- Email: [email protected]
Fifteen Years' Experience with the Aortic Homograft: The Conduit of Choice for Right Ventricular Outflow Tract Reconstruction
Fifteen Years’ Experience with the Aortic Homografk The Conduit of Choice for Right Ventricular Oufflow Tract Reconstruction P. H. Kay, F.R.C.S., and D. N. Ross, F.R.C.S, ABSTRACT Ninety-seven patients with pulmonary atresia underwent right ventricular outflow tract reconstruction using a homograft conduit. There were 46 hospital deaths (47%). Hospital mortality was significantly related to irreversible pulmonary hypertension (p < 0.001) and thoracotomy for ligation of bronchial collaterals f p < 0.01). The actuarial survival was 37 k 7% at 10 years. Sixteen patients undergoing recatheterization at a mean of 6 years had a mean transconduit gradient of 24 t 15 mm Hg. Obstructed conduits (i.e., with a gradient of greater than 50 mm Hg) were replaced in 3 patients, corresponding to 13 k 8% at 10 years. In each instance, the obstruction was due to neointimal hyperplasia in the Dacron tube rather than calcification of the homograft valve. The fresh, antibiotic-sterilized aortic homograft is the conduit of choice for right ventricular outflow tract reconstruction. The valve itself appears more resistant to calcification than its xenograft counterpart, and the absence of Dacron removes the problem of fibrinous peel obstructing the conduit. We now construct a tube of autologous pericardium to increase the length of the conduit and avoid complementary thoracotomy for ligation of bronchial collaterals.
The use of the aortic homograft for right ventricular outflow tract reconstruction was pioneered in 1966 by Ross and Somerville [l]. Initially the conduit was used for patients with pulmonary atresia, but subsequently the criteria were extended to include more complex anomalies [2, 31. However, early calcification leading to obstruction of the irradiated homografts caused several centers to abandon their use [4, 51. The alternative Dacron tube grafts containing a preserved xenograft aortic valve were introduced in 1973 [5] and have gained wide popularity since then [6-81. However, these conduits also suffer from late obstruction [9, 101. During the last 18 years we have continued to use fresh, antibiotic-sterilized aortic homografts for right ventricular outflow reconstruction [ 11-13]. Accordingly, we have analyzed our experience, focusing on the incidence of late conduit dysfunction.
From the National Heart Hospital, London, England Presented at the Twenty-first Annual Meeting of The society of Thoracic Surgeons, Phoenix, AZ,Jan 21-23, 1985. Address reprint requests to h4r. Ross, National Heart Hospital, Westmoreland St, London WlM BBA, England.
360
Material and Methods Between 1966 and 1984, 97 patients with pulmonary atresia underwent right ventricular outflow tract reconstruction using a homograft conduit. There were 47 male (48%)and 50 female patients, with ages ranging from 2 to 42 years (median, 11 years). Sixty-eight patients had undergone a systemicpulmonary shunt (22 Blalock-Taussig, 43 Waterston, and 3 Potts) as an initial procedure. Fifteen patients had undergone a second Blalock-Taussig shunt prior to correction.
Homograft Preservation It is our policy to use a fresh homograft whenever possible [13]. In this series, 84 patients received a fresh aortic homograft and 9, a frozen aortic homograft. The last 4 patients seen received a fresh pulmonary homograft. The fresh homografts are preserved in a nutrient medium to enhance tissue viability [14]. This medium is sterilized using an antibiotic mixture [15] and a fungicide [16]. After ten days if the valve has not been used in its fresh state, glycerol is added to the solution to a concentration of 15%. The valve is then frozen in the vapor phase of a liquid nitrogen refrigerator where it is stored. Operative Details Blalock-Taussig and Potts shunts were controlled before the onset of cardiopulmonary bypass. In those patients with a Waterston shunt, the right pulmonary artery was repaired with a patch of autologous pericardium. Standard techniques of cardiopulmonary bypass at moderate hypothermia (25°C) were used. A vertical incision was made in the right ventricle, thereby exposing the ventricular septa1 defect (VSD). In 86 patients the VSD was closed with a patch of knitted Dacron velour. The VSD was left open electively in 11 patients with hypoplastic pulmonary arteries and severe pulmonary hypertension [17]. A suitable homograft (19 to 23 mm) was then chosen and tailored to the existing anatomy. In 82 patients, the length of the conduit was increased by suturing the homograft to a tube fashioned from autologous pericardium (25 patients) or a preclotted Dacron tube (57 patients). The homograft conduit was then used to establish continuity between the right ventricle and the pulmonary artery. In the early part of the series, 12 patients underwent ligation of large bronchial collateral arteries. A complementary lateral thoracotomy was required in 10 of them.
361
Kay and Ross: Aortic Homograft for Right Ventricular Outflow Tract Reconstruction
follow-^^ All surviving patients have been followed up on an annual basis at the National Heart Hospital. The mean length of follow-up ( 2 standard deviation) is 8 ? 2 years. Eleven patients underwent elective cardiac catheterization 1 year after right ventricular outflow tract reconstruction. Sixteen patients, including 5 of those electively catheterized at 1 year, underwent late cardiac catheterization between 3 and 10 years (mean, 6 years) after operation.
Statistical Analysis Univariate analysis of discrete intraoperative variables affecting operative mortality was performed using the x2 test and Fischer’s exact test. The actuarial method 1181 was used to calculate 10-year survival of patients, together with the incidence of conduit obstruction.
