This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Karamlou, T. Articles by McCrindle, B. W. Search for Related Content PubMed PubMed Citation Articles by Karamlou, T. Articles by McCrindle, B. W. Related Collections CV surgery: aortic and vascular disease Pediatric and congenital heart disease, including cardiovascular surgery Related Article

(Circulation. 2005;112:3462-3469.)
© 2005 American Heart Association, Inc.

Pediatric Cardiology

Outcomes and Associated Risk Factors for Aortic Valve Replacement in 160 Children

A Competing-Risks Analysis

Tara Karamlou, MD; Karen Jang, MS; William G. Williams, MD; Christopher A. Caldarone, MD; Glen Van Arsdell, MD; John G. Coles, MD; Brian W. McCrindle, MD, MPH

From the Division of Cardiovascular Surgery (T.K., W.G.W., C.A.C., G.V.A., J.G.C.), Department of Surgery, and the Division of Cardiology (K.J., B.W.M.), Department of Pediatrics, University of Toronto, The Hospital for Sick Children, Toronto, Canada.

Correspondence to Dr Brian McCrindle, The Hospital for Sick Children, 555 University Ave, Toronto, Ontario, Canada M5G 1X8. E-mail brian.mccrindle{at}sickkids.ca

Received February 8, 2005; revision received August 22, 2005; accepted September 19, 2005.

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Background— We sought to define patient characteristics, outcomes, and associated risk factors after aortic valve replacement (AVR) in children.

Methods and Results— Clinical records from children undergoing AVR from 1974 to 2004 at our institution were reviewed. Competing-risks methodology determined the time-related prevalence of 3 mutually exclusive end states: death, repeated replacement, and survival without subsequent AVR and their associated risk factors. Longitudinal echocardiographic data were analyzed by mixed linear-regression models. Children (n=160) underwent 198 AVRs, with 33 having >1. Competing-risks analysis predicted that 10 years from the initial AVR, 19% had died without subsequent AVR, 34% underwent a second AVR, and 47% remained alive without replacement. Risk factors for death without a second AVR included lower weight (P<0.001) and younger age at AVR (P=0.04), performance of aortic arch reconstruction together with AVR (P=0.03), and nonautograft use (P=0.03). Risk factors for a second AVR included earlier operation year (P=0.04) and implantation of a bioprosthetic or homograft valve (P=0.004). Analysis of serial echocardiographic measurements showed that pulmonary autograft use was associated with slower progression of peak aortic gradient (P=0.002), smaller left ventricular dimension (P=0.04), and decreased prevalence of aortic regurgitation (P=0.04).

Conclusions— Mortality and repeated valve replacement are common after initial AVR in children, especially in younger patients and those with bioprosthetic or homograft valves. Pulmonary autograft use is associated with decreased mortality, slower gradient progression, and smaller left ventricular dimension.

Key Words: heart defects, congenital • follow-up studies • pediatrics • aortic valve • valvuloplasty

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Aortic valve disease is one of the most common congenital cardiac defects, occurring in &5% of all children with heart disease,^1,2 and aortic valve replacement (AVR) is sometimes required. AVR in children poses unique challenges that have hampered identification of the ideal prosthesis.^3–6 Homograft and bioprosthetic valves achieve superior hemodynamic result initially but at the cost of accelerated degeneration.^7–9 Small patient size and the risk of thromboembolism limit the usefulness of mechanical valves, and somatic outgrowth is a universal problem with all available prostheses. The Ross operation (use of the pulmonary valve as an autograft) introduced in 1967¹⁰ has emerged as an attractive alternative, but concerns persist about the fate of the neoaortic valve, the potential for development of neoaortic valve insufficiency, and the lack of consistent long-term outcome data.^11–14 In addition, autograft use often necessitates repeated interventions for right ventricular outflow tract obstruction.^3,12

In addition to prosthesis failure and reoperation, children are simultaneously at an increased risk of death after AVR. Although outcomes after AVR in children have been reported elsewhere,^{2–4,8,15,16} previous studies have not used competing-risks methodology nor examined longitudinal data to evaluate functional performance of different prostheses over time. We therefore sought to use competing-risks analysis to determine the time-related prevalence and associated risk factors of 3 mutually exclusive end states: death, repeated replacement, and survival without subsequent AVR. We also sought to define the time course of hemodynamic parameters and left ventricular dimensions as impacted by prosthesis type according to longitudinal data analysis.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Study Subjects
Patients 18 years of age who underwent initial AVR at our institution between January 1974 and June 2004 were identified from computerized databases.

