| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2004;128:703-709
© 2004 The American Association for Thoracic Surgery
Surgery for congenital heart disease |
a Division of Cardiology, Department of Pediatrics, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
b Division of Cardiovascular Surgery, Department of Surgery, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
Read at the Eighty-fourth Annual Meeting of The American Association for Thoracic Surgery, Toronto, Ontario, Canada, April 25-28, 2004.
Received for publication April 23, 2004; revisions received July 24, 2004; accepted for publication July 26, 2004. * Address for reprints: Brian McCrindle, MD, The Hospital for Sick Children, 555 University Ave, Toronto, Ontario, Canada M5G 1X8 (E-mail: brian.mccrindle{at}sickkids.ca).
 |
Abstract |
|---|
 Top Abstract Patients and methodsResultsDiscussionDiscussionReferences |
|---|
METHODS: We included 104 children undergoing at least one mitral valve replacement between 1980 and 2003 and reviewed clinical records. Competing-risks methodology was used to determine time-related prevalence and associated risk factors after initial mitral valve replacement for death and repeat replacement.
RESULTS: The underlying mitral valve disease was congenital in 83%, rheumatic in 13%, Marfan syndrome in 3%, and isolated endocarditis in 1%, with 64% having primarily regurgitation, 16% having stenosis, 20% having both, and 32% having undergone previous valvotomy, valvuloplasty, or repair. There were 137 valve replacements, with 26 patients having more than one. Valve prosthesis type was St Jude Medical in 37%, Björk-Shiley in 25%, Carbomedics in 20%, Ionescu-Shiley in 10%, and other types in 8%. Both early and late complications were common. Median age at the initial replacement was 5.9 years (range, birth to 19 years). Competing-risks analysis predicted 19% to have died at 15 years after initial replacement, with risk factors including noncongenital valve morphology, lower weight, and longer duration of cardiopulmonary bypass. A repeat replacement was predicted for 71%, with risk factors including the presence of multiple left-heart obstructive lesions and Ionescu-Shiley valve prosthesis.
CONCLUSIONS: Mitral valve replacement might be necessary in children with extremely dysplastic valves and severe hemodynamic impairment or after failed repair. However, with the appropriate selection of the prosthetic valve and reduction of cardiopulmonary bypass time, surgeons might decrease mortality and increase prosthesis longevity.
|
Patients and methods |
|---|
|
TopAbstractPatients and methodsResultsDiscussionDiscussionReferences |
|---|
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 echocardiographic measurements.
Data analysis
Data are described as frequencies, medians with ranges, and means with SDs. Where data are missing, the number of nonmissing values is given. Both replaced valve failure with subsequent 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 bagging.10 Competing-risks analysis was used in a manner previously reported.11 Post-MVR echocardiographic assessments of mean diastolic valve gradient and estimated right ventricular systolic pressure were modeled, and risk factors were sought by using mixed linear regression.
|
Results |
|---|
|
TopAbstractPatients and methodsResultsDiscussionDiscussionReferences |
|---|
MVR
Of the 104 patients, during the study interval, 78 patients had only 1 MVR (18 subsequently died), 20 patients had 2 MVRs (9 subsequently died), 5 patients had 3 MVRs, and 1 patient had 4 MVRs (all currently surviving) for a total of 137 MVR episodes. The characteristics at the MVR episode are shown in Table 1. The distribution of ages at the initial MVR was bimodal, with an early peak around the 25th percentile of 1.4 years and a later smaller peak around the 75th percentile of 12.3 years. The age at initial MVR was significantly younger in those with congenital heart disease versus those with other pathologic conditions (P= .004), although those with congenital heart disease predominated throughout all ages. Total mortality for all patients regardless of subsequent procedures is shown in Figure 1. The analysis of early and late complications relates to the total 137 episodes of MVR.
|
|
Late complications
There was also important late morbidity after the 117 MVRs not complicated by death or subsequent MVR within 30 days of MVR. There was an episode of important bleeding related to anticoagulation with 6 (5%) MVRs, with 2 episodes of embolism caused by thrombus and 3 with transient ischemic attacks. There were an additional 6 (5%) permanent pacemaker implantations. Late arrhythmias occurred with 16 (14%) MVRs. There were 6 (5%) later episodes of endocarditis, 3 (3%) of valve thrombosis, and 17 (15%) of perivalvular leak. Anticoagulation at latest follow-up assessment included warfarin in 86 MVRs, aspirin in 15 MVRs, persantin in 9 MVRs, and no anticoagulation in 7 MVRs. Only one patient with underlying congenital heart disease had cardiac transplantation at 9 years after a second MVR.