Results Operative Mortality Forty-six patients (47%)died within one month of operation. The most common cause of death was low cardiac output, which occurred in 26 patients. In each instance severe pulmonary hypertension resulted in a ratio of peak systolic right ventricular to left ventricular pressure of greater than 0.8:l following total correction. In 7 patients, the VSD was reopened but this failed to change the outcome. There were 6 deaths among the 11 patients in whom the VSD was electively left open. Twelve patients died of hemorrhage. In 7 it was directly related to ligation of large collateral bronchial arteries through a complementary thoracotomy. In addition, 2 patients died of respiratory failure following severe hemorrhage from bronchial collaterals into the lung. Four patients died of sepsis. In 3 of them, an empyema had developed following ligation of collateral branches. Two other patients died of arrhythmia. Hospital mortality was significantly related to irreversible pulmonary hypertension (p < 0.001) and thoracotomy for ligation of bronchial collaterals ( p < 0.01). Late Mortality Eleven (22%)of the 51 surviving patients died during the follow-up period (0.8% per year). Four patients died suddenly of documented recurrent ventricular arrhythmias, one of which occurred following a blow on the chest. There was no evidence of conduit obstruction at postmortem examination. Two other patients died suddenly within a year of operation, presumably of an arrhythmia. Postmortem examination showed normal function of the homograft valve. Two patients died following the development of Staphylococcus epidennidis endocarditis within two months of implantation of the fresh homograft. One patient died following excision and replacement of the Dacron tube, which had become occluded by fibrinous peel 4 years after implantation of the homograft conduit. Another patient died of respiratory failure due to pulmonary infarction following ligation of a large bronchial
Patient survival 100 1 hospital survivors
70?7”/, 5
60
o l , 0
1
2 97 46 41 1
I
I
I
I
3 4 5 38 33 31
I
I
6 7 8 30 26 23
I
I
9
10years 15 patients
18
Fig 1. Acfuarial survival following right ventricular outflow tract reconstruction for pulmonary atresia.
collateral one year after right ventricular outflow tract reconstruction. One patient died in childbirth. Thus, the actuarial survival (t standard error) of the 97 patients undergoing right ventricular outflow tract reconstruction for pulmonary atresia was 37 2 7% at 10 years. Of the 51 patients initially discharged from the hospital, 70 f 7% survived 10 years (Fig 1).
Postoperative Catheterization Data Eleven patients underwent elective cardiac catheterization 1 year after operation. Catheterization revealed a resting systolic gradient across the homograft conduit of 2 to 33 mm Hg (mean, 12 f 6 mm Hg). In the 16 patients undergoing late recatheterization, the transconduit gradient had increased to a mean of 24 f 15 mm Hg. Catheterization data are shown in the Table. There was no correlation between transconduit gradient and time after operation ( r = 0.13) (Fig 2). Further Reconstruction Three patients with a transconduit gradient of 50 mm Hg have undergone revision of the conduit. In each instance, the main obstruction was related to a fibrinous peel within the Dacron tube. Two patients had severe calcification in the wall of the homograft, but in only 1 of the 3 was there calcification in the valve cusps. Postoperative Catheterization Dutua Early (1 yr) (N = 11)
Variable
Late (mean, 6 yr)
(N = 16)
Systemic ventricular pressure (mm Hg)
98
22
101 f 18
Pulmonary ventricular pressure (mm Hg) Right pulmonary artery pressure (mm Hg) Transconduit gradient (mm Hg)
50 2 21
57 f 24
39 2 23
32
12
24 f 15
‘Data are shown as mean
f
2
2
6
standard deviation.
2
20
362 The Annals of Thoracic Surgery Vol 40 No 4 October 1985
Years after operation
Fig 2 . Relationship between transconduit gradient and time after operation.
13 f t i L v 0
0 51
1 46
2 41
5 4 3 38 33 31
6 7 30 26
8 23
9 18
10years 15 p;ilienls
Fig 3 . Actuarial incidence of conduit obstruction requiring further right ventricular outflow tract reconstruction.
Thus, the actuarial incidence of conduit obstruction requiring replacement with a further homograft was 13 f 8% at 10 years (Fig 3). Postoperative Functional Class Thirty-eight (95%) of the 40 survivors are asymptomatic and lead normal lives, including 2 who have had normal pregnancies. Two patients remain symptomatic. One of them has severe right heart failure and is currently awaiting heart-lung transplantation.
Comment The National Heart Hospital has become the main center for the surgical correction of pulmonary atresia in Great Britain. Many of the patients in this series were referred late in the natural history of the disease, and often had had placement of a palliative shunt several years previously at another center. Operative M o r t a l i t y
Though the overall operative mortality was high (47%), it was of the same order of magnitude as reported by Di Carlo [19], Fontan [ZO], and their associates. Initially an attempt was made to carry out repair in all patients, regardless of the pulmonary anatomy. However, it became apparent that irreversible pulmonary hypertension resulting in right ventricular failure was the most com-
mon cause of death, accounting for 26 (57%)of the hospital deaths. Furthermore, the 2 deaths due to arrhythmia were related to a tense, distended right ventricle, and 5 of the deaths caused by hemorrhage were due to uncontrolled bleeding from friable pulmonary arteries at the distal suture line. It is now our policy to perform a palliative shunt only in those patients in whom there is severe hypoplasia of part of the pulmonary arterial tree. Placement of the shunt is followed by an early corrective procedure, before the onset of increased pulmonary vascular resistance [ll].At the same time, we have adopted the formula of Alfion and colleagues [21] for predicting the ratio of right ventricular to left ventricular pressure after repair. With these guidelines, mortality has fallen to 17% in the 12 patients seen since 1980. The second major factor contributing to hospital mortality was the concomitant ligation of bronchial collateral arteries. In the majority of patients, this involved a tedious dissection through a lateral thoracotomy, often when tissue planes had been obliterated by adhesions from previous shunt procedures. Ten (83%)of the 12 patients died either directly of hemorrhage or of empyema and pulmonary sepsis. It is now our policy to ligate bronchial arteries causing pulmonary edema at a later date [17]. During the bypass procedure, the operative field is kept dry by inflating small balloon catheters positioned in the right and left pulmonary arteries.