Data Collection and Measurements
The study was approved by the institutional research ethics board. Data collected from medical record review included patient demographics, preoperative cardiac and noncardiac diagnoses, clinical condition, previous procedures and complications, echocardiographic and cardiac catheterization assessments and procedures, operative data, and postoperative and follow-up clinical status, including repeated echocardiographic measurements. Echocardiographic measurements included peak instantaneous gradient across the prosthetic valve, subjective grade of prosthesis insufficiency, and left ventricular end-diastolic dimension (LVEDD, which was converted into Z-scores from regression equations based on previously published nomograms^17,18). Prosthesis Z-score was likewise determined as the number of standard deviations from mean normal aortic valve size.¹⁹

Data Analysis
Data are given as frequency, median with range, or mean±SD as appropriate, with the number of nonmissing values indicated. All data analyses were performed with SAS statistical software (version 9.1, SAS Institute, Inc). Both replaced valve failure with subsequent repeated replacement and mortality were modeled as time-dependent events by using both Kaplan-Meier estimates and parametric methods, with the association with risk factors being explored in multivariable analysis with bootstrap bagging²⁰ to guide variable selection and assess reliability of inclusion in the final regression models. Competing-risks analysis was used in a manner as previously reported.^20–27 In brief, parametric probability estimates of freedom from repeated valve replacement and death were generated from models based on the decomposition of the hazard function into multiple, overlapping phases of risk (available for use with the SAS system at http://www.clevelandclinic.org/heartcenter/hazard). The HAZARD procedure uses the method of maximum likelihood to aid in resolving up to 3 phases of the distribution of times until an event (early decreasing or peaking hazard, constant hazard, and late increasing hazard). Although 3 phases are potentially present for any time-related event, model validity does not require that separate scaling parameters be specified for all 3 phases, because realistically, many events have only 1 or 2 discrete phases of risk. Once the characteristic equation (hazard function) for each competing state is specified, competing-risks analysis is used to integrate them (with the function for remaining alive without death or subsequent replacement being the mathematically derived remainder given by [1–{proportion experiencing mortality+proportion experiencing repeated replacement}]) to show the proportion of patients within all of the specified states at any given time point. Post-AVR serial echocardiographic assessments of peak instantaneous prosthetic valve gradient and LVEDD Z-score were modeled, and risk factors were sought by using mixed linear regression. Factors associated with the presence of prosthetic valve insufficiency after AVR were sought by repeated-measures ordinal outcome models.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Patient Characteristics
Initial demographic and morphological characteristics are listed in Table 1. During the study period, 160 children underwent an initial AVR. The underlying valve disease was congenital in 84% (59 of whom had a bicuspid valve), rheumatic in 7%, isolated endocarditis in 2%, or other valve pathology in 7%. Associated cardiac defects were found in 52%, the most prevalent of which included mitral valve dysfunction in 23%, truncus arteriosus or Shone’s syndrome in 8%, and ventricular septal defect in 7%. The level of left ventricular outflow tract (LVOT) obstruction was predominantly valvular in 76%, subvalvular in 10%, supravalvular in 9%, and secondary to a hypoplastic valvular annulus in 3%. Multiple levels of obstruction were found in 15%. Patients with predominant subvalvular or supravalvular obstruction all had concomitant aortic valve pathology, including important regurgitation, stenosis, or valvar dysplasia of sufficient magnitude to warrant AVR.

View this table: [in this window] [in a new window]   TABLE 1. Patient Characteristics (n=160)

AVR Episodes
A flow chart depicting events after the initial AVR is shown in Figure 1. Of the 160 patients, during the study interval 127 had only 1 AVR (24 subsequently died), 33 patients had a second AVR (2 subsequently died), and 5 patients had a third AVR (3 subsequently died), for a total of 198 AVR episodes. The characteristics at the initial and subsequent AVR episodes are shown in Table 2. Median age at initial AVR was 11 years (range, 2 days to 18 years). The age at initial AVR was significantly younger in those with associated cardiac anomalies versus those with isolated aortic valve disease (P=0.04), although those with associated defects predominated throughout all ages. Patients receiving tissue valves were significantly younger (8.9±6.0 versus 11.4±4.6 years, P=0.02) than those receiving mechanical prostheses. The need for concomitant aortic arch enlargement or aortic reconstruction was associated with slightly longer cardiopulmonary bypass time (166±57 versus 152±26 minutes for those without reconstruction; P=0.15). Overall time-related survival for all patients regardless of subsequent procedures was 85%, 78%, and 70% at 1, 5, and 15 years, respectively.

View larger version (11K): [in this window] [in a new window]   Figure 1. Flow chart depicting events after initial AVR. There were 160 children who underwent 198 AVRs during the study period. There were 29 total deaths, 24 occurring without re-replacement, 2 occurring after a second AVR, and 3 occurring after a third AVR.

View this table: [in this window] [in a new window]   TABLE 2. Procedural Characteristics at Initial and Subsequent AVRs Combined (n=198 Procedures in 160 Patients)

Postoperative Clinical Course
There were 13 deaths within 30 days of AVR. Postoperative complications in 73 patients included postoperative hemorrhage in 12 patients (8%), severe sepsis in 11 (7%), pericardial effusion in 11 (7%), temporary or permanent heart block in 6 (4%), arrhythmia requiring medical therapy in 7 (5%), stroke in 3 (2%), and other complications in the remainder. A reoperation before hospital discharge was performed in 12 patients (8%), including exploration for bleeding in 4 (3%) and repair and replacement of the prosthesis in 1 patient each.

