- Email: [email protected]
Electrical instability in patients undergoing surgery for atrioventricular septal defect
International Elsevier
CARD10
Journal of Cardiology,
15
30 (1991) IS-21
01179
Electrical
instability in patients undergoing for atrioventricular septal defect
surgery
Luciano Daliento ‘, Giulio Rizzoli 2, Maria Cristina Marchiori t, Gianfranco Buja t, Ornella Milanesi 3, Serafina Valente I, Giovanni Stellin 2 and Alessandro Mazzucco 2 Departments
of ’ Cardiology,
’ Cardiac Surgery and ’ Pediatrics,
University of Padua, Padua, Italy
(Received 1 June 1990; revision accepted 9 August 1990)
Daliento L, Rizzoli G, Marchiori MC, Buja G, Milanesi 0, Valente S, Stellin G, Mazzucco A. Electrical instability in patients undergoing surgery for atrioventricular septal defect. Int J Cardiol 1991;30:15-21. Postoperative 24-hour Holter monitoring was performed in 106 patients with atrioventrieuhu septal defect in order to identify the incidence of atrial and ventricular arrhythmias. Of the patients, 72 had separate atrioventricular orifices, in&ding 13 with a small ventricular component to the defect, and 34 patients had a common atrioventricular orifice. Two groups of abnormal patients were found. First, patients with good electrical stability characterized by isolated atrial (9 patients) and ventricular (25 patients) extrasystoles fatling in classes I and II of Town. Second, patients with marked electrical instability characterized in one patient by repetitive atrial extrasystoles, in another by atria1 flutter, in 2 by polymorphic ventricular extrasystole and in 8 by couplets or triplets. Electrical instability in individual patients was then compared, by means of logistic regression analysis, with operative, surgical and postoperative variables. There was no incidence of sudden death in our series. After surgical repair, ventricular arrhythmias were more frequent than atrial arrhythmias (33% vs. 10%) and were unrelated to the type of atrioventricular septal defect. Cardiac electrical instability after operation was significantly related to larger operative body size, higher postoperative end diastolic diameter of the right ventricle, larger size of the ventricular septal defect, coexistence of postoperative right bundle branch block and left anterior hemiblock. Conversely, the risk of arrhythmias was reduced by more recent operative data and by greater shortening fraction of the left ventricle.
Key words: Atrioventricular
septal defect; Holter monitoring, 24-hour; Electrical cardiac instability
Correspondence to: L. Daliento, Cattedra di Cardiologia, Pohclinico Universitario, Via Giustiniani, 2, 35128 Padova, Italy. Supported by the National Council for Research, Target Project F.A.T.M.A. “Fattori di malattia patologia matemo-infantile”, Rome, Italy. 0167-5273/91/$03.50
Introduction
Arrhythmias are commonly present after surgical repair of congenital heart diseases [l]. They can be secondary to structural alterations of
0 1991 Elsevier Science Publishers B.V. (Biomedical Division)
16
cardiac chambers due to different types of overload, or to a direct damage of atria1 and ventricular structures, intrinsic to the specific type of surgery. In the past, the most important problem in patients operated for atrioventricular septal defect was damage to the conduction tissues with resultant perioperative atrioventricular block [2]. Knowledge of the location of the atrioventricular node and bundle has dramatically decreased the incidence of this problem in recent years. In contrast, the incidence of electrocardiographic modifications and of tachyarrhythmias in patients operated for atrioventricular septal defect with an undamaged specialized conduction system remains unknown. The aim of this study, therefore, was to verify the incidence, type and causes of arrhythmias in patients undergoing surgical repair of atrioventricular septal defect utilizing ambulatory (Holter) electrocardiography, and to investigate the relationship between arrhythmias and several pre-, intra- and post-operative incremental risk factors.
Definition of Terms Atrioventricular septal defect is defined as a congenital heart defect characterized by absence of the central portion of the atrioventricular septum, also commonly referred to as “atrioventricular canal” [3]. According to Anderson et al. [4] we distinguished two forms of atrioventricular septal defect: (a) atrioventricular septal defect with separated atrioventricular valves, in the presence of two orifices, due to the fusion of bridging leaflets with (an) additional tongue of tissue. Usually a shunt occurs only at atria1 level, when the bridging leaflets are firmly attached to the ventricular septum. If an interventricular communication remains underneath the confluent bridging leaflets, a shunt can be found at both atria1 and ventricular levels. (b) Atrioventricular septal defect with common atrioventricular valve, when the bridging leaflets are unfused. A common orifice is associ-
ated with both atria1 and ventricular communications.
