of the Right Ventricular Outflow Tract

Transcription

1 Late Results of Reconstruction of the Right Ventricular Outflow Tract with Porcine Xenogafts in Children George S. Bisset 111, M.D., David C. Schwartz, M.D., George Benzing 111, M.D., James Helmsworth, M.D., J. Tracy Schreiber, M.D., and Samuel Kaplan, M.D. ABSTRACT Thirty-three children, aged 2.5 to 17.5 years (mean, 8.3 years), having xenograft external conduits placed between the right ventricle and pulmonary artery were followed for 1 to 6 years postoperatively (mean, 3.5 years). There were no late deaths in the study group, and no infection of a valved conduit has been demonstrated. Twenty of these children were catheterized during the followup period. The gradients from the right ventricle to pulmonary artery ranged from 8 to 90 mm Hg (average, 31 mm Hg). A total of 8 patients were classified as having hemodynamically documented conduit failure, and an additional 2 patients are clinically expected to have conduit failure. This represents a total incidence of 30% xenograft conduit failure in a 6-year follow-up. Although the etiology of this dysfunction is probably multifactorial, factors such as valve size, conduit angulation, and immunological competence bear special consideration. We conclude that although valved external conduits continue to play an important role in the treatment of complex congenital heart disease, a valved conduit with greater longevity is needed for use in children. Modem search for a functionally competent valved prosthesis to establish continuity between the right ventricle and pulmonary artery was initiated by Ross and Somerville [ll, who in 1966 described the use of an aortic homograft From the Departments of Pediatric Cardiology and Cardio-Thoracic Surgery, Children s Hospital Medical Center, Cincinnati, OH. Supported in part by Grant 5 T32 HL Presented at the Twenty-ninth Annual Scientific Session of the American College of Cardiology, Mar 9-13, 1980, Houston, TX. Accepted for publication Sept 29, Address reprint requests to Dr. Schwartz, Division of Pediatric Cardiology, Children s Hospital Medical Center, Elland and Bethesda Aves, Cincinnati, OH to functionally correct pulmonary atresia and ventricular septa1 defect (VSD). This bioprothesis was attractive because of its relative freedom from thromboembolic complications, which obviated the need for anticoagulation in young, physically active children. However, initial enthusiasm over the aortic homografts was dampened by findings of late valve dysfunction in a significant number of patients, especially those receiving irradiated, freezedried, or chemically treated homografts [2-51. There was also dissatisfaction with the fresh homografts, which demonstrated late dysfunction [6] and presented problems because of the inability to readily harvest these tissues. Because of these findings, alternative tissue valves, including dura mater, fascia lata, and pericardial tissue, have been utilized. They also have resulted in a high incidence of valve dysfunction [7, 81. In attempts to alleviate these problems and to reduce tissue antigenicity, various groups used glutaraldehyde-processed porcine xenografts [6, 91. This valve has proved to be durable in adults in the aortic and mitral positions [6, 101. However, in children there have been recent reports of porcine xenograft failure in the aortic, mitral, and pulmonary positions [5,11,12]. The purpose of this report is to present the longterm performance and complications associated with the use of porcine xenografts for reconstruction of the right ventricular outflow tract in children with congenital heart disease. Material and Methods All surviving patients in whom porcine aortic valve (16 to 35 mm) xenograft external conduits* were placed between the right ventricle and pulmonary artery at Children s Hospital *Hancock Laboratories, Inc., Anaheim, CA by The Society of Thoracic Surgeons

2 438 The Annals of Thoracic Surgery Vol 31 No 5 May 1981 Medical Center between December, 1973, and August, 1979, were included in the study. Thirty-three children, 2.5 to 17.5 years old (mean, 8.3 years) at the time of operation, were in the study group. There were 15 girls and 18 boys. The indications for operative intervention included progressive exercise intolerance, failure to thrive, increasing cyanosis with progressive polycythemia, and hypoxic spells. Twenty-eight patients had valved conduits inserted as part of the surgical repair of tetralogy of Fallot. Valved conduits were utilized for patients with an absent pulmonary valve, anomalous coronary artery distribution precluding incision of the right ventricular outflow tract, pulmonary hypertension secondary to previous systemic-pulmonary shunts, or a severely hypoplastic or atretic pulmonary valve ring associated with infundibular obstruction. In this last