The Fate of Small-Diameter Homografts in the Pulmonary Position

The Fate of Small-Diameter Homografts in the Pulmonary Position Nicodème Sinzobahamvya, MD, Jutta Wetter, MD, Hedwig C. Blaschczok, MD, Mi-Young Cho, MD, Anne Marie Brecher, MD, and Andreas E. Urban, MD Department of Paediatric Cardio-Thoracic Surgery, Deutsches Kinderherzzentrum, Sankt Augustin, Germany Background. Limited durability is expected for small homograft valves that are used to correct congenital cardiac disease. Methods. All 76 homograft valves with an internal annulus diameter ranging from 8 to 13 mm that were implanted from 1987 through 2000 in the pulmonary position were retrospectively analyzed. In each case, homograft size was normalized to the patient s body surface area: z-value. For 93% (14 of 15) of the 8 to 9 mm grafts, z was less than 2. For 56% (5 of 9) of the 10 mm grafts and 98% (51 of 52) of the 11 to 13 mm allografts, z was greater than 2. Survival and freedom from complications were estimated by the Kaplan-Meier method. Homograft failure was defined as homograft replacement or late death; significant dysfunction, as homograft obstruction with an echo-doppler gradient greater than 50 mm Hg or grade III or IV valvular insufficiency. The log-rank test was used to compare outcomes. Results. Seven patients died early after operation; three, late. Survival was 86.5% 3.8% at 1 year and remained stable during the succeeding years. Freedom from failure for all homografts was 90.6% 3.7%, 71.8% 6.9%, and 61.8% 9.0% at 1, 5, and 10 years, respectively. Corresponding freedom from significant dysfunction was 87.6% 4.1%, 51.2% 7.4%, and 10.1% 8.3%. The smaller homografts (z less than 2) failed and deteriorated faster (p < 0.0001): only 32.1% 13.0% were still functioning at 24 months. The larger grafts (z at least 2) retained function for the first 4 years, and 73.7% 10.4% had not yet failed at 10 years. Conclusions. Smaller (z less than 2) homografts (the great majority of 8 to 9 mm grafts) have to be replaced early, usually within 2 years of implantation. Larger (z at least 2) grafts (nearly all 11 to 13 mm grafts) show remarkable durability and are suitable valved conduits for establishing right ventricle to pulmonary artery continuity in neonates and young infants. (Ann Thorac Surg 2001;72:2070 6) 2001 by The Society of Thoracic Surgeons Small-diameter valved conduits used to repair congenital heart disease in neonates and young infants are not expected to have a long life, either because of accelerated degeneration or because of rapid outgrowth. Their failure leads to reoperation in the first years of life, somewhat tempering the benefits of early complete correction of those cardiac lesions. Since 1987, we have been implanting small homografts between the right ventricle and the pulmonary artery or arteries for total repair of various cardiac malformations in early infancy. We reviewed our experience in terms of lesions corrected, survival of patients, and durability and function of these small conduits. Patients and Methods Definitions Small-diameter homograft valves are those allografts with an internal annulus diameter ranging from 8 to 13 mm. The z-value or z-score of the homograft is given Accepted for publication July 30, 2001. Address reprint requests to Dr Urban, Department of Paediatric Cardio- Thoracic Surgery, Deutsches Kinderherzzentrum, Arnold-Janssen- Strasse 29, 53757 Sankt Augustin, Germany; e-mail: andreas.e.urban. md@t-online.de. by the following formula: Z (A B)/C, where A is the diameter of the implanted homograft, B the normal diameter of the pulmonary valve based on patient body surface area, and C the standard deviation around the normal dimension of the pulmonary valve. B and C refer to normal individuals of the same body surface area as the patient and are obtained using regression analysis and the equation of Rowlatt and associates [1]. Early survivors are patients who were discharged from the hospital and who survived at least 30 days. Homograft failure is defined as homograft replacement or late death; significant homograft dysfunction, as homograft obstruction causing a trans-homograft peak pressure gradient greater than 50 mm Hg on Doppler-echocardiography, or grade III or IV homograft regurgitation. Patients Diagnoses Between June 1987 and December 2000, 76 small homografts were implanted in 76 consecutive children who underwent operations involving reconstruction of the right ventricular outflow tract (RVOT). Table 1 shows the different types and sizes of the homografts used, and Table 2 indicates the corresponding homograft z-values. All homografts except four were cryopreserved, and those four were freshly sterilized by antibiotics. Ho- 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00 Published by Elsevier Science Inc PII S003-4975(01)03178-2