Competing risks for death or subsequent prosthesis replacement after initial MVR
The competing-risks analysis relates outcomes only after the initial MVR for each of the 104 patients. During follow-up of the 104 initial MVR episodes, 26 valves were subsequently replaced, and 18 patients died without a further MVR. The hazard function for time-related transition to a second MVR was characterized by a very brief and steep early hazard phase (4 early-phase events) and a prolonged late hazard phase (22 late-phase events). The hazard function for time-related transition to death without a second MVR was characterized by a very steep early hazard phase only. The competing risks for the 2 events showed that after 15 years from initial MVR, 71% of valves had been replaced, 19% of patients had died without a further MVR, and 10% were surviving without a further MVR (Figure 2). Incremental risk factors each for early-phase death and for late-phase subsequent replacement after initial MVR were sought and are shown in Table 2. A stratified graph of valve longevity by different types of initial prosthesis shows that mechanical valves (St Jude Medical [St Jude Medical, Inc, St Paul, Minn], Björk-Shiley [Shiley, Inc, Irvine, Calif], and Carbomedics [Sulzer Carbomedics, Inc, Austin, Tex)]had similar and better longevity than tissue valves (Ionescu-Shiley [Shiley] and other tissue valves, Figure 3). From the multivariable competing-risks analysis, predicted outcomes for a patient with good characteristics, including weight at initial MVR of 20 kg, underlying AVSD after repair, and a cardiopulmonary bypass (CPB) time of 120 minutes with placement of a mechanical valve prosthesis, would show a 6% chance of death with a second MVR, a 74% chance of a second MVR, and a 20% chance of survival with the original prosthesis in place 15 years later. In contrast, for a patient weighing 5 kg with underlying multiple left-heart obstructive lesions who undergoes placement of an Ionescu-Shiley valve prosthesis with a prolonged CPB time of 480 minutes, predicted outcomes 15 years later would show a 96% chance of death without a second MVR and a 4% chance of a second MVR with almost no chance of survival with the original prosthesis in place.
|
|
|
Factors associated with serial echocardiographic measurements after MVR
Mean echocardiographic gradient across the replaced MV was available after 54 MVRs, with a total of 256 measurements performed up to a maximum interval of 11.7 years (mean, 3.2 years). Significant independent factors associated with a higher mean gradient during follow-up included an increased interval from MVR (increase of 0.76 mm Hg per year; P< .001) and a smaller absolute size of the MV implanted (increase of 0.40 mm Hg per 1 mm decrease in diameter; P< .001). After controlling for these 2 factors, no other factor was significantly associated with follow-up gradient, including the age and size of the patient, the cardiac morphology, or the type of prosthesis that was implanted.
Estimated right ventricular systolic pressure was available after 41 MVRs, with a total of 157 measurements performed up to a maximum interval of 11.7 years (mean, 3.0 years). The only significant independent factor associated with a higher right ventricular systolic pressure during follow-up was an increased interval from MVR (increase of 1.03 mm Hg per year; P= .002).
|
Discussion |
|---|
|
TopAbstractPatients and methodsResultsDiscussionDiscussionReferences |
|---|
Mortality
Similar to our study, reported mortality in young children with MV disease has been high, particularly for patients undergoing MVR, with most of the deaths occurring early. Kadoba and colleagues5 reported a 36% likelihood of death in the first year of life and a 20% likelihood of late death. Hospital mortality has significantly decreased in recent years, ranging from 12% to 33%.1 High mortality has been associated with a greater ratio of prosthetic valve size to body weight. Ultimately, even the smallest available prosthetic valves are considered too large for placement in small patients, resulting in higher mortality.2,3 Additional surgical options for small patients include further attempts at valve repair, abandonment of the left ventricle, and performance of an atrial septectomy and Damus-Kaye-Stanzel–type anastomosis or transplantation.12
We noted that initial valve morphology other than congenital heart disease was a risk for mortality. This group included 14 patients with rheumatic heart disease (1 died), 1 with endocarditis (who died), and 3 with Marfan syndrome (2 died). The benefits of reconstructive surgery are important for young patients with Marfan syndrome, who usually have a higher risk of aortic dissection.13 Recent reports suggest that MV reconstruction in the setting of active or healed endocarditis is feasible, can be performed with low operative mortality and morbidity, and should be considered rather than MVR for mitral regurgitation caused by endocarditis.14 A 15-year study of rheumatic heart disease showed that the mean age at hospitalization was 11 ± 3 years.15 In our study, data suggest that older and younger patients might have different underlying diagnoses, with higher-risk noncongenital valve disease more likely in adolescents.
