Ann Thorac Surg 2002;74:232-234
© 2002 The Society of Thoracic Surgeons
Case report
LVAD support in patients with bioprosthetic valves
Alessandro Barbone, MDa,
Vivek Rao, MD, PhDa,
Mehmet C. Oz, MDa,
Yoshifumi Naka, MD, PhD*a
a Department of Surgery, Columbia University, New York, New York, USA
Accepted for publication October 20, 2001.
* Address reprint requests to Dr Naka, Mechanical Circulatory Support, New York-Presbyterian Hospital, MHB 7-435, 177 Fort Washington Ave, New York, NY 10032 USA
e-mail: yn33{at}columbia.edu
 |
Abstract
|
|---|
The presence of mechanical or bioprosthetic valves has traditionally excluded patients from mechanical circulatory support. However, several centers have now developed algorithms for the surgical management of native or prosthetic valve disease in patients requiring left ventricular assist device insertion. We report adverse events associated with bioprosthetic valves in the mitral and tricuspid positions in 2 patients who received long-term mechanical support. We recommend anticoagulation for all patients with prosthetic valves in the mitral or tricuspid position to avoid thromboembolism, inflow conduit occlusion, or valvular incompetence.
|
Introduction
|
|---|
The presence of a prosthetic valve is no longer considered a contraindication for the insertion of a ventricular assist device (VAD). Different protocols have been developed to guide surgeons involved in the management of patients requiring mechanical circulatory support [1]. Generally, a bioprosthetic valve is considered to be at low risk for thrombus formation, especially if exposed to high flow. Some authors [2] have even suggested replacing a mechanical valve with a biological valve to reduce the risk of thromboembolism during left VAD (LVAD) support. However, we present the case of a patient in whom we found that biological valves can become problematic because of the lack of cyclic opening and closing in the unloaded ventricle.
A 62-year-old diabetic man sustained an anterior myocardial infarction in 1997, following which symptoms of New York Heart Association class IV heart failure developed. He underwent a regular transplantation evaluation workup, which revealed a 40% proximal right coronary artery stenosis (dominant), an 80% proximal left circumflex artery stenosis, a 70% stenosis of the middle part of the left anterior descending coronary artery, and a 90% ostial stenosis of the first diagonal branch. Left ventricular ejection fraction was estimated to be 0.30 with mild to moderate mitral regurgitation. Right heart catheterization showed a right atrial pressure of 4 mm Hg and a pulmonary artery pressure of 62/18 mm Hg with a wedge pressure of 25 mm Hg. A preoperative thallium stress test revealed fixed anteroseptal and apical defects, with some evidence of redistribution at 24 hours, thus raising the possibility of hibernating myocardium.
The patient underwent triple coronary artery bypass grafting (left internal mammary artery-left anterior descending coronary artery, saphenous vein graft-first diagonal branch, and saphenous vein graft-ramus intermedius) and mitral valve replacement with a 29-mm Hancock II valve with LVAD backup. The postoperative course was complicated by persistent severe biventricular failure requiring nitric oxide and multiple inotropic agents. On postoperative day 9, a transesophageal echocardiogram revealed a left atrial thrombus occluding the pulmonary veins and an unchanged left ventricular ejection fraction compared with preoperative studies. Because of persistent low output syndrome and imminent risk of thromboembolism, the patient underwent thrombectomy and implantation of a HeartMate vented electric LVAD (Thoratec Corp, Pleasanton, CA). No anticoagulation apart from oral aspirin was initiated.
After 3 weeks of support, the patient was evaluated for possible functional recovery and LVAD explantation. Although all grafts were confirmed patent by catheterization, left ventricular function remained poor, and explantation was not thought to be an option. After 43 days of LVAD support, a donor heart became available, and the patient underwent LVAD explantation and orthotopic heart transplantation. At the time of operation, native heart performance was evaluated in the absence of LVAD support. Almost no cardiac output was generated despite reasonable contractility.
After explantation, the grafts were found to be patent, but the mitral valve was fixed in the open position (Fig 1).
The leaflets were trapped by thrombi that had formed under the cusps (Fig 2).
The thrombi were firmly adherent to the cusps and the cloth annulus.
View larger version (112K):
[in this window]
[in a new window]
|
Fig 1. Bovine pericardial mitral valve from a patient supported by a left ventricular assist device. At the time of transplantation, the valve was fixed in the open position.
|
|
|
Comment
|
|---|
The management of prosthetic valves in an LVAD-supported patient has been a source of concern. In a previous review of our institutional experience, we [1] found a major risk of thrombosis of prosthetic aortic valves, particularly mechanical devices. As a result of the unloading of the left ventricle, the valve usually remains in a closed position and the low flow favors thrombus formation. Because of the fibrinolytic state induced by the VAD [3], we prefer not to anticoagulate patients who receive the HeartMate LVAD. We find it easier to oversew mechanical valves in the aortic position with Dacron patch closure.