group, conduits were only utilized when patch-graft enlargement of the pulmonary valve ring was not feasible. The remaining 5 patients, undergoing placement of the xenograft, had the following lesions (one in each group): truncus arteriosus (type I), isolated absence of the pulmonary valve, VSD with pulmonary artery band, double-outlet right ventricle with pulmonary stenosis, and doubleoutlet left ventricle with pulmonary stenosis. The operative technique in each of the 33 patients included the use of cardiopulmonary bypass with systemic and local hypothermia. Cold cardioplegic solution was utilized for circulatory arrest. The coronary artery anatomy was delineated in order to prevent injury to an anomalous branch. A vertical right ventriculotomy was used to approach the VSD and explore the right ventricular outflow tract. Following closure of the VSD with a Dacron patch, the proximal end of the conduit was tailored for attachment to the right ventriculotomy. In 25 patients, intraoperative measurements of right ventricular and pulmonary arterial pressures were performed at the termination of the surgical procedure by direct cannulation in the operating room. Right ventricular peak systolic and systemic arterial pressures only, were obtained in 7 of the remaining 8 patients. Twenty children underwent cardiac catheterization, with hemodynamic evaluation one month to 4 years postoperatively (mean, 1.5 years). Results The duration of follow-up ranged from 1 to 6 years (mean, 3.5 years). Each patient has been clinically assessed at six-month intervals. There have been no late deaths. In the 25 patients in whom right ventricular and pulmonary arterial pressures were measured intraoperatively, there was a peak systolic pressure gradient between 0 and 38 mm Hg (mean, 10 mm Hg). Of the remaining 7 patients, 6 had a peak right ventricular to systemic arterial pressure ratio of 0.5:1, and the remaining child had a ratio of 0.6: 1. There have been no instances of thromboembolic complication detectable by history or physical examination. No infection of a valved conduit has been identified in the immediate or late postoperative period. Of 22 patients who have undergone repeat operation or cardiac catheterization or both, 6 were noted to have a small residual VSD (pulmonary to systemic flow ratio of less than 2 : l), and 2 had a small residual atrial septa1 defect. In each of the 4 patients who has undergone reoperation, the gradient has been relieved as demonstrated by intraoperative right ventricular and distal pulmonary arterial pressures. There have been no deaths in this group. Postopera tive Catheterization Data Right ventricle-pulmonary artery peak systolic pressure gradients were measured in 20 patients. The gradients ranged from 8 to 90 mm Hg (average, 31 mm Hg). Seven of these patients had a right ventricle-pulmonary artery gradient greater than 40 mm Hg or gross porcine valve incompetence or both conditions (Table). An additional patient had a large pseudoaneurysm of the right ventricular outflow tract, which was associated with a 20 mm Hg gradient across the valve. Overall, 11 patients demonstrated a gradient at the level of the valve, 5 had obstruction at both the proximal anastomosis and the valve level, 3 had evidence of distal anastomosis and valvular pressure gradients, and 1 patient had obstruction at all three levels. Two additional patients (not included in the Table) now have

3 ~~~ ~ ~ 439 Bisset et al: RVOT Reconstruction with Porcine Xenografts Hemodynamic Data in 8 Patients with Xenograft Conduit Failure Intraoperative Follow-up Age at Pressure Catheter- Pa- Opera- Conduit Measure- ization Postop tient tion Size ments (yrs RV-PA Sites of No. Diagnosis (yrs) (mm) (mm Hg) postop) Gradienta Obstruction 1 Tetralogy of 9 18 RV= RV = 81/10 Valve Fallot LV = MPA (distal) = Tetralogy of 7 20 RV = RV= 9017 Valve Fallot PA = MPA (distal) = 2018 Distal RPA 3 Double RV = Valve outlet right MPA (distal) = 1218 ventricle 4 Tetralogy of RV = RV = 6410 Valve Fallot LV = MPA (distal) = 4415 Distal anastomosis 5 Tetralogy of Fallot 7 16 RV = 4515 PA = RV = 8015 MPA (distal) = Valve 6 Tetralogy of 3 16 RV = RV = Distal Fallot PA = MPA (distal) = anastomosis Proximal anastomosis LPA (distal) = Valve 7 Truncus arteriosus RV = PA = RV = MPA (distal) = Valve 8 Tetralogy of RV = RV = 6019 Valve Fallot PA = MPA (distal) = acatheterization data. bsites listed in order of severity. RV = right ventricle; LV = left ventricle; MPA = main pulmonary artery; PA = pulmonary artery; RPA = right pulmonary artery; LPA = left pulmonary artery. clinical and echocardiographic evidence of severe prosthetic valve incompetence, and have not been recatheterized. Pa thology In every patient in whom the valve was examined pathologically, the cusps were shortened and irregularly thickened, with focal mineralization. Occasional