Materials and Methods We studied 106 patients (49 males and 57 females) who underwent surgical repair of atrioventricular septal defect between 1978-1988 at the Institute of Cardiovascular Surgery of the University of Padua. Of the patients, 72 had separate right and left atrioventricular valves, including 13 with a small ventricular component to the defect, while 34 had a common atrioventricular valve. The median interval for follow-up was 40 months (range 6-125). The mean age at operation was 103 * 124 months (range 3.7 months - 63 years) and the mean weight was 24 t- 20 kilograms (range 4-83). The mean pulmonary arteriolar resistance was 3 f 2.4 U/m2 (range l-14 U/m2). Most of our patients (60%) were repaired with the use of two separate patches and valvar incompetence was managed according to a three-leaflet concept for the left component of the effectively common valve [5]; 35% were repaired with a single patch and two-leaflet reconstruction of the left atrioventricular valve [6]. In 5% of our patients, the correction consisted simply of an isolated closure of the ostium primum defect. Cold potassium cardioplegia was utilized in 90% of the cases. The average aortic cross-clamping time was 58 k 21 minutes (range 27-140). Down’s syndrome was present in 41%. Post-operative long-term follow-up included chest X-ray, basic and 24-hour electrocardiogram recording (Holter), M-mode, cross-sectional and Doppler echocardiography. We graded the severity of supra-ventricular arrhythmias according to the results of Holter monitoring. The first grade included patients with isolated atria1 extrasystoles. The second grade included cases with repetitive atrial extrasystoles and patients with atria1 flutter. The severity of ventricular arrhythmias was graded according to Lown’s classification [7]. In order to summarize these findings we defined a state of electrical instability with two levels. A small loss of stability was defined as isolated atria1 extrasystoles and for ventricular extrasystoles in grades 1 and 2 of Lown’s classification. Electrical instability of high grade was considered present in
17
patients with repetitive atria1 arrhythmias, atria1 flutter and/or ventricular extrasystoles falling in Lown’s grade 3. Patients with coexisting but different types and grades of arrhythmias, were classified according to the most severe arrhythmia. Atrioventricular conduction disturbances and electrocardiographic modifications after surgery were studied by standard electrocardiography. Electrical stability of the heart in individual patients was compared by means of logistic regression analysis with pre-operative, surgical and post-operative variables. Pre-operative variables included: weight, body surface area (m2), age at operation (months), functional class (New York Heart Association) (l-5) presence of Down’s syndrome (y/n), grade of left atrioventricular incompetence (l-4) pulmonary resistances (U/m’), presence and size of the ventricular component of the defect (O-2), right or left ventricular dominance (y/n), Rastelli’s classification (type A, B, C = y/n) and associated cardiac lesions (y/n). The ventricular component was judged large in 23% of cases, small in 22%; while, in 55% of patients, a ventricular communication was not present. A patent arterial duct was found in 8 cases, a dual orifice in the left atrioventricular valve in 6. Parachute-type left atrioventricular valve, valvular pulmonary stenosis, and tetralogy of Fallot were present in one case each. The surgical variables were: the year of operation (1978 to 1988), the technique used for repair (single patch = y/n, two patch = y/n), the time of aortic cross clamping (minutes), and period of cardio-circulatory arrest (minutes). Post-operative variables were: functional class (New York Heart Association) at last clinical control (l-4), grade of residual incompetence of the left atrioventricular valve by echo-Doppler (l-4), reoperation (y/n), cardio-thoracic ratio, echocardiographic end diastolic diameter of the left atrium (mm), end diastolic diameter of left ventricle (mm), shortening fraction of left ventricle, ejection fraction of the left ventricle, end diastolic volume of the right ventricle (ml), grade of atrioventricular block (l-3), right bundle branch block (y/n), right bundle branch block and left anterior hemiblock (y/n), presence of permanent endo-
cavitary pacemaker (y/n), and duration of follow up (months). All echocardiographic parameters were normalized for body surface area. Cluster analysis was performed to identify correlations between these variables and to clarify the meaning of any variable eventually introduced in the stepwise logistic analysis. 24-hour Holter monitoring was available before surgery in only 30 patients. In all other cases preoperative arrhythmias were diagnosed from the standard electrocardiogram. This variable, therefore, was not considered in our multivariate logistic analysis.
Results In Table 1, we have summarized the mean values of roentgengraphic and echocardiographic measurements. They show an overall satisfactory functional status of the average patient after surgical correction. The type of tachyarrhythmias detected by Holter monitoring in the course of follow-up is reported in Table 2. They were classified according to their origin (supra-ventricular and ventricular) and severity, and to the anatomical type of atrioventricular septal defect (with two valves or a common valve). Atria1 arrhythmias: one episode of atria1 flutter
TABLE
1
Mean values and standard deviations echocardiographic measurements.
of
radiological
and
Parameter
Two Av valves
Common AV valve
Total
P
C/T
0.51+0.04 34.6Ok8.80 44.7Ok8.70 39.7Ok8.80 64.1Ozt7.60 22.50 + 6.20 1.70+0.70
0.53k 26.4Ok 32.6Ort 41.7Ok 67.2Ok 20.10 rt 1.70+
0.52kO.04 32.30+8.00 40.90+9.30 40.40+7.30 65.10+7.00 22.00 f 5.70 1.75+0.80
0.01 0.0001 0.0001 NS 0.06 NS NS
EDDLA EDDLV SFLV EFLV EDDRV LAWI
0.03 8.40 9.60 7.00 8.40 10.4 0.60
C/T = cardio-thoracic ratio; EDDLA = end-diastolic diameter left atrium; EDDLV = end-diastolic diameter left ventricle; SFLV = shortening fraction left ventricle; EFLV = ejection fraction left ventricle; EDDRV = end-diastolic diameter right ventricle; LAWI = left atrioventricular valve incompetence (grade l-4).
18 TABLE 2 Postoperative tachyarrhythmias detected by Holter monitoring. Supraventricular arrhythmias Grade
Two AV valves
Ventricular arrhythmias Common AV valve
Total
Grade Lown Lown Lown Lown Lown
1
5
4
9
2
2
0
2
7 (9.7%)
4 (11.7%)
Two AV valves 1 2 3 4 5
11 (10.4%)
was recorded in a 16-year-old boy, 7 years after repair of atrioventricular septal defect associated with partial anomalous venous connexion of the upper right pulmonary vein to the superior caval vein. At follow-up, he presented with cardiomegaly and moderate incompetence of the left atrioventricular valve. He showed, however, a satisfactory performance during a treadmill test. A second patient presented with recurrent repetitive atria1 extrasystoles. He had undergone repair at the age of 35. Despite good valvar function, he showed significative reduction of left ventricular function at echocardiogram, an increased right ventricular volume and performed a poor treadmill test. Ventricular arrhythmias: these were significantly more frequent than supra-ventricular arrhythmias (33% vs. 10%) independent of the type of atrioventricular septal defect. Association
14 2 1 5 2 24 (33.3%)
Common AV valve
Total
9 0 1 1 0
23 2 2 6 2
11 (32.3%)
35 (33%)
of supra-ventricular and ventricular arrhythmias was found in 9 cases (6 cases with separate atrioventricular valves and 3 with common atrioventricular valves). Among the 30 patients who were studied by preoperative ambulatory monitoring, we found atria1 arrhythmias of grade 1 in 1 case and ventricular arrhythmias in 3 cases (2 cases in Lown’s class 1; 1 case in Lown’s class 5). Table 3 shows the relationship between preoperative and postoperative arrhythmias. The results of multivariate stepwise logistic regression analysis are summarized in Table 4. Conduction disturbances: before surgery, 27 patients showed atrioventricular block of first degree (P-Q interval > 0.20 msec), 22 an isolated incom-
TABLE 4 Results of multivariate stepwise logistic regression analysis.
