AIIograft reconstruction of the right ventricular outflow tract

Transcription

1 Eur J Cardio-thorac Surg (1996) 10: Springer-Verlag 1996 T. P. Willems A. J. J. C. Bogers A. H. Cromme-Dijkhuis E. W. Steyerberg L. A. van Herwerden R. B. Hokken J. Hess E. Bos Received: 17 October 1995 Accepted: 8 January 1996 Presented at the Ninth Annual Meeting of the European Association for Cardiothoracic Surgery, Paris, France, September, 1995 T. E Willems ([])1. A. J. J. C. Bogers. L. A. van Herwerden - R. B. Hokken - E. Bos Department of Cardio-Pulmonary Surgery, University Hospital Sophia-Dijkzigt and Erasmus University, Rotterdam, The Netherlands A. H. Cromme-Dijkhuis. J. Hess Department of Pediatric Cardiology, University Hospital Sophia-Dijkzigt and Erasmus University, Rotterdam, The Netherlands E. W. Steyerberg Department of Public Health, University Hospital Sophia-Dijkzigt and Erasmus University, Rotterdam, The Netherlands Present address: Department of Thoracic Surgery, Bd 156, Dr. Molewaterplein 40, NL-3015 GD Rotterdam, The Netherlands AIIograft reconstruction of the right ventricular outflow tract Abstract Objective. Evaluation of allograft reconstruction of the right ventricular outflow tract (RVOT)_ Methods. From 1986 to April 1995, 201 allografts (146 pulmonary, 55 aortic) were implanted in 189 patients for conduit reconstruction of the RVOT in congenital heart disease or in the pulmonary autograft procedure. The mean age at allograft implantation was 16 years (range 2 weeks-54 years). The primary diagnoses of these patients were truncus arteriosus (n = 19, 10%), transposition of the great arteries (TGA) with ventricular Septal defect (VSD) and pulmonary atresia (PA) or stenosis (PS) (n = 14, 7%), PA with VSD (n=26, 14%), PA or PS with intact septum (n = 7, 4%), tetralogy of Fallot (n =44, 23%), corrected TGA with PA or PS (n = 11, 6%), tricuspid atresia (n=9, 5%), aortic valve pathology for pulmonary autograft procedure (n = 55, 29%), and miscellaneous (n = 4, 2%)_ The allograft implantation was a reoperation in 54 patients (29%). Results. The mean follow-up was 2.5 years (range 4 weeks-9 years). Six patients died in hospital (3.2%). Patient survival at 5 years was 91% (95% CL 86-95%). Freedom from all valve-related events (2 deaths, 17 reoperations, one endocarditis), as determined during reoperation or autopsy at 5 years was 78% (95% CL 65-86%). Freedom from structural allograft failure was 83% (2 deaths, 12 reoperations, 95% CL 70-90%). Allografts implanted for congenital right heart defects failed earlier than allografts used for pulmonary autograft procedures (P= 0.05). Aortic allografts showed structural failure more often than pulmonary allografts (P = 0.05). There were more valverelated events in patients of a younger age at implantation (P= 0.02) and in those allograft valves from younger donors (P = 0.004). Conclusions. Allograft RVOT reconstruction is an adequate surgical therapy. The allograft should preferably be pulmonary. A younger age at implantation is a risk factor for allograft failure. Donor age may be a thus-far underestimated risk factor for allograft degeneration. [Eur J Cardio-thorac Surg (1996) 10: ] Key words Allograft - Autograft - Right ventricular outflow tract

2 610 Introduction The use of allografts for reconstruction of the right ventricular outflow tract (RVOT) was first described in 1966 [13]. Since then, and especially since the development of preservation techniques, the use of allografts has become widespread. Nevertheless, donor shortage may still be a limiting factor in the use of allografts. At present, the pulmonary allograft appears to be the conduit of choice for reconstruction of the RVOT in congenital heart disease or in the pulmonary autograft procedure [1, 2, 7, 10, 11, 14]. The pulmonary allograft has shown less degeneration as compared to the aortic allograft [ 1, 2]. However, in younger patients an accelerated rate of degeneration is described for aortic as well as pulmonary allografts, possibly due to an enhanced immunologic mechanism [1, 2]. Because a variable rate of allograft dysfunction has been reported, with freedom from valve replacement or valverelated death ranging from 70 to 95% at 5 years [1, 2, 4], we examined our experience with allograft valved conduits in children and adults to classify further the possible risk factors for allograft failure and to contribute to the discussion on allograft application in cardiac surgery. Material and methods From August 1986 