| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;80:1151-1158
© 2005 The Society of Thoracic Surgeons
a Cardiothoracic Center, Bristol Royal Infirmary, Bristol
b Cardiothoracic Center, John Radcliffe Hospital, Oxford
c Heart Science Center, Harefield Hospital, Middlesex, United Kingdom
* Address reprint requests to Dr Mahesh, Transplant Immunology, Heart Science Center, Harefield Hospital, MiddlesexUB96JH, UK (Email: b.mahesh{at}imperial.ac.uk).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsConclusionsReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsConclusionsReferences |
|---|
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsConclusionsReferences |
|---|
Microbiology
Prosthetic valve endocarditis is defined as early if occurring within 12 months of valve replacement and late if greater than 12 months have elapsed [1, 4, 5, 9–11]. Most cases are caused by Staphylococcus epidermidis and aureus, followed by Streptococcus viridans [8, 12, 13]; early PVE is often due to methicillin-resistant Staphylococcus epidermidis, gram-negative bacilli, fungi, and other HACEK-group organisms (Haemophilus, Actinobacillus, Cardiobacterium, Eikinella, Kingella), suggesting nosocomial infection [1, 14]. Common sources of bacteremia include wound infections, intravascular catheter infections, urinary-tract infections, and pneumonia [2, 4, 13, 15]. Infrequent, potentially fatal pathogens include fungi and mycoplasma; fungal PVE mandates surgery and long-term antifungal treatment and is associated with worse prognosis [10, 16], while mycoplasma can be treated very effectively by doxycycline alone [17]. Late PVE occurs due to bacteremia from skin, orodental and abdominal infections, and invasive medical-dental procedures and therefore Streptococcus viridans and Staphylococcus epidermidis are common pathogens [2, 13].
Pathogenesis of PVE: Differences Between Mechanical and Bioprosthetic PVE
Important factors in the pathogenesis include type of prosthesis, previous native valve endocarditis, male gender, and long cardiopulmonary bypass (CPB) time [13]. Early PVE rarely remains restricted to leaflets alone; frequently it involves the junction of the sewing ring and annulus, leading to valve dehiscence and paravalvular abscesses, which predispose to paraprosthetic leak and fistulization into other cardiac chambers or the pericardium [2, 4, 18], invariably necessitating replacement irrespective of the type of prosthesis [19], and associated with worse prognosis than late PVE [13].
Mechanical prostheses
Prostheses from metal and pyrolyte do not allow adherence of microorganisms to leaflets as long as they are free of thrombotic material [14]. Herein lies the importance of biofilms produced by bacteria, which enable them to adhere to surfaces of valves in the presence of high blood flow, protecting them from antibiotics [13, 14]. Infections in mechanical valves generally involve the sewing ring or adherent thrombi, leading to paraprosthetic leaks, ring abscesses, and invasive infection, necessitating operative intervention [8, 13, 14, 16]. Incidence of myocardial and paravalvular abscesses in mechanical PVE is estimated to be 38% and 63%, respectively [8]. Staphylococci are the most common organisms causing paravalvular abscesses; these are extremely virulent, often requiring surgical treatment and associated with higher mortality rates [3].
Bioprostheses
Bioprostheses are less susceptible to early infection, which is often restricted to the leaflets [16], making cure with antibiotics more likely, but increasing the chances of late failure due to degeneration of the cusps. This could also predispose to late infection due to implantation of organisms at sites of leaflet degeneration [2, 12, 19, 20]. The risk of sewing-ring infection is less, but once involved the pathogenesis is the same as for mechanical prostheses [14], and may form annular abscesses [16]. Late PVE is more likely than early PVE to be restricted to the valve leaflets alone [4]. Even in the absence of previous PVE, areas of degenerative changes in biological valves predispose to implantation of organisms.
Clinical Presentation and Diagnosis
Common clinical presentations include persistent fever after valve replacement, profound anorexia, new (changing) murmur, heart block, congestive heart failure (CHF), and embolic events. Myocardial infarction (MI) due to emboli and sudden death from disruption of the valve are more dramatic events [2, 4, 7, 16].