Late Complications
There was also important morbidity in 147 patients whose course was not complicated by death or subsequent replacement within 30 days. Nine patients (6%) had an episode of bleeding related to anticoagulation, with stroke and transient ischemic attack occurring in 1 patient each. Late arrhythmia occurred in 7 patients (5%), and another patient underwent pacemaker implantation for complete heart block. Paravalvular leak and prosthesis thrombosis were uncommon, occurring in 1 patient each. Anticoagulation at latest follow-up assessment included warfarin in 37%, aspirin in 26%, both aspirin and warfarin in 8%, and no anticoagulation in 26%. Anticoagulation was more frequently used for patients receiving mechanical prostheses but was used in 23 patients with bioprosthetic prostheses and in 6 autograft recipients. There were 3 late deaths among the 33 patients who underwent repeated AVR; 1 occurred 6 years after the first repeated replacement, and 2 occurred >5 years after a third AVR.

Competing Risks for Death or Subsequent Prosthesis Replacement After Initial AVR
During follow-up of the 160 initial AVR episodes, 33 prostheses were subsequently replaced and 24 patients died without further AVR. The hazard function for time-related transition to a second AVR was characterized by a constant-hazard phase (32 constant-phase events) and a late phase (1 late-phase event). The hazard function for time-related transition to death without a second AVR was characterized by a very steep early hazard phase (7 early-phase events) and a prolonged late phase (17 late-phase events). Competing-risks analysis predicted that after 10 years from initial AVR, 19% had died without a second AVR, 34% underwent a second AVR, and 47% remained alive without repeated replacement (Figure 2). Late-phase risk factors for death without a second AVR and for constant-phase risk factors for a subsequent second AVR after the initial AVR were sought and are shown in Table 3. The unfavorable effect of lower weight on mortality after initial AVR but before subsequent repeated replacement is shown in Figure 3. The predicted outcomes displayed in Figure 3 represent a specific solution to the multivariable equations for death without a second AVR for a hypothetical patient without autograft implantation or concomitant aortic arch reconstruction, with specific values assigned for the 2 other significant explanatory variables found to influence mortality within this competing-risks model. Similarly, Figure 4 demonstrates the unfavorable effect of earlier year of operation on survival to a second AVR. This graph represents a specific solution to the multivariable equation for survival to a subsequent AVR for a hypothetical patient with implantation of a tissue prosthesis at 3 different time intervals. A stratified graph of valve longevity by different types of initial prosthesis shows that autografts had superior longevity when compared with mechanical prostheses (Carbomedics [Sulzer Carbomedics, Inc], Bjork-Shiley [Shiley, Inc]) cryopreserved allografts, or bioprosthetic valves (Hancock [Medtronic], Ionescu-Shiley [Shiley], and others) (Figure 5).

View larger version (25K): [in this window] [in a new window]   Figure 2. Competing-risks depiction of events after initial AVR in 160 children. All patients are represented in the graph as alive at the time of initial AVR and thereafter migrate to 1 of 3 mutually exclusive end states (death, subsequent AVR, or remaining alive without subsequent AVR) at a time-dependent rate defined by the underlying hazard functions. At any point in time, the sum of the proportion of children in each state is 100%. For example, the estimated prevalences after 10 years from initial AVR are as follows: 19%, death without subsequent AVR; 34%, subsequent AVR; and 47%, alive without subsequent AVR. Solid lines represent parametric point estimates; dashed lines enclose the 90% confidence interval; and numbers in parentheses indicate the estimated proportion of patients in each state at 10 years after initial AVR.

View this table: [in this window] [in a new window]   TABLE 3. Incremental Risk Factors for Time-Related Transition From Initial AVR (N=160) to Either Death or a Second AVR

View larger version (18K): [in this window] [in a new window]   Figure 3. The unfavorable effect of lower weight on mortality after initial AVR but before subsequent re-replacement. The multivariable equation for death was solved for a hypothetical patient 6 months of age without autograft implantation or concomitant aortic arch reconstruction for 3 different values for weight at initial AVR. The solid lines represent parametric point estimates, and the numbers in parentheses indicate the predicted survival for patients with these specific criteria at 10 years from initial AVR.

View larger version (15K): [in this window] [in a new window]   Figure 4. The favorable effect of more recent era of operation on survival to a second AVR. The multivariable equation was solved 3 times for a hypothetical patient without autograft implantation or concomitant aortic arch reconstruction for 3 different intervals for AVR interval (where AVR interval=date of initial AVR–first AVR in the series). The solid lines represent parametric point estimates, and the numbers in parentheses indicate the predicted survival for patients with these specific criteria at 10 years from initial AVR.

View larger version (21K): [in this window] [in a new window]   Figure 5. Freedom from reoperation after initial AVR stratified by prosthetic type. A multivariable equation was constructed for remaining alive after initial AVR without subsequent valve replacement according to the original competing risk model and forcing all valve types into the equation. The resulting model was then solved for a hypothetical 10-year-old patient of 40 kg undergoing operation in 1990. The autograft has superior longevity, whereas the tissue valves and allografts have considerably worse durability. The numbers in parentheses represent the total number of AVR episodes for each prosthesis type.

