ORIGINAL ARTICLES: CONGENITAL HEART SURGERY
CONGENITAL HEART SURGERY: The Annals of Thoracic Surgery CME Program is located online at http://www.annalsthoracicsurgery.org/cme/ home. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
CONGENITAL HEART
National Practice Patterns and Early Outcomes of Aortic Valve Replacement in Children and Teens Jennifer S. Nelson, MD, MS, Timothy M. Maul, PhD, CCP, Peter D. Wearden, MD, PhD, Sara K. Pasquali, MD, MS, and Jennifer C. Romano, MD, MS Department of Cardiovascular Services, Nemours Children’s Hospital, Orlando, Florida; and Departments of Pediatrics and Cardiac Surgery, University of Michigan C.S. Mott Children’s Hospital, Ann Arbor, Michigan
Background. Several options exist for aortic valve replacement (AVR) in children and teens, but contemporary practice patterns and outcome data are lacking. We describe national AVR practice patterns and early outcomes. Methods. Children (aged 1 to 18 years) in The Society of Thoracic Surgeons Congenital Heart Surgery Database undergoing AVR from 2000 to 2016 were included. Preoperative characteristics, operative data, and outcomes were described. To evaluate practice patterns, centers were assigned tertiles by aortic valve surgical volume. Statistical comparisons included Mann-Whitney U statistic, Kruskal-Wallis, c2 test, and gamma testing. Results. In total, 3446 operations (46% children aged 1 to 12 years; 54% teens aged 12 to 18 years) were included. Preoperative risk factors were present in 23%, and 46% had a prior sternotomy. Valve utilization included autograft (64% child, 37% teen), mechanical (19% child, 35% teen), bioprosthetic (8% child, 20% teen), and homograft (9% child, 7% teen). Autografts were utilized more often
for teenage girls than for teenage boys (odds ratio 1.3, 95% confidence interval: 1.05 to 1.66, P < .05). Overall, inpatient mortality and major complications affected 1% and 10%, respectively, and these rates were highest for homografts (5%, P < .001, and 13%, P < .05). Autograft utilization varied widely across centers (10th to 90th percentile: 21% to 71% of total AVR volume). More autografts were utilized at high-volume centers vs low- or medium-volume centers (53% ± 2.3% vs 46% ± 2.6%, P < .001). Conclusions. Practice patterns for AVR in children and teens vary across centers, age groups, and sexes. Although early outcomes were similar across valve types, homografts had higher morbidity and mortality. Valve choice was related to aortic valve surgical volume. Further efforts are needed to understand and optimize AVR practice patterns and long-term outcomes.
A
Preoperative counseling for patients and families is difficult when side-by-side comparisons of outcomes by valve type are largely unavailable, outdated, or not stratified by age group. Studies comparing valve types in children are limited by small numbers of non-Ross AVRs and differences in patient characteristics.1-4 Early pediatric AVR outcomes for specific valve types have been reported, but generally from single centers with mixed age-group populations.2,5-7 Therefore, the ideal valve type for children and teens remains undefined.
ortic valve replacement (AVR) in children is a common surgical treatment strategy for congenital and acquired aortic valve (AV) disease, with an estimated 2000 pediatric AVRs in the United States in the last decade. For neonates and infants, AVR options are limited owing to patient size. For older children and teens, pulmonary autograft, mechanical, bioprosthetic, or homograft valves may be used. Valve selection includes lifestyle and medical considerations involving the patient, family, and surgeon.
(Ann Thorac Surg 2019;108:544-51) Ó 2019 by The Society of Thoracic Surgeons
Accepted for publication Mar 25, 2019. Presented at the Fifty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 26-29, 2019. Address correspondence to Dr Nelson, Department of Cardiovascular Services, Nemours Children’s Hospital, 13535 Nemours Pkwy, Orlando, FL 32828; email:
[email protected].
Ó 2019 by The Society of Thoracic Surgeons Published by Elsevier Inc.
The Supplemental Tables can be viewed in the online version of this article [https://doi.org/10.1016/j. athoracsur.2019.03.098] on http://www.annalsthoracic surgery.org.
0003-4975/$36.00 https://doi.org/10.1016/j.athoracsur.2019.03.098
Ann Thorac Surg 2019;108:544-51
Because several AVR options exist for children and teens, it would be useful to understand practice patterns and outcomes for these age groups beyond small series reported at single centers. In addition, the relationship between center volume and patterns of valve utilization remains unexplored in the context of pediatric AVR. The purpose of the present study was to evaluate national practice patterns and early outcomes by valve type of children and teens undergoing AVR. In addition, we sought to better understand valve utilization patterns in relation to center surgical volume.