The management of mitral valve disease is more controversial. Even with LVAD support, the mitral valve is exposed to normal flows and should not be at risk for thrombus formation. Furthermore, as native cardiac systole is not synchronous with LVAD ejection, the valve may be subjected to closure multiple times per minute. Our major concern in patients with an LVAD is the risk of an incompetent mitral valve, which can lead to transmission of the V wave to the lung causing pulmonary edema. For this reason, a repair stitch, such as described by Maisano and associates [4], was suggested in cases of moderate to severe mitral insufficiency.
Our report indicates that the problem may be more difficult than initially expected. Normal flow is not enough to ensure absence of thrombus, and the unloaded ventricle may not generate enough pressure to ensure valve closure. There are two predictable consequences: the thrombi can dislodge from the leaflets, or the thrombus can lead to inflow obstruction, thereby preventing weaning and recovery in these patients.
The first issue could be addressed by systemic anticoagulation, but, as we have already stated, this is not ideal for a patient with an LVAD. Immediately after operation, when the thrombus is fresh and easier to dislodge, bleeding is a major concern. In another patient with a biological tricuspid valve and a mechanical mitral valve who was supported by a Thoratec biventricular VAD system, we instituted postoperative anticoagulation with heparin sodium for both the mechanical valve and the VADs. However, 3 days after uneventful VAD insertion, bleeding and cardiac tamponade requiring reexploration developed. At the time of transplantation, the biological valve in the tricuspid position was found to have stiffened leaflets in the open position, but no thrombus was found underneath the cusps (Fig 3).
The mechanical valve in the mitral position was functioning properly without evidence of thrombus. The absence of thrombus was attributed to anticoagulation.
View larger version (111K):
[in this window]
[in a new window]
|
Fig 3. Porcine bioprosthesis explanted from the tricuspid position in a patient supported by a Thoratec biventricular assist device. No thrombus was evident in the valvular or subvalvular spaces. Notice the proximity to the bioprosthesis of the ventricular inflow cannula of the right ventricular assist device. This raises concern about inflow obstruction in the event of valve thrombosis.
|
|
The possibility of thrombus causing inflow obstruction is of greater concern, especially for patients with the potential for myocardial recovery. Usually, a support duration of 30 days is considered necessary for the heart to recover function and the VAD to be explanted. During a similar time frame, the bioprosthetic mitral valve in our patient developed thrombosis. Thrombosis of a prosthetic valve can cause either major stenosis or regurgitation and affect LVAD function. Before the experience reported here, we thought we were able to manage a prosthetic mitral valve without anticoagulation, but we were wrong. Although the patient did not have thromboembolic events, if there had been acute mechanical dysfunction of the device, the patient would not have tolerated massive mitral regurgitation and probably would have died immediately. This problem could be important for patients facing the possibility of long-term support (bridge to transplantation or destination therapy) but might be even more important for patients whose valve is supposed to be functional after short support (weaning to recovery). Therefore the threshold for instituting systemic anticoagulation should be low, and a mechanical prosthesis may be preferable. An argument can be made to anticoagulate any LVAD recipient with prosthetic aortic, mitral, or tricuspid valves.
|
References
|
|---|
- Rao V., Slater J.P., Edwards N.M., Naka Y., Oz M.C. Surgical management of valvular disease in patients requiring left ventricular assist device support. Ann Thorac Surg 2001;71:1448-1453.[Abstract/Free Full Text]
- Stringham J.C., Bull D.A., Karawande S.V. Patch closure of the aortic annulus in a recipient of a ventricular assist device. J Thorac Cardiovasc Surg 2000;119:1293-1294.[Free Full Text]
- Spanier T.B., Chen J.M., Oz M.C., Stern D.M., Rose E.A., Schmidt A.M. Time-dependent cellular population of textured-surface left ventricular assist devices contributes to the development of a biphasic systemic procoagulant response. J Thorac Cardiovasc Surg 1999;118:404-413.[Abstract/Free Full Text]
- Maisano F., Torracca L., Oppizzi M., et al. The edge-to-edge technique: a simplified method to correct mitral insufficiency. Eur J Cardiothorac Surg 1998;13:240-245.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
J. A. Crestanello, D. A. Orsinelli, M. S. Firstenberg, and C. Sai-Sudhakar
Aortic valve thrombosis after implantation of temporary left ventricular assist device
Interactive CardioVascular and Thoracic Surgery,
June 1, 2009;
8(6):
661 - 662.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