Longitudinal Analysis of Serial Echocardiographic Measurements After AVR
Peak instantaneous echocardiographic Doppler prosthetic valve gradient was available for 124 patients, with a total of 524 measurements up to a maximum interval of 19 years (mean, 3.1 years). Progression of the prosthetic valve gradient over time was nonlinear, with a rapid initial rise followed by a prolonged gradual rise thereafter (P<0.001). Use of a prosthesis other than a pulmonary autograft was associated with more rapid gradient progression (P=0.002). Other significant independent factors associated with higher peak gradient at any time during follow-up included younger age at initial operation (P=0.02) and earlier year of operation (P=0.03). Figure 6 shows that implantation of an autograft effectively blunted the early accelerated progression of the peak prosthesis gradient that was evident after implantation of all other prosthesis types.

View larger version (20K): [in this window] [in a new window]   Figure 6. Predicted progression of the peak instantaneous prosthetic valve gradient after initial AVR stratified by prosthesis type (autograft vs nonautograft) for a hypothetical patient of 3 years undergoing AVR in 1990. Solid lines represent point estimates from the mixed linear-regression model (surrounded by their 90% confidence intervals in dashed lines). The numbers of echocardiographic measurements for each time period are as follows: 0 to 2 years, 257; 2 to 4 years, 120; 4 to 6 years, 105; 6 to 8 years, 27; and >8 years, 15. AoV indicates aortic valve.

LVEDDs (expressed as Z-scores) were similarly available for 121 patients, with a total of 524 measurements performed up to a maximum interval of 19 years (mean, 3.1 years). The relation between LVEDD and time was also nonlinear, with a rapid decline initially followed by a constant slope thereafter (P<0.001). Incremental risk factors associated with higher LVEDD Z-scores at any point during follow-up also included lower weight at initial operation (P=0.01), younger age at AVR (P<0.001), and use of a prosthesis type other than an autograft (P=0.04), as shown in Figure 7.

View larger version (16K): [in this window] [in a new window]   Figure 7. Predicted trajectory of LVEDD Z-score (ZLVED) stratified by prosthesis type (autograft vs nonautograft) for a hypothetical 2.7-kg patient aged 3 years undergoing operation in 1990. Solid lines represent point estimates from the mixed linear-regression model (surrounded by their 90% confidence intervals in dashed lines). The numbers of echocardiographic measurements for each time period are as follows: 0 to 2 years, 257; 2 to 4 years, 120; 4 to 6 years, 105; 6 to 8 years, 27; and >8 years: 15.

Data on the presence of important prosthetic valve insufficiency (subjective grade on echocardiography of more than mild) were available for 118 patients, with a total of 240 measurements performed up to a maximum interval of 18 years (mean, 3.1 years). The relation between the presence of prosthetic valve insufficiency and time was nonlinear, with an initial rise followed by a constant slope thereafter (P<0.001). Incremental risk factors associated with the presence of prosthetic valve insufficiency at any point during follow-up also included female sex (P=0.02), higher Z-score of implanted prosthesis size (P=0.05), and use of a prosthesis type other than an autograft (P=0.04).

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
Our study reports a single institution’s experience with AVR in 160 children. The results of the competing-risks analysis provide valuable information about the time-related prevalence of death without repeated replacement and prosthesis longevity. In addition, the results of the longitudinal data analysis corroborate those of the competing-risks models and provide insight for the superior outcomes associated with use of a pulmonary autograft.

Competing-risks analysis was chosen because our patients were simultaneously at risk for 2 mutually exclusive events: death and prosthesis replacement. Conventional time-related analyses consider individual events such as death or reoperation either in isolation or as a combined end point. Though useful, they do not address the question of how often an event may occur in the presence of other events for which a patient is at simultaneous risk. In addition, the incremental risk factors associated with each competing outcome (eg, death and reoperation) are often disparate, and therefore, using combined end points may be misleading.

Mortality
We identified younger age and lower weight at initial AVR to unfavorably influence mortality without repeated replacement, especially in the extreme case of very young age or very low weight. Previously published reports that have shown that neonates and those <6 months of age compose the highest-risk group agree with these findings.^2,28 There are several reasons for poor outcome in this population. First, young age at initial operation was significantly associated with the presence of other cardiac anomalies, including important mitral valve dysfunction, which accounted for substantial mortality in our series. Others have noted that those with concomitant cardiac lesions fare worse than those with isolated aortic valve disease.^2,3 Second, the preoperative clinical status of younger patients, especially neonates, is likely to be considerably worse than those undergoing later AVR.² Detailed preoperative information was not available in sufficient numbers in our cohort, but the correlation between poor preoperative left ventricular function (fractional shortening <25%) and late mortality has been established by others.⁷ Finally, younger patients (and those with lower weight) are at highest risk for prosthesis outgrowth necessitating subsequent repeated replacement or intervention, which may contribute to increased mortality.

