| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Cardio Thoracic Surgery, Erasmus Medical Center Rotterdam, the Netherlands
c Department of Cardiology, Erasmus Medical Center Rotterdam, the Netherlands
b Harefield Heart Science Center, National Heart and Lung Institute, Harefield, United Kingdom
Accepted for publication August 11, 2009.
* Address correspondence to Dr Klieverik, Department of Cardio Thoracic Surgery, Bd 571, Erasmus Medical Center Rotterdam, PO Box 2040, Rotterdam, 3000 CA, the Netherlands (Email: l.klieverik{at}erasmusmc.nl).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
Methods: Between 1980 and 2002, 138 patients (81% males) underwent aortic valve replacement for NVE in 2 centers (106 allografts; 32 MPs). Perioperative characteristics and early and late morbidity and mortality were analyzed.
Results: Mean age was 47 years (range, 14 to 76 years), and 34% required emergency surgery. Abscess rate was 38% for allografts vs 18% for MPs. Concomitant mitral valve replacement was required in 38% MP patients and in 5% allograft patients. Hospital mortality was 8% (n = 11; p = 0.25): 10 allograft patients (9%) and 1 MP patient (3%). During a mean 8-year follow-up (range, 0 to 25 years) 33 patients died: 22 allograft (24%) and 11 MP patients (21%; p = 0.14). Survival at 15 years was 59% ± 6% for allograft patients and 66% ± 9% for MP patients (p = 0.68). Late recurrent endocarditis developed in 6 allograft patients and 1 MP patient (p = 0.29). Overall 15-year freedom from reoperation was 76% ± 9% for allografts and 93% ± 6% for MPs (p = 0.02).
Conclusions: Mechanical prostheses have comparable rates of midterm survival and freedom from recurrent infection. However, this is in combination with extensive excision of destructive tissue in a specific patient subset. Allograft reoperation rates increase with time. The importance of the mechanical prosthesis in NVE might be established in the coming years.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
According to the American College of Cardiology (ACC)/American Heart Association (AHA) guidelines for management of patients with heart valve disease, surgical treatment of active native valve endocarditis (NVE) should preferably consist of valve repair because of the risk of infection of prosthetic materials [1]. Thus far, there are no firm specific recommendations for use of particular valve prosthesis for surgical treatment of active NVE besides the general criteria for aortic valve selection.
Currently, two commonly used replacements for the infected native aortic valve are allografts and mechanical prostheses [2]. Biologic prostheses could also be used in NVE but have limited durability and are not well established in active NVE. Allografts are a good option in patients with active NVE. Particularly when there is extensive destruction of the surrounding tissue, allografts can cover defects with preservation of the natural anatomy of the aortic valve and adjacent structures [3]. Furthermore, allograft patients do not require life-long anticoagulation, and reinfection rate is low, with a constant phase compared with mechanical prostheses and bioprostheses [2]. Allografts have limited durability, however, which makes reoperation inevitable in the long-term, are not always readily available, and implantation requires specific surgical training.
Mechanical prostheses, on the other hand, are designed to last a lifetime, are readily available, and are easier to implant. Furthermore, the risk of endocarditis reinfection is reported to be very low [4]. On the downside, these valves are thrombogenic and require life-long anticoagulation with a low, but considerable risk of bleeding and thromboembolic events [5–7].
It remains a matter of debate whether there is a preferred valve substitute for active NVE treatment. In this regard, outcome of patients who underwent aortic valve replacement in 2 centers with an allograft or a mechanical prosthesis for active NVE were analyzed.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Patients
Between March 1980 and December 2002, 138 consecutive patients underwent aortic valve replacement for active NVE, comprising allografts in 106 and mechanical protheses in 32. Of these, 86 patients were operated on by different surgeons at Erasmus University Medical Center Rotterdam (Rotterdam, the Netherlands), and received 58 allografts and 28 mechanical prostheses. The remaining 52 were operated on by a single surgeon (MHY) at Harefield Hospital (Harefield, United Kingdom) and received 48 allografts and 4 mechanical prostheses.
