|
|
||||||||
Eur J Cardiothorac Surg 2007;32:133-142. doi:10.1016/j.ejcts.2007.02.025
Copyright © 2007, European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved
a Department of Pediatric Cardiology and Intensive Care Medicine, Hannover Medical School, Hannover, Germany
b Department of Cardiac, Thoracic, Transplantation and Vascular Surgery, Hannover Medical School, Hannover, Germany
c Department of Cardiology and Angiology, Hannover Medical School, Hannover, Germany
Received 23 October 2006; received in revised form 20 February 2007; accepted 23 February 2007.
* Corresponding author. Address: Hannover Medical School, Department of Pediatric Cardiology and Intensive Care Medicine, Carl-Neuberg-Str.1, K 10, D-30625 Hannover, Germany. Tel.: +49 511 532 9424; fax: +49 511 532 9832. (Email: boethig.dietmar{at}mh-hannover.de).
| Abstract |
|---|
|
|
|---|
Key Words: Right ventricular outflow tract Valved conduit Homograft Hemodynamic performance Long-term follow-up
| 1. Introduction |
|---|
|
|
|---|
In the last decades, the approach to homograft degeneration has changed remarkably regarding methods of diagnosis (color Doppler echo, MRI) and treatment policy (e.g. earlier reoperation [4], catheter based interventions [5]) and thus has created difficulties for the evaluation of valuable long-term data.
This paper presents the follow- up results of 188 consecutive homografts that were implanted in pulmonary position at our institution between October 1985 and October 2004. The evaluation focuses on homograft degeneration. We have tried to minimize the bias that is inherent due to repeated changes in the medical environment.
| 2. Patients and methods |
|---|
|
|
|---|
Ninety percent of the implants were of pulmonary origin and 10% were aortic roots. All homografts except one was cryopreserved. Pulmonary homografts have been our first choice, and aortic homografts were implanted only if no appropriately sized pulmonary graft was available. A slight homograft oversizing was preferred in young patients. However, the limited availability of small homografts and the lack of alternatives led to a broad range of acceptance for homograft size. Most homografts were used in Ross procedures (58%). The remaining 42% were implanted for other diagnoses (48.1% tetralogy of Fallot, 13.9% truncus arteriosus, 11.4% transposition of the great arteries, 8.9% pulmonary atresia, 5.1% double outlet right ventricle, and 12.6% others).
Of the homograft implantations, 56.9% were primary procedures. 24.5% followed previous palliative operations such as aortopulmonary shunts or pulmonary artery patch plasties. 13.8% were conduit exchanges and 4.8% replaced valveless conduits. Tables 13 give details about the patients demography and their operations, mean values and 95% confidence limits. In order to facilitate comparisons with other patient populations, all analyses were split according to three criteria: by age at time of operation (Table 1, Figure series A), operative procedure (Table 2, Figure series A), and size of homografts (Table 3, Figure series C). The 10-year cutoff for age stratification was chosen because most 10-year-old patients can receive a conduit of approximately adult size. In contrast, for patients under 10 years of age, outgrowth is considered a significant risk factor for development of stenosis. Results were obtained by reviewing the patients records and retrieving information from external physicians and hospitals involved in the postoperative treatment. The follow-up of all survivors included results of 773 echocardiographic and 74 angiographic studies. Six patients were lost to follow-up after a mean observation period of 2.5 years; among them, five moved to foreign countries. Follow-up is thus 96.8% complete. We included only transvalvular or transconduit instantaneous peak gradients; postanastomotic pulmonary artery (branch) stenoses were not included as terminal events in the evaluation since they were not related to the homograft. One patient had a plasty of the right pulmonary artery 1 year after implantation without explanting the homograft; in all other patients, reoperation meant homograft explantation for degeneration.
|
|
|
Since the policies for explantation and percutaneous intervention had changed within the observation interval, we considered the freedom from combined degeneration as the most relevant outcome parameter. In each case, an explantation or intervention was preceded by an echo or angiographic examination. Therefore, we chose only the endpoint degeneration to evaluate the role of procedural history and the role of homograft origin.
In the past decades, conduit exchange was riskier, and reoperations were postponed as long as clinically tolerable. Permanent right ventricular damage was more likely to be accepted then, than nowadays. The indication for angiography to confirm homograft deterioration was often seen only in advanced stages of clinical deterioration. Today, routine echos and a more aggressive approach towards procedural therapy reduce both the time until detection of homograft dysfunction as well as the interval between dysfunction and therapy. We illustrate the changing approach by displaying the interval between detection of a dysfunction and the procedural therapy, stratified by the decade of dysfunction detection.