to April 1995, 201 allograft valves (146 pulmonary and 55 aortic) were implanted in 189 patients for RVOT reconstruction at the University Hospital Sophia-Dijkzigt, Rotterdam. The allografts were utilized as part of RVOT reconstruction in congenital heart disease or in the pulmonary autograft procedure. The primary diagnoses of the 189 patients are given in Table 1. The four most frequent diagnoses were aortic valve pathology for pulmonary autograft procedure in 55 patients (29%), tetralogy offallot in 44 patients (23%), pulmonary atresia (PA) with ventricular septal defect (VSD) in 26 patients (14%) and truncus arteriosus in 19 patients (10%). Two patients with PA and VSD and two patients with tetralogy of Fallot had a double outlet right ventricle. Two other patients with tetralogy of Fallot had absent pulmonary valve syndrome. The miscellaneous group comprised four patients. Two of these patients had pulmonary stenoses (PS) after previous Mustard operations, Another patient had aortic atresia with an adequate left ventricle in whom the left ventricle was tunneled via the pulmonary valve, which was used to reconstruct the aorta. The allograft was used to reconnect the right ventricle to the pulmonary artery. One patient had PS after an arterial switch operation for transposition of the great arteries (TGA). Our results after allograft implantation in patients who had previously undergone intracardiac repair for PS or PA have recently been described [6]. In this series allografts which were used as part of the RVOT reconstruction in congenital heart disease were classified as extra-anatomical in position and allograft implantations as part of the pulmonary autograft procedure were classified as anatomical. Fifty-four patients (29%) had undergone 76 prior cardiac operations for repair of the RVOT. The actual pathology of 65 previous RVOT reconstructions at the time of allograft implantation is given in Table 2. Eleven RVOTs were reconstructed more than once before the allograft implantation (with xenograft n = 5, transannulair patch n = 2, valvulotomy n = 4). Eleven patients received a second allograft and one pa- Table 1 Primary diagnosis of 189 patients with 201 allografts implanted (PA pulmonary atresia, PS pulmonary stenosis, TGA transposition of the great arteries, VSD ventricular septal defect, DORV double outlet right ventricle) No. of patients No. of allografts PA, VSD 26 (14%) 29 (11%) with DORV 2 3 PA or PS with intact septum 7 (4%) 7 (3%) Tetralogy of Fallot 44 (23%) 46 (23%) with DORV 4 5 with absent pulmonary valve 3 3 Truncus arteriosus 19 (10%) 22 (11%) TGA with PA or PS, VSD 14 (7%) 16 (8%) CorrectedTGAwithPAorPS 11 (6%) 13 (6%) Tricuspid atresia 9 (5%) 9 (4%) Aortic valve pathology 55 (29%) 55 (27%) (autograft procedure) Miscellaneous 4 (2%) 4 (2%) Total Table 2 Actual pathology at time of allograft implantation after previous RVOT reconstruction Pathology Transannulair patch dysfunction Allograft conduit dysfunction Xenograft conduit dysfunction Prosthetic non-valved conduit dysfunction Valve failure after valvulotomy Allograft monocusp dysfunction Pericard valved conduit dysfunction Total 65 No tient received a third at our department. The mean age at the time of allograft implantation was 16 years (median 12.9 years, range 14 days -54 years). Fifty-six patients (30%) were aged less than 5 years at the time of allograft implantation (Fig. 1). The mean weight of the patients was 40 kg (median 38_6 kg, kg). One hundred and seventy-two cryopreserved allograft valves were supplied by the Heart Valve Bank, Rotterdam, through Bio Implant Services, The Netherlands, The technique for preparation and storage has been described previously [15]. Twenty fresh allografts were obtained from the National Heart Hospital, London. The Hospital Clinic I Provincial, Barcelona, Spain, and the Karolinska Homograft Bank, Stockholm, Sweden, supplied 7 and 2 cryopreserved allografts, respectively. The choice of conduit size was made before operation on the basis of the patient's body surface area [8], In infants aged less than 1 year, an effort was made to insert allografts somewhat larger than predicted from the tables of normal valve sizes. No attempt was made to achieve ABO blood type matching. Pulmonary allografts were mostly used in congenital heart disease requiring a reconstruction of the pulmonary arteries. The mean internal diameter of allografts was 22 mm (range mm). The mean donor age was 29.3 years (range 1-55 years).