For Rotterdam allograft patients, information on patient characteristics, perioperative details, and follow-up was obtained from an ongoing prospective cohort study [8]. Data for mechanical prosthesis patients were collected retrospectively from hospital records, correspondence with treating physicians, and through the civil registry. Data were collected retrospectively in a similar fashion for the Harefield patients.
The anticoagulation regimen at Harefield Hospital included no anticoagulation for allografts. For mechanical prostheses, 24 hours of subcutaneous anticoagulation was used postoperatively, with a switch after 24 hours to oral anticoagulation aiming for an international normalization ratio (INR) of 2.0 to 2.5.
The anticoagulation regimen at Erasmus Medical Center Rotterdam also includes no anticoagulation for allografts. For mechanical prostheses, postoperative intravenous anticoagulation aimed for an activated partial thromboplastin time of 55 to 70 and simultaneously started with oral anticoagulation until an INR of 2.0 to 3.0 was obtained.
Diagnosis of endocarditis was determined from clinical criteria, including signs of fever, new or altering cardiac murmurs, positive blood cultures, and echocardiographic findings [9]. Endocarditis was considered active if patients underwent operations before completing a 6-week course of antibiotic treatment.
Operative Technique
All operations were performed on cardiopulmonary bypass with moderate hypothermia. Circulatory arrest with deep hypothermia was needed for 2 Rotterdam patients because of ascending aorta and arch interventions.
Follow-Up
Valve-related events were defined according the guidelines for reporting morbidity and mortality after cardiac valvular operations [10]. The database was frozen on January 1, 2006. Follow-up was 92.7% complete. Follow-up was incomplete for 10 patients because of emigration.
Statistical Analysis
Data were analyzed with SPSS 12.0.1 software (SPSS Inc, Chicago, IL). Continuous data are reported as the mean ± 1 standard deviation, and comparisons between groups were made using the unpaired t test or the Mann-Whitney U test, where appropriate. Discrete data are presented as proportions and were compared using the 
2 test or the Fisher exact test.
A propensity score for receiving either an allograft or a mechanical prosthesis was calculated for each patient. First, by means of univariate logistic regression, variables were identified that were potentially associated with valve substitute selection. Variables with p < 0.05 in the univariable analysis and variables that from a clinical point of view may be considered to affect prosthetic valve selection [1] were included in the multivariable model. Variables included in the model were age at operation, women of childbearing age (women aged < 45 years at operation), surgical center, time period; left ventricular function (LVF) according to ejection fraction (EF) as good (> 0.50), impaired (0.40 to 0.50), and moderate/bad (< 40%); presence of abscesses, emergency operation (< 24 hours after diagnosis), and concomitant mitral valve replacement. The propensity score was used as a covariable in logistic and Cox regression models that studied mortality after operation.
Univariable and multivariable logistic regression was used to determine factors associated with hospital mortality. The factors analyzed were age at operation, sex, women of childbearing age, New York Heart Association (NYHA) functional class (defined as I, II, III, IV, and cardiogenic shock, V), preoperative creatinine level (µmol/L), preoperative ventilation support, preoperative atrial fibrillation, left ventricular function, emergency operation, presence of abscesses, type of infection-causing microorganism, concomitant procedures, valve substitute used, cardiopulmonary bypass time (CPB) in minutes, and propensity score.
Cumulative survival, freedom from reoperation or reintervention or freedom from valve-related events, including reoperations, were analyzed using the Kaplan-Meier method. The Tarone-Ware test was used to compare Kaplan-Meier curves correcting for significant difference in follow-up time between allograft and mechanical prosthesis patients. Age-matched survival in the general population was calculated using the Dutch population life tables (http://statline.cbs.nl/). We used the life-tables method to calculate the hazard rate (HR) of structural failure for allografts over time by subdividing the follow-up period after operation into 5-year intervals.
The Cox proportional hazards model was used for univariable and multivariable (stepwise backward method) analysis of time-related events. The following factors were considered: age at operation, sex, women of childbearing age, surgical center, type of infection-causing microorganism, presence of atrial fibrillation preoperatively, left ventricular function, presence of abscesses, urgency of operation, valve substitute used, concomitant procedures, CPB time, and propensity score. A value of p 
0.05 was considered statistically significant. All testing was two-sided.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
|
|
|
|
Age at operation, sex, childbearing age for women, NYHA class, history of stroke, preoperative renal failure, the microorganism, preexisting atrial fibrillation, and operative urgency had no effect on valve selection.
