- Email: [email protected]
Detachment of the septal tricuspid leaflet during transatrial closure of isolated ventricular septal defect
J THORAC CARDIOVASC SURG 82:773-778, 1981
Detachment of the septal tricuspid leaflet during transatrial closure of isolated ventricular septal defect In 27 (/8%) of the lSI patients who underwent transatrial closure of isolated ventricular septal defect (VSD) between 1966 and 1979, the tricuspid valve was partially detached in order to achieve better exposure. All 27 patients had defects of the membranous or paramembranous type situated behind the tricuspid septal leaflet. In some cases, tight chordae tendineae crossed over the defect and inserted in the edge of the VSD. A IS to 20 mm incision in the septal leaflet was usually needed to expose the defect sufficiently. There were two operative deaths among the 27 patients, both unrelated to the tricuspid incision. The remaining patients had uncomplicated postoperative courses. There were no long-term complications or instances of significant tricuspid valve incompetence, major residual shunt, or heart block at follow-up. Three patients, operated upon at the ages of 3, 3, and 6 yeatl5, respectively, had residual pulmonary hypertension. In one patient, who died 4 years postoperatively in a traffic accident, the tricuspid valve was intact and the previous incision could hardly be seen. It is concluded that detachment of the septal tricuspid leaflet is a safe procedure during transatrial closure of a VSD.
B. P. Frenckner, M.D.,* C. L. Olin, M.D.,* V. Bomfim, M.D.,* B. Bjarke, M.D.,** C. G. Wallgren, M.D.,*** and V. O. Bjork, M.D.,* Stockholm, Sweden
Closure of isolated ventricular septal defect (VSD) usually is performed through a right ventricular or a right atrial incision. The advantage of the right ventricular incision is generally a better surgical exposure. However, the ventriculotomy involves a higher risk of postoperative cardiac arrhythmias and impaired right ventricular performance. i, 2 Therefore, we always have favored the right atrial incision as our primary approach in patients with common types of VSD. If the VSD has not been adequately visualized through this incision, we have preferred to incise the septal leaflet of the
From the Thoracic Surgical Clinic and the Departments of Pediatric Cardiology, Karolinska Hospital and SI. Goran's Hospital, Stockholm, Sweden. Received for publication Feb. 12, 1981. Accepted for publication April 9, 1981. Address for reprints: V. O. Bjork, M.D., Thoracic Surgical Clinic, Karolinska Hospital, S-I04 01 Stockholm, Sweden. *Thoracic Surgical Clinic, Karolinska Hsopital. **Department of Pediatric Cardiology, SI. Gorarr's Hospital. ***Department of Pediatric Cardiology, Karolinska Hospital.
Fig. 1. Exposure of the ventricular septal defect through the right atrium after retraction of the tricuspid valve. The caval cannulas were introduced posteriorly in order to provide space for the atrial incision.
tricuspid valve rather than resort to a ventriculotomy to get better exposure.v" There are also cases in which such a detachment is necessary to avoid damage to chordae tendineae crossing over the defect and inserting in the edge of the VSD. This paper evaluates
0022-5223/81/110773+06$00.60/0 © 1981 The C. V. Mosby Co.
773
The Journal of
774
Thoracic and Cardiovascular
Frenckner et al.
Surgery
• Fig. 3. The patch was anchored to the septum with the aid of isolated mattress sutures buttressed with small pieces of cut Teflon tubes. Damage to the conduction system was avoided by placing the sutures well away from the edge of the defect both posteriorly and inferiorly.
Table I. Systolic PAP in 27 patients undergoing transatrial closure of VSD with partial detachment of tricuspid valve
l
PAP (mm Hg)
No. of patients
<30 30-60 >60
4 ( + 2 banded)
8 13
Legend: PAP. Systolic pulmonary artery pressure. VSD. Ventricular septal defect.
whether the septal tricuspid leaflet can be detached safely during transatrial closure of isolated VSD.
•
Fig. 2. Technique of tricuspid valve incision and repair during trans atrial closure of ventricular septal defect. S, A, and P, Septal, anterior, and posterior leaflets.