Data Source The Society of Thoracic Surgeons (STS) Congenital Heart Surgery Database (CHSD) was used for this study. The STS-CHSD is the largest database for congenital heart surgery and contains records on more than 475,000 operations conducted at 143 centers worldwide.8 The STSCHSD includes perioperative, operative, and outcomes data on all children undergoing heart surgery at participating centers; these data currently represent approximately 96% of all US congenital heart surgery centers.8,9 Formal site visits and audits ensure data quality and reliability. This study was exempt from further review by the Nemours Children’s Hospital Institutional Review Board and was approved by the Access and Publications Taskforce of the STS Workforce on National Databases.
Patient Population Children (aged 1 to 18 years) undergoing AVR with a pulmonary autograft (Ross operation10), mechanical, bioprosthetic, or homograft valve between 2000 and 2016 were included. Neonates and infants were excluded to focus on patients theoretically large enough to receive any of the four valve types. Root replacements were included and grouped according to valve type. Truncal valve replacements and records that did not specify valve type or mortality status at time of hospital discharge were excluded.
Data Collection Patient variables included baseline characteristics, prior cardiothoracic operations (available 2010 to 2016), extracardiac malformations, genetic abnormalities/syndromes, and STS-defined preoperative risk factors. Operative variables included cardiopulmonary bypass time, aortic crossclamp time, and concomitant procedures such as aortic arch reconstruction and mitral surgery.
Outcomes Early outcomes were analyzed by valve type. The primary outcome measures were inhospital mortality and major complications (Supplemental Table 1). Secondary outcomes included postoperative length of stay, postoperative cardiac arrest, unplanned reoperation, and unplanned readmissions. Definitions of the individual variables can be found in the STS-CHSD specifications.11
545
Analysis Patient characteristics and outcomes were summarized and stratified by age group and valve type received. Because of nonnormal distributions, continuous variables were summarized with median and interquartile range (IQR). Categoric variables were presented as frequencies and percentages. Results were summarized using frequency and percentage with calculated 95% confidence interval (CI). Comparisons between groups were performed using the Mann-Whitney U statistic, KruskalWallis testing, and odds ratio (OR), as appropriate. To display autograft trends over time, a power trendline was fit using MSExcel (2016) trendline function. All other analyses were performed using IBM SPSS Stastistics 25 (IBM Corp, Armonk, NY). A P value less than .05 was considered statistically significant. We described practice patterns by individual center AV surgery volume (number of pediatric—0 to 18 years—AV repairs plus AVRs) during the study period. Center valvetype utilization (%) was calculated as: ([specific valve type AVRs] / [total AVRs (autograft þ mechanical þ bioprosthetic þ homograft)]). When all centers were included for valve utilization patterns, very small volume centers represented the extremes of percent utilization for particular valve types. To eliminate these outliers, centers that performed fewer than 20 pediatric AVRs plus AV repairs during the study period were excluded for center volume analyses. Tertiles representing low-volume (20 to 42 cases), medium-volume (44 to 82 cases), and highvolume (85 to 411 cases) centers were created. Patterns of valve utilization were then compared across low-, medium-, and high-volume centers using gamma testing.
Results Population Characteristics Inclusion criteria were met by 3446 patient records (1594 for children and 1852 for teens) representing 123 institutions. Boys represented 74%. Median age at surgery was 7 years (IQR: 4 to 10) for children and 15 years (IQR: 14 to 17) for teens (Table 1). The most common primary diagnosis was aortic insufficiency (37%), followed by aortic stenosis (22%) and mixed aortic stenosis/aortic insufficiency (22%). Mixed aortic stenosis/aortic insufficiency was a commonly reported associated diagnosis (40%). Genetic/ chromosomal abnormalities affected 44% of the study population, whereas Shone’s syndrome was rare (2%). Overall, preoperative shock (1%) and mechanical ventilation (2%) were uncommon, except among homograft recipients (5% and 6%, respectively, P < .05). For 1,994 case records (2010 to 2016), prior cardiothoracic operations were captured. Within that group, 46% had a sternotomy before AVR, 9% had prior AVR, 13% had prior balloon aortic valvuloplasty, and 22% had prior surgical valvuloplasty. Of prior AVRs, 19% were pulmonary autografts.