small tears were noted through calcified cusp tissue. The valve from Patient 2 is illustrated in Figure 1. Microscopically, there were advanced degenerative changes, characterized by ghost cells (secondary to severe autolysis), and extensive areas of calcification, which were usually closely adjacent to areas of collagen degeneration. There was a distinct scarcity of inflammatory cells (lymphocytes, macrophages, or plasma cells) in each of the tissues examined. Each of the four valves that have been re- moved has been cultured for bacterial and fungal pathogens. No bacteriological or histological evidence of infective endocarditis has been recognized. Comment In order to provide a valved conduit to bypass obstruction between the right ventricle and pulmonary artery, many different techniques and materials have been utilized. Our initial experience with freeze-dried aortic homografts was disappointing [131. We utilized aortic homografts in 22 patients with right ventricular outflow tract obstruction. Calcification of the valve developed in 10 patients within the first year postoperatively. Three of the patients in our present study group initially had aortic homografts, but severe obstruction, associated with gross calcification of the valve cusps, subsequently developed. However, each now has

4 440 The Annals of Thoracic Surgery Vol 31 No 5 May 1981 Fig 1. (Patient 2.) Aortic aspect of a 20 mm xenograft value removed from a 7-year-old girl 1.5 years after implantation in the right ventricular outflow tract. had a Hancock prosthesis for 2 to 4 years and has been free from complications. Because of our suboptimal results with the freeze-dried homografts, we have utilized only glutaraldehyde-preserved porcine xenografts in the right ventricular outflow tract since The experience at the Mayo Clinic [141 with homografts prepared by chemical sterilization and irradiation was similarly disappointing. In contradistinction, Ross and Somerville [151 reported excellent long-term results in 24 patients who underwent right ventricular outflow tract reconstruction with fresh antibiotic-sterilized homografts. In our 25 patients, a mean gradient of 10 mm Hg was demonstrated intraoperatively across the entire conduit. Yet during later hemodynamic studies in a randomly selected group of 20 patients, the right ventricle-pulmonary artery mean peak systolic gradient was 31 mm Hg. In all but 2 of these patients, a gradient greater than that measured intraoperatively developed across the conduit. In the 8 patients classified as having conduit failure (see Table), it appears that Patients 1, 2, 3, 5, 7, and 8 demonstrated failure of the actual valve. In Patients 4 and 6 there was a combination of porcine valve failure and dysfunction at the site of proximal or distal anastomosis of the conduit. Sometimes failure and dysfunction occurred at both sites. Patient 4, in fact, required reoperation because of a large pseudoaneurysm of the right ventricular outflow tract adjacent to the proximal conduit anastomosis. In the remaining 3 patients who have undergone reoperation (Patients 2, 3, and 7), indications included a gradient across the valve of 70 mm Hg in Patient 2,60 mm Hg in Patient 3, and 53 mm Hg in Patient 7. The remaining patients with conduit failure are currently awaiting surgical intervention. There were few problems at reoperation. Patient 2 was noted to have an obstructed right pulmonary artery at the site of the previously closed Waterston shunt. A resection of the stenotic area with an end-to-end anastomosis was used to relieve the obstruction. Patient 3 had an organized thrombus surrounding the conduit at the distal anastomosis, making dissection difficult, but no major complications were encountered.. In Patient 4, the pseudoaneurysm was resected and a small VSD was closed with direct sutures. Fibrous membranes were seen attached to the porcine valve. Patient 7 also required closure of a small residual VSD. The right ventricle was noted to be quite dilated at the time of reoperation. In each of these patients, clinical signs of conduit failure were subtle. No symptoms suggesting residual obstruction were elicited in any patient except Patient 8. Therefore, conduit failure was not identified until cardiac catheterization was performed at the times listed in the Table. However, all of the patients had auscultatory evidence of varying degrees of right ventricular outflow tract obstruction. The poor long-term performance of the valved conduit in this study group raises important questions about the durability of the prosthesis. There was clinical or hemodynamic evidence of late valve dysfunction in 10 of 33 surviving patients. This represents a 30% incidence of xenograft conduit failure in a 6-year follow-up. This number is considerably higher