Patients
Between 1966 and 1979, 151 patients with isolated VSD were operated upon at the Thoracic Surgical Clinic, Karolinska Hospital, Stockholm, Sweden. A right atrial incision alone was used for closure of the defect in 142 cases (94%). The septal leaflet of the
Volume 82 Number 5
Transatrial closure of isolated VSD
November , 1981
775
Fig. 4. Autopsy spec imen from the patient who died in a traffic accident 4 years after operation. Th e force ps holds the septal leaflet of the tricuspid valve. The arrows indicate the ends of the incision where the sutures can hardly be seen. There is an incisio n (indica ted by the star) through the anterior part of the encapsulated ventricular septal defect patch .
Table II. Causes of death after operation Age at operation (yr)
PAP (mm Hg)
Time lapse af ter operation
Cause of death
3 5
90 110
1 day 1'12 mo
Hemolysis Sepsis (Candida albicans}
Legend: PAP. Systolic pulmonary anery pressure.
tricuspid valve was incised in 27 of these patients (19%) to achieve better exposure. An additional right ventricular incision had to be made in seven patients (5%) and a left ventricular incision in two patients ( I %). The age of the 27 patients who had a tricuspid valve incision ranged from 2 months to 45 years (median age 5 years). Two patients were less than I year old . Associated cardiovascular malformations were present in seven patients (26%) . Four had atrial septal defect (ASD), two persistent ductus arteriosus, one coarctation of the aorta, and one a congenital atrioventricular block. The pulmonary artery pressure was markedly elevated in the majority of the patients (Table I). Two patients with normal pulmonary artery pressure had previously undergone banding of the pulmonary artery. Operative technique
The heart was always exposed via a median sternotomy. The ascending aorta was cannulated with
either a soft plastic cannula or a Sarns aortic arch cannula . The caval veins were cannulated via the right atrium . Care was taken to introduce the cannula posteriorly in order to provide space for the incision anteriorly (Fig. I) . Cardiopulmonary bypass was initiated and perfusion hypothermia (25° C) induced . Deep hypothermia (20° C) with a brief period of circulatory arrest was used in two cases. The aorta was cross-clamped and the heart locally cooled with cold (4° C) Ringer 's solution. After 1977, cardioplegia with cold oxygenated blood was used. The right atrium was opened wide and the edges held back with stay sutures. The VSD was visualized by retracting the septal leaflet of the tricuspid valve with soft-edged retractors (Fig. I). If adequate exposure was not obtained , the septal leaflet was incised about I mm from the anulus (Fig. 2, A ). Usually a 15 to 20 mm incision was needed to visualize the defect completely (Fig . 2, B ) . Care was taken to avoid the conduction
776
The Journal of Thoracic and Cardiovascular Surgery
Frenckner et al.
Table III. Hemodynamic findings in nine of 16 patients who underwent postoperative cardiac catheterization No. of patients Severe residual pulmonary hypertension (PAP> 60 mm Hg) Minor residual shunt Insignificant tricuspid valve incompetence
3 3 3
Legend: PAP. Systolic pulmonary artery pressure.
system. After repair of the VSD, the incision was closed with a running 5-0 Prolene suture supplemented with a few mattress sutures of 5-0 Ti-Cron (Fig. 2, C). All patients had a VSD of the so-called membranous or paramembranous type. In three patients, the VSD was small enough to be closed by direct suture. In the others, it was closed with a Dacron patch with isolated mattress sutures used to anchor the patch to the septum. The sutures were buttressed with small pieces of Teflon tubes (Fig. 3). In the inferior aspect of the defect, the sutures were placed well away from the edge of the defect to avoid damage to the conduction system. In the posterior part of the defect, the sutures were placed in the base of the septal tricuspid leaflet, also with the intention of avoiding damage to the conduction system. Before the right atrial incision was closed, the whole ventricular septum was examined carefully to rule out the presence of additional defects. In two patients, additional muscular defects were found and repaired with buttressed sutures. After bypass, when the circulation had stabilized, right and left ventricular pressures were measured by direct puncture. Follow-up All 27 patients with detached tricuspid valves were carefully observed. The follow-up period ranged from 1 to 11 years (mean 5 years). Cardiac catheterization was performed in 16 patients, and in the majority a right ventriculogram was also obtained. In the remainder, a clinical examination was carried out with special attention focused on the function of the tricuspid valve. Results Mortality and complications. Of the total group of 151 patients who underwent transatrial VSD closure, six (4%) died. Of the 27 patients in whom the tricuspid valve was partially detached, two (7%) died. Both patients had severe pulmonary hypertension preoperatively, and the causes of death were hemolysis and Can-