Longer duration of CPB was also identified as a risk factor for death without a repeat MVR. Bypass time might be acting as a proxy for disease complexity. All organ systems could be adversely affected by CPB, primarily the heart, lungs, brain, and kidneys. The risk of serious complications depends on the age of the patient, how ill they are at the time of the operation, and the complexity of the operation to be performed. The mean CPB time for the entire cohort was 154 ± 75 minutes compared with 317 ± 113 minutes for patients who died without repeat MVR. In 8 of the 10 MVRs with the longest duration of bypass, there were other operations being performed concomitantly, including aortic valve replacement or reconstruction with the Ross or Konno procedure, repair of the AVSD, bidirectional cavopulmonary anastomosis, and Fontan procedure.
Prosthetic valve longevity
We identified 2 risk factors for decreased time to a repeat MVR after initial MVR: (1) the presence of multiple left-heart obstructive anomalies and (2) Ionescu-Shiley valve type. Left-heart lesions included aortic coarctation, aortic valve stenosis or regurgitation, and subaortic obstruction. The presence of other left-heart anomalies might place an added burden on the prosthetic valve, particularly if there is associated aortic stenosis.
The choice of prosthetic valve type is still debated. In our study the majority of patients (37%) underwent MVR with a St Jude Medical prosthetic valve. The advantages of a mechanical prosthesis include better hemodynamic flow characteristics for valves that are small in size, with a lower profile associated with less potential for left ventricular outflow tract obstruction and longer durability.1,6,16 Anticoagulation regimen remains a matter of concern for patients with mechanical valves. We noted that 10% of patients received Ionescu-Shiley valves, which was characterized as a risk factor for reduced valve longevity. Zweng and associates6 report that 4 of 10 patients with a bioprosthesis underwent reoperation within 17 to 47 months of an initial MVR. Although bioprosthetic valves do not require anticoagulation, they do have problems of limited durability.1,4
The results of this study show that the need for multiple MVR might increase mortality. The most significant increase in mortality was seen after 2 operations. Of the 20 patients who had 2 MVRs, 9 died without a third MVR. Five patients had 3 MVRs with no deaths, and 1 patient is a current survivor after 4 MVRs. We also noted that patients who required a subsequent MVR within 4 months of the initial MVR had higher mortality. In addition, patients aged less than 1 year at the time of their initial MVR were at high risk for both early death and an early subsequent MVR (with associated high mortality).
Late complications were common. Prosthetic valve thrombosis and embolism are comparatively frequent complications after valve replacement operations.7,17,18 Previous studies have reported rates of thromboembolism of 0.7% to 2% per year.1,4 Bacterial endocarditis and bleeding after MVR were reported in previous studies.5,7,9,19,20
Echocardiography remains the mainstay of follow-up assessment. We noted that significant independent factors associated with higher mean echocardiographic diastolic gradient during follow-up included an increasing interval from MVR (gradient worsens in a linear manner with time) and a smaller absolute size of the prosthesis implanted. Matsuoka and coworkers21 noted that 3 children with MVRs with a St Jude Medical prosthesis of 23 mm or smaller had greater mean echocardiographic diastolic gradients than those who received prostheses of 25 mm or larger. We also noted that the only significant factor associated with higher right ventricular systolic pressure during follow-up was an increased interval from MVR (pressure increases with time).
Summary
The results of our study show that lower patient weight, valve morphology other than congenital heart disease, and longer duration of CPB at initial MVR are significant predictors of death without a repeat MVR. Furthermore, the presence of multiple left-heart obstructive anomalies and use of the Ionescu-Shiley or bioprosthetic valve types were identified as significant risk factors for earlier repeat MVR. Ultimately, valve replacement might be necessary in children with extremely dysplastic valves and severe hemodynamic impairment or after failed repair. However, with the appropriate selection of the prosthetic valve and reduction of CPB time, surgeons might decrease morbidity and mortality in pediatric patients undergoing MVR.
|
Discussion |
|---|
|
TopAbstractPatients and methodsResultsDiscussionDiscussionReferences |
|---|
Your results demonstrate some expected findings but also some unexpected findings as well. I have a few questions.