Diagnosis rests on multiple blood cultures drawn before commencement of antibiotics [2, 4, 7, 16]. Blood cultures are frequently negative if taken after commencement of antibiotics. Persistent bacteremia with no other indication of endocarditis is a strong predictor of the future development of endocarditis [15].
Echocardiography [21, 22] should be performed in all cases of suspected PVE, to characterize valvular hemodynamics and detect vegetations, abscesses, shunts, etc [21]. This may be used to assess the sewing ring stability and leaflet motion of bioprosthetic valves, but in mechanical PVE, intense reverberations from the valve limit its ability to detect vegetations, especially from the mitral valve [12]. Transesophageal echocardiography (TEE) has made a major impact on early recognition of PVE, with a twofold to threefold higher sensitivity compared to transthoracic echocardiography (TTE) [3]. Transesophageal echocardiography is particularly sensitive in detecting paravalvular abscesses and valvular dysfunction in the form of stenosis, regurgitation, and paravalvular leak, and vegetations as small as 1 to 2 mm in size, particularly in relation to mitral prostheses [12, 14]. Sensitivity and specificity of TEE approach 100% and 83%, respectively, compared to morphologic findings [22]. However, TEE may not be sufficient to assess the anterior aspect of an aortic prosthesis, especially in the presence of a mitral prosthesis, and thus both TTE and TEE may be needed for complete assessment [12].
Negative echocardiography does not exclude PVE. Leukocyte scans have been used to detect abscesses, with limited success [14]. Magnetic resonance imaging may be performed safely in PVE, and is useful for identifying infective periprosthetic tissue signals, even before the development of blood flow abnormalities such as periprosthetic leakage [4], and distinguishing between valvular and paravalvular regurgitation [12]. Coronary angiography to look for coexistent coronary artery disease may be needed [23], but may be hazardous in the presence of vegetations [20].
Medical Treatment
Broad-spectrum antibiotics started on suspicion of PVE, tailored subsequently to suit the sensitivity of the organisms, constitute the first line of treatment, and may reduce systemic embolization by shrinking existing vegetations and preventing formation of new ones [2]. In cases of culture negative, presumed bacterial endocarditis, broad-spectrum antibiotics need to be continued for 6 weeks [13, 21]. Culture negative PVE may be due to fungi (candida/aspergillus) [21], Mycoplasma, and Ureaplasma. The latter are only occasionally detected by serology, but more often by polymerase chain reaction on the explanted valve. Specific treatment with doxycycline for 4 weeks results in complete cure [17].
Patients must be monitored closely for progressive CHF, worsening conduction abnormalities, and complications such as annular abscesses, fistulas, or paravalvular leaks [1, 3, 16, 24, 25]. In these situations, medical therapy alone is associated with high mortality, thus making surgery mandatory [15, 24, 26]. Surgery provides material for isolation of the organism, and some groups have used the Histoplasma capsulatum Accuprobe in sonication of explanted valves for rapidly arriving at a diagnosis [27].
Indications for Operation
A cerebral embolic event is associated with a higher operative mortality due to systemic heparinization and hypotension associated with CPB, which may cause worsening of cerebral edema or conversion of an ischemic into a hemorrhagic stroke [28] (Table 1). Maruyama and colleagues [28] have demonstrated a significantly worse outcome with cardiac surgery in patients with cerebral emboli due to vegetations, mainly due to infective vascular injury at the site of the embolus; this leads to conversion of an ischemic into a hemorrhagic stroke due to systemic heparinization for CPB. Surgery in the setting of stroke becomes imperative in cases of recurrent thromboembolism, persistent vegetations, and hemodynamic instability [28, 29].
|
Myocardial protection maybe achieved by a combination of intermittent antegrade and retrograde blood cardioplegia [2, 4, 10, 30], keeping the myocardial temperature at 10 to 15°C, accompanied by systemic cooling at 25–28°C [1, 2, 4]. Others recommend continuous antegrade cardioplegia administered through self-inflated coronary cannulas at 150 to 250 mL/min [30]. However, a range of techniques can be used successfully.
Radical debridement with a margin of healthy tissue to eradicate intracardiac foci of infection remains the primary aim of surgery for PVE [1, 3, 4, 6, 7, 9–11, 18, 19, 23, 26, 30–38], enabling secure fixation of the new prosthesis, avoiding recurrent or residual infection, periprosthetic leak/dehiscence, or subannular aneurysm formation [34]. The superior vena cava may be divided to improve access to the fibrous trigone and regions where the aortic and mitral annuli are close to each other [4].