Hospital Mortality
There were 11 hospital deaths (8%), and 7 patients (63%) required urgent operations, comprising 10 (9%) who received allografts and 1 (3%) who received a mechanical prosthesis (p = 0.25). Three allograft patients died in the operating theater: 1of persistent bleeding and 2 of heart failure.
Postoperatively, 3 allograft patients died of intracranial hemorrhage at 8, 9, and 10 days. One allograft patient died of a stroke at 48 days, and 3 allograft patients died of multiorgan failure at 5, 8, and 11 days. The mechanical prosthesis patient died of heart failure 1 day after the operation.
Potential risk factors for hospital death were female gender (OR, 4.2; 95% CI, 1.2 to 5.1; p = 0.03), endocarditis caused by Staphylococcus aureus (OR, 5.6; 95% CI, 1.2 to 25.2; p = 0.03), preoperative increased creatinine level (OR, 1.01; 95% CI, 1.01 to 1.02; p < 0.001), NYHA class IV (OR, 6.1; 95% CI, 1.3 to 29.3; p = 0.02), emergency operation (OR, 3.8l 95% 1.05 to 13.7; p = 0.04), and longer perfusion time (OR, 1.01; 95% CI, 1.01 to 1.02; p = 0.002). After multivariable analysis, preoperative increased creatinine (OR, 1.01; 95% CI, 1.003 to 1.02; p = 0.006) and longer perfusion time (OR, 1.01; 95% CI, 1.003 to 1.02; p = 0.01) were risk factors for hospital mortality. The propensity score had no effect on the hospital mortality rate.
Follow-Up and Late Mortality Rate
Mean follow-up was 8.8 ± 6.5 years, (maximum follow-up, 25 years; 1223 patient-years). Mean follow-up for allograft patients was 7.7 ± 5.6 years (maximum follow-up, 25 years; 820 patient-years). Mean follow-up for mechanical prosthesis patients was 12.5 ± 7.8 years (maximum follow-up, 24 years; 403 patient-years). Mean follow-up time was significantly different between the 2 groups (p < 0.001).
During follow-up, 34 patients died (p = 0.14) with a linearized occurrence rate (LOR) of 2.8%/patient-year, comprising 23 allograft patients (LOR, 2.8%/patient-year) and 11 mechanical prosthesis patients (LOR, 2.7%/patient-year). Causes of death in allograft patients were recurrent endocarditis in 3, intracranial hemorrhage in 1, sudden death in 1, and non-valve-related death in 18. Causes of death in mechanical prosthesis patients were sudden death in 2 and non-valve-related death in 9.
Figure 1 shows survival for the valve substitute groups compared with 47-year-old men in the general Dutch population. Overall 1-year survival was 89.6% ± 3.0% and 15-year survival was 61.3% ± 5.2%.
|
For mechanical prosthesis patients, 1-year survival was 93.8% ± 4.3% and 15-year survival was 65.6% ± 9%. Survival at 1 year was 92.9% ± 4.9% for Rotterdam patients and 75.0% ± 21.7% for Harefield patients (p = 0.55). Survival at 15 years was 64.2% ± 9.7% for Rotterdam patients and 75.0% ± 21.7% for Harefield patients (p = 0.74).
Preexisting atrial fibrillation (HR, 4.0; 95% CI, 1.4 to 11.4; p = 0.01) and older patient age (HR 1.05; 95% CI, 1.02 to 10.7; p < 0.001) were potential risk factors for increased late death. No other factors were associated with increased late death. After multivariable analysis, preexisting atrial fibrillation (HR 4.2; 95% CI, 1.4 to 12.6; p = 0.01) and older patient age (HR 1.04; 95% CI, 1.02 to 1.07; p = 0.001) were significant risk factors for increased late death, even after including the propensity score in the model.
No survival differences were observed for allograft patients with and without root abscesses (p = 0.34). Also, when comparing these patients between the 2 centers, no survival differences were observed (p = 0.30).