Ann Thorac Surg SINZOBAHAMVYA ET AL 2001;72:2070 6 SMALL HOMOGRAFTS 2071 Table 1. Size and Type of Homografts mm Size Median z-value Aortic (n 1) Pulmonary (n 2) Total n 8 0.9 2 5 7 9 1.4 6 2 8 10 2.3 5 4 9 11 3.1 5 4 9 12 3.7 11 12 23 13 4.0 13 7 20 Total 42 34 76 mografts were supplied by 10 different homograft banks, but most (n 36) came from a single institution; 42 were aortic and 34 were pulmonic grafts. Homograft z-scores ranged from 0 to 5.2, with a mean of 2.9 1.3 and a median of 3.0. For 93% (14 of 15) of the 8 to 9 mm grafts, z was less than 2. For 56% (5 of 9) of the 10 mm and 98% (51 of 52) of the 11 to 13 mm allografts, z was greater than 2. Most patients in this series (n 46) were female; the median age was 46 days (range 2 to 420 days); 23 were neonates, and 87% (66 of 76) were under the age of 120 days. Median weight was 3.3 kg (range 1.7 to 9.1 kg), with a body surface area ranging from 0.15 to 0.46 m 2 (mean 0.23 0.05 m 2 ). Of the 76 patients, 11 weighed less than 2.5 kg; 19 had a body surface area smaller than 0.20 m 2. The cardiac malformations summarized in Table 3 included truncus arteriosus (n 60), various forms of the tetralogy of Fallot (n 15), and one case of aortic atresia (Ross procedure). Surgical Technique Management Standard surgical techniques were used. Cardiopulmonary bypass time averaged 175 52 minutes; aortic cross-clamping, 82 24 minutes. Hypothermic circulatory arrest for thirty-eight patients lasted 41 18 minutes. St Thomas II cardioplegic solution was always infused, and beginning in 1993, modified ultrafiltration was applied in all patients. The choice of homograft Table 2. Homograft z-value Compared to the Size of Implanted Homografts Size of homograft mm Ranges of Homograft z-values 0 0.9 1 1.9 2 2.9 3 3.9 4 5.2 z-scores (n 1) (n 2) (n 3) (n 4) (n 5) n 8 5 2 0 0 0 7 9 2 5 1 0 0 8 10 1 3 5 0 0 9 11 0 0 5 4 0 9 12 0 0 5 10 8 23 13 0 1 0 9 10 20 Total 8 11 16 23 18 76 Table 3. Cardiac Lesions, Early and Late Mortality Diagnosis Number Early Deaths (%) Late Deaths Truncus arteriosus (simple) 53 0 (0) 1 Truncus arteriosus with IAA 7 5 (71.4) 1 Tetralogy of Fallot/PA/VSD 13 2 (15.4) 1 Absent pulmonary valve 2 0 (0) 0 syndrome Aortic atresia aorticoventricular 1 0 (0) 0 tunnel Total 76 7 (9.2) 3 IAA interrupted aortic arch; ventricular septal defect. PA/VSD pulmonary atresia/ depended to a great extent on availability. Our preferred conduit size is 1.5 times the diameter of the patient s pulmonary artery, calculated from the body surface area of the patient [2] (which in this study corresponds to a homograft z-value of 3.1 to 3.8). But this criterion was met on only 15 occasions. ABO blood group compatibility was not required. At operation, once cardioplegic cardiac arrest was established, the homograft was sutured distally to the pulmonary artery or arteries. The proximal anastomosis was performed on the beating heart. In seventy-one patients the implant was attached to the right ventricle by means of (1) a hood made of pericardium (n 28), (2) a polytetrafluoroethylene (PTFE) patch (n 17), (3) a mitral leaflet from the aortic homograft donor (n 10), (4) a homograft patch (n 4), or (5) a dacron tube extension (n 12). To facilitate chest re-entry, if reoperation became necessary, we routinely approximated the pericardium and covered the heart with an artificial membrane. In 12 cases, the sternum was closed secondarily. The median duration of postoperative endotracheal intubation for early survivors was 5 days. Follow-up: Evaluation of Homograft Function Echo-Doppler studies were used to evaluate heart and homograft function at least every 6 months for the first 2 years and once a year thereafter. The trans-homograft peak pressure gradient was calculated using the simplified Bernoulli