After extensive debridement, major reconstruction of the left-ventricular outflow tract (LVOT), aortic and/or mitral annuli maybe required. This may be accomplished by Teflon sheets and/or biological tissue such as autologous pericardium, glutaraldehyde-fixed bovine pericardium [11], pulmonary autograft, or aortic homograft [20]. Tension free repair is essential. Patches help exclude attenuated areas from high pressures and provide fixation points for the prosthesis [34]. Biological tissue has been the material of choice for reconstruction after extensive debridement [1, 4, 8, 10, 26, 31–37]. Abscesses need to be curetted until all infected material has been removed; smaller ones may be closed with pericardium [18], or filled with gelatin-resorcin-formol glue [23] or gentamicin-saturated fibrin glue [38]. Larger abscess cavities may be exteriorized into pericardium and thus excluded from the LVOT and cardiac chambers [1, 31]. Fistulas extending into the atria or pericardium through the perimembranous septum or the aortic root may be repaired using pericardial patches [3, 30, 38]. Orthotopic valve replacement is preferred; if translocation is necessary, it should be performed leaving attenuated tissues on the low pressure side of the valve to prevent aneurysmal dilatation [34]. If there is aortoventricular discontinuity after extensive debridement, the aortic root needs to be replaced [34]. Abscesses involving the noncoronary sinus may need to be debrided together with the roof of the left atrium (LA) and/or a portion of the anterior mitral leaflet; these may be repaired with pericardial patches. Fibrous continuity between aortic and mitral valves may then be reestablished by securing a triangular piece of bovine pericardium and both valve prostheses may be secured to this and the rest of the annulus [3]. Results of surgery for PVE by different groups are summarized in Table 2.
|
Lund and colleagues [39] and Yacoub and colleagues [40] found that even in PVE, unprocessed "homovital" homografts provided better survival and freedom from reoperation compared to antibiotic/cryopreserved homografts when implanted by a full-root technique. They attributed this to absence of preservation-related biophysical changes and gradients across the valve, and normal aortic root and coronary flow dynamics, thus preserving LV function. They observed long-term degeneration in all types of homografts, which they felt did not adversely affect outcome because of the slow rate. Thus, they suggested use of homografts as opposed to pulmonary valve autografts in younger patients with PVE [40]. They found that replacement of the homograft, if required subsequently due to degeneration and calcification, despite being a technically very demanding operation, could be performed with relatively little morbidity and mortality, because of improved surgical and myocardial protection techniques [41], even in cases of homograft endocarditis [31]. Sadowski and colleagues had similar results in homograft endocarditis [42].
Homografts have the lowest incidence of early postoperative infection/reinfection, followed only by the low constant phase of infection, as opposed to the high early peak of infection seen with standard prostheses [4, 8, 20, 35, 39, 43]. This may be due to a combination of factors including absence of a prosthetic sewing ring, resistance of the homograft to infection [1, 2, 4, 8, 20, 37, 39], and antibiotics used to preserve the homograft [20].
Pulmonary autograft
After extensive debridement of the aortic root, and repair of VSDs and fistulas, the aortic annulus is assessed and the Ross procedure may be carried out if there is no size discrepancy between the aortic and pulmonary roots and if the disease process does not involve the pulmonary root/valve. The pulmonary autograft in the aortic position remains "alive," has growth potential especially in children, has excellent hemodynamics, has a low risk of thrombosis and embolic complications without anticoagulation, and has a low reinfection rate [44]. Malalignment and regurgitation are not a problem if the autograft is implanted as a root replacement, even to a fragile LVOT after debridement [37]. Pettersson and colleagues [36] felt that the aortic homograft was exposed to degeneration and calcification of the cusps, committing the surgeon to a redo root replacement. However, the autograft would heal with minimal scar tissue, degeneration, and calcification and thus not require a reoperation. Niwaya and colleagues [37] advocated this procedure for patients with endocarditis and limited involvement of the aortic root, no medical comorbidity, and life expectancy exceeding 20 years. Carr-White and colleagues [33] used the pulmonary autograft with good results in PVE, but acknowledged increased difficulty and longer operating times, with risk of injury to coronary arteries in the presence of dense adhesions around the aortic and pulmonary roots. After initial enthusiasm for the Ross procedure, several authors reported dilatation of the autograft in aortic position, mandating its replacement [44–46]. They postulated that this may be technique-dependent, with the full-root technique being more prone compared with the subcoronary technique, wherein the native root and sinuses would support the autograft and prevent its dilatation [44–46]. However, this may not be possible in PVE, wherein pathology often dictates root replacement. Furthermore, the pulmonary homograft is subject to long-term calcification and degeneration [44], stenosis, and fibrosis [47]. This led some groups to recommend using newer prostheses to reconstruct the RVOT after the Ross procedure, namely, xenopericardial valves/conduits treated with the No-React anticalcification process [48]. Thus, the Ross operation is still a viable alternative in PVE, but its indications are specific [37].