dida albicans infection, respectively (Table II). The tricuspid valve was not implicated in either of the deaths. All remaining patients had uneventful postoperative courses and were discharged from the hospital in good condition. There were no late deaths related to heart disease (one patient died in a traffic accident). One patient, aged 6 years at operation, subsequently developed mitral and tricuspid insufficiency. At reoperation 3 years later, a severly malformed mitral valve was excised and replaced with a Bjork-Shiley prosthesis. The tricuspid valve proved to be diffusely dilated and was reduced with a De Vega annuloplasty. The septal tricuspid leaflet was normal and the previous suture could hardly be seen. The same was also true in the patient who died 4 years after operation in a traffic accident (Fig. 4). Follow-up. All 24 surviving patients were in good condition at follow-up. None had signs of tricuspid valve insufficiency. Two patients, aged 42 and 45 years at operation, were disturbed by episodes of supraventricular arrhythmia already present before operation. The remaining patients had no subjective complaints and were free from cardiac disability. Sixteen patients underwent postoperative cardiac catheterization (Table III). Three patients had residual pulmonary hypertension (pulmonary artery pressure > 60 mm Hg). They had been operated upon at the ages of 3,3, and 6 years, respectively, and the postoperative catheterization was performed 2, 8, and 10 years after operation. Three patients had minimal residual left-to-right shunts without hemodynamic significance. Right atrial pressure was normal in all patients. Angiocardiograms showed a small regurgitation of contrast medium from the right ventricle to the right atrium in three cases. The role of the rather stiff catheter introduced from the inferior vena cava was difficult to assess. When the contrast medium was injected into the inferior vena cava, however, no definite backflow was observed during systole. Thus the regurgitation lacked hemodynamic significance.
Discussion The main advantage of the right atrial approach for closure of VSD is that a ventriculotomy can be avoided. This is of special importance in patients with severe pulmonary hypertension, in whom a right ventricular incision may impair right ventricular function postoperatively. If the VSD cannot be closed through the atrial incision, this incision will serve as an aid in locating the defect and determining the exact site of the
Volume 82 Number5 November, 1981
Transatrial closure of isolated VSD
777
Table A. Responses to poll
Surgeon E. Arciniegas (Detroit, Mich.) H. Bender (Nashville, Tenn.)
A. Castaneda (Boston, Mass.) D. Cooley (Houston, Texas) G. Danielson (Rochester, Minn.) D. Doty (Iowa City, Iowa) G. Kaiser (Miami, Fla.) J. W. Kirklin (Birmingham, Ala.)
G. Lindesmith (Los Angeles, Calif.) J. Maim (New York, N. Y.) A. Starr (Portland, Ore.)
Percent of patients having isolated paramembranous VSD repaired through right atrium only
Percent of operations through right atriotomy only, in which incision of septal tricuspid leaflet was performed
0.8%
0%
'71-'81
0%
'73-April, '81
100%
0%
'73-'80
15%
5%
20 years
8%
2% (in early experience)
'65-'81
93%
0%
'72-'81
98% (a few prior to '73)
1%
'73-'81
76%
2% to 5% (estimated)
'67·'79
33%
40% to 50% (estimated)
'77·'80
80%
0%
Past 10 years
100%
0%
'78-'80
65% (almost none prior to '77)
Time interval from which these data were accumulated
Remarks
Prefers transatrial route for small children in whom hypothermia and circulatory arrest are used
Multivariate analysis showed no difference in survival whether by RA or RV approach
Only LV·RA shunt defects require tricuspid takedown In about 10% of cases, divides one or two chordae of septal leaflet and reattaches to patch with single 6-0 suture
Legend: VSD, Ventricular septal defect. RA, Right atrial. RV, Right ventricular. LV, left ventricular.