5 441 Bisset et al: RVOT Reconstruction with Porcine Xenografts A B Fig 2. (A) Anteroposterior and (B) lateral chest roentgenograms demonstrating position of the xenograft in a child with minimal obstruction at the valve site. than the 12% failure rate of extracardiac conduits between the right ventricle and pulmonary artery reported by Geha and co-workers [ll]. However, gradients of greater than 40 mm were recorded across the porcine conduit in 45% of the children in the study group of Rocchini and colleagues [12]. In the absence of documented infection, the etiology for the higher incidence of xenograft failure in children is speculative at this time. We would agree with Geha and associates [ll] that the cause is probably multifactorial. The selection of conduit size is a difficult problem for the surgeon. It appears that no specific equation can be formulated to predict appropriate valve size since the selection is determined by the diameter of the distal pulmonary arteries as well as by the mediastinal capacity to accommodate the conduit. This capacity will depend on aortic size and position, thoracic anteroposterior diameter, and cardiac size. Since no patient with a xenograft conduit diameter greater than 20 mm (n = 10 patients) has shown clinical or hemodynamic evidence of major obstruction, we believe the use of oversized conduits might reduce the incidence of late obstructive complications. In children, because growth will parallel increases in stroke volume across a fixed valve orifice, a vicious cycle may develop whereby an increasing gradient across the valve will perpetuate a higher rate of valve fatigue. A second problem associated with longevity of the valve relates to angulation of the conduit. This may represent a particularly important factor in the high incidence of failure in our patient group. It has been well established that preservation of the geometric configuration of the valve to as nearly normal a valve position as possible is an important goal. In Figure 2, anteroposterior and lateral chest roentgenograms demonstrate the intracardiac position of the porcine reinforcement ring in a patient with minimal obstruction at the valve site. This ring arises near the expected position of the normal pulmonary valve, thus preserving near-laminar flow through the valve and conduit. In contrast, the roentgenograms in Figure 3 represent similar views in a patient with major obstruction of the xenograft. There is extensive perpendicular angulation of the porcine ring. The abnormal hemodynamic stress on the valve leads to

6 442 The Annals of Thoracic Surgery Vol 31 No 5 May 1981 A B Fig3. (Patient 1.) (A) Anteroposterior and (B) lateral experiments by Bajpai [171 have documented chest roentgenograms demonstrating position of the high antigenicity of the glutaraldehyde-treated xenograft in a child with major obstruction across the xenograft. xenograft in rabbits. The well-developed immune system in children might explain the increased rate of valve deterioration. However, in thrombus formation with subsequent fibrosis the 4 patients in whom the valve was replaced on the inferior portion of the conduit where the with another porcine xenograft, there has been velocity of blood flow is decreased. Mustard no clinical evidence of obstruction or incompeand associates [16] demonstrated in vivo me- tence. Since none of these children has been rechanical factors relating to thrombosis in catheterized, it will be necessary to document silicone vessel models in pigs. In the highly an- those findings in the future. However, we gled branches, thrombus formation occurred would expect markedly accelerated valve dewhere angulation and, thus, turbulence were struction in the reimplanted valves if a hostmaximal. This turbulence necessitates a greater versus-graft reaction brought about the damage pressure difference across the area in order to to the initial implant. force a given volume of blood through the con- Intraoperative pressure measurements have duit, thus leading to further valve stress. The been of little value in predicting development superior portion of the conduit and xenograft of late valve dysfunction. Patients characterized are subjected to a "jet" effect, thereby produc- as having dysfunction demonstrated hemoing collagen damage to the valve and sub- dynamic criteria indicating xenograft failure sequent tissue degeneration. Similar problems within 2 years of valve placement. Since all have also been seen relating to lateral angula- these patients (except 1 with gross valve intion, specifically when the conduit must be competence associated with severe exercise inplaced far to the left in an attempt to avoid a tolerance) were relatively asymptomatic, it aplarge anteriorly displaced aorta or to prevent pears necessary to recatheterize all patients at sternal compression of the conduit. least 1 to 2 years postoperatively, even in the The third problem may relate to immuno- absence of clinical symptomatology. Although logical factors. Although immunogenicity of the no firm guidelines can be established as to the xenograft has been kept to a minimum, recent prerequisites for reoperation, we believe that a