Stentless bioprostheses
Due to the restricted supply of homografts, stentless bioprostheses have been increasingly used, commonly the Toronto stentless porcine valve (St Jude, Minneapolis, MN) and the Freestyle valve (Medtronic, Minneapolis, MN) [49]. Depending on the prosthesis, pathology, and characteristics of the aortic root, various implantation techniques (subcoronary, supraannular, root inclusion, and full-root) have been recommended [49].
Fukui and colleagues [50] found that the Freestyle valve had excellent handling properties, similar to those of the homograft in aortic PVE with paravalvular abscesses, and could be implanted by the full-root technique, after radical debridement of all infected tissues and aortic annular reconstruction. Others reported similar observations [51, 52]. Results with the Toronto root in recent multicenter trials have been extremely promising [53]. This may be implanted as a full-root, in cases of PVE. The wall of the prosthesis was treated with ethanol and aluminum to retard calcification [53]. Stentless valves have more synthetic material compared with the homograft and those with the least synthetic material may be preferable in PVE.
New bioprosthetic conduits
These comprise porcine and xenopericardial valves with or without a conduit and may be suitable for complex cases. The arterial wall has been treated by the No-React process to slow down tissue degeneration and calcification. These are available in several sizes and lengths tailored to suit the needs of the individual patient [48, 54, 55]. These have been used successfully in aortic PVE [54, 56].
Aortic PVE: Mechanical and Stented Tissue Valves
Alternative to the homograft in PVE is the use of standard prostheses after aggressive debridement of all infected tissue and laying open of abscesses [3, 5, 7, 11, 18, 19, 38]. Several authors found no difference between homografts and composite valve grafts (CVGs) in PVE [16, 38]. Others found no significant differences between standard mechanical and bioprosthetic valves in PVE in terms of recurrence and long-term event free survival [5, 6, 19]. Moon and colleagues [19] observed that the linearized rate of recurrent endocarditis was 2.5 ± 0.7% after valve replacement for PVE, with no difference between mechanical and bioprosthetic valves. However, the number of homografts in this series was small and thus excluded from analysis. Bioprostheses in younger patients are subject to accelerated degenerative changes after 10 years; thus, bioprostheses may be advisable in patients older than 60 years, or in younger patients with PVE with limited life expectancy due to coexisting CAD, LV dysfunction, end stage renal disease, etc [19]. d’Udekem and colleagues [3] felt that the homograft alone was inadequate to completely reconstruct extensively destroyed aortic roots, which often needed reconstruction using pericardial patches. After this aggressive debridement and reconstruction, they had good results irrespective of the valve used. Hagl and colleagues [18] felt that, in the face of extensive tissue destruction and adhesions in PVE, an irregular proximal suture-line would make the homograft prone to distortion and incompetence, which would be less likely with CVGs. Furthermore, the problems of late degeneration seen with homografts would not be present in CVGs. Ergin [34] used standard CVGs for root replacement in aortoventricular reconstruction because of ease of implantation and ready availability. In difficult situations after debridement of extensive infection in a Bentall-root replacement, he performed root replacement using a conduit with a valve sutured 1 to 2 cm from the proximal end. Others [5, 11] have recommended use of homografts only when extensive aortic annular reconstruction is required and standard devices cannot be implanted in recurrent PVE, mainly due to limited availability.