equation: gradient 4V 2, where V is the peak instantaneous trans-homograft Doppler velocity. Homograft insufficiency was graded by mapping the dimensions of the regurgitation jet with pulsed or color flow Doppler echocardiography, analogous to the semiquantitative method described by Perry and colleagues for evaluation of aortic insufficiency [3]. If a problem was detected or suspected, cardiac catheterization with angiocardiography was performed. Right ventricular pressure at least 75% of the systemic pressure was considered an indication for conduit replacement regardless of the clinical status of the patient. Data Analysis Perioperative and early postoperative data were collected retrospectively. The actual follow-up assessments took

2072 SINZOBAHAMVYA ET AL Ann Thorac Surg SMALL HOMOGRAFTS 2001;72:2070 6 Fig 1. Kaplan-Meier estimates of freedom from homograft failure for all grafts over time. Vertical bars represent standard error of the mean (S.E.M.). place from May 2000 to January 2001. Probabilities of survival, or freedom from homograft failure, and of freedom from homograft dysfunction were estimated by the Kaplan-Meier method (see Fig 1). The following three parameters served as end points in our study: time of death, graft replacement, and first diagnosis of significant graft dysfunction. The probabilities are given as percent (%) SEM (standard error of the mean). The connecting line is stepwise for the graph depicting homograft failure. It is straight for homograft dysfunction. The log-rank test was used to calculate the statistical difference between two groups. The fate of the allografts was compared according to their z-scores. Other variables studied included age at operation (neonates versus older infants), type of graft (aortic versus pulmonary), and supplier of homograft (the institution that supplied the most homografts versus other homograft banks), and use (or not) of a dacron tube extension. The difference is considered statistically significant with a p value less than or equal to 0.05. Means SD and percentages with 95% confidence limits are given. Results Survival Seven patients died early, six of them within 24 hours after operation, for an operative and hospital mortality rate of 9.2% (3.8% to 18.1%). In no case was death attributable to homograft failure. Table 3 shows the operative risk per lesion. Follow-up for early survivors (median duration: 54 months) was 100% complete. Three patients died late: 3 months, 11 months, and 12 months after repair. Death resulted from heart failure caused by left ventricular outflow tract obstruction (n 1) or from persistence of pulmonary hypertension (n 2). Cardiac catheterization in these three patients showed no significant obstruction of the conduit; nevertheless, the valve was grade III incompetent in the third patient, a finding verified at post-mortem examination. Actuarial survival was 86.5% 4.0% at 1 year and remained constant thereafter. For patients with truncus arteriosus not associated with interrupted aortic arch, survival was 98.1% 1.9% from the thirteenth postoperative month. Homograft Failure Fourteen homografts were explanted 2 to 135 months after insertion (median, 24 months; mean, 40 38 months), four of them in other hospitals, with no operative deaths. The 10 patients who underwent reoperation in our unit had had cardiac catheterization 4 months (median) before operation. Graft failure was due to infection in two patients. In one, Candida albicans caused digestion of the homograft with pseudo-aneurysm formation that required emergency intervention 2 months after truncus arteriosus repair. In the other case infection was subacute to chronic and caused homograft obstruction that led to reoperation 60 months postoperatively. Right ventricular outflow tract obstruction was the indication for homograft replacement in 13 patients. Significant homograft dysfunction had been detected by echocardiography 2 to 105 months (mean: 21 29 months, median: 9 months) before graft replacement in these 13 patients. At the time of reoperation, the somatic growth in all patients was such that the z-scores of the original grafts were negative: median 2.3, mean 2.4 1.0 ( 0.87 to 4.1). The right ventricular pressures ranged from 75% to 200% of the systemic pressure (mean: 115% 36%, median: 104%). At reoperation, the main site of obstruction was at the proximal anastomosis (n 2), within the homograft itself (n 5), or at the distal anastomosis to the pulmonary arteries (n 2). The point of obstruction could not be clearly determined in four cases, suggesting outgrowth as the single etiology for obstruction. Other causative factors of RVOT obstruction were infection as reported above (n 1), probable graft degeneration (n