3 6ll Table 3 Early mortality Causes No. e- ~J 0..Q E e age (years) Fig. 1 Distribution of age at time of allograft implantation of 189 patients The mean follow-up was 2.5 years (range 4 weeks-9 years). With the exception of 12 patients living abroad, all patients are followed annually, including physical examination and echocardiography. At this moment the results of the echocardiographic examinations of the RVOT in different anatomic situations show insufficient reliability for end point analysis. However, echocardiographic data were used to decide on the distinction between valve-related and nonvalve-related events. Postoperative allograft events were described according to Edmunds [3]. The nature of valve deterioration was determined at reoperation or autopsy. In order to distinguish between structural and non-structural allograft failure, all events were extensively and repeatedly discussed within the project team including the responsible surgeon. The decision to replace an allograft was based on symptoms attributed to pulmonary regurgitation or stenosis, and confirmed by echocardiography or cardiac catheterization. Operative technique Surgical procedures were performed using standard cardiopulmonary bypass with moderate hypothermia, myocardial protection with crystalloid cardioplegia (St. Thomas solution) and topical cooling. In patients who did not require associated intracardiac procedures, the heart was kept beating and the reconstruction was made without cross-clamping the aorta. The allograft was prepared according to protocol [15]. hnplantation of the allograft was carried out using the interposition technique, with running polypropylene sutures for the distal anastomosis_ Bifurcated reconstructions using donor pulmonary confluence were performed in 16 procedures. In 13 procedures allograft 0z = 4), pericardial (n = 3) or prosthetic patches (n = 6) were used for distal extension. The proximal anastomosis was also made with running polypropylene sutures. A proximal extension of the conduit was considered indicated in 50 operations and this was made using an allograft patch (n = 12), or the anterior mitral valve leaflet (n = 17) in cases of aortic allografts. Pericardial patches and prosthetic patches were used in 14 and 7 cases, respectively. In general, attempts were made to position the allograft away from the sternum to prevent compression or distortion. Statistical analysis Survival and freedom from allograft failure were analyzed according to the method of Kaplan-Meier [9]. The survival of a patient start- Hypoxic encephalopathy 1 Arrhythmia 1 Heart failure 1 Respiratory insufficiency 1 Pulmonary thromboembolism 1 Mediastinitis 1 Total 6 Table 4 Late mortality Causes Non-valve-related: Autograft failure, juvenile rheumatoid arthritis Restrictive cardiomyopathy Right heart failure, pulmonary hypertension Endocardial fibroelastosis Arrhythmia Valve-related: Allograft regurgitation, arrhythmia Allograft obstruction, right heart failure Total 8 ed at the time of allograft implantation and ended at death (event) or at the last follow-up (censoring). The analysis of allograft survival started at implantation and ended with allograft failure (reoperation, valve-related death) or at the last follow-up. The differences between survival curves were evaluated using the log-rank test. The Cox proportional hazard model was used to identify the contribution of risk factors for mortality and allograft failure, while correcting for correlation between risk factors. The analyzed factors included patient age at operation, extra-anatomical and anatomical allograft position, type of allograft, donor age and allograft diameter. Results Six out of 189 patients (3.2%) died early postoperatively. All early deaths were due to non-valve-related causes. The primary causes of early deaths are shown in Table 3. There were eight late deaths in the remaining 187 patients (Table 4). Two of the eight late deaths were valve-related. One of these patients died 3 months after implantation due to sustained ventricular tachycardia. An echocardiogram performed 3 weeks before death showed dilation of the right ventricle and severe allograft regurgitation. The other patient died due to acute right ventricular failure with a severe obstruction at the proximal anastomosis. The remaining late deaths were considered to be non-valve-related. At 5 years the overall patient survival rate was 91% (95% CL %)_ Statistical analysis identified none of the factors No.