The silzone valve
Based on reports of safety and efficacy of silver for antimicrobial/antifungal protection, this valve (St. Jude Medical, Inc, St. Paul, MN) was used extensively until January 2000. Subsequently, the Artificial Valve Endocarditis Reduction Trial (AVERT) group found an increased incidence of explants due to paravalvular leak thought to be due to silver-mediated inhibition of host fibroblast response; enrollment of patients into the AVERT was stopped, and the product withdrawn [57]. They found no reduction in PVE, but an increase in multiple transient neurologic deficits related to platelet thrombi, and this was higher in mitral than in aortic valves [57]. Others observed similar findings, when used exclusively in PVE [58]. However, these views have not been uniformly supported [59, 60].
Mitral PVE
Mitral PVE is less common than aortic PVE [1, 2, 16]. Aggressive debridement of abscesses and removal of all infected tissue form the cornerstone of treatment [2–4, 6, 10, 23, 26, 61]. Subsequently the annulus and defects may be reconstructed with autologous/bovine pericardium. Choice of prosthesis remains a matter of debate, reflecting the lack of consensus.
Standard mechanical valves and bioprostheses
Mechanical valves have been used extensively in mitral PVE [3, 6, 7, 11, 19]. Several authors found no difference between tissue and mechanical prostheses in mitral PVE in terms of recurrence and long-term event-free survival [5–7, 11, 19, 23, 62].
Jault and colleagues [23] described intraatrial insertion of a mitral prosthesis by suturing it to the LA wall above the annulus, after modification of the sewing ring with a Dacron collar. They found this to be useful when annular and subannular abscesses made conventional valve replacement difficult. However, Ergin [34] felt that this could expose attenuated annular tissues to high ventricular pressures leading to atrioventricular disruption. He preferred reconstruction of the annulus with pericardium and fixation of the prosthesis to this. If necessary, the valve could be translocated to the ventricular side, leaving the attenuated annulus exposed to the low-pressure atrial system. Lytle [4] preferred standard bioprostheses for replacement after debridement and reconstruction of the annulus in mitral PVE, thus avoiding anticoagulation in these patients who may harbor metastatic cerebral abscesses. He reconstructed the annulus with pericardium after extensive debridement and did not translocate the valve into the LA. d’Udekem and colleagues [3] stressed the importance of radical debridement of abscesses; they reconstructed the annulus with autologous pericardium, or with bovine pericardium when circumferential annular reconstruction was required. This was sutured directly to the LV endocardium posteriorly and to the fibrous area underneath the aortic valve superiorly. Where abscesses involved the central fibrous body, pericardium was used to reconstruct both the mitral and aortic annuli [3, 34]. The prosthesis was then secured to this patch and remnants of the annulus, with good results [3].
Mitral bioprostheses are more prone to structural deterioration with time, more so in younger patients. However, they have less thromboembolic and bleeding complications in the absence of any rhythm disturbance requiring anticoagulation [19]. Even in the latter case, complications are less due to lower levels of anticoagulation required.
Mitral homograft
This was used based on the premise that by preserving a functional papillary muscle and chordal complex, ventricular function would be more efficient than with standard prostheses. Furthermore, absence of prosthetic material in an infected area would decrease the possibility of reinfection [63]. However, the experience has not been satisfactory due to limited experience in PVE [64], homograft degeneration, chordal thinning, and rupture [65].
Aortic homograft and pulmonary autograft
The aortic homograft was used in the past, incorporated into a semirigid Dacron tube with a pericardial cuff on top, covering all prosthetic material [66]. This prosthesis was discontinued due to problems encountered at reoperation for homograft degeneration and calcification [64]. Glutaraldehyde-treated, stent-mounted aortic homografts were then developed [67]. Kabbani and colleagues [68] improvised upon this, using the pulmonary autograft in a similar fashion, but these may not be suitable in PVE.
Postoperative Complications
Postoperative complications [1, 2, 4, 5, 16, 19] such as bleeding, respiratory complications, low cardiac output syndrome, and renal dysfunction are not infrequent. Invasive vascular-access devices required in these situations increase the risk of infection. Radical debridement of aortic root abscesses may cause complete heart block requiring a pacemaker [1, 3, 11, 18, 19, 35]. Neurologic complications may occur. Cerebral abscesses are uncommon but may require surgical attention. Disseminated metastatic abscesses may need to be drained.