ventriculotomy. Another advantage of the right atrial incision is that a concomitant ASD (present in about 20% of the cases) can easily be closed. The follow-up of our 27 patients shows that an incision in the tricuspid valve can be made safely if the VSD cannot be adequately visualized through the atrium. None of our patients had evidence of hemodynamically significant tricuspid incompetence at followup. We also have used this incision in other forms of congenital heart disease where a VSD is present (double-outlet right ventricle, atrioventricular discordance, and so on), and to the best of our knowledge there have been no instances of dehiscence of the incision or significant incompetence of the valve. In our total series of patients with isolated VSD, it was possible to close the VSD through the atrium in 94%. In a similar series of patients also routinely operated upon through the atrium (but without incision in
the tricuspid valve), the corresponding figure was 72%.6 Hence, tricuspid valve detachment avoids the
need for ventriculotomy in many cases. We routinely have used isolated buttressed mattress sutures to anchor the patch to the septum. We believe this is the most efficient technique to close the VSD securely. Instead of using prefabricated Teflon patches, which often tend to twist the sutures, we have employed small pieces of Teflon tubes to buttress the sutures. They are easier to manage, particularly in infants and small children. There were few early and late complications among the 27 patients with incision of the tricuspid valve, Two operative deaths occurred. Both patients had a large VSDs with pulmonary hypertension, and neither of the deaths was related to the tricuspid incision. There were no instances of permanent heart block. Two patients, operated upon at the ages of 42 and 45 years, were
778
The Journal of Thoracic and Cardiovascular Surgery
Frenckner et al.
disturbed by attacks of supraventricular arrhythmia. This is a widely recognized problem in patients operated upon late in life. 7 Three patients, operated upon at the ages of 3, 3, and 6 years, had residual pulmonary hypertension in the absence of residual left-to-right shunting postoperatively. They were all operated upon in the beginning of the series, when the risk of progressive pulmonary vascular disease was not so well understood. We now advise closure of large defects with pulmonary hypertension before the age of I year. In conclusion, our experience shows that an incision in the tricuspid valve can safely be made during transatrial closure of YSD in order to improve exposure. None of our patients had evidence of hemodynamically significant tricuspid insufficiency at follow-up. The technique of incising the tricuspid valve is not regarded as a safe procedure by some surgeons, but we consider it reliable and recommend it as an alternative to a ventriculotomy during transatrial closure of isolated YSD.
REFERENCES Stirling GR, Stanley PH, Lillehei CW: The effects of cardiac bypass and ventriculotomy upon right ventricular function. Surg Forum 8:433-438, 1957 2 Kay JH, Anderson RM, Tolentino P, Dykstra P, Shapiro MJ, Meihaus JE, Magidson 0: The surgical repair of high pressure ventricular septal defect through the right atrium. Surgery 48:65-74, 1960 3 Bjork YO: The transatrial approach to ventricular septal defect. J THORAC CARDIOVASC SURC 47: 178-185, 1964
4 Bjork Vf): Surgical treatment of ventricular septal defect. Thorax 20: 278-284, 1965 5 Bjork vo. Bomfim v, Olin C: Transatrial closure of ventricular septal defect. Scand J Thorac Cardiovasc Surg 8:161-165, 1974 6 Lincoln C: Transatrial vs ventricular closure of isolated ventricular septal defect, Paediatric Cardiology 1977, Edinburgh, 1978, Churchill Livingstone, pp 155-162 7 Hallidie-Smith KA: Natural history and long-term followup of ventricular septal defect, Paediatric Cardiology 1977, Edinburgh, 1978, Churchill Livingstone, pp 169-175
Reviewers' comment Both official reviewers of this manuscript raised the question as to how often detachment of the septal tricuspid leaflet is required during transatrial closure of an isolated VSD of the paramembranous type. Since they themselves did not use the technique, they suggested that the readers should be made aware of this difference of opinion. On the one hand, the authors have shown that in their experience the technique seemed useful and also free of complications; yet the reviewers have suggested that it may be an unnecessary maneuver. The only definitive answer to the question would require additional comparative analysis from several centers of still larger numbers of operations. For now it seemed of potential interest to conduct a small poll of arbitrarily selected authorities active in the surgery of congenital heart disease to see what their practices might be with respect to this issue. While doing so, it seemed appealing also to determine how frequently these surgeons use the transatrial approach for closing an isolated paramembranous VSD. Eleven of the twelve surgeons contacted responded, and their responses are tabulated (see Table A on page 777).