Valve Choice A pulmonary autograft was utilized most commonly overall and within each age group (Table 2). A Konno
CONGENITAL HEART
Patients and Methods
NELSON ET AL AORTIC VALVE REPLACEMENT IN CHILDREN
546
NELSON ET AL AORTIC VALVE REPLACEMENT IN CHILDREN
Table 1.
Patient Characteristics
Characteristics
CONGENITAL HEART
Ann Thorac Surg 2019;108:544-51
Age, y Male Weight at operation, kg Noncardiovascular major abnormalitya Syndrome/chromosomal abnormalityb Any STS-CHSD preoperative risk factor Mechanical ventilation at surgery Shock or acidosis MCS at surgery Primary diagnosisc Aortic insufficiency Aortic insufficiency and stenosis Aortic stenosis, valvar Aortic stenosis, subvalvar Other diagnoses Aortic stenosis/insufficiency Aortic insufficiency Aortic stenosisd Subvalvar Valvare Supravalvar Not otherwise specifiede Aortic valve, other Endocarditis Marfan syndrome Rheumatic valve disease Shone’s syndrome Previous interventionse Prior sternotomy Balloon valvuloplasty Surgical valvuloplasty Aortic valve replacement Coarctation/arch repair/IAA repair Subaortic stenosis resection Ventricular septal defect repair Mitral valve operation Arterial switch operation Konno Atrioventricular septal defect
All (N ¼ 3446) 13 2534 44 129/2005 1531 791 66 46 11
Child (n ¼ 1594)
(8-15) (74) (24-63) (6) (44) (23) (2) (1) (0.3)
7 1100 23 67/904 722 382 45 24 8
(4-10) (70) (16-32) (7) (45) (24) (3) (2) (0.5)
Teen (n ¼ 1852) 15 1434 60 62/1101 809 409 21 22 3
(14-17) (78) (50-74) (6) (44) (22) (1) (1) (0.2)
1275 752 744 175
(37) (22) (22) (5)
544 334 360 120
(34) (21) (23) (8)
731 418 384 55
(40) (23) (21) (3)
1377 1198 688 241/688 564/688 65/688 15/127 432 121 78 18 40/1994
(40) (35) (20) (35) (82) (9) (12) (12) (4) (2) (0.5) (2)
644 490 388 175/388 300/388 44/388 9/75 182 50 19 7 29/894
(40) (31) (24) (45) (77) (11) (12) (11) (3) (1) (0.4) (3)
733 708 300 66/300 264/300 21/300 6/52 250 71 59 11 11/1100
(40) (38) (16) (22) (88) (7) (12) (14) (4) (3) (0.6) (1)
915/1994 264/1994 428/1994 178/1994 190/1994 203/1994 115/1994 100/1994 59/1994 29/1994 8/1994
(46) (13) (22) (9) (10) (10) (6) (5) (3) (2) (0.4)
445/894 136/894 187/894 62/894 128/894 119/894 67/894 60/894 33/894 15/894 6/894
(50) (15) (21) (7) (14) (13) (8) (7) (4) (2) (0.7)
470/1100 128/1100 241/1100 116/1100 62/1100 84/1100 48/1100 40/1100 26/1100 14/1100 2/1100
(43) (12) (22) (10) (6) (8) (4) (4) (2) (1) (0.2)
a
Codes available 2010 to 201611; bIncludes The Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHSD) codes for syndromes and chromosomal abnormalities11; cOnly one primary diagnosis may be selected in the STS-CHSD; dData reported for aortic stenosis without insufficiency; 8% of patients reported multiple levels of stenosis; eAvailable 2010 to 2016.
Values are median (interquartile range) or n (%). IAA, interrupted aortic arch; MCS, mechanical circulatory support.
procedure was the most common concomitant operation overall (15%) and was performed with autograft AVRs more frequently in children than in teens (31% vs 12%, P < .001). For the overall study population, mechanical valve AVR was performed in 28%, bioprosthetic AVR in 15%, and homograft AVR in 8%. Valve utilization by age at surgery is shown in Figure 1. Teens were twice as likely to receive a mechanical valve, compared with children (OR 2.29, 95% CI: 1.96 to 2.68, P < .001). Bioprosthetic
valves were also used with increasing frequency as age at operation increased (Figure 1). Over the study period, autograft utilization in children remained stable (Figure 2). Among teens, autograft utilization decreased while utilization of mechanical and bioprosthetic valves increased. Valve utilization within each age group by sex is shown in Figure 3. Among teens receiving either a mechanical or autograft AVR, 63% of girls received an autograft vs 48%
Ann Thorac Surg 2019;108:544-51
Table 2.