Prognosis
Predictors of early mortality include increased age [6, 16, 19, 20], female sex, shorter duration of symptoms, longer duration of ischemia and CPB, emergency surgery [5], staphylococcal infection, positive intraoperative culture, liver failure [19], renal dysfunction [16, 19], higher NYHA functional class, multiple previous operations [20], poor hemodynamic status [5, 26], preoperative pulmonary edema [62], ongoing sepsis [5], fever, previous dental procedure lacking prophylaxis, time to diagnosis [16], and type and size of the explanted prosthesis [6].
Younger age has an adverse impact on the durability of biological valves [69, 70], but this may not be an issue in PVE patients whose life expectancy is considerably lower [19]. Young age of the recipient, donor age greater than 65 years, high donor-recipient age difference, and diabetes are associated with accelerated homograft degeneration [39, 40].
Factors that adversely influence long-term survival include age [6, 39], poor LV function, complexity of the procedure and associated procedures [39], male gender, extensive infection [26, 62], and fungal PVE [10]. Poor LV function and previous myocardial infarction may influence long-term function of homografts as turbulent flow due to a dilated, poorly contracting LV can traumatize leaflet edges [39].
Predictors of late complications include increased age, PVE, annular abscess, acute preoperative MI and emergency operation [19]. The overall 10-year survival rate has been around 60% in various series [2, 4, 5] (Table 2), with a late mortality of 16% and a reoperation rate of 20%.
PVE Prophylaxis
Prophylaxis plays a vital role in prevention of this dreaded complication [2, 13, 14, 21, 71, 72]. Before valve replacement, broad-spectrum antibiotics should be given during the induction of anesthesia to ensure adequate antibiotic levels at the time of skin incision [72]. Using antibiotic and chemical-impregnated valves may prevent implantation and growth of microorganisms [13, 71, 73]. New techniques to destroy bacterial biofilms are being developed [13].
The American Heart Association has recommended that patients with prosthetic valves should be administered appropriate broad-spectrum antibiotics to cover procedures causing significant bacteremia, which include orodental procedures associated with significant bleeding, rigid bronchoscopies, sclerotherapy of esophageal varices, and procedures on the gastrointestinal and genitourinary tracts, and an obstructed biliary system [72]. Metronidazole or clindamycin should be included to cover invasive gastrointestinal procedures. Procedures performed through surgically scrubbed skin not involving breach of the respiratory or gastrointestinal mucosae do not cause significant bacteremia and do not need periprocedural antibiotics [72].
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterial and MethodsConclusionsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsConclusionsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. M. Fedoruk, W. R. E. Jamieson, H. Ling, J. S. MacNab, E. Germann, S. S. Karim, and S. V. Lichtenstein Predictors of recurrence and reoperation for prosthetic valve endocarditis after valve replacement surgery for native valve endocarditis. J. Thorac. Cardiovasc. Surg., February 1, 2009; 137(2): 326 - 333. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N E MANGHAT, V RACHAPALLI, R V. LINGEN, A M VEITCH, C A ROOBOTTOM, and G J MORGAN-HUGHES Imaging the heart valves using ECG-gated 64-detector row cardiac CT Br. J. Radiol., April 1, 2008; 81(964): 275 - 290. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Carozza, A. Della Corte, F. Ursomando, and M. Cotrufo The Choice of Valve Prosthesis for Infective Endocarditis in Intravenous Drug Users: Between Evidence and Preference Ann. Thorac. Surg., March 1, 2008; 85(3): 1141 - 1141. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Lopes, P. Calvinho, F. de Oliveira, and M. Antunes Allograft aortic root replacement in complex prosthetic endocarditis Eur. J. Cardiothorac. Surg., July 1, 2007; 32(1): 126 - 130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. H.L. Tang, M. Maganti, T. E. David, C. M. Feindel, H. E. Scully, and M. A. Borger Effect of Prior Valve Type on Mortality in Reoperative Valve Surgery Ann. Thorac. Surg., March 1, 2007; 83(3): 938 - 945. [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 |
