|
|
||||||||
Eur J Cardiothorac Surg 2000;17:146-153
© 2000 Elsevier Science NL
a Department of Thoracic and Cardiovascular Surgery, German Paediatric Heart Centre, Sankt Augustin, Germany
b Department of Paediatric Cardiology, German Paediatric Heart Centre, Sankt Augustin, Germany
c Department of Paediatric Cardiology, University Hospital, Essen, Germany
Corresponding author. German Paediatric Heart Centre, Johanniter Kinderklinik, Arnold-Janssen-Strasse 29, 53757 Sankt Augustin, Germany. Tel.: +49-2241-2493-26; fax: +49-2241-249-233
e-mail: andreas.e.urban.md{at}t-online.de
| Abstract |
|---|
|
|
|---|
Key Words: Atrioventricular septal defect Surgery zone of apposition Outcome Cleft Bileaflet procedure Left AV valve incompetence
| 1. Introduction |
|---|
|
|
|---|
There is still no consensus concerning the nature of the area between the left superior and inferior bridging leaflet of the common AV valve. It is classically described as a cleft: a source of regurgitation, which has to be closed, thus resulting in a two leaflet mitral valve. However, Rastelli [1] showed that suturing of a cleft never produced a normal mitral valve. In 1979, Piccoli [2] and in 1982, Studer [3] described the cleft as a true commissure and the left AV valve as a three-leaflet valve. On this basis Carpentier [4] promoted the concept of not closing the cleft, rendering the left AV valve essentially a three-leaflet valve. Recently, R.H. Anderson presented his concept of the common AV valve, in which the space between the facing surfaces of the left ventricular components of the bridging leaflets of the common AV valve is labelled zone of apposition (ZoA). Anderson enumerates important morphologic features that distinguish the zone of apposition from an isolated mitral cleft [5]. When the surgeon closes a cleft, he restores the anatomy of a normal anterior leaflet of the mitral valve. When he closes the zone of apposition, he does not produce a normal anterior leaflet of a mitral valve. Since the postoperative left AV valve in AVSD anomalies can never be a mitral valve, it would be misleading to use the term cleft for the space between the facing surfaces of the left ventricular components of the bridging leaflets of the common AV valve. This space, on the other hand, is not a commissure either. A commissure of the mitral valve is defined as a breach in the leaflet skirt supported by a fan-shaped chord atop its prominent papillary muscle. It is a functional division between segments of leaflet tissue guarding either an atrioventricular or ventriculo-arterial junction. The area between the two bridging leaflets fulfils none of these criteria. Therefore, we have adopted the terminology of zone of apposition to stress the point that a cleft, a commissure and the zone of apposition are three separate, clearly discernible entities.
Over time, more and more paediatric cardiac surgeons have adopted the concept of closing the zone of apposition. It now seems to prevail over the opinion that the commissure should be left untouched. Since 1981, both techniques have been used in our Department in a large number of cases by one surgeon. After studying our results with the trileaflet procedure, we changed our standard protocol for AVSD [6]. Since 1987, we progressively suture-closed the zone of apposition in all patients with balanced or left dominant AVSD. Today it seems timely to compare the outcome of patients treated according to the two different protocols in terms of survival, rate of reoperation and late function of the left AV valve.
| 2. Patients and methods |
|---|
|
|
|---|
One hundred and fifty nine patients fulfilled the inclusion criteria. Median age at operation was 11 months (range 1161 months), median weight 6.3 kg (range 2.642.7 kg). They fell into two groups. In group I patients (n=63), the ZoA was left open. In group II patients (n=96) it was closed either completely (n=87) or partially (n=9). Preoperative and perioperative data were analysed using Chi-square test or Fisher's exact test for univariate analysis, and MannWhitney test to compare several variables (GraphPad PrismTM software). The difference was considered statistically significant with a P-value of 0.05 or less.
Table 1 summarises our findings. The two groups were comparable with regard to age and weight at operation, male and female ratio, association of Down's syndrome, preoperative surgical procedures, major associated cardiac lesions, incidence of pulmonary hypertension, distribution of Rastelli AVSD types, number of ventricular imbalance cases and duration of cardio-pulmonary bypass. For diagnosis of pulmonary hypertension, pulmonary pressure had to be greater than 50% of systemic mean blood pressure either measured during catheterization or estimated by echocardiography. The two groups however differ in terms of size (P<0.0001), since we changed in 1987 from a trileaflet to a standard bileaflet repair. Moreover more patients in group II were operated on at the age of 712 months (P=0.04). Prior to operation, there were more cases of moderate and severe incompetence of the left AV valve in group II (P=0.004). The incompetence of the left AV valve was graded according to the echocardiographic findings when available, otherwise in conformity with the results of heart catheterization.
|
The follow-up interval for patients in group I was 12198 months (median 107), in group II it was 10126 months (median 61). The current evaluation for the survivors took place in January 1998. Five patients (three in group I, two in group II) from abroad were lost to follow-up.