Late M o r t a l i t y The results for the hospital survivors were excellent with 70 f 7% living 10 years, all but 2 of whom are leading normal lives. Four of the late deaths were in patients having documented ventricular arrhythmias, and the 2 sudden deaths may also be presumed to have been caused by arrhythmias. All these patients were known to have high right ventricular pressures in the postoperative period. Though chest roentgenograms did show the presence of calcium in the homograft wall, there was no evidence of conduit obstruction at postmortem examination. Two patients died of early staphylococcal endocarditis. Repeated cultures taken from the valves both before and at the time of implantation were negative, suggesting that infection occurred during the operation. Conduit Obstruction
Early conduit obstruction in the irradiated homograft conduits led to virtual abandonment of this technique in the United States [4, 8, 101. Indeed, Schaff and associates [lo] reported a 59% incidence of conduit obstruction at 10 years. Only 3 patients in our series required conduit replacement for obstruction, corresponding to 13 f 8% at 10 years. Similar excellent results have also been reported by Di Carlo (191, Fontan [ZO], and their colleagues using antibiotic-sterilized homografts. The method of sterilization is of fundamental importance in determining the longevity of the homograft. Furthermore, in our 3 patients, the site of obstruction
363 Kay and Ross: Aortic Homograft for Right Ventricular Outflow Tract Reconstruction
was in the Dacron tube and not at the level of the valve. This pathological finding of an intimal peel has been elegantly described by Agarwal and co-workers [22, 231. The nonporous, tightly woven Dacron prevents adherence of the neointima, which is continuously dispersed by blood, resulting in a thickening of the fibrous peel (231. In replacing these conduits, we now approach the problem with caution by establishing femorofemoral bypass prior to repeat sternotomy [19]. However, our current approach is to avoid the use of Dacron altogether, either by extending the conduit with autologous pericardium or by using the rather longer pulmonary homograft. Like Fontan and associates [20], our recatheterization studies showed no correlation between transconduit gradient and time. Though calcification appeared in the wall of the homograft, it did not interfere with valve function [13]. This is in contrast to the experience of Schaff and colleagues [lo], who found that at least 34 (55%) of 62 excised xenograft valve conduits had severe obstruction at the level of the valve. This calcification represents the end point of mechanical [24] and immunological [25, 261 damage to the valve leaflets. As the pressure gradient across the valve itself and the pulmonary artery pressure is low, so mechanical damage should be reduced. In any event, it should be similar for both homograft and xenograft conduits. Thus the difference in the longevity of the two conduits may, in part, be related to differences in antigenicity between the homograft and the glutaraldehyde-treated xenograft [25, 261.
References 1. Ross DN, Somerville J: Correction of pulmonary atresia with a homograft aortic valve. Lancet 2:1446, 1966 2. McGoon DC, Rastelli GC, Ongley PA: An operation for the correction of truncus arteriosus. JAMA 205:69, 1968 3. Rastelli GC, McGoon DC, Wallace RB: Anatomic correction of transposition of the great arteries with ventricular septa1 defect and subpulmonary stenosis. J Thorac Cardiovasc Surg 58545, 1969 4. Merin G, McGoon DC: Reoperation after insertion of aortic homograft as a right Ventricular outflow tract. Ann Thorac Surg 16:122, 1973 5. Bowman RO, Hancock WD, Malm J R A valve containing Dacron prosthesis: its use in restoring pulmonary arteryright ventricular continuity. Arch Surg 107724, 1973 6. Bailey WW, Kirklin JW, Bargeron LM, et al: Late results with synthetic valved external conduits from venous ventricle to pulmonary arteries. Circulation 56Suppl 273, 1977 7. Ciaravella JM Jr, McGoon DC, Danielson GK, et al: Experience with the extracardiac conduit. J Thorac Cardiovasc Surg 78:920, 1979 8. Castaneda AR, Norwood W1: Valved conduits: a panacea for complex congenital heart defects. In Cohn LH, Gallucci V (eds): Cardiac Bioprostheses: Proceedings of the Second International Symposium. New York, Yorke, 1982, pp 205216 9. Norwood WI, Freed MD, Rocchini AP, et al: Experience with valved conduits for repair of congenital cardiac lesions. Ann Thorac Surg 24223, 1977
10. Schaff HV, DiDonato RM, Danielson GK, et al: Reoperation for obstructed pulmonary ventricle-pulmonary artery conduits: early and late results. J Thorac Cardiovasc Surg 88:334, 1984 11. Moore CH, Martelli B, Ross DN: Reconstruction of right ventricular outflow tract with a valved conduit in 75 cases of congenital heart disease. J Thorac Cardiovasc Surg 71:11, 1976 12. Saravalli OA, Somerville 1, Jefferson KE: Calcification of aortic homografts used for reconstruction of the right ventricular outflow tract. J Thorac Cardiovasc Surg 80:909, 1980 13. Shabbo FP, Wain WH, Ross DN: Right ventricular outflow reconstruction with aortic homograft conduit: analysis of the long-term results. Thorac Cardiovasc Surg 28:21, 1980 14. Al-Janabi N, Ross DN: Enhanced viability of fresh aortic homografts stored in nutrient medium. Cardiovasc Res 7:817, 1973 15. Yacoub M, Kittle CF: Sterilization of valve homografts by antibiotic solution. Circulation 4l:Suppl 2:29, 1970 16. Wain WH, Pearce HM, Riddell RW, Ross DN: A reevaluation of antibiotic sterilisation of heart valve allografts. Thorax 32:740, 1977 17. Somerville J, Saravalli 0, Ross DN: Complex pulmonary atresia with congenital systemic collaterals: classification and management. Arch Ma1 Coeur 71:322, 1978 18. Grunkemeier GL, Starr A: Actuarial analysis of surgical results: rationale and method. Ann Thorac Surg 24:404, 1977 19. Di Carlo D, de Leva1 MR, Stark J: “Fresh” antibiotic sterilised aortic homografts in extracardiac valved conduits: long-term results. Thorac