TABLE
Variable
3
Pre-operative and post-operative ( ) tachyarrhythmias patients with pre-operative Holter monitoring.
in 30
Supraventricular arrhythmias
Ventricular arrhythmias
Grade
Two AV valves
Common AV valve
Grade
Two AV valves
Common AV valve
1
l(1)
--
Lown 1
2 (0)
0 (5)
2
Lown2 Lown3-Lown4 Lown5
(3) - -(l) l-
-- -
Weight EDDRV Size VSD RBBB + LAH Surgery (time) SFLV
Beta
Standard error
chiP square
0.00243282 0.09204016 1.35146328
0.00113806 0.03584754 0.58389875
4.57 6.59 5.36
0.0325 0.0102 0.0206
0.119 0.160 0.136
1.31441065
0.51049380
6.63
0.0100
0.160
0.08141638 0.03575315
2.85 2.53
0.0916 0.1117
- 0.069 -0.054
-0.13735320 -0.95687228
r
EDDRV = end-diastolic diameter right ventricle; VSD = ventricular septal defect; RBBB = right bundle branch block; LAH = left anterior hemiblock; SFLV = shortening fraction left ventricle.
19
plete right bundle branch block (QRS < 0.10 msec), and 3 an isolated complete right bundle branch block (QRS > 0.10 msec). The incomplete right bundle branch block seen in 64 patients was associated with left anterior hemiblock. Intraventricular conduction delay was not present in 17 cases but the morphology of the QRS complex reflected right or biventricular hypertrophy. Four patients experienced complete atrioventricular block at the time of surgery and eventually required implantation of a definitive pacemaker. Of them, 3 had a common atrioventricular orifice of Rastelli type C and one had separate atrioventricular valves with a small ventricular septal defect. In addition, 2 patients with a common atrioventricular valve had also persistence of the left superior caval vein. The patient with separate atrioventricular valves and small ventricular septal defect had a double orifice of the left atrioventricular valve. Seven patients showed electrocardiographic evidence of progression of their right bundle branch block after surgery. Two of them had had isolated right bundle branch block prior to repair, while 5 had incomplete right bundle branch block associated with left anterior hemiblock. Ninety-five patients showed a reduction of the duration of their QRS complex. The mean reduction in left axis deviation was about 20” due to reduction of right ventricular volume.
Discussion A high incidence of chronic arrhythmias has previously been reported after repair of atria1 septal defects within the oval fossa (“secundum” defect), particularly in adult patients [8]. These are represented mainly by ectopic atria1 beats, or junctional escape rhythms, sick sinus syndrome, atria1 flutter or fibrillation. Ventricular arrhythmias were rare [9,10]. Preoperative chronic overload of the cavities of the right heart, the type and site of cannulation for cardio-pulmonary by-pass and the location of the septal defect have been identified as the main causes of post-operative arrhythmias in these cases [lo-121. Patients operated for atrioventricular septal de-
fect have also been studied in the past to detect perioperative and late disturbances of the atrioventricular conducting tissues, and some reports have shown that late arrhythmias were essentially similar to those detected after repair of those defects occurring within the oval fossa [13-151. According to Ongley et al. [16], 4% of perioperative survivors experienced significant late postoperative arrhythmias, including atria1 flutter, supraventricular tachycardia and premature ventricular complexes. Studies using long-term ambulatory monitoring suggested, however, that the incidence of those arrhythmias could be higher and could be related to function of the left atrioventricular valve
1171.
In our series, the ventricular function and the function of atrioventricular valves after correction was, on the average, satisfactory. Late postoperative ambulatory electrocardiography showed a low incidence of supra-ventricular arrhythmias in all types of atrioventricular septal defect, after repair, most of which were without clinical relevance. the incidence of ventricular In contrast, arrhythmias was surprisingly higher (33%) than expected. These arrhythmias were usually benign, as the majority of patients (77%) had ventricular arrhythmias included in the first three grades of Lown’s classification. Arrhythmias in grades 4 and 5 were recorded in only 8 patients. All of them but one, who had the same preoperative arrhythmia, showed poor ventricular function after correction. Pre-operative 24-hour monitoring was not selectively applied according to clinical indications, but was performed randomly. By comparison with postoperative findings, we noticed that the preoperative ventricular arrhythmias of the first class of Lown disappeared after surgery in two patients, both with atrioventricular septal defect and separate valvar orifices, who underwent operation at age 1 and 9 years, respectively. They showed a normal ventricular function after the procedure and mild left atrioventricular valve incompetence without cardiomegaly. The last patient, with preoperative ventricular arrhythmias in grade 5 of Lown, showed the persistence of severe arrhythmias in Lown’s class 4 at the latest followup. He had a septal defect with separate atrio-
20
ventricular valves, moderate incompetence of the left atrioventricular valve and interruption of the inferior caval vein. Postoperative echocardiography showed normal volumes and dynamics of the left ventricle, mild incompetence of both atrioventricular valves and severe enlargement of the right ventricular cavity. Five patients with common atrioventricular valves and normal preoperative Holter monitoring developed ventricular arrhythmias after surgery of the first class of Lown. Therefore, from our data, there is no evidence of correlation between preoperative and postoperative arrhythmias. This cannot, however, be excluded because preoperative Holter was performed in only 30 patients. The large number of patients lacking this datum justifies the exclusion of this variable from our multivariate analysis. With this limitation, post surgical cardiac electrical instability was significantly related to larger size at the time of operation (weight), increased right ventricular volumes, larger dimension of the ventricular component of the septal defect and the association of right bundle branch block with left anterior hemiblock. Conversely, the risk of arrhythmias decreased if operation was performed more recently and if the shortening fraction of the left ventricle was adequate. Post-surgical ventricular arrhythmias are usually more frequent in patients with congenital heart disease receiving a ventriculotomy. In contrast, our analysis shows that, in patients undergoing surgery for repair of atrioventricular septal defect through the right atrium, it correlates exclusively to the functional status of the cardiac cavities secondary to pre-existent or residual overload. This opinion is supported by the relationship of postoperative arrhythmias with increased right ventricular diameters measured by echocardiography, and the postoperative association with right bundle branch block and left anterior hemiblock. Indeed, it has been shown by Zevallos and coworkers [18] that atrioventricular septal defect, absence of regression of right bundle branch block and stability of left axial deviation, is an electrocardiographic sign of a persistently dilated right ventricular chamber. The significance of a large ventricular component of the defect, compared to absence of a
ventricular component or to a restrictive ventricular component is due to hemodynarnic consequences of ventricular overload. A later age at operation which, in this analysis, is substituted and expressed by the weight at operation, significantly increases the likelihood of arrhythmias. This is not surprising, because prolonged persistence of hemodynamic abnormalities is a well known cause of late postoperative arrhythmias [19]. The significance of the date of surgery in the multivariate analysis can be due to better operative and postoperative care which is related to the “learning curve”. On the other hand, it can be a substitute for a longer follow-up with a time-related onset of arrhythmias. Unlike actuarial methods, which analyze the onset of events in a time-related frame, logistic analysis does not permit solution of this question. Actuarial methods were not used because the exact time of onset of arrhythmias is unknown. We cannot exclude the possibility that arrhythmias were already present immediately after discharge from hospital, since they were revealed by a scheduled 2Chour Holter monitoring in all patients, except in those rare cases in whom they were clinically evident. Left ventricular function is another important variable. Its importance is confirmed by the low risk of arrhythmias in patients with high values of the shortening fraction as judged echocardiographically. In conclusion, in patients without any ventricular component, or with only a restrictive ventricular component, and with normal shortening fraction of the left ventricle and end diastolic diameter of the right ventricle, we may anticipate postoperative electrical instability as revealed by 24hour electrocardiography in 2-18% of cases according to the age at operation. This expectation rises to 70-9056 in patients with a large ventricular component, postoperative right bundle branch block associated with left anterior hemiblock, and severely depressed left ventricular function and an increased end diastolic diameter of the right ventricle (Fig. 1). Arrhythmias in these patients were not lifethreatening and no sudden deaths were observed in our series. Nevertheless, they can cause some
21 100
_,_._.-.-.-.-.-.-.*,_._ _.*.-’
t
_._____._._.-----’
90
1 7011
TB
00
_*--*--
I
=:
). .20
*.d’
A
_...