4 612 mentioned in the section materials and methods as a risk factor for early or late mortality. Seventeen reoperations for allograft dysfunction were performed in 16 patients. Twelve allograft failures were diagnosed as structural, on the basis of operative findings. Eleven of these reoperations were performed for severe allograft obstruction and one for regurgitation. In this regurgitant allograft a false aneurysm in one sinus was found at operation and confirmed microscopically. Five of the 11 obstructed allografts showed a conspicuously calcified conduit wall and had leaflets which were macroscopically fibrotic. There were four obstructions located at the proximal anastomosis and extending into the leaflets. Another two obstructed allografts in Fontan circulation showed calcified leaflets, which in one could be removed and in the other could be enlarged by a patch. Five allograft failures were diagnosed as non-structural. One of these required reoperation for stenosis in the distal Dacron extension and four reoperations were necessary for stenosis of the distal anastomosis, two of which were reconstructed by allograft patch enlargement. Thirteen allografts were explanted and these conduits were replaced by another allograft. One allograft was replaced in another hospital. One patient required three allografts for severe obstruction. During follow-up one case of endocarditis was diagnosed, which was successfully treated medically. In total, 20 valve-related events occurred in the follow-up of 201 allografts (2 deaths, 17 reoperations, 1 endocarditis); the overall freedom from valve-related events was 78% (95% CL 65-86%) at 5 years. Freedom from reoperation for structural failure was 83% (95% CL 70-90%) at 5 years (Fig. 2). 03 t- 100% > m 9o% 80% o -~ 70% E o E o 60% "o Q Structural failure... All events AIIografts at risk:... 83% [ 78% i i i i i Time since implantation (years) Fig. 2 Freedom from allograft valve-related events. The dotted line indicates structural failure (5 years, 95% CL 70-90%), the solid line indicates all valve-related events including structural failure, nonstructural failure and endocarditis (5 years, 95% CL 65-86%) Freedom from structural failure of the aortic allografts was significantly lower than that of pulmonary allografts, as shown in Table 5. The freedom from structural failure after a pulmonary autograft procedure (anatomical allograft position) was also significantly lower than after RVOT reconstruction in congenital heart disease (extraanatomical allograft position). Younger patient age and a smaller allograft diameter at the time of implantation, calculated as continuous variables, were risk factors for structural failure. Younger donor age was a strong predictor of structural failure at 5 years. The primary diagnosis of congenital heart disease, as grouped in Table 1, was not a significant predictor of structural failure. The Cox regression analysis indicated that patient age, extra-anatomical and anatomical allograft position, type of allograft, allograft diameter and donor age remain associated with failure, when corrected for their correlation. Discussion The use of allografts for reconstruction of the RVOT is widespread in congenital heart disease and in the pulmonary autograft procedure. The pulmonary allograft is currently the conduit of choice [1, 2]. Furthermore~ the donor bifurcation may be desirable for reconstruction of the central pulmonary arteries. In this series of RVOT reconstruction with allografts, the superior late results of pulmonary allografts is confirmed. Pulmonary allografts were mainly used and several risk factors for valve degeneration were identified_ These were known risk factors, such as patient age at the time of implantation, type and diameter of allograft, but we also found a strong influence of donor age on allograft failure. Recent reports have identified a disquieteningly high failure rate of the aortic allograft in the RVOT at 5 years [ 1, 2]. Albert et al. have reported a freedom from valve replacement or valve-related death due to aortic allograft failure at 5 years of 76% compared to pulmonary allograft failure of 96% [1]. Bando et al. have recently published similar results with a freedom from valve replacement or valve-related death due to aortic allograft failure at 5 years of 70%, compared to 94% in pulmonary allografts [2]. In our series we found a striking similarity in results with a freedom from reoperation for allograft failure or valve-related death of 73% for aortic allografts and 93% for pulmonary allografts at 5 years. Other investigators have already expressed their concern about aortic allograft implantation, especially in younger patients [1, 2]. This concern is further heightened by the data of this study; there is an accelerated degeneration, especially in aortic allografts implanted in young patients. A striking new finding was a higher incidence of the structural failure of allografts obtained from younger do-