Follow-up evaluation, consisting of clinical examination and echocardiography, took place either in our Department (n=70) or in the referring centres (n=56) on the basis of an accepted guideline. Echocardiographic examinations focussed on the function of the left AV valve. The degree of regurgitation was estimated by measuring the maximum width of the jet at the base and the size of the valve annulus from two orthogonal planes. If the width was less than 1/3 of the valve annulus, it was considered as mild, if less than 2/3, as moderate, and if 2/3 or more, as severe. KaplanMeier curves for actuarial survival and freedom from reoperation were calculated using the GraphPad PrismTM software. The log-rank test was used to assess the statistical differences between the two groups, with determination of P-value. Where appropriate, means are given with ±standard deviation and percentages with 95% confidence limits (CL) in brackets.
| 3. Results |
|---|
|
|
|---|
|
|
|
|
In group II, reintervention for severe left AV valve incompetence was necessary in two cases early after the initial repair on the 15th and 16th postoperative day, respectively. The suture approximating the ZoA was broken in one patient with several valve chordae ruptured. In the other patient, the anterior leaflet had dehisced from the patch and infarction of the left ventricle was suspected. Both patients died, the first late after valve repair, the second early after valve replacement.
3.3. Late left AV valve function
In 98 of the 126 surviving patients (78%, 95%-CL: 7085), including reoperation survivors, the left AV valve was competent or mildly incompetent (Table 4). On the whole, the left AV valve function was not significantly different (P=0.82) in both groups at follow-up evaluation. However, an increase in the degree of valve regurgitation since discharge from hospital was noted in 28% (11/39) in group I and in 9% (8/87) in group II (P=0.013).
|
| 4. Discussion |
|---|
|
|
|---|
Our study includes two different time frames. Until October 1987, we employed the trileaflet technique, leaving the ZoA untouched. We thereafter moved to the bileaflet repair, closing surgically the ZoA. During this transitional period of 14 months, some patients with no preopertive AV valve regurgitation still had the trileaflet repair. After 1988, all patientswhether they had or had not AV valve incompetence, had bileaflet correction with closure of the ZoA. This progressive change explains the 11 overlapping patients in group I and the 2 in group II (see Table 1: year of operation). We chose to include all those 13 transitional patients in the study, as they met selection criteria; their exclusion would have reduced the size of group I in particular, thereby decreasing the number of events and increasing the chances of statistical errors. The inclusion of our transitional patients hason the other handnot statistically altered outcome. When comparing the outcome of the 11 overlapping cases of group I with the remainders of their group in a statistical model, early (P=0.58) and late (P=0.46) mortality are not different and rate of reoperation for left AV valve incompetence is the same (P=1). As a consequence of the different time frames, the follow-up for patients with an open ZoA is longer. This difference was statistically neutralised by applying the log-rank test for comparison of actuarial survival rates and freedom from reoperation.
Preoperative left AV valve incompetence was found more frequently in group II. This fact may be attributed to a bias in methodology since there were fewer preoperative echocardiographic data in group I (15) than in group II (65). The more sensitive echocardiographic techniques that were available only after we had changed our standard protocol are likely to have increased the detection of left AV valve regurgitation. But since outcome in group II was not worse than in group I but, in fact, better, it seems safe to assume that this potential bias did not influence our results. Several authors [1416] failed to detect any relationship between the degree of preoperative and postoperative atrioventricular regurgitation.
Earlier or later date of operation seems to have had little influence on the results in this series. Although one would expect that increasing surgical experience over time and continual improvement of postoperative management favours group II patients, the 30-day mortality is not statistically different between the two groups. This absence of statistical difference might be due to the small number of events in relation to the size of the two groups.
If patients with double orifice or parachute left AV valve are taken into account, early mortality is significantly higher in group I [17]. This is due to the fact that in these patients an individualised approach is necessary where the zone of apposition cannot be closed completely in most of the cases because suture-closure would lead to a postoperative stenotic AV valve. Since the scope of this study was to evaluate the role of the ZoA in patients with otherwise normal left AVSD-AV valve morphology, we decided to exclude patients with left AV valve anomalies.
There were more patients aged 712 months in group II than in group I (univariate analysis). But since no difference exists between both groups with respect to the number of cases operated on during the first year of life and since cases of pulmonary hypertension were equally distributed, we presume that this fact is negligible. However, our policy in the last years has been to perform primary repair earlier in infancy.