Detachment of the septal tricuspid leaflet during transatrial closure of isolated ventricular septal defect In 27 (/8%) of the lSI patients who underwent transatrial closure of isolated ventricular septal defect (VSD) between 1966 and 1979, the tricuspid valve was partially detached in order to achieve better exposure. All 27 patients had defects of the membranous or paramembranous type situated behind the tricuspid septal leaflet. In some cases, tight chordae tendineae crossed over the defect and inserted in the edge of the VSD. A IS to 20 mm incision in the septal leaflet was usually needed to expose the defect sufficiently. There were two operative deaths among the 27 patients, both unrelated to the tricuspid incision. The remaining patients had uncomplicated postoperative courses. There were no long-term complications or instances of significant tricuspid valve incompetence, major residual shunt, or heart block at follow-up. Three patients, operated upon at the ages of 3, 3, and 6 yeatl5, respectively, had residual pulmonary hypertension. In one patient, who died 4 years postoperatively in a traffic accident, the tricuspid valve was intact and the previous incision could hardly be seen. It is concluded that detachment of the septal tricuspid leaflet is a safe procedure during transatrial closure of a VSD.
B. P. Frenckner, M.D.,* C. L. Olin, M.D.,* V. Bomfim, M.D.,* B. Bjarke, M.D.,** C. G. Wallgren, M.D.,*** and V. O. Bjork, M.D.,* Stockholm, Sweden
Closure of isolated ventricular septal defect (VSD) usually is performed through a right ventricular or a right atrial incision. The advantage of the right ventricular incision is generally a better surgical exposure. However, the ventriculotomy involves a higher risk of postoperative cardiac arrhythmias and impaired right ventricular performance. i, 2 Therefore, we always have favored the right atrial incision as our primary approach in patients with common types of VSD. If the VSD has not been adequately visualized through this incision, we have preferred to incise the septal leaflet of the
From the Thoracic Surgical Clinic and the Departments of Pediatric Cardiology, Karolinska Hospital and SI. Goran's Hospital, Stockholm, Sweden. Received for publication Feb. 12, 1981. Accepted for publication April 9, 1981. Address for reprints: V. O. Bjork, M.D., Thoracic Surgical Clinic, Karolinska Hospital, S-I04 01 Stockholm, Sweden. *Thoracic Surgical Clinic, Karolinska Hsopital. **Department of Pediatric Cardiology, SI. Gorarr's Hospital. ***Department of Pediatric Cardiology, Karolinska Hospital.
Fig. 1. Exposure of the ventricular septal defect through the right atrium after retraction of the tricuspid valve. The caval cannulas were introduced posteriorly in order to provide space for the atrial incision.
tricuspid valve rather than resort to a ventriculotomy to get better exposure.v" There are also cases in which such a detachment is necessary to avoid damage to chordae tendineae crossing over the defect and inserting in the edge of the VSD. This paper evaluates
0022-5223/81/110773+06$00.60/0 © 1981 The C. V. Mosby Co.
773
The Journal of
774
Thoracic and Cardiovascular
Frenckner et al.
Surgery
• Fig. 3. The patch was anchored to the septum with the aid of isolated mattress sutures buttressed with small pieces of cut Teflon tubes. Damage to the conduction system was avoided by placing the sutures well away from the edge of the defect both posteriorly and inferiorly.
Table I. Systolic PAP in 27 patients undergoing transatrial closure of VSD with partial detachment of tricuspid valve
l
PAP (mm Hg)
No. of patients
<30 30-60 >60
4 ( + 2 banded)
8 13
Legend: PAP. Systolic pulmonary artery pressure. VSD. Ventricular septal defect.
whether the septal tricuspid leaflet can be detached safely during transatrial closure of isolated VSD.
•
Fig. 2. Technique of tricuspid valve incision and repair during trans atrial closure of ventricular septal defect. S, A, and P, Septal, anterior, and posterior leaflets.