Recurrent Endocarditis
Endocarditis was recurrent postoperatively in 6 allograft patients (LOR, 0.73%/patient-year) and 1 mechanical prosthesis patient (LOR, 0.25%/patient year); all were late episodes (range, 1.5 to 13.5 years). Of these patients, 4 allograft patients underwent reoperation and survived. One mechanical prosthesis patient and 2 allograft patients received only antibiotic treatment and died before reoperation could be performed.
The 1-year freedom from recurrent endocarditis was 100% (Fig 2). The overall 15-year freedom from recurrent endocarditis was 89.1% ± 4.3%. For allograft patients, it was 85.9% ± 6.5%, and for mechanical prosthesis patients, 94.7% ± 5.1% (p = 0.29). No variables were identified to be associated with recurrent endocarditis.
|
The HR of structural failure for allografts increased with time since operation: from 0.23%/patient-year in the first 5 years after operation to 0.82%/patient-year at 5 to 10 years after operation to 1.0%/patient-year at 10 to 15 years after operation to, finally, 6.7%/patient year at 15 to 20 years after operation.
The overall 15-year freedom from reoperation was 81.4% ± 6.2% (Fig 3); and by group, 75.8% ± 9.0% for allograft patients and 92.9% ± 6.9% for mechanical prosthesis patients (p = 0.02). The only potential risk factor for increased reoperation rate was allograft use (HR, 10.7; 95% CI, 1.3 to 87.3; p = 0.03).
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Valve Selection
Selecting a valve substitute in active endocarditis seems, in our study, dependent on patient-related factors. A trend towards allograft selection as the preferable valve substitute in patients with root abscesses was observed. Others have confirmed that extensive destruction of the left ventricular outflow tract and adjacent structures is one of the indications for an allograft in active NVE [3, 11]. At Erasmus University Medical Center Rotterdam, the decision for using allografts or mechanical prostheses in active endocarditis now depends strongly on the resulting anatomy after extensive débridement and excision of infected material or abscesses, fistula, or annulus involvement. In the 1990s, allografts were almost exclusively used in any type of endocarditi, regardless of the lesion extent. This is also reflected by the covariable time period in our valve selection propensity score analyses.
Before 1990, few allografts were used, and mechanical prostheses were the valve substitutes of choice, simply because of limited experience with allografts and their scarce availability. Mechanical prostheses were also more often used in our study patients, in whom active NVE destroyed the mitral valve, necessitating mitral valve replacement, thus warranting lifelong anticoagulation. Mechanical prostheses were implanted after extensive débridement and excision of infected material.
Although patient age is important for valve selection according the ACC/AHA guidelines [1], it did not have an effect in our study patients. Moreover, our patients were relatively young (mean age, 47 years), and according to the guidelines, mechanical prostheses are the recommended valve substitute of choice in these patients; biologic prostheses are merely recommended in patients aged older than 65 years [1].
Early Death
Surgical intervention for active NVE remains challenging, with high operative mortality rates and the necessity of early operations being reported [2, 12–16]. The overall hospital mortality rate in this study was 8%, which is in line with the 9% to 31% mortality rates described in published articles [4, 11, 14, 15].
No significant difference in early death between allograft patients and mechanical prosthesis patients was observed in the present study. Other factors determined in this study that potentially influenced early death were preoperative increased creatinine, NYHA class IV, emergency operation, longer perfusion time, and endocarditis caused by S aureus. These variables were also reported by other authors to influence early mortality rates in active NVE [2, 4, 13, 15].
Late Mortality
Survival at 15 years was 58.7% for allograft patients and 65.6% for mechanical prosthesis patients, which is comparable with other reports [14, 17]. Furthermore, only 7 patients died of valve-related causes (5 allograft patients and 2 mechanical prosthesis patients). Most of our study patients died of non-valve-related causes, and only 3 allograft patients died of recurrent endocarditis.