Your study encompassed a 23-year period. Have you analyzed results, looked for any period differences, for example, comparing the last decade with the previous 13 years?
Can you explain why the underlying valve pathology, other than congenital, was identified as an incremental risk factor for death, particularly because these seem to occur mostly in the older and larger patients?
Why is the mortality after the second MVR higher despite potentially larger patient size? Were there a number of second MVRs during the same hospitalization as the initial MVR? In conjunction with this question, why is the hazard function for death without repeat MVR different for the initial MVR versus a second MVR?
Given the model that you have, what would you recommend for a 4-kg child with infantile Marfan syndrome and severe MV regurgitation, which I did not repair very well 2 days ago? What should I do?
Last, is there any role for a bioprosthetic MVR in the pediatric population? What would the longevity and mortality be for a bioprosthetic MVR in an 18-year-old woman?
I commend the authors for their efforts in studying this difficult subject.
Dr Caldarone. Thank you for your kind comments. We agree that the question of an era effect is appropriate. The year of operation was entered into the analysis and did not come out as a significant risk factor; however, that does not mean an era effect is absent. As you pointed out, a small series looking at a diverse population requires analysis of a large number of variables, and therefore a small effect in any one variable might not be detected. I would suspect there is an era effect, but we were unable to detect one.
With regard to the identification of noncongenital diagnosis as a risk factor for death, I share with you some concern with the use of a basket term encompassing a diverse group of diagnoses, especially a negative basket term meaning all diagnoses of noncongenital origin. This technique lumps together these diagnoses, and in aggregate they became statistically significant, when in isolation they did not. Once again, this underscores the problem with a large number of variables and a small data set.
The noncongenital deaths were basically related to rheumatic disease, endocarditis, Marfan syndrome, and Shone syndrome. In fact, 3 of the deaths were quite late deaths at 4, 4, and 10 years after the operation, respectively. Therefore, in terms of decision making in a patient with noncongenital MV disease, I really do not think that should necessarily sway your decision about whether to implant a valve.
With regard to the deaths after a second MVR, you have very nicely pointed out one of the problems with our analysis. Patients who had a short interval from the first to the second MVR and then subsequently died after the second MVR were not identified in the analysis looking at the first MVR. These were censored events because they recurred after a second MVR. That might be an inappropriate way to deal with this statistically because a patient who requires 2 MVRs in the same hospital admission clearly is not the same as a patient showing up electively for a second MVR years later. Therefore, your point in that respect is well taken.
With regard to the poor-risk patient you described, there is no easy answer. Repair, repair, and replace when necessary is what we all use as our decision making, and it seems appropriate in the patient you described. I think attempts at persistent replacement need to be tempered with the fact that a patient with poor physiologic status as a result of multiple futile attempts at repair might be worse off than just biting the bullet up front and replacing the valve. Therefore, in short, I would say the patient should have one trip to the operating room for attempted repair. If the patient subsequently requires a return to the operating room, I would advocate another attempt at repair, and if it does not look satisfactory in the operating room, the valve should be replaced.
Finally, what is the ideal patient for a tissue prosthesis? As patients get older, I suspect the tissue prostheses will last longer. We could run our model and see if that actually is true. Ideally, in a female patient you would want to skip the childbearing age with a bioprosthesis in place, and after childbearing age, you would want to be able to have a mechanical prosthesis for the rest of her life. Therefore, I would say that would be the ideal indication.
Dr Gerhard Ziemer (Tuebingen, Germany) I have 2 questions. Did you ever encounter any secondary tricuspid valve problems by distortion when you had to use valves in the 4 to 6 mm/kg range, such as in neonates or young infants?
Also, do you have any clue why the Carbomedics valve turned out to be better in your risk analysis? Has it something to do with the design?
Dr Caldarone. Personally, I have not run into distortion of the tricuspid valve from the MV prosthesis; however, I have run into patients with pulmonary hypertension, which causes tricuspid insufficiency and might require tricuspid repair at the time of MVR.
With regard to the Carbomedics valve, there were no statistically significant differences between the mechanical valve types. Although on the survival curve it looked like there were differences, the differences were not statistically significant.
Dr Carlo Marcelletti (Palermo, Italy) I would like to ask about the incidence of complete atrioventricular block in MVR in the setting of partial or complete AVSD.
Dr Caldarone. It was 10% in the entire series and 15% in patients with AVSDs.