Patients
Between 1966 and 1979, 151 patients with isolated VSD were operated upon at the Thoracic Surgical Clinic, Karolinska Hospital, Stockholm, Sweden. A right atrial incision alone was used for closure of the defect in 142 cases (94%). The septal leaflet of the
Volume 82 Number 5
Transatrial closure of isolated VSD
November , 1981
775
Fig. 4. Autopsy spec imen from the patient who died in a traffic accident 4 years after operation. Th e force ps holds the septal leaflet of the tricuspid valve. The arrows indicate the ends of the incision where the sutures can hardly be seen. There is an incisio n (indica ted by the star) through the anterior part of the encapsulated ventricular septal defect patch .
Table II. Causes of death after operation Age at operation (yr)
PAP (mm Hg)
Time lapse af ter operation
Cause of death
3 5
90 110
1 day 1'12 mo
Hemolysis Sepsis (Candida albicans}
Legend: PAP. Systolic pulmonary anery pressure.
tricuspid valve was incised in 27 of these patients (19%) to achieve better exposure. An additional right ventricular incision had to be made in seven patients (5%) and a left ventricular incision in two patients ( I %). The age of the 27 patients who had a tricuspid valve incision ranged from 2 months to 45 years (median age 5 years). Two patients were less than I year old . Associated cardiovascular malformations were present in seven patients (26%) . Four had atrial septal defect (ASD), two persistent ductus arteriosus, one coarctation of the aorta, and one a congenital atrioventricular block. The pulmonary artery pressure was markedly elevated in the majority of the patients (Table I). Two patients with normal pulmonary artery pressure had previously undergone banding of the pulmonary artery. Operative technique
The heart was always exposed via a median sternotomy. The ascending aorta was cannulated with
either a soft plastic cannula or a Sarns aortic arch cannula . The caval veins were cannulated via the right atrium . Care was taken to introduce the cannula posteriorly in order to provide space for the incision anteriorly (Fig. I) . Cardiopulmonary bypass was initiated and perfusion hypothermia (25° C) induced . Deep hypothermia (20° C) with a brief period of circulatory arrest was used in two cases. The aorta was cross-clamped and the heart locally cooled with cold (4° C) Ringer 's solution. After 1977, cardioplegia with cold oxygenated blood was used. The right atrium was opened wide and the edges held back with stay sutures. The VSD was visualized by retracting the septal leaflet of the tricuspid valve with soft-edged retractors (Fig. I). If adequate exposure was not obtained , the septal leaflet was incised about I mm from the anulus (Fig. 2, A ). Usually a 15 to 20 mm incision was needed to visualize the defect completely (Fig . 2, B ) . Care was taken to avoid the conduction
776
The Journal of Thoracic and Cardiovascular Surgery
Frenckner et al.
Table III. Hemodynamic findings in nine of 16 patients who underwent postoperative cardiac catheterization No. of patients Severe residual pulmonary hypertension (PAP> 60 mm Hg) Minor residual shunt Insignificant tricuspid valve incompetence
3 3 3
Legend: PAP. Systolic pulmonary artery pressure.
system. After repair of the VSD, the incision was closed with a running 5-0 Prolene suture supplemented with a few mattress sutures of 5-0 Ti-Cron (Fig. 2, C). All patients had a VSD of the so-called membranous or paramembranous type. In three patients, the VSD was small enough to be closed by direct suture. In the others, it was closed with a Dacron patch with isolated mattress sutures used to anchor the patch to the septum. The sutures were buttressed with small pieces of Teflon tubes (Fig. 3). In the inferior aspect of the defect, the sutures were placed well away from the edge of the defect to avoid damage to the conduction system. In the posterior part of the defect, the sutures were placed in the base of the septal tricuspid leaflet, also with the intention of avoiding damage to the conduction system. Before the right atrial incision was closed, the whole ventricular septum was examined carefully to rule out the presence of additional defects. In two patients, additional muscular defects were found and repaired with buttressed sutures. After bypass, when the circulation had stabilized, right and left ventricular pressures were measured by direct puncture. Follow-up All 27 patients with detached tricuspid valves were carefully observed. The follow-up period ranged from 1 to 11 years (mean 5 years). Cardiac catheterization was performed in 16 patients, and in the majority a right ventriculogram was also obtained. In the remainder, a clinical examination was carried out with special attention focused on the function of the tricuspid valve. Results Mortality and complications. Of the total group of 151 patients who underwent transatrial VSD closure, six (4%) died. Of the 27 patients in whom the tricuspid valve was partially detached, two (7%) died. Both patients had severe pulmonary hypertension preoperatively, and the causes of death were hemolysis and Can-