Atrial fibrillation and older patient age were associated with increased late mortality risk, reflecting once again that suboptimal cardiac function and aging have an important effect in predicting late survival rates. Although there was a trend towards allograft selection in patients with root abscesses, and patients with abscesses may be more severely ill and could have adverse outcome, no differences in survival were observed between patients with presence of abscesses and without. Furthermore, no differences in outcome were observed in these allograft patients between the 2 centers. Other underlying comorbidities of patients may have a more profound effect in the outcome.
Recurrent Endocarditis
Freedom from recurrent endocarditis after active NVE was approximately 90%, was comparable for both valve substitutes, and was comparable with other series [15, 18]. When our linearized rates of recurrent endocarditis for mechanical prostheses and allografts are compared with other observational studies on these valve substitutes, our results are in range with these reports [5, 13]. The recurrent infection rate was lower for mechanical prostheses than for allografts in our study, and although this difference was not significant, other authors have observed this as well [11, 19].
Furthermore, Haydock and colleagues [2] found a constant phase of recurrent endocarditis for allografts in contrast to an early peaking phase for mechanical prostheses. In contrast in our study, there was no early peaking phase of recurrent endocarditis in the first year postoperatively for the mechanical prosthesis (0%/patient-year) and only 1 mechanical prosthesis patient had an episode of recurrent endocarditis. The HR for recurrent endocarditis for allografts in the first 5 years after operation was 0.23%/patient-year and increased over time to 6.7%/patient-year. Although this seems to be significant, it may be biased due to the small number of events that were observed and may be due to chance.
Reoperations
Allograft reoperations are perceived to be complex procedures potentially carrying a high risk of morbidity and death. A low reoperative mortality rate for allografts was found in this series, a good result compared with other reports [20, 21]. Two main causes for reoperation in our study were recurrent endocarditis and structural failure. Reoperation rates for recurrent endocarditis were 0.61%/patient-year for allografts and 0% for mechanical prostheses. A study by McGriffin and colleagues [22] found that patients who had active NVE are more prone to have another episode of endocarditis. In the study of Tyers and colleagues [23], endocarditis was an important cause of reoperation in mechanical prosthesis patients, and O'Brien and colleagues [24] showed a low reoperation rate for recurrent endocarditis after allograft implantation. Although these two studies describe an incidence of reoperation for recurrent endocarditis favoring the allograft, reoperation rates for any cause are much higher in allograft patients than in those with a mechanical prosthesis [13, 20, 24–27] .The present study confirms these results by showing that the reoperation risk for allografts increases with time since operation and is mainly determined by the increasing hazard of structural failure over time.
Yankah and colleagues [3] describe also a high reoperation rate for structural failure in allograft patients, and especially in patients with undersized allografts. Contradicting those results is a recent report from our group in which a larger allograft diameter was an independent predictor for structural valve deterioration [25].
Younger patient age is another risk factor for reoperation [2, 13], which was not a significant factor in this report. This might be due to the small number of structural valve deterioration reoperations observed and limited mean follow-up duration of the study. Careful follow-up of allograft patients over time will prevent emergency reoperations in decompensated patients with degenerated allografts. Elective reoperations can be performed with good results and low mortality rates [25]. So, particularly in active NVE patients with relatively long life expectancy, reoperation is the most important limitation of allograft use and should be considered when selecting a valve substitute.
Other Valve-Related Complications
Bleeding complications during follow-up were rare in our study; only 1 allograft patient on anticoagulation medication had an intracranial hemorrhage. No events of valve thrombosis or thromboembolism were observed in our mechanical and allograft recipients. The low occurrence rates of valve-related events are in agreement with other reports [14, 17].
Limitations
The partially retrospective nature of study may have led to an underestimation of valve-related events during follow-up, in particular for mechanical prosthesis patients. Moreover, follow-up is slightly over 92% complete, leaving possible valve-related events during follow-up of approximately 8% of patients unresolved, which might have influenced our results.
This study is a combined series involving 2 centers with different policies for allograft and mechanical prosthesis use, and these results may not apply to other centers.
A further question remains whether all patients were eligible for an allograft or a mechanical prosthesis and vice versa. Patient selection bias can occur due to different treatment policies and because implanting allografts remains difficult and requires specific surgical skills. On the other hand, not all allograft patients could have been treated with a mechanical prosthesis, especially those with extensive destruction of the aortic root. A case-control study or a randomized trial could give a better insight about which valve is best to implant in active aortic valve endocarditis, although the last mentioned is hardly achievable. Finally, timing of the operation after antibiotic treatment was not thoroughly analyzed in this study, and this may have an influence on the outcomes we observed.