Dr C. Troconis (Caracas, Venezuela) Is there any role for an autologous tricuspid valve translocation from the tricuspid site for the MVR and implanting a cryopreserved bioprosthetic valve in the right side in the pediatric population?
Dr Caldarone. No, that was not done in this series.
Dr Ross M. Ungerleider (Portland, Ore) I assume that 5 kg was the smallest patient in your series, and yet sometimes we see patients who are smaller. Is there any role for the Ross II operation, maybe with an aortic homograft and a small polytetrafluoroethylene tube? Have you tried any novel approaches like that?
Dr Caldarone. Personally, I have not. There were either 2 or 3 homograft replacements in the series, but because there were so few, we could not really identify that as a risk factor or benefit.
|
References |
|---|
|
TopAbstractPatients and methodsResultsDiscussionDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. Alsoufi, C. Manlhiot, B. W. McCrindle, C. C. Canver, A. Sallehuddin, S. Al-Oufi, M. Joufan, and Z. Al-Halees Aortic and mitral valve replacement in children: is there any role for biologic and bioprosthetic substitutes? Eur. J. Cardiothorac. Surg., July 1, 2009; 36(1): 84 - 90. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. J. Hickey, G. Jung, C. Manlhiot, A. G. Sakopoulos, C. A. Caldarone, J. G. Coles, G. S. Van Arsdell, and B. W. McCrindle Infective endocarditis in children: native valve preservation is frequently possible despite advanced clinical disease Eur. J. Cardiothorac. Surg., January 1, 2009; 35(1): 130 - 135. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. S. Selamet Tierney, F. A. Pigula, C. I. Berul, J. E. Lock, P. J. del Nido, and D. B. McElhinney Mitral valve replacement in infants and children 5 years of age or younger: Evolution in practice and outcome over three decades with a focus on supra-annular prosthesis implantation J. Thorac. Cardiovasc. Surg., October 1, 2008; 136(4): 954 - 961. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. P. Malhotra, F. Lacour-Gayet, D. N. Campbell, S. Miyamoto, D. R. Clarke, M. L. Dines, D. D. Ivy, and M. B. Mitchell Outcomes of Reparative and Transplantation Strategies for Multilevel Left Heart Obstructions With Mitral Stenosis Ann. Thorac. Surg., October 1, 2008; 86(4): 1305 - 1310. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Oppido, B. Davies, D. M. McMullan, A. D. Cochrane, M. M.H. Cheung, Y. d'Udekem, and C. P. Brizard Surgical treatment of congenital mitral valve disease: Midterm results of a repair-oriented policy. J. Thorac. Cardiovasc. Surg., June 1, 2008; 135(6): 1313 - 1321.e4. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Monagle, E. Chalmers, A. Chan, G. deVeber, F. Kirkham, P. Massicotte, and A. D. Michelson Antithrombotic Therapy in Neonates and Children: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition) Chest, June 1, 2008; 133(6_suppl): 887S - 968S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. N. Anderson Comment on the "actual" problem editorial J. Thorac. Cardiovasc. Surg., February 1, 2008; 135(2): 457 - 459. [Full Text] [PDF]
|
|
|
|
 |
|
 R. P Akhtar, A. R Abid, H. Zafar, S. S Sheikh, M. A Cheema, and J. S Khan Prosthetic Valve Replacement in Adolescents with Rheumatic Heart Disease Asian Cardiovasc Thorac Ann, December 1, 2007; 15(6): 476 - 481. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. A. Qamar, C. S. Goldberg, E. J. Devaney, E. L. Bove, and R. G. Ohye Current Risk Factors and Outcomes for the Arterial Switch Operation Ann. Thorac. Surg., September 1, 2007; 84(3): 871 - 879. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Ackermann, G. Balling, A. Eicken, T. Gunther, C. Schreiber, and J. Hess Replacement of the systemic atrioventricular valve with a mechanical prosthesis in children aged less than 6 years: Late clinical results of survival and subsequent replacement J. Thorac. Cardiovasc. Surg., September 1, 2007; 134(3): 750 - 756. [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. Karamlou, K. Jang, W. G. Williams, C. A. Caldarone, G. Van Arsdell, J. G. Coles, and B. W. McCrindle Outcomes and Associated Risk Factors for Aortic Valve Replacement in 160 Children: A Competing-Risks Analysis Circulation, November 29, 2005; 112(22): 3462 - 3469. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