dida albicans infection, respectively (Table II). The tricuspid valve was not implicated in either of the deaths. All remaining patients had uneventful postoperative courses and were discharged from the hospital in good condition. There were no late deaths related to heart disease (one patient died in a traffic accident). One patient, aged 6 years at operation, subsequently developed mitral and tricuspid insufficiency. At reoperation 3 years later, a severly malformed mitral valve was excised and replaced with a Bjork-Shiley prosthesis. The tricuspid valve proved to be diffusely dilated and was reduced with a De Vega annuloplasty. The septal tricuspid leaflet was normal and the previous suture could hardly be seen. The same was also true in the patient who died 4 years after operation in a traffic accident (Fig. 4). Follow-up. All 24 surviving patients were in good condition at follow-up. None had signs of tricuspid valve insufficiency. Two patients, aged 42 and 45 years at operation, were disturbed by episodes of supraventricular arrhythmia already present before operation. The remaining patients had no subjective complaints and were free from cardiac disability. Sixteen patients underwent postoperative cardiac catheterization (Table III). Three patients had residual pulmonary hypertension (pulmonary artery pressure > 60 mm Hg). They had been operated upon at the ages of 3,3, and 6 years, respectively, and the postoperative catheterization was performed 2, 8, and 10 years after operation. Three patients had minimal residual left-to-right shunts without hemodynamic significance. Right atrial pressure was normal in all patients. Angiocardiograms showed a small regurgitation of contrast medium from the right ventricle to the right atrium in three cases. The role of the rather stiff catheter introduced from the inferior vena cava was difficult to assess. When the contrast medium was injected into the inferior vena cava, however, no definite backflow was observed during systole. Thus the regurgitation lacked hemodynamic significance.
Discussion The main advantage of the right atrial approach for closure of VSD is that a ventriculotomy can be avoided. This is of special importance in patients with severe pulmonary hypertension, in whom a right ventricular incision may impair right ventricular function postoperatively. If the VSD cannot be closed through the atrial incision, this incision will serve as an aid in locating the defect and determining the exact site of the
Volume 82 Number5 November, 1981
Transatrial closure of isolated VSD
777
Table A. Responses to poll
Surgeon E. Arciniegas (Detroit, Mich.) H. Bender (Nashville, Tenn.)
A. Castaneda (Boston, Mass.) D. Cooley (Houston, Texas) G. Danielson (Rochester, Minn.) D. Doty (Iowa City, Iowa) G. Kaiser (Miami, Fla.) J. W. Kirklin (Birmingham, Ala.)
G. Lindesmith (Los Angeles, Calif.) J. Maim (New York, N. Y.) A. Starr (Portland, Ore.)
Percent of patients having isolated paramembranous VSD repaired through right atrium only
Percent of operations through right atriotomy only, in which incision of septal tricuspid leaflet was performed
0.8%
0%
'71-'81
0%
'73-April, '81
100%
0%
'73-'80
15%
5%
20 years
8%
2% (in early experience)
'65-'81
93%
0%
'72-'81
98% (a few prior to '73)
1%
'73-'81
76%
2% to 5% (estimated)
'67·'79
33%
40% to 50% (estimated)
'77·'80
80%
0%
Past 10 years
100%
0%
'78-'80
65% (almost none prior to '77)
Time interval from which these data were accumulated
Remarks
Prefers transatrial route for small children in whom hypothermia and circulatory arrest are used
Multivariate analysis showed no difference in survival whether by RA or RV approach
Only LV·RA shunt defects require tricuspid takedown In about 10% of cases, divides one or two chordae of septal leaflet and reattaches to patch with single 6-0 suture
Legend: VSD, Ventricular septal defect. RA, Right atrial. RV, Right ventricular. LV, left ventricular.