The need for concomitant aortic arch reconstruction or augmentation was also identified as an incremental risk factor for death without a second AVR. Extensive aortic reconstruction in our series was associated with slightly longer cardiopulmonary bypass time. We have previously shown that prolonged cardiopulmonary bypass time is an important risk factor for death in children undergoing mitral valve replacement.²²

Use of a pulmonary autograft, compared with all other prostheses, was protective with respect to death without repeated prosthesis replacement. In our series, only 3 patients died after autograft implantation, and none required repeated replacement. From our mixed regression analysis, autograft use was associated with a decreased prevalence of prosthetic valve insufficiency, retarded early progression of the peak instantaneous prosthetic valve gradient, and a more rapid decrease in the LVEDD. More favorable hemodynamics in this context may be secondary to growth of the autograft, thereby neutralizing prosthesis outgrowth. Maintenance of low transvalvular gradients in patients receiving pulmonary autografts has been reported previously.^7,13,14,29 In their series of 105 patients undergoing the Ross operation, Savoye and colleagues²⁹ found no progression of autograft peak gradients as measured by echocardiography (5.0±2.8 mm Hg at discharge versus 5.5±3.5 mm Hg at last follow-up, P=NS) during a 7-year study interval and further noted that autograft gradients were similar to those of native aortic valves.³⁰ Normalization of ventricular dimensions and regression of ventricular hypertrophy have also been well described.^7,31 The longitudinal data analysis used in this study provides more direct evidence of the hemodynamic and functional impact of particular valve substitutes and facilitates an empirical approach to prosthesis selection.

Despite the advantages of autograft use, concern exists about dilatation of the neoaortic root and potential development of neoaortic valve insufficiency after the Ross operation. We have shown that autograft use is not associated with an increased prevalence of neoaortic regurgitation. This finding may reflect the absence of anatomic mismatch between the aortic and pulmonary valves in our series, because placement of an oversized prosthesis was correlated with the development of aortic insufficiency. Recent literature addressing these issues has generated conflicting reports.^7,14,32 Simon at el¹⁴ documented rapid adaptation of the pulmonary autograft to systemic pressure and normal growth trajectories thereafter, with 95% of patients free from important regurgitation at 4 years. A proportional increase in the LVOT root diameter (indexed to patient body surface area) was similarly documented in a larger study of 260 patients (136 of whom were <18 years of age).³² Other authors have reported progressive dilation of the autograft with differential prevalence, depending on the definition used (generally 4.0 cm or greater), the method used for detection, and the duration of follow-up.¹¹ Predisposing factors favoring dilation have included anatomic mismatch between the pulmonary and aortic root and a complete aortic root replacement technique without pericardial buttressing of the ascending aorta and annulus.^11,31,33

Valve Longevity
We identified 2 risk factors for decreased time to a repeated AVR after the initial AVR: (1) earlier year of operation and (2) use of a bioprosthetic or homograft valve. The protective effect of the more recent era has been documented in other complex congenital lesions²³ coincident with improvements in patient triage, preoperative and postoperative care, diagnostic modalities, myocardial protection, and operative technique. In addition, our recent experience favored autografts and avoided the use of bioprostheses, which likely contributed to improved outcomes over time.

The choice of initial prosthesis is still debated. In this report, bioprosthetic valves and allografts had decreased freedom from prosthesis replacement when compared with mechanical valves and with autografts after adjustment for all other significant variables. Patients receiving tissue valves in our series were significantly younger than those receiving mechanical prostheses, because the larger size of mechanical valves often precludes their use in this setting. In addition, bioprosthetic valves were more commonly used early in our experience, contributing to suboptimal outcomes. Accelerated degeneration, calcification, and structural failure of bioprostheses have led to inferior longevity compared with mechanical valves in most published series.^5,6 The improved durability of mechanical valves, however, must be tempered by the long-term requirement for anticoagulation and thromboembolism. Nine children with mechanical valves developed important bleeding related to warfarin use during follow-up, and thromboembolic complications occurred in 3 patients. Patient quality of life, though not directly assessed in this study, has been shown to be compromised by required thromboembolic prophylaxis.³⁴

The data on allograft longevity are somewhat more controversial, because important risk factors influencing durability, such as younger age and smaller prosthesis size, are often more prevalent in allograft recipients. A favorable effect on ventricular function and dimensions was noted by Hasnat et al⁷ in a study of 144 patients undergoing re-replacement with an allograft. Lupinetti and colleagues⁸ also noted that comparable freedom from death and reoperation were achieved with use of either an allograft or a pulmonary autograft. In contradistinction, other authors have reported less favorable outcomes with allograft use.^6,35–37 The unresolved issue of potential sensitization leading to accelerated degeneration of allografts further complicates decision making.³⁸

Limitations
The present report is a retrospective analysis of patients from a single institution with diverse anatomy who underwent operation during a 30-year period. Undoubtedly, indications for valve replacement, timing of operation, and operative technique were not constant over time. In addition, prosthesis selection and operative procedures were not standardized or randomized, and the comprehensive set of options was not available throughout the study period. Data on reinterventions directed against the right ventricular outflow tract, an important potential complication of pulmonary autograft use, were not collected in this study. Finally, although a comprehensive set of variables was used in all analyses, unmeasured covariates may have contributed to disparate outcomes in the recipient populations.