Ann Thorac Surg SINZOBAHAMVYA ET AL 2001;72:2070 6 SMALL HOMOGRAFTS 2073 Table 4. Probabilities (% SEM) of Freedom from Homograft Failure According to Homograft z-values Over Time Interval (months) Ranges of z-values and Probabilities of Freedom 0 0.9 1 1.9 2 2.9 3 3.9 4 5.2 12 100 56 17 100 95 4 100 24 50 25 42 17 100 95 4 100 48 50 25 42 17 92 8 95 4 100 60 50 25 28 16 79 14 82 13 100 120... 0 63 18 65 18... 4) (leaflet shrinkage, wall calcification, or sclerosis), unfavorable anatomy with retrosternal compression of the prosthesis (n 1), and mixed factors (n 3), with a combination of outgrowth, conduit degeneration, and peripheral stenosis of the right or left pulmonary artery. Freedom from explantation of the homograft and replacement by another conduit at 1, 5, and 10 years was 95.3% 2.7%, 75.5% 6.9%, and 65.0% 9.3%, respectively. Failure occurred in a total of seventeen grafts (leading to fourteen replacements and three late deaths). Figure 1 shows the freedom from failure over time for all valved conduits, and Table 4 details the freedom from failure according to the homograft z-scores. The freedom from failure for the grafts with 0 0.9 and 1 1.9 z-scores is not significantly different: p 0.23. It is also statistically similar for those with z-scores ranging from 2 to 5.2: p 0.30. Figure 2 compares the results of these two groups: z less than 2 versus z at least 2. The smaller normalized valve prostheses failed more rapidly ( p 0.0001): only 32.1% 13.0% were still functioning at 24 months. Almost all of the larger grafts retained their function for the first 4 years, and 73.7% 10.4% had not failed at 10 years. Replacement conduits included homografts (n 11) with a median size of 17 mm, heterografts (n 2: 18 mm and 22 mm), and a 20-mm valveless PTFE tube. Conduit z-scores ranged from 1.8 to 7.6 (mean: 4.4 1.5, median: 4.7). Homograft Dysfunction Fifty-two survivors have their original homografts. In sixteen of them, the trans-homograft peak pressure gradient is higher than 50 mm Hg, and four have grade III homograft insufficiency. In three cases the insufficiency is associated with significant obstruction; in the fourth case, the valve had been mechanically dilated 28 months earlier. Two of these four patients are scheduled for conduit replacement. One child with RVOT obstruction collapsed while playing at school 4 days before admission for surgery. He was resuscitated but suffered severe cerebral impairment so that conduit exchange was abandoned. Two children underwent interventional dilatation of the right and left pulmonary artery with stent insertion. The estimate for freedom from significant dysfunction for all grafts including those that failed is shown in Figure 3; progressive deterioration occurred over years, with 66 months as a median interval. As evidenced in Figure 4, the smaller normalized allografts deteriorated more rapidly ( p 0.0001), with 50% showing significant dysfunction at 16 months. In contrast, 50% of the larger grafts had satisfactory function at 89 months. Comparison of the other variables studied is presented in Table 5. There is no significant difference of behavior between the aortic and pulmonary prosthesis or between those implanted during the neonatal period and those implanted later. Use of a dacron extension did not statistically influence conduit performance. The prosthesis with a z-value greater than or equal to 2 from the institution that supplied the most homografts, lasted longer: none failed until 135 months. Fig 2. Probabilities of freedom from failure according to normalized homograft size over time. It is 64.3% 12.8%, 32.1% 13.0%, 24.1% 12.0%, and 24.1% 12.0% at 1, 3, 5, and 10 years for the z 2 grafts, and 98.1% 1.9%, 95.5% 3.2%, 86.7% 6.6%, and 73.7% 10.4% for the z 2 grafts at the same time intervals. Vertical bars represent SEM. The difference is highly significant.