Cardiovasc Surg 32:10, 1984 20. Fontan F, Choussat A, Deville C, et al: Aortic valve homografts in the surgical treatment of complex cardiac malformations. J Thorac Cardiovasc Surg 87649, 1984 21. Alfion A, Blackstone E, Kirklin JW, et al: Surgical treatment of tetralogy of Fallot with pulmonary atresia. J Thorac Cardiovasc Surg 76:321, 1978 22. Aganval KC, Edwards WD, Feldt RM. et al: Clinicopathological correlates of obstructed right-sided porcinevalved extracardiac conduits. J Thorac Cardiovasc Surg 81:591, 1981 23. Aganval KC, Edwards WD, Feldt RM, et al: Pathogenesis of nonobstructive fibrous peels in right-sided porcine-valved extracardiac conduits. J Thorac Cardiovasc Surg 83:584, 1982 24. Circio CA, Commerford PJ, Rose AG, et a 1 Calcification of glutaraldehyde-preserved porcine xenografts in young patients. J Thorac Cardiovasc Surg 81:621, 1981 25. Carpentier A, Lemaigne G, Robert L, et al: Biological factors affecting long-term results of valvular heterografts. J Thorac Cardiovasc Surg 58:467, 1967 26. Zuhdi N, Hawley W, Voehl V: Porcine aortic valves as replacement for human heart valves. Ann Thorac Surg 17479, 1974
Discussion DR. GORDON K . DANIELSON JR (Rochester, MN): It is a real pleasure to have the opportunity to discuss this fine presentation by Mr. Kay. Mr. Ross made a major contribution to the development of the extracardiac conduit with his report in 1966 of clinical application of a homograft aorta with integral valve. The patient had pulmonary atresia. The series presented by Mr. Kay gives us a 19-year follow-up of the experience that began with that historic case. At the Mayo Clinic since January, 1967, 260 patients have
364 The Annals of Thoracic Surgery Vol 40 No 4 October 1985
undergone definitive repair of pulmonary atresia with an overall mortality of 6.9%. Among the last 60 patients, there was 1 death, a mortality of 2%. We prefer to permanently control systemic-pulmonary collaterals at the time of definitive repair, currently nearly always through a median sternotomy. More than half of the early deaths in the study just presented were due to low cardiac output. My first question is whether the authors believe that persistence of collateral flow may have contributed to the development of low cardiac output in these patients. Admittedly, control of collaterals may be difficult, especially if the pleural space is obliterated by adhesions from a previous operation. In such instances, we have investigated preoperative or early postoperative closure of the collaterals by catheter balloon or coil techniques, and we hope that advances in this area will simplify the surgeon's task. We agree that with the authors that there is a place for right ventricular outflow tract reconstruction with the VSD left open. Actually, we believe this "first-stage" procedure is the best approach for enlarging hypoplastic confluent pulmonary arteries. With this approach, we have had no operative deaths in the last 2 years, and the overall mortality for 118 patients has been 7.6%, essentially the same as that for complete repair. The homograft has several disadvantages. These include a tendency to angulate at the proximal anastomosis (particularly in the repair of transposition and truncus arteriosus), a tendency for the wall to calcify making subsequent removal more hazardous, the frequent need for constructing distal extensions to reach or enlarge the pulmonary arteries (especially for complex reconstructions), and availability. However, with regard to durability, the authors have made an excellent case for the fresh aortic homograft. They have an enviable 87% probability of the conduit remaining failure-free at 10 years. We have employed several types of extracardiac conduits in our series of 1,100 patients. The Hancock conduit, used in the largest subset, has a failure-free rate of 94% in 5 years, but the 10-year rate may be as low as 75%. Both of these conduits are an improvement over the irradiated homograft, which
had a failure-free rate of only 41% at 10 years. Fortunately, reoperation is a low-risk event. In a review of our first 100 reoperations for conduit replacement, we found there were no hospital deaths among the 47 patients who had isolated conduit obstruction without associated defects that required concomitant procedures. My second question concerns long-term results. Mr. Kay, d o you have any evidence yet, from your experience with pulmonary atresia and other congenital anomalies, as to whether the long-term results of the frozen homograft valve will compare favorably with those for the fresh homograft? The search for the ideal conduit continues. In the meantime, I agree with the authors that the fresh homograft, when available and where applicable, can justifiably be considered a conduit of choice for right ventricular outflow tract reconstruction. MR. KAY: I congratulate Dr. Danielson on his excellent results and thank him for his kind remarks. I am pleased to hear that he agrees that the long-term results of aortic homografts are superior to those of other conduits in the right ventricular outflow position. In answer to the first question, the National Heart Hospital is one of the few centers in the United Kingdom performing a definitive correction of pulmonary atresia. At the start of the series, it was thought that all patients should undergo total correction, regardless of the degree and reversibility of their pulmonary hypertension. Since we adopted a more selective policy, the operative mortality has been dramatically reduced. Analysis of the patients dying of pulmonary hypertension suggests that ligature of large bronchial collaterals is not of major importance. We are very fortunate in having a homograft laboratory within the hospital and consequently have a large supply of fresh homografts. Thus, we have little experience with the use of frozen aortic homografts in children. However, examination of the long-term results of homograft aortic valve replacement in more than 700 adults suggests improved results in patients receiving fresh homografts.