Conclusions
We have used competing-risks methodology and longitudinal data analysis to define outcomes in 160 children after AVR. We have demonstrated that younger age, lower operative weight, concomitant performance of aortic root replacement or reconstruction, and use of prosthesis type other than a pulmonary autograft are significant predictors of death without a repeated AVR. Furthermore, the use of a bioprosthetic or allograft valve type and earlier year of operation were identified as significant risk factors for earlier repeated AVR. Autograft use was not associated with an increased prevalence of prosthetic valve insufficiency. Blunted early progression of the peak prosthetic valve gradient and a rapid decrease in the LVEDD may account for the favorable outcomes in autograft recipients.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Salomon NW, Stinson EB, Oyer P, Copeland JG, Shumway NE. Operative treatment of congenital aortic stenosis. Ann Thorac Surg. 1978; 26: 452–460.[Abstract]
   
2. Brown JW, Ruzmetov M, Vijay P, Rodefeld MD, Turrentine MW. Surgery for aortic stenosis in children: a 40-year experience. Ann Thorac Surg. 2003; 76: 1398–1411.[Abstract/Free Full Text]
   
3. Alexiou C, McDonald A, Langley SM, Dalrymple-Hay MJR, Haw MP, Monro JL. Aortic valve replacement in children:are mechanical prostheses a good option? Eur J Cardiothorac Surg. 2000; 17: 125–133.[Abstract/Free Full Text]
   
4. Mazzitelli D, Guenther T, Schreiber C, Wottke M, Michel J, Meisner H. Aortic valve replacement in children:are we on the right track? Eur J Cardiothorac Surg. 1998; 13: 565–571.
   
5. Elkins RC, Knott-Craig CJ, Ward KE, McCue C, Lane MM. Pulmonary autograft in children: realized growth potential. Ann Thorac Surg. 1994; 57: 1387–1394.[Abstract]
   
6. Turrentine MW, Ruzmetov M, Vijay P, Bills RG, Brown JW. Biological versus mechanical aortic valve replacement in children. Ann Thorac Surg. 2001; 71: S356–S360.[Abstract/Free Full Text]
   
7. Hasnat K, Birks EJ, Liddicoat J, Hon JFK, Edwards S, Glennon S, Yacoub MH. Patient outcome and valve performance following a second aortic valve homograft replacement. Circulation. 1999; 100 [suppl II]: II-42–II-47.
   
8. Lupinetti FM, Duncan BW, Lewin M, Dyamenahalli U, Rosenthal GL. Comparison of autograft and allograft aortic valve replacement in children. J Thorac Cardiovasc Surg. 2003; 126: 240–246.[Abstract/Free Full Text]
   
9. Aklog L, Carr-White GS, Birks EJ, Yacoub MH. Pulmonary autograft versus aortic homograft for aortic vale replacement: interim results from a prospective randomized trial. J Heart Valve Dis. 2000; 9: 176–188.[Medline] [Order article via Infotrieve]
   
10. Ross DN. Replacement of aortic and mitral valves with a pulmonary autograft. Lancet. 1967; 2: 956–958.[Medline] [Order article via Infotrieve]
    
11. Luciani GB, Casali G, Favaro A, Prioli MA, Barozzi L, Santini F, Mazzucco A. Fate of the aortic root late after Ross operation. Circulation. 2003; 108 [suppl II]: II-61–II-67.
    
12. Kouchoukos NT, Masetti P, Nickerson NJ, Castner CF, Shannon WD, Davila-Roman VG. The Ross procedure: long-term clinical and echocardiographic follow-up. Ann Thorac Surg. 2004; 78: 773–781.[Abstract/Free Full Text]
    
13. Kouchokos NT, Davila-Roman VG, Spray TL, Murphy SF, Perillo JB. Replacement of the aortic root with a pulmonary autograft in children and young adults with aortic-valve disease. N Engl J Med. 1994; 330: 1–6.[Abstract/Free Full Text]
    
14. Simon P, Aschauer C, Moidl R, Marx M, Keznickl FP, Eigenbauer E, Wolner E, Wollenek G. Growth of the pulmonary autograft after the Ross operation in childhood. Eur J Cardiothorac Surg. 2001; 19: 118–121.[Abstract/Free Full Text]
    
15. Lupinetti FM, Duncan BW, Scifres AM, Fearneyhough CT, Kilian K, Rosenthal GL, Cecchin F, Jones TK, Herndon SP. Intermediate-term results in pediatric aortic valve replacement. Ann Thorac Surg. 1999; 68: 521–525.[Abstract/Free Full Text]
    
16. Najm HK, Coles JG, Black MD, Benson L, Williams WG. Extended aortic root replacement with aortic allografts or pulmonary autografts in children. J Thorac Cardiovasc Surg. 1999; 118: 503–509.[Abstract/Free Full Text]
    
17. Daubeney PEF, Blackstone EH, Weintraub RG, Slavik Z, Scanlon J, Webber SA. Relationship of the dimension of cardiac structures to body size: an echocardiographic study in normal infants and children. Cardiol Young. 1999; 9: 402–410.[Medline] [Order article via Infotrieve]
    
18. Sluysmans T, Colan SD. Theoretical and empirical derivation of cardiovascular allometric relationships in children. J Appl Physiol. 2004: 1144–1152.
    