2074 SINZOBAHAMVYA ET AL Ann Thorac Surg SMALL HOMOGRAFTS 2001;72:2070 6 Fig 3. Probabilities of freedom from homograft dysfunction for all grafts over time. The dashed lines indicate the standard error of the mean (S.E.M.). Comment Among the factors limiting the longevity of valve allografts, small size of the graft is the most frequently cited critical factor. Reports of the fate of small homografts after implantation in RVOT are few, and results vary. In the series of Leblanc and colleagues [4], freedom from reoperation for allografts with a diameter up to 15 mm was 80% at 1 year and 53% at 5 years. In the study reported by Homann and associates [5], survival of grafts with a diameter less than 15 mm was 77% at 5 years and 49% at 10 years. Replacement statistics between homografts less than 15 mm and those at least 15 mm in the report of Stark and associates did not differ: 84% and 58% at 5 and 10 years, respectively [6]. In the publication of Perron and colleagues [7] the probabilities of freedom from failure (dilation, stenting, or reoperation) were 82% and 17%, respectively, at 1 year and 5 years for all grafts up to 15 mm. They were, as estimated from the published graph, 72% and 17% for those sized 6 to 8 mm, 95% and 17% for the 9 to 11 mm, and 100% and 50% for the 12, 14, 15 mm grafts at the same time intervals. Outcomes taking into consideration homograft z- values have recently been published by Caldarone and associates [8] and Tweddell and coauthors [9], the latter authors being the first to indicate a z-value less than 2 as a risk factor for homograft failure and dysfunction. Their work reports probabilities of freedom from failure (homograft explantation or late death) of 92%, 60%, and 31% at 1, 5, and 10 years, respectively, for grafts with a z-value less than 2, and of 98%, 86%, and 66% for those with a z-value of at least 2 at the same time intervals (estimated from the published graph). In our study, after normalizing homograft size to the patient s body surface area, we found that the great majority of the 8 to 9 mm homografts had a z-score less than 2 and that almost all of the 11 to 13 mm grafts had a z-value of at least 2. The 10-mm grafts seem to constitute a transitional group where about 50% scored either above Fig 4. Probabilities of freedom from homograft dysfunction according to normalized homograft size over time. It is 57.4% 13.2%, 32.8% 13.1%, 8.2% 7.8%, and 0% at 1, 3, 5, and 10 years for the z 2 grafts, and 96.0% 2.8%, 93.8% 3.4%, 64.8% 8.2%, and 15.7% 12.3% for the z 2 grafts at the same time intervals. The dashed lines indicate the standard error of the mean. The difference is highly significant.

Ann Thorac Surg SINZOBAHAMVYA ET AL 2001;72:2070 6 SMALL HOMOGRAFTS 2075 Table 5. Comparison of Freedom From Events: Log-Rank Test Homograft variables p for Homograft Events Failure Dysfunction z-value: 2/ 2 0.0001 0.0001 Type: aortic/pulmonary 0.19 0.76 Dacron extension: used/not used 0.66 0.91 Supplier: single bank/others 0.09 0.68 (all conduits) (z 2) 0.91 0.20 (z 2) 0.0038 0.39 Age of patients: 30/ 30 days 0.84 0.99 or below 2 (see Table 2). Studies that mix all small conduits may thus give contradictory results. The discrepancies in the literature may also be explained by differing definitions of valve failure and differing indications for replacement. Some authors (eg [8]) do not include late death as a valve failure, even if a mild conduit dysfunction may have contributed to clinical deterioration. Whereas florid infection of the valve dictates quick reintervention, there are various thresholds for replacement of a progressively dysfunctional prosthesis. We, like others [10, 11], consider that the conduit should be replaced if the right ventricular pressure approaches the systemic pressure, regardless of the clinical status of the patient. Indeed, most of these patients are in functional New York Heart Association class I, despite severe RVOT obstruction. A waiting policy in such circumstances carries the potential for acute right ventricular failure, as experienced in one of our patients. Reoperation is indicated whenever low right ventricular pressure occurs in the presence of deteriorating right ventricular function or significant tricuspid regurgitation. It must also be considered if cardiac rhythm disturbances occur. Interventional dilatation and stenting has emerged as an occasionally effective technique for reopening the conduit, thereby delaying the need for reoperation [12, 13]. Two of our patients benefited from this procedure. It is important to note, however, that most degraded or outgrown small allografts are not suitable for dilatation. The difference in quality of homografts supplied by tissue banks is an additional factor in varied middle and long-term