The use of the aortic homograft for right ventricular outflow tract reconstruction was pioneered in 1966 by Ross and Somerville [l]. Initially the conduit was used for patients with pulmonary atresia, but subsequently the criteria were extended to include more complex anomalies [2, 31. However, early calcification leading to obstruction of the irradiated homografts caused several centers to abandon their use [4, 51. The alternative Dacron tube grafts containing a preserved xenograft aortic valve were introduced in 1973 [5] and have gained wide popularity since then [6-81. However, these conduits also suffer from late obstruction [9, 101. During the last 18 years we have continued to use fresh, antibiotic-sterilized aortic homografts for right ventricular outflow reconstruction [ 11-13]. Accordingly, we have analyzed our experience, focusing on the incidence of late conduit dysfunction.
From the National Heart Hospital, London, England Presented at the Twenty-first Annual Meeting of The society of Thoracic Surgeons, Phoenix, AZ,Jan 21-23, 1985. Address reprint requests to h4r. Ross, National Heart Hospital, Westmoreland St, London WlM BBA, England.
360
Material and Methods Between 1966 and 1984, 97 patients with pulmonary atresia underwent right ventricular outflow tract reconstruction using a homograft conduit. There were 47 male (48%)and 50 female patients, with ages ranging from 2 to 42 years (median, 11 years). Sixty-eight patients had undergone a systemicpulmonary shunt (22 Blalock-Taussig, 43 Waterston, and 3 Potts) as an initial procedure. Fifteen patients had undergone a second Blalock-Taussig shunt prior to correction.
Homograft Preservation It is our policy to use a fresh homograft whenever possible [13]. In this series, 84 patients received a fresh aortic homograft and 9, a frozen aortic homograft. The last 4 patients seen received a fresh pulmonary homograft. The fresh homografts are preserved in a nutrient medium to enhance tissue viability [14]. This medium is sterilized using an antibiotic mixture [15] and a fungicide [16]. After ten days if the valve has not been used in its fresh state, glycerol is added to the solution to a concentration of 15%. The valve is then frozen in the vapor phase of a liquid nitrogen refrigerator where it is stored. Operative Details Blalock-Taussig and Potts shunts were controlled before the onset of cardiopulmonary bypass. In those patients with a Waterston shunt, the right pulmonary artery was repaired with a patch of autologous pericardium. Standard techniques of cardiopulmonary bypass at moderate hypothermia (25°C) were used. A vertical incision was made in the right ventricle, thereby exposing the ventricular septa1 defect (VSD). In 86 patients the VSD was closed with a patch of knitted Dacron velour. The VSD was left open electively in 11 patients with hypoplastic pulmonary arteries and severe pulmonary hypertension [17]. A suitable homograft (19 to 23 mm) was then chosen and tailored to the existing anatomy. In 82 patients, the length of the conduit was increased by suturing the homograft to a tube fashioned from autologous pericardium (25 patients) or a preclotted Dacron tube (57 patients). The homograft conduit was then used to establish continuity between the right ventricle and the pulmonary artery. In the early part of the series, 12 patients underwent ligation of large bronchial collateral arteries. A complementary lateral thoracotomy was required in 10 of them.
361
Kay and Ross: Aortic Homograft for Right Ventricular Outflow Tract Reconstruction
follow-^^ All surviving patients have been followed up on an annual basis at the National Heart Hospital. The mean length of follow-up ( 2 standard deviation) is 8 ? 2 years. Eleven patients underwent elective cardiac catheterization 1 year after right ventricular outflow tract reconstruction. Sixteen patients, including 5 of those electively catheterized at 1 year, underwent late cardiac catheterization between 3 and 10 years (mean, 6 years) after operation.
Statistical Analysis Univariate analysis of discrete intraoperative variables affecting operative mortality was performed using the x2 test and Fischer’s exact test. The actuarial method 1181 was used to calculate 10-year survival of patients, together with the incidence of conduit obstruction.