19. Kirklin JW, Blackstone EH, Jonas RA, Kouchoukos NT. Anatomy, dimensions, and terminology. In: Barrat-Boyes BG, ed. Cardiac Surgery, 3rd ed. Edinburgh, Scotland: Churchill Livingstone; 2004: 50.
    
20. Blackstone EH, Naftel DC, Turner ME Jr. The decomposition of time-varying hazard into phases, each incorporating a separate stream of concomitant information. J Am Stat Assoc. 1986; 81: 615–624.[CrossRef]
    
21. McGriffin DC, Naftel DC, Kirklin JK, Morrow WR, Towbin J, Shaddy R, Alejos J, Rossi A, and the Pediatric Heart Transplant Study Group. Predicting outcome after listing for heart transplantation in children: a comparison of Kaplan-Meier and parametric competing risk analysis; Pediatric Heart Transplant Study Group. J Heart Lung Transplant. 1997; 16: 713–722.[Medline] [Order article via Infotrieve]
    
22. Kojori F, Chen Rui, Caldarone CA, Merklinger SL, Azakie A, Williams WG, Van Arsdell GS, Coles G, McCrindle BW. Outcomes of mitral valve replacement in children: a competing-risks analysis. J Thorac Cardiovasc Surg. 2004; 128: 703–709.[Abstract/Free Full Text]
    
23. McCrindle BW, Tchervenkov CI, Konstantinov IE, Williams WG, Neirotti RA, Jacobs ML, Blackstone EH, for the members of the Congenital Heart Surgeons Society. Risk factors with mortality and interventions in 472 neonates with interrupted aortic arch: a Congenital Heart Surgeons Society study. J Thorac Cardiovasc Surg. 2005; 129: 343–350.[Abstract/Free Full Text]
    
24. Blackstone EH. Breaking down barriers: helpful breakthrough statistical methods you need to understand better. J Thorac Cardiovasc Surg. 2001; 122: 430–439.[Free Full Text]
    
25. Lofland GK, McCrindle BW, Williams WG, Blackstone EH, Tchervenkov CI, Sittiwangkul R, Jonas RA and the Congenital Heart Surgeons Society. Critical aortic stenosis in the neonate: a multi-institutional study of management, outcomes, and risk factors. J Thorac Cardiovasc Surg. 2001; 121: 10–27.
    
26. Ashburn DA, Blackstone EH, Wells WJ, Jonas RA, Pigula FA, Manning PB, Lofland GK, Williams WG, McCrindle BW and the Members of the Congenital Heart Surgeons Society. Determinants of mortality and type of repair in neonates with pulmonary atresia and intact ventricular septum. J Thorac Cardiovasc Surg. 2004; 127: 1000–1008.[Abstract/Free Full Text]
    
27. McCrindle BW, Blackstone EH, Williams WG, Sittiwangkul R, Spray TL, Azakie A, Jonas RA, and the members of the Congenital Heart Surgeons Society. Are outcomes of surgical versus transcatheter balloon valvotomy equivalent in neonatal critical aortic stenosis? Circulation. 2001; 104 [suppl I]: I-152–I-158.
    
28. Hawkins JA, Minich LL, Tani LY, Day RW, Judd VE, Shaddy RE, McGough EC. Late results and reintervention after aortic valvotomy for critical aortic stenosis in neonates and infants. Ann Thorac Surg. 1998; 65: 1758–1763.[Abstract/Free Full Text]
    
29. Savoye C, Auffray JL, Hubert E, Godart F, Francart C, Goullard L, Deklunder G, Rey C, Orat A. Echocardiographic follow-up after Ross procedure in 100 patients. Am J Cardiol. 2000; 85: 854–857.[CrossRef][Medline] [Order article via Infotrieve]
    
30. Hatle L, Angelson B. Doppler Ultrasound in Cardiology: Physical Principles and Clinical Applications, 2nd ed. Philadelphia: Lea & Fibiger; 1985: 74–96.
    
31. Hraska V, Krajci M, Haun C, Ntalakoura K, Razek V, Lacour-gayet, Weil J, Reichenspurner H. Ross and Ross-Konno procedure in children and adolescents: mid-term results. Eur J Cardiothorac Surg. 2004; 25: 742–747.[Abstract/Free Full Text]
    
32. Al-Halees Z, Pieters F, Qadoura F, Shahid M, Al-Amri M, Al-Fadley F. The Ross procedure is the procedure of choice for congenital aortic valve disease. J Thorac Cardiovasc Surg. 2002; 123: 437–442.[Abstract/Free Full Text]
    