Univariate risk factor illustrations as KaplanMeier plots are highly suggestive to consider the split criterion itself as a real clinically significant risk factor. However, the common association of several factors makes multivariable methods as the Cox regression analysis indispensable. The mentioned potential risk factors (other operations vs Ross procedure, previous homograft, homograft size group, patient age under 10 years, aortic vs pulmonary homograft) were included in multivariable analyses. Stenosis, insufficiency and degeneration were considered as endpoints.
| 3. Results |
|---|
|
|
|---|
|
Freedom from homograft explantation for patients older than 10 years at implantation was 81.8%, while 40% of the grafts in younger patients had to be explanted after the same time interval (Fig. 2A). Obviously, other factors than outgrowth contributed to the indication for explantation: beyond 10 years, the explantation rate continues to rise remarkably also in the older patient group. Patients with a Ross operation (Fig. 2B) reach 95% freedom from explantation at 10 years. The outcome thereafter is not encouraging, but determined by only few patients. Homografts smaller than 20 mm appear significantly less durable than the larger sizes (Fig. 2C).
|
|
|
|
All the above time related evaluations were univariate: they considered one variable only. Obviously, more than one factor permits to divide the entire group of homograft recipients into different strata that have significant differences in outcome. The Cox regression is meant to calculate the independent impact of each of several factors on the terminal event. The algorithm can be set to run forward, by successively including more and more variables and testing their significance; or backwards, by starting with the inclusion of all potential risk factors and eliminating step by step the least important one until only independent significant factors remain in the analysis. If the results of both directions, forward and backward, coincide, the result is considered more robust than if this is not the case.
We applied the Cox regression analysis including the following potential risk factors: age at time of implantation <10 years or not, no Ross procedure versus Ross procedure, homograft diameter <20 mm or larger, previously implanted homograft versus first homograft implantation, aortic versus pulmonary homograft origin. Examined endpoints were: freedom from gradient
50 mmHg, freedom from insufficiency of at least grade moderate, and freedom from degeneration defined as either of the latter.
All Cox regression models (backward and forward), for all of the three end points (stenosis, insufficiency, degeneration) resulted in indicating that a non-Ross procedure and implantation age <10 years were both independent risk factors. P- values ranged from <0.001 to 0.028, and odds ratios from 2.25 to 3.6. Homograft origin, size and previous homograft implantation were no significant risk factors in any of the regression models.
| 4. Discussion |
|---|
|
|
|---|
Our KaplanMeier curves on freedom from stenosis, insufficiency and their combination are conservative estimations: the events gradient of 50 mmHg and more or moderate or more insufficiency was considered to have taken place at the first time the event was documented by an examination. Delay in the detection of these parameters of degeneration either by late decision for angiography (in former times) or caused by patients who skipped examinations is common. Although the assumption that a valve dysfunction is present before its diagnosis is justified, it is neglected in this evaluation. If all patients would have been examined at regular intervals, the event free intervals would be shorter.
There is a (minor) concern for the inter-method-variability when gradients and insufficiency classifications from two different sources (such as the echo and the cath lab) are pooled. Nonetheless, the results of these examinations with all the systematic limitations are both commonly used to make a decision whether to reoperate on a patient or not. So we pooled the results of both methods for the present evaluation.
The kind and magnitude of long-term sequelae after homograft implantation at our institution is not exceptional, but can be considered representative and normal for homografts. With all the mentioned limitations, our results compare well to other large series: Dittrich et al. [8] found 44% freedom from more than moderate insufficiency at 4 years in his group of 23 homograft patients. Homann et al. [6] observed 70% freedom from homograft replacement at 10 years in his cohort of 215 patients. Lange et al. [9] described 40% and 60% freedom from explantation at 15 years for his recipients of homografts smaller and larger than 15 mm, respectively. LeBlanc et al. [10] stratified the freedom from explantation by patient age and found 89% free at 4 years for patients above 2 years of age at implantation versus 52% for the younger ones. Tatebe et al. [11] found 42% freedom from homograft failure after 5 years in his observation of 141 homografts in patients under 10 years. Feier et al. [12] reports a 30% incidence of mean transconduit gradient of 20 mmHg and more at 5 years; Böhm et al. [13] found 66% freedom from explantation or transvalvular gradient of 30 mmHg or more after the same time, Settepani et al. [14] communicates that 25.5% of his patients developed homograft stenoses (no time related estimate was given) and Williams et al. [15] reports 88% freedom from reoperation at 5 years. Stark et al. [4] describe 31% conduit survival after 15 years, while Niwaya et al. [16] indicated 76% freedom from dysfunction at 8 years in his population of 78% Ross procedure patients. A 10-year-freedom from explantation of under 20% has been described for a group of 52 homografts implanted for congenital malformations in patients with a median age of 1.5 years [7].
Authors agree broadly that homograft durability and availability are not ideal, especially in small patients [1720].
According to Meyns et al. [21] and Bielefeld et al. [22], a previously implanted homograft does not accelerate degeneration of the second homograft. Mohammadi et al. [23] found homografts to be a risk factor for a second reoperation. In our population, univariate analysis showed significantly earlier degeneration of a second homograft, but the results of the Cox regression model suggests to attribute earlier degeneration to the high rate of younger patients with non-Ross procedures.