CT *.x. ,O
_,_._.-.-.-. ___
0
--_________
__----_---
10
_________----:
1
-
240
I
Expected
340
postoperative
540
440
electrical
640
740
,
a40
instability
Fig. 1. Group A: Patients with any ventricular component, or with a restrictive ventricular component to the defect with normal shortening fraction of the left ventricle and end diastolic diameter of the left ventricle. Group B: Patients with a large ventricular component, postoperative right bundle branch block with left anterior hemiblock and severely depressed left ventricular function and increased diastolic diameter of the right ventricle.
discomfort, particularly atria1 flutter and supraventricular tachycardia. Such arrhythmias can be utilized as an index of the functional status of the cardiac cavities. 24-hour electrocardiographic recording is the most effective way of revealing the electrical instability of these patients.
11
12
13
14
15
16
References 17 Vetter VL, Horowitz L. Electrophysiologic residua and sequelae of surgery for congenital heart defect. In: Engle MA, Perloff JK, eds. Congenital heart disease after surgery: benefits, residua, sequelae. New York: York Medical Books, 1983;262-305. Fryda RJ, Kaplan S. Helmsworth JA. Postoperative complete heart block in ctuldren. Br Heart J 1971;33:456-462. Rastelli GC, Kirklin JW, Titus JL. Anatomic observations on complete form of persistent common atrioventricular canal with special reference to atrioventricular valves. Mayo Clin Proc 1966;41:296-308. Anderson RH, Ho SY, Macartney FJ, Becker AE. The anatomy of atrioventncular septal defects. In: Jimenez MQ,
18
19
Martinez MA, eds. Pediatric cardiology: atriaventricular septal defects. Madrid: Ediciones Norma, 1988;132-150. Carpentier A. Surgical anatomy and management of the mitral components of atrioventricular canal defects. In: Anderson RH, Shinebourne A, eds. Pediatric cardiology 1977. London: Churchill Livingstone. 1978;477-490. Mair DD, McGoon DC. Surgical correction of atrioventricular during the first year of life. Am J Cardiol 1977;40:6669. Lown B, Wolf M. Approaches to sudden death from coronary heart disease. Circulation 1971;44:130-142. Senly WC, Farmer JC, Yong WG, Brown IW. Atria1 dysrhythmia and atria1 secundum defects. J Thorac Cardiovasc Surg 1969;57:245-250. Brandegurg RO Jr, Holmes D Jr. Brandegurg RO, McGoon DC. Clinical follow-up study of paroxysmal supraventricular tachyarrhythmias after operative repair of a secundum type atria1 septal defect in adults. Am J Cardiol 1983;51:273-276. Karpawich PP, Antillon JR, Coppola PR, Agarwal KC. Pre- and postoperative electrophysiologic assessment of children with secundum atria1 septal defect. Am J Cardiol 1985;55:519-521. Bink-Boelkens MTE, Meuzelaar KJ, Eygelaar A. Arrhythmias after repair of secundum atria1 septal defect: the influence of surgical modihcation. Am Heart J 1988; 115:629-633. Bink-Boelkens MTE, Velvis H, Van der Heide H, Eygelaar A, Hardjowijono RA. Dysrhythmias after atrial surgery in children. Am Heart J 1983;106:125-130. Sasaki R, Theilen EO, January LE. Ehrenhaft JL. Cardiac arrhythmias associated with the repair of atria1 and ventricular septal defects. Circulation 1958;18:909-915. Chen SC, Arcilla RA, Moulder PV, Cassels DE. Postoperative conduction disturbances in aerial septal defect. Am J Cardiol 1968;22:636-644. Linde LM, Goldbert SJ, Siegel S. The natural history of arrhythmias following septal defect repair. J Thorac Cardiovasc Surg 1964;48:303-309. Ongley PA, Ponypanich B, Spangler J, Feldt RH. The electrocardiogram in atrioventricular canal. In: Feldt RH, ed. Atrioventricular canal defects. Philadelphia: WB Sanders, 1976;51-75. Sommerville J, Ravel-Chion R, Van der Cumusen T, Presbitero T. Atrioventricular defect natural and unnatural history. In: Godman MJ ed. Pediatric cardiology, Vol. 4. London: Churchill Livingstone, 1981;404. Zevallos JC, Daliento L. Car&n G, Andriolo L. Da Ruos F, Caporale C. Estudio ECG-VCG pre e post-operatorio de la correction de defeto de septum atrioventricular. Arch 1st Cardiol Mex 1988;58:121-126. Bharati S, Lev M. The myocardlum, the conduction system, and generale sequelae after surgery for congenital heart disease. In: Engle MA, Perloff JK, eds. Congenital heart disease after surgery: benefits, reesidue, sequelae. New York: York Medical Books. 1983;247-260.