outcomes. The single institution that provided conduits with a z-value of at least 2 that did not fail until 135 months generally obtains valves from cadavers. Such grafts are unlikely to contain viable cells and therefore would provoke less host immunological response. As noted by Stark and colleagues [6], if homografts are harvested at the time of heart transplantation, and the time between harvest and preservation is very short, it is possible that immunological factors will play a role. In the present era, freshly antibiotic-sterilized or cryopreserved small homografts constitute, in many centers, the preferred conduits for right ventricle to pulmonary artery application in the neonate. Because of increasing shortage, our ideal conduit size (1.5 times the normal expected diameter of the pulmonary artery: z-value 3.1 to 3.8) could be realized in only 20% (15 of 76) of all cases. The largest homograft we implanted had a z-score of 5.2 in a neonate weighing 2.3 kg who had truncus arteriosus and received a 13-mm prosthesis. Although some authors [14] advocate the use of large conduits, their implantation certainly demands a longer right ventriculotomy, which could lead to increased right ventricular dysfunction and, possibly increased morbidity and mortality. In our series, the freedom from failure for the oversized (z at least 2) allografts, ie, the 11 to 13 mm grafts, at 10 years after implantation matches the durability for those 18 mm or larger implanted in older patients reported by Homann and associates [5]. Outgrowth seems to be the only reason for their explantation. In contrast, time to failure is short for conduits with a z-score less than 2 (generally the 8 to 9 mm grafts), corresponding to the limited longevity usually assigned to homografts implanted during infancy [7, 9, 11, 15]. Limitations of This Study This study carries the usual drawbacks of a retrospective study conducted over a long period (13 years). There was some variability with regard to indications for homograft use and homograft replacement. Shortages compelled us to use smaller or bigger homografts than the preferred size and, moreover, to diversify graft suppliers. As a consequence, some variables related to the homografts cannot be analyzed adequately. The echo-doppler evaluation of dysfunction is not precise enough: pressure gradients are usually overestimated and quantification of valvular regurgitation varies from investigator to investigator. Time at which significant dysfunction was diagnosed depended on the periodicity of the follow-up. We did not obtain histological data for the explanted homografts. Despite these limitations, there is evidence to support the findings that small homografts scoring a z-value from 2 to 5.2 (practically all of the 11 to 13 mm grafts) enjoy longer durability. Conclusion Small homograft valves enable the surgeon to perform early and complete repair of cardiac malformations that require RVOT reconstruction and offer the best chance for survival and improved quality of life. Reoperation is inevitable as the child grows, usually in the first2to3 years for the very small (z-value less than 2) 8 to 9 mm grafts, and much later for the bigger (z-value at least 2) 11 to 13 mm grafts. The 11 to 13 mm homografts, whether aortic or pulmonary, constitute the best valved conduits in the pulmonary position for neonates and young infants. References 1. Rowlatt UF, Rimoldi HJA, Lev M. The quantitative anatomy of the normal child s heart. Pediatr Clin North Am 1963;10: 499 588. 2. Urban AE, Sinzobahamvya N, Brecher AM, Wetter J, Malorny S. Truncus arteriosus: ten-year experience with ho-

2076 SINZOBAHAMVYA ET AL Ann Thorac Surg SMALL HOMOGRAFTS 2001;72:2070 6 mograft repair in neonates and infants. Ann Thorac Surg 1998;66:S183 8. 3. Perry GJ, Helmcke F, Nanda NC, Byarad C, Soto B. Evaluation of aortic insufficiency by Doppler color flow mapping. J Am Coll Cardiol 1987;9:952 9. 4. Leblanc JG, Russell JL, Sett SS, Potts JE. Intermediate follow-up of right ventricular outflow tract reconstruction with allograft conduits. Ann Thorac Surg 1998;66:S174 8. 5. Homann M, Haehnel JC, Mendler N, et al. Reconstruction of the RVOT with valved biological conduits: 25 years experience with allografts and xenografts. Eur J Cardiothorac Surg 2000;17:624 30. 6. Stark J, Bull C, Stajevic M, et al. Fate of subpulmonary homograft conduits: determinants of late homograft failure. J Thorac Cardiovasc Surg 1998;115:506 16. 7. Perron J, Moran AM, Gauvreau K, del Nido PJ, Mayer JE, Jonas RA. Valved homograft conduit repair of the right heart in early infancy. Ann Thorac Surg 1999;68:542 8. 8. Caldarone CA, McCrindle BW, Van Arsdell GS, et al. Independent factors associated with longevity of prosthetic pulmonary valves and valved conduits. J Thorac Cardiovasc Surg 2000;120:1022 31. 