Results Operative Mortality Forty-six patients (47%)died within one month of operation. The most common cause of death was low cardiac output, which occurred in 26 patients. In each instance severe pulmonary hypertension resulted in a ratio of peak systolic right ventricular to left ventricular pressure of greater than 0.8:l following total correction. In 7 patients, the VSD was reopened but this failed to change the outcome. There were 6 deaths among the 11 patients in whom the VSD was electively left open. Twelve patients died of hemorrhage. In 7 it was directly related to ligation of large collateral bronchial arteries through a complementary thoracotomy. In addition, 2 patients died of respiratory failure following severe hemorrhage from bronchial collaterals into the lung. Four patients died of sepsis. In 3 of them, an empyema had developed following ligation of collateral branches. Two other patients died of arrhythmia. Hospital mortality was significantly related to irreversible pulmonary hypertension (p < 0.001) and thoracotomy for ligation of bronchial collaterals ( p < 0.01). Late Mortality Eleven (22%)of the 51 surviving patients died during the follow-up period (0.8% per year). Four patients died suddenly of documented recurrent ventricular arrhythmias, one of which occurred following a blow on the chest. There was no evidence of conduit obstruction at postmortem examination. Two other patients died suddenly within a year of operation, presumably of an arrhythmia. Postmortem examination showed normal function of the homograft valve. Two patients died following the development of Staphylococcus epidennidis endocarditis within two months of implantation of the fresh homograft. One patient died following excision and replacement of the Dacron tube, which had become occluded by fibrinous peel 4 years after implantation of the homograft conduit. Another patient died of respiratory failure due to pulmonary infarction following ligation of a large bronchial
Patient survival 100 1 hospital survivors
70?7”/, 5
60
o l , 0
1
2 97 46 41 1
I
I
I
I
3 4 5 38 33 31
I
I
6 7 8 30 26 23
I
I
9
10years 15 patients
18
Fig 1. Acfuarial survival following right ventricular outflow tract reconstruction for pulmonary atresia.
collateral one year after right ventricular outflow tract reconstruction. One patient died in childbirth. Thus, the actuarial survival (t standard error) of the 97 patients undergoing right ventricular outflow tract reconstruction for pulmonary atresia was 37 2 7% at 10 years. Of the 51 patients initially discharged from the hospital, 70 f 7% survived 10 years (Fig 1).
Postoperative Catheterization Data Eleven patients underwent elective cardiac catheterization 1 year after operation. Catheterization revealed a resting systolic gradient across the homograft conduit of 2 to 33 mm Hg (mean, 12 f 6 mm Hg). In the 16 patients undergoing late recatheterization, the transconduit gradient had increased to a mean of 24 f 15 mm Hg. Catheterization data are shown in the Table. There was no correlation between transconduit gradient and time after operation ( r = 0.13) (Fig 2). Further Reconstruction Three patients with a transconduit gradient of 50 mm Hg have undergone revision of the conduit. In each instance, the main obstruction was related to a fibrinous peel within the Dacron tube. Two patients had severe calcification in the wall of the homograft, but in only 1 of the 3 was there calcification in the valve cusps. Postoperative Catheterization Dutua Early (1 yr) (N = 11)
Variable
Late (mean, 6 yr)
(N = 16)
Systemic ventricular pressure (mm Hg)
98
22
101 f 18
Pulmonary ventricular pressure (mm Hg) Right pulmonary artery pressure (mm Hg) Transconduit gradient (mm Hg)
50 2 21
57 f 24
39 2 23
32
12
24 f 15
‘Data are shown as mean
f
2
2
6
standard deviation.
2
20
362 The Annals of Thoracic Surgery Vol 40 No 4 October 1985
Years after operation
Fig 2 . Relationship between transconduit gradient and time after operation.
13 f t i L v 0
0 51
1 46
2 41
5 4 3 38 33 31
6 7 30 26
8 23
9 18
10years 15 p;ilienls
Fig 3 . Actuarial incidence of conduit obstruction requiring further right ventricular outflow tract reconstruction.
Thus, the actuarial incidence of conduit obstruction requiring replacement with a further homograft was 13 f 8% at 10 years (Fig 3). Postoperative Functional Class Thirty-eight (95%) of the 40 survivors are asymptomatic and lead normal lives, including 2 who have had normal pregnancies. Two patients remain symptomatic. One of them has severe right heart failure and is currently awaiting heart-lung transplantation.
Comment The National Heart Hospital has become the main center for the surgical correction of pulmonary atresia in Great Britain. Many of the patients in this series were referred late in the natural history of the disease, and often had had placement of a palliative shunt several years previously at another center. Operative M o r t a l i t y
Though the overall operative mortality was high (47%), it was of the same order of magnitude as reported by Di Carlo [19], Fontan [ZO], and their associates. Initially an attempt was made to carry out repair in all patients, regardless of the pulmonary anatomy. However, it became apparent that irreversible pulmonary hypertension resulting in right ventricular failure was the most com-
mon cause of death, accounting for 26 (57%)of the hospital deaths. Furthermore, the 2 deaths due to arrhythmia were related to a tense, distended right ventricle, and 5 of the deaths caused by hemorrhage were due to uncontrolled bleeding from friable pulmonary arteries at the distal suture line. It is now our policy to perform a palliative shunt only in those patients in whom there is severe hypoplasia of part of the pulmonary arterial tree. Placement of the shunt is followed by an early corrective procedure, before the onset of increased pulmonary vascular resistance [ll].At the same time, we have adopted the formula of Alfion and colleagues [21] for predicting the ratio of right ventricular to left ventricular pressure after repair. With these guidelines, mortality has fallen to 17% in the 12 patients seen since 1980. The second major factor contributing to hospital mortality was the concomitant ligation of bronchial collateral arteries. In the majority of patients, this involved a tedious dissection through a lateral thoracotomy, often when tissue planes had been obliterated by adhesions from previous shunt procedures. Ten (83%)of the 12 patients died either directly of hemorrhage or of empyema and pulmonary sepsis. It is now our policy to ligate bronchial arteries causing pulmonary edema at a later date [17]. During the bypass procedure, the operative field is kept dry by inflating small balloon catheters positioned in the right and left pulmonary arteries.