NELSON ET AL AORTIC VALVE REPLACEMENT IN CHILDREN
547
Operative Data by Valve Type and Age Group
Variable
3446 1594 1852
1716 (50) 1024 (64) 692 (37)
Mechanical
963 (28) 307 (19) 656 (35)
Bioprosthetic
501 (15) 122 (8) 379 (20)
Homograft
266 (8) 141 (9) 125 (7)
174 (125-229) 182 (135-238) 168 (119-223)
192 (148-241) 192 (145-242) 193 (151-240)
148 (108-202) 153 (110-212) 146 (107-200)
139.5 (102-187) 153 (101-201) 137 (103-183)
197 (151-270) 199 (152-296) 192 (150-261)
119 (87-161) 124 (91-166) 113 (83-155)
135 (101-173) 133 (99-171) 137 (103-175)
98 (71-132) 100 (71-130) 97 (72-133)
99.5 (74-134) 106 (74-134) 98 (74-134)
136 (104-193) 144 (111-201) 125 (102-180)
18 (11-32) 14 (10-23) 25 (12-35)
14 (8-30) 13 (8-23) 24 (15-54)
22 (9-32) 17 (12-23) 24 (9-34)
23 (15-60) 19 (14-52) 28 (18-60)
20 (14-34) 19 (4-36) 24 (16-32)
18 (4) 8 (7) 10 (3)
13 (5) 10 (7) 3 (2)
510 (15) 374 (24) 136 (7)
400 (23) 315 (31) 85 (12)
79 (8) 41 (13) 38 (6)
56 (2) 34 (2) 22 (1)
32 (2) 24 (2) 8 (1)
12 (1) 5 (2) 7 (1)
180 (5) 104 (6) 76 (4)
67 (4) 55 (5) 12 (2)
63 (6) 25 (8) 38 (6)
24 (5) 8 (7) 16 (4)
26 (10) 16 (11) 10 (8)
24 (2) 6 (2) 18 (3)
23 (5) 4 (3) 19 (5)
1 (0.4) 1 (0.7) 0 (0)
51 (2) 13 (0.8) 38 (2)
3 (0.2) 2 (0.2) 1 (0.1)
a Includes Ross with or without Konno and valvuloplasty converted to Ross with or without Konno; (n ¼ 87; 41 children, 46 teens).
4 (0.8) 1 (0.8) 3 (0.8)
b
8 (3) 4 (3) 4 (3)
Non-zero deep hypothermic cardiac arrest times
Values are n (%) or median (interquartile range).
of boys (P < .001). When valve choice was restricted to mechanical, bioprosthetic, or autograft, and categorized by need for anticoagulation therapy postoperatively (mechanical, yes; bioprosthetic and autograft, no), teen girls were significantly more likely than teen boys to receive a valve that did not require anticoagulation treatment postoperatively (OR 1.94, 95% CI: 1.51 to 2.49, P < .001). These differences were not appreciated in the younger age group (P ¼ .11). Among redo AVRs, mechanical valves were selected most commonly (56%), followed by bioprosthetic valves, autografts, and homografts (18%, 15%, and 12%, respectively; Supplemental Table 2). Autografts were selected most frequently after aortic valvuloplasty (54%).
Early Outcomes In general, early mortality after AVR in children and teens was low (1% overall; Table 3). The highest hospital
mortality was associated with homograft AVR (5% overall). Postoperative complications for each valve type are shown by age group in Table 3. For all valve types, major complications occurred more frequently among children than teens (11% vs 9%, P < .05). The highest rate of major complications was associated with homograft AVR in children (16%). There was no statistical difference in major complications for autograft recipients when stratified by age group and sex.
Practice Patterns by Center Volume After the lowest volume centers (fewer than 20 AV cases over study period) were excluded from the analytic cohort, 3227 records from 86 centers remained. Autograft utilization varied widely across individual centers (10th to 90th percentile: 21% to 71% of total AVR volume). The percent utilization of each valve type by center volume is
CONGENITAL HEART
Operations performed All Child Teen Cardiopulmonary bypass time, min All Child Teen Cross-clamp time, min All Child Teen Circulatory arrest time, minb All Child Teen Concomitant procedure Konno All Child Teen Coarctation/AA repair All Child Teen Mitral valve operation All Child Teen Aortic annulus enlargement All Child Teen
Autografta
All Valve Types
CONGENITAL HEART
548
NELSON ET AL AORTIC VALVE REPLACEMENT IN CHILDREN
Ann Thorac Surg 2019;108:544-51
Figure 1. Valve type utilization by age. Autografts were utilized frequently across all ages, and mechanical and bioprosthetic valve usage increased with age.
shown in Figure 4. High-volume centers had significantly higher autograft utilization compared with low-/mediumvolume centers (P < .001).