9. Tweddell JS, Pelech AN, Frommelt PC, et al. Factors affecting longevity of homograft valves used in right ventricular outflow tract reconstruction for congenital heart disease. Circulation 2000;102(Suppl. III):130 5. 10. Stark J. The use of valved conduits in pediatric cardiac surgery. Pediatr Cardiol 1998;19:282 8. 11. Mayer JE, Jr. Uses of homograft conduits for right ventricle to pulmonary artery connections in the neonatal period. Semin Thorac Cardiovasc Surg 1995;7:130 2. 12. Powell AJ, Lock JE, Keane JF, Perry SB. Prolongation of RV-PA conduit life span by percutaneous stent implantation. Intermediate-term results. Circulation 1995;92:3282 8. 13. Ovaert C, Caldarone CA, McCrindle BW, et al. Endovascular stent implantation for the management of postoperative right ventricular outflow tract obstruction. Clinical efficacy. J Thorac Cardiovasc Surg 1999;118:886 93. 14. Tam RK, Tolan MJ, Zamvar VY, et al. Use of larger sized aortic homograft conduits in right ventricular outflow tract reconstruction. J Heart Valve Dis 1995;4:660 4. 15. Cleveland DC, Williams WG, Razzouk AJ, et al. Failure of cryopreserved homograft valved conduits in the pulmonary circulation. Circulation 1992;86(Suppl. III):150 3. INVITED COMMENTARY This article by Dr Sinzobahamvya and colleagues offers a detailed look at the fate of small diameter valved allografts inserted in the pulmonary position in neonates and infants. Much has been written about allograft conduits in recent years regarding the rapid degeneration and need for replacement of these valved conduits, particularly in neonates and small infants. The authors have demonstrated that conduits in the 8 mm to 10 mm size range degenerate and need replacement more rapidly than those in the 11 mm to 13 mm size range. This was best predicted in their study of z-scores based upon the child s surface area with valved conduits with z-scores greater than 2 having a freedom from failure rate of about 85% at 5 years, as compared to a freedom from failure at about 25% at 5 years for this with z-scores less than 2. This study offers better results for freedom from reoperation than I have seen in my own practice with 11 mm to 13 mm valves, and those that have generally been achieved in neonates in other series utilizing cryopreserved valved allografts in this size range. Why is this? One explanation might be more stringent indices for replacement of a conduit, but the authors have appropriately pointed out that they generally replaced the conduit when right ventricular pressure neared systemic levels, regardless of the patient s symptoms. Another explanation might be the very favorable outcome of valved allografts supplied from one institution that used only cadaver donors and avoided fresh or homovital allografts. Or, it is possible that these grafts may elicit less of an immunologic reaction and this may in turn translate into better longevity. We are recognizing that cryopreserved valved allografts do cause an immunologic reaction, particularly in neonates and small infants, but the role immunology plays in the durability of allograft valves is less clear. Certainly the fact that the earliest failure for a valve with a z-score value greater than 2 was supplied by the single institution was 135 months is an amazing feat! The most important message to obtain from this article is that small valves in small infants fail rapidly, and larger valves (z-score 2 to 3) fail less rapidly in this difficult population group. Although the authors did not see a difference for failure rate for z-scores greater than 2 in infants less than 30 days old, as compared to those older than 30 days, it is nearly impossible to separate similar or linked variables such as small infant size (weight or body surface area) and small conduit size. It is hard to expect much more than 3 or 4 years from a valved conduit in an infant 1 weighing less than 2.5 kilograms, due to the expected rapid growth and weight gain in the first few years of life of a child this size. Caution should also be used in oversizing valves too much. This can necessitate longer ventriculotomies, or may lead to compression from the sternum and contribute to postoperative morbidity. The authors should be commended on excellent results and for demonstrating that more careful sizing utilizing z-scores can contribute to better longevity of the cryopreserved valved allografts in this difficult group of patients. John A. Hawkins, MD Division of Cardiothoracic Surgery Primary Children s Medical Center 100 N Medical Dr Salt Lake City, UT 84113 e-mail: jhawkins@med.utah.edu. 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00 Published by Elsevier Science Inc PII S003-4975(01)03318-5