV pressure ratio and less PI. The rate of tricuspid pulmonary valves in this series is similar to that in other series, including the group at the University of California San Francisco who noted that 15 of 42 acyanotic infants (36%) had tricuspid pulmonary valves [28].
There are several obvious shortcomings of this study; among them is the lack of long-term follow-up. However, it is quite clear before discharge that 70% of the TAP patients have moderate or severe PI and that number only increases at the approximately 3-year follow-up. The 15% immediate and 36% late moderate or severe PI in the valve-sparing group compares favorably with the 70% and 85% incidence, respectively, in the TAP group. The RVOT gradient, on the other hand, was stable or improved with time in the valve-sparing group, which is consistent with others' experience [26].
Another issue is technology. Although we currently advocate preserving the pulmonary valve to avoid the long-term problems of PI, the future of percutaneous valve insertion may tip the scales in favor of aggressive relief of RVOT at the expense of more PI. A pulmonary valve could later be inserted if needed as an outpatient procedure. Of course, that technology has not yet proven itself effective and reliable in humans and remains experimental [29].
Shunting has an uncertain effect on the rate of valve preservation; however, we were able to perform a valve-sparing operation on 60% of the 25 patients with prior shunts. There was no shunt-related mortality and one instance of significant morbidity. In 4 of 25 cases (16%) a pulmonary arterioplasty was performed after shunt takedown to relieve local stenosis. Although primary neonatal repair can certainly be performed, the rate of TAP is very high. The Michigan experience with neonatal TOF repair involves 100% TAP [8]. Even in the best-case scenario involving acyanotic infants with a mean weight of 4.6 kg and median age of 62 days, Parry and colleagues [28] reported a 24% use of TAP.
The group from Toronto reported their experience from 1997 through 1999 with all children with TOF who were 18 months or younger at the time of repair [30]. The strategy during that period was elective operation at 6 months of age on acyanotic infants, and immediate primary repair on cyanotic infants. The overall rate of transannular repair was 36%, but was greatest (50%) for patients who were younger than 3 months of age. Although transient, the youngest group (<3 months) did experience a higher incidence of organ dysfunction.
Although we did not demonstrate that the pulmonary annulus grew after placement of the shunt, there is evidence for such growth. Sousa Uva and associates [31] reported on 56 patients with TOF repair younger than 6 months of age. Palliative shunts were performed in 15 patients of whom 8 had complete repairs at the time of reporting. The z-score of the pulmonary annulus in these 8 patients had increased by a mean of 2.2. The rate of TAP was 56% in the 41 primarily repaired infants, who had a mean age of 2.9 months and a mean pulmonary annulus z-score of –1.8. Only 1 of 8 infants repaired after shunt required a TAP.
Karl and colleagues [5] operated on 366 patients with TOF between 1980 and 1991 using a transatrial, transpulmonary approach with 0.5% hospital mortality. They attribute these outstanding results to delaying surgery until the infants weighed between 6 and 8 kg. A palliative shunt was used in 37% of patients to delay repair until they had grown.
We have evolved our TOF patient management strategy to maximize the preservation of pulmonary valves. For acyanotic infants we electively delay surgery until they are 4 to 6 months of age. For cyanotic infants, who are younger than 2 months of age or have never left the hospital because of other comorbidities, we perform a palliative shunt and repair at 6 to 9 months of life. This strategy has resulted in a very low operative mortality, a reasonable reoperation rate, and a high rate of pulmonary valve preservation. This review has demonstrated that markers for success for pulmonary valve preservation are a measured pulmonary annulus z-score of –4 or larger, a tricuspid pulmonary valve, and a postoperative RV/LV pressure ratio less than 0.7. This strategy should reduce PI, preserve RV function, and decrease the incidence of late arrhythmias, which are the determinants of long-term outcome.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Your strategy seems virtually identical to ours, and I congratulate you on what are really outstanding surgical results. And I predict that as we follow these patients, 10, 20, 30, 40 years, we're going to see that these right ventricles are holding up better.
But I wanted to explore with you or have you hypothesize about what's happening with these pulmonary valves. Even in the patients that have a morphologic tricuspid valve with a small annulus, it's hard for me to accept, really, that those valves are all that competent. And I personally believe that the real issue is maintaining a contractile competent infundibulum. If you have a ventriculotomy, or certainly a big outflow patch, you have this capacitance chamber. And I think when you're noting in your postoperative follow-up echocardiograms a really low incidence of pulmonary insufficiency, isn't that really what you're seeing, that you don't have this big capacitance sump sitting below the pulmonary valve?
DR STEWART: You make an excellent point, Dr Fraser, and this has been shown by Gatzoulis and others that maintaining a somewhat restrictive right ventricle decreases the pulmonary insufficiency (PI). So I agree that it may be as much the right ventricle as it is the valve that makes a difference in the degree of PI between patients with and without a transannular patch.
DR JEFFREY P. JACOBS (St. Petersburg, FL): First, I'd congratulate you for an excellent presentation in the typical Mavroudis and Backer fashion with excellent data, excellent presentation of the data.
I think that it seems that your strategy of choosing a shunt versus a complete repair in some of the children potentially leads to preservation of the pulmonary valve down the road. And I think in our program we would agree that preservation of the pulmonary valve is something that's important.
I would just ask you, what is your criteria for deciding on a baby when you're going to do a shunt and when you're going to do a complete repair, and if the goal of preserving the pulmonary valve down the road affects that decision?
DR STEWART: Dr Jacobs, we do not have a hard-and-fast guideline or absolute size criteria when to perform a shunt. Essentially, in the first 1 to 2 months of life, any symptomatic baby that we don't think we can at least attempt a pulmonary valve-sparing operation, we'll go ahead and perform a shunt.