ventriculotomy. Another advantage of the right atrial incision is that a concomitant ASD (present in about 20% of the cases) can easily be closed. The follow-up of our 27 patients shows that an incision in the tricuspid valve can be made safely if the VSD cannot be adequately visualized through the atrium. None of our patients had evidence of hemodynamically significant tricuspid incompetence at followup. We also have used this incision in other forms of congenital heart disease where a VSD is present (double-outlet right ventricle, atrioventricular discordance, and so on), and to the best of our knowledge there have been no instances of dehiscence of the incision or significant incompetence of the valve. In our total series of patients with isolated VSD, it was possible to close the VSD through the atrium in 94%. In a similar series of patients also routinely operated upon through the atrium (but without incision in
the tricuspid valve), the corresponding figure was 72%.6 Hence, tricuspid valve detachment avoids the
need for ventriculotomy in many cases. We routinely have used isolated buttressed mattress sutures to anchor the patch to the septum. We believe this is the most efficient technique to close the VSD securely. Instead of using prefabricated Teflon patches, which often tend to twist the sutures, we have employed small pieces of Teflon tubes to buttress the sutures. They are easier to manage, particularly in infants and small children. There were few early and late complications among the 27 patients with incision of the tricuspid valve, Two operative deaths occurred. Both patients had a large VSDs with pulmonary hypertension, and neither of the deaths was related to the tricuspid incision. There were no instances of permanent heart block. Two patients, operated upon at the ages of 42 and 45 years, were
778
The Journal of Thoracic and Cardiovascular Surgery
Frenckner et al.
disturbed by attacks of supraventricular arrhythmia. This is a widely recognized problem in patients operated upon late in life. 7 Three patients, operated upon at the ages of 3, 3, and 6 years, had residual pulmonary hypertension in the absence of residual left-to-right shunting postoperatively. They were all operated upon in the beginning of the series, when the risk of progressive pulmonary vascular disease was not so well understood. We now advise closure of large defects with pulmonary hypertension before the age of I year. In conclusion, our experience shows that an incision in the tricuspid valve can safely be made during transatrial closure of YSD in order to improve exposure. None of our patients had evidence of hemodynamically significant tricuspid insufficiency at follow-up. The technique of incising the tricuspid valve is not regarded as a safe procedure by some surgeons, but we consider it reliable and recommend it as an alternative to a ventriculotomy during transatrial closure of isolated YSD.
REFERENCES Stirling GR, Stanley PH, Lillehei CW: The effects of cardiac bypass and ventriculotomy upon right ventricular function. Surg Forum 8:433-438, 1957 2 Kay JH, Anderson RM, Tolentino P, Dykstra P, Shapiro MJ, Meihaus JE, Magidson 0: The surgical repair of high pressure ventricular septal defect through the right atrium. Surgery 48:65-74, 1960 3 Bjork YO: The transatrial approach to ventricular septal defect. J THORAC CARDIOVASC SURC 47: 178-185, 1964
4 Bjork Vf): Surgical treatment of ventricular septal defect. Thorax 20: 278-284, 1965 5 Bjork vo. Bomfim v, Olin C: Transatrial closure of ventricular septal defect. Scand J Thorac Cardiovasc Surg 8:161-165, 1974 6 Lincoln C: Transatrial vs ventricular closure of isolated ventricular septal defect, Paediatric Cardiology 1977, Edinburgh, 1978, Churchill Livingstone, pp 155-162 7 Hallidie-Smith KA: Natural history and long-term followup of ventricular septal defect, Paediatric Cardiology 1977, Edinburgh, 1978, Churchill Livingstone, pp 169-175
Reviewers' comment Both official reviewers of this manuscript raised the question as to how often detachment of the septal tricuspid leaflet is required during transatrial closure of an isolated VSD of the paramembranous type. Since they themselves did not use the technique, they suggested that the readers should be made aware of this difference of opinion. On the one hand, the authors have shown that in their experience the technique seemed useful and also free of complications; yet the reviewers have suggested that it may be an unnecessary maneuver. The only definitive answer to the question would require additional comparative analysis from several centers of still larger numbers of operations. For now it seemed of potential interest to conduct a small poll of arbitrarily selected authorities active in the surgery of congenital heart disease to see what their practices might be with respect to this issue. While doing so, it seemed appealing also to determine how frequently these surgeons use the transatrial approach for closing an isolated paramembranous VSD. Eleven of the twelve surgeons contacted responded, and their responses are tabulated (see Table A on page 777).