The recent paper of Karamlou et al. [24] does illustrate durability deficits in younger patients using a parametric estimation that shows the implant-age related probability for the first conduit related intervention within 2 years.
The optimal time for replacement of a degenerated homograft remains controversial. The degree and long-term consequences of right ventricular myocardial damage are both uncertain and have to be balanced against the risk of repeated conduit exchanges. The operative risk also for redo operations has been reduced in the passed decades, so from the surgical point of view the decision for a homograft replacement is taken easier and earlier today than 15 or 20 years ago. But patients themselves, their referring cardiologists and their home physicians are important links in the chain leading to the decision to exchange a degenerated homograft. Fig. 5F suggests that the reaction time to make this decision has indeed decreased. In contrast to the previously presented curves, this KaplanMeier plot does not start at the day of operation, but at the first detection of homograft degeneration defined as mentioned , and the terminal event is not the homograft failure, but the therapy catheter based or surgical intervention. Thus, the terminal event is a positive event, so a short therpeutical gap results in a small area under the curve. Indeed, in the recent era (20002006) the curve descends more steeply than in the previous one (19901999). The delay times in the earliest decade are explained by the relatively high threshold to perform the invasive angiography in children; this diagnostic procedure was often performed shortly before the reoperation, just to have a final confirmation or a legal justification for the homograft exchange.
A retrospective single centre report has intrinsic limitationsno randomization with an alternative device, no standardized outcome parameters for comparisons. Although if clear end point criteria are given, the accuracy of examinations depended on the diagnostic equipment, the examiner's subjectivity in the description of a finding, and the nervousness of the patient during the echo examination. As mentioned, we overestimated to a certain extent the homograft durability, because conduit degeneration is determined only at the time of the first available echo or catheter examination, although degeneration might have existed months or years before.
Despite all the limitations, homograft results from this large patient group cannot be called ideal. Therefore, we have to look for other options such as bovine jugular veins [25] or valved stents. Comparisons with alternative devices remain difficult. A prospectively organized registry [24], or better, a randomized study comparing homografts and newer valved conduits will provide more evidence for finding the optimal available conduit for the pulmonary position. Currently, we think that bovine pulmonary valved conduits are an alternative to homografts.
| 5. Conclusion |
|---|
|
|
|---|
| Appendix A |
|---|
|
|
|---|
Dr G. Gerosa (Padova, Italy): How do you explain such a huge difference between the Ross and non-Ross patients in terms of long-term durability?
Dr Goerler: I think the major reason is that the age of non-Ross patients was significantly lower than that of the Ross patients. They were adult patients compared to patients with congenital heart disease. The mean age of the non-Ross patients was 14 years. That means that one reason for valve replacement was outgrowth, simply outgrowth. It had nothing to do with the quality of the valve.
Dr Gerosa: And the homografts were all cryopreserved homografts coming from the same bank or from different institutions?
Dr Goerler: I cant say anything about the specific origin of the homografts because they came from different homograft banks.
Dr Gerosa: Nevertheless, they were cryopreserved homografts or fresh homografts?
Dr Goerler: All homografts except one were cryopreserved.
Dr B. Koul (Lund, Sweden): I dont know if I missed something that you said. Are we talking about the pulmonary homografts or they were mixed, both pulmonary, aortic?
Dr Goerler: All patients received homographs in the pulmonary position. Eighteen of them came from the aortic position in the donors, the rest were pulmonary homographs.
Dr M. Nathanson (San Jose, CA): Is it possible that in the congenital group the initial pulmonary vascular resistance was higher than the adult group or the non-congenital group?
Dr Goerler: That is possible. But, unfortunately, I dont have this data.
| Footnotes |
|---|
\#9734;\#9734; This study was supported by the Medtronic Bakken Research Center, Maastricht, The Netherlands.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. Horer, M. Vogt, U. Stierle, J. Cleuziou, Z. Prodan, C. Schreiber, and R. Lange A comparative study of mechanical and homograft prostheses in the pulmonary position. Ann. Thorac. Surg., November 1, 2009; 88(5): 1534 - 1539. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Horer, T. Hanke, U. Stierle, J. J.M. Takkenberg, A. J.J.C. Bogers, W. Hemmer, J. G. Rein, R. Hetzer, M. Hubler, D. R. Robinson, et al. Homograft performance in children after the Ross operation. Ann. Thorac. Surg., August 1, 2009; 88(2): 609 - 615. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Papadopoulos, A. Esmaeili, A. Zierer, F. Bakhtiary, F. Ozaslan, and A. Moritz Secondary Repair of Incompetent Pulmonary Valves. Ann. Thorac. Surg., June 1, 2009; 87(6): 1879 - 1884. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. E. David Ross Procedure at the Crossroads Circulation, January 20, 2009; 119(2): 207 - 209. [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 |