And although we don't have the data here to prove that the pulmonary annulus grows, this has been shown by others. Dr Lacour-Gayet reported that the pulmonary annulus in their shunted patients had increase by 2.2 z-scores by the time of complete repair, and also that there was a lower incidence of transannular patch after shunting.
DR HENRY L. WALTERS (Detroit, MI): One of the important outcome events in the future, as you follow this series of patients, is going to be comparing the frequency with which the pulmonary valve has to be replaced in the valve-sparing group compared with that of the group that received the transannular patch. There are also other important outcome events that you will undoubtedly be following in this valve-sparing group. They would be exercise tolerance, right ventricular volumes, incidence of important arrhythmias, et cetera. Do you have any preliminary data or future plans to examine these variables?
DR STEWART: Dr Walters, we haven't studied these patients in terms of those very important functional criteria. Right ventricular dysfunction has already been shown by many studies to be a long-term consequence of pulmonary insufficiency after tetralogy repair. Arrhythmias have also been closely associated with long-standing PI. This study focused on how often we could repair tetralogy with preserved pulmonary valve function to avoid these outcomes. I agree that it will be interesting to follow up on this cohort to see if valve preservation improves functional outcomes long-term.
DR MARSHALL L. JACOBS (Philadelphia, PA): My question relates to an area you didn't touch on, and that's the subsequent growth and development of the branch pulmonary arteries. You have a series of greater than 100 infants, median age of 6 months, undergoing tetralogy repair. And your not mentioning any subsequent reinterventions for branch pulmonary artery stenosis suggests either it wasn't the focus of the paper, which I understand to be the case, or that there is something about your technique that results in flow distribution and hemodynamics that is associated with preferable growth and development of the branch pulmonary arteries.
If we had a series of 100 infant tetralogies, a handful would be back for one or more balloon angioplasties or surgical augmentations of branch pulmonary arteries. Was that complication encountered at all in this series?
DR STEWART: Dr Jacobs, we did encounter this problem, although relatively infrequently. There were 4 patients of the 25 with prior shunts who required a right pulmonary arterioplasty at the time of repair. After complete repair, there was 1 patient who required early balloon dilatation of a right pulmonary artery stenosis. This patient did not have a prior shunt.
DR JACOBS: And no coarctation of the left pulmonary artery in the area of ductal insertion or any problems of that sort?
DR STEWART: Ductal coarctations or branch pulmonary stenoses occurred in several patients before complete repair. These patients were treated with branch pulmonary arterioplasties, usually as an extension of the pantaloon pericardial patch, at the time of initial repair.
DR RICHARD N. GATES (Orange, CA): I have a question for you with regards to right ventricular to left ventricular (RV/LV) pressure ratio. Intuitively I like the idea of moderate stenosis for severe insufficiency, but we know in patients who have had a ventriculotomy that an RV/LV pressure ratio, once it gets over 0.6, is associated with sudden death, particularly as children start exercising when they're older. Have you looked at any of these patients that were in the 0.7 to 0.8 RV-to-LV pressure ratio, have you done any exercise tolerance testings on them and any Holter monitorings when they get to be 4 or 5 or older?
DR STEWART: We haven't measured late pressure ratios so much as we've looked at right ventricular outflow tract (RVOT) gradients. We found that at the 3-year follow-up that RVOT gradients had diminished compared with the immediate postoperative period. This finding is consistent with what others have reported. In fact, the greatest decrease is seen in those patients with the higher gradients.
DR GERHARD ZIEMER (Tuebingen, Germany): What makes you sure that you actually do a valve-sparing operation but not just an annulus-sparing operation and leaving some torn valvular tissue in place? Do you really see that the valve or whatever you have left in the annulus stops the retrograde flow in the pulmonary arteries?
But as you just showed yourself, even when you have a transannular patch with cutout pulmonary valve, there is a certain set of patients who does not have regurgitation at all. And the smaller percentage of patients with regurgitation in the annulus-sparing operation may be just a matter of not having enlarged the outflow tract more than necessary as it is often the case in transannular patching.
DR STEWART: This is a bit of a semantics question. The true measure of what we're doing is in the late pulmonary insufficiency. So whether we're sparing a valve per se or sparing the annulus, it's effective as evidenced by the incidence of moderate or severe late PI; 36% in the "valve-sparing" group versus 84% in the transannular patch group.
DR ZIEMER: You cannot call it semantics when you compare your term when it is used in aortic surgery, where the aortic valve-sparing operation is really a very strict nomenclature that could not survive with the amount of regurgitation you allow for the spared or saved pulmonary valve.
DR STEWART: This is true.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Anagnostopoulos, A. Azakie, S. Natarajan, N. Alphonso, M. M. Brook, and T. R. Karl Pulmonary valve cusp augmentation with autologous pericardium may improve early outcome for tetralogy of Fallot J. Thorac. Cardiovasc. Surg., March 1, 2007; 133(3): 640 - 647. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Pozzi, A. Quarti, and A. F. Corno Tetralogy of Fallot MMCTS, October 9, 2006; 2006(1009): 1487. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Airan, S. K. Choudhary, H. V. J. Kumar, S. Talwar, J. Dhareshwar, R. Juneja, S. S. Kothari, A. Saxena, and P. Venugopal Total transatrial correction of tetralogy of fallot: no outflow patch technique. Ann. Thorac. Surg., October 1, 2006; 82(4): 1316 - 1321. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. P. Graham Jr The Year in Congenital Heart Disease J. Am. Coll. Cardiol., June 20, 2006; 47(12): 2545 - 2553. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