Comment This study examined AVR practice patterns and early outcomes in a nationwide cohort of children and teens, representing the largest series to date. We found AVR practice patterns vary widely across centers and patient groups, including important sex differences. Early mortality after AVR was low overall, but inhospital mortality and major complications were higher in the homograft group, compared with other AVR types. Early outcomes of autograft (Ross) AVR were excellent. This study also associated AVR practice patterns with center volume: high-volume centers were more likely to utilize an autograft, compared with low- or medium-volume centers.
Figure 2. Autograft utilization over time (shaded gray bars indicate totals). Although utilization among children (gray bars) remained stable during the study, autograft utilization in teens (black bars) decreased (teens, power trendline [dotted line] R2 ¼ 0.78). A power function a*xb provided the best R2 value for the trend.
Figure 3. Valve choice in (A) children and (B) teens by sex (black bars, boys; gray bars, girls). Teen girls received autograft and bioprosthetic valves more frequently than teen boys. *Significant difference (P < .05) between frequencies by odds ratios.
Valve Choice The most frequently utilized valve type among children was a pulmonary autograft. Pulmonary autografts and mechanical valves were used with similar frequency in teens, but teen girls were significantly more likely to receive an autograft than teen boys. Among redo AVRs, nearly one fifth had a prior autograft, and mechanical valves were selected most frequently. This study, which largely represents North American centers, demonstrates some differences from similar analyses from Europe and the Middle East. In a UK national database study of 568 children aged 1 to 16 years undergoing AVR,1 the most common operation was also Ross AVR, but the proportion of children undergoing Ross in the United Kingdom (77%) was higher than the same age range in our cohort (54%). Mechanical, bioprosthetic, and homograft valves were also used less frequently in the United Kingdom compared with the current study (17% vs 26%, 2% vs 12%, and 4% vs 8%, respectively).1 Valve choice is partially driven by native valve pathology, which differs by region. Whereas congenital heart disease is more common in Europe and North America, rheumatic disease predominates in other parts of the world, particularly those that are underrepresented in the STS-CHSD.2,12,13 A 2009 AVR study from Saudi Arabia demonstrated 58% of patients had rheumatic AV pathology, compared with 0.5% in the current study.2 Likewise, the proportion of patients with connective tissue disorders such as Marfan syndrome influences valve choice.3 Here, the large majority of patients with a diagnosis code for Marfan syndrome received a mechanical valve (51 of 78, 65%) or bioprosthetic valve (21 of 78, 27%).
Ann Thorac Surg 2019;108:544-51
Table 3.
NELSON ET AL AORTIC VALVE REPLACEMENT IN CHILDREN
549
Early Postoperative Outcomes by Age Group and Valve Type
Outcomes
Autografta (n ¼ 1716)
37 (1) 25 (2) 12 (0.7)
15 (0.9) 11 (1) 4 (0.6)
6 (4-8) 6 (4-8) 6 (4-8)
5 (4-7) 5 (4-7) 5 (4-6)
Mechanical (n ¼ 963)
Bioprosthetic (n ¼ 501)
7 (0.7) 4 (1) 3 (0.5)
3 (0.6) 3 (2) 0 (0)
7 (5-10) 7 (5-11) 6 (5-9)
5 (4-7) 6 (4-9) 5 (4-7) 31 (6) 14 (12) 17 (4)
Homograft (n ¼ 266) 12 (5) 7 (5) 5 (4) 7 (5-15) 7 (5-16) 6 (4-13)
334 (10) 174 (11) 160 (9)
153 (9) 95 (9) 58 (8)
115 (12) 42 (14) 73 (11)
35 (13) 23 (16) 12 (10)
58 (2) 37 (2) 21 (1)
34 (2) 27 (3) 7 (1)
15 (2) 3 (1) 12 (2)
4 (0.8) 4 (3) 0 (0)
5 (2) 3 (2) 2 (2)
52 (2) 33 (2) 19 (1)
26 (2) 18 (2) 8 (1)
13 (1) 7 (2) 6 (0.9)
3 (0.6) 2 (2) 1 (0.3)
10 (4) 6 (4) 4 (3)
186 (6) 94 (6) 92 (5)
93 (6) 52 (5) 41 (6)
60 (7) 23 (8) 37 (6)
13 (3) 7 (6) 6 (2)
20 (8) 12 (9) 8 (7)
94/1994 (5) 46/894 (5) 48/1100 (5)
43/953 (5) 30/572 (6) 13/381 (4)
31/558 (6) 9/164 (6) 22/394 (6)
10/340 (3) 3/85 (4) 7/255 (3)
10/143 (8) 4/73 (6) 6/70 (9)
a Includes Ross and Ross-Konno; bIncludes complications collected by The Society of Thoracic Surgeons Congenital Heart Surgery Database 2000 to 2016; cIncludes intraaortic balloon pump, ventricular assist device, extracorporeal membrane oxygenation, or centrifugal pump system; dIncludes cardiac and noncardiac reoperations, and reoperations for bleeding; eOnly available 2010 to 2016.