5 613 Table 5 Freedom from all valve-related events and freedom from structural failure at 5 years stratified for risk factors (189 patients, 201 allografts) Analyzed factors Freedom from 95% CL P value Freedom from 95% CL P value all events structural failure Age at implantation: <1 year (n=29) >_1 year (n = 172) Age inverse (=l/age)* Allograft position: Extra-anatomical (n = 146) 73 % Autograft procedure (n = 55) 97 % Type of allograft: Aortic (n=55) 66% Pulmonary (n = 146) 88% Preservation: Fresh (n = 20) 81% Cryopreservation (n = 18 l) 77 % Allograft diameter: mm (n=96) 73% mm (n= 105) 91% Diameter inverse (= t/diameter)* Donor age #: 1-29 years (n=95) 72% years (n = 94) 82% Donor age inverse (= 1/donor age)* * P values calculated with continuous variables (inverse transformation) # 12 missing values nors [1, 2]. In contrast, O'Brien has reported that older donor age is associated with a greater risk of valve degeneration [12]. No adequate explanation for our findings is available yet, though one might speculate that the younger donor age may be attended by a higher cellular content in the allograft leaflets and wall, which may predispose to a more pronounced immunologic reaction. Immunologic reactions may also be related to allograft preservation techniques. Histologic examination of cryopreserved allografts showed a diminished presence of valve leaflet endothelial cells, which is possibly associated with less immunostimulation. Endothelial cells present on the surface of fresh allografts could be responsible for initiating the immunologic response [5]. Therefore, a cryopreserved allograft might be preferable to a fresh allograft for reconstruction of the RVOT, especially in younger patients. In this regard the further development of preservation techniques which result in reduced immunogenicity is desirable. Blood group and HLA-DR matching is also an option for reducing the immunologic reaction of the recipient. However, matching between graft donor and recipi- 78% (66-86%) 83% (72-90%) (59-83%) } 79% (66-88%) } (81-99%) % (46-81%) } 73% (52-85%) } (58-83%) } 79% (65-88%) } oo2 (54-94%) % (53-96%) ent will lengthen heart valve bank waiting lists, and donor shortage is still a seriously limiting factor. In our view immunosuppression is not indicated, although in the future a short and early induction-immunotherapy may be considered with mismatched allografts. Allograft failures which were diagnosed during reoperations as structural had different macroscopic aspects. Obstructions located at the proximal anastomosis with severe calcification of the valve leaflets were observed. There were also allograft failures showing severe degeneration of the conduit with macroscopic fibrotic leaflets. Because of the small numbers of allograft failures, we could not establish a relationship between macroscopic aspects and recipient or donor age. It would be preferable for future research that the macroscopic aspects of the explanted valve leaflets and conduit be described separately. Therefore systematic follow-up of allografts necessitates additional guidelines for reporting allograft conduit failure based on the guidelines of Edmunds [3]. In summary, allograft RVOT reconstruction is an adequate surgical therapy. Allograft failure is higher in aortic

6 614 allografts, but is still acceptable. Allografts in anatomical positions do better than in extra-anatomical positions. In younger patients it may be desirable to limit the risk factors and to implant a pulmonary, cryopreserved, matched allograft. Donor shortage is, however, a limiting factor. Do- nor age may be a thus-far underestimated risk factor for allograft degeneration. Aeknnwledgements We gratefully acknowledge the excellent secretarial support of Mrs. Ada Matzer. This research was financially supported by the Netherlands Heart Foundation (grant ). References l. Albert JD, Bishop DA, Fullerton DA, Campbell DN, Clarke DR (1993) Conduit reconstruction of the right ventricular outflow tract: lessons learned in a twelve-year experience. J Thorac Cardiovasc Surg 106: Bando K, Danielson GK, Schaff HV, Mair DD, Julsrud PR, Puga FJ (1995) Outcome of pulmonary and aortic homografts for right ventricular outflow tract reconstruction. J Thorac Cardiovasc Surg 109: Edmunds LH, Clark RE, Cohn LH, Miller DC, Weisel RD (1988) Guidelines for reporting morbidity and mortality after cardiac valvular operation. J Thorac Cardiovasc Surg 96: Hawkins JA, Bailey WW, Dillon T, Schwartz DC (1992) Midterm results with cryopreserved allograft valved conduits from the right ventricle to pulmonary arteries. J Thorac Cardiovasc Surg 104: Discussion Dr. A. Piwniea (Paris, France): You show very well that the allografts failed because of age, but could you comment on how they failed? Did you analyze the type of failure and did you look at them? Dr. Willems: Yes, we looked at all our explantations, and there is a difference in macroscopic aspects of the explanted allografts. You can see very calcified leaflets together with a calcified conduit but, on the other hand, you can also see a calcified conduit with fibrotic leaflets. We checked if there was any relation with age but, because the number of explants is low, we were not able to find such a relation. Perhaps it would be wise to make a difference between valve degeneration and conduit degeneration in further research on allograft failure. 