7 443 Bisset et al: RVOT Reconstruction with Porcine Xenografts right ventricle-pulmonary artery peak systolic gradient greater than 40 mm Hg, a right ventricular peak systolic pressure greater than 80 mm Hg, or clinical and hemodynamic evidence of gross prosthetic valve incompetence would be sufficient evidence to recommend conduit replacement. It is to be hoped that this will prevent progressive dilatation, hypertrophy, and subsequent fibrosis of the compromised right ventricles. In the patients with late valve failure, it was difficult to implicate a residual shunt as a possible cause for the excessive valve fatigue, since all but 1 had intact ventricular septa, and 1 additional patient had only a small atrial septa1 defect. In the 2 patients with gross xenograft incompetence (not catheterized), clinical evaluation suggests that the ventricular septum is intact in both. Some optimistic results have been realized with the utilization of the valve-bearing xenograft conduit. None of the 33 survivors has been maintained on anticoagulant therapy, and we have recognized no thromboembolic complications. This has proved to be one very distinct advantage of the bioprosthetic valve. Also, we have seen no late deaths, including the group undergoing reoperation. There has not been a single documented episode of endocarditis and, specifically, we have seen no infections involving the conduit. Valved external conduits continue to be necessary in certain forms of complex congenital heart disease, and although surgical and valvehandling techniques have improved vastly in recent years, substantial problems still exist. The surgical risks associated with anatomical restoration of right ventricle-pulmonary artery continuity remain costly. Yet, even with the significant morbidity, reparative operations offer many advantages over less risky palliative procedures. There are no attendant hazards related to late repair of systemicpulmonary artery shunts, and there is virtually no risk of pulmonary hypertension developing secondary to the shunt procedures. However, the problems with residual gradients, valve deterioration, and the uncertainty of long-term performance of the porcine valve necessitate important management decisions. Therefore, we recommend caution in utilizing the porcine xenograft for reconstruction of the right ventricular outflow tract in children when an alternative surgical approach is available. Although the use of over-sized conduits may reduce the risk of long-term dysfunction of the valves, it appears that a valved conduit with greater longevity is needed for use in children. References 1. Ross DN, Somerville J: Correction of pulmonary atresia with a homograft aortic valve. Lancet 2:1446, Beach M: Discussion of Wallace et a1 [14] 3. Barratt-Boyes 8, Roche A, Agnew TM, et al: Homograft valves. Med J Aust 2:Suppl1:38, Wallace RB: Tissue valves. Am J Cardiol 35:866, Nonvood WI, Freed MD, Rocchini AP, et al: Experience with valved conduits for repair of congenital cardiac lesions. Ann Thorac Surg 24:223, Stinson EB, Griepp RB, Oyer PE, Shumway NE: Long-term experience with porcine aortic valve xenografts. J Thorac Cardiovasc Surg 7354, Mary DAS, Dakrash BC, Catahpole RQ, et al: Tissue valves in the mitral position: five years' experience. Br Heart J 37:1123, Yarbrough JW, Roberts WC, Reis RL: Structural alterations in tissue cardiac valves implanted in patients and in calves. J Thorac Cardiovasc Surg 65:364, Davila JC, Magilligan DJ Jr, Lewis JW Jr: Is the Hancock porcine valve the best cardiac valve substitute today? Ann Thorac Surg 26:303, Oyer PE, Stinson EB, Reitz BA, et al: Long-term evaluation of the porcine xenograft bioprosthesis. J Thorac Cardiovasc Surg 7833, 1979 Geha AS, Laks H, Stansel HC, et al: Late failure of porcine valve heterografts in children. J Thorac Cardiovasc Surg 78:351, 1979 Rocchini AP, Rosenthal A, Keane JF, et al: Hemodynamics after repair with right ventricle to pulmonary artery conduit. Circulation 54:951, 1976 Kaplan S, McKinivan CE, Helmsworth JA, et al: Complications following homograft replacement of the right ventricular outflow tract. Ann Thorac Surg 18:250, 1974 Wallace RB, Londe SP, Titus JL: Aortic valve replacement with preserved aortic valve homografts. J Thorac Cardiovasc Surg 6744, 1974 Ross D, Somerville J: Homograft reconstruction of right ventricular outflow tract in pulmonary atresia: late results. Br Heart J 38:316, 1976 Mustard JF, Murphy EA, Rowsell HC, Downie HG: Factors influencing thrombus formation in vivo. Am J Med 33:621, 1962 Bajpai PK: Discussion of Geha et al [ll]