CARD10
Journal of Cardiology,
15
30 (1991) IS-21
01179
Electrical
instability in patients undergoing for atrioventricular septal defect
surgery
Luciano Daliento ‘, Giulio Rizzoli 2, Maria Cristina Marchiori t, Gianfranco Buja t, Ornella Milanesi 3, Serafina Valente I, Giovanni Stellin 2 and Alessandro Mazzucco 2 Departments
of ’ Cardiology,
’ Cardiac Surgery and ’ Pediatrics,
University of Padua, Padua, Italy
(Received 1 June 1990; revision accepted 9 August 1990)
Daliento L, Rizzoli G, Marchiori MC, Buja G, Milanesi 0, Valente S, Stellin G, Mazzucco A. Electrical instability in patients undergoing surgery for atrioventricular septal defect. Int J Cardiol 1991;30:15-21. Postoperative 24-hour Holter monitoring was performed in 106 patients with atrioventrieuhu septal defect in order to identify the incidence of atrial and ventricular arrhythmias. Of the patients, 72 had separate atrioventricular orifices, in&ding 13 with a small ventricular component to the defect, and 34 patients had a common atrioventricular orifice. Two groups of abnormal patients were found. First, patients with good electrical stability characterized by isolated atrial (9 patients) and ventricular (25 patients) extrasystoles fatling in classes I and II of Town. Second, patients with marked electrical instability characterized in one patient by repetitive atrial extrasystoles, in another by atria1 flutter, in 2 by polymorphic ventricular extrasystole and in 8 by couplets or triplets. Electrical instability in individual patients was then compared, by means of logistic regression analysis, with operative, surgical and postoperative variables. There was no incidence of sudden death in our series. After surgical repair, ventricular arrhythmias were more frequent than atrial arrhythmias (33% vs. 10%) and were unrelated to the type of atrioventricular septal defect. Cardiac electrical instability after operation was significantly related to larger operative body size, higher postoperative end diastolic diameter of the right ventricle, larger size of the ventricular septal defect, coexistence of postoperative right bundle branch block and left anterior hemiblock. Conversely, the risk of arrhythmias was reduced by more recent operative data and by greater shortening fraction of the left ventricle.
Key words: Atrioventricular
septal defect; Holter monitoring, 24-hour; Electrical cardiac instability
Correspondence to: L. Daliento, Cattedra di Cardiologia, Pohclinico Universitario, Via Giustiniani, 2, 35128 Padova, Italy. Supported by the National Council for Research, Target Project F.A.T.M.A. “Fattori di malattia patologia matemo-infantile”, Rome, Italy. 0167-5273/91/$03.50
Introduction
Arrhythmias are commonly present after surgical repair of congenital heart diseases [l]. They can be secondary to structural alterations of
0 1991 Elsevier Science Publishers B.V. (Biomedical Division)
16
cardiac chambers due to different types of overload, or to a direct damage of atria1 and ventricular structures, intrinsic to the specific type of surgery. In the past, the most important problem in patients operated for atrioventricular septal defect was damage to the conduction tissues with resultant perioperative atrioventricular block [2]. Knowledge of the location of the atrioventricular node and bundle has dramatically decreased the incidence of this problem in recent years. In contrast, the incidence of electrocardiographic modifications and of tachyarrhythmias in patients operated for atrioventricular septal defect with an undamaged specialized conduction system remains unknown. The aim of this study, therefore, was to verify the incidence, type and causes of arrhythmias in patients undergoing surgical repair of atrioventricular septal defect utilizing ambulatory (Holter) electrocardiography, and to investigate the relationship between arrhythmias and several pre-, intra- and post-operative incremental risk factors.
Definition of Terms Atrioventricular septal defect is defined as a congenital heart defect characterized by absence of the central portion of the atrioventricular septum, also commonly referred to as “atrioventricular canal” [3]. According to Anderson et al. [4] we distinguished two forms of atrioventricular septal defect: (a) atrioventricular septal defect with separated atrioventricular valves, in the presence of two orifices, due to the fusion of bridging leaflets with (an) additional tongue of tissue. Usually a shunt occurs only at atria1 level, when the bridging leaflets are firmly attached to the ventricular septum. If an interventricular communication remains underneath the confluent bridging leaflets, a shunt can be found at both atria1 and ventricular levels. (b) Atrioventricular septal defect with common atrioventricular valve, when the bridging leaflets are unfused. A common orifice is associ-
ated with both atria1 and ventricular communications.