Conclusions
The results of this 2-center study indicate that besides the allograft, the mechanical prosthesis also provides a good solution for patients with no anticoagulation contraindication in active native aortic valve endocarditis in terms of survival and recurrent endocarditis. However, when it comes to inserting a mechanical prosthesis in active endocarditis, this should not only be combined with extensive excision of infected tissue but also performed in patients without aortic root abscesses.
More reoperations for structural valve deterioration can be expected for the allograft in the second decade after operation given the increasing reoperation hazard with time. Whether the mechanical prosthesis will play a more significant role in replacing the destructed aortic valve in active native endocarditis might be shown in the years to come.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. G. Svensson, D. H. Adams, R. O. Bonow, N. T. Kouchoukos, D. C. Miller, P. T. O'Gara, D. M. Shahian, H. V. Schaff, C. W. Akins, J. E. Bavaria, et al. Aortic Valve and Ascending Aorta Guidelines for Management and Quality Measures Ann. Thorac. Surg., June 1, 2013; 95(6_Supplement): S1 - S66. [Full Text] [PDF]
|
|
|
|
 |
|
 V. Le Guillou, F. Bouchart, A. Gay, C. Nafeh-Bizet, C. Hubscher, A. Tabley, J. P. Bessou, and F. Doguet The Ross procedure in endocarditis: a report of 28 cases Eur J Cardiothorac Surg, April 26, 2013; (2013) ezt206v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. G. Svensson, D. H. Adams, R. O. Bonow, N. T. Kouchoukos, D. C. Miller, P. T. O'Gara, D. M. Shahian, H. V. Schaff, C. W. Akins, J. Bavaria, et al. Aortic Valve and Ascending Aorta Guidelines for Management and Quality Measures: Executive Summary Ann. Thorac. Surg., April 1, 2013; 95(4): 1491 - 1505. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. H. McCarthy, J. E. Bavaria, A. Pochettino, Z. Fox, P. Moeller, W. Y. Szeto, and N. D. Desai Comparing Aortic Root Replacements: Porcine Bioroots Versus Pericardial Versus Mechanical Composite Roots: Hemodynamic and Ventricular Remodeling at Greater Than One-Year Follow-Up Ann. Thorac. Surg., December 1, 2012; 94(6): 1975 - 1982. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. E. David Aortic valve repair for active infective endocarditis Eur J Cardiothorac Surg, July 1, 2012; 42(1): 127 - 128. [Full Text] [PDF]
|
|
|
|
|
|
K. Mayer, D. Aicher, S. Feldner, T. Kunihara, and H.-J. Schafers Repair versus replacement of the aortic valve in active infective endocarditis Eur J Cardiothorac Surg, July 1, 2012; 42(1): 122 - 127. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. S. Jassar, J. E. Bavaria, W. Y. Szeto, P. J. Moeller, J. Maniaci, R. K. Milewski, J. H. Gorman III, N. D. Desai, R. C. Gorman, and A. Pochettino Graft Selection for Aortic Root Replacement in Complex Active Endocarditis: Does It Matter? Ann. Thorac. Surg., February 1, 2012; 93(2): 480 - 487. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. D. Desai, F. McCarthy, W. Moser, W. Y. Szeto, A. Zeeshan, D. Brown, Y. J. Woo, A. Pochettino, P. Moeller, and J. E. Bavaria Durability of Porcine Bioroots in Younger Patients With Aortic Root Pathology: A Propensity-Matched Comparison With Composite Mechanical Roots Ann. Thorac. Surg., December 1, 2011; 92(6): 2054 - 2061. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Mokhles, I. Ciampichetti, S. J. Head, J. J. M. Takkenberg, and A. J. J. C. Bogers Survival of Surgically Treated Infective Endocarditis: A Comparison With the General Dutch Population Ann. Thorac. Surg., May 1, 2011; 91(5): 1407 - 1412. [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 |