Values are median (interquartile range) or n (%). MCS, mechanical circulatory support.
Early Mortality We found much lower inhospital mortality for AVRs in children and teens compared with pooled 30-day mortality estimates in a 2015 systematic review and metaanalysis of pediatric AVR, but studies in that analysis included neonates and infants.14 Early mortality was similar between Ross and mechanical AVR in a young adult population from Australia and New Zealand, but the investigators further reported a significant 20-year survival advantage for Ross.15 A 2016 UK national database analysis indicated early mortality among children aged 1 to 16 years was higher after mechanical compared with Ross AVR, and only Ross patients enjoyed long-term survival approaching that of the general population.1 A survival advantage for children with congenital AV disease undergoing the Ross operation compared with mechanical AVR has been reported by others, as well.1,3 In a pediatric AVR competing risks analysis, the use of a prosthesis other than an autograft was a significant predictor of death.5
Excellent autograft durability and superior quality of life has also been reported for young patients after autograft compared with mechanical AVR.2,5,6,16,17 Mechanical valves require life-long anticoagulation therapy and therefore present lifestyle challenges, risk of bleeding and thromboembolism, and the risk of pregnancy-related complications.18 For these reasons, mechanical valves are often avoided for children pursuing contact sports and girls approaching child-bearing years. In this analysis, teen girls were significantly more likely than teen boys to receive an autograft, possibly owing to child-bearing potential and a desire to avoid anticoagulation therapy. Homograft AVR was utilized infrequently, and the inhospital mortality was higher than other valve types. These results were not unexpected, as homografts are generally reserved for “bail out” operations when the autograft is not acceptable or available, and a mechanical valve is not preferred for infectious, size, or anticoagulation reasons. Supporting this theory, patients with endocarditis were approximately eight times more likely to
CONGENITAL HEART
Hospital mortality All Child Teen Postoperative length of stay, d All Child Teen Any major complicationb All Child Teen Postoperative cardiac arrest All Child Teen Postoperative MCSc All Child Teen Unplanned reoperationd All Child Teen Unplanned readmissione All Child Teen
All (N ¼ 3446)
550
NELSON ET AL AORTIC VALVE REPLACEMENT IN CHILDREN
Ann Thorac Surg 2019;108:544-51
CONGENITAL HEART
collect long-term follow-up data, so late reinterventions and outcomes are unknown. Nevertheless, these results are useful as a tool for counseling patients and families and should provide a new benchmark for early outcomes after AVR in children and teens. Although we noted a decline in utilization of the Ross operation between the ages of 17 and 18 years in this database, some AVR patients in this age group are treated by adult cardiac surgeons who do not contribute to the STS-CHSD. Future work should evaluate both the STSCHSD and the STS Adult Heart Surgery Database to better estimate national practice patterns for this population. Figure 4. Frequency of valve choice by center volume, categorized into low (gray bars), medium (black bars), and high (open bars) tertiles, based on aortic valve surgery volume during the study. For each valve type there was a correlation between center volume and frequency of valve utilization (gamma test P < .05 for all).
receive a homograft valve than any other type (OR 8.15, 95% CI: 5.79 to 11.48, P < .001). We also found patients undergoing urgent, emergent, or salvage operations (compared with elective) were significantly more likely to receive a homograft valve (OR 8.29, 95% CI: 4.61 to 14.94, P < .001; n ¼ 856; urgency data available 2014 to 2016). Patients receiving homografts were also more likely to be on mechanical ventilation or in shock before surgery, compared with recipients of other valve types, but neither of those risks had an effect on mortality in homograft recipients (data not shown). These observations may point to appropriate clinical decision-making. Although it has been suggested that homograft AVR may be more expedient compared with autograft AVR, cross-clamp and bypass times for those operations were similar in this study, and that observation has been noted by others.19
Center Volume and Utilization Patterns Center volume for pediatric AV surgery has not been previously studied in the context of valve choice. Here, correlations between center volume and valve choice were found for all valve types. Further study is needed to understand these differences.