To grade the regurgitation of the left AV valve at follow-up, we decided to use a semiquantitative colour Doppler method (measuring the maximum width of the jet and the size of the valve annulus in two orthogonal planes) which is a modification of the colour Doppler assessment described by Helmcke [18]. This method proved to be sufficiently sensitive and specific in the identification of left AV valve regurgitation and sufficiently accurate in estimating its severity. We are aware of the fact that there are several other echocardiographic methods to assess the severity of left AV valve incompetence with similar reliability.
4.2. Survival
Surgical repair of complete AVSD can be accomplished with acceptable early and late mortality. Early mortality in this series was 5.7% and corresponds well to the range of 216% that has been reported in the literature, [3,10,14,15,19,20]. Persistent high postoperative pulmonary arterial pressure is a well-known risk factor for early mortality [19]. Remarkably, four of the six patients in group I who died early died of persistent pulmonary hypertension whereas none of the three patients in group II who died early had pulmonary hypertension. Three of the four patients were older than 1 year at the time of repair. Our findings substantiate the widely held view that early intervention is a crucial factor for successful surgical repair and prevention of the development of irreversible pulmonary vascular changes.
Group I had a longer follow-up time than group II. To be able to make use of all the data available and to compare the resulting KaplanMaier curves, we applied the log-rank test. Although closure of the ZoA did not significantly reduce the rate of death in the first 30 postoperative days, it improved later survival considerably. This reduction in late mortality has been demonstrated for 9-years of follow-up of group II. It seems very likely that the improvement will be sustained, considering the stabilisation of the survival curves. Contrary to the less favourable results reported by others [3,14,21], the subset of patients with coexisting major cardiac anomalies did remarkably well in both groups of our series.
4.3. Reoperation
Regurgitation of the left AV valve is reported to lead to reoperation in 620% of patients after repair of complete AVSD [1416,22]. Other indications such as residual shunts and patch dehiscence are infrequent [10,15,23]. Our perioperative analysis showed in most patients that the incompetence was due to an unsutured or reopened ZoA, sometimes in combination with additional anomalies, and that there was no chordal support in this area. Similar observations were mentioned in various publications [3,15,24,25].
Several authors have shown that the failure to achieve a competent left AV valve at AVSD correction is the most important risk factor associated with mortality and reoperation [3,14,15,21,22]. This is confirmed by our study. We were able to demonstrate that there were more postoperative cases with an incompetent AV valve in group I and that their outcome was worse than for group II patients.
During reoperation, every effort was made to preserve the native valve. We rather accepted to retain a mild to moderate incompetent postoperative valve than to replace it. Koboda and Jonas [26] have indicated that valve replacement in patients with AVSD (complete or partial), either during primary repair or at reoperation, was associated with elevated early mortality (22%), and carried a high risk of creating complete heart block. We confirm their observation: four out of five of our patients who underwent valve replacement eventually died. The only complete heart block in this series occurred after valve replacement.
4.4. Left AV valve function at follow-up
Over the course of time, the function of the reconstructed left AV valve is not necessarily stable; a moderately incompetent valve after reconstruction may improve, an initially competent left AV valve may deteriorate. Usually, mild regurgitation is stable and severely compromised left AV valve function is rare [20]. In our series, in a majority of valves where the ZoA was left untouched, valve function deteriorated significantly more than in those with a closed ZoA and early after AVSD correction. The reintervention included closure of the ZoA whenever possible, which in turn protected valvular function in the survivors: hence the similarity of valve function in both groups at follow-up echocardiography.
The main goal of left AV valve reconstruction in patients with complete AVSD is to create a competent valve without creating valve stenosis. Stenosis of the left AV valve after surgical reconstruction was rare among patients with a closed ZoA. This could be explained by the fact that during reconstruction an aged-related minimal normal mitral valve diameter was used as a guide during ZoA closure to prevent valve stenosis.
| 5. Conclusion |
|---|
|
|
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. Suzuki, E. L. Bove, E. J. Devaney, T. Ishizaka, C. S. Goldberg, J. C. Hirsch, and R. G. Ohye Results of definitive repair of complete atrioventricular septal defect in neonates and infants. Ann. Thorac. Surg., August 1, 2008; 86(2): 596 - 602. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. L. Backer, R. D. Stewart, F. Bailliard, A. M. Kelle, C. L. Webb, and C. Mavroudis Complete Atrioventricular Canal: Comparison of Modified Single-Patch Technique With Two-Patch Technique Ann. Thorac. Surg., December 1, 2007; 84(6): 2038 - 2046. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Boening, J. Scheewe, K. Heine, J. Hedderich, D. Regensburger, H.-H. Kramer, and J. Cremer Long-term results after surgical correction of atrioventricular septal defects Eur. J. Cardiothorac. Surg., August 1, 2002; 22(2): 167 - 173. [Abstract] [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 |