Late M o r t a l i t y The results for the hospital survivors were excellent with 70 f 7% living 10 years, all but 2 of whom are leading normal lives. Four of the late deaths were in patients having documented ventricular arrhythmias, and the 2 sudden deaths may also be presumed to have been caused by arrhythmias. All these patients were known to have high right ventricular pressures in the postoperative period. Though chest roentgenograms did show the presence of calcium in the homograft wall, there was no evidence of conduit obstruction at postmortem examination. Two patients died of early staphylococcal endocarditis. Repeated cultures taken from the valves both before and at the time of implantation were negative, suggesting that infection occurred during the operation. Conduit Obstruction
Early conduit obstruction in the irradiated homograft conduits led to virtual abandonment of this technique in the United States [4, 8, 101. Indeed, Schaff and associates [lo] reported a 59% incidence of conduit obstruction at 10 years. Only 3 patients in our series required conduit replacement for obstruction, corresponding to 13 f 8% at 10 years. Similar excellent results have also been reported by Di Carlo (191, Fontan [ZO], and their colleagues using antibiotic-sterilized homografts. The method of sterilization is of fundamental importance in determining the longevity of the homograft. Furthermore, in our 3 patients, the site of obstruction
363 Kay and Ross: Aortic Homograft for Right Ventricular Outflow Tract Reconstruction
was in the Dacron tube and not at the level of the valve. This pathological finding of an intimal peel has been elegantly described by Agarwal and co-workers [22, 231. The nonporous, tightly woven Dacron prevents adherence of the neointima, which is continuously dispersed by blood, resulting in a thickening of the fibrous peel (231. In replacing these conduits, we now approach the problem with caution by establishing femorofemoral bypass prior to repeat sternotomy [19]. However, our current approach is to avoid the use of Dacron altogether, either by extending the conduit with autologous pericardium or by using the rather longer pulmonary homograft. Like Fontan and associates [20], our recatheterization studies showed no correlation between transconduit gradient and time. Though calcification appeared in the wall of the homograft, it did not interfere with valve function [13]. This is in contrast to the experience of Schaff and colleagues [lo], who found that at least 34 (55%) of 62 excised xenograft valve conduits had severe obstruction at the level of the valve. This calcification represents the end point of mechanical [24] and immunological [25, 261 damage to the valve leaflets. As the pressure gradient across the valve itself and the pulmonary artery pressure is low, so mechanical damage should be reduced. In any event, it should be similar for both homograft and xenograft conduits. Thus the difference in the longevity of the two conduits may, in part, be related to differences in antigenicity between the homograft and the glutaraldehyde-treated xenograft [25, 261.
References 1. Ross DN, Somerville J: Correction of pulmonary atresia with a homograft aortic valve. Lancet 2:1446, 1966 2. McGoon DC, Rastelli GC, Ongley PA: An operation for the correction of truncus arteriosus. JAMA 205:69, 1968 3. Rastelli GC, McGoon DC, Wallace RB: Anatomic correction of transposition of the great arteries with ventricular septa1 defect and subpulmonary stenosis. J Thorac Cardiovasc Surg 58545, 1969 4. Merin G, McGoon DC: Reoperation after insertion of aortic homograft as a right Ventricular outflow tract. Ann Thorac Surg 16:122, 1973 5. Bowman RO, Hancock WD, Malm J R A valve containing Dacron prosthesis: its use in restoring pulmonary arteryright ventricular continuity. Arch Surg 107724, 1973 6. Bailey WW, Kirklin JW, Bargeron LM, et al: Late results with synthetic valved external conduits from venous ventricle to pulmonary arteries. Circulation 56Suppl 273, 1977 7. Ciaravella JM Jr, McGoon DC, Danielson GK, et al: Experience with the extracardiac conduit. J Thorac Cardiovasc Surg 78:920, 1979 8. Castaneda AR, Norwood W1: Valved conduits: a panacea for complex congenital heart defects. In Cohn LH, Gallucci V (eds): Cardiac Bioprostheses: Proceedings of the Second International Symposium. New York, Yorke, 1982, pp 205216 9. Norwood WI, Freed MD, Rocchini AP, et al: Experience with valved conduits for repair of congenital cardiac lesions. Ann Thorac Surg 24223, 1977
10. Schaff HV, DiDonato RM, Danielson GK, et al: Reoperation for obstructed pulmonary ventricle-pulmonary artery conduits: early and late results. J Thorac Cardiovasc Surg 88:334, 1984 11. Moore CH, Martelli B, Ross DN: Reconstruction of right ventricular outflow tract with a valved conduit in 75 cases of congenital heart disease. J Thorac Cardiovasc Surg 71:11, 1976 12. Saravalli OA, Somerville 1, Jefferson KE: Calcification of aortic homografts used for reconstruction of the right ventricular outflow tract. J Thorac Cardiovasc Surg 80:909, 1980 13. Shabbo FP, Wain WH, Ross DN: Right ventricular outflow reconstruction with aortic homograft conduit: analysis of the long-term results. Thorac Cardiovasc Surg 28:21, 1980 14. Al-Janabi N, Ross DN: Enhanced viability of fresh aortic homografts stored in nutrient medium. Cardiovasc Res 7:817, 1973 15. Yacoub M, Kittle CF: Sterilization of valve homografts by antibiotic solution. Circulation 4l:Suppl 2:29, 1970 16. Wain WH, Pearce HM, Riddell RW, Ross DN: A reevaluation of antibiotic sterilisation of heart valve allografts. Thorax 32:740, 1977 17. Somerville J, Saravalli 0, Ross DN: Complex pulmonary atresia with congenital systemic collaterals: classification and management. Arch Ma1 Coeur 71:322, 1978 18. Grunkemeier GL, Starr A: Actuarial analysis of surgical results: rationale and method. Ann Thorac Surg 24:404, 1977 19. Di Carlo D, de Leva1 MR, Stark J: “Fresh” antibiotic sterilised aortic homografts in extracardiac valved conduits: long-term results. Thorac