33. Svensson G, Aljassim O, Svensson SE, Bech-Hanssen O, Kjellman U. Anatomical mismatch of the pulmonary autograft in the aortic root may be the cause of early aortic insufficiency after the Ross procedure. Eur J Cardiothorac Surg. 2002; 21: 1049–1054.[Abstract/Free Full Text]
    
34. van Doorn C, Yates R, Tunstill A Elliott M. Quality of life in children following mitral valve replacement. Heart. 2000; 84: 643–647.[Abstract/Free Full Text]
    
35. Clarke DR. Extended aortic root replacement with cryopreserved allografts: do they hold up. Ann Thorac Surg. 1991; 52: 669–675.[Abstract]
    
36. Clarke DR, Campbell DN, Hayward AR, Bishop DA. Degneration of aortic valve allografts in young recipients. J Thorac Cardiovasc Surg. 1993; 105: 934–942.[Abstract]
    
37. Gallo R, Kumar N, Prabhakar G, Al-Halees Z, Duran CMG. Accelerated degeneration of aortic homograft in an infant. J Thorac Cardiovasc Surg. 1994; 107: 1161–1162.[Free Full Text]
    
38. Smith JD, Ogino H, Hunt D, Laylor RM, Rose ML, Yacoub MH. Humoral response to human aortic valve homografts. Ann Thorac Surg. 1995; 60: S127–S130.
    

Related Article:

Issue Highlights
Circulation 2005 112: 3365. [Full Text]

This article has been cited by other articles:

				M. Masuda, H. Kado, Y. Ando, A. Shiose, T. Nakano, K. Fukae, Y. Tanoue, and R. Tominaga Intermediate-term results after the aortic valve replacement using bileaflet mechanical prosthetic valve in children. Eur. J. Cardiothorac. Surg., July 1, 2008; 34(1): 42 - 47. [Abstract] [Full Text] [PDF]

				E. A. Bacha, D. B. McElhinney, K. J. Guleserian, S. D. Colan, R. A. Jonas, P. J. del Nido, and G. R. Marx Surgical aortic valvuloplasty in children and adolescents with aortic regurgitation: Acute and intermediate effects on aortic valve function and left ventricular dimensions J. Thorac. Cardiovasc. Surg., March 1, 2008; 135(3): 552 - 559. [Abstract] [Full Text] [PDF]

				R. Arnold, J. Ley-Zaporozhan, S. Ley, T. Loukanov, C. Sebening, J.-B. Kleber, B. Goebel, S. Hagl, M. Karck, and M. Gorenflo Outcome After Mechanical Aortic Valve Replacement in Children and Young Adults Ann. Thorac. Surg., February 1, 2008; 85(2): 604 - 610. [Abstract] [Full Text] [PDF]

				T. J. Bradley, T. Karamlou, A. Kulik, B. Mitrovic, T. Vigneswaran, S. Jaffer, P. D. Glasgow, W. G. Williams, G. S. Van Arsdell, and B. W. McCrindle Determinants of repair type, reintervention, and mortality in 393 children with double-outlet right ventricle. J. Thorac. Cardiovasc. Surg., October 1, 2007; 134(4): 967 - 973.e6. [Abstract] [Full Text] [PDF]

				T. Karamlou, R. Gurofsky, A. Bojcevski, W. G. Williams, C. A. Caldarone, G. S. Van Arsdell, T. Paul, and B. W. McCrindle Prevalence and Associated Risk Factors for Intervention in 313 Children With Subaortic Stenosis Ann. Thorac. Surg., September 1, 2007; 84(3): 900 - 906. [Abstract] [Full Text] [PDF]

				B. Alsoufi, T. Karamlou, B. W. McCrindle, and C. A. Caldarone Management options in neonates and infants with critical left ventricular outflow tract obstruction Eur. J. Cardiothorac. Surg., June 1, 2007; 31(6): 1013 - 1021. [Abstract] [Full Text] [PDF]

				T. Karamlou, R. Gurofsky, E. Al Sukhni, J. G. Coles, W. G. Williams, C. A. Caldarone, G. S. Van Arsdell, and B. W. McCrindle Factors Associated With Mortality and Reoperation in 377 Children With Total Anomalous Pulmonary Venous Connection Circulation, March 27, 2007; 115(12): 1591 - 1598. [Abstract] [Full Text] [PDF]

				B. Alsoufi, T. Karamlou, T. Bradley, W. G. Williams, G. S. Van Arsdell, J. G. Coles, J. Smallhorn, M. Nii, V. Guerra, and C. A. Caldarone Short and midterm results of aortic valve cusp extension in the treatment of children with congenital aortic valve disease. Ann. Thorac. Surg., October 1, 2006; 82(4): 1292 - 1300. [Abstract] [Full Text] [PDF]

				T. P. Graham Jr The Year in Congenital Heart Disease J. Am. Coll. Cardiol., June 20, 2006; 47(12): 2545 - 2553. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Karamlou, T. Articles by McCrindle, B. W. Search for Related Content PubMed PubMed Citation Articles by Karamlou, T. Articles by McCrindle, B. W. Related Collections CV surgery: aortic and vascular disease Pediatric and congenital heart disease, including cardiovascular surgery Related Article