Materials and Methods We studied 106 patients (49 males and 57 females) who underwent surgical repair of atrioventricular septal defect between 1978-1988 at the Institute of Cardiovascular Surgery of the University of Padua. Of the patients, 72 had separate right and left atrioventricular valves, including 13 with a small ventricular component to the defect, while 34 had a common atrioventricular valve. The median interval for follow-up was 40 months (range 6-125). The mean age at operation was 103 * 124 months (range 3.7 months - 63 years) and the mean weight was 24 t- 20 kilograms (range 4-83). The mean pulmonary arteriolar resistance was 3 f 2.4 U/m2 (range l-14 U/m2). Most of our patients (60%) were repaired with the use of two separate patches and valvar incompetence was managed according to a three-leaflet concept for the left component of the effectively common valve [5]; 35% were repaired with a single patch and two-leaflet reconstruction of the left atrioventricular valve [6]. In 5% of our patients, the correction consisted simply of an isolated closure of the ostium primum defect. Cold potassium cardioplegia was utilized in 90% of the cases. The average aortic cross-clamping time was 58 k 21 minutes (range 27-140). Down’s syndrome was present in 41%. Post-operative long-term follow-up included chest X-ray, basic and 24-hour electrocardiogram recording (Holter), M-mode, cross-sectional and Doppler echocardiography. We graded the severity of supra-ventricular arrhythmias according to the results of Holter monitoring. The first grade included patients with isolated atria1 extrasystoles. The second grade included cases with repetitive atrial extrasystoles and patients with atria1 flutter. The severity of ventricular arrhythmias was graded according to Lown’s classification [7]. In order to summarize these findings we defined a state of electrical instability with two levels. A small loss of stability was defined as isolated atria1 extrasystoles and for ventricular extrasystoles in grades 1 and 2 of Lown’s classification. Electrical instability of high grade was considered present in
17
patients with repetitive atria1 arrhythmias, atria1 flutter and/or ventricular extrasystoles falling in Lown’s grade 3. Patients with coexisting but different types and grades of arrhythmias, were classified according to the most severe arrhythmia. Atrioventricular conduction disturbances and electrocardiographic modifications after surgery were studied by standard electrocardiography. Electrical stability of the heart in individual patients was compared by means of logistic regression analysis with pre-operative, surgical and post-operative variables. Pre-operative variables included: weight, body surface area (m2), age at operation (months), functional class (New York Heart Association) (l-5) presence of Down’s syndrome (y/n), grade of left atrioventricular incompetence (l-4) pulmonary resistances (U/m’), presence and size of the ventricular component of the defect (O-2), right or left ventricular dominance (y/n), Rastelli’s classification (type A, B, C = y/n) and associated cardiac lesions (y/n). The ventricular component was judged large in 23% of cases, small in 22%; while, in 55% of patients, a ventricular communication was not present. A patent arterial duct was found in 8 cases, a dual orifice in the left atrioventricular valve in 6. Parachute-type left atrioventricular valve, valvular pulmonary stenosis, and tetralogy of Fallot were present in one case each. The surgical variables were: the year of operation (1978 to 1988), the technique used for repair (single patch = y/n, two patch = y/n), the time of aortic cross clamping (minutes), and period of cardio-circulatory arrest (minutes). Post-operative variables were: functional class (New York Heart Association) at last clinical control (l-4), grade of residual incompetence of the left atrioventricular valve by echo-Doppler (l-4), reoperation (y/n), cardio-thoracic ratio, echocardiographic end diastolic diameter of the left atrium (mm), end diastolic diameter of left ventricle (mm), shortening fraction of left ventricle, ejection fraction of the left ventricle, end diastolic volume of the right ventricle (ml), grade of atrioventricular block (l-3), right bundle branch block (y/n), right bundle branch block and left anterior hemiblock (y/n), presence of permanent endo-
cavitary pacemaker (y/n), and duration of follow up (months). All echocardiographic parameters were normalized for body surface area. Cluster analysis was performed to identify correlations between these variables and to clarify the meaning of any variable eventually introduced in the stepwise logistic analysis. 24-hour Holter monitoring was available before surgery in only 30 patients. In all other cases preoperative arrhythmias were diagnosed from the standard electrocardiogram. This variable, therefore, was not considered in our multivariate logistic analysis.
Results In Table 1, we have summarized the mean values of roentgengraphic and echocardiographic measurements. They show an overall satisfactory functional status of the average patient after surgical correction. The type of tachyarrhythmias detected by Holter monitoring in the course of follow-up is reported in Table 2. They were classified according to their origin (supra-ventricular and ventricular) and severity, and to the anatomical type of atrioventricular septal defect (with two valves or a common valve). Atria1 arrhythmias: one episode of atria1 flutter
TABLE
1
Mean values and standard deviations echocardiographic measurements.
of
radiological
and
Parameter
Two Av valves
Common AV valve
Total
P
C/T
0.51+0.04 34.6Ok8.80 44.7Ok8.70 39.7Ok8.80 64.1Ozt7.60 22.50 + 6.20 1.70+0.70
0.53k 26.4Ok 32.6Ort 41.7Ok 67.2Ok 20.10 rt 1.70+
0.52kO.04 32.30+8.00 40.90+9.30 40.40+7.30 65.10+7.00 22.00 f 5.70 1.75+0.80
0.01 0.0001 0.0001 NS 0.06 NS NS
EDDLA EDDLV SFLV EFLV EDDRV LAWI
0.03 8.40 9.60 7.00 8.40 10.4 0.60
C/T = cardio-thoracic ratio; EDDLA = end-diastolic diameter left atrium; EDDLV = end-diastolic diameter left ventricle; SFLV = shortening fraction left ventricle; EFLV = ejection fraction left ventricle; EDDRV = end-diastolic diameter right ventricle; LAWI = left atrioventricular valve incompetence (grade l-4).
18 TABLE 2 Postoperative tachyarrhythmias detected by Holter monitoring. Supraventricular arrhythmias Grade
Two AV valves
Ventricular arrhythmias Common AV valve
Total
Grade Lown Lown Lown Lown Lown
1
5
4
9
2
2
0
2
7 (9.7%)
4 (11.7%)
Two AV valves 1 2 3 4 5
11 (10.4%)
was recorded in a 16-year-old boy, 7 years after repair of atrioventricular septal defect associated with partial anomalous venous connexion of the upper right pulmonary vein to the superior caval vein. At follow-up, he presented with cardiomegaly and moderate incompetence of the left atrioventricular valve. He showed, however, a satisfactory performance during a treadmill test. A second patient presented with recurrent repetitive atria1 extrasystoles. He had undergone repair at the age of 35. Despite good valvar function, he showed significative reduction of left ventricular function at echocardiogram, an increased right ventricular volume and performed a poor treadmill test. Ventricular arrhythmias: these were significantly more frequent than supra-ventricular arrhythmias (33% vs. 10%) independent of the type of atrioventricular septal defect. Association
14 2 1 5 2 24 (33.3%)
Common AV valve
Total
9 0 1 1 0
23 2 2 6 2
11 (32.3%)
35 (33%)
of supra-ventricular and ventricular arrhythmias was found in 9 cases (6 cases with separate atrioventricular valves and 3 with common atrioventricular valves). Among the 30 patients who were studied by preoperative ambulatory monitoring, we found atria1 arrhythmias of grade 1 in 1 case and ventricular arrhythmias in 3 cases (2 cases in Lown’s class 1; 1 case in Lown’s class 5). Table 3 shows the relationship between preoperative and postoperative arrhythmias. The results of multivariate stepwise logistic regression analysis are summarized in Table 4. Conduction disturbances: before surgery, 27 patients showed atrioventricular block of first degree (P-Q interval > 0.20 msec), 22 an isolated incom-
TABLE 4 Results of multivariate stepwise logistic regression analysis.