5. Hoekstra F, Knoop C, Aghai Z, Jutte N, Mochtar B, Bos E, Weimar W (1995) Stimulation of immune competent cells in vitro by human cardiac valve derived endothelial cells. Ann Thorac Surg 60: Hokken RB, Bogers AJJC, Spitaels SEC, Hess J, Bos E (1995) Pulmonary homograft insertion after repair of pulmonary stenosis. J Heart Valve Dis 4: Hokken RB, Bogers AJJC, Taams MA, Willems TP, Cromme-Dijkhuis AH, Witsenburg M, Spitaels SEC, Van Herwerden LA, Bos E (1995) Aomc root replacement with a pulmonary autograft. Eur J Cardiothorac Surg 9: Hopkins RA (1989) Cardiac reconstructions with allograft valves. 1st edn. Springer, New York, pp Kaplan EL, Meier P (1958) Nonparametric estimation from incomplete observations. J Am Stat Assoc 53: Kouchoukos NT, Davilla-Roman VG, Spray TL, Murphy SF, Perrillo JB (1994) Replacement of the aortic root with a pulmonary autograft in children and young adults with aortic valve disease. N Engl J Med 330:1-6 Dr. D. Metras (Marseille, France): I want to congratulate Dr. Willems on your results, and on a very nice presentation interestingly and unfortunately showing that the homograft is not a permanent solution. I want to ask a question about your indication for implanting a valve homograft in tetralogy of Fallot. Twenty-five percent of your homografts were implanted for a tetralogy of Fallot. I found the proportion of your Fallot tetralogies that had an homograft implanted was high. Could you tell us a little about what were the indications. Dr. Willems: Most of the patients with tetralogy of Fallot had one or more repairs before they had a homograft implantation. They had been operated before with a transannulat patch or xenograft, and the actual indication was pulmonary regurgitation or stenosis. 11. McGrath LB, Gonzalez-Lavin L, Graf D (1988) Pulmonary homograft implantation for ventricular outflow tract reconstruction: early phase results. Ann Thorac Surg 45: O'Brien ME Stafford EG, Gardner MAH, Pohlner PG, McGiffin DC (1987) A comparison of aortic valve replacement with viable cryopreserved and fresh allograft valves, with a note on chromosomal studies. J Thorac Cardiovasc Surg 94: Ross DN, Somerville J (1966) Correction of pulmonary atresia with a homograft aortic valve. Lancet II: Schoof PH, Cromme-Dijkhuis AH, Bogers AJJC, Thijssen EJM, Witsenburg M, Hess J, Bos E (1994) Aortic root replacement with pulmonary autograft in children. J Thorac Cardiovasc Surg 107: Thijssen HJM, Bos E, Konertz W, Van Suylen RJ, De By TMMH (1992) Kryokonservierung humaner Spenderherzklappen in der Herzklappenbank in Rotterdam. Z Herz-, Thorax-, Gef~igchir 6 [Suppl 1]:49-55 Dr. B. Messmer (Aachen, Germany): You didn't say anything about the procurement of your allografts in storage, Did you procure them yourself, did you store them yourself, did you treat them yourself, and was the same procedure held over the whole time? Dr. Willems: We implanted 201 allografts, 181 were cryopreserved and 20 were fresh. Most of the allografts, 172, were supplied from the Heart Valve Bank, Rotterdam. We have a standardized heart valve bank. Analysis in our series showed no difference between fresh and cryopreserved allografts.

7 615 Dr. T. David (Toronto, Ontario. Canada): I'm interested in those patients who had allograft valves in the anatomical position; you had 50, or go. How many of those failed and what were the ages of the patients? Dr. Willems: In our series there were 55 patients with anatomical position of the allograft in the autograft procedure, with an age range of 4 months-52 years. One of the anatomically positioned allograft failed at an age of 36 years. The age distribution of the patients with allografts in extra-anatomical positions is almost the same as that of the patients with allografts in the puhnonary position for the autograft procedure. Dr. David: In patients older than 15 years of age the allograft seldom fails in the first decade when a valve larger than 25 mm is implanted. My main concern is regarding the Ross procedure. This is an extensive operation for aortic valve disease, and if the pulmonary allograft is going to fail within 8 or 10 years, it almost defeats the purpose of the operation. The information regarding the fate of allografts in the pulmonary position after a Ross procedure remains unknown. Dr. W. Daenen (Leuven, Belgium): We have had experience now in about 140 homografts in the right-sided position, and we have seen that in about 10-15% we had fever of unknown origin where all possible etiologies were ruled out. Have you had the same experience, and if this is true, do you think that is an immune response of the host? Dr. Willems: Actually we don't see many patients with fever after right ventricular reconstruction with an allograft. But we see fever after allograft implantation in the left-sided position. We don't have a clinical solution for this problem yet. Maybe there is an immunologic response after allograft implantation.