Study Limitations This study offers a contemporary, multiinstitutional analysis of practice patterns and early outcomes in children and teens undergoing AVR, and its limitations are common to others utilizing the STS-CHSD. Although we excluded records that were missing key operative or outcome variables, specific valve manufacturer and valve size were frequently not reported. In addition, this study was unable to assess reasons for valve selection. Because detailed anatomic, echocardiographic, and complete prior operative data are not captured in the STS-CHSD, we were unable to determine which patients were truly eligible for a particular valve type (ie, pulmonary autograft). Likewise, in this cohort of patients treated with AVR, it was impossible to comment on the appropriateness of surgical strategy, timing, or potential use of catheter-based options. Lastly, the STS-CHSD does not
Conclusion Practice patterns for AVR vary widely across centers and patient groups, but the majority of children receive pulmonary autografts. Highest utilization of autografts was observed in high-volume centers. Among teens, there are important sex differences in valve selection. In children and teens, AVR is associated with low inhospital mortality. Future work is needed to understand valve choice decision making and longer-term outcomes by valve type. The data for this research were provided by The Society of Thoracic Surgeons National Database Participant User File Research Program. Data analysis was performed at the investigators’ institutions.
References 1. Alsoufi B, Al-Halees Z, Manlhiot C, et al. Mechanical valves versus the Ross procedure for aortic valve replacement in children: propensity-adjusted comparison of long-term outcomes. J Thorac Cardiovasc Surg. 2009;137:362-370.e9. 2. Karamlou T, Jang K, Williams WG, et al. Outcomes and associated risk factors for aortic valve replacement in 160 children: a competing-risks analysis. Circulation. 2005;112: 3462-3469. 3. Brown JW, Patel PM, Ivy Lin JH, et al. Ross versus non-Ross aortic valve replacement in children: a 22-year single institution comparison of outcomes. Ann Thorac Surg. 2016;101: 1804-1810. 4. Sharabiani MT, Dorobantu DM, Mahani AS, et al. Aortic valve replacement and the Ross operation in children and young adults. J Am Coll Cardiol. 2016;67:2858-2870. 5. Turrentine MW, Ruzmetov M, Vijay P, et al. Biological versus mechanical aortic valve replacement in children. Ann Thorac Surg. 2001;71(Suppl):356-360. 6. Nelson JS, Pasquali SK, Pratt CN, et al. Long-term survival and reintervention after the Ross procedure across the pediatric age spectrum. Ann Thorac Surg. 2015;99:2086-2094 [discussion 2094-2095]. 7. Tan Tanny SP, Yong MS, d’Udekem Y, et al. Ross procedure in children: 17-year experience at a single institution. J Am Heart Assoc. 2013;2:e000153. 8. Jacobs JP, Shahian DM, D’Agostino RS, et al. The Society of Thoracic Surgeons National Database 2018 annual report. Ann Thorac Surg. 2018;106:1603-1611. 9. Morales DL, Khan MS, Turek JW, et al. Report of the 2015 Society of Thoracic Surgeons congenital heart surgery practice survey. Ann Thorac Surg. 2017;103:622-628. 10. Ross DN. Replacement of aortic and mitral valves with a pulmonary autograft. Lancet. 1967;2:956-958. 11. The Society of Thoracic Surgeons. STS Congenital Heart Surgery Database data specifications, version 3.3. 2015.
Ann Thorac Surg 2019;108:544-51
12. Liu FZ, Xue YM, Liao HT, et al. Five-year epidemiological survey of valvular heart disease: changes in morbidity, etiological spectrum and management in a cardiovascular center of Southern China. J Thorac Dis. 2014;6:1724-1730. 13. Triki F, Jdidi J, Abid D, et al. Characteristics, aetiological spectrum and management of valvular heart disease in a Tunisian cardiovascular centre. Arch Cardiovasc Dis. 2017;110: 439-446. 14. Etnel JR, Elmont LC, Ertekin E, et al. Aortic valve replacement in children: a systematic review and meta-analysis. J Thorac Cardiovasc Surg. 2016;151:143-152.e1-3. 15. Buratto E, Shi WY, Wynne R, et al. Improved survival after the Ross procedure compared with mechanical aortic valve replacement. J Am Coll Cardiol. 2018;71:1337-1344.