TABLE
Variable
3
Pre-operative and post-operative ( ) tachyarrhythmias patients with pre-operative Holter monitoring.
in 30
Supraventricular arrhythmias
Ventricular arrhythmias
Grade
Two AV valves
Common AV valve
Grade
Two AV valves
Common AV valve
1
l(1)
--
Lown 1
2 (0)
0 (5)
2
Lown2 Lown3-Lown4 Lown5
(3) - -(l) l-
-- -
Weight EDDRV Size VSD RBBB + LAH Surgery (time) SFLV
Beta
Standard error
chiP square
0.00243282 0.09204016 1.35146328
0.00113806 0.03584754 0.58389875
4.57 6.59 5.36
0.0325 0.0102 0.0206
0.119 0.160 0.136
1.31441065
0.51049380
6.63
0.0100
0.160
0.08141638 0.03575315
2.85 2.53
0.0916 0.1117
- 0.069 -0.054
-0.13735320 -0.95687228
r
EDDRV = end-diastolic diameter right ventricle; VSD = ventricular septal defect; RBBB = right bundle branch block; LAH = left anterior hemiblock; SFLV = shortening fraction left ventricle.
19
plete right bundle branch block (QRS < 0.10 msec), and 3 an isolated complete right bundle branch block (QRS > 0.10 msec). The incomplete right bundle branch block seen in 64 patients was associated with left anterior hemiblock. Intraventricular conduction delay was not present in 17 cases but the morphology of the QRS complex reflected right or biventricular hypertrophy. Four patients experienced complete atrioventricular block at the time of surgery and eventually required implantation of a definitive pacemaker. Of them, 3 had a common atrioventricular orifice of Rastelli type C and one had separate atrioventricular valves with a small ventricular septal defect. In addition, 2 patients with a common atrioventricular valve had also persistence of the left superior caval vein. The patient with separate atrioventricular valves and small ventricular septal defect had a double orifice of the left atrioventricular valve. Seven patients showed electrocardiographic evidence of progression of their right bundle branch block after surgery. Two of them had had isolated right bundle branch block prior to repair, while 5 had incomplete right bundle branch block associated with left anterior hemiblock. Ninety-five patients showed a reduction of the duration of their QRS complex. The mean reduction in left axis deviation was about 20” due to reduction of right ventricular volume.
Discussion A high incidence of chronic arrhythmias has previously been reported after repair of atria1 septal defects within the oval fossa (“secundum” defect), particularly in adult patients [8]. These are represented mainly by ectopic atria1 beats, or junctional escape rhythms, sick sinus syndrome, atria1 flutter or fibrillation. Ventricular arrhythmias were rare [9,10]. Preoperative chronic overload of the cavities of the right heart, the type and site of cannulation for cardio-pulmonary by-pass and the location of the septal defect have been identified as the main causes of post-operative arrhythmias in these cases [lo-121. Patients operated for atrioventricular septal de-
fect have also been studied in the past to detect perioperative and late disturbances of the atrioventricular conducting tissues, and some reports have shown that late arrhythmias were essentially similar to those detected after repair of those defects occurring within the oval fossa [13-151. According to Ongley et al. [16], 4% of perioperative survivors experienced significant late postoperative arrhythmias, including atria1 flutter, supraventricular tachycardia and premature ventricular complexes. Studies using long-term ambulatory monitoring suggested, however, that the incidence of those arrhythmias could be higher and could be related to function of the left atrioventricular valve
1171.
In our series, the ventricular function and the function of atrioventricular valves after correction was, on the average, satisfactory. Late postoperative ambulatory electrocardiography showed a low incidence of supra-ventricular arrhythmias in all types of atrioventricular septal defect, after repair, most of which were without clinical relevance. the incidence of ventricular In contrast, arrhythmias was surprisingly higher (33%) than expected. These arrhythmias were usually benign, as the majority of patients (77%) had ventricular arrhythmias included in the first three grades of Lown’s classification. Arrhythmias in grades 4 and 5 were recorded in only 8 patients. All of them but one, who had the same preoperative arrhythmia, showed poor ventricular function after correction. Pre-operative 24-hour monitoring was not selectively applied according to clinical indications, but was performed randomly. By comparison with postoperative findings, we noticed that the preoperative ventricular arrhythmias of the first class of Lown disappeared after surgery in two patients, both with atrioventricular septal defect and separate valvar orifices, who underwent operation at age 1 and 9 years, respectively. They showed a normal ventricular function after the procedure and mild left atrioventricular valve incompetence without cardiomegaly. The last patient, with preoperative ventricular arrhythmias in grade 5 of Lown, showed the persistence of severe arrhythmias in Lown’s class 4 at the latest followup. He had a septal defect with separate atrio-
20
ventricular valves, moderate incompetence of the left atrioventricular valve and interruption of the inferior caval vein. Postoperative echocardiography showed normal volumes and dynamics of the left ventricle, mild incompetence of both atrioventricular valves and severe enlargement of the right ventricular cavity. Five patients with common atrioventricular valves and normal preoperative Holter monitoring developed ventricular arrhythmias after surgery of the first class of Lown. Therefore, from our data, there is no evidence of correlation between preoperative and postoperative arrhythmias. This cannot, however, be excluded because preoperative Holter was performed in only 30 patients. The large number of patients lacking this datum justifies the exclusion of this variable from our multivariate analysis. With this limitation, post surgical cardiac electrical instability was significantly related to larger size at the time of operation (weight), increased right ventricular volumes, larger dimension of the ventricular component of the septal defect and the association of right bundle branch block with left anterior hemiblock. Conversely, the risk of arrhythmias decreased if operation was performed more recently and if the shortening fraction of the left ventricle was adequate. Post-surgical ventricular arrhythmias are usually more frequent in patients with congenital heart disease receiving a ventriculotomy. In contrast, our analysis shows that, in patients undergoing surgery for repair of atrioventricular septal defect through the right atrium, it correlates exclusively to the functional status of the cardiac cavities secondary to pre-existent or residual overload. This opinion is supported by the relationship of postoperative arrhythmias with increased right ventricular diameters measured by echocardiography, and the postoperative association with right bundle branch block and left anterior hemiblock. Indeed, it has been shown by Zevallos and coworkers [18] that atrioventricular septal defect, absence of regression of right bundle branch block and stability of left axial deviation, is an electrocardiographic sign of a persistently dilated right ventricular chamber. The significance of a large ventricular component of the defect, compared to absence of a
ventricular component or to a restrictive ventricular component is due to hemodynarnic consequences of ventricular overload. A later age at operation which, in this analysis, is substituted and expressed by the weight at operation, significantly increases the likelihood of arrhythmias. This is not surprising, because prolonged persistence of hemodynamic abnormalities is a well known cause of late postoperative arrhythmias [19]. The significance of the date of surgery in the multivariate analysis can be due to better operative and postoperative care which is related to the “learning curve”. On the other hand, it can be a substitute for a longer follow-up with a time-related onset of arrhythmias. Unlike actuarial methods, which analyze the onset of events in a time-related frame, logistic analysis does not permit solution of this question. Actuarial methods were not used because the exact time of onset of arrhythmias is unknown. We cannot exclude the possibility that arrhythmias were already present immediately after discharge from hospital, since they were revealed by a scheduled 2Chour Holter monitoring in all patients, except in those rare cases in whom they were clinically evident. Left ventricular function is another important variable. Its importance is confirmed by the low risk of arrhythmias in patients with high values of the shortening fraction as judged echocardiographically. In conclusion, in patients without any ventricular component, or with only a restrictive ventricular component, and with normal shortening fraction of the left ventricle and end diastolic diameter of the right ventricle, we may anticipate postoperative electrical instability as revealed by 24hour electrocardiography in 2-18% of cases according to the age at operation. This expectation rises to 70-9056 in patients with a large ventricular component, postoperative right bundle branch block associated with left anterior hemiblock, and severely depressed left ventricular function and an increased end diastolic diameter of the right ventricle (Fig. 1). Arrhythmias in these patients were not lifethreatening and no sudden deaths were observed in our series. Nevertheless, they can cause some
21 100
_,_._.-.-.-.-.-.-.*,_._ _.*.-’
t
_._____._._.-----’
90
1 7011
TB
00
_*--*--
I
=:
). .20
*.d’
A
_...