Cardiovasc Surg 32:10, 1984 20. Fontan F, Choussat A, Deville C, et al: Aortic valve homografts in the surgical treatment of complex cardiac malformations. J Thorac Cardiovasc Surg 87649, 1984 21. Alfion A, Blackstone E, Kirklin JW, et al: Surgical treatment of tetralogy of Fallot with pulmonary atresia. J Thorac Cardiovasc Surg 76:321, 1978 22. Aganval KC, Edwards WD, Feldt RM. et al: Clinicopathological correlates of obstructed right-sided porcinevalved extracardiac conduits. J Thorac Cardiovasc Surg 81:591, 1981 23. Aganval KC, Edwards WD, Feldt RM, et al: Pathogenesis of nonobstructive fibrous peels in right-sided porcine-valved extracardiac conduits. J Thorac Cardiovasc Surg 83:584, 1982 24. Circio CA, Commerford PJ, Rose AG, et a 1 Calcification of glutaraldehyde-preserved porcine xenografts in young patients. J Thorac Cardiovasc Surg 81:621, 1981 25. Carpentier A, Lemaigne G, Robert L, et al: Biological factors affecting long-term results of valvular heterografts. J Thorac Cardiovasc Surg 58:467, 1967 26. Zuhdi N, Hawley W, Voehl V: Porcine aortic valves as replacement for human heart valves. Ann Thorac Surg 17479, 1974
Discussion DR. GORDON K . DANIELSON JR (Rochester, MN): It is a real pleasure to have the opportunity to discuss this fine presentation by Mr. Kay. Mr. Ross made a major contribution to the development of the extracardiac conduit with his report in 1966 of clinical application of a homograft aorta with integral valve. The patient had pulmonary atresia. The series presented by Mr. Kay gives us a 19-year follow-up of the experience that began with that historic case. At the Mayo Clinic since January, 1967, 260 patients have
364 The Annals of Thoracic Surgery Vol 40 No 4 October 1985
undergone definitive repair of pulmonary atresia with an overall mortality of 6.9%. Among the last 60 patients, there was 1 death, a mortality of 2%. We prefer to permanently control systemic-pulmonary collaterals at the time of definitive repair, currently nearly always through a median sternotomy. More than half of the early deaths in the study just presented were due to low cardiac output. My first question is whether the authors believe that persistence of collateral flow may have contributed to the development of low cardiac output in these patients. Admittedly, control of collaterals may be difficult, especially if the pleural space is obliterated by adhesions from a previous operation. In such instances, we have investigated preoperative or early postoperative closure of the collaterals by catheter balloon or coil techniques, and we hope that advances in this area will simplify the surgeon's task. We agree that with the authors that there is a place for right ventricular outflow tract reconstruction with the VSD left open. Actually, we believe this "first-stage" procedure is the best approach for enlarging hypoplastic confluent pulmonary arteries. With this approach, we have had no operative deaths in the last 2 years, and the overall mortality for 118 patients has been 7.6%, essentially the same as that for complete repair. The homograft has several disadvantages. These include a tendency to angulate at the proximal anastomosis (particularly in the repair of transposition and truncus arteriosus), a tendency for the wall to calcify making subsequent removal more hazardous, the frequent need for constructing distal extensions to reach or enlarge the pulmonary arteries (especially for complex reconstructions), and availability. However, with regard to durability, the authors have made an excellent case for the fresh aortic homograft. They have an enviable 87% probability of the conduit remaining failure-free at 10 years. We have employed several types of extracardiac conduits in our series of 1,100 patients. The Hancock conduit, used in the largest subset, has a failure-free rate of 94% in 5 years, but the 10-year rate may be as low as 75%. Both of these conduits are an improvement over the irradiated homograft, which
had a failure-free rate of only 41% at 10 years. Fortunately, reoperation is a low-risk event. In a review of our first 100 reoperations for conduit replacement, we found there were no hospital deaths among the 47 patients who had isolated conduit obstruction without associated defects that required concomitant procedures. My second question concerns long-term results. Mr. Kay, d o you have any evidence yet, from your experience with pulmonary atresia and other congenital anomalies, as to whether the long-term results of the frozen homograft valve will compare favorably with those for the fresh homograft? The search for the ideal conduit continues. In the meantime, I agree with the authors that the fresh homograft, when available and where applicable, can justifiably be considered a conduit of choice for right ventricular outflow tract reconstruction. MR. KAY: I congratulate Dr. Danielson on his excellent results and thank him for his kind remarks. I am pleased to hear that he agrees that the long-term results of aortic homografts are superior to those of other conduits in the right ventricular outflow position. In answer to the first question, the National Heart Hospital is one of the few centers in the United Kingdom performing a definitive correction of pulmonary atresia. At the start of the series, it was thought that all patients should undergo total correction, regardless of the degree and reversibility of their pulmonary hypertension. Since we adopted a more selective policy, the operative mortality has been dramatically reduced. Analysis of the patients dying of pulmonary hypertension suggests that ligature of large bronchial collaterals is not of major importance. We are very fortunate in having a homograft laboratory within the hospital and consequently have a large supply of fresh homografts. Thus, we have little experience with the use of frozen aortic homografts in children. However, examination of the long-term results of homograft aortic valve replacement in more than 700 adults suggests improved results in patients receiving fresh homografts.