NELSON ET AL AORTIC VALVE REPLACEMENT IN CHILDREN
551
16. Notzold A, Huppe M, Schmidtke C, et al. Quality of life in aortic valve replacement: pulmonary autografts versus mechanical prostheses. J Am Coll Cardiol. 2001;37:1963-1966. 17. Saleeb SF, Newburger JW, Geva T, et al. Accelerated degeneration of a bovine pericardial bioprosthetic aortic valve in children and young adults. Circulation. 2014;130:51-60. 18. van Hagen IM, Roos-Hesselink JW, Ruys TP, et al. Pregnancy in women with a mechanical heart valve: data of the European Society of Cardiology Registry of Pregnancy and Cardiac Disease (ROPAC). Circulation. 2015;132:132-142. 19. Woods RK, Pasquali SK, Jacobs ML, et al. Aortic valve replacement in neonates and infants: an analysis of The Society of Thoracic Surgeons Congenital Heart Surgery Database. J Thorac Cardiovasc Surg. 2012;144:1084-1089.
When a child requires aortic valve replacement (AVR), prosthetic valve options are limited and suboptimal because there is no valve substitute that is capable of providing lifelong durability, optimal hemodynamics, zero thrombogenicity, and the potential to grow with the child. The only option that provides a living valve substitute is the Ross procedure, at the cost of double prosthetic valve disease. Children should grow up being able to live life to the fullest, but the harsh reality is that kids who undergo AVR—regardless the valve type implanted—face many challenges in doing so. In this issue of The Annals of Thoracic Surgery, the study by Nelson and colleagues1 provides an insightful overview of current United States practice patterns and early outcomes of AVR in children ages 1 to 18 years. Half of AVRs concern the Ross procedure; almost 3 of 10 children receive a mechanical valve, and the remainder 23% receive bioprostheses and homografts. The study also illustrates that practice variation is obvious across centers, patient age, and sex. Practice variation across patient age and sex is probably a good thing, but practice variation across centers may not, because it possibly reflects unequal access to all available valve substitutes. Aortic valve repair was not included in this study, but it would be worthwhile to investigate whether this surgical approach was underused given that more than one-third of patients presented with aortic regurgitation. The observation that the Ross procedure is performed more often in high-volume centers is comforting because it requires specific surgical skills and expertise,2 and one may expect that high-volume centers will perform better than those where only a few Ross operations take place every year. Interestingly, Ross procedure use in teens seems to have decreased in the United States over time (Figure 2), and in patients between the ages of 14 and 18 years, a mechanical prosthesis is the most often implanted valve substitute (Figure 4).1 The study does not provide possible reasons for this shift, but it may be because several valve
Ó 2019 by The Society of Thoracic Surgeons Published by Elsevier Inc.
guidelines in the past decade have not been supportive of the Ross procedure in young adult patients—without a solid evidence base—in some instances even removing the mention of the Ross procedure as an option, while at the same, time a growing body of evidence is accumulating in favor of the Ross procedure.3 Although a mechanical prosthesis is easier to implant and provides a durable solution, it comes with the burden of anticoagulation, suboptimal hemodynamics, and a ticking sound, all translating to suboptimal life expectancy and quality of life. In particular, these are severe limitations in the most active phase of life (teen years, young adulthood) that need to be addressed, together with the pros and cons of other prosthetic valve options, in the process of shared prosthetic valve selection. The mechanisms underlying the observed practice variation over time and across centers deserve further investigation, and perhaps it is time for evidence-based guidelines to drive equal access to care and optimization of patient-tailored treatment choices for children in need of AVR. Johanna J. M. Takkenberg, MD, PhD Department of Cardio-Thoracic Surgery, RG633 Erasmus University Medical Center Rotterdam PO Box 2040 Rotterdam 3000CA, The Netherlands email:
[email protected]
References 1. Nelson JS, Maul TM, Wearden PD, Pasquali SK, Romano JC. National practice patterns and early outcomes of aortic valve replacement in children and teens. Ann Thorac Surg. 2019;108: 544-551. 2. Bouhout I, Ba PS, El-Hamamsy I, Poirier N. Aortic valve interventions in pediatric patients. Semin Thorac Cardiovasc Surg. 2019;31:277-287. 3. Misfeld M, Borger MA. The Ross procedure: time to reevaluate the guidelines. J Thorac Cardiovasc Surg. 2019;157:211-212.
0003-4975/$36.00 https://doi.org/10.1016/j.athoracsur.2019.05.012
CONGENITAL HEART
INVITED COMMENTARY