CT *.x. ,O
_,_._.-.-.-. ___
0
--_________
__----_---
10
_________----:
1
-
240
I
Expected
340
postoperative
540
440
electrical
640
740
,
a40
instability
Fig. 1. Group A: Patients with any ventricular component, or with a restrictive ventricular component to the defect with normal shortening fraction of the left ventricle and end diastolic diameter of the left ventricle. Group B: Patients with a large ventricular component, postoperative right bundle branch block with left anterior hemiblock and severely depressed left ventricular function and increased diastolic diameter of the right ventricle.
discomfort, particularly atria1 flutter and supraventricular tachycardia. Such arrhythmias can be utilized as an index of the functional status of the cardiac cavities. 24-hour electrocardiographic recording is the most effective way of revealing the electrical instability of these patients.
11
12
13
14
15
16
References 17 Vetter VL, Horowitz L. Electrophysiologic residua and sequelae of surgery for congenital heart defect. In: Engle MA, Perloff JK, eds. Congenital heart disease after surgery: benefits, residua, sequelae. New York: York Medical Books, 1983;262-305. Fryda RJ, Kaplan S. Helmsworth JA. Postoperative complete heart block in ctuldren. Br Heart J 1971;33:456-462. Rastelli GC, Kirklin JW, Titus JL. Anatomic observations on complete form of persistent common atrioventricular canal with special reference to atrioventricular valves. Mayo Clin Proc 1966;41:296-308. Anderson RH, Ho SY, Macartney FJ, Becker AE. The anatomy of atrioventncular septal defects. In: Jimenez MQ,
18
19
Martinez MA, eds. Pediatric cardiology: atriaventricular septal defects. Madrid: Ediciones Norma, 1988;132-150. Carpentier A. Surgical anatomy and management of the mitral components of atrioventricular canal defects. In: Anderson RH, Shinebourne A, eds. Pediatric cardiology 1977. London: Churchill Livingstone. 1978;477-490. Mair DD, McGoon DC. Surgical correction of atrioventricular during the first year of life. Am J Cardiol 1977;40:6669. Lown B, Wolf M. Approaches to sudden death from coronary heart disease. Circulation 1971;44:130-142. Senly WC, Farmer JC, Yong WG, Brown IW. Atria1 dysrhythmia and atria1 secundum defects. J Thorac Cardiovasc Surg 1969;57:245-250. Brandegurg RO Jr, Holmes D Jr. Brandegurg RO, McGoon DC. Clinical follow-up study of paroxysmal supraventricular tachyarrhythmias after operative repair of a secundum type atria1 septal defect in adults. Am J Cardiol 1983;51:273-276. Karpawich PP, Antillon JR, Coppola PR, Agarwal KC. Pre- and postoperative electrophysiologic assessment of children with secundum atria1 septal defect. Am J Cardiol 1985;55:519-521. Bink-Boelkens MTE, Meuzelaar KJ, Eygelaar A. Arrhythmias after repair of secundum atria1 septal defect: the influence of surgical modihcation. Am Heart J 1988; 115:629-633. Bink-Boelkens MTE, Velvis H, Van der Heide H, Eygelaar A, Hardjowijono RA. Dysrhythmias after atrial surgery in children. Am Heart J 1983;106:125-130. Sasaki R, Theilen EO, January LE. Ehrenhaft JL. Cardiac arrhythmias associated with the repair of atria1 and ventricular septal defects. Circulation 1958;18:909-915. Chen SC, Arcilla RA, Moulder PV, Cassels DE. Postoperative conduction disturbances in aerial septal defect. Am J Cardiol 1968;22:636-644. Linde LM, Goldbert SJ, Siegel S. The natural history of arrhythmias following septal defect repair. J Thorac Cardiovasc Surg 1964;48:303-309. Ongley PA, Ponypanich B, Spangler J, Feldt RH. The electrocardiogram in atrioventricular canal. In: Feldt RH, ed. Atrioventricular canal defects. Philadelphia: WB Sanders, 1976;51-75. Sommerville J, Ravel-Chion R, Van der Cumusen T, Presbitero T. Atrioventricular defect natural and unnatural history. In: Godman MJ ed. Pediatric cardiology, Vol. 4. London: Churchill Livingstone, 1981;404. Zevallos JC, Daliento L. Car&n G, Andriolo L. Da Ruos F, Caporale C. Estudio ECG-VCG pre e post-operatorio de la correction de defeto de septum atrioventricular. Arch 1st Cardiol Mex 1988;58:121-126. Bharati S, Lev M. The myocardlum, the conduction system, and generale sequelae after surgery for congenital heart disease. In: Engle MA, Perloff JK, eds. Congenital heart disease after surgery: benefits, reesidue, sequelae. New York: York Medical Books. 1983;247-260.